Issuu on Google+


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

Editor-in-Chief Jan Cools (Leuven)

Deputy Editor Luca Malcovati (Pavia)

Managing Director Antonio Majocchi (Pavia)

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

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

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 (Rotterdam); 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 March 2016.


haematologica calendar of events

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

42nd EBMT Annual Meeting European Society for Bone and Marrow Transplantation Chairs: MS Sanz, A Urbano, JL DĂ­ez April 3-6, 2016 Valencia, Spain

8th International Conference on Thrombosis and Hemostasis Issues in Cancer - ICTHIC April 8-10, 2016 Chairs: A Falanga, B Brenner, FR Rickles Bergamo, Italy

Hematology Tutorial on managing complications in patients with hematologic malignancies in the era of new drugs EHA-ROHS-RSH Chairs: E Parovichnikova, I Poddubnaya, R FoĂ  July 1-3, 2016 Moscow, Russian Federation

ISTH Advanced Training Course in Thrombosis and Hemostasis International Society on Thrombosis and Haemostasis (ISTH)Chair: S Watson September 6-9, 2016 Oxford, UK

ESH-EBMT 20th Training Course on Haemopoietic Stem Cell Transplantation ESH EBMT May 11-14, 2016 Chairs: J Apperley, E Carreras, E Gluckman, T Masszi Budapest, Hungary

EHA Scientific Conference on Bleeding Disorders Scientific Program Committee: C Balduini (Chair), A Falanga (Chair), F Rodeghiero, I Pabinger, M Makris September 14-17, 2016 Barcelona, Spain

21st Congress of the European Hematology Association European Hematology Association June 9-12, 2016 Copenhagen, Denmark

Calendar of Events updated on March 1, 2016


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

Table of Contents Volume 101, Issue 4: April 2016 Cover Figure Artistic impression of multiple myeloma cells (Image created by www.somersault1824.com)

Editorials 391

On mice and humans: the role of thymic stromal lymphopoietin in human B-cell development and leukemia Angela Maria Savino and Shai Izraeli

394

Innovation in hematology: morphology and flow cytometry at the crossroads. Marie C. Béné, et al.

Review Articles 396

Treatment of relapsed and refractory multiple myeloma - Leaders in Hematology review series Pieter Sonneveld and Annemiek Broijl

407

New and emerging prognostic and predictive genetic biomarkers in B-cell precursor acute lymphoblastic leukemia Anthony V. Moorman

Articles Hematopoiesis

417

A novel xenograft model to study the role of TSLP-induced CRLF2 signals in normal and malignant human B lymphopoiesis Olivia L. Francis,

Platelet Biology & Its Disorders

427

Coated platelets function in platelet-dependent fibrin formation via integrin αIIbb3 and transglutaminase factor XIII Nadine J.A. Mattheij, et al.

Bone Marrow Failure

437

Extracellular vesicle miR-7977 is involved in hematopoietic dysfunction of mesenchymal stromal cells via poly(rC) binding protein 1 reduction in myeloid neoplasms Hiroto Horiguchi, et al.

Acute Myeloid Leukemia

448

Helicase-like transcription factor is a RUNX1 target whose downregulation promotes genomic instability and correlates with complex cytogenetic features in acute myeloid leukemia Chi Keung Cheng, et al.

Chronic Lymphocytic Leukemia

458

Prevalence and characteristics of central nervous system involvement by chronic lymphocytic leukemia Paolo Strati, et al.

Haematologica 2016; vol. 101 no. 4 - April 2016 http://www.haematologica.org/


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

Hodgkin Lymphoma

466

Impact of post-brentuximab vedotin consolidation on relapsed/refractory CD30+ Hodgkin lymphomas: a large retrospective study on 240 patients enrolled in the French Named-Patient Program Aurore Perrot, et al.

474

Single or tandem autologous stem-cell transplantation for first-relapsed or refractory Hodgkin lymphoma: 10-year follow-up of the prospective H96 trial by the LYSA/SFGM-TC study group David Sibon, et al.

Cell Therapy & Immunotherapy

482

Expression of a dominant T-cell receptor can reduce toxicity and enhance tumor protection of allogeneic T-cell therapy Angelika Holler, et al.

Stem Cell Transplantation

491

High-risk HLA alleles for severe acute graft-versus-host disease and mortality in unrelated donor bone marrow transplantation Satoko Morishima, et al.

499

Infused total nucleated cell dose is a better predictor of transplant outcomes than CD34+ cell number in reduced-intensity mobilized peripheral blood allogeneic hematopoietic cell transplantation Paul S. Martin, et al.

506

Multi-state analysis illustrates treatment success after stem cell transplantation for acute myeloid leukemia followed by donor lymphocyte infusion Matthias Eefting, et al.

Letters to the Editor Letters are available online only at www.haematologica.org/content/101/4.toc

e126

The role of Matriptase-2 during early postnatal development in humans Luigia De Falco, et al. http://www.haematologica.org/content/101/4/e126

e129

Mutation status of essential thrombocythemia and primary myelofibrosis defines clinical outcome Julia Asp, et al. http://www.haematologica.org/content/101/4/e129

e133

NUP214-ABL1 fusion defines a rare subtype of B-cell precursor acute lymphoblastic leukemia that could benefit from tyrosine kinase inhibitors Nicolas Duployez, et al. http://www.haematologica.org/content/101/4/e133

e135

Clinical impact of small subclones harboring NOTCH1, SF3B1 or BIRC3 mutations in chronic lymphocytic leukemia Silvia Rasi, et al. http://www.haematologica.org/content/101/4/e135

e139

Brentuximab vedotin followed by ABVD +/- radiotherapy in patients with previously untreated Hodgkin lymphoma: final results of a pilot phase II study Massimo Federico, et al. http://www.haematologica.org/content/101/4/e139

Haematologica 2016; vol. 101 no. 4 - April 2016 http://www.haematologica.org/


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

e142

Frequency, distribution and clinical management of incidental findings and extramedullary plasmacytomas in whole body diffusion weighted magnetic resonance imaging in patients with multiple myeloma Anita Wale, et al. http://www.haematologica.org/content/101/4/e142

e145

Survival in multiple myeloma patients who develop second malignancies: a population-based cohort study Gudbjรถrg Jonsdottir, et al. http://www.haematologica.org/content/101/4/e145

e149

Phase I/II trial of weekly bortezomib with lenalidomide and dexamethasone in first relapse or primary refractory myeloma Annemiek Broijl, et al. http://www.haematologica.org/content/101/4/e149

e153

IL-15-activated cytokine-induced killer cells may sustain remission in leukemia patients after allogeneic stem cell transplantation: feasibility, safety and first insights on efficacy Eva Rettinger, et al. http://www.haematologica.org/content/101/4/e153

e157

Droplet digital polymerase chain reaction for DNMT3A and IDH1/2 mutations to improve early detection of acute myeloid leukemia relapse after allogeneic hematopoietic stem cell transplantation Chiara Brambati, et al. http://www.haematologica.org/content/101/4/e157

Comments Comments are available online only at www.haematologica.org/content/101/4.toc

e162

Endotoxemia as a trigger of thrombosis in cirrhosis Francesco Violi, et al. http://www.haematologica.org/content/101/4/e162

Haematologica 2016; vol. 101 no. 4 - April 2016 http://www.haematologica.org/


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

The origin of a name that reflects Europe’s cultural roots.

Ancient Greek

aÂma [haima] = blood a·matow [haimatos] = of blood lÒgow [logos]= reasoning

Scientific Latin

haematologicus (adjective) = related to blood

Scientific Latin

haematologica (adjective, plural and neuter, used as a noun) = hematological subjects

Modern English

The oldest hematology journal, publishing the newest research results. 2014 JCR impact factor = 5.814

Haematologica, as the journal of the European Hematology Association (EHA), aims not only to serve the scientific community, but also to promote European cultural identify.


EDITORIALS On mice and humans: the role of thymic stromal lymphopoietin in human B-cell development and leukemia Angela Maria Savino and Shai Izraeli Leukemia Research Section, Edmond and Lily Safra Children Hospital, Sheba Medical Center, Tel Hashomer, Israel and Department of Human Molecular Genetics and Biochemistry, Sackler Medical School, Tel Aviv University, Israel E-mail: sizraeli@sheba.health.gov.il

T

doi:10.3324/haematol.2016.142448

he development of immunodeficient mouse models has revolutionized the ability to study human disease in vivo. One of the difficulties is that many mouse and human cytokines are not cross-reactive. For this reason, the in vivo models currently used are not optimized enough to recapitulate the correct environment for every type of human hematopoietic or leukemic cell. This problem has been recently addressed by the development of several models of humanized mice expressing human cytokines, as described in detail by the recent excellent review in Haematologica by Theocarides et al.1 In this issue of Haematologica, Francis et al.2 created a

chimeric mouse able to produce human thymic stromal lymphopoietin (TSLP) to study the effect of human TSLP on human B-cell development and human B-cell precursor acute lymphoblastic leukemias (BCP-ALLs) with abnormal expression of the TSLP receptor (TSLPR) (Figure 1). The generation of this mouse was required since the homology between mouse and human TSLP is only 43% and 35% for the TSLPR, with no cross reactivity between the species.3 Thymic stromal lymphopoietin is a cytokine that was first described as a constituent of medium conditioned by thymic stromal cells. TSLP is important in many allergic disorders,

Figure 1. Schematic representation of the humanized mouse model and the effect of human TSLP on normal and malignant hemopoiesis. Only human TSLP is able to induce a conformational change in human heterodimeric receptor and to activate the JAK/STAT or PI3K/mTOR downstream pathway (red dots indicate phosphorylation). The injection of human stromal cells transduced to produce TSLP in NSG mice, induced marked increase of B-cell lymphopoiesis upon transplantation with CD34+ cells derived from human cord blood and, after transplantation with CRLF2 positive leukemia, induced a gene expression similar to the original patient. The increased expression of genes involved in mTOR signaling could be associated with chemoresistance.

haematologica | 2016; 101(4)

391


Editorials

Figure 2. TSLP and CRLF2 positive leukemias: a hypothetical model. In the pre-leukemic phase, B-cell progenitors over-expressing the TSLP receptor are expanded in response to TSLP. In the leukemic phase, acquired activating mutations in the pathway (e.g. in JAK2) make the cells hypersensitive to TSLP, causing their dramatic expansion.

including asthma and atopic dermatitis (reviewed by Tal et al.4). In addition to the thymic medulla, its major sites of production are the lung, bronchial tree, intestine and the skin.5 It is currently not known to what degree it is produced by bone marrow stroma, although some studies suggest its expression in some mesenchymal stroma cells,6 for example, during co-culture with myeloma cells.7 Its receptor (TSLPR) is a heterodimer composed of one TSLP-binding subunit (CRLF2) and the alpha subunit of interleukin 7 receptor (IL7RA). The TSLPR is expressed on a variety of cells (lymphocytes, mast cells, basophils, monocytes and dendritic cells) and, upon the binding of the ligand, it activates the JAK/STAT signaling pathway. While TSLP has a role in the development of some subsets of T lymphocytes,8 its role in mouse B-cell development is unclear. Mice unresponsive to both IL7 and TSLP (lacking IL2RG and CRLF2) have neither B nor T cells. However, in mice lacking only the receptor to IL7, transgenic expression of TSLP partially restored B and T lymphopoiesis, showing a sort of redundancy with IL7.9 Even less clear is the role, if any, of TSLP in human B-cell development. While some levels of CRLF2 RNA are detected in normal B-cell progenitors, the surface expression of TSLPR is barely detectable. To date, there has only been one in vitro study, reporting the co-operation between TSLP and IL7 in the proliferation and differentiation of human fetal B-cell precursors.10 The model created by Francis et al. allowed the first in vivo examination of the role of human TSLP in B-cell development (Figure 1).2 Human bone marrow stroma cell line (HS27) transduced to secrete human TSLP were given by intra-peritoneal injection to NSG mice and survived in the peritoneal cavity (they were not detectable either in the spleen or bone marrow (BM) of the mice) reaching serum levels of TSLP similar to those of human blood. The advantage of this model consists in the ability to modulate the serum levels of hTSLP based on the timing and number of stromal cells injected. This aspect could be biologically relevant, as underlined by the authors, since the in vivo production of TSLP is usually increased by environmental factors. The authors demonstrated that the TSLP produced from the stromal cells was functional, being able to induce ex vivo the phosphorylation of STAT5 and p70S6K, which 392

are targets of the two major TSLPR-related pathways. Importantly, they showed that the presence of human TSLP was associated with a marked increase of human Bcell lymphopoiesis in NSG mice transplanted with CD34+ cells derived from human cord blood. While this observation clearly supports the view that TSLP can affect human B-cell lymphopoiesis, its physiological role remains unclear. NSG mice express mouse IL7 that can only partially stimulate the human IL7 receptor,11 thus the observed TSLP effect is on the background of a relative lack of IL7. In addition, for reasons that are still unclear, human hematopoiesis in these mice is skewed toward the B-cell lineage. Regardless of these limitations, the findings of Francis et al. clearly demonstrate that, despite the almost absent expression of the receptor, TSLP can support human B-cell lymphopoiesis.2 Unlike the uncertainty of its importance in normal B-cell development, a clear role has emerged for the TSLP pathway in BCP-ALL. Up to two-thirds of BCP-ALL in children with Down syndrome and 5%-10% of BCP-ALL children and adults without Down syndrome have acquired genomic aberrations leading to a markedly increased expression of CRLF2, and hence the receptor to TSLP, in the malignant blast cells.12-14 The increased expression is often accompanied by additional somatic activating mutations in the TSLPR pathway, including the CRLF2 or IL7R receptors or the downstream JAK signaling molecules.15,16 These mutations induce the cytokine independent growth of leukemic cells that nevertheless present extreme hypersensitivity to TSLP. These genetic events suggest a potential model for leukemogenesis consisting of two stages: initial TSLP dependent expansion of a (pre-leukemic) B cell over-expressing CRLF2, followed by the further acquisition of cytokine independence by activating mutations leading to frank leukemia (Figure 2). While the model suggests that TSLP is essential to the evolution to leukemia, it could also be possible that the extreme sensitivity of the frank leukemic cells to TSLP may protect them from chemotherapy, explaining the bad prognosis of these leukemias. The model created by Francis et al. may be used to study these questions. The authors examined the effect of systemic human TSLP on growth of only two primografts of haematologica | 2016; 101(4)


Editorials

human CRLF2 positive BCP-ALL. While they could not observe any effects of human TSLP on the growth of the leukemic cells, their gene expression analysis was more like the original patient sample showing an increased expression of genes involved in mTOR signaling that is known to be associated with chemoresistance.17 It would have been interesting to see if the presence of TSLP changed the response of these leukemias to chemotherapy. The new mouse model suffers from two major limitations. It is laborious, requiring weekly injections of a large number of genetically modified stromal cells secreting human TSLP. More importantly, it does not recapitulate the normal sites of production of TSLP and its normal regulation. This could be achieved by either creating an NSG transgenic mouse with human BAC of TSLP or by a “knock-in� of the human TSLP into the mouse TSLP locus. It is important to create such a mouse model, especially in the light of the possibility that TSLP is made by some BM stromal cells.6,7 Recent studies underline the importance of the human BM niche in providing signals of survival and proliferation of normal and malignant hematopoietic cells (reviewed by Schepers et al.18). Furthermore, it has already been demonstrated in transgenic mice that the local increase in TSLP production induced systemic alterations in B-cell development.19 It might, therefore, be reasonable to hypothesize that, in addition to the systemic levels of TSLP, local production of TSLP in the bone marrow may protect residual leukemic cells from chemotherapy. A more physiological humanized TSLP mouse model will be required to test this hypothesis. The study of the role of human TSLP in BCP-ALL has an important practical significance. If, indeed, CRLF2 positive BCP-ALL depends on TSLP, then drugs targeting TSLP, developed for allergic disorders,20 may have a role in the treatment of these leukemias. Acknowledgments We acknowledge the financial support of Fondazione Italiana per la Ricerca sul Cancro-FIRC to AMS, the Israel Science Foundation, ERA-NET and the Israel Cancer Research Foundation to SI

References 1. Theocharides AP, Rongvaux A, Fritsch K, Flavell RA, Manz MG. Humanized hemato-lymphoid system mice. Haematologica. 2016;101(1):5-19.

haematologica | 2016; 101(4)

2. Francis OL, Milford TAM, Martinez SR, et al. A novel xenograft model to study the role of TSLP-induced CRLF2 signals in normal and malignant human B lymphopoiesis. Haematologica 2016;101(4):417-426. 3. Quentmeier H, Drexler HG, Fleckenstein D, et al. Cloning of human thymic stromal lymphopoietin (TSLP) and signaling mechanisms leading to proliferation. Leukemia. 2001;15(8):1286-1292. 4. Tal N, Shochat C, Geron I, Bercovich D, Izraeli S. Interleukin 7 and thymic stromal lymphopoietin: from immunity to leukemia. Cell Mol Life Sci. 2014;71(3):365-378. 5. Dewas C, Chen X, Honda T, et al. TSLP expression: analysis with a ZsGreen TSLP reporter mouse. J Immunol. 2015;194(3):1372-1380. 6. Siracusa MC, Saenz SA, Tait Wojno ED, et al. Thymic Stromal Lymphopoietin-Mediated Extramedullary Hematopoiesis Promotes Allergic Inflammation. Immunity. 2013;39(6):1158-1170. 7. Nakajima S, Fujiwara T, Ohguchi H, et al. Induction of thymic stromal lymphopoietin in mesenchymal stem cells by interaction with myeloma cells. Leuk Lymphoma. 2014;55(11):2605-2613. 8. Al-Shami A, Spolski R, Kelly J, et al. A role for thymic stromal lymphopoietin in CD4(+) T cell development. J Exp Med. 2004;200(2):159168. 9. Chappaz S, Flueck L, Farr AG, Rolink AG, Finke D. Increased TSLP availability restores T- and B-cell compartments in adult IL-7 deficient mice. Blood. 2007;110(12):3862-3870. 10. Scheeren FA, van Lent AU, Nagasawa M, et al. Thymic stromal lymphopoietin induces early human B-cell proliferation and differentiation. Eur J Immunol. 2010;40(4):955-965. 11. Johnson SE, Shah N, Panoskaltsis-Mortari A, LeBien TW. Murine and human IL-7 activate STAT5 and induce proliferation of normal human pro-B cells. J Immunol. 2005;175(11):7325-7331. 12. Hertzberg L, Vendramini E, Ganmore I, et al. Down syndrome acute lymphoblastic leukemia, a highly heterogeneous disease in which aberrant expression of CRLF2 is associated with mutated JAK2: a report from the International BFM Study Group. Blood. 2010;115(5):10061017. 13. Mullighan CG, Collins-Underwood JR, Phillips LA, et al. Rearrangement of CRLF2 in B-progenitor- and Down syndrome-associated acute lymphoblastic leukemia. Nat Genet. 2009;41(11):12431246. 14. Russell LJ, Capasso M, Vater I, et al. Deregulated expression of cytokine receptor gene, CRLF2, is involved in lymphoid transformation in B-cell precursor acute lymphoblastic leukemia. Blood. 2009;114(13):26882698. 15. Shochat C, Tal N, Bandapalli OR, et al. Gain-of-function mutations in interleukin-7 receptor-{alpha} (IL7R) in childhood acute lymphoblastic leukemias. J Exp Med. 2011;208(5):901-908. 16. Yoda A, Yoda Y, Chiaretti S, et al. Functional screening identifies CRLF2 in precursor B-cell acute lymphoblastic leukemia. Proc Natl Acad Sci USA. 2010;107(1):252-257. 17. Allegretti M, Ricciardi MR, Licchetta R, et al. The pan-class I phosphatidyl-inositol-3 kinase inhibitor NVP-BKM120 demonstrates antileukemic activity in acute myeloid leukemia. Sci Rep. 2015;5:18137. 18. Schepers K, Campbell TB, Passegue E. Normal and leukemic stem cell niches: insights and therapeutic opportunities. Cell Stem Cell. 2015;16(3):254-267. 19. Astrakhan A, Omori M, Nguyen T, et al. Local increase in thymic stromal lymphopoietin induces systemic alterations in B cell development. Nat Immunol. 2007;8(5):522-531. 20. Gauvreau GM, O'Byrne PM, Boulet LP, et al. Effects of an anti-TSLP antibody on allergen-induced asthmatic responses. N Engl J Med. 2014;370(22):2102-2110.

393


Editorials

Innovation in hematology: morphology and flow cytometry at the crossroads. Marie C. Béné,1 Gina Zini,2 and EHA Scientific Working Group “Diagnosis”, European LeukemiaNet WP10 1

Hematology Biology, University Hospital, Nantes, France; and 2Medicine Transfusion Department, Institute of Hematology, Catholic University, Rome, Italy. E-mail: mariecbene@gmail.com

doi:10.3324/haematol.2016.141861

I

n a rapidly changing world, the concept of innovation moves quickly from far fetched ideas, springing from a curious mind, to daily objects that nobody would ever dream of living without. One of the most obvious example is the smartphone, so much more than a telephone. It has, for instance, considering only the world of hematology, changed the mere concept of attending an international meeting. Gone are the days when we had to carry heavy books and frenetically search for that abstract which caught our eye late last night the paper of which is just about to be presented. Now, when the time arrives, the phone application pops up an alarm with the location, speaker, abstract, social network links, and even note-taking tools! And we take it all for granted! Such exciting changes and ideas really do make themselves felt everywhere, even in such seasoned and validated methods as morphological and flow cytometry (FCM) analysis of pe-ripheral blood (PB) or bone marrow (BM). And if we go back to the smartphone, have you noticed how young (and not so young!) morphologists are swiftly replacing the now almost redundant CCD cameras and their sometimes complex software and share their doubts by directly capturing microscopy fields by placing the camera eye of their smartphone on the ocular of their microscope? Indeed, laboratory diagnosis is rapidly changing from what it was in the past century, and this perspective essay will try and delineate where we are and where we are heading in these two fields, where “fast”, “accurate” and “traceable” seem to be the keywords (…or perhaps these should be “faster”, “more accurate” and “accreditable”?) As reported in a collective European LeukemiaNet (ELN) review of the methodologies of leukemia diagnosis in 2015,1 the future of morphology, although already moving forward in many places, still has digital wonders in store.2 One of the pitfalls of such new approaches is the extent to which it still relies on the quality of smears and staining. However, huge progress has been made with more controlled slide dipping and reagent spray staining.3 Although adjustments will still be necessary, it is to be expected that, by scanning large numbers of cells, robotized smearing and staining will allow image analysis systems to provide reproducible information. The circulation of digitalized white blood cell (WBC) images for teaching or teleconsultation has long been in use.1,4 But besides requiring time-consuming acquisition, it is eating-up large amounts of computer storage space. Increasingly in routine hematology, PB slide scanning and the generation of cell libraries is, however, gaining ground for computerassisted differentials. It enables the storage of patient-related galleries of pre-analyzed and categorized single cells, allowing not only fast assessment but also quick retrieval of rare but crucial cells for concerted assessment (i.e. the elusive 394

Auer rod-containing faggot-blast). Another advantage is it can provide documented/traceable information in a way that complies with the rules of accreditation. Yet, a limitation of these galleries is that, for any individual patient, at the moment, they do not allow storage and display of much more than the 100-200 cells manually reviewed in optical microscopy. Recent advances, such as whole slide imaging (WIS), mostly implemented by pathologists,5,6 allow us to browse a whole sample and electronically store the location of cells of interest, instead of each identified image in a gallery. Given that enough storage space for WIS files is available, this technology could ultimately remove the need for those heavy slide cabinets with their smelly coverslipmounted slides and old-looking numbered and dated labels, where unfailingly THE slide we are looking for has either disappeared or is the only one broken in its drawer. Instead, tomorrow’s morphologists could open a given patient file and immediately get an instant self-generated gallery of cells on a specific slide retrieved from short files of cell co-ordinates automatically pre-identified by the expert system, while being allowed to take a closer look at whichever area the robot neglected, if needed. It is, however, expected that learning software, gathering knowledge with the accumulation of data in a Bayesian approach,7 will be able to act as an intelligent expert system. Based on such information and thorough clinical data, further exploration and therapeutic options of a given patient will be more accurately prescribed by a more self-assured morphologist. It is also likely that such expert systems will be able to concomitantly process morphology and FCM data, since, as detailed below, both methods produce results in the same time-frame, providing proper sampling has allowed both methods to be run. Of course, slide review of PB smears flagged during WBC assessment by cell analyzers is the most obvious of these combined approaches, since a liquid sample was, by definition, at the origin of the flag. It should, however, be noted that modern instruments increasingly combine the Coulter principle of impedance variation for cell counting and flow cytometry technologies, either on unstained cells or by taking advantage of auto-fluorescence or fluorescent staining of nucleic acids or proteins.8 Currently used instruments already have so-called “research parameters” which store a large array of potential innovative applications for an early orientation of patients’ clinical conditions.9 But let’s go back to conventional differentials and we see that in FCM, the Hematoflow® solution,10 with a simple and robust 6-antibody / 5-color combination already provides information on at least 13 leukocyte subsets. In the settings where it is used, it has drastically decreased the number of blood slides to be reviewed manually or digitalized down to about 4%.10 One big advantage of FCM is that it can provide the higher precision of large numbers of counted cells, haematologica | 2016; 101(4)


Editorials

which, by the way, is also exploited in modern automated analyzers, in line with the increased accuracy predicted by the Rümke table.11 Currently, solutions such as Hematoflow® or similar home-designed cocktails,12 with fixed controlled panels and adapted software, already provide a large amount of information and modify the algorithms of further manual / digitalized slide examination, guiding the expert towards the anomaly under investigation and a quick diagnostic / therapeutic approach. It is to be expected that such applications will be developed into selfcontained systems, where the primary tube will be directly processed for all FCM steps of sampling, antibody-staining, lysis-no-wash, histogram analy-sis and computer-assisted validation of normal samples or oriented flagging. Self-contained cost-effective solutions of this type could revolutionize the early steps of hematologic diagnosis in both malignant and non-malignant disorders. Another impressive incursion of innovation in FCM is the appearance of LED-based multiparameter instruments using large panels well over currently established 8-10-color solutions. Such new generation instruments bypass the cumbersome step of compensations and are likely to allow much smaller amounts of both sample and antibodies to be used. Another characteristic of PB that is often taken for granted, but not so well exploited, is that each microliter in an average sample of approximately10x109/L leukocytes contains one hundred more cells than those investigated in a blood smear. Yet flow cytometry can provide detailed and accurate information about each of these cells individually. Between 50 and 100 mL aliquotes are currently used, and this amount could easily be reduced for routine applica-tions, even allowing for a very poor sample of 1x109/L neutrophils, to assess 104 such cells in a 10 mL sample way beyond what even digitalized morphology can provide. In the future, the huge progress achieved in sophisticated automated flow paths for molecular assays of single-cell sequencing could be applied to the less demanding needs of morphology and FCM, circumventing the current limitations of manual pipetting, early syringebased robots, and void volumes. A lot of innovation has also already been put to work for electronically assisted investigation of red blood cells, i.e. stomatocytes, sickle cells, schiztocytes and Jolly bodies, both by smear digitalizing or use of developed research parameters of automated counters.9,13,14 Platelets are also coming under increasing scrutiny.9 Moreover, although at early stages, it is more than likely that combined morphology / FCM techniques will provide information on these cells. In fact, this is already the case in the above mentioned “research parameters” of last generation cell analyzers.9 One last issue where imagination and innovation still have to be put to the test is that of BM sampling. This still suffers from two major hindrances: it is highly operator-dependent and it is marred by the increasing need to perform more assays, i.e. morphology, flow cytometry, cytogenetics, molecular assays, thus requiring more volume. This results in hemodiluted samples that are responsible for variations between initial smears and further analyses in samples less and less representative of the patient’s BM. Some solutions, haematologica | 2016; 101(4)

however, seem to be gaining in maturity with training15 and perhaps the development of assisted puncture devices, allowing for better reproducibility.16 This would lead to huge progress being made also in the other fast-developing field of minimal residual disease, where there is still no real consensus as to the number of cells to be examined for accurate evaluation, while innovative solutions could be developed especially in lysis-no-wash procedures of BM samples. In conclusion, while clearly being at the crossroads between classical and futuristic technologies, both morphology and FCM need to better integrate the results they can quickly provide at the bedside within efficient algorithms to refine diagnostic and therapeutic indications in the global aim of personalized medicine. Even more far fetched, as mentioned at the beginning of this paper, the dream of non-traumatic in vivo cell counts, at least for basic parameters, is already becoming a reality.8,17 But that’s another story…

References 1. Béné MC, Grimwade D, Haferlach C, Haferlach T, Zini G. European LeukemiaNet. Leukemia diagnosis: today and tomorrow. Eur J Haematol. 2015;95(4):365-373. 2. Chung J, Ou X, Kulkarni RP, Yang C. Counting White Blood Cells from a Blood Smear Using Fourier Ptychographic Microscopy. PLoS ONE 2015;10:e0133489 3. http://www2.mlo-online.com/features/201107/current-buzz/whatsnew-in-slide-makers-and-slide-stainers.aspx [Last accessed 26 Dec 2015] 4. Horiuchi Y, Tabe Y, Kasuga K, et al. The efficacy of an internet-based elearning system using the CellaVision Competency Software for continuing professional development. Clin Chem Lab Med. 2015 Oct 17. [Epub ahead of print] 5. Pantanowitz L, Sinard JH, Henricks WH, et al. College of American Pathologists Pathology and Laboratory Quality Center. Validating whole slide imaging for diagnostic purposes in pathology: guideline from the College of American Pathologists Pathology and Laboratory Quality Center. Arch Pathol Lab Med 2013;137(12):1710-1722. 6. Chen ZW, Kohan J, Perkins SL, Hussong JW, Salama ME. Web-based oil immersion whole slide imaging increases efficiency and clinical team satisfaction in hematopathology tumor board. J Pathol Inform. 2014;5:41. 7. Prinyakupt J, Pluempitiwiriyawej C. Segmentation of white blood cells and comparison of cell morphology by linear and naïve Bayes classifiers. Biomed Eng Online. 2015;14:63. 8. Green R, Wachsmann-Hogiu S. Development, history, and future of automated cell counters. Clin Lab Med. 2015;35(1):1-10. 9. Lecompte TP, Bernimoulin MP. Novel parameters in blood cell counters. Clin Lab Med. 2015;35(1):209-224. 10. Allou K, Vial JP, Béné MC, Lacombe F. The routine leukocyte differential flow cytometry HematoFlow™ method: A new flagging system for automatic validation. Cytometry B Clin Cytom. 2015;88(6):375-384. 11. Rümke CL. Uncertainty as to the acceptance or rejection of the presence of an effect in relation to the number of observations in an experiment. Triangle. 1968;8(7):284-289. 12. Roussel M, Davis BH, Fest T, Wood BL; International Council for Standardization in Hematology (ICSH). Toward a reference method for leukocyte differential counts in blood: comparison of three flow cytometric candidate methods. Cytometry A. 2012;81(11):973-982. 13. Horn CL, Mansoor A, Wood B, et al. Performance of the CellaVision(®) DM96 system for detecting red blood cell morphologic abnormalities. J Pathol Inform. 2015;6:11. 14. Da Costa L. Digital image analysis of blood cells. Clin Lab Med. 2015;35(1):105-122. 15. Yap ES, Koh PL, Ng CH, de Mel S, Chee YL. A Bone Marrow Aspirate and Trephine Simulator. Simul Healthc. 2015:10(4):245-248. 16. The Arrow OnControl powered bone marrow biopsy system for bone marrow aspiration and biopsy. NICE advice [MIB19] February 2015. ISBN: 978-1-4731-0980-3. 17. http://www.sciencedaily.com/releases/2015/09/150930074436.htm [Last accessed 26 Dec 2015]

395


REVIEW ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Haematologica 2016 Volume 101(4):396-406

Leaders in Hematology review series

Treatment of relapsed and refractory multiple myeloma Pieter Sonneveld and Annemiek Broijl

Erasmus MC Cancer Institute, Department of Hematology, Rm Na824, Rotterdam, the Netherlands

ABSTRACT

T

he approach to the patient with relapsed or relapsed/refractory multiple myeloma (RRMM) requires a careful evaluation of the results of previous treatments, the toxicities associated with them and an assessment of prognostic factors. Since the majority of patients will have received prior therapy with drug combinations including a proteasome inhibitor and/or an immunomodulatory drug (IMiD), it is the physician’s task to choose the right moment for the start of therapy and define with the patient which goals need to be achieved. The choice of regimen is usually based on prior responsiveness, drugs already received, prior adverse effects, the condition of the patient and expected effectiveness and tolerability. Many double and triple drug combinations are available. In addition, promising new drugs like pomalidomide, carfilzomib and monoclonal antibodies are, or will be, available shortly, while other options can be tried in clinical studies. Finally, supportive care and palliative options need to be considered in some patients. It is becoming increasingly more important to consider the therapeutic options for the whole duration of the disease rather than take a step by step approach, and to develop a systematic approach for each individual patient.

Correspondence: p.sonneveld@erasmusmc.nl

Introduction

Received: December 9, 2015. Accepted: January 13, 2016.

doi:10.3324/haematol.2015.129189

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/4/396

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

396

Multiple myeloma (MM) is a hematologic disorder which is characterized by a proliferation of malignant, monoclonal plasma cells in the bone marrow (BM) and/or extramedullary sites.1 Symptomatic MM is characterized by typical manifestations of organ damage named CRAB, such as lytic bone lesions, hypercalcemia, anemia and renal impairment.1 Recently the criteria for MM were redefined by the International Myeloma Working Group (IMWG), which are summarized in Table 1.2 Progress in the first progression-free survival (PFS) and overall survival (OS) has been achieved through the introduction of high-dose therapy (HDT) with autologous stem cell transplantation (ASCT), and by the introduction of thalidomide, bortezomib and lenalidomide.3 Despite the recent progress in OS rates, MM remains an incurable disease and the majority of patients will relapse and will require treatment.

Definitions of relapsed and relapsed/refractory risease The IMWG published definitions of relapsed MM as well as treatment indications in 2006, 2009 and 2011.4-6 Relapsed MM is regarded as a recurrence of the disease after prior response, and has been defined based on objective laboratory and radiological criteria: ≥ 25 % increase of the serum or urine monoclonal protein (Mprotein) or ≥ 25 % difference between involved and uninvolved serum free light chains from its nadir, respectively, or the development of new plasmacytomas or hypercalciemia. In patients with non-secretory disease, relapse is defined as an increase of the bone marrow plasma cells. In general, an indication for relapse treatment has been defined as either the appearance or reappearance of one or more CRAB criteria or a rapid and consistent biochemical relapse. Relapsed/refractory MM (RRMM) is defined as a disease which becomes non-responsive or progressive on therapy or within 60 days of the last treatment in patients who had achieved a minimal response (MR) or better on prior therapy.7 haematologica | 2016; 101(4)


Treatment of relapsed and refractory multiple myeloma

Indications for relapse treatment The aim of relapse treatment is to relieve disease symptoms and/or to prevent the development of CRAB symptoms. Second and later remissions tend to be shorter because of more aggressive tumor behavior at each relapse due to the selection of resistant clones and the development of refractory disease.8 In the case of relapse presenting with new or worse CRAB symptoms, immediate treatment is mandatory. A biochemical relapse or progression may require immediate treatment, or in the case of indolent disease, careful monthly monitoring of M-protein levels until significant progression.9 The indications for starting treatment at clinical and /or biochemical relapse were recently defined in a consensus paper by the IMWG.10 These are summarized in Table 2. In brief, the treatment of biochemical relapse is indicated if any of the following is present: a doubling of the serum M-protein, an increase of serum M-protein by ≥ 10 g/L, an increase of urine M-protein by ≥ 500 mg/24h or an increase of involved serum free light chains (FLC) level by ≥ 200 mg/L (plus abnormal ratio) by 2 measurements, 2 months apart. In the presence of high-risk factors, such as aggressive disease at diagnosis, a short treatment-free interval with a suboptimal response to the previous treatment line, imminent risk for organ dysfunction such as previous light chain-induced renal impairment, aggressive bone lesions or unfavorable cytogenetics t(4;14) or del17p, treatment should be initiated at the stage of biochemical relapse before serious symptomatic disease develops.11

with combination regimens. Intermediate-risk (60 %) and low-risk (20 %) patients may benefit from less intensive regimens or even monotherapy, respectively. Patient-related factors such as age and performance status may limit the number and type of treatment options. It is important to consider the previous line(s) of therapy, the disease responses and prior adverse events (Figure 1). The choices are also influenced by age, frailty status and the expectations of the patients and their families. The physical and emotional impact of hospitalization or frequent hospital visits should be considered. In many relapse trials the occurrence of any grade treatment-related adverse events is approximately 50% and serious adverse events (SAE) 20%. Treatmentrelated AEs are therefore a frequent cause of premature discontinuation, which will influence the outcome. It is important that the patient completes the planned treatment in order to achieve control of the disease.14 Quality of life is an important goal, which reflects the patient's ability to continue their activities.15,16 In Table 3 these factors are listed. As of 2015 the EMA has approved for the treatment of relapsed MM i) lenalidomide in combination with dexamethasone; ii) bortezomib alone or in combination with pegylated liposomal doxorubicin. Treatment with carfilzomib combined with lenalidomide and dexamethasone was also recently approved, while approval is expected for therapy with panobinostat combined with bortezomib and dexamethasone. For RRMM pomalidomide combined with dexamethasone has been approved. However, many other drugs are used that were not formally approved for this indication, such as thalidomide and bendamustine.

Treatment considerations at relapse When the indication for treatment has been made, various factors may be considered in order to make the optimal treatment choice. First, a risk estimation has to be made since the clinical course of the disease varies widely between patients. Twenty percent of patients have highrisk disease, which is associated with unfavorable cytogenetics del17p, t(4;14), add 1q, t(14;16), a high-risk gene expression profile, ISS3 and high serum lactodehydrogenase at diagnosis, or plasmacel leukemia or a rapid progression/relapse.12,13 These patients require immediate treatment

Prognostic factors at relapse and impact of response Results of prior treatment. The majority of patients with NDMM receive a triple-drug induction regimen that contains at least one novel agent plus dexamethasone and a third agent. The European Medicine Agency (EMA) has approved bortezomib and thalidomide for first-line treatment in transplant eligible and non-transplant eligible patients. Continuous lenalidomide plus dexamethasone has recently received approval by the EMA for use in nontransplant eligible NDMM patients based on the MM-020

Table 1. Revised International Myeloma Working Group diagnostic criteria for multiple myeloma and smouldering myeloma. Definition of multiple myeloma Clonal bone marrow plasma cells ≥10% or biopsy-proven bony or extramedullary plasmacytoma* and any one or more of the following myeloma defining events: • Myeloma defining events: • Evidence of end organ damage that can be attributed to the underlying plasma cell proliferative disorder, specifically: • Hypercalcaemia: serum calcium >0·25 mmol/L (>1 mg/dL) higher than the upper limit of normal or >2·75 mmol/L (>11 mg/dL) • Renal insufficiency: creatinine clearance <40 mL per min† or serum creatinine >177 mol/L (>2 mg/dL) • Anaemia: haemoglobin value of >20 g/L below the lower limit of normal, or a haemoglobin value <100 g/L • Bone lesions: one or more osteolytic lesions on skeletal radiography, CT, or PET-CT‡ • Any one or more of the following biomarkers of malignancy: • Clonal bone marrow plasma cell percentage* ≥60% • Involved:uninvolved serum free light chain ratio§ ≥100 • >1 focal lesions on MRI studies¶ Definition of smouldering multiple myeloma Both criteria must be met: • Serum monoclonal protein (IgG or IgA) ≥30 g/L or urinary monoclonal protein ≥500 mg per 24 h and/or clonal bone marrow plasma cells 10–60% • Absence of myeloma defining events or amyloidosis PET-CT: 18F-fluorodeoxyglucose PET with CT. *Clonality should be established by showing / -light-chain restriction on flow cytometry, immunohistochemistry, or immunofluorescence. Bone marrow plasma cell percentage should preferably be estimated from a core biopsy specimen; in case of a disparity between the aspirate and core biopsy, the highest value should be used. †Measured or estimated by validated equations. ‡If bone marrow has less than 10% clonal plasma cells, more than one bone lesion is required to distinguish from solitary plasmacytoma with minimal marrow involvement. §These values are based on the serum Freelite assay (The Binding Site Group, Birmingham, UK).The involved free light chain must be ≥100 mg/L. ¶Each focal lesion must be 5 mm or more in size.

haematologica | 2016; 101(4)

397


P. Sonneveld et al. Table 2. Indications for treatment at relapse.

Indications for treatment at relapse Clinical relapse • Development of new soft-tissue plasmacytomas or bone lesions • Definite increase (≥50%) in size of existing plasmacytomas or bone lesions • Hypercalcemia (≥11.5 mg/dL; 2.875 mmol/L) • Decrease in hemoglobin of ≥2 g/dL (1.25 mmol/L), or to < 10g/dL because of myeloma • Rise in serum creatinine by ≥2 mg/dL or more (≥177 mmol/L), due to myeloma • Hyperviscosity requiring therapeutic intervention Significant biochemical relapse in patients without clinical relapse • Doubling of the M-component in two consecutive measurements separated by 2 months with the reference value of 5 g/L, or • In two consecutive measurements any of the following increases: o the absolute levels of serum M protein by ≥10 g/L, or o an increase of urine M protein by ≥500 mg per 24 hours, or o an increase of involved FLC level by ≥20 mg/dL (plus an abnormal FLC ratio) or 25% increase (whichever is greater)

trial.17 Consequently, more patients will receive lenalidomide in first-line treatment, and this will reduce or change the number of options in relapsed MM. The IMWG has conducted a survey of the risk of progression in patients relapsing after prior therapy with IMiDs and bortezomib, showing that that the median OS and event-free survival (EFS) rates were 9 months and 5 months, respectively.18 In two individual studies single-agent bortezomib or bortezomib plus pegylated liposomal doxorubicin and dexamethasone were evaluated. A lower response rate was observed in thalidomide-exposed compared to thalidomide-naïve patients with bortezomib monotherapy as salvage treatment.19 When bortezomib was combined with pegylated liposomal doxorubicin and dexamethasone this difference was not observed.20 In an analysis of the MM009 and MM-010 trials comparing lenalidomide plus dexamethasone with dexamethasone in relapsed MM, thalidomide-naïve patients had a significantly better overall response rate (ORR) than patients who had received thalidomide. Also, time to progression (TTP, 13 vs. 8 months) and PFS (13 vs. 8 months) were superior in thalidomide-naïve patients, while OS was not different.21 In other trials a similar difference was observed between thalidomide-naïve versus thalidomide exposed patients who received subsequent treatment with VRD (bortezomib, lenalidomide, dexamethasone), VCD (bortezomib, cyclophosphamide, dexamethasone) or VTD (bortezomib, thalidomide, dexamethasone).22-24 In a recent update of the VISTA trial, patients who had been treated upfront with VMP (bortezomib, melphalan, prednisone) had a similar or better outcome following relapse treatment with bortezomib-based or IMiD-based regimens as compared to MP (melphalan, prednisone) treated patients.25 Other studies of the impact of upfront bortezomib treatment on the outcome of relapse treatment with lenalidomide plus dexamethasone showed less favorable results.26-28 Bortezomib retreatment has been investigated in a recent prospective Italian trial, where 40% of patients achieved a response with a TTP of 8 months and there was a better outcome in patients who achieved a complete response (CR) following prior bortezomib (see below).29 398

Achievement of response. In first relapse clinically relevant responses can be achieved in 40 – 50% of patients. An important question is whether the depth of response affects long-term outcome in relapsed MM (for review see Lonial et al.30) In the APEX trial the achievement of CR with bortezomib was associated with a longer time to next treatment compared with VGPR or PR (24 vs. 13 vs. 6 months).31 In the MM-009 and MM-010 trials using lenalidomide plus dexamethasone TTP and OS were significantly longer in patients who achieved VGPR or better compared with PR (TTP: 27 vs. 12 months; OS: not reached vs. 44 months).32 At subsequent relapses and in RRMM virtually no impact of CR/VGPR on OS or TTP is observed as was demonstrated in the MM-003 trial with pomalidomide plus low-dose dexamethasone.33 These data indicate that in first relapse CR may be a relevant treatment goal which can be actively pursued in patients. In second relapse and beyond the goal of treatment is to prevent organ impairment and to control the disease. Cytogenetics. In contrast to newly diagnosed MM the impact of unfavorable cytogenetics on the outcome of relapse treatment has not been well documented, due to the lack of consistent data in large clinical trials. In addition, there is a selection bias of patients able to participate in trials with experimental agents and the number of patients in Phase 2 trials is too small to allow subanalyses. Therefore, the results of cytogenetic subgroups should be considered with caution. In a trial in relapsed MM patients treated with bortezomib, gain of 1(q21) was associated with a worse PFS (7 vs. 2 months) and OS (24 vs. 5 months), while del17p, del(1p21) and del13q had no impact.34 In a trial with escalated dose bortezomib plus RD followed by lenalidomide maintenance in patients in first relapse no impact of unfavorable cytogenetics on PFS or OS was observed.35 In a prospective trial comparing VRD vs. lenalidomide plus dexamethasone (RD) in relapsed MM, the presence of +1(q21) was associated with shorter OS in the RD group, while the impact of t(4;14) and del13q was less clear. In this trial del17p had a poor outcome in both groups.24 In a study by the Intergroupe Francophone du Myelome (IFM), t(4;14) had a negative haematologica | 2016; 101(4)


Treatment of relapsed and refractory multiple myeloma

Figure 1. Global strategy for treatment strategy at relapse.

impact on OS in relapsed MM patients treated with lenalidomide plus dexamethasone.26 More recently, in a trial with single-agent carfilzomib, relapsed MM patients with del17p or t(4;14) or t(14;16) had a worse OS than patients without these karyotypes (19 vs. 9 months).36 In contrast, the Aspire trial using carfilzomib plus lenalidomide and dexamethasone showed no difference in PFS and TTP across subgroups with high-risk versus standardrisk cytogenetics.37 Finally, a subgroup analysis of the MM003 trial showed a benefit of response and PFS (3 months vs. 1 month) in patients with t(4;14) or del17p who received pomalidomide plus LD dexamethasone compared to dexamethasone alone.38 From these few studies one may conclude that at present there is no convincing evidence that patients with relapsed MM who have highrisk cytogenetics should be excluded from relapse treatment.

activity in patients with relapsed MM and are generally well tolerated.39-44 These agents have set the stage for the development of next-generation immunomodulatory drugs (IMiDs) and proteasome inhibitors, i.e. pomalidomide and carfilzomib in relapsed and/or refractory disease.33,37 In general, doublet or triplet regimens are preferred above single agents for optimal effect. Recently the IMWG published recommendations for global myeloma care including treatment options at relapse.45 Figure 1 gives a general strategy for treatment selection, which will be discussed below. In Figure 2 the European Myeloma Network guideline for myeloma care shows specific treatment choices, which are endorsed by the authors.11,14 The reader is also referred to recently published ESMO guidelines, other publications11,14,46,47 and to published overviews.48,49

Retreatment Treatment options at relapsed MM New regimens which include the novel agents thalidomide, bortezomib and lenalidomide as single-agents or in combination with dexamethasone have shown significant haematologica | 2016; 101(4)

Retreatment with an agent used previously is considered feasible, provided the treatment produced a clinically meaningful response of adequate duration and acceptable toxicity. In general, the minimal depth of the initial 399


P. Sonneveld et al.

response should be partial response, while the minimal duration should be at least 6 months. Trials and retrospective analyses have shown that retreatment with bortezomib is feasible and effective and does not incur cumulative toxicity.50-52 Lenalidomide retreatment is also feasible and may induce responses in up to 44 % of relapsed patients, and is better than retreatment with thalidomide.51,53 Tolerability is another important consideration, and factors like neuropathy, myelosuppression and thrombosis may influence the choice of therapy. For some patients a change to a less intensive schedule or dose may make the treatment better tolerated. For bortezomib changing to a weekly and subcutaneous schedule will reduce toxicity.54 If an effective alternative treatment is available at relapse, switching drug class is preferable, while previously used agents may then be considered in later lines.29,50 Patients may even become sensitive to (escalated dosages of) drugs to which they were previously refractory, based on the appearance of different tumor clones during subsequent stages of the disease.55

Thalidomide Thalidomide monotherapy has long been considered a valuable treatment option for patients with relapsed MM. However, there are no randomized phase III clinical trials supporting the use of thalidomide in this indication. In a systematic review of 42 clinical studies of thalidomide monotherapy in 1629 intent-to-treat patients, thalidomide was associated with a partial response or better in 29% of patients (CR 1.6%) and led to a median OS of 14 months.56 These results were validated by a meta-analysis limited to trials of ≥50 patients, which reported a similar ORR of 28% (CR 2% and PR 26%).57 Median doses were in the range of 200–800 mg/day, indicating the lack of standards regarding the optimal dose of thalidomide treatment. A later phase III comparison of low-dose (100 mg) and highdose (400 mg) thalidomide showed similar activity, with improved safety associated with low-dose thalidomide indicating that 100 mg should be the preferred dose.58 Retreatment with thalidomide was associated with a 30% ORR. Combining thalidomide with dexamethasone improves its efficacy in relapsed MM and phase II studies have reported an ORR of 41–56%.59,60 The efficacy is further improved when thalidomide is used in triple or even quadruple combinations, including thalidomide with dexamethasone and bortezomib (VTD), dexamethasone and cyclophosphamide (CTD), bortezomib plus melphalan and prednisone (VMPT) or dexamethasone (VMDT), lenalidomide, melphalan and prednisone (RMPT), or VTD with pegylated liposomal doxorubicin.23,61-66 These regimens have been associated with an ORR of 63–90% with CR being reported in 2–35% of patients. Conclusion. Thalidomide combined with dexamethasone is still an option for relapsed patients especially if they are thalidomide-naïve, for whom an oral treatment schedule is needed and who are not eligible for bortezomib or lenalidomide-based treatment. However, the use of thalidomide has diminished because its continuous use is hampered by poor tolerability, the risk of thrombotic events and/or peripheral neuropathy as well as the availability of effective alternative treatments.

Bortezomib Bortezomib is an effective drug for patients with relapsed MM. It can be safely and effectively adminis400

Table 3. Patient- and treatment-related factors in the selection of treatment at relapse MM.

Patient- and treatment-related factors in the selection of treatment at relapse MM Patient-related Factors Age and Frailty WHO Performance Status Comorbidities Transplant eligibility Residual or late effects of prior therapies Pre-existing (peripheral) neuropathy and/or thrombotic events Disease-related Factors Type and risk status of initial disease Response and response duration following prior therapies Presence of refractory disease Aggressiveness of current relapse Treatment-related factors Response and/or refractoriness to prior therapies Previous use of IMiDs and proteasome inhibitor Previous use of alkylators Prior HDT with ASCT Single, dual or triple drug combination Type and severity of adverse events related to prior therapies Bone marrow reserve Expected efficacy & toxicity of proposed therapy Availability, cost and management requirements Expectations of the patient

tered to patients with renal impairment.67 The drug is now routinely administered subcutaneously and may be used in a weekly schedule in elderly patients. Its main toxicities include peripheral neuropathy which may preclude further treatment68, gastrointestinal symptoms and transient thrombocytopenia. The phase III APEX trial reported that bortezomib, as compared with dexamethasone, improves outcomes in patients with RRMM.69 Patients treated with bortezomib had higher ORRs (38% vs. 18%; CR 6% vs. <1%; P<0.001 for each), longer TTP (6.2 months vs. 3.5 months; P<0.001), and better 1-year OS (80% vs. 66%; P=0.003) than those treated with dexamethasone. An updated analysis, based on a median follow-up period of 22 months for surviving patients, confirmed a survival benefit of 6 months for bortezomib compared with dexamethasone (median OS 29.8 months vs. 23.7 months) despite a 62% crossover of patients from the dexamethasone group to the bortezomib group.70 Retreatment with bortezomib has clinical value if the patients were responsive previously, and if the response lasted more than 6 months.29 Bortezomib is effective in combination with other agents and studies have reported improved ORR and good tolerability in combination with dexamethasone and/or pegylated liposomal doxorubicin.20,71-74 Bortezomib is also effective as part of quadruple drug salvage regimens (ORR 56–88%; CR 6–46%;(VGPR or better 34–55%), and favorable response rates have been reported of bortezomib with CTD (VCTD), VMPT, and bortezomib with doxorubicin, dexamethasone, and lenalidomide (DVD-R).62,75,76 Bortezomib also appears to enhance the effects of lenalidomide, and combinations of these two agents are currently under clinical evaluation. Conclusion: Bortezomib combined with dexamethasone is an effective treatment for RRMM. Its efficacy is increased when combined with thalidomide, cyclophoshaematologica | 2016; 101(4)


Treatment of relapsed and refractory multiple myeloma

Figure 2. Guideline for treatment choices in patients with relapsed and/or refractory multiple myeloma developed by the European Myeloma Network.

phamide or an anthracyclin. Patients should be carefully monitored for neuropathy, in case the dose and schedule need to be reduced.

Lenalidomide Lenalidomide as a single agent is active and well tolerated in patients with relapsed MM, and adding dexamethasone to lenalidomide has been shown to further improve response rates by up to 30%.77 In the MM-009 and MM-010 trials patients were treated with lenalidomide plus dexamethasone until disease progression or unacceptable toxicity was observed.78,79 In the MM-010 trial lenalidomide plus dexamethasone significantly improved ORR (60.2% vs. 24.0%), TTP (11.3 months vs. 4.7 months), and OS (not reached vs. 20.6 months; HR = 0.66) compared with dexamethasone alone. An analysis of pooled data from the MM-009 and MM-010 trials haematologica | 2016; 101(4)

after a follow-up of 48 months confirmed the improved outcomes with lenalidomide plus dexamethasone, which significantly improved ORR (60.6% vs. 21.9%), duration of response (15.8 vs. 7 months), and median TTP (13.4 vs. 4.6 months). OS was also significantly longer in patients treated with lenalidomide plus dexamethasone compared with dexamethasone alone (38 vs. 31.6 months).80 A later analysis suggested that to achieve a maximum PFS benefit, patients should be treated for at least 12 months with full-dose lenalidomide plus dexamethasone, followed by lower-dose continued therapy.81 Because of the good tolerability of lenalidomide, high clinical activity and its oral formulation various combination regimens have been evaluated. Among these, the addition of cyclophosphamide to lenalidomide and dexamethasone for continuous treatment has clinical value in patients with suboptimal response.82 As lenalidomide 401


P. Sonneveld et al.

appears to enhance the anti-myeloma effects of bortezomib, these two agents have been evaluated as combination therapy in patients with relapsed MM.83 Retreatment with lenalidomide has shown a significant ORR. The combination lenalidomide, bortezomib and dexamethasone (RVD) is an active and well-tolerated regimen in patients with relapsed MM and can overcome drug resistance in patients previously treated with lenalidomide, bortezomib, thalidomide, or ASCT. With a follow-up of >2 years, MR or better was achieved by 78% of patients, including PR or better in 64% and CR or near CR in 25% of patients; the median PFS was 9.5 months and the median OS was 26 months. A four-agent combination regimen (RMPT) followed by maintenance with lenalidomide was effective and well tolerated in patients with relapsed MM with a 1-year PFS and OS of 51% and 72%, respectively.84 The European Myeloma Network has defined a consensus statement for the use of lenalidomide.85 Conclusion. Lenalidomide combined with dexamethasone is currently the most efficient option for relapsed MM and may be combined with bortezomib, cyclophosphamide or other agents. It is necessary to give the full-dose with dexamethasone during reinduction and continue with a lower-dose until progression.

Pomalidomide Recently, pomalidomide alone or combined with dexamethasone was approved by the EMA, based on the results of the MM-003 trial for use in patients who have received at least 2 prior therapies including lenalidomide and bortezomib and have demonstrated disease progression. Dosing variations in early clinical trials defined the dose of pomalidomide at 4 mg on days 1-21 in a 28-day schedule.86 Pomalidomide has demonstrated an ORR of 33% and a median response duration of 8 months in heavily pretreated patients.87 In the extended Phase 2 part of the MM-002 trial, PFS and OS were 4 and 16 months, respectively.88 The pivotal trial demonstrating the superiority of pomalidomide with dexamethasone was MM-003, which showed a better PFS (4 vs. 1.9 months) and OS (12 vs. 8 months) in spite of crossover from the dexamethasone arm for those patients who did not respond.33 Recently an expert panel consensus statement was published on the optimal use of pomalidomide in RRMM.89 Conclusion. Pomalidomide is the only agent with demonstrated clinical activity in end-stage disease in patients who have received bortezomib and lenalidomide. It should be given until progression, preferably combined with dexamethasone.

Corticosteroids and conventional agents Dexamethasone is added to most other therapies at a weekly dose of 20-40 mg. Its toxicities, i.e. osteoporosis and infections, limit prolonged use of the drug. However, in patients who have exhausted other options weekly dexamethasone or continuous low-dose (20 mg) prednisolone may be considered. Cyclophosphamide is an alkylating agent that is usually well tolerated and can be given orally or intravenously. It is often combined with bortezomib in the VCD or CyBorD schedules or with lenalidomide, but it can also be taken alone in a 300 mg/m2 weekly regimen. Standard-dose intermittent oral 402

melphalan may also be a valuable option for economic reasons or when patients have no other treatment options. High-dose chemotherapy such as DCEP and DT-PACE can be given in RRMM, although responses are usually short.90

Bendamustine This is a bi-functional alkylating agent that was approved for the treatment of chronic lymphocytic leukemia, indolent B-cell non-Hodgkin lymphoma and NDMM in patients who cannot tolerate thalidomide or bortezomib because of neuropathy. Bendamustine is currently undergoing clinical evaluation as an anti-myeloma treatment. Early studies have reported promising efficacy and good tolerability with a regimen of bendamustine in combination with thalidomide, in combination with lenalidomide and dexamethasone, or in combination with bortezomib and dexamethasone in patients with relapsed MM.91-93

High-Dose Therapy (HDT) with autologous stem cell transplantation (ASCT) or Allogeneic SCT In transplant eligible patients with relapse or progression, HDT followed by ASCT may be considered. In patients who did not receive HDT before, it is the treatment of choice if stem cells can be obtained.94 In general, HDT plus ASCT is valuable if patients had responded to a previous HDT and have achieved at least 2 years of PFS. This has also become an option in patients of a more advanced age.95 As shown recently in a randomized trial by the NCRI a second high-dose melphalan plus ASCT is a better option than weekly cyclophosphamide in patients who received ASCT at least 18 months before, resulting in a longer time to progression (19 vs. 11 months).96 Two international trials by IFM with an American consortium and by the European Myeloma Network are currently investigating the benefit of HDT in newly diagnosed MM versus first relapse. Allogeneic transplantation is an experimental option for use in clinical trials in patients with high-risk disease and unfavorable karyotypes.97 The European Group for Blood and Marrow Transplantation published a long follow-up of NDMM and relapsed MM patients treated with autologous/reduced-intensity allogeneic SCT vs. autologous SCT, showing a superior OS with the former combination (47% vs. 31% at 96 months).98 The same group more specifically analyzed the results of reduced-intensity allogeneic SCT in relapsed MM.99 At the present time, most investigators currently regard allogeneic SCT as an option for younger, fit patients with high-risk disease in first relapse.

Novel agents under clinical development In addition to approved novel agents, ongoing trials show promising activity of various agents of different drug classes. While these drugs are not available as yet, they are considered experimental in Europe. Because of their interest for patients wishing to participate in clinical trials they will be briefly discussed. An extensive, recent summary of this topic was published by the IMWG group.100 Carfilzomib when combined with lenalidomide and dexamethasone has recently demonstrated significant clinical activity in the ASPIRE trial, leading to an unprecedented PFS of 26 months versus 17 months in the control group.37 The combination was well tolerated and there was a clinical benefit in the quality of life. In the ENDEAVhaematologica | 2016; 101(4)


Treatment of relapsed and refractory multiple myeloma

OUR study, carfilzomib plus dexamethasone was superior to bortezomib plus dexamethasone in relapsed MM, including patients who were treated with but not refractory to bortezomib.101 The EMA evaluated the drug in 2015 and full access to the market is expected in 2016. Another trial comparing single agent carfilzomib with dexamethasone (FOCUS) was not positive because of good results in the control arm, where > 90 % received cyclophosphamide/dexamethasone. This also suggests that carfilzomib must be combined with other agents. HDAC inhibitors. Panobinostat and vorinostat are epigenetic drugs that can be combined with other agents.102,103 Vorinostat combined with bortezomib and dexamethasone showed a PFS advantage of only 1 month compared with bortezomib and dexamethasone. Panobinostat had a significant advantage when combined with bortezomib and dexamethasone over those drugs with placebo, however, at the cost of gastrointestinal symptoms and fatigue. The EMA and FDA have the drug combination under consideration. Antibodies. Amongst novel agents the monoclonal antibodies elotuzumab and daratumumab carry a high promise for the treatment of relapsed MM. Elotuzumab is a monoclonal antibody which targets SLAMF-7, which is present on the surface of plasma cells. While it has little single-agent activity in RRMM, elotuzumab combined with lenalidomide and low-dose dexamethasone showed more than 80 % ORR without significant toxicity.104 In Phase 2 studies the lower-dose of 10 mg/kg was associated with a longer PFS than the dose of 20 mg/kg (33 versus 19 months).105 The results of the Phase 3 trial ELOQUENT-2 confirm the effect of elotuzumab in relapsed MM.106 Daratumumab is an antibody that targets CD38 and kills plasma cells through ADCC and CDC mediated mechanisms.107 In a study in patients of whom 75 % were refractory to bortezomib and lenalidomide, single agent daratumumab resulted in an ORR of 36% at the 16mg/kg dose, while 65% of responding patients were progressionfree at 12 months.108 Based on preclinical studies lenalidomide has been identified as a synergistic partner for daratumumab.109 Clinical trials have been initiated using various combinations of daratumumab with other drugs. SAR650984 is another anti-CD38 antibody which is currently being investigated in clinical trials. Conclusions. Monoclonal antibodies are a novel class of agents which specifically target plasma cells, have strong clinical activity in particular when combined with other agents, and show limited toxicity. It is expected that antibodies will have an important place in the future treatment of relapsed MM.

References 1. Palumbo A, Anderson K. Multiple myeloma. N Engl J Med. 2011;364(11):1046-1060. 2. Rajkumar SV, Dimopoulos MA, Palumbo A, et al. International Myeloma Working Group updated criteria for the diagnosis of multiple myeloma. Lancet Oncol. 2014;15(12):e538-e548. 3. Rajkumar SV. Treatment of multiple myeloma. Nat Rev Clin Oncol. 2011;8(8):479-491.

haematologica | 2016; 101(4)

Supportive care Patients with relapsed MM are vulnerable due to the presence of disease, prior exposure to chemotherapy, myelosuppression, prolonged exposure to corticosteroids and organ impairment. Frequent infections and bone disease are common and should be adequately prevented and treated. Intravenous zoledronate or pamidronate should be started or restarted at relapse, with calcium and vitamin D supplements. Low-dose local radiotherapy (20-40 Gy) may be administered to local bone lesions in case of pain or imminent fractures. Infections should be managed proactively, especially those of encapsulated pathogens. Prophylactic vaccination is recommended for influenza A and B, pneumococci and hemophilus influenza. Anemia may be treated with erythropoietin (40.000 U weekly) or darbopoietin (500 mcg q 3 weeks) or transfusion. Patients with increased risk for venous thrombolic events and those who will be treated with thalidomide or lenalidomide should receive prophylaxis with aspirin (1 risk factor) or LMWH (≥2 risk factors). Treatment for polyneuropathy and pain should be administered according to published guidelines.110 For detailed guidelines the reader is referred to the IMWG consensus.45

Conclusions and future directions Treatment with thalidomide, lenalidomide, and bortezomib has improved responses in patients with relapsed MM. Continuous or repeated therapy with lenalidomide and bortezomib is well tolerated and leads to durable clinical responses, resulting in improved PFS and OS. The majority of adverse events associated with these novel agents are hematologic and can be managed using dose modifications and/or growth factor support; however, patients on long-term thalidomide or bortezomib should be regularly monitored for the development of peripheral neuropathy. Long-term treatment with novel agents could result in the emergence of drug-resistant MM clones, especially in patients with adverse FISH cytogenetics. Therefore, well-designed phase III studies with long-term follow-up are required to confirm the clinical benefits and safety of the continuous treatment approaches in MM patients. Approaches for patients relapsing on long-term treatment are yet to be defined. Continuous therapy from first relapse to disease progression has the potential to maintain the suppression of residual disease, prolong the time to subsequent relapse and extend OS rates. In effect, this treatment strategy may generate prolonged control of the disease. If this can be achieved in RRMM patients, it will represent a paradigm shift, allowing MM to be managed as a chronic illness.

4. Durie BG, Harousseau JL, Miguel JS, et al. International uniform response criteria for multiple myeloma. Leukemia. 2006;20(9): 1467-1473. 5. Kyle RA, Rajkumar SV. Criteria for diagnosis, staging, risk stratification and response assessment of multiple myeloma. Leukemia. 2009;23(1):3-9. 6. Rajkumar SV, Harousseau JL, Durie B, et al. Consensus recommendations for the uniform reporting of clinical trials: report of the International Myeloma Workshop

Consensus Panel 1. Blood. 2011;117(18): 4691-4695. 7. Anderson KC, Kyle RA, Rajkumar SV, et al. Clinically relevant end points and new drug approvals for myeloma. Leukemia. 2008;22 (2):231-239. 8. Magrangeas F, Avet-Loiseau H, Gouraud W, et al. Minor clone provides a reservoir for relapse in multiple myeloma. Leukemia. 2013;27(2):473-481. 9. Fernández de Larrea C, Jiménez R, Rosiñol L, et al. Pattern of relapse and progression

403


P. Sonneveld et al.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

404

after autologous SCT as upfront treatment for multiple myeloma. Bone Marrow Transplant. 2014. 49(2):223-227. Palumbo A, Rajkumar SV , San Miguel JF, et al. International Myeloma Working Group consensus statement for the management, treatment, and supportive care of patients with myeloma not eligible for standard autologous stem-cell transplantation. J Clin Oncol. 2014;32(6):587-600. Ludwig H, Sonneveld P, Davies F, et al., European perspective on multiple myeloma treatment strategies in 2014. Oncologist. 2014;19(8):829-844. Moreau P, Cavo M, Sonneveld P, et al., Combination of international scoring system 3, high lactate dehydrogenase, and t(4;14) and/or del(17p) identifies patients with multiple myeloma (MM) treated with front-line autologous stem-cell transplantation at high risk of early MM progressionrelated death. J Clin Oncol. 2014;32(20):2173-2180. Mikhael JR, Dingli D, Roy V, et al., Management of newly diagnosed symptomatic multiple myeloma: updated Mayo Stratification of Myeloma and Risk-Adapted Therapy (mSMART) consensus guidelines 2013. Mayo Clin Proc. 2013;88(4):360-376. Ludwig H, Sonneveld P. Disease control in patients with relapsed and/or refractory multiple myeloma: what is the optimal duration of therapy? Leuk Res. 2012;36 Suppl 1:S27-34. Sonneveld P, Verelst SG, Lewis P, et al. Review of health-related quality of life data in multiple myeloma patients treated with novel agents. Leukemia. 2013;27(10):19591969. Dimopoulos MA, Palumbo A, Hajek R,et al. Factors that influence health-related quality of life in newly diagnosed patients with multiple myeloma aged >/= 65 years treated with melphalan, prednisone and lenalidomide followed by lenalidomide maintenance: results of a randomized trial. Leuk Lymphoma. 2014;55(7):1489-1497. Benboubker L, Dimopoulos MA ,Dispenzieri A, et al., Lenalidomide and dexamethasone in transplant-ineligible patients with myeloma. N Engl J Med. 2014;371(10):906-917. Kumar SK, Lee JH, Lahuerta JJ, et al. Risk of progression and survival in multiple myeloma relapsing after therapy with IMiDs and bortezomib: a multicenter international myeloma working group study. Leukemia. 2012;26(1):149-157. Vogl DT, Stadtmauer EA, Richardson PG, et al. Impact of prior therapies on the relative efficacy of bortezomib compared with dexamethasone in patients with relapsed/refractory multiple myeloma. Br J Haematol. 2009;147(4):531-534. Sonneveld P, Hajek R, Nagler A, et al. Combined pegylated liposomal doxorubicin and bortezomib is highly effective in patients with recurrent or refractory multiple myeloma who received prior thalidomide/lenalidomide therapy. Cancer. 2008; 112(7):1529-1537. Wang M, Dimopoulos MA , Chen C, et al. Lenalidomide plus dexamethasone is more effective than dexamethasone alone in patients with relapsed or refractory multiple myeloma regardless of prior thalidomide exposure. Blood. 2008;112(12):4445-4451. Kropff M, Bisping G, Schuck E, et al. Bortezomib in combination with intermediate-dose dexamethasone and continuous low-dose oral cyclophosphamide for relapsed multiple myeloma. Br J Haematol.

2007;138(3):330-337. 23. Pineda-Roman M, Zangari M, van Rhee F, et al. VTD combination therapy with bortezomib-thalidomide-dexamethasone is highly effective in advanced and refractory multiple myeloma. Leukemia. 2008;22(7): 14191427. 24. Dimopoulos MA, Kastritis E, Christoulas D, et al. Treatment of patients with relapsed/refractory multiple myeloma with lenalidomide and dexamethasone with or without bortezomib: prospective evaluation of the impact of cytogenetic abnormalities and of previous therapies. Leukemia. 2010;24(10):1769-1778. 25. Mateos MV, Richardson PG, Schlag R, et al., Bortezomib plus melphalan and prednisone compared with melphalan and prednisone in previously untreated multiple myeloma: updated follow-up and impact of subsequent therapy in the phase III VISTA trial. J Clin Oncol. 2010;28(13):2259-2266. 26. Avet-Loiseau H, Soulier J, Fermand JP, et al. Impact of high-risk cytogenetics and prior therapy on outcomes in patients with advanced relapsed or refractory multiple myeloma treated with lenalidomide plus dexamethasone. Leukemia. 2010;24(3):623628. 27. Reece D, Song KW, Tu T, et al. Influence of cytogenetics in patients with relapsed or refractory multiple myeloma treated with lenalidomide plus dexamethasone: adverse effect of deletion 17p13. Blood. 2009;114 (3):522-525. 28. Klein U, Neben K, Hielscher T, et al., Lenalidomide in combination with dexamethasone: effective regimen in patients with relapsed or refractory multiple myeloma complicated by renal impairment. Ann Hematol. 2011;90(4):429-439. 29. Petrucci MT, Giraldo P, Corradini P, et al. A prospective, international phase 2 study of bortezomib retreatment in patients with relapsed multiple myeloma. Br J Haematol. 2013;160(5):649-659. 30. Lonial S, Anderson KC. Association of response endpoints with survival outcomes in multiple myeloma. Leukemia. 2014;28 (2):258-268. 31. Niesvizky R, Richardson PG, Rajkumar SV, et al. The relationship between quality of response and clinical benefit for patients treated on the bortezomib arm of the international, randomized, phase 3 APEX trial in relapsed multiple myeloma. Br J Haematol. 2008;143(1):46-53. 32. Harousseau JL, Dimopoulos MA, Wang M, et al. Better quality of response to lenalidomide plus dexamethasone is associated with improved clinical outcomes in patients with relapsed or refractory multiple myeloma. Haematologica. 2010;95(10):1738-1744. 33. San Miguel J, Weisel K, Moreau P, et al. Pomalidomide plus low-dose dexamethasone versus high-dose dexamethasone alone for patients with relapsed and refractory multiple myeloma (MM-003): a randomised, open-label, phase 3 trial. Lancet Oncol. 2013;14(11):1055-1066. 34. Chang H, Trieu Y, Qi X, Jiang NN, Xu W, Reece D. Impact of cytogenetics in patients with relapsed or refractory multiple myeloma treated with bortezomib: Adverse effect of 1q21 gains. Leuk Res. 2011;35(1):95-98. 35. Jakubowiak AJ, Siegel DS, Martin T, et al. Treatment outcomes in patients with relapsed and refractory multiple myeloma and high-risk cytogenetics receiving singleagent carfilzomib in the PX-171-003-A1 study. Leukemia. 2013;27(12):2351-2356. 36. Stewart AK, Rajkumar SV, Dimopoulos MA,

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

et al. Carfilzomib, Lenalidomide, and Dexamethasone for Relapsed Multiple Myeloma. N Engl J Med. 2015;372(2):142152. Glasmacher A, von Lilienfeld-Toal M. The current status of thalidomide in the management of multiple myeloma. Acta Haematol. 2005;114 Suppl 1:3-7. Richardson PG, Boral AL, Anderson KC. Bortezomib in multiple myeloma (vol 352, pg 2487, 2005). N Engl J Med. 2005; 353(12):1297-1298. Richardson PG, Barlogie B, Berenson J, et al., A phase 2 study of bortezomib in relapsed, refractory myeloma. N Engl J Med. 2003;348(26):2609-2617. Richardson P, Jagannath S, Schlossman R, et al. Thalidomide for relapsed multiple myeloma after high dose therapy and stem cell transplantation: Results of an open-label, multicenter phase two study of efficacy, toxicity and correlation with biological activity. Blood. 2002;100(11):813a-813a. Dimopoulos M, Spencer A, Attal M, et al. Lenalidomide plus dexamethasone for relapsed or refractory multiple myeloma. N Engl J Med. 2007;357(21):2123-2132. Weber DM, Chen C, Niesvizky R, et al., Lenalidomide plus dexamethasone for relapsed multiple myeloma in North America. N Engl J Med. 2007;357(21):21332142. Ludwig H, Miguel JS, Dimopoulos MA, et al., International Myeloma Working Group recommendations for global myeloma care. Leukemia. 2014;28(5):981-992. van de Donk NW, Lokhorst HM, Dimopoulos M, et al. Treatment of relapsed and refractory multiple myeloma in the era of novel agents. Cancer Treat Rev. 2011;37(4):266-283. Moreau P, San Miguel J, Ludwig H, et al. Multiple myeloma: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up. Ann Oncol. 2013. 24 Suppl 6:vi13337. Gentili S, Lonial S. Relapsed and refractory multiple myeloma: new therapeutic strategies. Hematol Oncol Clin North Am. 2014;28(5):861-890. Engelhardt M, Terpos E, Kleber M, et al. European Myeloma Network recommendations on the evaluation and treatment of newly diagnosed patients with multiple myeloma. Haematologica. 2014;99(2):232242. Knopf KB, Duh MS, Lafeuille MH, et al. Meta-analysis of the efficacy and safety of bortezomib re-treatment in patients with multiple myeloma. Clin Lymphoma Myeloma Leuk. 2014;14(5):380-388. Mohty B, El-Cheikh J, Yakoub-Agha I, et al. Treatment strategies in relapsed and refractory multiple myeloma: a focus on drug sequencing and 'retreatment' approaches in the era of novel agents. Leukemia. 2012;26(1):73-85. Taverna C, Voegeli J, Trojan A, et al. Effective response with bortezomib retreatment in relapsed multiple myeloma--a multicentre retrospective survey in Switzerland. Swiss Med Wkly. 2012; 142:w13562. Madan S, Lacy MQ, Dispenzieri A, et al. Efficacy of retreatment with immunomodulatory drugs (IMiDs) in patients receiving IMiDs for initial therapy of newly diagnosed multiple myeloma. Blood. 2011;118(7): 1763-1765. Moreau P, Coiteux V, Hulin C, et al. Prospective comparison of subcutaneous versus intravenous administration of bortezomib in patients with multiple myeloma.

haematologica | 2016; 101(4)


Treatment of relapsed and refractory multiple myeloma Haematologica. 2008;93(12):1908-1911. 53. Keats JJ, Chesi M, Egan JB, et al. Clonal competition with alternating dominance in multiple myeloma. Blood. 2012;120(5): 10671076. 54. Glasmacher A, Hahn C, Hoffmann F, et al. A systematic review of phase-II trials of thalidomide monotherapy in patients with relapsed or refractory multiple myeloma. Br J Haematol. 2006;132(5):584-593. 55. Prince HM, Schenkel B, Mileshkin L. An analysis of clinical trials assessing the efficacy and safety of single-agent thalidomide in patients with relapsed or refractory multiple myeloma. Leuk Lymphoma. 2007;48 (1):4655. 56. Yakoub-Agha I, Mary JY, Hulin C, et al. Low-dose vs. high-dose thalidomide for advanced multiple myeloma: a prospective trial from the Intergroupe Francophone du Myelome. Eur J Haematol. 2012;88(3):249259. 57. Dimopoulos MA, Zervas K, Kouvatseas G, et al. Thalidomide and dexamethasone combination for refractory multiple myeloma. Ann Oncol. 2001;12(7):991-995. 58. Palumbo A, Bertola A, Falco P, et al. Efficacy of low-dose thalidomide and dexamethasone as first salvage regimen in multiple myeloma. Hematol J. 2004;5(4):318-324. 59. Kyriakou C, Thomson K, D'Sa S, et al. Lowdose thalidomide in combination with oral weekly cyclophosphamide and pulsed dexamethasone is a well tolerated and effective regimen in patients with relapsed and refractory multiple myeloma. Br J Haematol. 2005;129(6):763-770. 60. Palumbo A, Ambrosini MT, Benevolo G, et al. Bortezomib, melphalan, prednisone, and thalidomide for relapsed multiple myeloma. Blood. 2007;109(7):2767-2772. 61. Terpos E, Kastritis E, Roussou M, et al. The combination of bortezomib, melphalan, dexamethasone and intermittent thalidomide is an effective regimen for relapsed/refractory myeloma and is associated with improvement of abnormal bone metabolism and angiogenesis. Leukemia. 2008;22(12):2247-2256. 62. Offidani M, Bringhen S, Corvatta L, et al. Thalidomide-dexamethasone plus pegylated liposomal doxorubicin vs. thalidomidedexamethasone: a case-matched study in advanced multiple myeloma. Eur J Haematol. 2007;78(4):297-302. 63. Palumbo A, Larocca A, Genuardi M, et al. Melphalan, prednisone, thalidomide and defibrotide in relapsed/refractory multiple myeloma: results of a multicenter phase I/II trial. Haematologica. 2010;95(7):1144-1149. 64. Garderet L, Iacobelli S, Moreau P, et al. Superiority of the triple combination of bortezomib-thalidomide-dexamethasone over the dual combination of thalidomidedexamethasone in patients with multiple myeloma progressing or relapsing after autologous transplantation: the MMVAR/IFM 2005-04 Randomized Phase III Trial from the Chronic Leukemia Working Party of the European Group for Blood and Marrow Transplantation. J Clin Oncol. 2012;30(20):2475-2482. 65. Scheid C, Sonnveld P, Schmidt-Wolf IG, et al. Bortezomib before and after autologous stem cell transplantation overcomes the negative prognostic impact of renal impairment in newly diagnosed multiple myeloma: a subgroup analysis from the HOVON65/GMMG-HD4 trial. Haematologica. 2014;99(1):148-154. 66. Delforge M, BladĂŠ J, Dimopoulos MA, et al. Treatment-related peripheral neuropathy in

haematologica | 2016; 101(4)

67.

68.

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

80.

multiple myeloma: the challenge continues. Lancet Oncol. 2010;11(11):1086-1095. Richardson PG, Sonneveld P, Schuster MW, et al. Bortezomib or high-dose dexamethasone for relapsed multiple myeloma. N Engl J Med. 2005;352(24):2487-2498. Richardson PG, Sonneveld P, Schuster MW, et al. Extended follow-up of a phase 3 trial in relapsed multiple myeloma: final time-toevent results of the APEX trial. Blood. 2007;110(10):3557-3560. Jagannath S, Barlogie B, Berenson JR, et al. Updated survival analyses after prolonged follow-up of the phase 2, multicenter CREST study of bortezomib in relapsed or refractory multiple myeloma. Br J Haematol. 2008;143(4):537-540. Mikhael JR, Belch AR, Prince HM, et al. High response rate to bortezomib with or without dexamethasone in patients with relapsed or refractory multiple myeloma: results of a global phase 3b expanded access program. Br J Haematol. 2009;144(2):169175. Orlowski RZ, Nagler A, Sonneveld P, et al. Randomized phase III study of pegylated liposomal doxorubicin plus bortezomib compared with bortezomib alone in relapsed or refractory multiple myeloma: combination therapy improves time to progression. J Clin Oncol. 2007;25(25):38923901. Dimopoulos MA, Orlowski RZ, Facon T, et al. Retrospective matched-pairs analysis of bortezomib plus dexamethasone versus bortezomib monotherapy in relapsed multiple myeloma. Haematologica. 2015;100 (1):100-106. Kim YK, Sohn SK, Lee JH, et al. Clinical efficacy of a bortezomib, cyclophosphamide, thalidomide, and dexamethasone (Vel-CTD) regimen in patients with relapsed or refractory multiple myeloma: a phase II study. Ann Hematol. 2010;89(5): 475-482. Berenson JR, Yellin O, Kazamel T, et al. A phase 2 study of pegylated liposomal doxorubicin, bortezomib, dexamethasone and lenalidomide for patients with relapsed/refractory multiple myeloma. Leukemia. 2012;26(7):1675-1680. Richardson PG, Blood E, Mitsiades CS, et al. A randomized phase 2 study of lenalidomide therapy for patients with relapsed or relapsed and refractory multiple myeloma. Blood. 2006;108(10):3458-3464. Dimopoulos M, Spencer A, Attal M, et al. Lenalidomide plus dexamethasone for relapsed or refractory multiple myeloma. N Engl J Med. 2007;357(21):2123-2132. Weber DM, Chen C, Niesvizky R, et al. Lenalidomide plus dexamethasone for relapsed multiple myeloma in North America. N Engl J Med. 2007;357(21):21332142. Dimopoulos MA, Chen C, Spencer A, et al. Long-term follow-up on overall survival from the MM-009 and MM-010 phase III trials of lenalidomide plus dexamethasone in patients with relapsed or refractory multiple myeloma. Leukemia. 2009;23(11): 21472152. Dimopoulos MA, Hussein M, Swern AS, Weber D. Impact of lenalidomide dose on progression-free survival in patients with relapsed or refractory multiple myeloma. Leukemia. 2011;25(10):1620-1626. van de Donk NW, Wittebol S, Minnema MC, et al. Lenalidomide (Revlimid) combined with continuous oral cyclophosphamide (endoxan) and prednisone (REP) is effective in lenalidomide/dexamethasonerefractory myeloma. Br J Haematol.

2010;148 (2):335-337. 81. Richardson PG, Xie W, Jagannath S, et al. A phase 2 trial of lenalidomide, bortezomib, and dexamethasone in patients with relapsed and relapsed/refractory myeloma. Blood. 2014;123(10):1461-1469. 82. Palumbo A, Larocca A, Falco P, et al. Lenalidomide, melphalan, prednisone and thalidomide (RMPT) for relapsed/refractory multiple myeloma. Leukemia. 2010;24(5): 1037-1042. 83. Dimopoulos MA, Palumbo A, Attal M, et al. Optimizing the use of lenalidomide in relapsed or refractory multiple myeloma: consensus statement. Leukemia. 2011;25 (5):749-760. 84. Leleu X, Attal M, Arnulf B, et al. Pomalidomide plus low-dose dexamethasone is active and well tolerated in bortezomib and lenalidomide-refractory multiple myeloma: Intergroupe Francophone du Myelome 2009-02. Blood. 2013;121(11): 1968-1975. 85. Richardson PG, Siegel D, Baz R, et al. Phase 1 study of pomalidomide MTD, safety, and efficacy in patients with refractory multiple myeloma who have received lenalidomide and bortezomib. Blood. 2013;121(11):19611967. 86. Richardson PG, Siegel DS, Vij R, et al. Pomalidomide alone or in combination with low-dose dexamethasone in relapsed and refractory multiple myeloma: a randomized phase 2 study. Blood. 2014;123 (12):18261832. 87. Dimopoulos MA, Leleu X, Palumbo A, et al. Expert panel consensus statement on the optimal use of pomalidomide in relapsed and refractory multiple myeloma. Leukemia. 2014;28(8):1573-1585. 88. Gerrie AS, Mikhael JR, Cheng L, et al. D(T)PACE as salvage therapy for aggressive or refractory multiple myeloma. Br J Haematol. 2013;161(6):802-810. 89. Ramasamy K, Hazel B, Mahmood S, et al. Bendamustine in combination with thalidomide and dexamethasone is an effective therapy for myeloma patients with end stage renal disease. Br J Haematol. 2011;155(5):632-634. 90. Rodon P, Hulin C, Pegourie B, et al. Phase 2 study of bendamustine, bortezomib and dexamethasone as second-line treatment for elderly patients with multiple myeloma: the Intergroupe Francophone du Myelome 2009-01 trial. Haematologica. 2015;100 (2):e56-59. 91. Lentzsch S, O'Sullivan A, Kennedy RC, et al. Combination of bendamustine, lenalidomide, and dexamethasone (BLD) in patients with relapsed or refractory multiple myeloma is feasible and highly effective: results of phase 1/2 open-label, dose escalation study. Blood. 2012;119(20):4608-4613. 92. Fermand JP, Ravaud P, Chevret S, et al. Highdose therapy and autologous peripheral blood stem cell transplantation in multiple myeloma: up-front or rescue treatment? Results of a multicenter sequential randomized clinical trial. Blood. 1998;92(9): 31313136. 93. Auner HW, Szydlo R, Hoek J, et al. Trends in autologous hematopoietic cell transplantation for multiple myeloma in Europe: increased use and improved outcomes in elderly patients in recent years. Bone Marrow Transplant. 2015;50(2):209-215. 94. Cook G, Williams C, Brown JM, et al. Highdose chemotherapy plus autologous stemcell transplantation as consolidation therapy in patients with relapsed multiple myeloma after previous autologous stem-cell trans-

405


P. Sonneveld et al.

95.

96.

97.

98.

406

plantation (NCRI Myeloma X Relapse [Intensive trial]): a randomised, open-label, phase 3 trial. Lancet Oncol. 2014;15(8):874885. Lokhorst H, Einsele H, Vesole D, et al. International Myeloma Working Group consensus statement regarding the current status of allogeneic stem-cell transplantation for multiple myeloma. J Clin Oncol. 2010;28(29):4521-4530. Gahrton G, Iacobelli S, Bjรถrkstrand B, et al. Autologous/reduced-intensity allogeneic stem cell transplantation vs autologous transplantation in multiple myeloma: longterm results of the EBMT-NMAM2000 study. Blood. 2013;121(25):5055-5063. Auner HW, Szydlo R, van Biezen A, et al. Reduced intensity-conditioned allogeneic stem cell transplantation for multiple myeloma relapsing or progressing after autologous transplantation: a study by the European Group for Blood and Marrow Transplantation. Bone Marrow Transplant. 2013;48(11):1395-1400. Ocio EM, Richardson PG, Rajkumar SV, et al. New drugs and novel mechanisms of action in multiple myeloma in 2013: a report

from the International Myeloma Working Group (IMWG). Leukemia. 2014;28(3):525542. 99. Dimopoulos M, Siegel DS, Lonial S, et al. Vorinostat or placebo in combination with bortezomib in patients with multiple myeloma (VANTAGE 088): a multicentre, randomised, double-blind study. Lancet Oncol. 2013;14(11):1129-1140. 100. San-Miguel JF, Hungria VT, Yoon SS, et al. Panobinostat plus bortezomib and dexamethasone versus placebo plus bortezomib and dexamethasone in patients with relapsed or relapsed and refractory multiple myeloma: a multicentre, randomised, double-blind phase 3 trial. Lancet Oncol. 2014;15(11):1195-1206. 101. Lonial S, Vij R, Harousseau JL, et al. Elotuzumab in combination with lenalidomide and low-dose dexamethasone in relapsed or refractory multiple myeloma. J Clin Oncol. 2012;30(16):1953-1959. 102. Richardson PG, Jagannath S, Moreau P, et al. A Phase 2 Study of Elotuzumab (Elo) in Combination with Lenalidomide and LowDose Dexamethasone (Ld) in Patients (pts) with Relapsed/Refractory Multiple

Myeloma (R/R MM): Updated Results. Blood. 2012;120(21):202 abstr. 103. Lonial S, Dimopoulos M, Palumbo A, et al. Elotuzumab Therapy for Relapsed or Refractory Multiple Myeloma. N Engl J Med. 2015;373(7):621-631. 104. de Weers M, Tai YT, van der Veer MS, et al. Daratumumab, a novel therapeutic human CD38 monoclonal antibody, induces killing of multiple myeloma and other hematological tumors. J Immunol. 2011;186(3):18401848. 105. Nijhof IS, Groen RW, Lokhorst HM, et al. Upregulation of CD38 expression on multiple myeloma cells by all-trans retinoic acid improves the efficacy of daratumumab. Leukemia. 2015;29(10):2039-2049. 106. Nijhof IS, Lammerts van Bueren JJ, et al. Daratumumab-mediated lysis of primary multiple myeloma cells is enhanced in combination with the human anti-KIR antibody IPH2102 and lenalidomide. Haematologica. 2015;100(2):263-268. 107. Richardson PG, Delforge M, Beksac M, et al. Management of treatment-emergent peripheral neuropathy in multiple myeloma. Leukemia. 2012;26(4):595-608.

haematologica | 2016; 101(4)


REVIEW ARTICLE

New and emerging prognostic and predictive genetic biomarkers in B-cell precursor acute lymphoblastic leukemia

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Anthony V. Moorman

Leukaemia Research Cytogenetics Group, Northern Institute for Cancer Research, Newcastle University, Newcastle-upon-Tyne, UK

Haematologica 2016 Volume 101(4):407-416

ABSTRACT

A

cute lymphoblastic leukemia (ALL) is a heterogeneous disease at the genetic level. Chromosomal abnormalities are used as diagnostic, prognostic and predictive biomarkers to provide subtype, outcome and drug response information. t(12;21)/ETV6-RUNX1 and high hyperdiploidy are good-risk prognostic biomarkers whereas KMT2A (MLL) translocations, t(17;19)/TCF3-HLF, haploidy or low hypodiploidy are highrisk biomarkers. t(9;22)/BCR-ABL1 patients require targeted treatment (imatinib/dasatinib), whereas iAMP21 patients achieve better outcomes when treated intensively. High-risk genetic biomarkers are four times more prevalent in adults compared to children. The application of genomic technologies to cases without an established abnormality (B-other) reveals copy number alterations which can be used either individually or in combination as prognostic biomarkers. Transcriptome sequencing studies have identified a network of fusion genes involving kinase genes - ABL1, ABL2, PDGFRB, CSF1R, CRLF2, JAK2 and EPOR. In vitro and in vivo studies along with emerging clinical observations indicate that patients with a kinaseactivating aberration may respond to treatment with small molecular inhibitors like imatinib/dasatinib and ruxolitinib. Further work is required to determine the true frequency of these abnormalities across the age spectrum and the optimal way to incorporate such inhibitors into protocols. In conclusion, genetic biomarkers are playing an increasingly important role in the management of patients with ALL.

Clinical epidemiology of acute lymphoblastic leukemia Acute lymphoblastic leukemia (ALL) is a heterogeneous disease at the demographic, clinical and genetic levels. Although ALL can occur at any age, it is more prevalent among children, particularly those aged 3-6 years old. More than 50% of the 600 patients diagnosed annually in England and Wales will be aged 0-14 years old, and fewer than 20% will be over 60 years old (A Moorman, unpublished observations, 2016). Males are diagnosed with ALL more frequently than females, resulting in a sex ratio of 1.4:1, respectively. Survival rates from ALL have improved dramatically over the past four decades but vary significantly with age. Children treated on modern protocols have survival rates exceeding 90%.1,2 In contrast, survival from adult ALL is approximately 40% for those patients aged between 25 and 59 years old and is significantly lower (<20%) for older adults.3-5 Improvements in outcome have resulted from optimizing the use of a relatively small number of anti-leukemic drugs, better supportive care, and the introduction of treatment stratification based on risk factors. Traditional risk factors include sex, age, white cell count (WCC) and immunophenotype (B-cell/T-cell) with males, older patients and those with higher white cell count or T-cell ALL having a greater risk of relapse and death. More recently, treatment response (reduction in leukemic burden) has been used to direct treatment.1,2,6 Measuring treatment response or minimal residual disease is performed either by tracking the leukemic clone in serial samples by PCR or flow cytometry looking for specific Ig/TCR rearrangements or immunophenohaematologica | 2016; 101(4)

Correspondence: anthony.moorman@newcastle.ac.uk Received: January 22, 2016 Accepted: January 25, 2016. doi:10.3324/haematol.2015.141101 This manuscript is based on an educational manuscript from the annual congress of EHA (June 2015). Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/4/407

Š2016 Ferrata Storti Foundation Material published in Haematologica is covered by copyright. All rights reserved to Ferrata Storti Foundation. Copies of articles are allowed for personal or internal use. A permission in writing by the publisher is required for any other use.

407


A.V. Moorman et al.

typic profiles. Minimal residual disease (MRD) is a useful tool for treatment stratification and has been adopted by many clinical study groups in order to risk-stratify patients. One of the major advantages of MRD is that it is applicable to the majority of patients (>90%). However, as MRD measures treatment response, it is protocol dependent, and MRD time points and thresholds need to be carefully assessed for each type of protocol. There is ongoing debate about how to best integrate genetic risk factors and MRD into a cohesive clinical strategy for improving outcome in ALL and different models are emerging (see below). However, one important advantage of genetic risk factors is that they can also act as useful therapeutic targets; for example, the recently identified network of gene fusions which are sensitive to tyrosine kinase inhibitors.7

Genetic landscape of acute lymphoblastic leukemia Like all cancers, ALL is characterized by the sequential acquisition of genetic aberrations which drive the initiation and maintenance of the leukemic clone.8,9 Broadly

speaking, genetic abnormalities can be considered as primary or secondary events. Primary abnormalities are responsible for the initiation of a pre-leukemic clone which, upon the acquisition of additional secondary or cooperating genetic changes, converts into overt ALL. Elegant studies have demonstrated that the pre-leukemic clone can lie dormant for several years prior to activation.10 Primary abnormalities in ALL are often chromosomal translocations, resulting in chimeric fusion genes, or gross aneuploidy (gain or loss of multiple whole chromosomes); whereas secondary abnormalities are usually copy number alterations (CNA) (frequently micro-deletions) and point mutations. Primary abnormalities are, by definition, present in all the cells comprising the leukemic clone and define the key features of the leukemia. In contrast, secondary abnormalities are present only in a subset of the leukemic cells and give rise to a complex branching subclonal architecture.11 In ALL, there is a strong correlation between the primary chromosomal abnormality and the spectrum of secondary or co-operating mutations observed in that subtype (Figure 1).12 The vast majority of aberrations act either as primary or secondary abnormali-

Figure 1. Overview of key co-operating mutations in relation to distinct genetic subtypes of B-cell precursor acute lympholastic leukemia.

408

haematologica | 2016; 101(4)


Genetic biomarkers in B-cell precursor ALL

ties; however, a few have been reported as both types in different contexts. The comprehensive genetic testing of patients suspected of having ALL can confirm the diagnosis of ALL and identify important prognostic and predictive biomarkers which can be used to tailor therapy. Primary genetic abnormalities are more reliable prognostic markers than secondary aberrations, probably due to the fact that they define the key features of the clone and are ubiquitous. Therefore, the focus of most screening algorithms in ALL is on the reliable detection of key primary chromosomal abnormalities used to stratify patients into different risk groups. This can be achieved using a combination of cytogenetics, FISH (fluorescence in situ hybridization) and RT-PCR (reverse transcription polymerase chain reaction) but more modern techniques such as multiplex ligation-dependent probe amplification (MLPA), DNA copy number arrays and targeted gene resequencing are increasingly being used to screen for new and emerging genetic biomarkers. In this article, I will describe and discuss established, new, and emerging diagnostic, prognostic and predictive genetic biomarkers in the major subtype of B-cell precursor ALL.

Good-risk prognostic genetic biomarkers t(12;21)/ETV6-RUNX1 and high hyperdiploidy (51-65 chromosomes) are well recognized diagnostic and prog-

nostic biomarkers in both pediatric and adult ALL. ETV6RUNX1 results from the chromosomal translocation, t(12;21)(p13;q22) which is cytogenetically cryptic and therefore FISH or RT-PCR is required for its accurate detection. High hyperdiploidy (51-65 chromosomes) is readily detectable by cytogenetics but also by the application of locus specific and centromeric FISH probes as the pattern of chromosomal gain is non-random with eight chromosomes accounting for more than 75% gains; namely chromosomes X, 4, 6, 10, 14, 17, 18 and 21.13-17 These two biomarkers account for approximately 60% of pediatric and adolescent ALL but less than 15% of adult ALL (Figure 2) with ETV6-RUNX1 being virtually non-existent among adults over 30 years of age.18 Patients with either of these abnormalities have a very good outcome compared to their age-matched counterparts, and overall survival (OS) rates at five years is over 90% in pediatric ALL and 55% in adult ALL (Figure 3A).19,20 Although many studies have examined the prognostic relevance of secondary abnormalities (including IKZF1 deletion, ETV6 deletions, RAS pathway mutations) within these two subgroups, no reliable additional biomarkers have emerged.21-25 In addition, within high hyperdiploidy, many studies have assessed specific trisomies, modal chromosome number and structural abnormalities as additional prognostic markers. Although specific trisomies (+4, +10, +17 and +18) have emerged as clinically relevant biomarkers within particular treatment protocols, they have not proved to

Figure 2. Frequency of primary chromosomal abnormalities in children and adults with B-cell precursor acute lymphoblastic leukemia.

haematologica | 2016; 101(4)

409


A.V. Moorman et al.

be universally applicable.13-17 Given the excellent outcome of patients with ETV6-RUNX1 and high hyper-diploidy in pediatric ALL, it is difficult to envisage further clinically actionable biomarkers emerging from within this risk group.

Five chromosomal abnormalities [KMT2A (MLL) translocations, t(9;22)/BCR-ABL1, t(17;19)/TCF3-HLF, near haploidy and low hypodiploidy] are well recognized prognostic biomarkers of high-risk disease at all ages.8 The KMT2A (MLL) gene located at 11q23 undergoes rearrangements, usually translocations, with a plethora of partner genes; with AFF1 (AF4), MLLT1 (ENL), MLLT4 (AF6), MLLT3 (AF9) and MLLT10 (AF10) accounting for more than 85% ALL cases.26 Near haploidy and low hypodiploidy are defined by massive chromosomal loss resulting in a modal number of less than 30 chromosomes and 30-39 chromosomes, respectively.27 Recent studies have identified the RAS gene and TP53 mutations, respectively, as key additional drivers of these distinct ploidy subgroups.28 Both subgroups display a propensity to undergo chromosome doubling which can create a diagnostic dilemma if only the doubled-up subclone is dividing and hence masquerades as high hyperdiploidy.27 However, the pattern of chromosomal loss/gain is distinctive, and these two subgroups are usually distinguishable from one another. All MLL translocations, as well as BCR-ABL1 and TCF3-HLF, are readily detectable by cytogenetics, FISH and RT-PCR. Given the promiscuity of the MLL gene, FISH with a dual-color break-apart directed to the 11q23

locus is the most convenient method of detection. The frequency of these five high-risk genetic aberrations is four times higher in adults compared to children and adolescents (Figure 2), explaining, in part, the strong correlation between age and outcome (Figure 3A and B). KMT2A (MLL) translocations are also highly prevalent in infant ALL (<1 year) where they account for approximately 80% of patients.29 Patients with one of these five abnormalities are classified as high risk in UK protocols and treated with the most intensive regimens.19 If treated as standard risk, patients with one of these aberrations have an approximately 3-fold increased risk of relapse and/or death compared to intermediate-risk patients.19 BCR-ABL1 is a predictive biomarker for targeted therapy with a tyrosine kinase inhibitor, such as imatinib or dasatinib.30 Tyrosine kinase inhibitors directly inhibit the leukemogenic effect of the BCR-ABL1 oncoprotein and, in combination with standard chemotherapy, produce significantly superior outcomes in patients of all ages.31,32 Recent studies of MLLrearranged ALL have highlighted the importance of epigenetic dysregulation in this sub-group and, in particular, the requirement for the histone methyltransferase, DOT1L.33 These studies raise the possibility of developing a targeted therapy for these high-risk patients using DOT1L inhibitors such as EPZ004777.34 In adult ALL, patients without an established chromosomal translocation or ploidy sugroups are classified as having a complex karyotype if their karyotype harbors five or more chromosomal abnormalities.35 This definition identified approximately 5% patients which have a higher risk of relapse or death in some, but not all, treatment protocols (Figure 3B).35-37 Given the subjective nature of chromosomal analysis and the vagaries of cytogenetic nomencla-

A

B

High-risk prognostic genetic biomarkers

Figure 3. Outcome of patients with acute lymphoblastic leukemia (ALL) by genetic risk group. (A) Event-free survival of children and adolescents with B-cell precursor ALL treated on ALL2003 and stratified by cytogenetics and copy number alterations profile. (B) Survival of adults treated on UKALLXII stratified by genetic risk group.

410

haematologica | 2016; 101(4)


Genetic biomarkers in B-cell precursor ALL

ture, complex karyotype is not an ideal biomarker for the reliable identification of high-risk patients. The intensive research described below aimed at unraveling the genetics of B-other ALL. This will, in time, provide a more reliable biomarker for this subset of patients.

t(1;19)(q23;p13)/TCF3-PBX1 Approximately 3% of children/adolescents and 6% of adults harbor the translocation t(1;19) which in more than 95% cases results in TCF3-PBX1 fusion.38 It is an intriguing biomarker in ALL: as a diagnostic biomarker it correlates tightly with a pre-B immunophenotype with the leukemic cells expressing cytoplasmic m. It is readily and reliably identified by cytogenetics, FISH and RT-PCR, making it an ideal prognostic biomarker in practical terms. However, outcome studies have produced variable results and there is a stark contrast between pediatric and adult ALL regarding how these patients are viewed in terms of risk. In pediatric cohorts, early studies reported TCF3-PBX1 as a biomarker of poor prognosis, but most recent studies from protocols delivering more intensive chemotherapy have reported considerably improved outcome (OS >80%).19,39-41 Interestingly, some studies have reported an association with central nervous system (CNS) relapse and a poor outcome after first relapse, indicating substantial clinical heterogeneity, which may point to the presence of additional prognostic biomarkers in this subgroup.40 There is a similar story in adult ALL, with the larger and more recent studies showing TCF3-PBX1 to be associated with intermediate risk.35,42,43 However, some clinical study groups classify adult patients with TCF3-PBX1 as high risk and treat these patients more aggressively.44

Intrachromosomal amplification of chromosome 21 (iAMP21) In the past few years, iAMP21 has become an important prognostic and predictive biomarker in childhood ALL. iAMP21 is a grossly abnormal chromosome generated via breakage-fusion-bridge cycles and chromothripsis.45 The result of these rearrangements is the amplification and loss of multiple regions along the length of chromosome 21. The consistent feature of all iAMP21 cases is the amplification of the RUNX1 locus located at 21q22.12, which provides the basis for a convenient and reliable detection assay: FISH using locus-specific probes.46 The internationally accepted definition of iAMP21 is three or more extra copies of the RUNX1 gene on a single abnormal chromosome 21, equating to more than 5 signals per cell.46 Patients with iAMP21 are older than other children with ALL (median age 9 years) but a lower median WCC.47,48 Studies by the UK and the Childrenâ&#x20AC;&#x2122;s Oncology Group (COG) have reported that iAMP21 patients treated as standard risk have a very high rate of relapse (>80%) but that this is significantly reduced (<20%) when the pa-tients are treated intensively, notably on the most intensive arm of UKALL2003 (regimen C).49,50 Thus, iAMP21 can be considered both a prognostic and predictive biomarker in pediatric ALL. However, the Associazione Italiana di Ematologia ed Oncologia Pediatrica (AEIOP) and BerlinFrankfurt-Munster (BFM) study groups have reported that MRD can also be used to identify iAMP21 patients at risk of relapse.51,52 As iAMP21 is extremely rare among adult haematologica | 2016; 101(4)

patients (â&#x2030;Ľ25 years), its prognostic effect in this age group is unclear.

Translocation involving the IGH locus IGH translocations are well recognized in lymphoid malignancies where the juxtaposition of an oncogene to the IGH enhancer drives its overexpression.53 IGH translocations are frequent in lymphomas and mature leukemias. However, recent studies have revealed an extensive network of IGH translocations specific to BCP-ALL, which drive the expression of a variety of oncogenes.54 The most common IGH translocation involves CRLF2, accounting for approximately 25% of cases. Other recurrent translocation partners include five members of the CEBP gene family and ID4, accounting for approximately 10% and 7% of cases, respectively. Although numerous partner genes have yet to be identified, there does not appear to be any functional link between the partner genes of IGH translocations. Given the wide spectrum of these partners and the finding that several, including IGH-CRLF2, are cytogenetically cryptic, FISH using a break-apart probe directed to the IGH locus provides a reliable detection method. The most notable clinical feature of patients with IGH translocations is their age profile. Their frequency is low among children under ten years of age (<3%) but considerably higher (10%) among adolescents and young adults (15-24 years).54 Patients with IGH translocations have been shown to have an inferior outcome compared to other patients in both the adolescent and young adult setting.54

Dissecting B-other acute lymphoblastic leukemia Approximately 70% of pediatric and 60% of adult patients with ALL harbor a genetic abnormality regarded as an established diagnostic and/or prognostic biomarker (Figure 2). Patients without one of these abnormalities are collectively referred to as B-other ALL. Unraveling the genetic landscape of B-other ALL has been the main focus of research efforts over the past ten years. A wide variety of techniques, ranging from cytogenetics to whole genome sequencing, have been applied in order to define and assess biomarkers in this subgroup. Two main approaches have shaped the current research strategy. Firstly, a plethora of micro-deletions affecting genes in key pathways, including lymphoid differentiation, cell-cycle differentiation and proliferation were discovered using SNP arrays in both childhood and adult ALL.55-58 Numerous studies have assessed the role of individual lesions and copy number alteration (CNA) profiles as prognostic biomarkers. Almost all of these CNA are secondary aberrations which are sub-clonal and can be acquired, lost or enriched for between diagnosis and relapse.59-61 Secondly, gene expression profiling has been used to define cytogenetic subgroups and at the same time identify novel subgroups of patients.62,63

IKZF1 and ERG deletions IKZF1 deletions occur in 15% of pediatric and 30% of adult ALL, but are more prevalent among patients with 411


A.V. Moorman et al.

BCR-ABL1 (>60%).14,64-66 In addition, they are associated with other high-risk features such as older age, high WCC, persistent MRD, and Down syndrome.12,67 Initial reports suggested that patients harboring an IKZF1 deletion had a significantly inferior outcome, implying that it was a reliable prognostic marker.68 However, more recent studies based on larger and more representative cohorts have suggested that its effect is pleiotropic.69-73 In particular, a COG study found that IKZF1 deletions were prognostic among National Cancer Institute (NCI) high-risk (age >10 years or WCC >50x109/L) patients but not NCI standard-risk patients.74 In addition, studies have found that the presence of an IKZF1 deletion does not abrogate the prognosis associated with other good risk genetic abnormalities such as ETV6-RUNX1 and ERG deletions (see below).21,72,73,75 These results correlate with those from studies that have examined the interaction of IKZF1 deletions and MRD. Several studies have now reported that IKZF1 deletions are not prognostic among patients who clear their disease rapidly. Instead the prognostic effect is restricted to, or at least strongest in, patients with higher levels of disease burden after initial chemotherapy.69.70,76 Furthermore, IKZF1 deletions were not associated with a greater risk of second relapse or death after a first marrow relapse.77 However, within the context of BCR-ABL1 ALL, it appears that IKZF1 deletions are strongly prognostic even when these patients are treated with imatinib-containing regimens.66 Therefore, it is likely that IKZF1 deletions, which are secondary abnormalities, are also a â&#x20AC;&#x153;secondaryâ&#x20AC;? marker of poor outcome rather than being a key independent prognostic biomarker. Three studies have now identified a distinct subgroup of pediatric B-other ALL patients characterized by a monoallelic deletion of the ERG gene.63,72,73 The frequency is 10%15% of B-other ALL which equates to 3%-5% of BCPALL overall. Interestingly, these patients have an excellent outcome of over 90% at five years despite having a very high incidence of IKZF1 deletions (approx. 40%).63,72,73 Even though the presence of an ERG deletion appears to define a distinct subgroup of B-other ALL, there is evidence that it is a sub-clonal event which can be lost or gained between diagnosis and relapse. Its role as a robust prognostic marker requires urgent further investigation, particularly within adult ALL.

unsurprisingly the results of these studies have been conflicting with some studies concluding that it was a prognostic marker of poor outcome82,83 while others concluded that it was not relevant in the context of other risk factors.52,67,75,84 Interestingly, three studies found discordant results depending on whether CRLF2 involvement was determined by genetics or expression,74,82,85 thereby emphasizing the requirement for standardized definitions. In addition, P2RY8-CRLF2 is a secondary abnormality and often present only within low level subclones which do not drive relapse.86 Although CRLF2 does not appear to be a robust prognostic marker, it is an attractive therapeutic target particularly within the context of Down syndrome patients who are prone to the toxic side-effects of chemotherapy. Therefore, inhibition of the JAK and PI3K pathways represent potential therapeutic strategies in these cases.7,87

CRLF2 deregulation

BCR-ABL1-like

The interstitial deletion in the PAR1 region of chromosome X and Y gives rise to dysregulation of CRLF2 via juxtaposition of this gene to the P2RY8 promoter.78,79 Overexpression of CRLF2 can also arise from an IGH translocation and, more rarely, an activating mutation.80 In addition, high CRLF2 expression can occur in patients who lack a clear genetic alteration at this locus. CRLF2 rearrangements have been associated with activation of the JAK-STAT, ERK and mTOR/PI3K pathways and it is noteworthy that approximately 50% of cases also harbor a JAK2 mutation.7,78,79,81 The overall frequency of CRLF2 rearrangements in BCP-ALL is 5% but is higher in B-other (30%) and patients with Down syndrome (>50%).67,78 Studies investigating the prognostic relevance of CRLF2 have varied in the method used to identify cases. Some studies have focused on genetic rearrangements whilst others have measured CRLF2 mRNA expression. Perhaps

Four independent studies reported a subgroup of Bother ALL patients with a gene expression profile similar to BCR-ABL1 positive ALL.68,89-91 Although these patients lacked the fusion gene they shared the same poor outcome. The subgroups were termed BCR-ABL1-like or Phlike and both accounted for approximately 50% of B-other ALL cases. There were significant differences in the genetic make-up of these subgroups; especially with respect to the prevalence of IKZF1 deletions, CRLF2 rearrangements and JAK2 mutations.68,89 Crucially, the gene expression signatures were not transferable, and when applied to the same cohort of patients, the signatures classified different patients as BCR-ABL1-like/Ph-like.92 Therefore, gene expression signatures do not provide an ideal prognostic biomarker for use between different clinical trials. However, the common features of these subgroups are poor outcome and enrichment of CRLF2 rearrangements

412

Copy number alteration profiling The major limitation to assessing the prognostic relevance of individual CNAs is that it does not take into account the fact that many cases will harbor more than one deletion while others may have none. An alternative approach has been to integrate a CNA profile into the existing established cytogenetic risk-group classification. Here a CNA profile based on the presence or absence of the eight most frequently deleted genes segregates patients with intermediate-risk cytogenetics (mostly Bother) into two new genetic risk groups (Figure 4).88 The prognosis of patients with good- or high-risk cytogenetics was unaffected by their CNA profile. However, intermediate cytogenetic risk patients, separated into two subgroups (good risk vs. intermediate-/high-risk CNA profile) with differential OS rates (98% vs. 87%).88 Thus, this approach has identified a group of B-other ALL patients with a good-risk CNA profile and a very low risk of relapse who potentially could be considered for treatment deintensification. The validity of such an approach is supported by observations that the prognostic effect of IKZF1 deletions depends on the presence/absence of other deletions (e.g. ERG and CDKN2A/B deletions) and MRD levels.63,72,73,76

haematologica | 2016; 101(4)


Genetic biomarkers in B-cell precursor ALL

Figure 4. Proposal for a new integrated risk classification of childhood B-cell precursor acute lymphoblastic leukemia based on primary chromosomal abnormalities and copy number alteration (CNA) profile (yellow indicates deleted; blue indicates not deleted).

and IKZF1 deletion, albeit to varying extents. Although it is unlikely that any gene signature expression will become a robust prognostic biomarker within B-other ALL, there is clearly a subset of patients with B-other ALL who have a poor outcome, and it is now emerging that members of this group can be identified from their distinctive genomic abnormalities.

Kinase activating gene fusions In 2012, Roberts and colleagues performed RNA sequencing of a small cohort of BCR-ABL1-like cases and discovered chimeric fusions involving PDGFRB, ABL1 and JAK2.87 Since this initial study, a complex network of kinase-activating aberrations has been revealed, with many occurring in single patients.7,93,94 The common feature of these chimeric genes is the fusion of the 3â&#x20AC;&#x2122; end of a kinase gene with the 5â&#x20AC;&#x2122; portion of so-called activating gene, of which over 30 have now been reported. haematologica | 2016; 101(4)

Theoretically, all these kinase-activating lesions can be targeted with appropriate small molecule inhibitors, and hence become useful predictive biomarkers. For example, in vitro and in vivo studies have demonstrated that ABL1, ABL2, PDGFRB and CSF1R fusions are sensitive to imatinib and dasatinib, and that CRLF2, JAK2 and EPOR are sensitive to JAK inhibitors (e.g. ruxolitinib).7 Moreover, a small number of children harboring ABL1, ABL2, PDGFRB or CSF1R fusions with refractory disease achieved a complete remission following treatment with imatinib or dasatinib.7,95-98 However, it should be noted that these patients were highly selected and that follow up was extremely limited. Nonetheless, these laboratory and clinical observations provide encouraging evidence that targeted therapies could be offered routinely to patients with one of these predictive biomarkers in the near future. Their prevalence has yet to be firmly established in large representative cohorts. Initial screening of patients in the UK has indicated that ABL1, ABL2, PDGFRB or CSF1R fusions collectively occur in 1%-2% of B-cell precursor ALL inde413


A.V. Moorman et al.

pendent of age; whereas CRLF2, JAK2 and EPOR fusions together occur in 3%-5% of childhood ALL rising to approximately 15% in adolescent and young adult ALL (A Moorman, unpublished observations, 2016). Current published studies have suggested that IGH-CRLF2, PR2Y8CRLF2 and EBF1-PDGFRB are the most prevalent of these gene fusions. The optimal detection method is challenging given the number of genes involved and the complex nature of some of the chromosomal rearrangements which give rise to these fusion genes. FISH using probes to target the kinase gene provides a simple and efficient strategy for detection of many of the fusions, especially ABL1, PDGFRB, CSF1R, JAK2 and CRLF2, and can readily be incorporated into current screening algorithms. However, assays based on next generation sequencing technology are on the horizon and are likely to provide a more comprehensive approach.

Conclusion and future perspectives The extensive genetic heterogeneity found within ALL provides a wealth of potential genetic biomarkers that

References 1. Vora A, Goulden N, Mitchell C, et al. Augmented post-remission therapy for a minimal residual disease-defined high-risk subgroup of children and young people with clinical standard-risk and intermediate-risk acute lymphoblastic leukaemia (UKALL 2003): a randomised controlled trial. Lancet Oncol. 2014;15(8):809-818. 2. Vora A, Goulden N, Wade R, et al. Treatment reduction for children and young adults with low-risk acute lymphoblastic leukaemia defined by minimal residual disease (UKALL 2003): a randomised controlled trial. Lancet Oncol. 2013;14(3):199-209. 3. Goldstone AH, Richards SM, Lazarus HM, et al. In adults with standard-risk acute lymphoblastic leukemia, the greatest benefit is achieved from a matched sibling allogeneic transplantation in first complete remission, and an autologous transplantation is less effective than conventional consolidation/maintenance chemotherapy in all patients: final results of the International ALL Trial (MRC UKALL XII/ECOG E2993). Blood. 2008;111(4):18271833. 4. Moorman AV, Chilton L, Wilkinson J, Ensor HM, Bown N, Proctor SJ. A populationbased cytogenetic study of adults with acute lymphoblastic leukemia. Blood. 2010;115(2):206-214. 5. Sive JI, Buck G, Fielding A, et al. Outcomes in older adults with acute lymphoblastic leukaemia (ALL): results from the international MRC UKALL XII/ECOG2993 trial. Br J Haematol. 2012;157(4):463-471. 6. Conter V, Bartram CR, Valsecchi MG, et al. Molecular response to treatment redefines all prognostic factors in children and adolescents with B-cell precursor acute lymphoblastic leukemia: results in 3184 patients of the AIEOP-BFM ALL 2000 study. Blood. 2010;115(16):3206-3214. 7. Roberts KG, Li Y, Payne-Turner D, et al.

414

8.

9. 10. 11.

12.

13.

14.

15.

16.

could be used to assist patient management. Prognostic biomarkers such as ETV6-RUNX1 and high hyperdiploidy can be used to define a cohort of patients with a low risk of relapse on standard therapy whereas patients with high-risk cytogenetics require more intensive or targeted therapy. In the past ten years, genomic analysis has revolutionized the way researchers and clinicians think about the biology of ALL, and new therapeutic strategies and options are beginning to emerge. Additional research is required to assess the clinical utility of some of these discoveries, as a number of questions remain unanswered. For example: 1) what is the optimal way to use copy number alterations as prognostic biomarkers in Bother ALL and within the context of MRD-driven protocols? 2) Which kinase-activating abnormalities are predictive biomarkers for treatment with an appropriate inhibitor? 3) What is the role of these new genetic biomarkers in directing therapy after first relapse? In addition to addressing these translational questions, the largescale application of whole genome and exome sequencing will undoubtedly identify new genetic biomarkers which may add to or replace our current repertoire of prognostic and predictive biomarkers.

Targetable kinase-activating lesions in Phlike acute lymphoblastic leukemia. N Engl J Med. 2014;371(11):1005-1015. Moorman AV. The clinical relevance of chromosomal and genomic abnormalities in Bcell precursor acute lymphoblastic leukaemia. Blood Rev. 2012;26(3):123-135. Inaba H, Greaves M, Mullighan CG. Acute lymphoblastic leukaemia. Lancet. 2013;381 (9881):1943-1955. Greaves M. Infection, immune responses and the aetiology of childhood leukaemia. Nat Rev Cancer. 2006;6(3):193-203. 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. Schwab CJ, Chilton L, Morrison H, et al. Genes commonly deleted in childhood Bcell precursor acute lymphoblastic leukemia: association with cytogenetics and clinical features. Haematologica. 2013;98(7):10811088. Moorman AV, Richards SM, Martineau M, et al. Outcome heterogeneity in childhood high-hyperdiploid acute lymphoblastic leukemia. Blood. 2003;102(8):2756-2762. Heerema NA, Raimondi SC, Anderson JR, et al. Specific extra chromosomes occur in a modal number dependent pattern in pediatric acute lymphoblastic leukemia. Genes Chromosomes and Cancer. 2007;46(7):684693. Sutcliffe MJ, Shuster JJ, Sather HN, et al. High concordance from independent studies by the Children's Cancer Group (CCG) and Pediatric Oncology Group (POG) associating favorable prognosis with combined trisomies 4, 10, and 17 in children with NCI Standard-Risk B-precursor Acute Lymphoblastic Leukemia: a Children's Oncology Group (COG) initiative. Leukemia. 2005;19(5):734-740. Dastugue N, Suciu S, Plat G, et al. Hyperdiploidy with 58-66 chromosomes in childhood B-acute lymphoblastic leukemia is highly curable: 58951 CLG-EORTC

results. Blood. 2013;121(13):2415-2423. 17. Paulsson K, Forestier E, Andersen MK, et al. High modal number and triple trisomies are highly correlated favorable factors in childhood B-cell precursor high hyperdiploid acute lymphoblastic leukemia treated according to the NOPHO ALL 1992/2000 protocols. Haematologica. 2013;98(9):14241432. 18. Jabbar Al Obaidi MS, Martineau M, Bennett CF, et al. ETV6/AML1 fusion by FISH in adult acute lymphoblastic leukemia. Leukemia. 2002;16(4):669-674. 19. Moorman AV, Ensor HM, Richards SM, et al. Prognostic effect of chromosomal abnormalities in childhood B-cell precursor acute lymphoblastic leukaemia: results from the UK Medical Research Council ALL97/99 randomised trial. Lancet Oncol. 2010;11(5):429438. 20. Chilton L, Buck G, Harrison CJ, et al. High hyperdiploidy among adolescents and adults with acute lymphoblastic leukaemia (ALL): cytogenetic features, clinical characteristics and outcome. Leukemia. 2014;28(7):15111518. 21. Enshaei A, Schwab CJ, Konn ZJ, et al. Longterm follow-up of ETV6-RUNX1 ALL reveals that NCI risk, rather than secondary genetic abnormalities, is the key risk factor. Leukemia. 2013;27(11):2256-2259. 22. Bhojwani D, Pei D, Sandlund JT, et al. ETV6RUNX1-positive childhood acute lymphoblastic leukemia: improved outcome with contemporary therapy. Leukemia. 2012;26(2):265-270. 23. Barbany G, Andersen MK, Autio K, et al. Additional aberrations of the ETV6 and RUNX1 genes have no prognostic impact in 229 t(12;21)(p13;q22)-positive B-cell precursor acute lymphoblastic leukaemias treated according to the NOPHO-ALL-2000 protocol. Leuk Res. 2012;36(7):936-938. 24. Paulsson K, Johansson B. High hyperdiploid childhood acute lymphoblastic leukemia. Genes Chromosomes Cancer. 2009;48(8): 637-660.

haematologica | 2016; 101(4)


Genetic biomarkers in B-cell precursor ALL 25. Boer JM, Veer Avd, Rizopoulos D, et al. Prognostic Value of Rare IKZF1 deletions in Childhood B-Cell Precursor Acute Lymphoblastic Leukemia: An International Collaborative Study. Blood. 2014;124(21): 368. 26. Meyer C, Hofmann J, Burmeister T, et al. The MLL recombinome of acute leukemias in 2013. Leukemia. 2013;27(11):2165-2176. 27. Harrison CJ, Moorman AV, Broadfield ZJ, et al. Three distinct subgroups of hypodiploidy in acute lymphoblastic leukaemia. Br J Haematol. 2004;125(5):552-559. 28. Holmfeldt L, Wei L, Diaz-Flores E, et al. The genomic landscape of hypodiploid acute lymphoblastic leukemia. Nat Genet. 2013;45 (3):242-252. 29. Pieters R, Schrappe M, De LP, et al. A treatment protocol for infants younger than 1 year with acute lymphoblastic leukaemia (Interfant-99): an observational study and a multicentre randomised trial. Lancet. 2007;370(9583):240-250. 30. Druker BJ. Translation of the Philadelphia chromosome into therapy for CML. Blood. 2008;112(13):4808-4817. 31. 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. 32. Biondi A, Schrappe M, De Lorenzo P, et al. Imatinib after induction for treatment of children and adolescents with Philadelphiachromosome-positive acute lymphoblastic leukaemia (EsPhALL): a randomised, openlabel, intergroup study. Lancet Oncol. 2012;13(9):936-945. 33. Bernt KM, Zhu N, Sinha AU, et al. MLLrearranged leukemia is dependent on aberrant H3K79 methylation by DOT1L. Cancer Cell. 2011;20(1):66-78. 34. Bernt KM, Armstrong SA. Targeting epigenetic programs in MLL-rearranged leukemias. Hematology Am Soc Hematol Educ Program. 2011;2011:354-360. 35. 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. 36. Motllo C, Ribera JM, Morgades M, et al. Prognostic significance of complex karyotype and monosomal karyotype in adult patients with acute lymphoblastic leukemia treated with risk-adapted protocols. Cancer. 2014;120(24):3958-3964. 37. Pullarkat V, Slovak ML, Kopecky KJ, Forman SJ, Appelbaum FR. Impact of cytogenetics on the outcome of adult acute lymphoblastic leukemia: results of Southwest Oncology Group 9400 study. Blood. 2008;111(5):25632572. 38. Barber KE, Harrison CJ, Broadfield ZJ, et al. Molecular Cytogenetic Characterisation of TCF3 (E2A) / 19p13.3 Rearrangements in BCell Precursor Acute Lymphoblastic Leukaemia. Genes Chromosomes Cancer. 2007;46(5):478-486. 39. Raimondi S, Behm F, Robertson PK, et al. Cytogenetics of pre-B-cell acute lymphoblastic leukemia with emphasis on prognostic implications on the t(1;19). J Clin Oncol. 1990;8:1380-1388. 40. Jeha S, Pei D, Raimondi SC, et al. Increased risk for CNS relapse in pre-B cell leukemia with the t(1;19)/TCF3-PBX1. Leukemia. 2009;23(8):1406-1409.

haematologica | 2016; 101(4)

41. Kager L, Lion T, Attarbaschi A, et al. Incidence and outcome of TCF3-PBX1-positive acute lymphoblastic leukemia in Austrian children. Haematologica. 2007;92(11):1561-1564. 42. GFCH. Cytogenetic Abnormalities in Adult Acute Lymphoblastic Leukemia: Correlations with Hematologic Findings and Outcome. A Collaborative Study of the Groupe Francais de Cytogenetique Hematologique. Blood. 1996;87(8):3135-3142. 43. Garg R, Kantarjian H, Thomas D, et al. Adults with acute lymphoblastic leukemia and translocation (1;19) abnormality have a favorable outcome with hyperfractionated cyclophosphamide, vincristine, doxorubicin, and dexamethasone alternating with methotrexate and high-dose cytarabine chemotherapy. Cancer. 2009;115(10):21472154. 44. Beldjord K, Chevret S, Asnafi V, et al. Oncogenetics and minimal residual disease are independent outcome predictors in adult patients with acute lymphoblastic leukemia. Blood. 2014;123(24):3739-3749. 45. Li Y, Schwab C, Ryan SL, et al. Constitutional and somatic rearrangement of chromosome 21 in acute lymphoblastic leukaemia. Nature. 2014;508(7494):98-102. 46. Harrison CJ, Haas O, Harbott J, et al. Detection of prognostically relevant genetic abnormalities in childhood B-cell precursor acute lymphoblastic leukaemia: recommendations from the Biology and Diagnosis Committee of the International BerlinFrankfurt-Munster study group. Br J Haematol. 2010;151(2):132-142. 47. Moorman AV, Richards SM, Robinson HM, et al. Prognosis of children with acute lymphoblastic leukaemia (ALL) and intrachromosomal amplification of chromosome 21 (iAMP21). Blood. 2007;109:2327-2330. 48. Harrison CJ, Moorman AV, Schwab C, et al. An international study of intrachromosomal amplification of chromosome 21 (iAMP21): cytogenetic characterization and outcome. Leukemia. 2014;28(5):1015-1021. 49. Moorman AV, Robinson H, Schwab C, et al. Risk-directed treatment intensification significantly reduces the risk of relapse among children and adolescents with acute lymphoblastic leukemia and intrachromosomal amplification of chromosome 21: a comparison of the MRC ALL97/99 and UKALL2003 trials. J Clin Oncol. 2013;31(27):3389-3396. 50. Heerema NA, Carroll AJ, Devidas M, et al. Intrachromosomal amplification of chromosome 21 is associated with inferior outcomes in children with acute lymphoblastic leukemia treated in contemporary standardrisk children's oncology group studies: a report from the children's oncology group. J Clin Oncol. 2013;31(27):3397-3402. 51. Attarbaschi A, Panzer-Grumayer R, Mann G, et al. Minimal residual disease-based treatment is adequate for relapse-prone childhood acute lymphoblastic leukemia with an intrachromosomal amplification of chromosome 21: the experience of the ALLBFM 2000 trial. Klin Padiatr. 2014;226(67):338-343. 52. Attarbaschi A, Morak M, Cario G, et al. Treatment outcome of CRLF2-rearranged childhood acute lymphoblastic leukaemia: a comparative analysis of the AIEOP-BFM and UK NCRI-CCLG study groups. Br J Haematol. 2012;158(6):772-777. 53. Dyer MJ, Akasaka T, Capasso M, et al. Immunoglobulin heavy chain locus chromosomal translocations in B-cell precursor acute lymphoblastic leukemia: rare clinical curios or potent genetic drivers? Blood.

2010;115(8):1490-1499. 54. Russell LJ, Enshaei A, Jones L, et al. IGH@ translocations are prevalent in teenagers and young adults with acute lymphoblastic leukemia and are associated with a poor outcome. J Clin Oncol. 2014;32(14):1453-1462. 55. Kawamata N, Ogawa S, Zimmermann M, et al. Molecular allelokaryotyping of pediatric acute lymphoblastic leukemias by high-resolution single nucleotide polymorphism oligonucleotide genomic microarray. Blood. 2008;111(2):776-784. 56. Paulsson K, Cazier JB, Macdougall F, et al. Microdeletions are a general feature of adult and adolescent acute lymphoblastic leukemia: Unexpected similarities with pediatric disease. Proc Natl Acad Sci USA. 2008;105(18):6708-6713. 57. Strefford JC, Worley H, Barber K, et al. Genome complexity in acute lymphoblastic leukemia is revealed by array-based comparative genomic hybridization. Oncogene. 2007;26(29):4306-4318. 58. Mullighan CG, Goorha S, Radtke I, et al. Genome-wide analysis of genetic alterations in acute lymphoblastic leukaemia. Nature. 2007;446(7137):758-764. 59. Mullighan CG, Phillips LA, Su X, et al. Genomic analysis of the clonal origins of relapsed acute lymphoblastic leukemia. Science. 2008;322(5906):1377-1380. 60. Hogan LE, Meyer JA, Yang J, et al. Integrated genomic analysis of relapsed childhood acute lymphoblastic leukemia reveals therapeutic strategies. Blood. 2011;118(19):52185226. 61. Yang JJ, Bhojwani D, Yang W, et al. Genome-wide copy number profiling reveals molecular evolution from diagnosis to relapse in childhood acute lymphoblastic leukemia. Blood. 2008;112(10):4178-4183. 62. Yeoh EJ, Ross ME, Shurtleff SA, et al. Classification, subtype discovery, and prediction of outcome in pediatric acute lymphoblastic leukemia by gene expression profiling. Cancer Cell. 2002;1(2):133-143. 63. Harvey RC, Mullighan CG, Wang X, et al. Identification of novel cluster groups in pediatric high-risk B-precursor acute lymphoblastic leukemia with gene expression profiling: correlation with genome-wide DNA copy number alterations, clinical characteristics, and outcome. Blood. 2010;116 (23):4874-4884. 64. Mullighan CG, Miller CB, Radtke I, et al. BCR-ABL1 lymphoblastic leukaemia is characterized by the deletion of Ikaros. Nature. 2008;453(7191):110-114. 65. Martinelli G, Iacobucci I, Storlazzi CT, et al. IKZF1 (Ikaros) deletions in BCR-ABL1-positive acute lymphoblastic leukemia are associated with short disease-free survival and high rate of cumulative incidence of relapse: a GIMEMA AL WP report. J Clin Oncol. 2009;27(31):5202-5207. 66. van der Veer A, Zaliova M, Mottadelli F, et al. IKZF1 status as a prognostic feature in BCR-ABL1-positive childhood ALL. Blood. 2014;123(11):1691-1698. 67. Buitenkamp TD, Pieters R, Gallimore NE, et al. Outcome in children with Down's syndrome and acute lymphoblastic leukemia: role of IKZF1 deletions and CRLF2 aberrations. Leukemia. 2012;26(10):2204-2211. 68. Mullighan CG, Su X, Zhang J, et al. Deletion of IKZF1 and Prognosis in Acute Lymphoblastic Leukemia. N Engl J Med. 2009;360(5):470-480. 69. Palmi C, Valsecchi MG, Longinotti G, et al. What is the relevance of Ikaros gene deletions as prognostic marker in pediatric Philadelphia negative B-cell precursor acute

415


A.V. Moorman et al.

70.

71.

72.

73.

74.

75.

76.

77.

78.

416

lymphoblastic leukemia? Haematologica. 2013;98(8):1226-1231. Waanders E, van der Velden VH, van der Schoot CE, et al. Integrated use of minimal residual disease classification and IKZF1 alteration status accurately predicts 79% of relapses in pediatric acute lymphoblastic leukemia. Leukemia. 2011;25(2):254-258. Moorman AV, Schwab C, Ensor HM, et al. IGH@ translocations, CRLF2 deregulation, and microdeletions in adolescents and adults with acute lymphoblastic leukemia. J Clin Oncol. 2012;30(25):3100-3108. Clappier E, Auclerc MF, Rapion J, et al. An intragenic ERG deletion is a marker of an oncogenic subtype of B-cell precursor acute lymphoblastic leukemia with a favorable outcome despite frequent IKZF1 deletions. Leukemia. 2014;28(1):70-77. Zaliova M, Zimmermannova O, Dorge P, et al. ERG deletion is associated with CD2 and attenuates the negative impact of IKZF1 deletion in childhood acute lymphoblastic leukemia. Leukemia. 2014;28(1):182-185. Chen IM, Harvey RC, Mullighan CG, et al. Outcome modeling with CRLF2, IKZF1, JAK, and minimal residual disease in pediatric acute lymphoblastic leukemia: a Children's Oncology Group study. Blood. 2012;119(15):3512-3522. van der Veer A, Waanders E, Pieters R, et al. Independent prognostic value of BCR-ABL1like signature and IKZF1 deletion, but not high CRLF2 expression, in children with Bcell precursor ALL. Blood. 2013;122(15): 2622-2629. Dagdan E, Zaliova M, Dรถrge P, et al. The Strong Prognostic Effect of Concurrent Deletions of IKZF1 and PAX5, CDKN2A, CDKN2B or PAR1 in the Absence of ERG Deletions (IKZF1plus) in Pediatric Acute Lymphoblastic Leukemia Strongly Depends on Minimal Residual Disease Burden after Induction Treatment Blood. 2014;124(21): 131. Moorman AV, Erhorn A, Parker C, et al. The Clinical Relevance Of Genetics In Predicting Outcome After a First Relapse In Children With B-Cell Precursor Acute Lymphoblastic Leukaemia. Blood. 2013;122(21):2566. Russell LJ, Capasso M, Vater I, et al. Deregulated expression of cytokine receptor gene, CRLF2, is involved in lymphoid transformation in B-cell precursor acute lym-

79.

80.

81.

82.

83.

84.

85.

86.

87.

88.

phoblastic leukemia. Blood. 2009;114(13): 2688-2698. Mullighan CG, Collins-Underwood JR, Phillips LA, et al. Rearrangement of CRLF2 in B-progenitor- and Down syndrome-associated acute lymphoblastic leukemia. Nat Genet. 2009;41(11):1243-1246. Chapiro E, Russell L, Lainey E, et al. Activating mutation in the TSLPR gene in Bcell precursor lymphoblastic leukemia. Leukemia. 2010;24(3):642-645. Tasian SK, Doral MY, Borowitz MJ, et al. Aberrant STAT5 and PI3K/mTOR pathway signaling occurs in human CRLF2rearranged B-precursor acute lymphoblastic leukemia. Blood. 2012;120(4):833-842. Cario G, Zimmermann M, Romey R, et al. Presence of the P2RY8-CRLF2 rearrangement is associated with a poor prognosis in non-high-risk precursor B-cell acute lymphoblastic leukemia in children treated according to the ALL-BFM 2000 protocol. Blood. 2010;115(26):5393-5397. Harvey RC, Mullighan CG, Chen IM, et al. Rearrangement of CRLF2 is associated with mutation of JAK kinases, alteration of IKZF1, Hispanic/Latino ethnicity, and a poor outcome in pediatric B-progenitor acute lymphoblastic leukemia. Blood. 2010;115(26):5312-5321. Ensor HM, Schwab C, Russell LJ, et al. Demographic, clinical, and outcome features of children with acute lymphoblastic leukemia and CRLF2 deregulation: results from the MRC ALL97 clinical trial. Blood. 2011;117(7):2129-2136. Palmi C, Vendramini E, Silvestri D, et al. Poor Prognosis for P2RY8-CRLF2 Fusion but not for CRLF2 Over-expression in Children with Intermediate Risk B-Cell Precursor Acute Lymphoblastic Leukemia. Leukemia. 2012;26(10):2245-2253. Morak M, Attarbaschi A, Fischer S, et al. Small sizes and indolent evolutionary dynamics challenge the potential role of P2RY8-CRLF2-harboring clones as main relapse-driving force in childhood ALL. Blood. 2012;120(26):5134-42. Roberts KG, Morin RD, Zhang J, et al. Genetic alterations activating kinase and cytokine receptor signaling in high-risk acute lymphoblastic leukemia. Cancer Cell. 2012;22(2):153-166. Moorman AV, Enshaei A, Schwab C, et al. A

89.

90. 91.

92.

93.

94.

95.

96.

97.

98.

novel integrated cytogenetic and genomic classification refines risk stratification in pediatric acute lymphoblastic leukemia. Blood. 2014;124(9):1434-1444. Den Boer ML, van Slegtenhorst M, De Menezes RX, et al. A subtype of childhood acute lymphoblastic leukaemia with poor treatment outcome: a genome-wide classification study. Lancet Oncol. 2009;10(2):125134. Kiyokawa N, Iijima K, Yoshihara H, et al. An Analysis Of Ph-Like ALL In Japanese Patients. Blood. 2013;122(21):352. Kronnie Gt, Silvestri D, Vendramini E, et al. Philadelphia-Like Signature In Childhood Acute Lymphoblastic Leukemia: The AIEOP Experience. Blood. 2013;122(21):353. Boer JM, Marchante JRM, Horstmann MA, et al. BCR-ABL1-Like Cases In Pediatric Acute Lymphoblastic Leukemia: A Comparison Between COG/St. Jude and Dutch DCOG Signatures. Blood. 2013;122(21):2633. Kobayashi K, Mitsui K, Ichikawa H, et al. ATF7IP as a novel PDGFRB fusion partner in acute lymphoblastic leukaemia in children. Br J Haematol. 2014;165(6):836-841. Lilljebjorn H, Agerstam H, Orsmark-Pietras C, et al. RNA-seq identifies clinically relevant fusion genes in leukemia including a novel MEF2D/CSF1R fusion responsive to imatinib. Leukemia. 2014;28(4):977-979. Lengline E, Beldjord K, Dombret H, Soulier J, Boissel N, Clappier E. Successful tyrosine kinase inhibitor therapy in a refractory B-cell precursor acute lymphoblastic leukemia with EBF1-PDGFRB fusion. Haematologica. 2013;98(11):e146-148. Weston BW, Hayden MA, Roberts KG, et al. Tyrosine kinase inhibitor therapy induces remission in a patient with refractory EBF1PDGFRB-positive acute lymphoblastic leukemia. J Clin Oncol. 2013;31(25):e413416. Schwab C, Andrews R, Chilton L, et al. EBF1-PDGFR Fusion in Paediatric Acute Lymphoblastic Leukaemia (ALL): Genetic Profile and Clinical Implications. Blood. 2014;124(21):1068. Schwab C, Andrews R, Chilton L, et al. SSBP2-CSF1R Is a Recurrent Fusion in BOther Acute Lymphoblastic Leukaemia with Variable Clinical Outcome. Blood. 2014;124(21):3773.

haematologica | 2016; 101(4)


ARTICLE

Hematopoiesis

A novel xenograft model to study the role of TSLP-induced CRLF2 signals in normal and malignant human B lymphopoiesis

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Olivia L. Francis,1* Terry-Ann M. Milford,1* Shannalee R. Martinez,1 Ineavely Baez,1 Jacqueline S. Coats,1 Karina Mayagoitia,1 Katherine R. Concepcion,1 Elizabeth Ginelli,1 Cornelia Beldiman,1 Abigail Benitez,1 Abby J. Weldon,1 Keshav Arogyaswamy,1 Parveen Shiraz,1 Ross Fisher,1 Christopher L. Morris,1 Xiao-Bing Zhang,1 Valeri Filippov,1 Ben Van Handel,2 Zheng Ge,3,4 Chunhua Song,4 Sinisa Dovat,4 Ruijun Jeanna Su,1* and Kimberly J. Payne1*

1 Loma Linda University, CA, USA; 2Novogenix Laboratories, LLC, 1425 San Pablo, CA, USA; 3The First Affiliated Hospital of Nanjing Medical University, Jiangsu Province Hospital, Department of Hematology, Nanjing 210029, China; and 4Pennsylvania State University Medical College, Department of Pediatrics, Hershey, PA, USA *

OLF and TMM contributed equally to this work.

Haematologica 2016 Volume 101(4):417-426

ABSTRACT

T

hymic stromal lymphopoietin (TSLP) stimulates in vitro proliferation of human fetal B-cell precursors. However, its in vivo role during normal human B lymphopoiesis is unknown. Genetic alterations that cause overexpression of its receptor component, cytokine receptor-like factor 2 (CRLF2), lead to high-risk B-cell acute lymphoblastic leukemia implicating this signaling pathway in leukemogenesis. We show that mouse thymic stromal lymphopoietin does not stimulate the downstream pathways (JAK/STAT5 and PI3K/AKT/mTOR) activated by the human cytokine in primary high-risk leukemia with overexpression of the receptor component. Thus, the utility of classic patient-derived xenografts for in vivo studies of this pathway is limited. We engineered xenograft mice to produce human thymic stromal lymphopoietin (+T mice) by injection with stromal cells transduced to express the cytokine. Control (–T) mice were produced using stroma transduced with control vector. Normal levels of human thymic stromal lymphopoietin were achieved in sera of +T mice, but were undetectable in –T mice. Patient-derived xenografts generated from +T as compared to –T mice showed a 3-6-fold increase in normal human B-cell precursors that was maintained through later stages of B-cell development. Gene expression profiles in high-risk B-cell acute lymphoblastic leukemia expanded in +T mice indicate increased mTOR pathway activation and are more similar to the original patient sample than those from –T mice. +T/–T xenografts provide a novel pre-clinical model for understanding this pathway in B lymphopoiesis and identifying treatments for high-risk B-cell acute lymphoblastic leukemia with overexpression of cytokine-like factor receptor 2.

Introduction Thymic stromal lymphopoietin (TSLP) is an IL-7-like cytokine that plays key roles at several points in normal hematopoietic cell development and function.1-4 The role of TSLP in B lymphopoiesis has been evaluated almost exclusively in mice.5-8 Knowledge of TSLP in normal human B-cell development is limited to a single in vitro study showing that TSLP increases production of fetal B-cell precursors.9 Genetic alterations that cause overexpression of the TSLP receptor component, CRLF2, have been linked to B-cell acute lymphoblastic leukemia haematologica | 2016; 101(4)

Correspondence: kpayne@llu.edu

Received: February 11, 2015. Accepted: November 24, 2015. Pre-published: November 25, 2015. doi:10.3324/haematol.2015.125336

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/04/417

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

417


O.L. Francis et al.

(ALL), thus implicating the TSLP pathway in leukemogenesis.10-13 The low homologies of murine versus human TSLP and CRLF2 (approx. 40%)14,15 suggest the need for novel models to study the in vivo role of TSLP in normal and malignant human B lymphopoiesis. Acute lymphoblastic leukemia is the most common

A

B

C

D

E

Figure 1. Mouse TSLP does not activate the human TSLP receptor complex. (A) Pathways activated downstream of TSLP receptor in human cells. (B) CRLF2 BALL cell lines (MUTZ5, MHH-CALL4) and primary CRLF2 B-ALL cells used to produce patient-derived xenografts (PDX) used in the studies described here were stained for flow cytometry to detect the TSLP receptor components (IL-7Rα and CRLF2). Plotted in red are CRLF2 B-ALL cells within living cell light scatter. Quadrants shown are set based on unstained controls (blue overlay) (C-E) CRLF2 B-ALL cell lines and primary CRLF2 B-ALL cells were stimulated with human TSLP (hTSLP), mouse TSLP (mTSLP), or no cytokine and evaluated for phosphorylated STAT5 (pSTAT5) AKT (pAKT), and S6 (pS6) by phospho-flow cytometry.

418

childhood malignancy and primarily involves the B lineage (B-ALL). Although cure rates approach 90%, approximately 10%-20% of patients still relapse.16,17 Genomic profiling has identified several high-risk B-ALL subtypes that are chemoresistant.18-21 These include CRLF2 B-ALL, a leukemia with genetic alterations causing overexpression of the CRLF2 protein on the surface of B-ALL cells.10-13 CRLF2 and the IL-7 receptor alpha chain (IL-7Rα) together form the TSLP receptor signaling complex (Figure 1A).8, 22 Binding of TSLP induces CRLF2 and IL-7Rα dimerization leading to activation of the JAK-STAT523,24 and the PIK3/AKT/mTOR pathways,25,26 as demonstrated in CRLF2 B-ALL.27 The finding that JAK kinases are mutated in CRLF2 B-ALL28 suggested that CRLF2 and mutated JAK co-operate to induce constitutive STAT5 activation in CRLF2 B-ALL.29,30 However, approximately half of CRLF2 B-ALL lack JAK mutations. Thus, the role of TSLP in the leukemogenesis of CRLF2 B-ALL remains unclear and the mechanisms for its contribution to chemoresistance are unknown. The genetic landscape produced by inherited germline variations contributes to leukemogenesis and disease outcome,31 and is a biological component that contributes to racial, ethnic and other health disparities in ALL.32 This is particularly relevant in CRLF2 B-ALL which occurs five times more often in Hispanic children than others28 and comprises more than half of the ALL cases in children with Down Syndrome.11,18,33 Patient-derived xenograft (PDX) models produced by injecting human cells into immune deficient mice provide in vivo pre-clinical models for understanding disease mechanisms and identifying effective therapies in the context of the range of genetic landscapes present in the human population. However, engineered cellular models8 suggest that mouse TSLP (mTSLP) is species-specific and unlikely to stimulate CRLF2-mediated signaling in human cells.8,22 Given the role of TSLP in activating the CRLF2 pathway,27 and the identification of CRLF2 as a biological component of health disparities in CRLF2 B-ALL,11,18,28,33 it is important that studies to identify disease mechanisms and potential therapies for this leukemia be performed in pre-clinical models that provide human TSLP (hTSLP). Our goal was to develop and validate a xenograft model that can be used to study the role of hTSLP in normal and malignant B lymphopoiesis. Here we describe the development of a novel xenograft model system comprised of mice that provide hTSLP (+T mice) and mice that do not (–T mice). PDX generated from +T mice show functional in vivo hTSLP effects, expanding the production of normal B-cell precursors from hematopoietic stem cells (HSCs) and inducing changes in mTOR-regulated gene expression in primary CRLF2 B-ALL cells. These data validate the +T/–T xenograft model system as an important new preclinical model for understanding the role of hTSLP in normal and malignant B lymphopoiesis and for identifying therapies to treat CRLF2 B-ALL.

Methods Human samples and cell lines Human CRLF2 B-ALL cell lines MUTZ5 and MHH-CALL4 were from DSMZ (Braunschweig, Germany) and the human stroma line HS-27a (HS27) from ATCC (Manassas, VA, USA). Primary human CRLF2 B-ALL cells (Online Supplementary Table S1) and haematologica | 2016; 101(4)


Xenograft to study TSLP in human B lymphopoiesis

umbilical cord blood (CB) were obtained in accordance with protocols approved by the Loma Linda University Institutional Review Board (IRB) and with the 1975 Declaration of Helsinki, as revised in 2008. Details of HS27 stroma characteristics, CB CD34+ cell isolation, and culture media are described in the Online Supplementary Methods.

Mice Studies were performed using NOD.Cg-Prkdcscid Il2rgtm1Wjl/SzJ (NSG) mice (Jackson Laboratory) housed under pathogen-free conditions at Loma Linda University Animal Care Facility. Studies were conducted in accordance with the Loma Linda University Institutional Animal Care and Use Committee (IACUC) approved protocols.

Flow cytometry Cells were prepared for flow cytometry using specific antibody clones as described in the Online Supplementary Methods. Flow cytometry analysis was performed using a MACSQuant analyzer (Miltenyi Biotec Inc., Auburn, CA, USA) and FlowJo analysis software (FlowJo, Ashland, OR, USA).

Stromal cell transduction Human HS27 stromal cells were transduced to express hTSLP (+T stroma) or with control vector (–T stroma) at a multiplicity of infection of 0.5-5/cell overnight. After transduction, cells were washed, expanded in culture, and frozen in aliquots. For the generation of +T and –T mice, stromal cells were thawed, cultured, and injected into mice within 15 passages after thawing. Details of lentiviral vector production and ELISAs to determine hTSLP levels are available in the Online Supplementary Methods.

Xenograft transplantation HS27 stroma transduced to be +T or –T were injected intraperitoneally into NSG mice at doses of 5 or 10x106 cells at time points described in the Results section. Mice were sub-lethally irradiated with 225 cGy, then transplanted by tail vein injection with freshly thawed human umbilical CB CD34+ cells (1x105) or primary CRLF2 B-ALL cells (5x105 Patient 1 and 1.5x106 Patient 2).

Gene expression analysis in CRLF2 B-ALL cells Microarray analyses were performed by Miltenyi Biotec’s Genomic Services (Miltenyi Biotec GmbH, Bergisch Gladbach)

A E

F

B G

C

haematologica | 2016; 101(4)

D

Figure 2. Engineering xenograft mice to express normal serum levels of human TSLP. (A) Strategy for developing the –/+hTSLP xenograft model. (B) ELISA assays of hTSLP protein in culture supernatant from HS27 +T and HS27 –T stromal cells collected after multiple passages following the initial transduction (Transduction 1) or at early (1-5), mid (6-10) and late (11-15) postthaw passages following Transduction 2. (C-D) Elisa assays of serum hTSLP in blood collected at indicated time points from +T or –T mice injected with stroma as follows: (C) loading dose (3 doses of 1x107 in first week) followed by weekly doses of 1x107 stromal cells (Transduction 1), graphed are the means ± SEM of n = 2-7 mice each time point. (D) injection of 5x106 stromal cells (Transduction 2) at weekly intervals; graphed are the means ± SEM n ≥ 24 mice at each time point. (E-G) CRLF2 B-ALL cell lines were stimulated with stromal cell supernatant (Transduction 2) and levels of pSTAT5, pAKT and pS6 were measured by phospho-flow cytometry.

419


O.L. Francis et al.

using Agilent Whole Human Genome Oligo Microarrays (Design ID 039494, 8x60K v.2). Libraries for RNA sequencing were prepared and sequenced by the UCLA Clinical Microarray Core. Data discussed in this publication are deposited in the NCBI Gene Expression Omnibus (GEO).34 Microarray and RNA sequencing data are available through accession numbers GSE65274 and GSE74339, respectively, at http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc= GSExxx. Gene set enrichment analysis (GSEA) was performed by Novogenix Laboratories, LLC (Los Angeles, CA, USA). Details of microarray, RNA sequencing, data processing, differential gene expression analysis, GSEA, and quantitative RT-PCR analysis are available in the Online Supplementary Appendix. Further details are available in the Online Supplementary Methods.

Human HS27 stromal cells38 were transduced with lentiviral vector containing hTSLP (+T stroma) or with control vector (–T stroma). Culture supernatant collected from stroma after the initial transduction showed hTSLP levels of approximately 500 pg/mL (Figure 2B, Transduction 1). The ability of +T stroma to generate measureable serum levels of hTSLP in vivo was evaluated. NSG mice receiving weekly intraperitoneal injections of

A

Results Mouse TSLP does not activate the human TSLP receptor complex Activation of the human TSLP receptor complex can be induced by hTSLP, but not by mTSLP, in murine Ba/F3 cells transduced to express human CRLF2 and IL-7Rα.8 However, the ability of mTSLP to activate CRLF2-signaling in human cells, particularly those expressing high levels of CRLF2 observed in CRLF2 B-ALL, has not been reported. To determine if mTSLP can induce CRLF2-mediated signals, we tested its ability to activate downstream pathways in human CRLF2 B-ALL cells. First, we verified that the CRLF2 B-ALL cell lines and primary CRLF2 B-ALL cells used in our studies express the TSLP receptor components. Expression of IL-7Rα, as well as high levels of CRLF2, was observed in both CRLF2 B-ALL cell lines and in the 2 primary CRLF2 B-ALL samples used in studies described here (Figure 1B). We then used phospho-flow cytometry to determine whether mTSLP can activate the JAK-STAT5 and PI3K/AKT/mTOR pathways activated by hTSLP in CRLF2 B-ALL cells.27 hTSLP induced robust increases in phosphorylated STAT5 (pSTAT5) and S6 (pS6, activated downstream of mTOR) and more modest increases in phosphorylated AKT (pAKT) in CRLF2 B-ALL cell lines and primary CRLF2 B-ALL cells (Figure 1C-E). In contrast, stimulation of CRLF2 B-ALL cells with mTSLP showed no effect (Figure 1C-E).35 These data show that mTSLP fails to induce the STAT5, AKT, or S6 phosphorylation that are characteristic of CRLF2-mediated signals induced by hTSLP in human CRLF2 B-ALL cells. Thus, TSLP produced in the existing mouse xenograft models does not induce human CRLF2-mediated signals.

Engineering xenograft mice to express normal serum levels of human TSLP We developed a strategy for engineering xenograft mice that provide hTSLP to activate CRLF2-mediated signals (Figure 2A) by building on a previous cytokine delivery approach.36 Our goal was to produce stable, sustained serum levels of hTSLP, similar to normal levels reported in children (range 13-32 pg/mL)37 by intraperitoneal injection of stromal cells transduced to express high levels of hTSLP (+T mice) (Figure 2A). Control mice (–T mice) were produced by injecting stroma transduced with empty vector. Because the only difference between +T and –T mice is the presence of hTSLP in +T mice, this model would allow us to identify the in vivo effects of hTSLP by comparing hematopoiesis in +T and –T PDX generated from these mice. 420

B

C

D

Figure 3. hTSLP produced in xenograft mice exerts in vivo functional effects on normal human B-lineage cells. (A) –T and +T NSG mice were transplanted with 1x105 CD34+ human cord blood (CB) cells and injected with weekly doses of 5x105 stromal cells (Transduction 2 stroma). (B) Five weeks post transplant, mice were euthanized, BM harvested and cells were stained for flow cytometry to detect mouse CD45 (mCD45), human CD45 (hCD45), and hCD19. Total living cells were gated and mCD45 versus hCD45 is plotted with gates to identify mouse and human leukocytes, respectively, in the BM. (C) hCD45+ cells were gated and hCD19 expression is shown by histogram. Percentages are mean ± SEM. Data shown in A-C were obtained from CB2, n=2 –T PDX and n=4 +T PDX; data from two additional CB samples (CB1 and CB3) are shown in Online Supplementary Figures S3-S4. (D) Graphed are the fold change in total human cells (hCD45+ ), human B-lineage cells (CD19+), and human non-B-lineage cells (CD19-). Fold change was obtained by normalizing cell numbers in +T animals to the average cell numbers in –T animals for each experiment. (D) Data pooled from 3 different experiments performed with CB samples 1-3. Total n=9 –T PDX mice and n=14 +T PDX mice. Statistical significance was calculated using a two-tailed, unpaired t-test (*P≤0.05; ** P≤0.05).

haematologica | 2016; 101(4)


Xenograft to study TSLP in human B lymphopoiesis

5x106+T stroma showed serum hTSLP levels that were low but detectable (4-10 pg/mL) (data not shown). To achieve higher levels of sustained hTSLP expression, NSG mice were injected with 3 doses of 1x107 +T or –T stroma over a 1-week period (loading dose) followed by weekly injections of 1x107 stromal cells. Serum collected at weekly time points showed serum hTSLP levels of approximately 80 pg/mL following the initial 3 doses of +T stroma but quickly dropped to 15-20 pg/mL with +T stroma injec-

tions (Figure 2C, left panel). hTSLP was not detectable in the sera of mice transplanted with –T stroma (Figure 2C). These data show that +T stroma can generate normal serum levels of hTSLP and that serum levels of hTSLP correspond to the number of stromal cells injected during the previous 1-2 week period. These data suggested that stable, sustained hTSLP production might be more likely with weekly injections of stroma that could deliver slightly higher doses of hTSLP.

A

B

C

Figure 4. B-cell precursor production is increased in +T PDX. Bone marrow (BM) and spleen were harvested from –T and +T PDX mice generated from 3 different CB samples as shown in Figure 3, and in Online Supplementary Figures S3 and S4. (A) Harvested cells were counted and stained for flow cytometry to identify progenitors and B-cell subsets in the BM and spleen based on indicated immunophenotypes. (B) The number of cells and fold change in each subset were obtained by normalizing cell numbers in +T animals to the average cell numbers in –T animals for each experiment. (C) Graphed is the frequency of each subset within the total CD19+ B-lineage pool. Data shown are pooled from experiments performed using PDX generated from CB 1-3. n=9 –T PDX mice and n=15 +T PDX mice. Statistical significance was calculated using a two-tailed, unpaired t-test (**P≤0.01; ***P ≤0.001).

haematologica | 2016; 101(4)

421


O.L. Francis et al.

A second stromal cell transduction produced hTSLP levels averaging approximately 900 pg/mL in culture supernatants harvested at early, mid and late passages from +T stroma across multiple thaws of frozen aliquots (Figure 2, Transduction 2). (hTSLP was undetectable in supernatant from similarly passaged –T stroma (Figure 2, Transduction 2). NSG mice transplanted at weekly time points with 5x106 stroma from Transduction 2 showed serum hTSLP levels of approximately 20 pg/mL that were stable for up to 11 weeks (Figure 2D and data not shown). Evaluation of BM and spleen of +T and –T mice by flow cytometry showed that HS27 stroma (CD29+CD90+) were not detectable in either the spleen or BM of PDX mice (Online Supplementary Figure S1). These data provide evidence that the HS27 +T stroma in the peritoneal cavity give rise to normal serum levels of hTSLP in PDX mice. To verify that the hTSLP produced in our model induces CRLF2-mediated signals, CRLF2 B-ALL cell lines were assessed for activation of the JAK/STAT5 and PI3K/AKT/mTOR pathways. Supernatant from +T stroma induced robust pSTAT5 and pS6 and minimal pAKT (Figure 2E-G). In contrast, CRLF2 B-ALL cells stimulated with supernatant from –T stroma showed background levels of pSTAT5, pAKT, and pS6, similar to cells incubated with medium alone (Figure 2E-G). These data show that +T stroma, but not –T stroma, used in our model produce hTSLP that can activate signaling in human CRLF2 BALL cells.

A

B

C

D

hTSLP produced in xenograft mice exerts in vivo functional effects on normal human B-lineage cells We evaluated the ability of hTSLP in +T PDX to induce in vivo functional effects on human B-lineage cells. hTSLP has been shown to increase the in vitro production of normal human B-cell precursors9 despite their low levels of CRLF2 expression (Online Supplementary Figure S2). Thus, expansion of normal human B-cell precursors provides a sensitive in vivo functional test of the hTSLP produced in our xenograft model. We compared normal B-cell production in +T and –T NSG mice transplanted with umbilical CB CD34+ cells from 3 different donors (Figure 3A and Online Supplementary Figures S3A and S4A). Five weeks post-transplant mice were euthanized and bone marrow (BM) harvested for flow cytometry to evaluate the production of B-lineage and other hematopoietic cells. Mouse and human cells were identified based on expression of mouse CD45 (mCD45) or human CD45 (hCD45), respectively (Figure 3B and Online Supplementary Figures S3B and S4B). Flow cytometry analysis of hCD19 expression on gated human cells showed an increase in the percentage of B-lineage cells in +T as compared to –T PDX mice generated from all 3 CB donors (Figure 3C and Online Supplementary Figures S3C and S4C). Pooled data comparing BM cellularity in all 3 experiments showed that the total number of human cells (hCD45+) was increased by 34-fold in +T as compared to –T PDX and this was due to significant increases in CD19+ (B-lineage) cells but not CD19– cells (Figure 3D). These data suggest that hTSLP in +T PDX selectively expands human B-lineage cells rather than exerting global effects on hematopoiesis.

B-cell precursor production is increased in +T PDX mice To confirm that normal human B-cell precursors are expanded in +T PDX, we compared the production of 422

E

Figure 5. hTSLP in +T xenograft mice induces mTOR-regulated genes in primary CRLF2 B-ALL. (A) –T and +T NSG mice were transplanted with primary CRLF2 B-ALL cells from Patient 1 and injected weekly with 5x106 stromal cells (transduction 1). (B) Nine weeks post transplant, mice were euthanized, BM was harvested and cells were stained with mCD45 and hCD19 to identify human BALL cells. CRLF2 staining (black overlay) on gated hCD19+ cells is shown in bottom panels. Isotype control is shown in gray. (C) Human hCD19+ B-ALL cells were isolated by magnetic separation and analyzed using whole genome microarray to identify genes that are differentially regulated in +T as compared to –T PDX. Graphed are the numbers of Agilent gene reporters up or down-regulated (≥2 fold, P≤0.05) in CRLF2 B-ALL cells from +T as compared to –T PDX. (D) qRT-PCR validation of whole genome microarray. Regression analysis of fold changes measured by microarray versus fold changes measured by qRT-PCR. (E) Gene Set Enrichment Analysis (GSEA) of whole genome microarray data shows that CRLF2 B-ALL cells harvested from +T PDX exhibit an increase in expression of genes regulated downstream of mTOR signaling as compared to cells from –T PDX. The top half of the GSEA enrichment plot shows the enrichment score for each gene and the bottom half shows the values of the ranking metric moving down the list of the ranked genes.

haematologica | 2016; 101(4)


Xenograft to study TSLP in human B lymphopoiesis

human B-cell subsets (Figure 4A) and non-B-cell lineages in –T and +T PDX. Staining of PDX BM for markers of human non-B cells showed that human CD19– cells are comprised primarily of monocytes (CD14+), but also include granulocytes (CD66B+), NK cells (CD56+), and CD34+ cells. The number of cells in each non-B lineage was similar (data not shown) in the BM from –T and +T PDX, indicating that the production of these cells is not skewed. However, the production of B-cell precursors was increased 3-6-fold at each stage of B-cell development in +T PDX as compared to –T PDX, beginning with the earliest CD34+ pro-B cells (Figure 4B). This increase was maintained through the mature B-cell stage in BM and spleen (Figure 4B) with no skewing of B-cell subsets in the BM (Figure 4C) or spleen (data not shown). These data provide evidence that hTSLP produced in vivo in +T PDX has a functional effect on normal human B-cell precursors that express low levels of CRLF2.

A

hTSLP in xenograft mice induces mTOR-regulated genes in primary CRLF2 B-ALL To determine the ability of hTSLP in +T xenografts to activate CRLF2 downstream pathways in CRLF2-B-ALL cells, we used primary samples obtained from 2 different Hispanic pediatric patients. Patients' characteristics are shown in Online Supplementary Table S1. First, we evaluated the ability of NSG mice transplanted with HS27 stroma to support engraftment of primary CRLF2 B-ALL cells. PDX from Patient 1 were generated using NSG mice injected with +T stroma from transduction 1 (stroma injection schedule) (Figure 5A). Serum levels of hTSLP produced in these mice were detectable but below normal levels (4-10 pg/mL) and therefore provided a sensitive test of in vivo hTSLP activity in +T PDX. +T and –T NSG mice were transplanted with primary CRLF2 B-ALL cells (Figure 1B, right panel) from Patient 1. Mice were euthanized at nine weeks when peripheral blood chimerism reached

D

B E

C

Figure 6. +T PDX show gene expression profiles that are more similar to the original patient sample. (A) –T and +T NSG mice were transplanted with primary CRLF2 B-ALL cells from Patient 2. Beginning at ten weeks after leukemia transplant mice received two injections with 5x106 stromal cells (Transduction 2 stroma) one week apart. (B) Twelve weeks post transplant, mice were euthanized, BM was harvested stained with mCD45 and hCD19 to identify human B-ALL cells. CRLF2 staining (black overlay) on gated hCD19+ cells is shown in bottom panels. Isotype control is shown in gray. (C-E) RNA sequencing was performed on CRLF2 B-ALL cells isolated by magnetic separation from 2 +T PDX mice (+T PDX M1 and M2) and 2 –T PDX mice (–T PDX M1 and M2) and from the original Patient 2 primary sample used to generate them. (C) Differential gene expression analysis was performed and graphed are the numbers of significantly up- or down-regulated genes (≥2-fold, P≤0.05) in CRLF2 B-ALL cells from +T as compared to –T PDX. (D) GSEA of RNA sequencing data shows that CRLF2 B-ALL cells from +T PDX exhibit an increase in expression of genes regulated downstream of mTOR signaling as compared to cells from –T PDX. (E) Graphed is a dendrogram of unsupervised hierarchical clustering and heatmap of Spearman rank correlation values showing that gene expression in +T PDX is more closely correlated (P<0.0001) to the original sample than is –T PDX. Spearman rank coefficients are shown in Online Supplementary Figure S5.

haematologica | 2016; 101(4)

423


O.L. Francis et al.

80%. BM was harvested and stained for mCD45 and human-specific markers (CRLF2 and CD19) to identify CRLF2 B-ALL cells (Figure 5B). Flow cytometry analysis showed BM chimerism of human CRLF2 B-ALL cells was more than 90% in +T and –T PDX (Figure 5B). These data show that NSG mice support engraftment of primary BALL cells when injected with –T or with +T stroma. The leukemia cell burden in the BM of –T and +T PDX generated from Patient 1 and from Patient 2 showed no consistent differences (Online Supplementary Figure S6). Thus, our next question was whether hTSLP in +T animals induces gene expression profiles associated with pathway activation downstream of CRLF2-mediated signaling in primary CRLF2 B-ALL cells. Human CRLF2 B-ALL cells were isolated from the BM of +T and –T PDX generated from Patient 1 by magnetic separation. Gene expression was compared by whole genome microarray performed in triplicate. Differential gene expression analysis identified 280 gene reporters that were up-regulated and 281 gene reporters that were down-regulated at least 2-fold (Figure 5C). Microarray was validated by quantitative RT-PCR (qRT-PCR) analysis of selected differentially expressed genes (Online Supplementary Table S2). Changes in gene expression by microarray and qRT-PCR were strongly correlated as indicated by linear regression (Figure 5D). GSEA39,40 of whole genome microarray data showed a strong enrichment for genes regulated downstream of mTOR activation in CRLF2 B-ALL cells from +T PDX as compared to –T PDX (FDR q-value=0.022) (Figure 5E). These data provide evidence that the hTSLP in +T PDX, even when present at low levels, acts on CRLF2 B-ALL cells to increase mTOR pathway activation downstream of CRLF2.

+T PDX show gene expression profiles that are more similar to the original patient sample Next we performed experiments to determine whether +T PDX provide an in vivo environment more like the patient than –T PDX and if this could be achieved with short-term in vivo hTSLP exposure. PDX were established using CRLF2 B-ALL cells from Patient 2. Injection of stromal cells from transduction 2 were not initiated until two weeks prior to euthanasia (Figure 6A). At euthanasia, BM was harvested and stained for flow cytometry to verify CRLF2 B-ALL engraftment (Figure 6B). Human leukemia cells were isolated from the BM of +T and –T PDX by magnetic separation for RNA sequencing. Differential gene expression analysis identified 503 genes that were up-regulated and 117 that were down-regulated at least 2fold in CRLF2 B-ALL cells from +T as compared to –T PDX (Figure 6C). Similar to what we observed with Patient 1, GSEA analysis of RNA sequencing data showed an enrichment for genes regulated downstream of mTOR in CRLF2 B-ALL cells from +T as compared to –T PDX (FDR q-value = 0.00048) (Figure 6D). A comparison of +T PDX to –T PDX showed similar immunophenotypes to each other and to that reported for the respective original Patient 1 and Patient 2 samples (Online Supplementary Table S1 and Figure S7), indicating that the differences in mTORregulated gene expression are unlikely to be due to selective survival of different clones. These data demonstrate that in vivo hTSLP-induced changes in mTOR regulated gene expression can be observed after short-term stimulation and validate our model in a second primary CRLF2 B-ALL patient sample. 424

To determine whether +T PDX provide an in vivo environment that is more similar to the patient than –T PDX, we compared gene expression profiles of +T and –T PDX to the original ex vivo CRLF2 B-ALL patient sample. Human leukemia cells were isolated from the BM of +T and –T PDX generated from Patient 2, and from the original primary Patient 2 sample by magnetic separation. RNA sequencing was performed to determine whole genome expression. Analysis of gene expression by unsupervised clustering and Spearman rank correlation show that gene expression in +T PDX, as compared to –T PDX, is significantly more closely correlated to the original patient sample (Figure 6E). These data validate the use of +T/–T xenograft mice as an in vivo pre-clinical model system for studies to understand the role of hTSLP in normal human B-cell production and in the development and progression of CRLF2 B-ALL.

Discussion Patient-derived xenografts provide a critical complement to other models in understanding normal and malignant human B lymphopoiesis in the context of the broad background of human genetic variation. A limitation of PDX is the reduced cross-species activity of some cytokines. Classic immune deficient mice, including NSG mice, selectively support B-lineage differentiation. This is because IL-7, a cytokine important for human B-cell production,41 shows high species cross reactivity42 while myeloid supportive cytokines do not. Here we show that mTSLP does not induce the increases in STAT5 and PI3K/AKT/mTOR phosphorylation induced by hTSLP in CRLF2 B-ALL cells (Figure 1). CRLF2 B-ALL cells express abundant IL-7Rα and CRLF2 as compared to normal B-cell precursors (Figure 1 and Online Supplementary Figure S1) and thus provide a sensitive test for mTSLP activity. These data establish that the classic mouse xenograft models do not provide cross-species activation of the human TSLP receptor. The goals of the studies described here were to develop a xenograft model that expresses hTSLP and to validate the effects of hTSLP produced in the model on normal and malignant B-cell precursors. We engineered NSG mice that express normal serum levels of hTSLP (+T mice) or that lack hTSLP (–T mice) (Figure 2) by intraperitoneal injection with stroma transduced to express hTSLP (+T stroma) or with a control vector (–T stroma) (Figure 2). hTSLP produced by +T stroma induced in vitro STAT5 and PI3K/AKT/mTOR pathway activation (Figure 2) and showed in vivo functional effects as indicated by selective expansion of normal B-cell precursors (Figures 3 and 4 and Online Supplementary Figures S3 and S4) and upregulation of the mTOR pathway in primary CRLF2 B-ALL cells (Figures 5 and 6). The studies described here establish and validate a novel hTSLP+/– xenograft model system while providing the first data on the in vivo impact of hTSLP on normal B-cell development and primary CRLF2+ B-ALL cells. Normal serum levels of hTSLP were achieved in our model; however, a novel aspect of our model is the ability to modulate serum levels of hTSLP based on the timing and number of stromal cells injected intraperitoneally. The ability to vary hTSLP levels or initiate hTSLP at time points after transplant of hematopoietic cells offers advantages for experimental design. The ability to modulate haematologica | 2016; 101(4)


Xenograft to study TSLP in human B lymphopoiesis

serum levels of hTSLP is also biologically relevant because TSLP production is increased by environmental factors such as pollutants and allergens, and is up-regulated in some autoimmune diseases and solid tumors.43 TSLP is primarily produced by cells residing outside the BM (immune cells, smooth muscle cells, gut and skin epithelial cells),44 although human BM stroma do produce hTSLP.45 Thus, the major source of hTSLP that acts on normal and malignant B cells in vivo in patients is likely to be produced outside the marrow as it is in our model. Little is known about cellular mechanisms that contribute to the initiation, progression and maintenance of CRLF2 B-ALL or the biological factors that contribute to the health disparities associated with this disease. Studies of the relationship between inherited genetic variation and susceptibility to ALL provide some clues to the relationship of ALL to ethnicity.46 Patient-derived xenograft (PDX) models are particularly critical for health disparities in diseases where inherited genetic variation or alterations such as Down Syndrome are likely to play a role in disease progression.47,48 The ability to generate PDX mice that provide hTSLP makes it possible to study the initiation, progression and maintenance of CRLF2 B-ALL from normal human progenitors transduced to over-express CRLF2 on a range of genetic backgrounds. The +T and –T PDX generated in the studies described here did not show consistent differences in the leukemia cell burden (Online Supplementary Figure S6). However, changes in gene expression that contribute to chemoresistance would not necessarily be expected to impact leukemia burden in untreated PDX, but would instead be revealed following chemotherapy. CRLF2 B-ALL is a leukemia at high risk for relapse and the mTOR pathway is known to play a role in chemoresistance in multiple malignancies.49 The ability of +T PDX to model TSLPinduced mTOR activation makes this an important model for understanding and identifying therapeutic strategies for overcoming chemoresistance in CRLF2 B-ALL. Our finding that gene expression in +T PDX is significantly

References 1. Ito T, Wang YH, Duramad O, et al. TSLPactivated dendritic cells induce an inflammatory T helper type 2 cell response through OX40 ligand. J Exp Med. 2005;202(9):1213-1223. 2. Ying S, O'Connor B, Ratoff J, et al. Thymic stromal lymphopoietin expression is increased in asthmatic airways and correlates with expression of Th2-attracting chemokines and disease severity. J Immunol. 2005;174(12):8183-8190. 3. Zhou BH, Comeau MR, De Smedt T, et al. Thymic stromal lymphopoietin as a key initiator of allergic airway inflammation in mice. Nat Immunol. 2005;6(10):1047-1053. 4. Siracusa MC, Saenz SA, Hill DA, et al. TSLP promotes interleukin-3-independent basophil haematopoiesis and type 2 inflammation. Nature. 2011;477(7363):229-233. 5. Ray RJ, Furlonger C, Williams DE, Paige CJ. Characterization of thymic stromal-derived lymphopoietin (TSLP) in murine B cell development in vitro. Eur J Immunol. 1996;26(1):10-16. 6. Sims JE, Williams DE, Morrissey PJ, et al.

haematologica | 2016; 101(4)

7.

8.

9.

10.

11.

more like the original patient sample than –T PDX (Figure 6E and Online Supplementary Figure S5) supports the use of +T PDX mice as a pre-clinical model of CRLF2 B-ALL. PDX generated from +T and –T mice and CRLF2 B-ALL samples obtained from Hispanic or Down Syndrome pediatric patients will allow us to study disease mechanisms: 1) in the context of a genetic background associated with CRLF2 B-ALL health disparities; and 2) in the context of TSLP-CRLF2-mediated signals present in vivo in the patient. The model developed here provides an important new tool for understanding the mechanisms of TSLP activity in normal and malignant B lymphopoiesis. It also provides a pre-clinical model for identifying therapies to effectively treat CRLF2 B-ALL in an environment that can induce the TSLP-mediated signals present in vivo in patients and in the context of the genetic background that leads to health disparities in this disease. Acknowledgments The authors would like to thank Batul T. Suterwala, Jonathon L. Payne, and Michelle Choe for manuscript critique. Funding This work was supported by NIH R21CA162259, NIH P20 MD006988, NIH 2R25 GM060507, a St. Baldrick’s Research Grant and a Hyundai Hope on Wheels Scholar Hope Grant, the Department of Pathology and Human Anatomy, the Department of Basic Sciences, and the Center for Health Disparities and Molecular Medicine at Loma Linda University School of Medicine (to KJP) and by a Grant to Promote Collaborative and Translational Research from Loma Linda University (to KJP and CLM). This work was also supported by NIH R01 HL095120, a St Baldrick’s Foundation Career Development Award, a Hyundai Hope on Wheels Scholar Grant Award, the Four Diamonds Fund of the Pennsylvania State University, and the John Wawrynovic Leukemia Research Scholar Endowment (to SD) The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.

Molecular cloning and biological characterization of a novel murine lymphoid growth factor. J Exp Med. 2000;192(5):671-680. Osborn MJ, Ryan PL, Kirchhof N, Panoskaltsis-Mortari A, Mortari F, Tudor KS. Overexpression of murine TSLP impairs lymphopoiesis and myelopoiesis. Blood. 2004;103(3):843-851. Reche PA, Soumelis V, Gorman DM, et al. Human thymic stromal lymphopoietin preferentially stimulates myeloid cells. J Immunol. 2001;167(1):336-343. Scheeren FA, van Lent AU, Nagasawa M, et al. Thymic stromal lymphopoietin induces early human B-cell proliferation and differentiation. Eur J Immunol. 2010;40(4):955965. Yoda A, Yoda Y, Chiaretti S, et al. Functional screening identifies CRLF2 in precursor B-cell acute lymphoblastic leukemia. Proc Natl Acad Sci USA. 2010;107(1):252-257. Russell LJ, Capasso M, Vater I, et al. Deregulated expression of cytokine receptor gene, CRLF2, is involved in lymphoid transformation in B-cell precursor acute lymphoblastic leukemia. Blood. 2009;

114(13):2688-2698. 12. Chapiro E, Russell L, Lainey E, et al. Activating mutation in the TSLPR gene in B-cell precursor lymphoblastic leukemia. Leukemia. 2010;24(3):642-645. 13. Cario G, Zimmermann M, Romey R, et al. Presence of the P2RY8-CRLF2 rearrangement is associated with a poor prognosis in non-high-risk precursor B-cell acute lymphoblastic leukemia in children treated according to the ALL-BFM 2000 protocol. Blood. 2010;115(26):5393-5397. 14. Quentmeier H, Drexler HG, Fleckenstein D, et al. Cloning of human thymic stromal lymphopoietin (TSLP) and signaling mechanisms leading to proliferation. Leukemia. 2001;15(8):1286-1292. 15. Tonozuka Y, Fujio K, Sugiyama T, Nosaka T, Hirai M, Kitamura T. Molecular cloning of a human novel type I cytokine receptor related to delta1/TSLPR. Cytogenet Cell Genet. 2001;93(1-2):23-25. 16. Raetz EA, Bhatla T. Where do we stand in the treatment of relapsed acute lymphoblastic leukemia? Hematology Am Soc Hematol Educ Program. 2012;2012:129136.

425


O.L. Francis et al. 17. Hunger SP, Lu X, Devidas M, et al. Improved survival for children and adolescents with acute lymphoblastic leukemia between 1990 and 2005: a report from the children's oncology group. J Clin Oncol. 2012;30(14):1663-1669. 18. Mullighan CG, Collins-Underwood JR, Phillips LA, et al. Rearrangement of CRLF2 in B-progenitor- and Down syndrome-associated acute lymphoblastic leukemia. Nat Genet. 2009;41(11):1243-1246. 19. Mullighan CG, Goorha S, Radtke I, et al. Genome-wide analysis of genetic alterations in acute lymphoblastic leukaemia. Nature. 2007;446(7137):758-764. 20. Mullighan CG. Genomic profiling of Bprogenitor acute lymphoblastic leukemia. Best Pract Res Clin Haematol. 2011;24 (4):489-503. 21. Chen IM, Harvey RC, Mullighan CG, et al. Outcome modeling with CRLF2, IKZF1, JAK, and minimal residual disease in pediatric acute lymphoblastic leukemia: a Children's Oncology Group study. Blood. 2012;119(15):3512-3522. 22. Park LS, Martin U, Garka K, et al. Cloning of the murine thymic stromal lymphopoietin (TSLP) receptor: Formation of a functional heteromeric complex requires interleukin 7 receptor. J Exp Med. 2000;192(5):659-670. 23. Wohlmann A, Sebastian K, Borowski A, Krause S, Friedrich K. Signal transduction by the atopy-associated human thymic stromal lymphopoietin (TSLP) receptor depends on Janus kinase function. Biol Chem. 2010;391(2-3):181-186. 24. Rochman Y, Kashyap M, Robinson GW, et al. Thymic stromal lymphopoietin-mediated STAT5 phosphorylation via kinases JAK1 and JAK2 reveals a key difference from IL-7-induced signaling. Proc Natl Acad Sci USA. 2010;107(45):19455-19460. 25. Zhong J, Kim MS, Chaerkady R, et al. TSLP signaling network revealed by SILACbased phosphoproteomics. Mol Cell Proteomics. 2012;11(6):M112.017764. 26. Zhong J, Sharma J, Raju R, et al. TSLP signaling pathway map: a platform for analysis of TSLP-mediated signaling. Database (Oxford). 2014;2014:bau007. 27. Tasian SK, Doral MY, Borowitz MJ, et al. Aberrant STAT5 and PI3K/mTOR pathway signaling occurs in human CRLF2-

426

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

rearranged B-precursor acute lymphoblastic leukemia. Blood. 2012;120(4):833-842. Harvey RC, Mullighan CG, Chen IM, et al. Rearrangement of CRLF2 is associated with mutation of JAK kinases, alteration of IKZF1, Hispanic/Latino ethnicity, and a poor outcome in pediatric B-progenitor acute lymphoblastic leukemia. Blood. 2010;115(26):5312-5321. Roll JD, Reuther GW. CRLF2 and JAK2 in B-progenitor acute lymphoblastic leukemia: a novel association in oncogenesis. Cancer Res. 2010;70(19):7347-7352. Malin S, McManus S, Cobaleda C, et al. Role of STAT5 in controlling cell survival and immunoglobulin gene recombination during pro-B cell development. Nat Immunol. 2010;11(2):171-179. Moriyama T, Relling MV, Yang JJ. Inherited genetic variation in childhood acute lymphoblastic leukemia. Blood. 2015;125(26): 3988-3995. Lim JY, Bhatia S, Robison LL, Yang JJ. Genomics of racial and ethnic disparities in childhood acute lymphoblastic leukemia. Cancer. 2014;120(7):955-962. Hertzberg L, Vendramini E, Ganmore I, et al. Down syndrome acute lymphoblastic leukemia, a highly heterogeneous disease in which aberrant expression of CRLF2 is associated with mutated JAK2: a report from the International BFM Study Group. Blood. 2010;115(5):1006-1017. Edgar R, Domrachev M, Lash AE. Gene Expression Omnibus: NCBI gene expression and hybridization array data repository. Nucleic Acids Res. 2002;30(1):207-210. Dann SG, Selvaraj A, Thomas G. mTOR Complex1-S6K1 signaling: at the crossroads of obesity, diabetes and cancer. Trends Mol Med. 2007;13(6):252-259. Dao MA, Pepper KA, Nolta JA. Long-term cytokine production from engineered primary human stromal cells influences human hematopoiesis in an in vivo xenograft model. Stem Cells. 1997;15(6):443-454. Lee EB, Kim KW, Hong JY, Jee HM, Sohn MH, Kim KE. Increased serum thymic stromal lymphopoietin in children with atopic dermatitis. Pediatr Allergy Immunol. 2010;21(2 Pt 2):e457-460. Graf L, Iwata M, Torok-Storb B. Gene expression profiling of the functionally dis-

39.

40.

41.

42.

43.

44. 45.

46.

47.

48.

49.

tinct human bone marrow stromal cell lines HS-5 and HS-27a. Blood. 2002;100(4):15091511. Mootha VK, Lindgren CM, Eriksson KF, et al. PGC-1alpha-responsive genes involved in oxidative phosphorylation are coordinately downregulated in human diabetes. Nat Genet. 2003;34(3):267-273. Subramanian A, Tamayo P, Mootha VK, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci USA. 2005;102(43):15545-15550. Parrish YK, Baez I, Milford TA, et al. IL-7 Dependence in human B lymphopoiesis increases during progression of ontogeny from cord blood to bone marrow. J Immunol. 2009;182(7):4255-4266. Johnson SE, Shah N, Panoskaltsis-Mortari A, LeBien TW. Murine and human IL-7 activate STAT5 and induce proliferation of normal human pro-B cells. J Immunol. 2005;175(11):7325-7331. Ziegler SF, Roan F, Bell BD, Stoklasek TA, Kitajima M, Han H. The biology of thymic stromal lymphopoietin (TSLP). Adv Pharmacol. 2013;66:129-155. Takai T. TSLP expression: cellular sources, triggers, and regulatory mechanisms. Allergol Int. 2012;61(1):3-17. Allakhverdi Z, Comeau MR, Armant M, et al. Mast Cell-Activated Bone Marrow Mesenchymal Stromal Cells Regulate Proliferation and Lineage Commitment of CD34(+) Progenitor Cells. Front Immunol. 2013;4(461. Xu H, Yang W, Perez-Andreu V, et al. Novel susceptibility variants at 10p12.31-12.2 for childhood acute lymphoblastic leukemia in ethnically diverse populations. J Natl Cancer Inst. 2013;105(10):733-742. Liem NL, Papa RA, Milross CG, et al. Characterization of childhood acute lymphoblastic leukemia xenograft models for the preclinical evaluation of new therapies. Blood. 2004;103(10):3905-3914. Lock RB, Liem NL, Papa RA. Preclinical testing of antileukemic drugs using an in vivo model of systemic disease. Methods Mol Med. 2005;111:323-334. Jiang BH, Liu LZ. Role of mTOR in anticancer drug resistance: perspectives for improved drug treatment. Drug Resist Updat. 2008;11(3):63-76.

haematologica | 2016; 101(4)


ARTICLE

Platelet Biology & Its Disorders

Coated platelets function in platelet-dependent fibrin formation via integrin αIIbb3 and transglutaminase factor XIII

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Nadine J.A. Mattheij,1 Frauke Swieringa,1* Tom G. Mastenbroek,1* Michelle A. Berny-Lang,2 Frauke May,3 Constance C.F.M.J. Baaten,1 Paola E.J. van der Meijden,1 Yvonne M.C. Henskens,4 Erik A.M. Beckers,5 Dennis P.L. Suylen,1 Marc W. Nolte,3 Tilman M. Hackeng,1 Owen J.T. McCarty,2 Johan W.M. Heemskerk,1 and Judith M.E.M. Cosemans1

Department of Biochemistry, Cardiovascular Research Institute Maastricht (CARIM), Maastricht University, The Netherlands; 2Department of Biomedical Engineering, Oregon Health & Science University, Portland, OR, USA; 3CSL Behring GmbH, Marburg, Germany; 4Central Diagnostic Laboratory, Maastricht University Medical Center, The Netherlands; and 5Department of Internal Medicine, Maastricht University Medical Center, The Netherlands

1

*FS and TGM contributed equally to this work

Haematologica 2016 Volume 101(4):427-436

ABSTRACT

C

oated platelets, formed by collagen and thrombin activation, have been characterized in different ways: i) by the formation of a protein coat of α-granular proteins; ii) by exposure of procoagulant phosphatidylserine; or iii) by high fibrinogen binding. Yet, their functional role has remained unclear. Here we used a novel transglutaminase probe, Rhod-A14, to identify a subpopulation of platelets with a cross-linked protein coat, and compared this with other platelet subpopulations using a panel of functional assays. Platelet stimulation with convulxin/thrombin resulted in initial integrin αIIbb3 activation, the appearance of a platelet population with high fibrinogen binding, (independently of active integrins, but dependent on the presence of thrombin) followed by phosphatidylserine exposure and binding of coagulation factors Va and Xa. A subpopulation of phosphatidylserine-exposing platelets bound Rhod-A14 both in suspension and in thrombi generated on a collagen surface. In suspension, high fibrinogen and Rhod-A14 binding were antagonized by combined inhibition of transglutaminase activity and integrin αIIbb3. Markedly, in thrombi from mice deficient in transglutaminase factor XIII, platelet-driven fibrin formation and RhodA14 binding were abolished by blockage of integrin αIIbb3. Vice versa, star-like fibrin formation from platelets of a patient with deficiency in αIIbb3 (Glanzmann thrombasthenia) was abolished upon blockage of transglutaminase activity. We conclude that coated platelets, with initial αIIbb3 activation and high fibrinogen binding, form a subpopulation of phosphatidylserine-exposing platelets, and function in platelet-dependent star-like fibrin fiber formation via transglutaminase factor XIII and integrin αIIbb3.

Introduction Platelet activation and blood coagulation are highly reciprocally interacting processes and both are essential for hemostasis and thrombosis. Activated platelets support and steer the coagulation process by at least four mechanisms: i) by releasing coagulation factors like factor V and XIII; ii) by exposing the procoagulant phospholipid phosphatidylserine (PS) at their outer surface to support thrombin generation; iii) by providing a scaffold for the formation of fibrin fibers; and iv) by causing haematologica | 2016; 101(4)

Correspondence: judith.cosemans@maastrichtuniversity.nl

Received: May 29, 2015. Accepted: December 23, 2015. Pre-published: December 31, 2015. doi:10.3324/haematol.2015.131441

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/4/427

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

427


N.J.A. Mattheij et al.

retraction of the fibrin clot.1,2 In a growing thrombus, aggregated and procoagulant platelets form two distinct populations,3,4 which is at least partly explained by the high Ca2+ response required for PS exposure and coagulation factor binding, and by the calpain-dependent closure of active αIIbb3 integrins after PS exposure, thus antagonizing inclusion of procoagulant platelets into a platelet aggregate.5,6 However, another platelet population has also been identified, usually referred to as coated platelets,7 which may partly overlap with the two other platelet populations described above.3 In the initial paper, Dale et al. describe COAT platelets (later renamed as coated platelets) as a population of platelets arising after combined stimulation with collagen and thrombin, which bind α-granule proteins, including factor V, fibrinogen, von Willebrand factor, thrombospondin, fibronectin and α2-antiplasmin, in a transglutaminase-dependent way via the formation of covalent serotonin conjugations.8 Since this first description, coated platelets have been invariably considered as platelets formed after combined stimulation of collagen receptors (e.g. with collagen, convulxin or collagen-related peptide) and thrombin receptors (e.g. with thrombin or thrombin receptor-activating peptides), but there is no uniform definition of this platelet population in the literature. The Dale group has been using the retention of secreted proteins, including platelet-

A

C

derived serotonin-derivatized proteins, factor V and tissue factor pathway inhibitor on the platelet surface as a characteristic.9-11 Another definition has been used by the Jobe group, i.e. platelets containing high surface levels of fibrinogen, likely through cross-linking via the transglutaminase factor XIII.12 However, in recent years, it has become common practice to consider coated platelets more or less equivalent to fibrinogen binding platelets or PS-exposing platelets. For example, platelet subpopulations in patient studies have recently been characterized using biotin-fibrinogen.13-15 This ambiguity in definition and described properties raises questions as to whether coated platelets form a platelet subpopulation (after collagen/thrombin receptor stimulation) that is distinguishable from that of fibrinogen and/or PS-exposing platelets and whether they fulfill a specific function. In the present paper, we used a specific transglutaminase substrate, i.e. the α2-antiplasmin-derived peptide RhodA14, as a tool to identify transglutaminase-active platelets. We compared the binding of Rhod-A14 to platelets, stimulated via the collagen and thrombin receptors, with other platelet activation markers. The results indicate that transglutaminase active platelets form as a subpopulation of PS exposing platelets. We also provide evidence that the transglutaminase activity along with integrin αIIbb3 activation is required for fibrin anchoring at the platelet surface and starlike platelet-dependent fibrin formation.

B

D

E

Figure 1. Transglutaminase activity confined to a subpopulation of phospholipid phosphatidylserine (PS)-exposing platelets. Washed human platelets (5x107/mL) were stimulated with convulxin (Cvx, 100 ng/mL) plus thrombin (Thr, 4 nM) in the presence of 0.2 mM GPRP and 2 mM CaCl2. Samples, taken at given time points, were labeled for 5 min with Rhod-A14 and AF647-annexin A5. Platelets were pre-treated with K9-DON (100 mM), where indicated. (A) Representative histograms obtained by flow cytometry. (B) Platelet subpopulations staining with Rhod-A14 (white bars) or AF647-annexin A5 (black bars) after 60-min stimulation with Cvx (100 ng/mL), Thr (4 nM), SFLLRN (15 mM) and/or ionomycin (10 mM), as indicated. (C-E) Effects of factor XIII and transglutaminase inhibitors on platelet subpopulations staining with Rhod-A14 (C) or AF647-annexin A5 (E) after stimulation with Cvx (100 ng/mL) plus Thr (4 nM). Platelets were pre-incubated with vehicle (○, solid line), 100 unit/mlLfactor XIII (●, dotted line), 100 mM K9-DON (■, dotted line), or 200 mM Boc-DON (♦, dotted line). Solid gray line, unstimulated platelets. (D) Dose-dependent inhibition of Rhod-A14 binding (60 min) by K9-DON or Boc-DON. Mean ± SEM (n=3-6); *P<0.05.

428

haematologica | 2016; 101(4)


Coated platelets in star-like fibrin formation

Methods Blood collection and platelet preparation Experiments were approved by the local Medical Ethics Committees. Blood was taken from healthy volunteers and from a patient with Glanzmann thrombasthenia, with established deficiency in integrin αIIbb3,16 after informed consent and in accordance with the Declaration of Helsinki. Platelet-rich plasma (PRP), defibrinated platelet-free plasma (PFP) and washed platelets were prepared from whole blood as described in the Online Supplementary Appendix. Animal studies were approved by the local animal experimental committees. Mice deficient in factor XIII A1 subunit (F13a1tm1Gdi, abbreviated as F13a1–/–)17 were bred on a mixed 129Sv/CBA background and were compared to F13a1+/+ mice of the same background (Harlan Laboratories). Murine blood was taken on trisodium citrate for whole-blood flow experiments; other blood samples were taken on acid-citrate-dextrose anticoagulant to isolate washed platelets, as previously described.18

Flow cytometric platelet analyses Washed human or mouse platelets (5x107/mL) were pre-incubated with indicated inhibitors or Me2SO vehicle for 10 min, and stimulated in the presence of 2 mM CaCl2. In the activations, 0.2 mM Gly-Pro-Arg-Pro (GPRP) was added to prevent formation of large fibrin fibers.19 Platelet sub-populations were distinguished by probing with Rhod-A14 (10 mg/mL), AF647-annexin A5 (1:200), AF488-factor V (20 nM), OG488-factor Xa (100 nM), AF647-fibrinogen (100 mg/mL) and FITC-PAC-1 (1.25 mg/mL). After staining for 5 min, samples were analyzed with a FACScan flow cytometer (BD Accuri Cytometer).6 In a separate set of experiments, reconstituted PRP was activated with tissue factor (2 pM) and CaCl2 (16.7 mM) in the presence of GPRP (2 mM) at 37°C, after which samples were taken for fluorescent labeling. Analysis was by flow cytometry as described above.

Thrombin generation Thrombin generation was measured in citrate-anticoagulated

A

C

haematologica | 2016; 101(4)

human PRP as previously described.20 First-derivative curves were converted into curves of nanomolar thrombin concentrations using a calibrator for human α-thrombin.21 All analyses were in triplicate.

Thrombus formation on collagen under flow Whole blood thrombus formation on collagen was assayed in the absence or presence of coagulation, as previously described.22 Thrombi formed on collagen were then stained with Rhod-A14 and AF647-annexin A5 in Hepes buffer, supplemented with CaCl2 (2 mM) and heparin (1 unit/mL). Phase-contrast and fluorescence images were captured and analyzed with Metamorph software v.7.5.0.0 (MDS Analytical Technologies), as detailed elsewhere.23 Statistical difference of quantitative colocalization of 2-color confocal images was determined using the ZEN software 2010 B SP1. Star-like fibrin formation was assessed as described in the Online Supplementary Appendix.

Statistical analysis Significance of differences between control and experimental groups, as well as changes between groups over time, was determined by 1-way or 2-way analysis of variance followed by a Bonferroni post hoc test. Distribution of fibrin fibers was evaluated by χ2 analysis.24 P<0.05 was considered significant. Further details on the materials and methods used are available in the Online Supplementary Appendix.

Results Coated platelets as transglutaminase active platelets First we assessed the appearance of a coated platelet population following the initial description of formation of a transglutaminase-dependent protein coat on the platelet surface.8 Therefore, we synthesized a fluorescentlabeled 14-amino acid transglutaminase peptide substrate, GNQEQVSPLTLLK(C-tetramethylrhodamine)W (RhodA14), derived from the N-terminal peptide sequence of

B

D

Figure 2. Transglutaminase-positive platelets are not identical to procoagulant platelets. (A and B) Washed platelets were stimulated with convulxin plus thrombin, as indicated for Figure 1. At given time points, samples were taken and labeled with AF647-fibrinogen (○), OG488-factor Xa (▲), AF647-annexin A5 ( ●), AF488-factor V(a) (■), or Rhod-A14 (♦). (A) Fractions of platelets binding the indicated probes. (B) Effect of platelet treatment with K9-DON (100 mM) (gray bars) on fractions of positive platelets. (C and D) Citrate-anticoagulated PRP (1x108 platelets/mL) was pre-incubated with vehicle or K9-DON (100 mM), and used for measurement of thrombin generation; coagulation was triggered with tissue factor/CaCl2. Representative thrombin generation curves (C) and quantification of peak height (D). Mean ± SEM (n=3-4), *P<0.05.

429


N.J.A. Mattheij et al. α2-antiplasmin. This peptide contains a glutamine at Q3, which serves as a transglutamination site for cross-linking to available ε-amino lysine residues.25 The Rhod-A14 peptide was site-specifically labeled with a tetramethylrhodamine group, which did not interfere with its substrate properties. Control experiments indicated that the peptide could act as a substrate for both tissue-type transglutaminase and factor XIIIa (data not shown, but see below). Flow cytometric analysis indicated that the Rhod-A14 label bound to a population of convulxin/thrombin-activated platelets, which was gradually formed after approximately 15 min of activation and increased until 60 min to approximately 50% Rhod-A14 positive platelets (Figure 1A and B). This differed markedly from the appearance of PS-exposing platelets analyzed by annexin A5 binding, which was already maximal after 15 min. Two-color flow cytometry indicated that the Rhod-A14 positive platelets represented a discrete subpopulation: the majority of these platelets were PS positive (binding of annexin A5), but had inactive αIIbb3 integrins (no binding of PAC-1 mAb) (Online Supplementary Figure S1). When comparing various strong platelet agonists, we found that the Ca2+-ionophore ionomycin, while causing nearly complete PS exposure, resulted in Rhod-A14 labeling of a subpopulation of only approximately 25% platelets (Figure 1B). Dual platelet stimulation with convulxin and the thrombin receptor activating peptide SFLLRN resulted in a similar Rhod-A14 positive fraction, whereas stimulation with thrombin alone did not lead to either Rhod-A14 incorporation or to PS exposure. To specifically test the contribution of the transglutaminase factor XIII in Rhod-A14 binding to platelets, we analyzed platelets from F13a1–/– mice, which are deficient in the A-subunit of factor XIII. Washed platelets from

F13a1–/– mice, when stimulated with convulxin/thrombin, showed greatly reduced Rhod-A14 binding in comparison to wild-type platelets (Online Supplementary Figure S2A). However, PS exposure, factor Xa, and fibrin(ogen) binding were unchanged in the F13a1–/– platelets (Online Supplementary Figure S2B-D). Vice versa, pre-treatment of human washed platelets with purified factor XIII prior to stimulation with convulxin/thrombin resulted in an enhanced appearance of the Rhod-A14 binding cells (Figure 1C). Additional control experiments again confirmed the involvement of transglutaminase activity, since Rhod-A14 binding to washed human platelets was antagonized in a dose-dependent way by the general transglutaminase inhibitor K9-DON and the tissue transglutaminase inhibitor Boc-DON (Figure 1D). In sharp contrast, neither the addition of factor XIII nor these inhibitors affected the PS exposure evoked by convulxin/thrombin (Figure 1E). Taken together, these results indicate that the transglutaminase-active Rhod-A14 positive platelets form as a subpopulation of PS-exposing platelets after collagen/thrombin receptor stimulation.

Coated platelets as high fibrinogen-binding platelets Platelet activation with convulxin/thrombin has been demonstrated to result in a fraction of platelets displaying high fibrinogen binding, which has also been referred to as coated platelets.12 In the present study, after 5 min of stimulation with convulxin/thrombin, the binding of fluorescently labeled fibrinogen reached a maximal high level, and this occurred faster than the appearance of PS-exposing platelets (Online Supplementary Figure S3A-C). Other strong platelet agonists, ionomycin and convulxin/ SFLLRN induced only low fibrinogen binding. However, thrombin alone induced similarly high fibrinogen binding

A

C

B

D

Figure 3. High fibrinogen binding of platelets is independent of αIIbb3 activation and phospholipid phosphatidylserine (PS) exposure in tissue factor-activated plasma. Defibrinated PFP was reconstituted with platelets and triggered with tissue factor (2 pM) and CaCl2 (16.7 mM). Samples were taken for fluorescent labeling at indicated time points. (A and B) Populations of platelets with high binding of OG488-fibrinogen and binding AF647-annexin A5. Total fractions (A) and sub-fractions (B) of high-fibrinogen and annexin A5 binding platelets. (C and D) Populations of platelets binding FITC-PAC-1 mAb and AF647-annexin A5. Total fractions (C) and subfractions (D) of platelets positive for FITC-PAC-1 mAb and AF647-annexin A5. Mean ± SEM (n=3).

430

haematologica | 2016; 101(4)


Coated platelets in star-like fibrin formation

as in combination with convulxin. This suggests that the high fibrinogen binding relies on the presence of thrombin, but not on PS exposure. Notably, the transglutaminase inhibitors K9-DON or Boc-DON did not affect either fibrinogen binding or PS exposure (Online Supplementary Figure S3A and D). Interestingly, the fibrinogen binding per platelet was slightly reduced upon prolonged incubation with convulxin/thrombin (Online Supplementary Figure S3B), while the fraction of fibrinogen-positive platelets remained stable (Figure 2A). This suggests that a portion of ÎąIIbb3 integrins on the surface Cvx/Thr activated platelets can close in time, which is in agreement with labeling studies using PAC-1 mAb (Online Supplementary Figure S1D). Dual-color flow cytometry indicated that Rhod-A14 binding platelet fraction was a subpopulation of the high fibrinogen-binding cells (Online Supplementary Figure S1C).

Coated platelets as procoagulant PS-exposing platelets In recent papers, it has been proposed that coated platelets are equivalent to coagulation factor binding platelets or PS-exposing platelets.7,26 To investigate this in more detail, we compared the fractions of convulxin/thrombin-stimulated platelets that bind AF647annexin A5, OG488-factor Xa, FITC-factor V or RhodA14. The appearance of a platelet population with binding sites for annexin A5, factor V and factor Xa occurred

quickly with similar kinetics, whereas the formation of Rhod-A14-binding platelets was considerably slower (Figure 2A). Pre-incubation with transglutaminase inhibitor K9-DON did not influence the binding of annexin A5 nor the coagulation factors, but greatly impaired the binding of Rhod-A14 (Figure 2B). In agreement with this result, the presence of K9-DON did not influence the process of thrombin generation in PRP (Figure 2C and D). This indicated that the transglutaminase activity of platelets was independent of the interaction of factor Va and Xa, and of subsequent plateletdependent thrombin generation, which relies on this interaction.20

Coated platelets as fibrin-forming platelets Our earlier work had demonstrated that platelet stimulation with strong agonists leads to integrin ÎąIIbb3 activation, followed by PS exposure and calpain-dependent closure of the integrin.6 Furthermore, we found that PS exposure instigates the formation of a fibrin network linked to the platelet surface.27 In the present study, we investigated this again by assessing the determinants of high fibrin(ogen) binding and PS exposure of platelets in tissue factor-activated blood plasma. For this purpose, plasma samples were defibrinated to prevent massive clot formation,19 and then reconstituted with washed platelets. After triggering coagulation with tissue factor and calcium, the

A

B

Figure 4. Subpopulation of phospholipid phosphatidylserine (PS)-exposing platelets in thrombi is transglutaminase positive. Human whole blood was perfused over collagen under anti-coagulated (A) or coagulated (B) conditions at shear rate of 1000 s-1. Blood samples were pre-incubated with vehicle (control) or K9-DON (100 mM). (A and B) Representative microscopic images of differential interference contrast (DIC), and confocal images of AF647-annexin A5 or Rhod-A14 fluorescence (bars, 20 mm). Right bar graphs indicate co-localization coefficient of AF647-annexin A5 with Rhod-A14 fluorescence and of Rhod-A14 with AF647-annexin A5 under vehicle-treated (non)coagulant conditions, as assessed with Zen software. Mean Âą SEM (n=4).

haematologica | 2016; 101(4)

431


N.J.A. Mattheij et al.

majority of platelets showed rapid and sustained high fibrinogen binding, followed by the appearance of a platelet subpopulation with PS exposure, as determined by duallabeling with OG488-fibrinogen and AF647-annexin A5 (Figure 3A). Detailed analysis indicated that only a proportion of the high fibrinogen-binding platelets exposed PS, whereas the majority of PS-exposing platelets showed high fibrinogen binding (Figure 3B). Once again, the high fibrin(ogen) binding was dependent on the presence of thrombin, as these platelets stained with an anti-fibrin Ab and the thrombin inhibitor hirudin substantially suppressed fibrinogen binding (data not shown). High fibrin(ogen) binding of a subpopulation of the PSexposing cells was observed not only with labeled fibrinogen, but also with an antibody (FITC-WAK) recognizing platelet-bound fibrin(ogen) (Online Supplementary Figure S4A and B). This is in agreement with the results with washed platelets described above that showed that, in the presence of thrombin, the high fibrin(ogen) binding is regulated differently from PS exposure. Staining of the tissue factor-triggered platelets in plasma with FITC-PAC-1 mAb, recognizing activated integrin αIIbb3, pointed to rapid integrin activation that subsequently declined (Figure 3C). However, when these platelets were stained for the integrin b3-chain, fluorescence remained high (data not shown), indicating that the integrin was still present and accessible at the platelet surface. In comparison to integrin activation, exposure of PS rose more slowly over time (Figure 3C). Similar to our findings as reported before,6 the majority of PAC-1 mAb positive platelets were annexin A5 negative, while the annexin A5 positive platelets were mostly PAC-1 mAb negative (Figure 3D). Dot plots of platelets in plasma, double labeled with AF647-annexin A5 and FITC-PAC-1 mAb,

A

B

confirmed that most PS-exposing cells were unable to bind PAC-1 mAb (Online Supplementary Figure S4C). In this plasmatic condition, a population of PS-positive platelets showed a mean 2.8-fold increase in Rhod-A14 fluorescence above background. Taken together, these results suggest that the high fibrin(ogen)-binding platelet fraction contains a fibrin coat, which precedes PS exposure and primes for binding or Rhod-A14.

Transglutaminase activity in platelet thrombus formation under flow conditions To investigate where coated platelets localize in platelet thrombi, we performed whole blood perfusion experiments over a collagen surface under arterial shear conditions. Double staining with Rhod-A14 and AF657-annexin A5 indicated that, under non-coagulant conditions (with thrombin inhibitor present), Rhod-A14 binding was restricted to a subpopulation of PS-exposing platelets (Figure 4A). Co-localization analysis demonstrated that the majority of Rhod-A14 positive pixels also stained with AF647-annexin A5. Addition of K9-DON completely blocked the Rhod-A14 staining, but did not affect PS exposure. Zoomed images indicated that the Rhod-A14 labeling was confined to balloon-shaped, PS-exposing platelets, but often did not label the complete platelet (Figure 4A, overlay and bar graph). When thrombi were allowed to form on collagen under coagulant conditions (recalcification without thrombin inhibitor), this resulted in increased Rhod-A14 staining, which again was antagonized by K9DON. Also a higher AF657-annexin A5 staining was found, which was unaffected by K9-DON (Figure 4B). Under these coagulant conditions, PS-exposing balloonshaped platelets still bound Rhod-A14, while the majority of the Rhod-A14 label was located on fibrin fibers.

C

D

Figure 5. Absence of factor XIII results in fibrin formation without transglutaminase activity. Whole blood from F13a1+/+ or F13a1–/– mice was perfused over collagen under non-coagulant (A and B) or coagulant (C and D) conditions during 4 min at shear rate of 1000 s-1. Blood samples were pre-incubated with vehicle (control) or K9-DON (100 mM), and post-stained with Rhod-A14. Microscopic images of brightfield and Rhod-A14 fluorescence are shown (bars, 20 mm); also quantification of thrombus size [(% of surface area coverage, (SAC)] and Rhod-A14 fluorescence. Mean ± SEM (n=4); *P<0.05.

432

haematologica | 2016; 101(4)


Coated platelets in star-like fibrin formation

Next, we performed flow experiments with F13a1+/+ or F13a1–/– mouse blood over collagen under both non-coagulant and coagulant conditions. In the non-coagulant condition, citrate anti-coagulated whole blood was recalcified in the presence of PPACK to block thrombin generation while physiological calcium and magnesium levels were restored. Under these conditions factor XIII activation is averted (no thrombin), whereas tissue type transglutaminase can be active (physiological calcium levels). The thrombi obtained with blood from F13a1+/+ or F13a1–/– mice were similar in size under these conditions with no more than low Rhod-A14 labeling (Figure 5A and B). These data suggest no major role for tissue type transglutaminase in thrombus formation and Rhod-A14 binding in the absence of coagulation, which is in line with the low expression of tissue-type transglutaminase in mouse platelets.28 Under coagulant conditions, large fibrin-containing thrombi were formed with both F13a1+/+ and F13a1–/– blood (Figure 5C and D). However, Rhod-A14 incorporated only into the fibrin fibers of wild-type thrombi and not of factor XIII-deficient thrombi. The Rhod-A14 staining again was antagonized by K9-DON, but not by tirofiban. These results hence indicate that the process of fibrin fiber formation can occur in the absence of transglutaminase activity.

Synergistic roles of transglutaminase and integrin αIIbb3 in platelet fibrin formation under flow conditions

We hypothesized that transglutaminase and integrin αIIbb3 could be involved in the high fibrin(ogen) binding to coated platelets. Given this, we investigated the contribution of both of these in human and mouse systems. We used inhibitors against transglutaminases (K9-DON) and αIIbb3 integrin (tirofiban) in concert with blood from F13a1–/– mice and a Glanzmann thrombasthenia patient with established deficiency in αIIbb3. In washed platelets from control subjects stimulated with convulxin/thrombin, the combined presence of the K9-DON and tirofiban inhibited the binding of fibrin(ogen) and Rhod-A14 (P<0.05) (Figure 6A and B). A similar reduction of RhodA14 binding was observed when pre-treating Glanzmann platelets with K9-DON (P<0.05) (data not shown). To investigate the formation of fibrin fibers, blood from F13a1+/+ and F13a1–/– mice was pre-labeled with AF647-fibrinogen and perfused over a collagen surface under coagu-

A

lating conditions. Detailed examination by confocal fluorescence microscopy indicated that, for wild-type thrombi, fibrin fibers were connected to the platelet surface and were oriented in all directions (Figure 7A and B). This staining pattern resembled the star-like formation of fibrin at the platelet surface that had been previously observed.27 In wild-type thrombi treated with the integrin blocker tirofiban, and also in factor XIII-deficient thrombi, the orientation of fibrin fibers from platelets was typically more frequent in the direction of the blood flow (Figure 7A and B). A similar shift in orientation of fibrin fibers was observed on human thrombi formed with whole blood from a Glanzmann thrombasthenia patient (Online Supplementary Figure S5). The addition of tirofiban to factor XIII-deficient blood markedly and nearly completely blocked the star-like orientation of fibrin fibers, but still allowed the plasmatic formation of fibrin fibers rectilinear with the blood flow. In subsequent experiments, platelet-dependent star-like fibrin formation was examined on the level of single platelets by using a previously established method.27 Whereas on the surface of control platelets a star-like fibrin network was formed upon post-perfusion with normal pool plasma, the orientation of the fibrin fibers was significantly altered upon post-perfusion with factor XIII-deficient plasma (Figure 7C and D). A similar shift in orientation of fibrin fibers was observed on the surface of Glanzmann thrombasthenia platelets. Post-perfusion of Glanzmann thrombasthenia platelets with factor XIIIdeficient plasma completely blocked fibrin formation during more than 20 min. However, rectilinear plasmatic fibrin formation did occur after approximately 30 min of plasma perfusion, but no platelet-dependent fibrin formation could be observed. A similar absence of plateletdependent fibrin formation was seen upon addition of K9DON to Glanzmann thrombasthenia platelets (P<0.001). Taken together, these data indicate a non-redundant and synergistic role of the transglutaminase factor XIII and integrin αIIbb3 in platelet-dependent fibrin formation.

Discussion The original paper by Dale et al. describes COAT (later renamed coated) platelets as a population of platelets that result from combined stimulation with collagen and

B

Figure 6. High fibrinogen binding of platelets requires both transglutaminase and α IIbb3 activities. Washed platelets were stimulated with convulxin (100 ng/mL) and thrombin (4 nM) (see Figure 1). Pre-incubation was with vehicle, K9-DON (100 mM) and/or tirofiban (5 mg/mL). Flow cytometric data were assessed in the presence of Rhod-A14 and AF647-fibrinogen. Percentages of mean fluorescence intensity of fibrinogen binding (A) and platelets positive for Rhod-A14 are shown (B). Mean ± SEM (n=3); *P<0.05 versus vehicle.

haematologica | 2016; 101(4)

433


N.J.A. Mattheij et al. thrombin, and that can bind secreted α-granule proteins in a transglutaminase-dependent way.8 In the present paper, we used a novel fluorescent-labeled transglutaminase peptide substrate, Rhod-A14, to identify transglutaminaseactive platelets and characterize the formation and functional responses of these platelets. In washed human platelets, stimulated by convulxin/thrombin, we identified a slowly formed (1-h) subpopulation of PS-exposing platelets capable of binding Rhod-A14 by flow cytometry. The general transglutaminase inhibitor K9-DON and the tissue-type transglutaminase inhibitor Boc-DON prevented Rhod-A14 incorporation, whereas the transglutaminase factor XIII enhanced incorporation. Rhod-A14-positive platelets appeared as a discrete population of platelets with high fibrin(ogen) binding and PS exposure, while only a subpopulation of

PS-positive and high fibrin(ogen)-binding platelets bound Rhod-A14. We further found that PS exposure, but not Rhod-A14 binding, was accompanied by the binding of coagulation factors Va and Xa to platelets. This is in agreement with results from other groups, also reporting high similarity of the annexin A5- and the coagulation factorbinding platelet populations.29,30 Furthermore, both in time of appearance and in subpopulation, the high fibrin(ogen)binding platelets differed from the Rhod-A14-binding platelets. Hence, in contrast to others,13,14 we consider (transglutaminase-active) coated platelets as non-identical to fibrinogen-binding platelets. Whereas convulxin/thrombin stimulated washed platelets showed persistently high fibrin(ogen) binding and slowly increasing RhodA14 binding, PAC-1 mAb binding gradually decreased after an initial peak. This is

A

B

C

D

Figure 7. Absence of factor XIII and blockage/absence of integrin αIIbb3 impairs platelet-derived fibrin formation. (A and C) Blood from F13a1+/+ or F13a1–/– mice was perfused over collagen under coagulant conditions (4 min, shear rate 1000 s-1). Tirofiban (5 µg/mL) was added where indicated. (A) Confocal images of AF647fibrinogen fluorescence (bars, 20 mm). (C) Bar graph with distribution of number of fibrin fibers per region (region 1=0-45°, 2=45-90°, 3=90-135°, 4=135-180°, 5=180-225°, 6=225-270°, 7=270-315°, 8=315-360°), indicating the orientation of the formed fibrin fibers relative to the flow direction. The orientation of fibrin fibers was determined from thrombi traversing the arbitrarily drawn dotted white line (n=4); *P<0.001. (B and D) Washed platelets from a control subject or a patient with Glanzmann thrombasthenia were allowed to adhere to a glass coverslip, blocked and then post-perfused under coagulating conditions with normal pool plasma (NPP) or FXIII-deficient plasma (FXIII-DP). (B) Representative phase contrast images taken during the first minute of fibrin formation; bar is 5 mm. Note that no fibrin formation was observed with Glanzmann platelets during 20 min of post-perfusion with FXIII-DP. (D) Bar graph with distribution of number of fibrin fibers per region relative to the flow direction (region 1=0-45°, 2=45-90°, 3=90-135°, 4=135-180°, 5=180-225°, 6=225-270°, 7=270-315°, 8=315-360°), indicating the orientation of the formed fibrin fibers on the level of individual platelets. Independent experiments for control (n=4) and Glanzmann (n=1, triplicate); *P<0.001.

434

haematologica | 2016; 101(4)


Coated platelets in star-like fibrin formation indicative of secondary inactivation of integrin αIIbb3, which is in agreement with earlier results.6,31 One explanation for the persistent fibrin(ogen) binding to platelets with apparently inactive integrins, is that the activated αIIbb3 displays distinct binding sites for fibrinogen and fibrin, allowing interaction with fibrin, independently of the classical integrin activation domains.32 However, in the present study, we observed that the high fibrin(ogen) binding of washed platelets was antagonized by combined inhibition of αIIbb3 and transglutaminases. This suggests initial binding of fibrinogen to the activated integrin, followed by secondary binding of a growing fibrin fiber to platelet surface proteins via transglutaminase activity. In contrast to washed platelets, Rhod-A14 binding platelets were formed much faster (minutes) during human whole-blood thrombus formation on collagen. This can be explained by a rapid availability of transglutaminases like human factor XIII in the blood plasma, but it may also point to a relatively slow externalization and/or activation of platelet-derived transglutaminases in the absence of plasma. Given that transglutaminases are broad-spectrum enzymes, cross-linking ε-(g-glutamyl)lysine isopeptide bonds of multiple proteins,33,34 it can safely be assumed that Rhod-A14binding platelets are covered (coated) with a large set of cross-linked proteins at their surface. In our previous work, we had shown that platelets act as a scaffold for star-like fibrin formation.27 Here we provide evidence of a combined role of integrin αIIbb3 and factor XIII in this process. In agreement with the results from others,12 we find that murine factor XIII is not essential for high fibrinogen binding and PS exposure. Although platelet-dependent fibrin fiber formation could still occur in blood from F13a1–/– mice (which points to the availability of factor XIII for the formation of fibrin fibers as such), we did observe a significant change in the attachment of fibrin fibers formed on the thrombus surface. Additional blockage of integrin αIIbb3 led to a further shift from platelet-dependent star-like fibrin formation to plateletindependent fiber orientation rectilinear to the blood flow. In the human system, the star-like fibrin fiber formation on the platelet surface was also synergistically mediated by integrin αIIbb3 activation (Glanzmann thrombasthenia

References 7. 1. Cosemans JM, Angelillo-Scherrer A, Mattheij NJ, Heemskerk JW. The effects of arterial flow on platelet activation, thrombus growth, and stabilization. Cardiovasc Res. 2013;99(2):342-352. 2. de Witt SM, Swieringa F, Cavill R, et al. Identification of platelet function defects by multi-parameter assessment of thrombus formation. Nat Commun. 2014;5:4257. 3. Munnix IC, Cosemans JM, Auger JM, Heemskerk JW. Platelet response heterogeneity in thrombus formation. Thromb Haemost. 2009;102(6):1149-1156. 4. Versteeg HH, Heemskerk JW, Levi M, Reitsma PH. New fundamentals in hemostasis. Physiol Rev. 2013;93(1):327-358. 5. Heemskerk JW, Mattheij NJ, Cosemans JM. Platelet-based coagulation: different populations, different functions. J Thromb Haemost. 2013;11(1):2-16. 6. Mattheij NJ, Gilio K, van Kruchten R, et al. Dual mechanism of integrin αIIbb3 closure

haematologica | 2016; 101(4)

8.

9. 10.

11.

12. 13.

patient) and factor XIII activity. This functional redundancy of integrin αIIbb3 and factor XIII might explain the poor correlation of the defect in αIIbb3 with the bleeding tendency observed in Glanzmann thrombasthenia patients.35 Although earlier studies found tissue type transglutaminase to be present in human platelets,8,36 more recent studies showed it could not be detected on the platelet protein level.37,38 Interestingly, in mouse platelets, tissue type transglutaminase (Tgm2) could be detected, albeit with a 25-fold lower expression in comparison to factor XIII (F13a).28 Our present findings point to a predominant role of factor XIII, and not of tissue type transglutaminase, in both the human and murine systems, which is in line with the recent studies. A role for platelet-derived FXIII was recently proposed in regulating anti-fibrinolysis by crosslinking α2-antiplasmin to fibrin.38 Our present findings significantly extend this study by demonstrating a role for factor XIII earlier in the process of hemostasis and thrombosis, i.e. by mediating the formation of a star-like coat of fibrin fibers on the platelet surface. In summary, our study shows that the mechanism for formation of coated Rhod-A14-binding platelets is that initial collagen/thrombin receptor-induced αIIbb3 activation leads to fibrin(ogen) binding, which on PS-exposing platelets can be consolidated by factor XIII transglutaminase activity, and act as a scaffold for star-like platelet-driven fibrin fiber formation. Since many of these coated platelets still bind coagulation factors, they could also have a role in stimulating prothrombinase activity and thrombin generation, leading to stabilization of the platelet-fibrin clot. Funding This work was supported by research funding from ZonMW MKMD (114021004) to JWMH and JMEMC; the Dutch Heart Foundation (2011T6) to JMEMC; the Dutch Landsteiner Foundation for Blood Transfusion Research (1006) to FS, PEJvdM and JWMH; the Cardiovascular Center Maastricht to PEJvdM; the National Institutes of Health (R01HL101972) to OJTM; the American Heart Association (13EIA12630000) to OJTM; the Whitaker International Fellows Program (MAB-L); and Cyttron II (LSH framework: FES0908) to TMH.

in procoagulant platelets. J Biol Chem. 2013;288(19):13325-13336. Dale GL. Coated-platelets: an emerging component of the procoagulant response. J Thromb Haemost. 2005;3(10):2185-2192. Dale GL, Friese P, Batar P, et al. Stimulated platelets use serotonin to enhance their retention of procoagulant proteins on the cell surface. Nature. 2002;415(6868):175-179. Alberio LJ, Clemetson KJ. All platelets are not equal: COAT platelets. Curr Hematol Rep. 2004;3(5):338-343. Szasz R, Dale GL. Thrombospondin and fibrinogen bind serotonin-derivatized proteins on COAT-platelets. Blood. 2002;100(8):2827-2831. Maroney SA, Haberichter SL, Friese P, et al. Active tissue factor pathway inhibitor is expressed on the surface of coated platelets. Blood. 2007;109(5):1931-1937. Jobe SM, Leo L, Eastvold JS, et al. Role of FcRg and factor XIIIA in coated platelet formation. Blood. 2005;106(13):4146-4151. Kirkpatrick AC, Tafur AJ, Vincent AS, Dale GL, Prodan CI. Coated-platelets improve

14.

15.

16.

17.

prediction of stroke and transient ischemic attack in asymptomatic internal carotid artery stenosis. Stroke. 2014;45(10):29953001. Prodan CI, Stoner JA, Cowan LD, Dale GL. Higher coated-platelet levels are associated with stroke recurrence following nonlacunar brain infarction. J Cereb Blood Flow Metab. 2013;33(2):287-292. Prodan CI, Vincent AS, Dale GL. Coatedplatelet levels are persistently elevated in patients with mild traumatic brain injury. J Head Trauma Rehabil. 2014;29(6):522-526. Rosado JA, Meijer EM, Hamulyak K, Novakova I, Heemskerk JW, Sage SO. Fibrinogen binding to the integrin αIIbb3 modulates store-mediated calcium entry in human platelets. Blood. 2001;97(9):26482656. Lauer P, Metzner HJ, Zettlmeissl G, et al. Targeted inactivation of the mouse locus encoding coagulation factor XIII-A: hemostatic abnormalities in mutant mice and characterization of the coagulation deficit. Thromb Haemost. 2002;88(6):967-974.

435


N.J.A. Mattheij et al. 18. Gilio K, Harper MT, Cosemans JM, et al. Functional divergence of platelet protein kinase C (PKC) isoforms in thrombus formation on collagen. J Biol Chem. 2010;285(30):23410-23419. 19. Van der Meijden PE, Feijge MA, Swieringa F, et al. Key role of integrin αIIbb3 signaling to Syk kinase in tissue factor-induced thrombin generation. Cell Mol Life Sci. 2012;69(20):3481-3492. 20. Vanschoonbeek K, Wouters K, van der Meijden PE, et al. Anticoagulant effect of dietary fish oil in hyperlipidemia: a study of hepatic gene expression in APOE2 knock-in mice. Arterioscler Thromb Vasc Biol. 2008;28(11):2023-2029. 21. Hemker HC, Giesen P, Al Dieri R, et al. Calibrated automated thrombin generation measurement in clotting plasma. Pathophysiol Haemost Thromb. 2003;33 (1):4-15. 22. Gilio K, van Kruchten R, Braun A, et al. Roles of STIM1 and Orai1 in glycoprotein VI- and thrombin-dependent procoagulant activity and thrombus formation. J Biol Chem. 2010;285:23629-23638. 23. Van Kruchten R, Cosemans JMEM, Heemskerk JWM. Measurement of whole blood thrombus formation using parallelplate flow chambers: a practical guide. Platelets. 2012;23(31):229-242. 24. Cosemans JM, Munnix IC, Wetzker R, Heller R, Jackson SP, Heemskerk JW. Continuous signaling via PI3K isoforms b and g is required for platelet ADP receptor function in dynamic thrombus stabiliza-

436

tion. Blood. 2006;108(9):3045-3052. 25. Miserus RJ, Herías MV, Prinzen L, et al. Molecular MRI of early thrombus formation using a bimodal alpha2-antiplasminbased contrast agent. JACC Cardiovasc Imaging. 2009;2(8):987-996. 26. Abaeva AA, Canault M, Kotova YN, et al. Procoagulant platelets form an α-granule protein-covered "cap" on their surface that promotes their attachment to aggregates. J Biol Chem. 2013;288(41):29621-29632. 27. Cosemans JM, Schols SE, Stefanini L, et al. Key role of glycoprotein Ib/V/IX and von Willebrand factor in platelet activationdependent fibrin formation at low shear flow. Blood. 2011;117(2):651-660. 28. Zeiler M, Moser M, Mann M. Copy number analysis of the murine platelet proteome spanning the complete abundance range. Mol Cell Proteomics. 2014;13(12):3435-3445. 29. London FS, Marcinkiewicz M, Walsh PN. A subpopulation of platelets responds to thrombin- or SFLLRN-stimulation with binding sites for factor IXa. J Biol Chem. 2004;279(19):19854-19859. 30. Panteleev MA, Ananyeva NM, Greco NJ, Ataullakhanov FI, Saenko EL. Two subpopulations of thrombin-activated platelets differ in their binding of the components of the intrinsic factor X-activating complex. J Thromb Haemost. 2005;3(11): 2545-2553. 31. Kulkarni S, Jackson SP. Platelet factor XIII and calpain negatively regulate integrin αIIbb3 adhesive function and thrombus

32.

33.

34.

35. 36.

37.

38.

growth. J Biol Chem. 2004;279(29):3069730706. Podolnikova NP, Yakovlev S, Yakubenko VP, Wang X, Gorkun OV, Ugarova TP. The interaction of integrin αIIbb3 with fibrin occurs through multiple binding sites in the αIIbb-propeller domain. J Biol Chem. 2014;289(4):2371-2783. Nikolajsen CL, Dyrlund TF, Poulsen ET, Enghild JJ, Scavenius C. Coagulation factor XIIIa substrates in human plasma: identification and incorporation into the clot. J Biol Chem. 2014;289(10):6526-6534. Siebenlist KR, Meh DA, Mosesson MW. Protransglutaminase (factor XIII) mediated crosslinking of fibrinogen and fibrin. Thromb Haemost. 2001;86(5):1221-1228. Nurden AT, Nurden P. Inherited disorders of platelet function: selected updates. J Thromb Haemost. 2015;13 Suppl 1:S2-9. Puszkin EG, Raghuraman V. Catalytic properties of a calmodulin-regulated transglutaminase from human platelet and chicken gizzard. J Biol Chem. 1985;260(29):1601216020. Burkhart JM, Vaudel M, Gambaryan S, et al. The first comprehensive and quantitative analysis of human platelet protein composition allows the comparative analysis of structural and functional pathways. Blood. 2012;120(15):e73-82. Mitchell JL, Lionikiene AS, Fraser SR, Whyte CS, Booth NA, Mutch NJ. Functional factor XIII-A is exposed on the stimulated platelet surface. Blood. 2014;124 (26):3983-3990.

haematologica | 2016; 101(4)


ARTICLE

Bone Marrow Failure

Extracellular vesicle miR-7977 is involved in hematopoietic dysfunction of mesenchymal stromal cells via poly(rC) binding protein 1 reduction in myeloid neoplasms

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Hiroto Horiguchi,1 Masayoshi Kobune, 1 Shohei Kikuchi,1 Masahiro Yoshida,1 Masaki Murata,2 Kazuyuki Murase,1 Satoshi Iyama,1 Kohichi Takada,1 Tsutomu Sato,1 Kaoru Ono,1 Akari Hashimoto,1 Ayumi Tatekoshi,1 Yusuke Kamihara,1 Yutaka Kawano,1 Koji Miyanishi,1 Norimasa Sawada,2 and Junji Kato 1 Department of Medical Oncology and Hematology, Sapporo Medical University School of Medicine; and 2Department of Pathology, Sapporo Medical University School of Medicine, Japan

1

Haematologica 2016 Volume 101(4):437-447

ABSTRACT

T

he failure of normal hematopoiesis is observed in myeloid neoplasms. However, the precise mechanisms governing the replacement of normal hematopoietic stem cells in their niche by myeloid neoplasm stem cells have not yet been clarified. Primary acute myeloid leukemia and myelodysplastic syndrome cells induced aberrant expression of multiple hematopoietic factors including Jagged-1, stem cell factor and angiopoietin-1 in mesenchymal stem cells even in noncontact conditions, and this abnormality was reverted by extracellular vesicle inhibition. Importantly, the transfer of myeloid neoplasmderived extracellular vesicles reduced the hematopoietic supportive capacity of mesenchymal stem cells. Analysis of extracellular vesicle microRNA indicated that several species, including miR-7977 from acute myeloid leukemia cells, were higher than those from normal CD34+ cells. Remarkably, the copy number of miR-7977 in bone marrow interstitial fluid was elevated not only in acute myeloid leukemia, but also in myelodysplastic syndrome, as compared with lymphoma without bone marrow localization. The transfection of the miR-7977 mimic reduced the expression of the posttranscriptional regulator, poly(rC) binding protein 1, in mesenchymal stem cells. Moreover, the miR-7977 mimic induced aberrant reduction of hematopoietic growth factors in mesenchymal stem cells, resulting in decreased hematopoietic-supporting capacity of bone marrow CD34+ cells. Furthermore, the reduction of hematopoietic growth factors including Jagged-1, stem cell factor and angiopoietin-1 were reverted by target protection of poly(rC) binding protein 1, suggesting that poly(rC) binding protein 1 could be involved in the stabilization of several growth factors. Thus, miR-7977 in extracellular vesicles may be a critical factor that induces failure of normal hematopoiesis via poly(rC) binding protein 1 suppression.

Correspondence: mkobune@sapmed.ac.jp

Received: August 6, 2015. Accepted: January 15, 2016. Pre-published: January 22, 2016. doi:10.3324/haematol.2015.134932

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/4/437

Introduction In myeloid neoplasms (MNs) including acute myeloid leukemia (AML) and myelodysplastic syndrome (MDS), a variety of mechanisms could be involved in the failure of normal hematopoiesis.1-3 In these disorders, neoplastic clones eventually take over the bone marrow (BM) niche even in lower-risk MDS and hypoplastic MDS.4 It has been suggested that normal hematopoiesis could be compromised in the development of AML/MDS as well as the growth advantage of AML/MDS cells.5-8 However, the precise molecular mechanisms governing the replacement of normal hematopoietic stem/progenitor cells by AML/MDS stem/progenitor cells haematologica | 2016; 101(4)

Š2016 Ferrata Storti Foundation Material published in Haematologica is covered by copyright. All rights reserved to Ferrata Storti Foundation. Copies of articles are allowed for personal or internal use. A permission in writing by the publisher is required for any other use.

437


H. Horiguchi et al.

have not yet been clarified. Recently, it has been shown that BM stromal cells, including mesenchymal stem/stromal cells (MSCs), cooperate to maintain normal hematopoietic9-12 and leukemic stem cells via several molecules, including adhesion molecules, gap junction proteins, cytokines and morphogens.13 More recently, studies using mesenchymal progenitor-specific knockout mice demonstrated impaired microRNA (miRNA) biogenesis in BM MSCs and the development of MDS.14 In patients with AML/MDS, it has been shown by our group and others that abnormal protein expression, such as that of hedgehog-interacting protein15 or aurora kinase A/B,16 occurs in MSCs. These findings suggest that the dysfunction of MSCs could be associated with the development of AML/MDS. Recently, extracellular vesicles (EVs) released from hematopoietic and BM stromal cells have been found and regarded as novel factors that modulate communication between stem cells and their niche.17 The EVs have been roughly classified into three types including apoptotic body, microvesicle and exosome, according to their size and production mechanism.18 EVs are extracellular nanoshuttles of RNA, protein and lipids that facilitate communication between cells and tissues. However, little is known about the precise molecular mechanisms and involvement of EVs that govern the induction of stromal abnormalities.19-21 In the present study, we first conducted comparative analyses between normal MSCs and those derived from AML/MDS patients to gain insight into the comprehensive changes in gene expression and cell function. We further attempted to identify effectors that were correlated with alterations in AML/MDS-derived MSCs. Consequently, we focused on EV miR-7977 released from AML/MDS cells. We found that the copy number of miR7977 in the plasma of the BM cavity (BM fluid) was elevated not only in AML patients, but also in MDS patients. Moreover, transfection of a miR-7977 mimic induced the reduction of hematopoietic growth factors in BM MSCs, resulting in a decreased hematopoietic-supporting capacity of BM CD34+ cells.

Methods Reagents and human BM MSCs GW4869 (inhibitor of the neutral sphingomyelinase, SMPD2) was purchased from Cayman Chemical (Ann Arbor, MI, USA). Anti-Jagged 1 (JAG1) (ab109536) and anti- PCBP1 (poly(rC) binding protein 1) antibodies (ab168377) were purchased from Abcam® (Tokyo, Japan). StemPro®-34 (Life Technologies, Carlsbad, CA, USA) was used as a serum-free medium. This study was approved by the Institutional Review Board at our university and conducted according to the Declaration of Helsinki. The patients with lymphoma stage I/II and those with AML/MDS in this study were also fully informed of the experimental protocol. Human BM CD34+ cells and four different lot numbers of MSCs from healthy volunteers (HVs) were purchased from AllCells, LLC (Toronto, Canada) (HV-derived MSCs #1, #2, #3 and #4). Human primary MSCs and human BM CD34+ cells derived from lymphoma patients stage I/II without bone marrow localization (control MSCs) and patients with AML/MDS (Online Supplementary Table S1 and Table S2) were prepared as previously described.15, 22 The MSCs were cultured in MSCGM™ hMSC basal medium with the addition of supplements from the MSCGM hMSC SingleQuot Kit 438

(Lonza Japan Ltd, Tokyo, Japan). The diagnostic criteria of AML/MDS were based according to the World Health Organization (WHO) 2008 Classification of MNs. MDS was further evaluated by the Revised International Prognostic Scoring System (IPSS-R).

Coculture of BM CD34+ cells with human MSCs We modified our previous cord blood-based coculture system to a BM CD34+-based system.22,23 In this new coculture system, BM CD34+ cells were cocultured with human control- or AML/MDSderived MSCs in serum-free StemPro®-34 medium in the presence of a cytokine cocktail consisting of 50 ng/mL human thrombopoietin (TPO), 10 ng/mL human stem cell factor (SCF), 50 ng/mL human Flk2/Flt3 ligand (FLT3LG) and 100 ng/mL human delta-like protein 4 (DLL4) (all from R&D Systems, Minneapolis, MN, USA). In this system, the primary MSC layer could be maintained for over 8 weeks even in a serum-free medium (Online Supplementary Methods).

Clonogenic analysis of cocultured hematopoietic cells The clonogenic assay was performed using MethoCult GF H4434V (StemCell Technologies, Vancouver, Canada) as described previously.23

Contact and non-contact culture systems Contact and non-contact culture systems were conducted using Polyester Membrane Transwell Clear Inserts and Companion Plates (BD Biosciences, San Jose, CA, USA; pore size: 0.4 mm, pore density: 1x108/cm2, 12 well) as reported previously (Online Supplementary Methods).24

EV transfer assay using cells labeled with GFP, PKH26 and SYTO RNAselect Green fluorescent protein (GFP)-transduced leukemic cells were established using the retroviral vector, pRx-IRES-hrGFP, as described previously.23 The leukemic cells were stained with PKH26 (Sigma-Aldrich, St. Louis, MO, USA), a red fluorescent membrane cell linker, before coculture according to the manufacturer’s instructions as previously reported.25 Total RNA in leukemic cells was stained with SYTO RNAselect (Life Technologies). 2 x 104 to 2 x 105 labeled leukemic cells were added into the transwell insert and cocultured for 3 or 14 days with or without 10 mM GW4869 or nSMase2 siRNA (Stealth RNAi SMPD2 human (s13170), Life Technologies) to inhibit EV secretion. Target MSCs were transferred onto Lab-Tek II Chamber Slides (Thermo Scientific, Waltham, MA, USA), and visualized using ZEISS/ELYRAS 1LSM780 confocal microscope (ZEISS, Oberkochen, Germany).

EV preparation EVs were isolated from the supernatant of hematopoietic cell lines or BM fluid by centrifugation, filtration and the Exosome Precipitation Solution (ExoQuick-TC; System Biosciences, Mountain View, CA, USA). Briefly, the supernatant of hematopoietic cells or the BM fluid was centrifuged at 3,000 g for 15 min to remove cells and apoptotic bodies.26 Subsequently, the sample was passed through a 0.45 mm pore size Millipore Hydrophilic Durapore filter (Merck Millipore, Tokyo, Japan).27,28 The largersized microvesicular particles were deposited onto the filter membrane. The resulting filtrate was transferred to a sterile vessel, and the appropriate volume of ExoQuick-TC was added. After incubation at 4°C for 2 hours, the mixed solution was centrifuged at 1,500 g for 30 minutes and the supernatant was removed. The pellet was kept on ice and used as the EV fraction after centrifugation at 1,500 g for 5 minutes to remove residual solution. haematologica | 2016; 101(4)


miR-7977 modulates PCBP1 level in BM MSCs

Microarray analysis for EV miRNA

Results

To harvest EVs from 2 x 105 primary AML CD34+ cells, normal BM CD34+ cells and leukemic cell lines, including TF-1 and Kasumi-1, cells were cultured in serum-free StemPro®-34 medium with a cytokine cocktail on plates coated with FN fragments (Retronectin®: Takara Bio, Tokyo, Japan) instead of MSCs as reported earlier.29 EV miRNA from the supernatant of CD34+ hematopoietic and leukemic cells were prepared as reported previously.30 Microarray analysis of the miRNA profiles was done with the human miRNA Oligo chip (Human_miRNA_V20) and the 3D-Gene® miRNA labeling kit (TORAY, Kanagawa, Japan).

Statistical analysis Each data set was first evaluated for normality of distribution by the Komolgorov-Smirnov test to decide whether a nonparametric rank-based analysis or a parametric analysis should be used. The significance of differences between groups was assessed by ANOVA, followed by Dunnett’s multiple comparison tests. Results are expressed as the mean ± standard deviation (SD). The significance of differences was assessed by either the Student’s t-test or the Mann-Whitney U-test, and a P-value < 0.05 was considered as statistically significant.

A

Analysis of gene expression and hematopoietic-supporting capacity of MSCs derived from normal or AML/MDS BM In the present study, we first screened for differences in gene expression between MSCs derived from control and AML/MDS BM by quantitative real-time PCR (qRT-PCR) array (Online Supplementary Methods). The expression of multiple hematopoietic factors was reduced in AML/MDS-derived MSCs as compared with control MSCs, although the expression of some genes such as tolllike receptor 4 (TLR4) and CD44 was elevated in AML/MDS-derived MSCs (Figure 1A). To assess the hematopoietic-supporting capacity of AML/MDS-derived MSCs, normal BM CD34+ cells were cocultured with human control- or AML/MDS-derived MSCs in serumfree StemPro®-34 medium in the presence of a cytokine cocktail consisting of human SCF, TPO, FLT3LG and DLL4.22 Using this system, we assessed the clonogenicity and in vivo repopulating activity of BM CD34+ cells 14 days after their coculture with primary MSCs derived from

B

C

Figure 1. Comparative analysis of mRNA expression between HV- and AML/MDS-derived MSCs. (A) Analysis of hematopoietic factors in MSCs by qRT-PCR array. Results of cluster analysis in MSCs-derived from control (MSC ID 10, 11 and 12), RCUD (MSC ID 8), RAEB-2 (MSC ID 6), AML M2 (MSC ID 2) and AML M4 (MSC ID 1) patients (Online Supplementary Table S1). (B) Clonogenic assay after coculture with human stromal cells. Y-axis indicates the number of colonies after ex vivo coculture of 2x104 BM CD34+cells on MSC layer. X-axis indicates the individual MSCs derived from control and AML/MDS patients. Pre-coculture indicates the number of colonies derived from fresh BM CD34+ cells. MSCs derived from HV were used as normal control. *P<0.01, colony-forming units (CFU)-Mix: primary AML/MDSderived MSCs vs. normal control-derived MSCs (Student’s t-test, two-tailed). BFU-E, burst forming units of erythroid; CFU-GM, CFU-granulocyte/monocyte; CFU-Mix, CFU-mixed. Results are expressed as means ± SD. Similar results were obtained in 3 independent experiments, performed in triplicate. AML/MLD: AML with multilineage dysplasia. (C) All hematopoietic cells that were cocultured with AML/MDS-derived MSCs were transplanted into immunodeficient mice. Representative results, each performed in triplicate of control MSCs (ID 10 (●), ID 11 (■), ID 12 ( )), MSCs derived from AML (ID 1 (○), ID 3 (□), ID 4 (△)) and MSCs derived from MDS (ID 9 (▲), ID 8 ( )), are shown (Online Supplementary Table S1). Y-axis indicates the percentage of human specific CD45+ cells. X-axis indicates weeks after transplantation. *P<0.01 controls vs. AML/MDS. Similar results were obtained in two independent experiments.

haematologica | 2016; 101(4)

439


H. Horiguchi et al.

AML/MDS patients. The number of colony forming unit (CFU)-Mix was remarkably reduced as compared with control MSCs (Figure 1B). Moreover, BM CD34+ cells cocultured with AML/MDS-derived MSCs, but not control MSCs, lost in vivo repopulating activity in immunodeficient mice (Figure 1C). Furthermore, comparative analyses between control- (ID 11) and AML-derived MSCs (ID 2) using age-matched patients showed identical results (Online Supplementary Figure S1). These results indicated that AML/MDS-derived MSCs could not support normal hematopoietic progenitor/stem cells.

We further investigated whether primary AML and MDS cells could induce a similar effect on primary BM MSCs cultured in a non-contact system using hematopoietic PCR array. Surprisingly, even primary MDS cells and AML cells induced altered mRNA expression of multiple hematopoietic growth factors in non-contact conditions (Figure 2C, Highlighted). Remarkably, key hematopoietic factors such as JAG1 and SCF in MSCs were significantly reduced (Figure 2D). Collectively, certain soluble/humoral factors may be involved in the reduction of JAG1 and SCF mRNA expression.

Elucidation of the effectors that alter the hematopoietic factors derived from AML/MDS stromal cells

The release of EVs from leukemic cells and transfer to MSCs

In an attempt to identify these effectors, we employed non-contact and contact coculture systems to determine whether the effectors are soluble or not (Figure 2A). In this analysis, we employed the change in stromal JAG1 expression as an indicator, and used AML cell lines such as TF-1 (AML M6) and Kasumi-1 (AML M2) in this screening. Unexpectedly, in both non-contact and contact coculture systems, decreased JAG1 and SCF expression in MSCs was observed with both leukemic cell lines (Figure 2B), indicating that the effectors were at least in part soluble or humoral.

It has been shown that hematopoietic cells, including those in MNs, release a variety of soluble/humoral factors such as cytokines, membrane-anchored mediators including Wnt/Hh,23,31 shed receptors and EVs.32 Among them, recent studies have focused on the involvement of EVs including exosomes in transcriptome alteration and cellular phenotype switching. Hence, we investigated whether leukemic cells secrete EVs in vitro into the supernatant. The fraction of EVs was examined by transmission electron microscopy, and 30-50 nm vesicles were mainly observed

A

C

B

D

Figure 2. Analysis of the effect of MN cells on BM MSCs. (A) Non-contact and contact culture systems using transwell clear inserts with 0.4 μm pore membrane filter and companion plates in serum-free StemPro®-34 medium. (B) The effect of non-contact and contact coculture of TF-1 or Kasumi-1 leukemic cells with BM MSCs on JAG1 mRNA expression. Results are expressed as means ± SD. *P<0.05, normal MSCs cocultured with AML cells vs. normal MSCs without coculture (Control). (C) Effect of primary AML cells and MDS cells on the expression of hematopoietic factors in BM MSCs. CD34+ fraction of primary AML cells and MDS cells were cocultured with MSCs in the presence of cytokines including SCF, TPO and DLL4. The hematopoietic factors in MSCs were analyzed by qRT-PCR array. Sample ID: Cont-1, ID 1; Cont-2, ID 3; Cont-3, ID 5; RCUD, ID 11; RAEB-1, ID 23; AML M1, ID 35; AML M2, ID 37; AML M4, ID 40 (Online Supplementary Table S2). (D) The expression levels of JAG1 and SCF mRNAs were further confirmed by qRT-PCR. *P<0.01 controls vs. AML/MDS. Data represent three independent experiments, each done in triplicate.

440

haematologica | 2016; 101(4)


miR-7977 modulates PCBP1 level in BM MSCs

(Figure 3A). Subsequently, the tetraspanin, CD63, and endosomal markers, ALIX and TSG101, were examined by immunoblotting analysis. As a result, CD63, ALIX and TSG101 were detected in the fraction of EVs released from TF-1 and Kasumi-1 cells (Figure 3B), suggesting that EVs could contain exosomes as well as other microvesicles. Subsequently, we determined whether EV transfer from leukemic cells to MSCs could be achieved. Leukemic cells labeled with GFP and PKH26 (cell membrane) or SYTO RNAselect (total RNA) were cocultured in a non-contact system with MSCs. Diffuse GFP (green) and multiple and sporadic PKH26 (red) signals were detected in MSCs by confocal microscopy (Figure 3C, upper left panel). Superresolution analysis of these confocal images showed that PKH26 signals exhibited multiple aggregations of hollow shell pattern (Figure 3C, lower left panel). Further, sporadic signals for SYTO RNAselect (green) were observed in MSCs (Figure 3C, right panel). These findings clearly indi-

A

B

cated that the EVs derived from leukemic cells were efficiently transferred into MSCs.

The effect of inhibition on microvesicular transfer and gene expression in MSCs To confirm the transfer of EVs and determine its effect on MSCs, an EV inhibitor which functions through the suppression of neutral sphingomyelinases (GW4869), as well as siRNA and short hairpin (sh)RNA against the neutral sphingomyelinase, sphingomyelin phosphodiesterase 2 (SMPD2), were used to treat MSCs cultured in a noncontact system with leukemic cells. PKH26 (red) and GFP (green) signals were largely decreased in the presence of the inhibitor of EVs by confocal microscopic analysis, indicating a reduction in the transfer of EVs from leukemic cells to MSCs (Figure 3D and Online Supplementary Figure S2A,S2B). We subsequently examined whether the EV inhibitor protected against changes in mRNA expression by qRT-PCR array. Acute leukemic cell lines in non-con-

E

C

D

Figure 3. The effect of inhibition of EV biogenesis by the nSMase2 inhibitor, GW4869, on BM stromal cells cocultured with leukemic cells. (A) Transmission electron microscopy. The EV fraction was prepared from the supernatant of TF-1 cells. Scale bar: 100 nm. (B) The EV fraction was analyzed by immunoblotting using anti-CD63, anti-ALIX and anti-TSG101 antibodies. The EV fraction was prepared from 5 ml supernatant of TF-1 or Kasumi-1 cells, or serumfree medium. (C) EV transfer to MSCs was analyzed by super resolution confocal microscopy (ZEISS/ELYRAS 1LSM780). Left panels (upper and lower): Image of an MSC cocultured with Kasumi-1 leukemic cells labeled with GFP (green) and PKH26 (red) shows presence of microvesicles in the cell (yellow, arrows). Scale bar: 10 mm (left upper), 2.0 mm (left lower). Right panel: Image of an MSC cocultured with leukemic cells labeled with SYTO RNAselect™ (green) shows transfer of RNA from leukemic cells to the MSC. The signals were obtained by 5 rotations, and the pictures were reconstituted as a wide-field structured illumination microscopy (SIM) image. DAPI (blue) was used for nuclear staining. Scale bar: 10 mm. (D) EV transfer assay was conducted in the presence or absence of an EV inhibitor (GW4869) in MSCs cocultured with leukemic cells labeled with PKH26 (red) and GFP (green), and visualized by confocal microscopy. Merged images are shown, and DAPI (blue) was used for nuclear staining. Scale bar: 20 mm. (E) Changes in JAG1 and SCF mRNA levels in MSCs cocultured with TF-1 or Kasumi-1 cells in serum-free StemPro®34 medium and treated with or without GW4869 were examined by qRT-PCR. Data shown are from one representative experiment of three showing similar results, each done in sextuplicate. Y-axis indicates fold change relative to the control or control + GW4869 (=1.0) after normalization to 18S expression level. Results are expressed as means ± SD. †P<0.05, HV-derived MSCs without coculture vs. HV-derived MSCs cocultured with AML cells. *P<0.05, HV-derived MSCs cocultured with AML cells in the absence of GW4869 vs. HV-derived MSCs cocultured with AML cells in the presence of GW4869 (Student’s t-test, two-tailed).

haematologica | 2016; 101(4)

441


H. Horiguchi et al.

tact coculture with MSCs reduced the gene expression of multiple hematopoietic factors including SCF and JAG1 in MSCs (Figure 3E). Importantly, the EV inhibitor reverted the reduction of a large number of genes including JAG1 and SCF (Figure 3E and Online Supplementary Figure S3A). We additionally confirmed the effect of EV inhibition on gene expression of SCF and JAG1 by qRT-PCR. The decrease in JAG1 expression was clearly inhibited by SMPD2 shRNA (Online Supplementary Figure S2B) and the EV inhibitor (Online Supplementary Figure S3A). SCF expression was partially restored by shRNA (Online Supplementary Figure S2C), suggesting that SCF expression was regulated by not only EVs, but other factors as well. We further examined whether these effects on MSCs could be induced by primary AML cells. Although normal CD34+ cells did not alter JAG1 expression in MSCs (Online Supplementary Figure S3B), myeloblastic leukemia (M2) and myelomonocytic leukemia (M4) cells induced the reduction in JAG1 expression, which was restored by EV inhibition. Collectively, the alterations in certain mRNA expressions such as JAG1 in MSCs were induced via EV transfer from leukemic cells.

A

C

Purification of EVs from primary AML and analysis of miRNA It has been revealed that EVs contain multiple components including miRNA, mRNA, fragments of DNA, peptides and lipids. Moreover, our results demonstrated that detectable EV RNA from leukemic cells was transferred into MSCs (Figure 3C). Balakrishnan et al. have recently shown that miRNA interacts with mRNA and regulates gene expression in BM stromal cells.33 Therefore, we decided to analyze EV miRNA in our model. To obtain EV miRNA, we utilized a fibronectin (FN)-based, MSC-free culture system (Figure 4A). The CD34+ fraction of normal BM or AML cells was cultured on FN-coated dishes in 10 mL of serum-free medium. Subsequently, the EV fraction was enriched from the supernatant, and EV miRNA was extracted. We compared the contents of miRNA harvested from CD34+ cells derived from HV and AML patients using a human miRNA Oligo chip. Scatter plot and cluster analysis revealed that EVs derived from primary AML (M1 and M4) cells contained an elevated fraction of miRNAs as compared with normal BM CD34+ cells (Figure 4B,C). In particular, EV miR-4286, miR-7977 and miR-8073 from

D

B

E

Figure 4. Comparison of miRNA between primary hematopoietic and leukemic CD34+ cells. (A) The culture system using normal BM or AML CD34+ cells plated on fibronectin substratum in the presence of cytokines including SCF, TPO, FLT3LG, IL-3 and DLL4. In this system, fibronectin fragment instead of MSCs was utilized to purify the EVs derived from hematopoietic cells. (B) miRNA was analyzed by miRNA chip and compared. Scattered plot demonstrates that EV miRNA is released from normal CD34+ and primary AML cells (M4, Sample ID 40). (C) Heatmap of elevated EV miRNAs derived from leukemic CD34+ cells as compared with those from normal CD34+ cells. Highlighted EV miRNAs show more than 2-fold elevation in both AML M1 (Sample ID 35) and AML M4 (Sample ID 40) as compared with normal CD34+ cells (Online Supplementary Table S2). CD34: commercially purchased normal CD34+ cells. (D) The levels of miR-7977, miR-4286 and miR-8073 were examined by qRT-PCR in MSCs cocultured with BM CD34+ , primary AML M1 and AML M4 cells. SNORD61 was used as an internal standard. Results are expressed as means Âą SD. *P<0.05, normal MSCs cocultured with AML cells vs. normal MSCs cocultured with CD34+ cells (Studentâ&#x20AC;&#x2122;s t-test). (E) Cy5-labeled miR-7977 was transfected into AML M1 cells, which were cocultured with MSCs for 3 days. The labeled miR-7977 that was transferred into MSCs was analyzed by super resolution confocal microscopy (ZEISS/ELYRAS 1LSM780). Scale bar: 10 mm.

442

haematologica | 2016; 101(4)


miR-7977 modulates PCBP1 level in BM MSCs

primary leukemic cells were more than 2-fold higher than those from normal CD34+ cells. We examined whether these miRNAs could be transferred into MSCs after coculture with normal CD34+ cells and primary AML (M1 and M4) cells. Expectedly, the levels of miR-7977, miR-4286 and miR-8073 in MSCs after coculture with primary AML (M1 and M4) cells were significantly higher than in those cocultured with normal CD34+ cells (Figure 4D). Moreover, Cy5-labeled miR-7977 in AML M1 cells was transferred into MSCs 3 days after non-contact coculture (Figure 4E). These results indicated that miRNAs in AML cells could be transferred via EVs.

The level of miRNA in human BM cavity We further investigated whether the level of EV miRNA was elevated in BM fluid in patients with AML/MDS obtained by BM aspiration. According to the Revised International Prognostic Scoring System (IPSS-R), MDS patients were classified into lower-risk MDS (very low and low) and higher-risk MDS (intermediate, high and very high). Electron microscopy revealed that multiple 3050 nm vesicles could be observed in BM fluid (Figure 5A), and CD63, ALIX and TSG101 were detected in these vesicles (Figure 5B). Subsequently, we analyzed the copy number of miR-7977, miR-4286 and miR-8073 in EVs of BM from lymphoma stage I/II (as control), MDS and AML patients (Online Supplementary Table S2). We found that miR-7977 was significantly increased even in lower-risk MDS in addition to higher-risk MDS and AML patients (Figure 5C). Moreover, miR-7977 was significantly correlated with the percentage of blasts in BM (Figure 5D). However, miR-4286 and miR-8073 were not significantly increased in lower-risk MDS patients although miR-4286 was significantly increased in AML patients, and miR8073 was significantly increased in higher-risk MDS and AML patients (Online Supplementary Figure S4A,S4B). Collectively, these results indicated that aberrant expression of miR-7977 in the BM cavity could be involved in the disturbance of normal hematopoiesis in patients with MDS and AML.

A

The effect of miR-7977 on MSCs Using an online database for miRNA target prediction (miRDB), we found that miR-7977 can potentially interact with JAG1 mRNA and primarily interacts with PCBP1 mRNA, which is involved in posttranscriptional control.34 Hence, we employed a miR-7977 mimic to analyze the effect on BM MSCs. The levels of JAG1 and PCBP1 mRNA were decreased after transfection of the miR-7977 mimic (Online Supplementary Figure S5A), and target protection of JAG1 and PCBP1 reverted their reduction (Figure 6A,B). Moreover, luciferase assay with the 3’ untranslated region (3’UTR) of JAG1 and PCBP1 indicated that miR-7977 directly interacted with PCBP-1 and JAG1 mRNAs (Figure 6C,D). Unexpectedly, target protection of PCBP1 partially reverted the reduction in JAG1 mRNA after transfection with the miR-7977 mimic (Figure 6E). It has been revealed that the K-homologous (KH) domain of PCBP1 binds to the 3’UTR with a C-rich motif of mRNAs and enhances the efficiency of 3’ processing, thereby altering the levels of expression of subsets of mRNAs in the mammalian transcriptome.34 Thus, JAG1 mRNA may be a target of PCBP1. These findings indicated that miR-7977 regulated JAG1 expression at the translational and posttranscriptional levels via PCBP1. Importantly, the levels of mRNAs of multiple growth factors were reduced after transfection of the miR-7977 mimic into MSCs (Online Supplementary Figure S5B). Moreover, the reductions in SCF and ANGPT1 (Angiopoietin 1) proteins were reverted by target protection of PCBP1 (Figure 6F), suggesting that PCBP1 could be involved in the stabilization of multiple growth factors. Collectively, transfection of a miR-7977 mimic could induce disturbance of the expression of hematopoietic factors in BM MSCs.

Evaluation of hematopoietic-supporting capacity of MSCs after transfection of a miR-7977 mimic In an attempt to assess the hematopoietic-supporting capacity of negative control or miR-7977 mimic-transfected MSCs, we analyzed the expression of surface markers on hematopoietic cells 7 days after coculture with BM CD34+

C

Lymphoma Lower-risk Higher-risk AML stage I/II MDS MDS

B

D r=0.473 P=0.0021

haematologica | 2016; 101(4)

Figure 5. The copy number of EV miR-7977 in BM fluid. (A) The EVs in BM fluid after BM aspiration (Sample ID 3) were analyzed by transmission electron microscopy. Scale bar: 50 nm. (B) The EV fraction derived from patients with MALT lymphoma stage Ι (Sample ID 3), RCUD (ID 11), RAEB-1 (ID 23) and AML (ID 28) was analyzed by immunoblotting using anti-CD63, anti-ALIX and anti-TSG101 antibodies (Online Supplementary Table S2). (C) EV miR-7977 in BM fluid was analyzed by qRT-PCR. A miR-7977 mimic was used as internal positive control during miRNA quantification. Y-axis indicates the copy number of miR-7977 in 1 mL of BM fluid obtained by BM aspiration. Lymphoma stage I/II without BM localization as control (n=9), lower-risk MDS (n=10), higher-risk MDS (n=8) and AML (n=13) patients. Results are expressed as means ± SD. *P<0.05, the copy number of miR7977 in lower-risk MDS, higher-risk MDS or AML vs. the copy number of miR-7977 in control (ANOVA, followed by Dunnett’s multiple comparison tests). (D) Correlation between the copy number of miR-7977 in BM fluid and percentage of blasts in BM among control, MDS and AML patients. Correlation coefficient: r=0.473. P=0.0021.

443


H. Horiguchi et al.

cells. The number of CD34+CD38– cells and CFU-Mix in coculture with miR-7977 mimic-transfected MSCs was significantly lower than that in coculture with negative control-transfected MSCs (Figure 7A,B and Online Supplementary Figure S6A). The percentage of CD34CD38– and CD11b+ cells that were cocultured with miR7977 mimic transfected MSCs was higher than that cocultured with negative control transfected MSCs (Figure 7A,B and Online Supplementary Figure S6B). These results suggested that the hematopoietic-supporting capacity of miR-7977 mimic transfected MSCs was reduced as compared with that of negative control transfected MSCs. Importantly, the reduction in the hematopoietic-supporting capacity of miR7977 mimic transfected MSCs was reverted with the cotransfection of the PCBP1 protector, or ANGPT1 or SCF

expression vector (Online Supplementary Figure S7A,S7B).35 Subsequently, we prepared EVs from 5 mL of BM fluid derived from HV or AML/MDS patients and labeled them with PKH26. AML/MDS-derived EVs contained abundant miR-7977 while control EVs had a drastically lower level of it (Figure 5C). The transfer efficiency of both control- and AML/MDS-derived EVs into MSCs was around 50% (Online Supplementary Figure S8). The percentage of CD34+ cells and the number of clonogenic cells in coculture with MSCs harboring AML/MDS-derived EVs were significantly lower than those in coculture with MSCs harboring control EVs (Figure 7C and Online Supplementary Figure S9). These results strongly suggest that miR-7977 modulates the hematopoietic-supporting capacity of BM MSCs via reduction of PCBP1.

A

D

B

E

C

F

Figure 6. The effect of a miR-7977 mimic on MSCs. (A) The reduction in JAG1 was analyzed by western blotting analysis. Control, 5 nM negative control siRNA; miR7977, 5 nM miR-7977 mimic; Control protector, 250 nM negative control miScript Target Protector; JAG1 protector, 250 nM Target Protector for JAG1. (B) The reduction in PCBP1 was analyzed by western blotting analysis. PCBP1 protector, Target Protector for PCBP1. Direct interaction between miR-7977 and target genes was evaluated by luciferase assay using pLuc-JAG1 (C) and pLuc-PCBP1 (D). Data are presented as the ratio of the normalized value to the light emission observed in the cells transduced with the control vector (pLuc-Cont). pMIR-Cont or pMIR-7977 was cotransfected with pLuc-JAG1 or pLuc-PCBP1 into MSCs 48 hrs before analysis (Online Supplementary Methods). Data shown are from one representative experiment of three showing similar results, each done in quadruplicate. Results are expressed as means ± SD. *P<0.05, pMIR-Cont with pLuc-Cont vs. pMIR-7977 with pLuc-JAG1. **P<0.01, pMIR-Cont with pLuc-Cont vs. miR-7977 with pLuc-PCBP1. (E) The expression of JAG1 mRNA in MSCs after transfer of 5 nM miR7977 mimic and 250 nM PCBP1 protector was examined by qRT-PCR. *P<0.05, HV-derived MSCs transduced with negative control vs. HV-derived MSCs transduced with miR-7977 mimic. †P<0.05 MSCs transfected with miR-7977 mimic vs. MSCs cotransfected with miR-7977 mimic and PCBP1 protector (Student’s t-test). Y-axis indicates fold change relative to the control or control + GW4869 (=1.0) after normalization to 18S expression level in MSCs. Data shown are from one representative experiment of three showing similar results, each done in quadruplicate. (F) The recovery of SCF and ANGPT1 expression after PCBP1 protector transfer into MSCs was analyzed by western blotting analysis.

444

haematologica | 2016; 101(4)


miR-7977 modulates PCBP1 level in BM MSCs

Discussion

mutations in hematopoietic cells and eventually the development of AML/MDS.14,36 Further, activating b-catenin knock-in cells exhibited elevated JAG1 expression. Conversely, Dicer1 knockout cells exhibited a reduction in JAG1 expression (GDS3404 and GDS4504). These findings suggest that osteoblasts and MSCs play different roles in supporting normal hematopoiesis,37 and JAG1 expression could be regulated by b-catenin signaling and miRNA biogenesis. Consistent with these previous reports, we and others reported that mRNA expression of hematopoietic factors in MDS-derived MSCs was significantly disturbed.15,38 Moreover, we revealed in the present study that the mRNA expression of several hematopoietic factors in BM MSCs was decreased (Figure 1A), and these reductions correlated with the dysfunction of hematopoietic support in AML/MDS-derived MSCs (Figure 1B,C). Collectively, the disturbance of MSC function in AML/MDS could be involved in the failure of normal hematopoiesis. From these results, an ensuing intriguing question is

In the present study, we found that the expression of multiple growth factors was significantly reduced in AML/MDS-derived BM MSCs as compared with those from control BM MSCs. Functionally, AML/MDS-derived BM MSCs exhibited lower hematopoietic-supporting capacity of stem/progenitor cells. Moreover, EVs derived from the CD34+ fraction of AML/MDS cells could be transferred to MSCs, and EV inhibition partially restored aberrant expression of hematopoietic growth factors including JAG1, SCF and ANGPT1 in AML/MDS-derived MSCs. In addition, EV miR-7977 derived from AML/MDS cells was remarkably enriched in vitro and in BM cavity, and was found to be involved in aberrant expression of mRNAs and reduction in the hematopoietic-supporting capacity of BM MSCs. Recently, mesenchymal progenitor-specific conditional Dicer1 knockout or osteoblast-specific activating b-catenin knock-in increased the frequency of genetic

A

B

C

Figure 7. Hematopoietic-supporting capacity of miR-7977 mimic-transfected MSCs. (A) Expression of CD34 and CD38 in ex vivo cocultured hematopoietic cells. The X-axis indicates CD34 labeled with a FITC-conjugated monoclonal antibody (CD34-FITC). The Y-axis indicates CD38 labeled with a PE-conjugated monoclonal antibody (CD38-PE). Positivity for a surface antigen was defined using the isotype control monoclonal antibody. Left panel: Isotype control. Middle left panel: Expression of CD34/CD38 in ex vivo cultured hematopoietic cells without MSCs in the presence of SCF, TPO, FLT3LG and DLL4. Middle right panel: Expression of CD34/CD38 in cocultured hematopoietic cells with negative control-transfected MSCs. Right panel: Expression of CD34/CD38 in cocultured hematopoietic cells with miR-7977 mimic-transfected MSCs. Data shown are from one representative experiment of three showing similar results. (B) The summary of CD34/CD38 positive cells after culture with or without MSCs. *P<0.01 negative control vs. miR-7977 mimic. MSC free: CD34/CD38 positive cells among cultured hematopoietic cells without MSCs. (C) Expression of CD34 in ex vivo hematopoietic cells cocultured with HV-derived MSCs harboring EVs derived from lymphoma stage I/II (control), MDS or AML patients. Non-treated HV-derived MSCs (n=4, lot number #3), MSCs with control EVs (n=4, sample ID 1, 2, 3 and 4), MSCs with MDS-derived EVs (n=4, sample ID 10, 11, 22 and 23) and MSCs with AML-derived EVs (n=4, sample ID 28, 29, 31 and 33). MDS- and AML-derived EVs contained abundant miR-7977. The Y-axis indicates absolute number of CD34+CD38– cells. *P<0.01, Non-treated MSCs vs. MSCs with AML/MDS-derived EVs. †P<0.05, Non-treated MSCs vs. MSCs with HV-derived EVs (Student’s t-test, two-tailed). Data shown are from one representative experiment of three showing similar results.

haematologica | 2016; 101(4)

445


H. Horiguchi et al.

how the disturbance of stromal function is induced. One possible explanation is that stromal dysfunction occurs spontaneously with advancing age and/or genetic damage mediated by reactive oxygen species. Consistent with these notions, it was previously reported that chromosomal abnormalities such as loss of heterozygosity and uniparental disomy (UPD), which can result from doublestranded breaks, are sometimes observed in MSCs derived from AML/MDS patients.39-41 Another possible explanation is that certain factors can directly induce functional abnormality in MSCs. Recently, it was demonstrated that the transplantation of chronic myelogenous leukemia (CML) repopulating cells into immunodeficient mice altered the microenvironmental regulation of the stem cell niche.12,42,43 In accordance with these findings, we also found that primary CD34+ leukemic cells, but not normal CD34+ cells, induced a decrease in JAG1 and SCF expression. These findings indicated that leukemic cells could induce the dysfunction of BM stromal cells. However, it is difficult to identify the effectors in leukemic cells which mediate these effects. To resolve this problem, we utilized the non-contact and contact culture systems using primary leukemic cells and MSCs. We found that even in a non-contact condition, primary leukemic cells induced the alteration of mRNA expression in MSCs, suggesting that the effectors are soluble or humoral factors (Figure 2). Recently, it was demonstrated that EVs derived from BM MSCs facilitated the progression of multiple myeloma,44 and those derived from CML and chronic lymphocytic leukemia cells facilitated the progression through an autocrine mechanism.19,21 Moreover, primary AML cells released EVs which were possibly enriched for both coding and non-coding RNAs.17,45 These findings led us to explore the possibility of EVs-mediated communication between leukemic cells and MSCs. In the present study, comparative analysis revealed that miRNA species, including miR-7977, were elevated in AMLderived EVs (Figure 4C), and miR-7977 was significantly elevated in MSCs cocultured with primary AML cells (Figure 4D) as well as BM fluid of AML, in lower-risk and higher-risk MDS patients (Figure 5C). These results suggested that BM EVs work as nanoshuttles to carry various biological elements including miRNA. Importantly, transfection of a miR-7977 mimic reduced the levels of JAG1 and PCBP1 (Figure 6A,B). It has been revealed that the KH-domain of PCBP1 binds to a 3’UTR with C-rich motif of mRNAs and enhances the efficiency of 3’ processing, thereby altering the levels of

References 1. Gupta P, Niehans GA, LeRoy SC, et al. Fas ligand expression in the bone marrow in myelodysplastic syndromes correlates with FAB subtype and anemia, and predicts survival. Leukemia. 1999;13(1):44-53. 2. Contini P, Zocchi MR, Pierri I, Albarello A, Poggi A. In vivo apoptosis of CD8(+) lymphocytes in acute myeloid leukemia patients: involvement of soluble HLA-I and Fas ligand. Leukemia. 2007;21(2):253-260. 3. Matsunaga T, Takemoto N, Sato T, et al. Interaction between leukemic-cell VLA-4 and stromal fibronectin is a decisive factor for minimal residual disease of acute myel-

446

expression of subsets of mRNAs in the mammalian transcriptome.34,46 In the present study, the mRNA levels of multiple growth factors were reduced after transfection of a miR-7977 mimic into MSCs (Figure 6C). Collectively, miR-7977 could alter the transcriptome in BM MSCs, suggesting that excess miR-7977 may have an impact on the hematopoietic function of MSCs. In fact, the hematopoietic-supporting capacity was significantly reduced in MSCs after transfection of a miR-7977 mimic and miR7977-enriched EVs (Figure 7). In the present study, although the regulation of EV miRNA levels, including miR-7977, miR-4286 and miR8073, was not clarified, one plausible possibility could be that intracellular miRNAs were increased by differential regulation of the miRNA promoters via certain transcription factors.47,48 In order to determine this, biological databases were used, and several transcription factor binding sites, including those for Evi-1, GATA-2 and PAX-6, were detected in the miR-7977 promoter. Another possible explanation could be that the release of EV miRNAs was elevated.49 In fact, endosomal markers including TSG101 and ALIX were elevated in BM EVs derived from AML and MDS as compared with control EVs (Figure 5B). The third possibility is that certain unknown long non-coding RNAs such as HOTAIR or PCBP1-1:1 in hematopoietic cells may sponge several miRNAs.50 Further studies to investigate the miRNA promoter and level of non-coding RNAs in hematopoietic cells are required to understand the precise regulation of EV miRNAs in AML and MDS. In conclusion, we found that EV miR-7977 derived from AML/MDS cells was transferred into BM MSCs and could reduce stem/progenitor cell-supporting capacity of MSCs via PCBP1 reduction. EV miR-7977 could be involved in the dysfunction of normal hematopoiesis in AML and MDS. Acknowledgments We would like to thank the staff at the division of electron microscopy in Sapporo Medical University for their technical support. We also thank Yumiko Kaneko for preparation of the BM MNSs from patient samples. The manuscript has been carefully reviewed by an experienced medical editor at NAI Inc. The extracellular vesicle miRNAs have been deposited into NCBI’s Gene Expression Omnibus under the accession code GSE64029. Funding This work was supported in part by a grant from the Ministry of Health, Labour and Welfare of Japan to M.K. (ID: 15K09482).

ogenous leukemia. Nat Med. 2003;9(9): 1158-1165. 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. Guidetti F, Grazioli S, Capelli F, et al. Primitive hematopoietic stem cells shows a polyclonal pattern in myelodysplastic syndromes. Haematologica. 2004;89(1):21-28. 6. Karasawa M, Tsukamoto N, Sakai H, et al. Clinical outcome in three patients with myelodysplastic syndrome showing polyclonal hematopoiesis. Acta Haematol. 1999;101(1):46-49.

7. Cancer Genome Atlas Research N. Genomic and epigenomic landscapes of adult de novo acute myeloid leukemia. N Engl J Med. 2013;368(22):2059-2074. 8. Busque L, Patel JP, Figueroa ME, et al. Recurrent somatic TET2 mutations in normal elderly individuals with clonal hematopoiesis. Nat Genet. 2012;44(11): 1179-1181. 9. Avecilla ST, Hattori K, Heissig B, et al. Chemokine-mediated interaction of hematopoietic progenitors with the bone marrow vascular niche is required for thrombopoiesis. Nat Med. 2004;10(1):64-71. 10. Ding L, Saunders TL, Enikolopov G, Morrison SJ. Endothelial and perivascular

haematologica | 2016; 101(4)


miR-7977 modulates PCBP1 level in BM MSCs

11. 12. 13.

14.

15.

16.

17. 18.

19.

20.

21.

22.

23.

24.

cells maintain haematopoietic stem cells. Nature. 2012;481(7382):457-462. Asada N, Katayama Y. Regulation of hematopoiesis in endosteal microenvironments. Int J Hematol. 2014;99(6):679-684. Morrison SJ, Scadden DT. The bone marrow niche for haematopoietic stem cells. Nature. 2014;505(7483):327-334. Lanotte M, Allen TD, Dexter TM. Histochemical and ultrastructural characteristics of a cell line from human bone-marrow stroma. J Cell Sci. 1981;50:281-297. Raaijmakers MH, Mukherjee S, Guo S, et al. Bone progenitor dysfunction induces myelodysplasia and secondary leukaemia. Nature. 2010;464(7290):852-857. Kobune M, Iyama S, Kikuchi S, et al. Stromal cells expressing hedgehog-interacting protein regulate the proliferation of myeloid neoplasms. Blood Cancer J. 2012;2:e87. Oliveira FM, Lucena-Araujo AR, Favarin Mdo C, et al. Differential expression of AURKA and AURKB genes in bone marrow stromal mesenchymal cells of myelodysplastic syndrome: correlation with G-banding analysis and FISH. Exp Hematol. 2013;41(2):198-208. Huan J, Hornick NI, Shurtleff MJ, et al. RNA trafficking by acute myelogenous leukemia exosomes. Cancer Res. 2013;73(2):918-929. Gyorgy B, Szabo TG, Pasztoi M, et al. Membrane vesicles, current state-of-the-art: emerging role of extracellular vesicles. Cell Mol Life Sci. 2011;68(16):2667-2688. Raimondo S, Saieva L, Corrado C, et al. Chronic myeloid leukemia-derived exosomes promote tumor growth through an autocrine mechanism. Cell Commun Signal. 2015;13:8. Boyiadzis M, Whiteside TL. Information transfer by exosomes: A new frontier in hematologic malignancies. Blood Rev. 2015; 29(5):281-290. Yeh YY, Ozer HG, Lehman AM, et al. Characterization of CLL exosomes reveals a distinct microRNA signature and enhanced secretion by activation of BCR signaling. Blood. 2015;125(21):3297-3305. Kawano Y, Kobune M, Yamaguchi M, et al. Ex vivo expansion of human umbilical cord hematopoietic progenitor cells using a coculture system with human telomerase catalytic subunit (hTERT)-transfected human stromal cells. Blood. 2003;101(2):532-540. Kobune M, Ito Y, Kawano Y, et al. Indian hedgehog gene transfer augments hematopoietic support of human stromal cells including NOD/SCID-beta2m-/- repopulating cells. Blood. 2004;104(4): 1002-1009. Kobune M, Kawano Y, Takahashi S, et al. Interaction with human stromal cells enhances CXCR4 expression and engraft-

haematologica | 2016; 101(4)

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

ment of cord blood Lin(-)CD34(-) cells. Exp Hematol. 2008;36(9):1121-1131. Kawano Y, Kobune M, Chiba H, et al. Ex vivo expansion of G-CSF-mobilized peripheral blood CD133+ progenitor cells on coculture with human stromal cells. Exp Hematol. 2006;34(2):150-158. Jeppesen DK, Hvam ML, Primdahl-Bengtson B, et al. Comparative analysis of discrete exosome fractions obtained by differential centrifugation. J Extracell Vesicles. 2014;3: 25011. Grant R, Ansa-Addo E, Stratton D, et al. A filtration-based protocol to isolate human plasma membrane-derived vesicles and exosomes from blood plasma. J Immunol Methods. 2011;371(1-2):143-151. Channavajjhala SK, Rossato M, Morandini F, et al. Optimizing the purification and analysis of miRNAs from urinary exosomes. Clin Chem Lab Med. 2014;52(3):345-354. Kobune M, Xu Y, Baum C, Kelley MR, Williams DA. Retrovirus-mediated expression of the base excision repair proteins, formamidopyrimidine DNA glycosylase or human oxoguanine DNA glycosylase, protects hematopoietic cells from N,N',N"-triethylenethiophosphoramide (thioTEPA)induced toxicity in vitro and in vivo. Cancer Res. 2001;61(13):5116-5125. Kosaka N, Iguchi H, Yoshioka Y, Takeshita F, Matsuki Y, Ochiya T. Secretory mechanisms and intercellular transfer of microRNAs in living cells. J Biol Chem. 2010;285(23): 17442-17452. Chiba H, Kobune M, Kato J, et al. Wnt3 modulates the characteristics and cobblestone area-supporting activity of human stromal cells. Exp Hematol. 2004;32(12): 1194-1203. Yao Y, Wang C, Wei W, et al. Dendritic cells pulsed with leukemia cell-derived exosomes more efficiently induce antileukemic immunities. PLoS One. 2014;9(3):e91463. Balakrishnan I, Yang X, Brown J, et al. Genome-wide analysis of miRNA-mRNA interactions in marrow stromal cells. Stem Cells. 2014;32(3):662-673. Ji X, Wan J, Vishnu M, Xing Y, Liebhaber SA. alphaCP Poly(C) binding proteins act as global regulators of alternative polyadenylation. Mol Cell Biol. 2013;33 (13):2560-2573. Yamauchi A, Ito Y, Morikawa M, et al. Preadministration of angiopoietin-1 followed by VEGF induces functional and mature vascular formation in a rabbit ischemic model. J Gene Med. 2003;5(11):994-1004. Kode A, Manavalan JS, Mosialou I, et al. Leukaemogenesis induced by an activating beta-catenin mutation in osteoblasts. Nature. 2014;506(7487):240-244. Boulais PE, Frenette PS. Making sense of

38.

39.

40.

41.

42.

43.

44.

45. 46.

47. 48.

49.

50.

hematopoietic stem cell niches. Blood. 2015;125(17):2621-2629. Medyouf H, Mossner M, Jann JC, et al. Myelodysplastic cells in patients reprogram mesenchymal stromal cells to establish a transplantable stem cell niche disease unit. Cell Stem Cell. 2014;14(6):824-837. Deeg HJ, Beckham C, Loken MR, et al. Negative regulators of hemopoiesis and stroma function in patients with myelodysplastic syndrome. Leuk Lymphoma. 2000;37(3-4):405-414. Blau O, Baldus CD, Hofmann WK, et al. Mesenchymal stromal cells of myelodysplastic syndrome and acute myeloid leukemia patients have distinct genetic abnormalities compared with leukemic blasts. Blood. 2011;118(20):5583-5592. Santamaria C, Muntion S, Roson B, et al. Impaired expression of DICER, DROSHA, SBDS and some microRNAs in mesenchymal stromal cells from myelodysplastic syndrome patients. Haematologica. 2012;97(8): 1218-1224. Schepers K, Pietras EM, Reynaud D, et al. Myeloproliferative neoplasia remodels the endosteal bone marrow niche into a selfreinforcing leukemic niche. Cell Stem Cell. 2013;13(3):285-299. Zhang B, Ho YW, Huang Q, et al. Altered microenvironmental regulation of leukemic and normal stem cells in chronic myelogenous leukemia. Cancer Cell. 2012;21(4):577592. Roccaro AM, Sacco A, Maiso P, et al. BM mesenchymal stromal cell-derived exosomes facilitate multiple myeloma progression. J Clin Invest. 2013;123(4):1542-1555. Guo WT, Wang XW, Wang Y. Micro-management of pluripotent stem cells. Protein Cell. 2014;5(1):36-47. Huo LR, Zhong N. Identification of transcripts and translatants targeted by overexpressed PCBP1. Biochim Biophys Acta. 2008;1784(11):1524-1533. Ruvolo PP. The Interplay between PP2A and microRNAs in Leukemia. Front Oncol. 2015;5:43. Wang G, Wang Y, Shen C, et al. RNA polymerase II binding patterns reveal genomic regions involved in microRNA gene regulation. PLoS One. 2010;5(11):e13798. Saito Y, Suzuki H, Taya T, et al. Development of a novel microRNA promoter microarray for ChIP-on-chip assay to identify epigenetically regulated microRNAs. Biochem Biophys Res Commun. 2012;426(1):33-37. Xing CY, Hu XQ, Xie FY, et al. Long noncoding RNA HOTAIR modulates c-KIT expression through sponging miR-193a in acute myeloid leukemia. FEBS Lett. 2015;589(15):1981-1987.

447


ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Acute Myeloid Leukemia

Ferrata Storti Foundation

Haematologica 2016 Volume 101(4):448-457

Helicase-like transcription factor is a RUNX1 target whose downregulation promotes genomic instability and correlates with complex cytogenetic features in acute myeloid leukemia Chi Keung Cheng,1 Natalie P. H. Chan,1 Thomas S. K. Wan,1 Lai Ying Lam,1 Coty H. Y. Cheung, 1 Terry H. Y. Wong, 1 Rosalina K. L. Ip,1 Raymond S. M. Wong,2,3 and Margaret H. L. Ng 1,4

Department of Anatomical and Cellular Pathology, Prince of Wales Hospital, The Chinese University of Hong Kong; 2Department of Medicine and Therapeutics, Prince of Wales Hospital, The Chinese University of Hong Kong; 3Sir Y. K. Pao Centre for Cancer, Prince of Wales Hospital, Hong Kong; and 4State Key Laboratory in Oncology in South China, The Chinese University of Hong Kong, Cina 1

ABSTRACT

H

Correspondence: margaretng@cuhk.edu.hk

Received: September 23, 2015. Accepted: January 13, 2016. Pre-published: January 22, 2016. doi:10.3324/haematol.2015.137125

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/4/448

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

448

elicase-like transcription factor is a SWI/SNF chromatin remodeling factor involved in various biological processes. However, little is known about its role in hematopoiesis. In this study, we measured helicase-like transcription factor mRNA expression in the bone marrow of 204 adult patients with de novo acute myeloid leukemia. Patients were dichotomized into low and high expression groups at the median level for clinicopathological correlations. Helicaselike transcription factor levels were dramatically reduced in the low expression patient group compared to those in the normal controls (n=40) (P<0.0001). Low helicase-like transcription factor expression correlated positively with French-American-British M4/M5 subtypes (P<0.0001) and complex cytogenetic abnormalities (P=0.02 for ≥3 abnormalities; P=0.004 for ≥5 abnormalities) but negatively with CEBPA double mutations (P=0.012). Also, low expression correlated with poorer overall (P=0.005) and event-free (P=0.006) survival in the intermediate-risk cytogenetic subgroup. Consistent with the more aggressive disease associated with low expression, helicase-like transcription factor knockdown in leukemic cells promoted proliferation and chromosomal instability that was accompanied by downregulation of mitotic regulators and impaired DNA damage response. The significance of helicase-like transcription factor in genome maintenance was further indicated by its markedly elevated expression in normal human CD34+ hematopoietic stem cells. We further demonstrated that helicase-like transcription factor was a RUNX1 target and transcriptionally repressed by RUNX1-ETO and site-specific DNA methylation through a duplicated RUNX1 binding site in its promoter. Taken together, our findings provide new mechanistic insights on genomic instability linked to helicase-like transcription factor deregulation, and strongly suggest a tumor suppressor function of the SWI/SNF protein in acute myeloid leukemia.

Introduction SWI/SNF proteins are chromatin remodeling factors that use the energy of ATP hydrolysis to remodel nucleosomes and regulate transcription. These proteins are essential for lineage specification and stem cell maintenance.1 Emerging evidence haematologica | 2016; 101(4)


HLTF regulates genomic stability in AML

has also suggested non-redundant roles of these proteins in normal and malignant hematopoiesis. It has been shown that BAF53a is required for hematopoietic stem/progenitor cell maintenance,2 whereas BRG1 is necessary for erythroid and granulocytic differentiation.3,4 On the other hand, ARID1A and SNF5 are disrupted in Burkitt lymphoma and chronic myeloid leukemia, respectively.5,6 The significance of SWI/SNF proteins in hematopoiesis is further exemplified by their interaction with key hematopoietic transcription factors to govern lineage-specific gene expression.7 Helicase-like transcription factor (HLTF) is a SWI/SNF protein initially identified as a transcription factor that binds to several gene promoters and enhancers.8,9 However, growing evidence has indicated its involvement in DNA repair. HLTF has been suggested as the yeast Rad5 homolog, which governs postreplication repair of damaged DNA and its downregulation increases mutagenesis and chromosome abnormalities.10,11 Consistent with this tumor suppressive function, HLTF is frequently silenced by promoter hypermethylation in colon and gastric cancer and its restoration inhibits proliferation.12,13 In addition, HLTF expression is severely reduced in certain melanoma and lung cancer cell lines.14 However, it has also been shown that HLTF upregulation is associated with tumor progression in hypopharyngeal and cervical cancers.15,16 Accordingly, variable prognostic implications of HLTF expression have been reported.17,18 Unlike in solid tumors, the role of HLTF in hematological malignancies remains largely unclear. In this study, we sought to investigate potential HLTF deregulation and its clinicopathological and functional implications in acute myeloid leukemia (AML). We demonstrated that HLTF expression was reduced in adult AML patients with more aggressive disease phenotypes, including complex cytogenetic abnormalities that were recapitulated in HLTF knocked down leukemic cells exhibiting impaired mitotic and DNA repair functions. Further investigation of the HLTF transcriptional control delineated mechanisms involving altered RUNX1 functions that contribute to the HLTF downregulation. Our findings indicate HLTF as a RUNX1 target and that deregulated HLTF may represent a new molecular pathway underlying genomic instability in AML.

Methods

The AML cohort from The Cancer Genome Atlas (TCGA) was used for data validation.19

Cell lines OCI-AML3 was provided by Prof. M.D. Minden (Princess Margaret Cancer Centre, University Health Network, Toronto, Canada). U937T and U937T-AE lines were provided by Prof. D.E. Zhang (Department of Pathology, University of California, San Diego, USA). Other cell lines were obtained commercially. Cells were maintained in RPMI-1640 medium containing 10% fetal bovine serum.

Real-time quantitative PCR (RQ-PCR) Total RNA was extracted and reverse transcribed. RQ-PCR was performed using TaqMan assays (Life Technologies). Each sample was measured in triplicate and expression levels were determined by 2−ΔΔCt. GAPDH was used for normalization. Primer/probe sequences are provided in the Online Supplementary Table S1.

Immunohistochemistry Immunohistochemical staining was performed using BM biopsies from 24 AML patients, with 12 having low and the other 12 having high HLTF mRNA levels. Details are provided in the Online Supplementary Methods.

Generation and analysis of stable HL-60 cell lines HLTF knockdown in HL-60 cells was performed using short hairpin RNA (shRNA). Details of cell line generation and the analysis of cell proliferation, G-banded metaphases and histone H2AX phosphorylation are provided in the Online Supplementary Methods.

DNA constructs, transient transfection and reporter gene assays Details of the constructs are provided in the Online Supplementary Methods. Transient transfection was performed with Lipofectamine 2000 (Life Technologies). Each transfection contained 50ng of pGL3 reporter constructs, 150ng of pCMV expression plasmids, and 4ng of Renilla luciferase plasmid pRL-CMV (Promega). Luciferase activities were measured 24 hours after transfection using the Dual-Glo Luciferase Assay System (Promega).

Methylation-sensitive high resolution melting (MS-HRM) analysis, bisulfite sequencing and methylation-specific PCR (MSP) Bisulfite-modified DNA was prepared using the EZ DNA Methylation Gold Kit (Zymo Research). Detailed procedures are provided in the Online Supplementary Methods.

Patient samples Diagnostic bone marrow (BM) samples from 204 adult patients (≥18 years old) with de novo AML were studied (Table 1). Acute promyelocytic leukemia, therapy-related AML, or AML arising from a prior myelodysplastic syndrome (MDS)/myeloproliferative neoplasm were excluded. Normal BM and peripheral blood samples were obtained from individuals who had no prior history of malignancy. All subjects gave informed consent for the study, which was approved by the Joint CUHK-NTEC Clinical Research Ethics Committee and was in accordance with the Declaration of Helsinki. Treatment details, cytogenetic and mutational studies and immunophenotyping are provided in the Online Supplementary

Statistical analysis To investigate the clinicopathological significance of HLTF expression, the median HLTF mRNA level was used as the cut-off to dichotomize the patient cohort into low and high expression groups. Unpaired t and Fisher’s exact tests were used to analyze the relation with continuous and categorical variables between groups, respectively. Overall survival (OS) and event-free survival (EFS) were defined as previously described.20 Kaplan-Meier curves were compared by log-rank test. Multivariate Cox regression analysis was performed to test the significance of HLTF levels with adjustment for other potential prognostic factors. Two-sided P<0.05 was considered statistically significant. Statistical analyses were performed using SPSS 21.0 (IBM).

Methods Mononuclear cells were isolated from the patient and control samples for nucleic acid extraction. haematologica | 2016; 101(4)

Other methods are provided in the Online Supplementary Methods. 449


C.K. Cheng et al.

Results Adult AML patients with low HLTF expression have distinctive clinicopathological features associated with poorer prognosis Compared to the normocellular BM group (n=40), HLTF mRNA levels were dramatically reduced (P<0.0001) in the low expression patient group but similar to the high expression patient group (Online Supplementary Figure S1). HLTF protein expression measured by immunohistochemical staining of BM biopsies correlated with HLTF mRNA expression in the 24 patients studied (P=0.032 by Pearson correlation) (Online Supplementary Figure S2). Low HLTF

expression in the AML patients correlated positively with elevated lactate dehydrogenase levels (P=0.02) and the French-American-British (FAB) M4/M5 subtypes (P<0.0001), but negatively with CEBPA double mutations (P=0.012) (Table 1; Online Supplementary Table S2). Additionally, low HLTF expression was associated with CD14 (P=0.003), CD15 (P=0.036) and CD20 (P=0.034) positivity on leukemic blasts (Online Supplementary Table S3). Since HLTF downregulation has been shown to induce chromosome aberrations,10 we investigated if HLTF expression correlated with specific cytogenetic features in AML. Notably, patients with low HLTF expression had more frequent multiple chromosome abnormalities (≥ 3

Table 1. Comparison of laboratory, cytogenetic and molecular features between AML patients with low and high HLTF expression.

Parameters

Entire cohort (n=204)

Low HLTF expression (n=102)

High HLTF expression (n=102)

P

Mean age, years (range) Sex, n (% males) Mean hemoglobin, g/dL (range) Mean platelets, 109/L (range) Mean WBC, 109/L (range) Mean LDH, U/L (range) Mean BM blast, % (range) FAB subtypesa, n (%) M0 M1 M2 M4 M5 M6 M7 Missing data Cytogenetic abnormalitiesb, n (%) t(8;21) inv(16)/t(16;16) Normal karyotype 11q23 inv(3)/t(3;3) +8 +11 +13 -5/-5q -7/-7q -17 -X/Y del(9q) Complex (≥3 unrelated abnormalities) Complex (≥5 unrelated abnormalities) Missing data Molecular markers, n (%) FLT3-ITD FLT3-D835/I836 KIT mutations CEBPA double mutations CEBPA single mutations NPM1 mutations DNMT3A mutations WT1 mutations IDH1 mutations IDH2 mutations

52.3 (18-91) 107 (52.5%) 8.1 (2.9-13.6) 61.3 (2-328) 43.3 (0.3-517) 705 (101-5860) 61 (12-98)

52.0 (18-91) 54 (52.9%) 8.1 (2.9-13.6) 62.9 (8-328) 48.5 (0.6-282.2) 845 (132-5860) 62 (12-95)

52.6 (20-86) 53 (52%) 8.1 (4.1-12.9) 59.6 (2-247) 38.1 (0.3-517) 572 (101-3180) 61 (15-98)

0.756 >0.999 0.926 0.642 0.242 0.020* 0.784

8 (4%) 48 (24.1%) 57 (28.6%) 38 (19.1%) 42 (21.1%) 5 (2.5%) 1 (0.5%) 5

3 (3%) 19 (19.2%) 22 (22.2%) 24 (24.2%) 30 (30.3%) 1 (1%) 0 (0%) 3

5 (5%) 29 (29%) 35 (35%) 14 (14%) 12 (12%) 4 (4%) 1 (1%) 2

0.721 0.136 0.060 0.073 0.002* 0.369 >0.999

15 (8.2%) 12 (6.6%) 100 (54.9%) 4 (2.2%) 2 (1.1%) 14 (7.7%) 3 (1.6%) 3 (1.6%) 7 (3.8%) 6 (3.3%) 3 (1.6%) 10 (5.5%) 3 (1.6%) 17 (9.3%) 11 (6.0%) 22

10 (11.5%) 7 (8%) 43 (49.4%) 3 (3.4%) 2 (2.3%) 6 (6.9%) 0 (0%) 3 (3.4%) 5 (5.7%) 5 (5.7%) 1 (1.1%) 7 (8%) 1 (1.1%) 13 (14.9%) 10 (11.5%) 15

5 (5.3%) 5 (5.3%) 57 (60%) 1 (1.1%) 0 (0%) 8 (8.4%) 3 (3.2%) 0 (0%) 2 (2.1%) 1 (1.1%) 2 (2.1%) 3 (3.2%) 2 (2.1%) 4 (4.2%) 1 (1.1%) 7

0.178 0.555 0.180 0.350 0.227 0.785 0.247 0.107 0.262 0.606 >0.999 0.105 >0.999 0.020* 0.004*

43 (21.1%) 12 (5.9%) 9 (4.4%) 22 (10.8%) 11 (5.4%) 52 (25.5%) 40 (19.6%) 12 (5.9%) 12 (5.9%) 25 (12.3%)

22 (21.6%) 6 (5.9%) 6 (5.9%) 5 (4.9%) 6 (5.9%) 28 (27.5%) 23 (22.5%) 5 (4.9%) 6 (5.9%) 10 (9.8%)

21 (20.6%) 6 (5.9%) 3 (2.9%) 17 (16.7%) 5 (4.9%) 24 (23.5%) 17 (16.7%) 7 (6.9%) 6 (5.9%) 15 (14.7%)

>0.999 >0.999 0.498 0.012* >0.999 0.630 0.378 0.768 >0.999 0.394

WBC: white blood cell; LDH, lactate dehydrogenase; BM: bone marrow; FAB: French-American-British; ITD: internal tandem duplications. aP<0.0001 when FAB M4 and M5 are combined. bAll patients with a specific abnormality were counted irrespective of the presence of additional abnormalities. *Statistically significant.

450

haematologica | 2016; 101(4)


HLTF regulates genomic stability in AML

unrelated abnormalities) than patients with high HLTF expression (P=0.02) (Table 1). This association was even more obvious when at least 5 abnormalities were considered as complex (P=0.004). Complex karyotypes in AML have been consistently associated with a very poor prognosis, and defined as the presence of multiple chromosome abnormalities (≥3 or ≥5) in the absence of the prognostically favorable t(8;21), inv(16)/t(16;16) and t(15;17).20 The association of low HLTF expression with a complex karyotype remained significant when patients with the core-binding factor (CBF) translocations were excluded (≥3 abnormalities: 13% vs. 5%, P=0.085; ≥5 abnormalities: 11% vs. 1%, P=0.011). Clonal heterogeneity has recently been identified as an additional unfavorable cytogenetic marker in AML.21,22 The incidence of cytogenetic heterogeneity in the forms of distinct subclones or composite karyotypes was also significantly higher in the low HLTF expression group than in the high expression group (16% vs. 4%, P=0.012).

Survival data were available from 154 of the 204 AML patients who had received standard chemotherapy. Complete remission was achieved in 122 of them (79%). With a mean follow-up time of 24 months, patients with low HLTF expression tended to have worse OS (P=0.127) and EFS (P=0.168) than patients with high HLTF expression when all cases were analyzed. When stratified by cytogenetics, patients with intermediate-risk cytogenetics with low HLTF expression had significantly shorter OS (mean, 16 vs. 24 months, P=0.005) and EFS (mean, 11 vs. 20 months, P=0.006) than those patients with high HLTF expression (Online Supplementary Figure S3). Similar trends were also noted when only cytogenetically normal AML patients were analyzed (OS: mean, 17 vs. 23 months, P=0.037; EFS: mean, 13 vs. 18 months, P=0.069) (Online Supplementary Figure S3). The favorable and adverse cytogenetic risk groups were not analyzed separately because of small sample sizes. In multivariate analysis, low HLTF expression remained as an independent adverse prognos-

A

C B

D

E

haematologica | 2016; 101(4)

Figure 1. HLTF knockdown promotes proliferation and chromosomal instability in HL-60 cells. (A) Western blot analysis of HLTF expression in HL-60_control (empty vector control) and HL-60_shRNA1-3 lines. GAPDH served as a loading control. (B) WST-1 (top) and colony forming assays (bottom) for cell proliferation analysis of HL-60_control and HL60_shRNA1-3 lines. For the latter assay, colonies were lysed and detected by the CyQuant GR Dye in a fluorescence plate reader after 10-day culture in a semi-solid methylcellulose medium. Results are expressed as mean ± SE from triplicate experiments. RFUs represent relative fluorescence units. (C) Western blotting of various mitotic regulators in HL-60_control and HL-60_shRNA1-3 lines. GAPDH served as a loading control. Representative blots from repeated experiments are shown. (D) Phosphorylation of histone H2AX at serine 139 in HL-60_control and HL-60_shRNA1-3 lines after 1-hour MMS (500mM and 1000mM) treatment. The phosphorylation levels were normalized to total histone H2AX levels in each sample. Results are expressed as mean ± SE from triplicate experiments and are relative to the untreated control in each cell line. (E) RQ-PCR analysis of HLTF expression in 14 pairs of normal human CD34+ and CD34–/CD11b+ cell populations. Expression levels were relative to the U937 cell line. As expected, analysis of the same 13 pairs revealed a drastic increase in PU.1 mRNA expression in the CD34–/CD11b+ mature myeloid compartment. * indicates P<0.05 vs. HL-60_control.

451


C.K. Cheng et al. A

B

E

C

F

D

Figure 2. Transcriptional control of HLTF by RUNX1 and RUNX1-ETO. (A) A schematic representation of the human HLTF gene. The location of the duplicated and single RUNX1 binding sites on the HLTF promoter is shown. Dashed, filled and grey arrows indicate the primers used for ChIP assay, bisulfite sequencing and MS-HRM analysis, respectively. Filled boxes represent exons and the exon numbers are indicated. Horizontal arrows below the figure indicate the deletion and mutant promoter constructs analyzed in subsequent panels. The cross indicates a mutation introduced into the duplicated RUNX1 binding site. The first nucleotide of the HLTF mRNA (NM_003071.3) is assigned as +1. (B, top) HLTF (-769/+43) promoter-luciferase construct was co-transfected with RUNX1 (encodes AML1c), RUNX1_R201Q, RUNX1ETO (RE) or CBFB-MYH11 (CM) expression plasmid together with pRL-CMV into K562 cells. Co-transfection with the same amount of empty pCMV vector was done in parallel. Results are presented as fold change by comparing the normalized firefly luciferase activity of the construct co-transfected with the expression plasmid with that co-transfected with the empty vector control. (B, bottom) Western blot analysis of RUNX1, RUNX1-ETO and CBFB-MYH11 expression in the K562 transfectants. GAPDH served as a loading control. (C) HLTF promoter-luciferase constructs were co-transfected with pCMV-RUNX1-ETO and pRL-CMV into K562 cells. Co-transfection with the same amount of empty pCMV was done in parallel. The duplicated RUNX1 site was mutated from 5’-ACCGCAGGCACCGCA-3’ to 5’-GTCGACGGCGTCGAC-3’ in -424_RUNX1mut. Results are expressed as relative promoter activity by comparing the normalized firefly luciferase activity of the construct with the respective promoterless pGL3-Basic control. In all experiments in panels B and C, transfection efficiency was normalized according to the co-transfected pRL-CMV Renilla luciferase activity and results are expressed as mean ± SE from at least triplicate experiments. (D) Chromatin from t(8;21)-positive (Kasumi-1 and SKNO-1) and t(8;21)-negative (THP-1) cell lines was immunoprecipitated with the indicated antibodies and then analyzed by PCR using oligo 1 and oligo 2 primers. The oligo 1 primers amplified a 181-bp HLTF promoter fragment encompassing the duplicated RUNX1 binding site. The oligo 2 primers amplified a distal HLTF region lacking RUNX1 binding sites. Immunoprecipitation with anti-IgG was done as a negative control. (E, top) Confirmation of RUNX1-ETO induction following tetracycline withdrawal (-Tet) in U937T-AE cells by RT-PCR. Amplification of GAPDH was done as internal control. (E, middle) HLTF, SHPRH and RUNX3 levels were measured by RQ-PCR after 24 hours of tetracycline withdrawal in U937T and U937T-AE cell lines. Relative mRNA levels were calculated by comparing them with the expression levels before tetracycline withdrawal and results are presented as mean ± SE from triplicate experiments. (E, bottom) HLTF protein expression was analyzed by Western blotting following tetracycline withdrawal in U937T and U937T-AE cell lines. GAPDH served as a loading control. Representative blots from repeated experiments are shown. *indicates P<0.05. (F) HLTF mRNA levels in diagnostic BM samples from 59 patients with FAB-M2 AML were measured by RQ-PCR. Each circle represents one patient and the number of patients in each group is shown. Horizontal lines indicate the mean HLTF/GAPDH levels. Expression levels were relative to the U937 myeloid cell line.

452

haematologica | 2016; 101(4)


HLTF regulates genomic stability in AML

A

B

C

E

D

F

G

Figure 3. Aberrant CpG methylation at the duplicated RUNX1 binding site of the HLTF promoter causes HLTF repression in AML. (A) A representative normalized melting curve of MS-HRM analysis of HLTF promoter methylation in the ME-1 cell line (indicated by an arrow). The MS-HRM curves derived from the methylation standards (0%, 2.5%, 5%, 10% and 100%) are also shown. (B) Bisulfite sequencing of the HLTF promoter in OCI-AML3 and ME-1 cell lines. Each row of squares represents one PCR clone. Open and filled squares represent unmethylated and methylated CpG dinucleotides, respectively. Each CpG dinucleotide is numbered (no. 1-18). (C) Nucleotide sequence of the HLTF promoter region analyzed by bisulfite sequencing. The duplicated RUNX1 binding site is boxed. Arrows indicate the locations of the MSP primers. (D, top) RQ-PCR analysis of HLTF expression in AML cell lines after 4-day treatment with 2mM of 5’-AZA. Drugs were replenished at day 2 of the treatment. Results are presented as relative expression levels by comparing the normalized HLTF levels in the 5’-AZA-treated groups with those in the respective control group (DMSO-treated). Results are expressed as mean ± SE from duplicate experiments, each done in triplicate. *indicates P<0.05. (D, bottom) MSP analysis revealed partial demethylation of the duplicated RUNX1 site after 4-day 5’-AZA treatment (2mM) in OCI-AML3 and ME-1 cells. M and U represent PCR using primers specific for the methylated and unmethylated sequences, respectively. (E) EMSA analysis of RUNX1 binding to the HLTF promoter. Left, the specificity of the DNAprotein complex (indicated by an arrow) was confirmed by the addition of cold competitors and a RUNX1 antibody. For the supershift assays, nuclear extract (NE) was pre-incubated with 1ml of antibody at room temperature for 30 mins before probe addition. Right, the use of methylated EMSA probe (m-probe) abrogated RUNX1 binding to the HLTF promoter sequence. u-probe indicates the unmethylated probe. (F, top) Representative MSP analysis of CpG methylation of the duplicated RUNX1 site in the HLTF promoter in BM samples from 8 AML patients. N and P represent normocellular BM and the positive control OCI-AML3 cells, respectively. (F, bottom) Validation of the MSP results by bisulfite sequencing in an AML patient sample showing aberrant CpG methylation at the duplicated RUNX1 binding site. (G) HLTF mRNA levels in FAB-M1 patients with respect to the DNA methylation status of the duplicated RUNX1 binding site in the HLTF promoter. HLTF expression was determined by RQ-PCR and normalized to GAPDH. Expression levels were relative to the U937 myeloid cell line. Each triangle represents one patient, and the number of patients in each group is shown. Patients with RUNX1 mutations (n=6) were excluded from the analysis. Horizontal lines indicate the mean HLTF/GAPDH levels. M and U represent patients with and without methylation of the duplicated RUNX1 site, respectively.

haematologica | 2016; 101(4)

453


C.K. Cheng et al.

tic factor for EFS in patients with intermediate-risk cytogenetics (P=0.032, hazard ratio=1.755, 95% confidence interval=1.05-2.931) (Online Supplementary Table S4). Low HLTF expression was also found to correlate with poorer OS in cytogenetically normal AML patients from the TCGA validation cohort (P=0.006) (Online Supplementary Figure S4).

HLTF knockdown promotes proliferation and chromosomal instability in HL-60 cells To study the functional role of HLTF in AML, HLTF expression was stably knocked down in HL-60 cells. As shown in Figure 1A, HLTF protein levels were barely detectable in cells transfected with HLTF shRNAs (HL60_shRNA1-3) as compared to the empty vector control (HL-60_control). Downregulation of HLTF expression significantly increased HL-60 proliferation in both WST-1 and colony-forming assays (Figure 1B). No apparent effect on HL-60 differentiation, as determined by MayGrunwald-Giemsa-stained morphological assessment, myeloperoxidase staining and flow cytometric analysis of cell surface marker expression (CD11b, CD14 and CD15), was observed following the knockdown (data not shown). Cytogenetic analysis of the HL-60_control and HL60_shRNA1-3 lines revealed similar chromosome abnormalities including +18, del(9)(p21) and two rearrangements involving chromosome 5 as reported in the parental line (Online Supplementary Table S5; Online Supplementary Figures S5-S8).23 Interestingly, knockdown of HLTF expression led to the emergence of a cytogenetic subclone (HL60_shRNA1) or new clone (HL-60_shRNA2) with the acquisition of common clonal aberrations such as add(10)(p11.2) as well as the loss of der(10;13) and trisomy 13 (Online Supplementary Table S5; Online Supplementary Figures S5-S8). Moreover, a tetraploid subclone developed in the HL-60_shRNA3 line. Western blot analysis showed that HLTF knockdown in the three HL-60 lines was associated with downregulated expression of STAG2, RAD21, SMC2 and AURKB (Figure 1C), which play important roles in chromosome packaging and segregation during cell division. A key early event in DNA damage response is phosphorylation of histone H2AX at serine 139, which marks the damaged sites and signals the recruitment of DNA repair machinery.24 Methyl methanesulfonate (MMS), an alkylating agent, can induce DNA breakage and H2AX phosphorylation. As shown in Figure 1D, knockdown of HLTF expression attenuated H2AX phosphorylation in HL-60 cells after MMS treatment, indicating an impaired DNA damage response. To gain insights into the role of HLTF in myelopoiesis, we measured HLTF mRNA levels in human CD34+ hematopoietic stem cells (HSCs) and CD34â&#x20AC;&#x201C;/CD11b+ cells isolated from 14 normocellular BM samples. In all the paired samples, HLTF expression was high in the HSC population but decreased dramatically (3.8- to 18.7-fold) in mature myeloid cells (Figure 1E). Using HemaExplorer,25 we found that HLTF is also highly expressed in common myeloid progenitors, granulocyte-monocyte progenitors and promyelocytes. Consistently, HLTF expression drops sharply as promyelocytes differentiate into myelocytes and decreases further in polymorphonuclear cells.

HLTF is a RUNX1 target and transcriptionally repressed by RUNX1-ETO in AML Since deregulated HLTF mRNA expression had signifi454

cant clinicopathological implications in AML, we investigated the transcriptional control of HLTF. Previous studies using ChIP-sequencing have suggested HLTF as a putative RUNX1 target in t(4;11) leukemias.26 Accordingly, analysis of the HLTF 5â&#x20AC;&#x2122;-flanking sequences revealed a single and a duplicated RUNX1 consensus binding site in the proximal promoter region (Figure 2A). The leukemogenic RUNX1ETO fusion protein associated with t(8;21)-AML interacts with the same consensus binding sites as RUNX1 and typically acts as a dominant mutant to repress RUNX1 targets.27 To examine potential regulation of the HLTF promoter by RUNX1 and RUNX1-ETO, we constructed a luciferase reporter plasmid carrying an 812-bp HLTF promoter fragment (nucleotide (nt) -769 to +43) that encompasses the two RUNX1 binding sites. Overexpression of RUNX1 (the longest AML1c isoform) activated the HLTF promoter in K562 cells (Figure 2B). This activation was impaired when the RUNX1_R201Q mutant with defective transcriptional activity28 was used (Figure 2B). On the other hand, ectopic expression of RUNX1-ETO repressed the HLTF promoter (Figure 2B) and similar results were obtained from HeLa cells (data not shown). In contrast, cotransfection with pCMV-CBFB-MYH11 encoding the related inv(16) fusion protein failed to repress the HLTF promoter (Figure 2B). Also, the co-transfection did not affect the RUNX1-mediated HLTF activation (Online Supplementary Figure S9). Deletion mapping showed that sequences between nt -769 and -424, which contain the single RUNX1 binding site, were not important for basal HLTF promoter activity and RUNX1-ETO repression (Figure 2C). However, further deletion to nt -196 or mutation of the duplicated RUNX1 site (-424_RUNX1mut) drastically reduced HLTF transcription and attenuated the RUNX1-ETO effect (Figure 2C). ChIP assays in t(8;21)negative (THP-1) and t(8;21)-positive (Kasumi-1 and SKNO-1) cell lines confirmed the specific binding of RUNX1 and RUNX1-ETO to the HLTF promoter (Figure 2D). We next employed the inducible U937T-AE model29 to study HLTF repression by RUNX1-ETO. The induction of RUNX1-ETO reduced the mRNA levels of HLTF and RUNX3 (a known transcriptional repression target of the fusion protein30) but not the related SHPRH (Figure 2E), which is another Rad5 homolog in human cells.10 The downregulation of HLTF protein expression was also confirmed in U937T-AE cells after tetracycline removal (Figure 2E). We found no obvious changes in RAD21, SMC2, STAG2 and AURKB protein expression in U937TAE cells upon RUNX1-ETO induction, suggesting that these genes may not be directly regulated by the fusion protein (Online Supplementary Figure S10). t(8;21)-AML is associated with the FAB M2 subtype. The comparison of HLTF mRNA expression revealed significantly reduced levels in t(8;21)-FAB M2 patients than non-t(8;21)-FAB M2 patients (P=0.025) (Figure 2F). Collectively, these findings indicate that RUNX1-ETO represses HLTF through direct binding to the duplicated RUNX1 site in the HLTF promoter.

The duplicated RUNX1 binding site in the HLTF promoter is aberrantly methylated in AML Since HLTF is silenced by promoter hypermethylation in certain human cancers,12,13 we evaluated this epigenetic change in AML using a MS-HRM approach with a sensitivity of about 2.5%. In total, we analyzed BM samples haematologica | 2016; 101(4)


HLTF regulates genomic stability in AML

from 133 adult AML patients in the original cohort (Online Supplementary Table S6), 7 AML cell lines and 10 normocellular BM samples. Aberrant HLTF promoter methylation was detected in one AML patient (~41% CpG methylation) (Online Supplementary Figure S11). None of the 10 normocellular BM samples exhibited the methylation. Among the 7 AML cell lines (THP-1, Kasumi-1, CMK, MV4-11, HL-60, ME-1 and OCI-AML3), two of them (OCI-AML3 and ME-1) also displayed abnormal MS-HRM melting profiles (Figure 3A). Intriguingly, in both cell lines, the methylation was restricted to four CpG sites (no. 1-4), which cluster along the duplicated RUNX1 binding site (Figure 3B,C). Treatment with 5-aza-2’-deoxycytidine (5’-AZA) increased HLTF mRNA expression in the methylated cell lines but not the unmethylated THP-1 cells (Figure 3D). MSP assays showed that the treatment caused partial demethylation of the duplicated RUNX1 site in both OCIAML3 and ME-1 cells (Figure 3D). We next investigated whether such site-specific DNA methylation affects the binding of the cognate transcription factor RUNX1 using EMSAs. As shown in Figure 3E, a specific DNA-protein complex was formed when the duplicated RUNX1 site in the probe was unmethylated. However, the complex formation was profoundly inhibited when the four CpG sites clustered along the duplicated RUNX1 site were methylated (Figure 3E). Collectively, these findings indicate that methylation of the duplicated RUNX1 site blocks RUNX1 binding to the HLTF promoter and results in transcriptional inhibition of the HLTF gene. Using the MSP assays, we found aberrant CpG methylation of the duplicated RUNX1 site in 4.5% (6 out of 133) of the adult AML patients (Figure 3F). All the positive cases belong to the FAB M1 subtype, accounting for 17% (6 out of 36) of the M1 cases. Such methylation was undetected in 16 normal (8 normocellular BM and 8 peripheral blood) samples examined. Those M1-AML patients with this sitespecific methylation had reduced HLTF expression compared to those patients without the methylation, although the difference was not statistically significant in this small cohort (P=0.167) (Figure 3G). To validate our findings, we analyzed HLTF promoter methylation data (HumanMethylation450) in 100 AML patients from TCGA (Online Supplementary Table S7). Concordantly, the duplicated RUNX1 binding site was found to be methylated in 7 cases (Online Supplementary Table S8), with 6 of them being classified as FAB-M1 (representing 26% of the M1 cases in the validation cohort) and the remaining case as FAB-M0. Other CpG sites in the proximal promoter and 5’-untranslated region were unmethylated in all the samples. Since RUNX1 is mutated in AML and the mutations generally abolish RUNX1 transactivation potential,28,31 we examined the correlation of HLTF expression with RUNX1 mutations in the 133 adult AML patients. RUNX1 mutations were detected in 12% (16 out of 133) of the cases. Patients with mutated RUNX1 tended to have lower HLTF mRNA levels than patients with wild-type RUNX1 (P=0.099) (Online Supplementary Figure S12).

HLTF is rarely mutated in AML Since SWI/SNF mutations are found in human cancers,1 we investigated whether HLTF is mutated in AML. Of the 133 adult AML patient samples tested for HLTF promoter methylation, 132 cases were available for mutation screening. Excluding 5 single nucleotide polymorphisms haematologica | 2016; 101(4)

(p.D269N, p.T303T, p.N311S, p.A776A and p.T900S) detected, 8 HLTF sequence variations were identified with 6 of them being novel (p.M1_W3del, p.D282N, p.G465W, p.W582*, p.T641Nfs*3 and p.T862T) (Online Supplementary Table S9). Only the frameshift mutation (p.T641Nfs*3) was confirmed to be somatic as the change disappeared in the remission BM sample.

Discussion Chromosomal abnormalities are considered the most significant prognostic determinants in AML, yet the underlying mechanisms have remained elusive. TP53 (encoding p53) alterations have been regarded as an important pathway responsible for the marked genomic instability in complex karyotype AML, as they are frequent and correlate with genomic complexity in this AML subtype.32 Herein we showed that reduced expression of the SWI/SNF gene HLTF was also closely associated with complex cytogenetic abnormalities in adult AML patients. Consistently, we showed that HLTF knockdown in leukemic cells induced structural and numerical chromosome defects. The increased chromosomal instability associated with HLTF downregulation may lead to clonal heterogeneity, with individual subclones developing drug resistance and causing relapse, thereby exacerbating AML outcomes. Complex cytogenetic abnormalities are also found in CBF-AML and have been shown to adversely affect long-term survival in some studies.33,34 Interestingly, we observed that among our CBF-AML patients (n=27), lower HLTF expressors also had more secondary chromosome abnormalities (1.46 vs. 0.5) and higher chances of carrying a complex karyotype (23% vs. 0%) involving three or more additional aberrations than higher HLTF expressors. These observations suggest broad impacts of HLTF reduction on chromosomal instability across AML subtypes. Previous studies have shown that submicroscopic copy number alterations are present in about half of the cytogenetically normal AML patients,35 and whether HLTF also impacts on these alterations warrants further investigation. Regarding prognostication, we revealed in this study an inverse correlation between reduced HLTF and CEBPA double mutations in adult AML patients. Since CEBPA double mutations are associated with a favorable prognosis,36,37 these confounding interactions may partly contribute to the adverse prognostic effects of low HLTF expression in AML, particularly in the intermediate-risk cytogenetic subgroup, where CEBPA mutations are most prevalent.37 Larger studies are needed to confirm these associations. Previous studies have revealed that the mRNA expression of cohesin and condensin genes is downregulated in Hltf-deficient cells,38 suggesting that HLTF may regulate these genes at the transcriptional level. The consensus HLTF DNA-binding sequence has been reported to be 5’MCWTDK3’.39 Interestingly, the 5’-flanking region of STAG2, RAD21 and SMC2 contains multiple HLTF binding sites. Moreover, we observed a strong positive correlation between the mRNA expression of HLTF and the three genes in 173 AML patient samples from TCGA. These findings suggest that HLTF may transcriptionally regulate STAG2, RAD21 and SMC2 in a coordinated manner to govern genome functions. Also, our present data suggested that HLTF regulates DNA damage response as its down455


C.K. Cheng et al.

regulation impairs H2AX phosphorylation in the presence of DNA damage. This defect will likely compromise DNA repair and induce mutations and/or chromosome aberrations, thereby promoting genomic instability. Thus, it is conceivable that AML leukemic cells deficient in HLTF are prone to acquiring genetic alterations. ATM and ATR are two major kinases that phosphorylate H2AX,24 and interestingly, their expressions are found to be reduced in Hltfdepleted cells.40 Of note, the cohesin and condensin complexes are also strongly implicated in other processes including DNA repair and gene regulation, indicating that HLTF deregulation has diverse effects on genome functions. It could be postulated that the high-level HLTF expression in normal human CD34+ cells is required for HSC functions by preventing genomic instability and malignant transformation. Herein we demonstrated that HLTF is a direct transcriptional target of RUNX1. Like HLTF, RUNX1 impairment has been linked to an increase in genomic instability.41 Interestingly, emerging evidence has indicated a functional interplay between RUNX1 and p53. It has been shown that RUNX1 can stimulate p53 activity and is recruited onto p53 target promoters in response to DNA damage.42 Likewise, RUNX1 and p53 synergistically activate the transcription of GADD45A, encoding a sensor of DNA stress, and MDS/AML patients with mutated RUNX1 show decreased GADD45A expression.43 These observations are further corroborated by the inability of RUNX1 to induce cellular senescence in murine fibroblasts with defective p53.44 The role of HLTF in genome maintenance and its intimate relationship with RUNX1 raise the possibility that a RUNX1/HLTF axis may collaborate with p53 in AML leukemogenesis. It has been shown that the RUNX1-ETO fusion protein induces a mutator phenotype by downregulating DNA repair genes mainly in the base excision repair pathway.45 Reduced HLTF expression has been shown to increase spontaneous and damage-induced mutagenesis.10,11 Our findings that HLTF is transcriptionally repressed by RUNX1-ETO thus provide further mechanistic insights on how the fusion protein predisposes to the acquisition of additional mutations. On the other hand, our data suggested that the duplicated RUNX1 binding site in the HLTF promoter was not repressed by the related CBFB-MYH11 protein. In fact, it was recently found that CBFB-MYH11 cooperated with RUNX1 and a distinct set of regulators to preferentially activate gene expression.46 More interestingly, our data revealed differential patterns of HLTF promoter methylation in solid tumor and blood cancers. While the methylation spans along the entire HLTF promoter in solid tumors,12,13 HLTF promoter methylation in AML clusters along the duplicated RUNX1 binding site. Importantly, such methylation seems sufficient to abrogate RUNX1 transcriptional functions and thus may represent a novel mechanism deregulating RUNX1 in AML. Although the mechanisms underlying this site-specific DNA methylation are unclear, it was noted that all the six M1-AML cases

References 1. Wilson BG, Roberts CW. SWI/SNF nucleosome remodellers and cancer. Nat Rev Cancer. 2011;11(7):481-492. 2. Krasteva V, Buscarlet M, Diaz-Tellez A, Bernard MA, Crabtree GR, Lessard JA. The

456

with the methylation harbor IDH1/2 mutations as compared to only 16.7% mutation in M1-AML cases lacking the methylation. Since IDH1/2 mutations induce DNA hypermethylation,47 it is possible that the aberrant HLTF methylation is related to mutant IDH1/2 activities. RUNX1 activates transcription by recruiting coactivators including p300.48 Interestingly, a putative p300 binding site is found adjacent to the duplicated RUNX1 site. Mutation of this p300 site repressed the HLTF promoter, whereas p300 overexpression activated the promoter (Cheng et al., 2016, unpublished observations). These findings strongly suggest that RUNX1 and p300 cooperatively regulate HLTF transcription. It is worth noting that p300 and MOZ (another RUNX1 coactivator) are predominantly disrupted in FABM4/M5 AML.49 Since HLTF expression is also reduced in these AML subtypes, the reduction may be due to altered RUNX1 coactivator activities. On the other hand, the reduction may be due to microRNA-mediated gene repression as the 3â&#x20AC;&#x2122;-untranslated region of HLTF contains two putative binding sites for microRNA-155, which is specifically over-expressed in FAB-M4/M5 AML.50 These observations suggest that mechanisms other than altered RUNX1 functions also contribute to HLTF downregulation in AML. While SWI/SNF mutations are common in certain human cancers,1 HLTF is rarely mutated in AML and only detected in 1 out of 132 (~0.8%) adult AML patients. The mutation is expected to produce a truncated protein lacking the RING domain and the last three helicase domains. Since the RING domain is implicated in proliferating cell nuclear antigen ubiquitination,10 the loss of this domain may impair the postreplication repair function. Interestingly, the nonsense c.1746G>A variant (Online Supplementary Table S9) also seems to affect the DNA repair function by protein truncation. However, the significance of other identified variants is unclear. It remains to be determined whether HLTF variants affect leukemia risk and phenotype. In conclusion, our findings indicate that HLTF is a novel RUNX1 target and a potentially important regulator of genomic stability in AML. HLTF contains multiple discrete domains, in particular a unique DNA-binding domain that is absent in most SWI/SNF proteins.8 Characterizing the molecular functions of these domains shall help decipher its role in leukemogenesis and identify targets for therapeutic intervention. Acknowledgments The authors would like to thank Prof. D.E. Zhang, Prof. S.W. Hiebert, and Prof. M.D. Minden for providing valuable expression plasmids and cell lines. The authors also thank Yonna Leung, Alan Yau, T.K. Kwan and Simon Yau for their technical assistance. Funding This work was supported in part by a direct research grant from The Chinese University of Hong Kong, Hong Kong (Project No. 2041392).

BAF53a subunit of SWI/SNF-like BAF complexes is essential for hemopoietic stem cell function. Blood. 2012;120(24):4720-4732. 3. Bultman SJ, Gebuhr TC, Magnuson T. A Brg1 mutation that uncouples ATPase activity from chromatin remodeling reveals an essential role for SWI/SNF-related complex-

es in beta-globin expression and erythroid development. Genes Dev. 2005;19(23):28492861. 4. Vradii D, Wagner S, Doan DN, et al. Brg1, the ATPase subunit of the SWI/SNF chromatin remodeling complex, is required for myeloid differentiation to granulocytes. J

haematologica | 2016; 101(4)


HLTF regulates genomic stability in AML

Cell Physiol. 2006;206(1):112-118. 5. Giulino-Roth L, Wang K, MacDonald TY, et al. Targeted genomic sequencing of pediatric Burkitt lymphoma identifies recurrent alterations in antiapoptotic and chromatinremodeling genes. Blood. 2012;120(26): 5181-5184. 6. Grand F, Kulkarni S, Chase A, Goldman JM, Gordon M, Cross NC. Frequent deletion of hSNF5/INI1, a component of the SWI/SNF complex, in chronic myeloid leukemia. Cancer Res. 1999;59(16):3870-3874. 7. Bakshi R, Hassan MQ, Pratap J, et al. The human SWI/SNF complex associates with RUNX1 to control transcription of hematopoietic target genes. J Cell Physiol. 2010;225(2):569-576. 8. Sheridan PL, Schorpp M, Voz ML, Jones KA. Cloning of an SNF2/SWI2-related protein that binds specifically to the SPH motifs of the SV40 enhancer and to the HIV-1 promoter. J Biol Chem. 1995;270(9):4575-4587. 9. Ding H, Benotmane AM, Suske G, Collen D, Belayew A. Functional interactions between Sp1 or Sp3 and the helicase-like transcription factor mediate basal expression from the human plasminogen activator inhibitor1 gene. J Biol Chem. 1999;274(28):1957319580. 10. Motegi A, Liaw HJ, Lee KY, et al. Polyubiquitination of proliferating cell nuclear antigen by HLTF and SHPRH prevents genomic instability from stalled replication forks. Proc Natl Acad Sci USA. 2008;105(34):12411-12416. 11. Lin JR, Zeman MK, Chen JY, Yee MC, Cimprich KA. SHPRH and HLTF act in a damage-specific manner to coordinate different forms of postreplication repair and prevent mutagenesis. Mol Cell. 2011;42 (2):237-249. 12. Moinova HR, Chen WD, Shen L, et al. HLTF gene silencing in human colon cancer. Proc Natl Acad Sci U S A. 2002;99(13):4562-4567. 13. Leung WK, Yu J, Bai AH, et al. Inactivation of helicase-like transcription factor by promoter hypermethylation in human gastric cancer. Mol Carcinog. 2003;37(2):91-97. 14. MacKay C, Toth R, Rouse J. Biochemical characterisation of the SWI/SNF family member HLTF. Biochem Biophys Res Commun. 2009;390(2):187-191. 15. Capouillez A, Decaestecker C, Filleul O, et al. Helicase-like transcription factor exhibits increased expression and altered intracellular distribution during tumor progression in hypopharyngeal and laryngeal squamous cell carcinomas.Virchows Arch. 2008;453 (5):491-499. 16. Capouillez A, Noël JC, Arafa M, et al. Expression of the helicase-like transcription factor and its variants during carcinogenesis of the uterine cervix: implications for tumour progression. Histopathology. 2011;58(6):984-988. 17. Wallner M, Herbst A, Behrens A, et al. Methylation of serum DNA is an independent prognostic marker in colorectal cancer. Clin Cancer Res. 2006;12(24):7347-7352. 18. Capouillez A, Debauve G, Decaestecker C, et al. The helicase-like transcription factor is a strong predictor of recurrence in hypopharyngeal but not in laryngeal squamous cell carcinomas. Histopathology. 2009;55(1):7790. 19. Cancer Genome Atlas Research Network. Genomic and epigenomic landscapes of adult de novo acute myeloid leukemia. N Engl J Med. 2013;368(22):2059-2074. 20. 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.

haematologica | 2016; 101(4)

21. Bochtler T, Stölzel F, Heilig CE, et al. Clonal heterogeneity as detected by metaphase karyotyping is an indicator of poor prognosis in acute myeloid leukemia. J Clin Oncol. 2013;31(31):3898-3905. 22. Medeiros BC, Othus M, Fang M, Appelbaum FR, Erba HP. Cytogenetic heterogeneity negatively impacts outcomes in patients with acute myeloid leukemia. Haematologica. 2015;100(3):331-335. 23. Liang JC, Ning Y, Wang RY, et al. Spectral karyotypic study of the HL-60 cell line: detection of complex rearrangements involving chromosomes 5, 7, and 16 and delineation of critical region of deletion on 5q31.1. Cancer Genet Cytogenet. 1999;113(2):105-109. 24. Kinner A, Wu W, Staudt C, Iliakis G. Gamma-H2AX in recognition and signaling of DNA double-strand breaks in the context of chromatin. Nucleic Acids Res. 2008;36(17):5678-5694. 25. Bagger FO, Rapin N, Theilgaard-Mönch K, et al. HemaExplorer: a database of mRNA expression profiles in normal and malignant haematopoiesis. Nucleic Acids Res. 2013;41:D1034-1039. 26. Wilkinson AC, Ballabio E, Geng H, et al. RUNX1 is a key target in t(4;11) leukemias that contributes to gene activation through an AF4-MLL complex interaction. Cell Rep. 2013;3(1):116-127. 27. Downing JR. The AML1-ETO chimaeric transcription factor in acute myeloid leukaemia: biology and clinical significance. Br J Haematol. 1999;106(2):296-308. 28. Michaud J, Wu F, Osato M, et al. In vitro analyses of known and novel RUNX1/AML1 mutations in dominant familial platelet disorder with predisposition to acute myelogenous leukemia: implications for mechanisms of pathogenesis. Blood. 2002;99(4):1364-1372. 29. Burel SA, Harakawa N, Zhou L, Pabst T, Tenen DG, Zhang DE. Dichotomy of AML1-ETO functions: growth arrest versus block of differentiation. Mol Cell Biol. 2001;21(16):5577-5590. 30. Cheng CK, Li L, Cheng SH, et al. Transcriptional repression of the RUNX3/AML2 gene by the t(8;21) and inv(16) fusion proteins in acute myeloid leukemia. Blood. 2008;112(8):3391-3402. 31. Harada H, Harada Y, Niimi H, Kyo T, Kimura A, Inaba T. High incidence of somatic mutations in the AML1/RUNX1 gene in myelodysplastic syndrome and low blast percentage myeloid leukemia with myelodysplasia. Blood. 2004;103(6):23162324. 32. Rücker FG, Schlenk RF, Bullinger L, et al. TP53 alterations in acute myeloid leukemia with complex karyotype correlate with specific copy number alterations, monosomal karyotype, and dismal outcome. Blood. 2012;119(9):2114-2121. 33. Mosna F, Papayannidis C, Martinelli G, et al. Complex karyotype, older age, and reduced first-line dose intensity determine poor survival in core binding factor acute myeloid leukemia patients with long-term follow-up. Am J Hematol. 2015;90(6):515-523. 34. Appelbaum FR, Kopecky KJ, Tallman MS, et al. The clinical spectrum of adult acute myeloid leukaemia associated with core binding factor translocations. Br J Haematol. 2006;135(2):165-173. 35. Bullinger L, Krönke J, Schön C, et al. Identification of acquired copy number alterations and uniparental disomies in cytogenetically normal acute myeloid leukemia using high-resolution single-nucleotide polymorphism analysis. Leukemia. 2010;24(2): 438-449. 36. Wouters BJ, Lowenberg B, Erpelinck-

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

Verschueren CA, van Putten WL, Valk PJ, Delwel R. Double CEBPA mutations, but not single CEBPA mutations, define a subgroup of acute myeloid leukemia with a distinctive gene expression profile that is uniquely associated with a favorable outcome. Blood. 2009;113(13):3088-3091. Green CL, Koo KK, Hills RK, Burnett AK, Linch DC, Gale RE. Prognostic significance of CEBPA mutations in a large cohort of younger adult patients with acute myeloid leukemia: impact of double CEBPA mutations and the interaction with FLT3 and NPM1 mutations. J Clin Oncol. 2010;28 (16):2739-2747. Helmer RA, Foreman O, Dertien JS, Panchoo M, Bhakta SM, Chilton BS. Role of helicase-like transcription factor (hltf) in the G2/m transition and apoptosis in brain. PLoS One. 2013;8(6):e66799. Hewetson A, Chilton BS. Progesteronedependent deoxyribonucleic acid looping between RUSH/SMARCA3 and Egr-1 mediates repression by c-Rel. Mol Endocrinol. 2008;22(4):813-822. Helmer RA, Martínez-Zaguilán R, Dertien JS, et al. Helicase-like transcription factor (Hltf) regulates G2/M transition, Wt1/Gata4/Hif-1a cardiac transcription networks, and collagen biogenesis. PLoS One. 2013;8(11):e80461. Michaud J, Simpson KM, Escher R, et al. Integrative analysis of RUNX1 downstream pathways and target genes. BMC Genomics. 2008;9:363.doi: 10.1186/1471-2164-9-363. Wu D, Ozaki T, Yoshihara Y, Kubo N, Nakagawara A. Runt-related transcription factor 1 (RUNX1) stimulates tumor suppressor p53 protein in response to DNA damage through complex formation and acetylation. J Biol Chem. 2013;288(2):1353-1364. Satoh Y, Matsumura I, Tanaka H, et al. C-terminal mutation of RUNX1 attenuates the DNA-damage repair response in hematopoietic stem cells. Leukemia. 2012;26(2):303311. Linggi B, Müller-Tidow C, van de Locht L, et al. The t(8;21) fusion protein, AML1 ETO, specifically represses the transcription of the p14(ARF) tumor suppressor in acute myeloid leukemia. Nat Med. 2002;8(7):743750. Alcalay M, Meani N, Gelmetti V, et al. Acute myeloid leukemia fusion proteins deregulate genes involved in stem cell maintenance and DNA repair. J Clin Invest. 2003;112(11): 1751-1761. Mandoli A, Singh AA, Jansen PW, et al. CBFB-MYH11/RUNX1 together with a compendium of hematopoietic regulators, chromatin modifiers and basal transcription factors occupies self-renewal genes in inv(16) acute myeloid leukemia. Leukemia. 2014;28(4):770-778. Figueroa ME, Abdel-Wahab O, Lu C, et al. Leukemic IDH1 and IDH2 mutations result in a hypermethylation phenotype, disrupt TET2 function, and impair hematopoietic differentiation. Cancer Cell. 2010; 18(6):553567. Kitabayashi I, Yokoyama A, Shimizu K, Ohki M. Interaction and functional cooperation of the leukemia-associated factors AML1 and p300 in myeloid cell differentiation. EMBO J. 1998;17(11):2994-3004. Katsumoto T, Yoshida N, Kitabayashi I. Roles of the histone acetyltransferase monocytic leukemia zinc finger protein in normal and malignant hematopoiesis. Cancer Sci. 2008;99(8):1523-1527. O'Connell RM, Rao DS, Chaudhuri AA, et al. Sustained expression of microRNA-155 in hematopoietic stem cells causes a myeloproliferative disorder. J Exp Med. 2008; 205(3):585-594.

457


ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Chronic Lymphocytic Leukemia

Ferrata Storti Foundation

Prevalence and characteristics of central nervous system involvement by chronic lymphocytic leukemia Paolo Strati,1 Joon H. Uhmš, Timothy J. Kaufmann,1 Chadi Nabhan,2 Sameer A. Parikh,1 Curtis A. Hanson,1 Kari G. Chaffee,1 Timothy G. Call,1 and Tait D. Shanafelt1

Haematologica 2016 Volume 101(4):458-465

1

Mayo Clinic College of Medicine, Rochester, MN; and 2University of Chicago, IL, USA

ABSTRACT

A

Correspondence: shanafelt.tait@mayo.edu

Received: September 12, 2015. Accepted: January 22, 2016. Pre-published: January 27, 2016. doi:10.3324/haematol.2015.136556

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/4/458

Š2016 Ferrata Storti Foundation Material published in Haematologica is covered by copyright. All rights reserved to Ferrata Storti Foundation. Copies of articles are allowed for personal or internal use. A permission in writing by the publisher is required for any other use.

458

broad array of conditions can lead to neurological symptoms in chronic lymphocytic leukemia patients and distinguishing between clinically significant involvement of the central nervous system by chronic lymphocytic leukemia and symptoms due to other etiologies can be challenging. Between January 1999 and November 2014, 172 (4%) of the 4174 patients with chronic lymphocytic leukemia followed at our center had a magnetic resonance imaging of the central nervous system and/or a lumbar puncture to evaluate neurological symptoms. After comprehensive evaluation, the etiology of neurological symptoms was: central nervous system chronic lymphocytic leukemia in 18 patients (10% evaluated by imaging and/or lumbar puncture, 0.4% overall cohort); central nervous system Richter Syndrome in 15 (9% evaluated, 0.3% overall); infection in 40 (23% evaluated, 1% overall); autoimmune/inflammatory conditions in 28 (16% evaluated, 0.7% overall); other cancer in 8 (5% evaluated, 0.2% overall); and another etiology in 63 (37% evaluated, 1.5% overall). Although the sensitivity of cerebrospinal fluid analysis to detect central nervous system disease was 89%, the specificity was only 42% due to the frequent presence of leukemic cells in the cerebrospinal fluid in other conditions. No parameter on cerebrospinal fluid analysis (e.g. total nucleated cells, total lymphocyte count, chronic lymphocytic leukemia cell percentage) were able to offer a reliable discrimination between patients whose neurological symptoms were due to clinically significant central nervous system involvement by chronic lymphocytic leukemia and another etiology. Median overall survival among patients with clinically significant central nervous system chronic lymphocytic leukemia and Richter syndrome was 12 and 11 months, respectively. In conclusion, clinically significant central nervous system involvement by chronic lymphocytic leukemia is a rare condition, and neurological symptoms in patients with chronic lymphocytic leukemia are due to other etiologies in approximately 80% of cases. Analysis of the cerebrospinal fluid has high sensitivity but limited specificity to distinguish clinically significant chronic lymphocytic leukemia involvement from other etiologies.

Introduction Chronic lymphocytic leukemia (CLL) is a clonal disorder of B lymphocytes, characterized by proliferation and accumulation of small mature-appearing lymphocytes in the blood, bone marrow, and lymphoid tissues.1 The presence of clinically significant infiltration of CLL lymphocytes outside of these sites is relatively rare and is defined as extramedullary CLL. Although the central nervous system (CNS) haematologica | 2016; 101(4)


CNS involvement in CLL

is one of the most commonly observed extramedullary manifestations of CLL,2 less than 100 cases of CNS involvement by CLL are described in the literature; all are case reports or small case series.3-16 Despite the low frequency of clinically significant CNS involvement by CLL, post-mortem studies of patients with CLL indicate that occult CNS involvement by CLL is a relatively common finding, with a prevalence of 7%-71%.17-20 This discrepancy between clinical manifestations and autopsy findings illustrates the fact that, while CLL cells may frequently be present in the CNS, they rarely cause clinically significant manifestations. This makes the evaluation of neurological symptoms in patients with CLL challenging, since the mere identification of CLL cells in CNS does not necessarily indicate that CLL is the etiology of the patients’ neurological symptoms. In addition, the spectrum of neurological conditions that occur in patients with CLL is broad and includes infections, other Table 1. Distribution of final diagnosis among patients evaluated for neurological symptoms.

Patients (n=172)

N (%)

Clinically significant CNS involvement by CLL Richter syndrome involving CNS Infections PML Cryptococcus meningitis Aseptic meningitis Sinusitis VZV encephalitis EBV encephalitis WNL encephalitis HHV6 encephalitis Coccidioides meningitis Endophthalmitis Autoimmune/inflammatory Autoimmune radiculopathy Autoimmune myelopathy CNS vasculitis Autoimmune encephalopathy CIDP Multiple sclerosis Eaton-Lambert Sarcoidosis Uveitis Other cancer Primary brain tumor Metastatic disease Non-CLL-related Delirium Non-autoimmune radiculopathy Migraine Epilepsy/seizure Dementia Progressive supranuclear palsy Fibromyalgia Atrophic lateral sclerosis Bell’s palsy Arteriovenous malformation Transient ischemic attack Trigeminal pain Refractive eye disorder

18 (10) 15 (9) 40 (23) 11 9 7 5 3 1 1 1 1 1 28 (16) 6 5 5 4 3 2 1 1 1 8 (5) 5 3 63 (37) 19 9 9 6 4 3 3 2 2 1 1 1 1

PML: progressive multifocal leukoencephalopathy; VZV: varicella zoster virus; EBV: Epstein-Barr virus; WNL: West Nile virus; HHV6: human herpesvirus 6; CIDP: chronic inflammatory demyelinating polyneuropathy.

haematologica | 2016; 101(4)

malignancies, autoimmune/inflammatory diseases, and non-CLL-related medical conditions. Here, we report the first cohort study of patients with CLL undergoing evaluation of neurological symptoms, and describe the prevalence and characteristics of patients diagnosed with clinically significant CNS involvement by CLL or Richter syndrome (RS).

Methods Study population This study was reviewed and approved by the Institutional Review Board of the Mayo Clinic and was conducted in accordance with the principles of the Declaration of Helsinki. Between January 1999 [when flow cytometry on cerebrospinal fluid (CSF) was introduced in our institution] and November 2014, 4174 patients with CLL/SLL were cared for and monitored in the Division of Hematology at our institution and consented to be registered on the Mayo Clinic database. Information regarding base-line evaluation, medical history, laboratory findings, and prognostic factors were obtained from clinical and research records. The latter included fluorescent in situ hybridization (FISH) for common CLL chromosome abnormalities, analysis of the mutation status of the immunoglobulin heavy chain variable (IGHV) gene, and CD38, ZAP70 and CD49d expression by flow cytometry.

Diagnostic work up for CNS involvement The present analysis focused on all patients with CLL who underwent evaluation of neurological symptoms (including headache, confusion, and strength or sensitivity deficit) with a magnetic resonance imaging (MRI) of the CNS (e.g. brain, spine) and/or underwent lumbar puncture with CSF analysis. Patients with established active systemic RS at time of neurological symptoms were not included in the study. CSF analysis included: basic chemistry (glucose, protein), total nucleated cells (TNC) quantification, and microbiology/autoimmune studies. CSF studies included cytology and flow cytometry. All CSF and brain biopsies were reviewed by Mayo Clinic pathologists. After comprehensive evaluation, patients in whom the etiology of neurological symptoms was determined to be due to CLL and/or RS, and who required treatment of CLL/RS for their CNS involvement, were considered to have clinically significant involvement of the CNS by CLL. Patients who had an alternative etiology for their neurological symptoms identified on evaluation but who had incidental presence of CLL B cells in the CSF as a bystander were in effect not considered to have clinically significant involvement by CLL.

Statistical analysis Descriptive statistics were used to summarize base-line characteristics. Categorical and continuous variables were evaluated using the χ2 or Fisher exact tests and the Mann-Whitney test, as appropriate. Overall survival (OS) was defined as time from CNS diagnosis to death or last follow up. Survival curves were plotted using the Kaplan-Meier method, and comparisons were made using the log rank test. All P-values were 2-sided; P≤0.05 was considered significant. Statistical analysis was performed using SPSS 21.

Results Diagnostic work up Of the 4174 patients with CLL cared for at our center 459


P. Strati et al.

between January 1999 and November 2014, 1115 (28%) underwent MRI of the brain and/or the spine for evaluation of neurological symptoms during the course of their disease. We next focused on the 50 patients who had MRI findings suspicious for possible CNS malignancy (23 with meningeal involvement, 27 with parenchymal brain involvement) and 122 patients whose MRI was not suspicious for CNS malignancy, but who underwent a lumbar puncture with CSF analysis to further evaluate their neurological symptoms. The neurological symptoms that initially prompted evaluation in these 172 were: cognitivebehavioral symptoms in 47 (27%), headache in 34 (20%), cranial nerve deficits in 24 (14%), paresthesia in 21 (12%), weakness in 19 (11%), gait disorder in 15 (9%), and seizure in 12 (7%). A more detailed presentation of the diagnostic work up in these patients is shown in Figure 1.

Final diagnosis Among the 50 patients whose MRI of the brain/spine was suspicious for CNS involvement by malignancy, 34 underwent a CNS tissue biopsy. Biopsy revealed histological features consistent with a small lymphocytic lymphoma/CLL confirmed by immunostains in 9 cases, RS in 12, progressive multifocal leukoencephalopathy (PML) in 2, vasculitis in 3, other cancers in 6 (3 primary brain tumors, 3 metastatic tumors), and normal brain tissue in 2 patients. Although patients with established systemic RS at the time of CNS evaluation were not included in this series (see Methods), 11 of 12 included patients whose ini-

tial manifestation of RS was CNS involvement were later found to have systemic disease. Additional diagnostic work up of CSF and blood for infectious and autoimmune etiologies was based on patientsâ&#x20AC;&#x2122; clinical profile. After comprehensive evaluation in all 172 patients, the etiology of neurological symptoms was determined to be clinically significant CNS involvement by CLL in 18 (10%; 0.4% of overall cohort), CNS involvement by RS in 15 (9% evaluated patients; 0.3% overall), infection in 40 (23% evaluated patients; 1% overall), autoimmune/inflammatory disease in 28 (16% evaluated patients; 0.7% overall), other cancer in 8 (5% evaluated patients; 0.2% overall) and nonCLL-related conditions in 63 (37% evaluated patients; 1.5% overall). Further details regarding non-CLL/RS diagnoses in these patients are provided in Table 1. Among the 13 patients with clinically significant CNS involvement by CLL, the involvement was parenchymal in 7 patients (Figure 2A) and meningeal in 6 (Figure 2B). Five of the 13 patients ultimately found to have clinically significant CNS involvement by CLL had no evidence of CNS disease on MRI. Diagnosis of clinically significant CNS involvement by CLL was based on both CSF analysis and tissue biopsy in 4 of these patients, on tissue biopsy only (CSF analysis negative or not performed) in 5, on CSF analysis only (tissue biopsy not performed) in 7, and was presumed after exhaustive work up to exclude other causes in 2 patients in whom the location of radiographic disease did not allow a tissue biopsy to be performed. Among the 15 patients with CNS involvement by RS,

Figure 1. Diagnostic work up performed for neurological symptoms in this cohort of patients with chronic lymphocytic leukemia. CLL: chronic lymphocytic leukemia; SLL: small lymphocytic lymphoma; MRI: magnetic resonance imaging; LP: lumbar puncture; CNS: central nervous system.

Table 2. Concordance of lumbar puncture and tissue biopsy for diagnosis of CNS CLL/RS.

Patients (n=162)

N

CNS biopsy positive for CLL

CNS biopsy positive for RS

Final diagnosis of CLL/RS

Negative for CLL/RS by CSF analysis Positive for CLL/RS by CSF analysis

131 31

4/19 (21%) 4/7 (57%)

9/19 (47%) 0/7 (0%)

15*/132 (11%) 13/31 (42%)

CNS biopsy was performed in 21 of 145 patients with LP negative for CLL/RS, and in 7 of 31 patients with LP positive for CLL/RS. CNS: central nervous system; LP: lumbar puncture; RS: Richter syndrome; CLL: chronic lymphocytic leukemia. *1 patient with negative LP and CNS lesions not amenable to biopsy was considered to have CNS CLL and treated accordingly; 1 patient had extra-CNS biopsy positive for RS and was considered to have CNS RS.

460

haematologica | 2016; 101(4)


CNS involvement in CLL

the CNS involvement was parenchymal in 8 (Figure 2C) and meningeal in 7 (Figure 2D) patients. Diagnosis of CNS involvement by RS was confirmed by tissue biopsy in 12 patients and by CSF analysis in 3. Tissue biopsy was frequently necessary to confirm the diagnosis of CNS RS. Among the 12 patients with biopsy proven RS, 9 had no findings to suggest RS on CSF analysis while the remaining 3 patients did not undergo CSF analysis.

Lumbar puncture and tissue biopsy concordance in patients with CNS CLL/RS Overall, 162 patients underwent a CSF analysis: 44 between 1999 and 2004, 54 between 2005 and 2009, and 64 between 2010 and 2015. Of the 131 patients who had no evidence of CLL/RS on CSF analysis, 15 (11%) were ultimately diagnosed with clinically significant CNS involvement by CLL or RS based on tissue biopsy, most of whom had a parenchymal brain lesion on MRI (Table 2). Among the 31 patients with CSF analysis demonstrating the presence of CLL B cells, 18 (58%) had an alternative diagnosis for their neurological symptoms identified after tissue biopsy and/or other work up. This included: infections in 8 patients, an autoimmune/inflammatory process in 5, metastatic cancer in 1, and non-CLL-related etiology in 4.

Factors associated with CNS CLL/RS Chronic lymphocytic leukemia characteristics of the 33 patients with a final diagnosis of clinically significant involvement of the CNS by CLL or RS are shown in Table

3. CSF characteristics associated with clinically significant CNS involvement by CLL or RS on univariate analysis were: low CSF glucose (53 vs. 60 mg/dL; P=0.04), elevated CSF total nucleated cell count (14 vs. 2 cells/mL; P=0.007), elevated CSF lymphocyte count (10 vs. 2 cells/mL; P=0.02), and elevated CSF CLL cell percentage by flow cytometry (5% vs. 0%; P=0.05). Due to substantial overlap in all of these CSF parameters between groups, however, none of these characteristics reliably discriminated between patientâ&#x20AC;&#x2122;s whose neurological symptoms were due to clinically significant CNS involvement by CLL/RS or another etiology.

Treatment and survival of patients with CNS CLL/RS Of 33 patients with clinically significant CNS involvement by CLL or RS, 30 received therapy for the CNS disease, whereas 3 received best supportive care only. The initial therapy for patients with clinically significant CNS involvement by CLL was a purine nucleoside-analogbased regimen in 8 (42%) patients, radiation therapy in 3 (16%), anthracycline-based regimens in one (5%), highdose methotrexate (MTX) in 2 (13%), high-intensity regimens (e.g. fractionated cyclophosphamide, vincristine, doxorubicin, and dexamethasone or ifosfamide, carboplatin, etoposide, variably combined with intrathecal MTX or cytarabine) in 2 (13%), and best supportive care in 2 (11%). Among patients with CNS involvement by RS, the initial therapy was an anthracycline-based regimen in 5 (33%) patients, high-dose MTX in 3 (20%), high-

Figure 2. Magnetic resonance findings. (A) Chronic lymphocytic leukemia (CLL) parenchymal. (B) CLL meningeal. (C) Richter syndrome (RS) parenchymal. (D) RS meningeal.

haematologica | 2016; 101(4)

461


P. Strati et al.

intensity regimens in 6 (40%), and best supportive care in one (7%). After a median follow up of 14 months, the median overall survival (OS) for all 172 CLL patients evaluated for neurological symptoms was 24 months. After a median follow up from diagnosis of CNS involvement by CLL or RS of 12 months (range 1-178 months), 78% of patients with clinically significant CNS involvement by CLL and 80% of patients with CNS RS have died. The median OS of patients with clinically significant CNS involvement by CLL or RS were 12 and 11 months, respectively (Figure 3).

Discussion Although the CNS is one of the most common extramedullary manifestations of CLL,2 less than 100 cases

have been reported in the world literature to date.3-16 Among the 4174 CLL patients cared for at our center over a 17-year interval, 33 (19% of evaluated patients; 0.8% overall cohort) developed clinically significant CNS involvement by CLL or RS. While clinically significant CNS involvement by CLL is rare, neurological symptoms frequently occur in patients with CLL, and distinguishing whether or not these symptoms are due to CLL or other etiologies can be challenging. When a mass lesion is identified on imaging and a tissue biopsy can be performed, the evaluation and diagnosis for CNS involvement by CLL or RS is relatively straightforward. Biopsy is needed in these cases to distinguish between CLL, RS, primary brain tumors, or metastatic malignancy, as illustrated in our series. The evaluation of neurological symptoms becomes far more challenging when no lesion is identified on imaging. In fact, interpret-

Table 3. Clinical characteristics of patients with clinically significant CNS CLL/RS as compared to other etiologies of neurological symptoms.

Patients (n= 172) Age (years) Males Females ALC (x109/L) Hemoglobin (mg/dL) Platelets(109/L) Rai stage 0 I-II III-IV CD49d positive negative CD38 positive negative ZAP70 positive negative IGHV unmutated mutated FISH del13q negative +12 del11q del17p Previously treated Untreated CSF glucose (mg/dL) CSF protein (mg/dL) CSF TNC (cells/mL) CSF neutrophils (%) CSF lymphocytes (%) CSF lymphocyte count (cells/mL) CSF CLL (%)

CNS CLL/RS (33)

N (%), median [range] No CNS CLL/RS (139)

63 [48-83] 24 (73) 9 (27) 5 [0-238] 13 [9-16] 181 [9-431]

68 [30-90] 95 (68) 44 (32) 4 [0-325] 12 [7-17] 163 [22-658]

12 (36) 14 (42) 7 (22) 4 (44) 5 (56) 8 (40) 12 (60) 9 (60) 6 (40) 3 (30) 7 (70)

53 (38) 53 (38) 33 (24) 41 (59) 29 (41) 45 (39) 69 (61) 39 (50) 39 (50) 43 (67) 21 (33)

3 (33) 1 (11) 4 (45) 1 (11) 0 (0) 17 (52) 16 (48) 53 [19-90] 62 [31-161] 14 [1-443] 1 [0-97] 72 [8-97] 10 [1-97] 5 [0-93]

6 (14) 6 (14) 15 (37) 9 (21) 6 (14) 76 (55) 63 (45) 60 [15-176] 60 [11-301] 2 [1-11850] 1 [0-96] 70 [1-100] 2 [1-11731] 0 [0-74]

P-uni 0.44 0.68 0.44 0.44 0.70

0.89 0.49 1 0.58 0.04

0.49

0.85 0.03 0.78 0.007 0.94 0.95 0.02 0.05

CNS: central nervous system; CLL: chronic lymphocytic leukemia; RS: Richter syndrome; ALC: absolute lymphocyte count; IGHV: immunoglobulin heavy chain variable region gene; FISH: fluorescence in situ hybridization; CSF: cerebral spinal fluid; TNC: total nucleated cells.

462

haematologica | 2016; 101(4)


CNS involvement in CLL

ing the results of CSF analysis can be particularly difficult. Although flow cytometry is often needed to distinguish between reactive lymphocytes and CLL B cells,5,12 the presence of CLL B cells in the CSF does not necessarily indicate that CLL is the etiology of the patient’s neurological symptoms. In fact, CLL cells traffic to sites of inflammation and can often be present as a bystander in patients with infections or other inflammatory conditions.21 Contamination of the CSF by peripheral blood (e.g. “bloody tap”) can also detect CLL B cells in the CSF. As noted, autopsy series have demonstrated clinically insignificant CNS involvement in up to 70% of patients, indicating asymptomatic CNS involvement by CLL is relatively common.17-20 This underscores the fact that a complete comprehensive work up is critical to exclude other etiologies of neurological symptoms rather than assuming the mere presence of CLL B cells in the CSF is an indication that CLL is the etiology of the patient’s neurological symptoms. Indeed, the specificity of the presence of CLL B cells in the CSF study was relatively poor. As shown in our series, the sensitivity of CSF analysis is also low for those patients with parenchymal brain involvement by CLL/RS. An overview of our approach to the evaluation of neurological symptoms in patients with CLL is provided in Figure 4. As the first step, CLL should be re-staged, and transformation considered. Tests performed on peripheral blood should include cryptococcal antigen, but also erythrocyte sedimentation rate, vasculitis panel and ACE levels; tests performed on CSF should include Gram and fungal staining, bacterial and fungal cultures, and viral studies, including evaluation for all herpetic viruses and JC virus. It is well recognized that patients with CLL have abnormalities in both cellular and humoral immune function and are at increased risk for infection.22 Purine nucleosideanalogs, the historical mainstay of treatment of CLL, significantly decrease CD4+ count and also contribute to risk of infection.23 The most common CNS infection reported in our series was PML, caused by the JC virus. PML occurred in 11 (0.3%) out of the 4174 in our series. The reported incidence of PML in patients with CLL varies from 0.5% to 11.1%. However, most previous reports were heavily weighted toward patients receiving treatment, where advanced age, male sex and use of purine analogs appeared to be the primary risk factors.8,24-26 In our series, all 11 patients with PML had received treatment, with 8 of 11 having received prior anti-CD20 therapy (e.g. rituximab, ofatumumab, obinutuzumab), 8 of 11 had received purine nucleoside analog-based therapy, and 4 of 11 had received prior alemtuzumab. Although some statistically significant differences were observed in CSF parameters (e.g. glucose, number of nucleated cells) among patients with clinically significant involvement of the CNS by CLL compared to those with other etiologies for their neurological symptoms there was wide variability, and none reliably distinguished clinically significant involvement of the CNS by CLL from other causes. It is also worth noting that no clinical feature of CLL (e.g. Rai stage, prior CLL therapy) or CLL prognostic parameter (e.g. IGHV, ZAP-70, FISH, CD49d) was clearly associated with CNS involvement. In our series, only 52% of patients had been previously treated at the time of CNS CLL/RS diagnosis, and 78% had early stage disease by physical examination and complete blood count. Interestingly, most cases of CNS involvement by CLL reported in the literature also occurred in early stage and haematologica | 2016; 101(4)

Figure 3. Overall survival (OS) by final diagnosis of etiology of neurological symptoms. *Clinically significant chronic lymphocytic leukemia (see Methods).

previously untreated patients.5,14,27,28 While the treatment of other CNS lymphomas has been extensively described and studied, there is no standard initial regimen for CNS CLL/RS due to the rarity of these conditions.29 Although intrathecal therapy is widely used in the treatment of leptomeningeal diffuse large B-cell lymphoma and some other non-Hodgkin lymphomas, this is primarily because the medications that comprise the definitive treatment for these histologies do not penetrate the CNS. In contrast, a number of case reports have demonstrated the activity of systemic fludarabine-based therapy for CLL patients with CNS disease, which is typically our initial treatment approach.30-32 Despite treatment, the prognosis for patients with clinically significant CNS involvement by CLL or RS in our series was dismal, with a median OS of 12 and 11 months, respectively, concordant with other published data.2 The role of newer agents, inhibiting the B-cell receptor pathway and potentially altering the migration of CLL cells in extramedullary tissue, as a treatment for CNS involvement by CLL is unknown. Both lenalidomide and ibrutinib have shown promising CNS penetration in lymphoma, and further studies in patients with CNS involvement are warranted.34,35 Our study has some major limitations that need to be highlighted. First of all, this is a retrospective study; the diagnostic work up was not standardized but was based on physician’s judgement, with consequent variability and missing values. Although, in our series, more biologically aggressive CLL characteristics were not associated with clinically significant CNS involvement by CLL, the sample size is limited for this analysis and may not have adequate power to detect a modest increase in risk. In addition, the diagnostic tools used have technical limitations; for example, MRI has limited sensitivity in the detection of leptomeningeal disease,36-38 while the sensitivity of flow cytometry has improved over the 16-year period of our study. In conclusion, although neurological symptoms occur 463


P. Strati et al.

Figure 4. Diagnostic work up of neurological symptoms in patients with chronic lymphocytic leukemia. CNS: central nervous system; LP: lumbar puncture; PCR: polymerase chain reaction; CSF: cerebrospinal fluid; Ab: antibodies; NLP: neoplastic.¹PCR for viruses to include JC, WNL, HHV-6, EBV, VZV and CMV. Fungal antigens (both on CSF and blood) to include cryptococcus and coccidioides. ²CSF evaluation to include oligoclonal bands; serum markers to include p-ANCA, c-ANCA, antiacetylcholine receptor and ACE.

relatively frequently in patients with CLL, clinically significant CNS involvement by CLL or RS is a rare condition. The presence of CLL B cells in the CSF has reasonable sensitivity but low specificity and is insufficient to establish CLL as the etiology of patients’ neurological symp-

References 1. Chiorazzi N, Rai KR, Ferrarini M. Chronic lymphocytic leukemia. N Engl J Med. 2005; 352(8):804-815. 2. Ratterman M, Kruczek K, Sulo S, Shanafelt TD, Kay NE, Nabhan C. Extramedullary chronic lymphocytic leukemia: systematic analysis of cases reported between 1975 and 2012. Leuk Res. 2014;38(3):299-303. 3. Kalac M, Suvic-Krizanic V, Ostojic S, Kardum-Skelin I, Barsic B, Jaksica B. Central nervous system involvement of previously undiagnosed chronic lymphocytic leukemia in a patient with neuroborreliosis. Int J Hematol. 2007;85(4):323-325. 4. Denier C, Tertian G, Ribrag V, et al. Multifocal deficits due to leukemic meningoradiculitis in chronic lymphocytic leukemia. J Neurol Sci. 2009;277(1-2): 130-132. 5. Harrison PB, Cripps AW. Exclusion of leukaemic meningitis by quantitation of T and B lymphocytes in cerebrospinal fluid.

464

toms. A comprehensive neurological evaluation including imaging, CSF analysis, and evaluation for a wide range of infectious and inflammatory conditions is needed to determine the etiology of neurological symptoms in these patients.

Aust N Z J Med. 1982;12(3):286-287. 6. Solal-Celigny P, Schuller E, Courouble Y, Gislon J, Elghozi D, Boivin P. [Cerebromeningeal location of chronic lymphoid leukemia. Rapid immunochemical diagnosis and complete remission by intrathecal chemotherapy]. Presse Med. 1983;12(37):2323-2325. 7. Ruiz P, Moezzi M, Chamizo W, Ganjei P, Whitcomb CC, Rey LC. Central nervous system expression of a monoclonal paraprotein in a chronic lymphocytic leukemia patient. Acta Haematol. 1992;88(1):37-40. 8. Bower JH, Hammack JE, McDonnell SK, Tefferi A. The neurologic complications of B-cell chronic lymphocytic leukemia. Neurology. 1997;48(2):407-412. 9. Majumdar G. Chronic lymphocytic leukaemia with cerebral infiltration. Leuk Lymphoma. 1998;28(5-6):603-605. 10. Wang ML, Shih LY, Dunn P, Kuo MC. Meningeal involvement in B-cell chronic lymphocytic leukemia: report of two cases. J Formos Med Assoc. 2000;99(10):775-778.

11. Brick WG, Majmundar M, Hendricks LK, Kallab AM, Burgess RE, Jillella AP. Leukemic leptomeningeal involvement in stage 0 and stage 1 chronic lymphocytic leukemia. Leuk Lymphoma. 2002; 43(1):199-201. 12. Akintola-Ogunremi O, Whitney C, Mathur SC, Finch CN. Chronic lymphocytic leukemia presenting with symptomatic central nervous system involvement. Ann Hematol. 2002;81(7):402-404. 13. Zhu J, Wu Z, Fan L, Xu W, Li J. [Chronic lymphocytic leukemia with central nervous system invasion: one case report and literature review]. Zhonghua Xue Ye Xue Za Zhi. 2014;35(7):592-595. 14. Weizsaecker M, Koelmel HW. Meningeal involvement in leukemias and malignant lymphomas of adults: incidence, course of disease, and treatment for prevention. Acta Neurol Scand. 1979;60(6):363-370. 15. Grisold W, Jellinger K, Lutz D. Human neurolymphomatosis in a patient with chronic lymphatic leukemia. Clin Neuropathol. 1990;9(5):224-230.

haematologica | 2016; 101(4)


CNS involvement in CLL 16. Pohar S, deMetz C, Poppema S, Hugh J. Chronic lymphocytic leukemia with CNS involvement. J Neurooncol. 1993;16(1):35-37. 17. Reske-Nielsen E, Petersen JH, Sogaard H, Jensen KB. Letter: Leukaemia of the central nervous system. Lancet. 1974;1(7850): 211-212. 18. Bojsen-Moller M, Nielsen JL. CNS involvement in leukaemia. An autopsy study of 100 consecutive patients. Acta Pathol Microbiol Immunol Scand A. 1983; 91(4):209-216. 19. Cramer SC, Glaspy JA, Efird JT, Louis DN. Chronic lymphocytic leukemia and the central nervous system: a clinical and pathological study. Neurology. 1996; 46(1):19-25. 20. Barcos M, Lane W, Gomez GA, et al. An autopsy study of 1206 acute and chronic leukemias (1958 to 1982). Cancer. 1987; 60(4):827-837. 21. Nowakowski GS, Call TG, Morice WG, Kurtin PJ, Cook RJ, Zent CS. Clinical significance of monoclonal B cells in cerebrospinal fluid. Cytometry B Clin Cytom. 2005;63 (1):23-27. 22. Tsiodras S, Samonis G, Keating MJ, Kontoyiannis DP. Infection and immunity in chronic lymphocytic leukemia. Mayo Clin Proc. 2000;75(10):1039-1054. 23. Garcia-Suarez J, de Miguel D, Krsnik I, Banas H, Arribas I, Burgaleta C. Changes in the natural history of progressive multifocal leukoencephalopathy in HIV-negative lymphoproliferative disorders: impact of novel therapies. Am J Hematol. 2005; 80(4):271-281.

haematologica | 2016; 101(4)

24. D'Souza A, Wilson J, Mukherjee S, Jaiyesimi I. Progressive multifocal leukoencephalopathy in chronic lymphocytic leukemia: a report of three cases and review of the literature. Clin Lymphoma Myeloma Leuk. 2010;10(1):E1-9. 25. Gonzalez H, Bolgert F, Camporo P, Leblond V. Progressive multifocal leukoencephalitis (PML) in three patients treated with standard-dose fludarabine (FAMP). Hematol Cell Ther. 1999;41(4):183-186. 26. Amend KL, Turnbull B, Foskett N, Napalkov P, Kurth T, Seeger J. Incidence of progressive multifocal leukoencephalopathy in patients without HIV. Neurology. 2010;75(15):1326-1332. 27. Bagic A, Lupu VD, Kessler CM, Tornatore C. Isolated Richter's transformation of the brain. J Neurooncol. 2007;83(3):325-328. 28. O'Neill BP, Habermann TM, Banks PM, O'Fallon JR, Earle JD. Primary central nervous system lymphoma as a variant of Richter's syndrome in two patients with chronic lymphocytic leukemia. Cancer. 1989;64(6):1296-1300. 29. Ferreri AJ, Marturano E. Primary CNS lymphoma. Best Pract Res Clin Haematol. 2012;25(1):119-130. 30. Elliott MA, Letendre L, Li CY, Hoyer JD, Hammack JE. Chronic lymphocytic leukaemia with symptomatic diffuse central nervous system infiltration responding to therapy with systemic fludarabine. Br J Haematol. 1999;104(4):689-694. 31. Knop S, Herrlinger U, Ernemann U, Kanz L,

32.

33.

34.

35.

36. 37.

38.

Hebart H. Fludarabine may induce durable remission in patients with leptomeningeal involvement of chronic lymphocytic leukemia. Leuk Lymphoma. 2005;46(11): 1593-1598. Gonsalves WI, Zent CS, Pulido JS, Patnaik MM. Visual loss in early-stage chronic lymphocytic leukemia. J Clin Oncol. 2013; 31(17):e280-282. Hoang-Xuan K, Bessell E, Bromberg J, et al. Diagnosis and treatment of primary CNS lymphoma in immunocompetent patients: guidelines from the European Association for Neuro-Oncology. Lancet Oncol. 2015; 16(7):e322-332. Bernard S, Goldwirt L, Amorim S, et al. Activity with ibrutinib in mantle cell lymphoma patients with central nervous system relapse. Blood. 2015;126(14):1695-1698. Houillier C, Choquet S, Touitou V, et al. Lenalidomide monotherapy as salvage treatment for recurrent primary CNS lymphoma. Neurology. 2015;84(3):325-326. Yousem DM, Patrone PM, Grossman RI. Leptomeningeal metastases: MR evaluation. J Comput Assist Tomogr. 1990; 14(2):255-261. Fischer L, Martus P, Weller M, et al. Meningeal dissemination in primary CNS lymphoma: prospective evaluation of 282 patients. Neurology. 2008;71(14):1102-1108. Kiewe P, Fischer L, Martus P, Thiel E, Korfel A. Meningeal dissemination in primary CNS lymphoma: diagnosis, treatment, and survival in a large monocenter cohort. Neuro Oncol. 2010;12(4):409-417.

465


ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Hodgkin Lymphoma

Ferrata Storti Foundation

Impact of post-brentuximab vedotin consolidation on relapsed/refractory CD30+ Hodgkin lymphomas: a large retrospective study on 240 patients enrolled in the French Named-Patient Program Aurore Perrot,1 Hélène Monjanel,2 Réda Bouabdallah,3 Philippe Quittet,4 Clémentine Sarkozy,5 Marc Bernard,6 Aspasia Stamatoullas,7 Cécile Borel,8 Krimo Bouabdallah,9 Emmanuelle Nicolas-Virelizier,10 Marion Fournier,11 Franck Morschhauser,12 and Pauline Brice2 for the Lymphoma Study Association (LYSA)

Service d'Hématologie, CHU de Nancy, Université de Lorraine, Vandoeuvre les Nancy; Hôpital Saint Louis AP-HP, Université Diderot, Paris; 3Département d'Hématologie, CLCC Institut Paoli Calmettes, Marseille; 4Département d'Hématologie Clinique, CHRU Hôpital Lapeyronie, Montpellier; 5Service Hématologie, Hospices Civils de Lyon, Université Claude Bernard, Pierre-Bénite; 6Service d'Hématologie, Hôpital Pontchaillou, Rennes; 7 Département d'Hématologie, Centre Henri Becquerel, Rouen; 8Fédération d'Hématologie, Institut Universitaire du Cancer, Toulouse; 9Service d'Hématologie, CHU de Bordeaux, Hôpital Haut-Lévèque, Pessac; 10Service de Cancérologie Médicale, CLCC Léon Bérard, Lyon; 11The Lymphoma Academic Research Organisation (LYSARC), Centre Hospitalier Lyon Sud, Pierre-Bénite; and 12 Service d'Hématologie, Hôpital Claude Huriez, Unité GRITA, Université de Lille, Lille, France 1 2

Haematologica 2016 Volume 101(4):466-473

ABSTRACT

Correspondence: au.perrot@chu-nancy.fr

Received: July 22, 2015. Accepted: December 22, 2015. Pre-published: January 14, 2016. doi:10.3324/haematol.2015.134213

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/4/466

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

466

B

rentuximab vedotin was reported to be effective and safe against refractory/relapsed Hodgkin lymphoma in cohorts of between 12 to 102 patients. Herein we report our retrospective analysis of the French experience with brentuximab vedotin used alone to treat 240 refractory/relapsed Hodgkin lymphoma patients enrolled in a named patient program between 2011 and 2013. All patients had histologically documented CD30+ Hodgkin lymphoma; 74% had refractory disease or early relapses. After a median of 3 lines of chemotherapy, brentuximab vedotin was infused intravenously (1.8 mg/kg every 3 weeks). The primary endpoint was best response. Response at the end of treatment, its duration, survival data and toxicity profile were secondary endpoints. Patients received a median of 6 cycles; 68 underwent a consolidation thereafter. The best response was observed after a median of 4 cycles in 145 (60.4%) patients: 33.8% complete response/unconfirmed complete response, 26.7% partial response. Objective responses were observed as decreased (39.3%) in the 28 patients >60 years. The median response duration was 8.4 months. With median follow-up at 16.1 months, median progression-free survival was 6.8 months and this was significantly longer for patients transplanted after brentuximab vedotin (a median of 18,8 months); median overall survival was not reached. No death has been linked to brentuximab vedotin toxicity. The most common adverse events were peripheral sensory neuropathy (29.3%) and hematological toxicity. The results of this analysis support the previously reported brentuximab vedotin efficacy with manageable toxicity. Because of the short-term responses in most patients, a high-dose therapy with stem cell transplantation for responders should be considered as quickly as possible.

haematologica | 2016; 101(4)


Post-brentuximab vedotin consolidation for CD30+ HL

Introduction

Table 1. Demographics and baseline clinical characteristics of 240 HL patients.

Conventional chemotherapy (with radiation for localized disease) has made Hodgkin lymphoma (HL) a curable malignancy for most patients. However, approximately 20% of HL patients have refractory disease or relapse after currently available first-line treatments. Patients with chemosensitive HL after salvage therapy should receive high-dose chemotherapy supported by autologous stem cell transplantation (auto-SCT) as standard.1 However, about 50% of patients relapse after auto-SCT, underscoring the need for effective agents for post-autograft failures. Alternating non-cross-resistant chemotherapy lines, exploring the role of novel drugs and ensuring early referral for allogeneic SCT (allo-SCT) seem to be the best current options. CD30 is a transmembrane receptor protein expressed on activated B or T lymphocytes and on the surface of malignant Hodgkin Reed-Sternberg cells. Through CD30–CD30 ligand interaction, eosinophils and mast cells may stimulate those malignant cells, promoting their survival.2 Brentuximab vedotin (BV, SGN-35) is a chimeric antiCD30 antibody conjugated via a protease-cleavable linker to a microtubule-disrupting agent, monomethyl auristatin E. After binding to the CD30 cell-surface, that agent is internalized, traffics to the lysosome and is released to disrupt microtubules, inducing cell cycle arrest and apoptosis.3 In the phase I, dose-escalation trial on 12 relapsed/refractory CD30+ HL patients, the maximum tolerated BV dose was determined to be 1.8 mg/kg every 3 weeks.4 Tumor regression was observed in 50% of the patients. Subsequently, BV was tested in a phase II trial on 102 patients with relapsed/refractory HL who had already undergone auto-SCT.5 In that pivotal, phase II, multicenter trial, 75% of the patients achieved an objective response and 34% a complete response (CR). Others have reported their retrospective experiences with between 25 to 65 patients:6-9 overall response rate (ORR) varied between 60% and 72%, with the CR rate between 17% and 34%, the median response duration was 8 to 9 months, with a median progression-free survival (PFS) rate of between 5 and 7 months. From 2011-2013, BV was available in France for a Millennium/Takeda Named Patient Program (NPP). Herein, we describe objective responses, adverse events, PFS and overall survival (OS) rates of 240 patients among the 313 enrolled in this NPP.

Demographic/clinical characteristic

Methods Study design The BV NPP was open from January 2011 until October 2013. All patients had histologically confirmed CD30+ HL and all patients had relapsed after prior auto-SCT, or after two chemotherapy lines (if the patients were not SCT candidates due to age, insufficient stem cell collection or chemorefractory disease). BV, 1.8 mg/kg (capped at 100 kg of body weight) in 250 mL of 0.9% saline was infused intravenously over 30 minutes once every 3 weeks (for a maximum of 16 cycles). The LYSA group analyzed the outcome of HL patients treated with BV during this NPP. Takeda provided the list of patients authorized to receive BV treatment. Inclusion criteria were diagnosis of relapsed/refractory CD30+ HL and treatment with BV in the haematologica | 2016; 101(4)

Median age at diagnosis: years (range) Sex, no. (%) Male Female ECOG performance status at diagnosis (n=160) 0/1 Ann Arbor stage at diagnosis (n=238) I/II III/IV Extranodal disease at diagnosis Bulky mediastinum at diagnosis (n=117) Prior radiation First-line chemotherapy ABVD BEACOPP/VABEM Others Primary refractory disease† Median prior chemotherapy regimens: no. (range) >3 First remission <2 months Disease status after last therapy‡ Relapsed Refractory No. of patients with at least one auto-SCT No. of prior auto-SCTs 1 2 No. of patients with prior allo-SCT Median auto-SCT to relapse months: (range) Median initial diagnosis to first study drug dose: months (range) Median last treatment to first study drug dose: months (range)

Value 30 (14-78) 156 (65%) 84 (35%) 146 (91.3%) 96 (40.3%) 142 (59.7%) 98 (40.8%) 54 (46.2%) 56 (23.3%) 198 (82.8%) 30 (12.5%) 12 (5%) 117 (48.7%) 3 (1-13) 94 (39.2%) 167 (69.6%) 102 (42.5%) 134 (55.8%) 142 (59.2%) 116 26 37 (15.4%) 6.5 (1-123) 30.8 (3.1-335.5) 2.5 (0.1-53.1)

ECOG: Eastern Cooperative Oncology Group; ECOG performance scores range from 0 (normal activity) to 5 (death), with higher scores indicating more severe disability; ABVD: Doxorubicin, Bleomycin, Vinblastine, Dacarbazine; BEACOPP/VABEM: Bleomycin, Etoposide, Doxorubicin, Cyclophophamide, Vincristine, Procarbazin, Prednisone/Vindesin, Adrimaycin, BCNU, Etoposide, Methylprednisolone; auto/allo-SCT, autologous/allogeneic stem cell transplantation. †Defined as failure to obtain a complete remission with front-line therapy or relapse within 3 months of it. ‡Relapsed indicates a best response of complete or partial remission to last therapy before BV, and refractory indicates a best response of stable.

NPP non-trial setting. We asked all participating centers that included more than one patient to complete a dedicated case report form (CRF). Among the 313 patients included in the NPP, 240 patients with completed CRFs from 89 French centers were included in this retrospective analysis. No data were available for the remaining patients due to small centers having included only a few patients. The study was approved by competent authorities, namely the Comité Consultatif sur le Traitement de l'Information en matière de Recherche dans le domaine de la Santé (CCTIRS) and the Commission Nationale de l'Informatique et des Libertés (CNIL). All CRFs were included in a LYSARC (The Lymphoma Academic Research Organization) specific database.

Study assessments Response was assessed by computerized tomography (CT) or FDG positron emission tomography (FDG-PET/CT) scans accord467


A. Perrot et al.

Figure 1. Frequencies of BV-related adverse events.

ing to the revised response criteria for malignant lymphoma.10,11 It was specified in the Temporary Recommendations for Use of the NPP to perform an evaluation of the disease after 4 cycles of BV. Safety and tolerance were evaluated by recording the frequencies and severities of all adverse events according to the National Cancer Institute's Common Terminology Criteria for Adverse Events (NCI CTCAE) v4.0. BV dose reduction to 1.2 mg/kg was recommended for grade 3 toxicity.

Statistical analyses The primary objective was the best response to BV. Response at the end of BV treatment, its duration, disease-free survival (DFS), PFS, OS and toxicity profile were secondary objectives. Response duration was measured from the date of the first response evaluated as CR/unconfirmed CR (uCR) or partial response (PR) to the date of the first documented disease progression, relapse or death from any cause, whichever occurred first. DFS was measured from the date of the first response evaluated as CR/CRu to the date of the first documented disease progression or relapse or death from any cause, whichever occurred first. PFS was measured from the date of the first BV cycle to the date of the first documented disease progression, relapse or death from any cause, whichever occurred first. OS was defined as the time from BV onset to death from any cause and was censored at the date of the last update. OS and PFS were estimated with the Kaplan-Meier method, using exact 95% confidence intervals (95% CIs), and with P<0.05 defining statistical significance. Analyses were computed using SAS software version 9.2.

Results Patient characteristics Table 1 summarizes the characteristics of the 240 patients (156 (65%) males and 84 (35%) females), whose median age at HL diagnosis was of 30 (range: 14-78) years. At diagnosis, performance status was 0/1 for 146 patients, almost 60% had stage III/IV HL, 46.2% had bulky mediastinum and about 40% had extranodal disease. Eighteen patients had previous malignancies, including six non-HL 468

and six patients with human immunodeficiency virus seropositive. In this large cohort, patients were young (<60 years) and first-line chemotherapy was based on an ABVD regimen for 82.8% of them; only around 24% of patients had previously received radiotherapy. Thirty-seven patients (24 men and 13 women) had undergone allo-SCT before BV. Their median age at HL diagnosis was 27 (range: 14-56), and 32 (range: 18-60) years at BV initiation. Patients had received a median of four (range: 1-13) prior chemotherapy regimens and 75% of them had already undergone radiotherapy. The median remission duration was 4.6 months (range: 0.3-27.6) and 32 out of 37 (86.5%) of them underwent auto-SCT before allo-SCT. Conditioning regimens were mostly reducedintensity conditioning (RIC), mainly fludarabine-based RIC for 25 out of 37 patients

Safety All patients enrolled in this retrospective analysis received at least one BV infusion with a median of six (range: 1-16) cycles; only 10 patients received all 16 cycles. Dose intensity was >75% for 91.6% of the patients. Only 40 patients had at least one dose reduction from 1.8 to 1.2 mg/kg (because of neuropathy for 11 patients and hepatic cytolysis for 2 patients). The most common treatment-related adverse events (â&#x2030;Ľ10%) were anemia (39%), peripheral neuropathy (29.3%), thrombocytopenia (28%), neutropenia (23%), diarrhea (14%) and infections (10%) (Figure 1). Peripheral neuropathy was present before BV treatment in 5% of the patients and increased with the number of cycles, peaking at cycle 7. Other observed grade 3 or 4 toxicities concerned 8 patients: 2 with hepatic cytolysis, 1 hypersensitivity requiring a desensitization procedure, 4 patients with bone pain and 1 with viral meningoencephalitis. BV was discontinued for 229 patients, but for only 17 (7.4%) due to toxicity. Other causes were HL progression in 126 (55%) patients, planned transplant for 58 (25.3%), donor lymphocyte infusion (DLI) for 6 patients, radiotherhaematologica | 2016; 101(4)


Post-brentuximab vedotin consolidation for CD30+ HL

apy for 11, 7 deaths and 4 patients lost to follow-up. Safety was highly acceptable for the 37 patients with prior allo-SCT. Anemia, thrombocytopenia and neutropenia were reported in 27.6%, 28.6% and 36.7% of them, respectively. Infections occurred less frequently (6.9%) probably because of the frequent use of anti-infectious agents in the post-transplant context. Vomiting and diarrhea (both grade <3) occurred in 14.3% and 17.2% of them, respectively. Peripheral neuropathy, no grade 3 or 4 neuropathy, was reported in 46.9% of patients.

Efficacy Responses and outcomes (DFS, PFS, OS) are summarized in Table 2 and Figure 2, respectively. The median follow-up was 16.1 (95% CI [0.2-32.6]) months. Primary endpoint: best response to BV treatment. Response was assessed using CT and the 1999 Cheson criteria for 19.9% of the patients, with PET-CT and the 2007 Cheson criteria in 78.7% and the others were evaluated clinically. The best response to BV treatment was observed after a median of four cycles of BV (range: 1-16). It was a CR for 70 (29.2%) patients, uCR for 11 (4.6%), PR for 64 (26.7%), stable disease (SD) for 18 (7.5%) and progressive disease (PD) for 68 (28.3%), for an objective response rate of 60.4%. The best response rate was not worse for the 37 patients with prior allo-SCT, with a CR at 39.5%, uCR at 5.3% and PR at 18.4%, corresponding to an objective response rate of 60.5%. Best responses did not differ significantly for several subgroups of patients, with 51.2% objective responses for patients with more than three treatment lines of therapy before BV, 64.3% for patients with more than one auto- or allo-SCT before BV, 62.9% for patients relapsing

<6 months after a first transplant and 58.2% for patients with HL refractory to first-line therapy (or with a first median response duration <12 months). Notably, older patients had poorer response rates: objective responses were observed for only 39.3% of the 28 patients >60 years and 40.8% of the 49 >50 year old patients, with CR rates of 21.4% and 22.4%, respectively. Response at the end of BV treatment. Patients received a median of six BV cycles (range: 1-16). Responses to BV treatment were assessed using the 1999 Cheson criteria for 22.3% and the 2007 Cheson criteria for 77.2% of patients. Response rates at the end of BV treatment cycles were dramatically lower than the best response rates after a median of four cycles, because of relapses or progression. For the whole cohort, the objective response rate after a median of six BV cycles was 33.5% (versus 60.4% best response after four cycles), with CR for 52 (21.8%), uCR for 5 (2.1%), PR for 23 (9.6%), SD for 8 (3.3%) and PD for 129 (54%) patients. Response duration. The median response duration was 8.4 [95% CI, 5.7-13.8] months for the 145 patients in CR/uCR/PR. Focusing on the 23 responders who had relapsed after allo-SCT before BV, the median response duration was slightly longer at 10.3 [95% CI, 3.9-20.2] months. DFS. For the whole cohort, 81 patients achieved CR/uCR as best responses after BV. The median DFS for these 81 patients was 18.7 months. For the 37 patients with prior allo-SCT before BV, median DFS was 15.4 (95% CI [6.8-not reached (NR)]) months. PFS. The median PFS was 6.8 months for 239 assessable patients and 11.3 months for 145 patients with objective responses (CR/uCR/PR). The quality of the best response

Table 2. Key responses in all patients and those relapsing after allo-SCT.

Response

Whole cohort

Patients with prior allo-SCT

Best response (after 4 cycles), n Patients with an objective response, n (%) Complete response (CR) Unconfirmed complete response (uCR) Partial response (PR) Stable disease (SD) Progressive disease (PD) Not assessable Response after end of BV cycles, n Objective response (CR/uCR/PR) Duration of best objective response, n Median response duration: months [95% CI] Disease-free survival (patients in CR/uCR), n Median DFS: months [95% CI] Progression-free survival, n Median PFS: months [95% CI] 1-year PFS probability: % [95% CI] 2-year PFS probability: % [95% CI] Overall survival, n Median OS: months [95% CI] 1-year OS probability: % [95% CI] 2-year OS probability: % [95% CI]

240 145 (60.4%) 70 (29.2%) 11 (4.6%) 64 (26.7%) 18 (7.5%) 68 (28.3) 9 (3.8%) 240 80 (33.3%) 145 8.4 [5.7-13.8] 81 18.7 [13.2-NR] 239 6.8 [5.7-8.0] 31.7 [25.6-38] 13.1 [7.4-20.5] 237 NR [20.9-NR] 76.4 [70.1-81.6] 57.8 [48.7-65.7]

37 23 (62.2%) 15 (40.5%) 2 (5.4%) 6 (16.2%) 5 (13.5%) 6 (16.2%) 3 (8.1%) â&#x20AC;&#x201C; â&#x20AC;&#x201C; 23 10.3 [3.9-20.2] 17 15.4 [6.8-NR] 37 7.9 [6.0-10.8] 32.3 [17.7-47.8] 7.7 [0.6-27.2] 36 NR [16.6-NR] 80.2 [62.8-90.0] 63.2 [42.1-78.3]

CI: confidence interval; NR: not reached.

haematologica | 2016; 101(4)

469


A. Perrot et al.

after BV was associated with a longer median PFS: 20.9 (95% CI [16.2-NR]) months for patients in CR versus 6.9 [95% CI, 5.8-8.3] months (P<0.0001). The median PFS did not differ significantly for patients who had received more than three treatment lines before BV (6.3 months), patients with more than one auto- or allo-SCT before BV (7.6 months), patients relapsing <6 months after a first SCT (6.3 months) and patients with HL refractory to first-line therapy or with a first median response duration <12 months (6.3 months). For the 37 patients given BV for relapse after allo-SCT, the median PFS was 7.9 [95% CI 6.0-10.8] months. In contrast, age was a prognostic factor for PFS after BV: the median PFS was significantly shorter for patients >60 years: at 4.4 versus 7.1 months (P<0.021). Progression/relapse after BV. Disease progressed or relapsed after BV in 156 (65.8%) patients: 123 (80.4% of them) received chemotherapy, 39 (25.7%) underwent transplantation (auto-SCT for 18 patients, with 3 having received tandem auto-/allo-SCT, and allo-SCT for 21 patients) and 17 (11.3%) were given radiotherapy.

OS. After a median follow-up of 16.1 months, median OS was not achieved; estimated 1-year OS was 76.4% [95% CI 70.1-81.6] and estimated 2-year OS was 57.8% [95% CI 48.7-65.7]. The most common causes among the 76 deaths were HL progression (68%), concurrent illness (9.3%) and consolidation toxicity (5.3%). OS did not differ significantly for patient subsets according to the number of prior treatment lines received or transplants before BV, for patients with lasting first transplant to relapse interval <6 months or for those with HL refractory to first-line therapy or with remission lasting <12 months. For the 37 patients given BV for relapse after allo-SCT, their respective estimated 1-year and 2-year OS rates were, interestingly, slightly longer than for the entire cohort: 80.2% [62.8-90.0] and 63.2% [42.1-78.3], respectively.

Impact of consolidation post-BV Eleven patients underwent consolidation radiotherapy after obtaining objective best responses to BV (5 CR and 6 PR); 2 patients in PR achieved CR after radiotherapy. Fifty-

A

B

Figure 2. Kaplan-Meier plots of probabilities of PFS and OS for the whole cohort.

470

haematologica | 2016; 101(4)


Post-brentuximab vedotin consolidation for CD30+ HL

seven (18.2%) patients received post-BV SCT consolidation, after obtaining objective responses for 54 patients and in a setting of SD for 3. Response was assessed using the 1999 Cheson criteria for 12.5% and the 2007 Cheson criteria for 87.5% of patients. Consolidation was auto-SCT for 29 patients and alloSCT for 28 patients, whose median age was 28 years, with only 4 patients >50 years, and the majority were men (61.8%). Before BV, 30 out of 57 patients had undergone

auto- or allo-SCT(s), predominantly auto-SCT. Twentyfour SCT-consolidated patients achieved CR, with no difference between auto-SCT and allo-SCT (Table 3). Both transplant groups were comparable with minor differences, especially in terms of the median number of BV cycles before transplantation (five cycles for the auto-SCT group versus six for the allo-SCT group, not significant). No significant difference was found for the median durations of responses, neither median PFS (at 18.8 [95% CI

Figure 3. PFS according to auto- or allo-SCT consolidation for 145 patients with objective responses to BV.

Table 3. Baseline characteristics and outcomes of 57 patients who received consolidation after BV.

Characteristic Median age at diagnosis: years [range] Age >60 years Age >50 years Male/female: n (%) 1 prior auto-SCT 2 prior auto-SCTs 1 prior allo-SCT Tandem auto/auto- SCT Tandem auto/allo- SCT Total BV cycles: n [range] Median BV-consolidation interval: months [95% CI] Cycles for best response: n [range] Objective best response: n CR/uCR: n PR: n SD/PD: n Median response duration (n=54): months [95% CI] Median PFS: months Estimated 1-year OS: % [95% CI] Estimated 2-year OS: % [95% CI]

Consolidation n=57

Autologous-SCT n=29

Allogeneic-SCT n=28

28 [15-63] 1 4 35 (61.4)/22 (38.6) 25 2 3 2 2 6 [2-13] 4.4 [2.1-11.8] 4 [2-12] 54 34/3 17 /0 NR 18.8 [15.0-NR]

29 [19-52] 0 2 17 (58.6)/12 (41.4) 2 0 0 0 0 5 [2-13] 3.7 [2.1-10.3] 4 [2-12] 27 17/2 8 2/0 NR [7.2-NR] NR [9.2-NR] 88.7 [68.8-96.3] 79.9 [50.9-92.8]

27.5 [15-63] 1 2 18 (64.3)/10 (35.7) 23 2 3 2 2 6 [2-11] 6.2 [2.5-11.8] 4 [2-8] 27 17/1 9 1/0 17.3 [8.5-NR] 18.8 [12.6-NR] 87.1 [65.0-95.7] 81.3 [56.9-92.7]

CI: confidence interval; NR: not reached.

haematologica | 2016; 101(4)

471


A. Perrot et al.

15.0 NR] months) nor median OS (not reached). Probabilities of 1-year and 2-year OS were, respectively, 88.7% and 79.9% after auto-SCT consolidation and 87.1% and 81.3% after allo-SCT (not significant). Patients who underwent auto- or allo-SCT as consolidation after BV had better outcomes than those not prescribed a consolidation. Among the 145 objective responders to BV, 54 patients transplanted as consolidation had a median PFS of 18.8 versus 8.7 months for the 91 patients without transplant (P<0.0001) (Figure 3). These results confirmed the importance of considering BV as a bridge to transplant. Consolidation feasibility was also demonstrated for the 57 patients with allo-SCT before BV: 15 of them underwent consolidation after BV, either an auto-SCT (n=5), a second allo-SCT (n=4), or one or more DLIs (n=6). For 41 very high-risk patients with refractory HL (refractory to first-line therapy or remission lasting <12 months) who received BV followed by transplant as consolidation, the median PFS was a very promising 18.5 [95% CI, 3.928.8] months.

Discussion After a median of four BV cycles, the ORR for refractory/relapsed HL patients was 60.4%, including 33.8% CR/uCR. However, at the end of BV treatment (a median of six cycles), the response rate declined to 23.9% CR/uCR and 9.6% PR. For the 145 responders, the median response duration was 8.4 months. After a median followup of 16 months, the median PFS slightly exceeded 6 months, while median OS had not been reached, with the estimated 1-year and 2-year OS of 76.4% and 57.8%, respectively. BV had an acceptable toxicity profile, with no treatment-related deaths, while grade 1/2 sensory neuropathy occurred in 27% of the patients. These results for a heavily pretreated patient cohort are comparable to those obtained in the pivotal phase II trial (75% ORR, 34% CR, median response duration 6.7 months and median PFS 5.6 months).5 In our cohort, 48.7% of the patients had primary refractory HL and did not undergo auto-SCT prior to BV, which indicates a more resistant disease than that of patients in the phase II trial.5 BV achieved CR or stable PR that served as a bridge to SCT. Among the 57 patients who received a consolidation transplant after BV, 30 had never been transplanted before BV: 27 auto-SCT and 3 allo-SCT. The median PFS for this subgroup was about 18 months. The results of other studies have suggested this bridge to transplant concept, with a respective median PFS at 5.1 months, 1-year PFS at 92.3% and PFS at 50% with a 13 month follow-up for primary refractory HL.7,12,13 The findings of a comparative study indicated that BV before alloSCT improved 2-year PFS.14 As a targeting agent, BV is an ideal treatment before transplant, that enabled some patients to proceed to transplantation with a CR and no organ toxicity. The results of the pivotal phase II trial were updated with a median follow-up of 3 years.15 Among their 34 patients who obtained CR, 47% remained progressionfree after a median of 53.3 months. A third of 18 patients remaining in CR or PR underwent allo-SCT consolidation. Even though durable remissions post-BV without additional treatment are possible, our results for our large 472

Post-brentuximab vedotin consolidation for CD30+ HL

cohort demonstrated the strong impact of auto- or alloSCT consolidation on the outcomes of high-risk HL patients responding to BV: median PFS was markedly prolonged from 8.7 to 18.8 months after transplant. Consolidation radiotherapy could be an alternative option for patients not eligible for transplant or in PR after BV: in our experience, 2 patients in PR achieved CR after additional radiotherapy, as supported by a case report.16 In the pivotal phase II trial, patients had not undergone allo-SCT before BV.5 BV efficacy and safety after allo-SCT were evaluated in 24 and 16 HL patients17,18 with ORRs of between 50% and 69%. Thirty-seven patients received BV in the French NPP, with an ORR of 60.5% (including 39.5% CR) and a median PFS of 7.9 months; there was no difference compared to patients without previous alloSCT. The possibility of combining BV with DLI had previously been shown for 4 patients,19 wherein the authors hypothesized that concomitant BV and DLI could potentially further enhance the graft-versus-lymphoma effect, perhaps through the release of cytokines such as interleukin-6, chemokine ligand 17 and/or tumor necrosis factor-α. In our experience with 57 patients relapsing after allo-SCT, BV was followed by one or several DLIs for 6 patients and by a second allo-SCT for 5; all of the DLI recipients and 4 of the 5 second allo-SCT patients were alive at the last follow-up. Although BV showed a remarkable efficacy against high-risk relapsed, refractory HL, 7.5% and 28.3% of our cohort patients achieved only SD and PD, respectively. The use of BV as first-line salvage therapy seems to have improved the response rates of 37 phase II trial patients (ORR 75%, CR 34%).20 Promising associations are BV-bendamustine or a sequential combination of gemcitabine, vinorelbine, pegylated liposomal doxorubicin and BV: a phase I-II trial enrolling 45 patients and a retrospective analysis of 11 patients obtained, respectively, 94% ORR and 82% CR, or 100% ORR and 72.5% CR, with both regimens having manageable adverse events.21,22 Combining BV with PD-1 blockade therapies could also warrant being assessed in the future. In conclusion, the results of our large retrospective analysis of 240 patients with refractory/relapsed HL from 89 treatment centers (including 37 patients with prior alloSCT) in a real-life setting support the previously reported BV efficacy in heavily pretreated CD30+ HL patients with manageable toxicity. Due to the short response duration of about 8.4 months, and the improved outcomes with consolidation, we recommend early auto- or allo-SCT or even radiotherapy to consolidate responses and to cure HL. Resorting to DLI or a second allo-SCT after BV achieves durable remissions in this difficult-to-treat setting of post-allograft relapses. Acknowledgments The authors would like to thank Sophie Pallardy and Sami Boussetta for statistical analysis, the Lymphoma Study Academic Research Organisation (LYSARC) for study management and the clinicians of the LYSA who contributed to the collection of clinical data, especially M. Larmarque, A. Contejean & C. Haioun (Créteil), A. Quinquenel (Reims), A. Chauchet (Besançon), C. Tomowiak (Poitiers), S. François (Angers), L. Fornecker (Strasbourg), S. Garciaz (Marseille), M. Lenoble (Montfermeil), G. Salmeron (Versailles), L. Terriou (Lille). haematologica | 2016; 101(4)


Post-brentuximab vedotin consolidation for CD30+ HL

References 1. Van Den Neste E, Casasnovas O, André M, et al. Classical Hodgkin’s lymphoma: the Lymphoma Study Association guidelines for relapsed and refractory adult patients eligible for transplant. Haematologica. 2013;98(8):1185-1195. 2. Küppers R. The biology of Hodgkin’s lymphoma. Nat Rev Cancer. 2009;9(1):15-27. 3. Katz J, Janik JE, Younes A. Brentuximab vedotin (SGN-35). Clin Cancer Res. 2011;17(20):6428-6436. 4. Younes A, Bartlett NL, Leonard JP, et al. Brentuximab vedotin (SGN-35) for relapsed CD30-positive lymphomas. N Engl J Med. 2010;363(19):1812-1821. 5. Younes A, Gopal AK, Smith SE, et al. Results of a pivotal phase II study of brentuximab vedotin for patients with relapsed or refractory Hodgkin’s lymphoma. J Clin Oncol. 2012;30(18):2183-2189. 6. Rothe A, Sasse S, Goergen H, et al. Brentuximab vedotin for relapsed or refractory CD30+ hematologic malignancies: the German Hodgkin Study Group experience. Blood. 2012;120(7):1470-1472. 7. Gibb A, Jones C, Bloor A, et al. Brentuximab vedotin in refractory CD30+ lymphomas: a bridge to allogeneic transplantation in approximately one quarter of patients treated on a Named Patient Programme at a single UK center. Haematologica. 2013;98(4): 611-614. 8. Zinzani PL, Viviani S, Anastasia A, et al. Brentuximab vedotin in relapsed/refractory Hodgkin’s lymphoma: the Italian experience and results of its use in daily clinical practice outside clinical trials. Haematologica.

haematologica | 2016; 101(4)

2013;98(8):1232-1236. 9. Monjanel H, Deville L, Ram-Wolff C, et al. Brentuximab vedotin in heavily treated Hodgkin and anaplastic large-cell lymphoma, a single centre study on 45 patients. Br J Haematol. 2014;166(2):306-308. 10. Cheson BD, Horning SJ, Coiffier B, et al. Report of an international workshop to standardize response criteria for non-Hodgkin’s lymphomas. NCI Sponsored International Working Group. J Clin Oncol. 1999;17(4): 1244. 11. Cheson BD, Pfistner B, Juweid ME, et al. Revised response criteria for malignant lymphoma. J Clin Oncol. 2007;25(5):579-586. 12. Chen R, Palmer JM, Thomas SH, et al. Brentuximab vedotin enables successful reduced-intensity allogeneic hematopoietic cell transplantation in patients with relapsed or refractory Hodgkin lymphoma. Blood. 2012;119(26):6379-6381. 13. Garciaz S, Coso D, Peyrade F, et al. Brentuximab vedotin followed by allogeneic transplantation as salvage regimen in patients with relapsed and/or refractory Hodgkin’s lymphoma. Hematol Oncol. 2014;32(4):187-191. 14. Chen R, Palmer JM, Tsai N-C, et al. Brentuximab vedotin is associated with improved progression-free survival after allogeneic transplantation for Hodgkin lymphoma. Biol Blood Marrow Transplant. 2014;20(11):1864-1868. 15. Gopal AK, Chen R, Smith SE, et al. Durable remissions in a pivotal phase 2 study of brentuximab vedotin in relapsed or refractory Hodgkin lymphoma. Blood. 2015;125(8):1236-1243. 16. Dozzo M, Zaja F, Volpetti S, Sperotto A, Magli A, Fanin R. Brentuximab vedotin in

17.

18.

19.

20.

21.

22.

combination with extended field radiotherapy as salvage treatment for primary refractory Hodgkin lymphoma. Am J Hematol. 2015;90(4):E73. Gopal AK, Ramchandren R, O’Connor OA, et al. Safety and efficacy of brentuximab vedotin for Hodgkin lymphoma recurring after allogeneic stem cell transplantation. Blood. 2012;120(3):560-568. Carlo-Stella C, Ricci F, Dalto S, et al. Brentuximab vedotin in patients with Hodgkin lymphoma and a failed allogeneic stem cell transplantation: results from a named patient program at four Italian centers. Oncologist. 2015;20(3):323-328. Theurich S, Malcher J, Wennhold K, et al. Brentuximab vedotin combined with donor lymphocyte infusions for early relapse of Hodgkin lymphoma after allogeneic stemcell transplantation induces tumor-specific immunity and sustained clinical remission. J Clin Oncol. 2013;31(5):e59-63. Chen RW, Palmer J, Martin P, et al. Results of a phase II trial of brentuximab vedotin as first line salvage therapy in relapsed/refractory HL prior to AHCT. Blood. 2014;124 (21):abst 501. LaCasce A, Bociek RG, Matous J, et al. Brentuximab vedotin in combination with bendamustine for patients with Hodgkin lymphoma who are relapsed or refractory after frontline therapy. Blood. 2014;124 (21):abst 293. Michallet A-S, Guillermin Y, Deau B, et al. Sequential combination of gemcitabine, vinorelbine, pegylated liposomal doxorubicin and brentuximab as a bridge regimen to transplant in relapsed or refractory Hodgkin lymphoma. Haematologica. 2015;100(7):e269-271.

473


ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Hodgkin Lymphoma

Ferrata Storti Foundation

Single or tandem autologous stem-cell transplantation for first-relapsed or refractory Hodgkin lymphoma: 10-year follow-up of the prospective H96 trial by the LYSA/SFGM-TC study group David Sibon,1 Franck Morschhauser,2 Matthieu Resche-Rigon,3 David Ghez,4 Jehan Dupuis,5 Ambroise Marçais,1 Bénédicte Deau-Fischer,6 Reda Bouabdallah,7 Catherine Sebban,8 Gilles Salles,9 and Pauline Brice3

Hématologie, Hôpital Universitaire Necker – Enfants Malades, Assistance Publique – Hôpitaux de Paris (AP–HP), Université Paris Descartes – Sorbonne Paris Cité, Imagine Institute; 2Hôpital Claude Huriez, Lille; 3Hôpital Saint-Louis, AP–HP, Université Paris Diderot, Sorbonne Paris Cité; 4Institut Gustave Roussy, Villejuif; 5Hôpital Henri-Mondor, AP–HP, Université Paris-Est, Créteil; 6Hôpital Cochin, AP–HP, Université Paris Descartes, Sorbonne Paris Cité; 7Institut Paoli-Calmettes, Marseille; 8Centre Léon Bérard, Lyon; and 9 Centre Hospitalier Lyon-Sud, Hospices Civils de Lyon, Université Claude-Bernard Lyon 1, Pierre-Bénite, France 1

Haematologica 2016 Volume 101(4):474-481

ABSTRACT

W

Correspondence: david.sibon@aphp.fr

Received: September 9, 2015. Accepted: December 17, 2015 Pre-published: December 31, 2015. doi:10.3324/haematol.2015.136408

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/4/474

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

474

e assessed the long-term results of autologous stem-cell transplantation for patients with first-relapsed or refractory Hodgkin lymphoma included in the prospective Lymphoma Study Association/Société Française de Greffe de Moelle H96 trial. This large multicenter phase II trial evaluated a risk-adapted strategy with single or tandem autologous stem-cell transplantation for 245 Hodgkin lymphoma patients. Poor-risk patients (n=150) had primary refractory Hodgkin lymphoma (n=77) or ≥2 risk factors at first relapse (n=73) and were eligible for tandem autologous stem-cell transplantation. Intermediate-risk patients (n=95) had one risk factor at first relapse and were eligible for single autologous stemcell transplantation. With a median follow-up of 10.3 years, 10-year freedom from second failure and overall survival rates were, respectively: 64% (95% CI, 54% to 74%) and 70% (95% CI, 61% to 80%) for the intermediate-risk group, and 41% (95% CI, 33% to 49%) and 47% (95% CI, 39% to 55%) for the poor-risk group. Considering only patients who did not relapse after completing autologous stem-cell transplantation, the 15-year cumulative incidences of second primary malignancies were 24% for the 70 intermediate-risk patients and 2% for the 75 poor-risk ones. With long-term follow-up, the risk-adapted strategy remains appropriate. Tandem autologous stem-cell transplantation can still be considered an option for poor-risk patients, but integration of positron-emission tomography findings and new drugs may help to refine the need for a second autologous stem-cell transplant and possibly improve outcomes of patients with first-relapsed or refractory Hodgkin lymphoma.

Introduction The true efficacy of a given treatment is only evident after prolonged follow-up. To determine it, analyses of long-term prospective rather than retrospective data are needed. Two randomized studies established the advantage of autologous stem-cell transplantation (ASCT) over standard-dose salvage treatment for patients with relapsed Hodgkin lymphoma (HL) sensitive to chemotherapy.1,2 However, longhaematologica | 2016; 101(4)


Long-term results of the H96 trial

term prospective data on the efficacy and late effects of ASCT are lacking. Moreover, the long-term benefit of ASCT for patients with primary refractory HL has not been studied prospectively. In 2008, our group published a prospective analysis, the H96 trial, whose primary end-point was to evaluate freedom from second failure (FF2F) for poor- and intermediate-risk HL groups.3 The results of this trial showed the interest of a risk-adapted strategy with single or tandem ASCT. The aim of the present study was to assess prospectively the long-term results and late effects of ASCT for first-relapsed or refractory HL in H96 trial patients.

Methods Information on the methods has already been published in detail3 and is briefly summarized below. The study protocol was approved by the Ethics Committee of Saint-Louis Hospital (Paris, France).

(SD) or progressive disease (PD) – was defined according to the 1999 international response criteria.4 Left ventricular ejection fraction was assessed at baseline, after salvage treatment and was recommended yearly thereafter.

Statistical analysis Analyses were performed on an intent-to-treat basis (inclusion of every patient who received at least one cycle of salvage treatment) and per-protocol (restricted to the patients who completed ASCT). The primary end-point was FF2F, measured from the date of inclusion until progression, relapse or death from any cause, or the date of last contact for those who were failure-free. Overall survival (OS) was measured from the date of inclusion until death from any cause, or the date of last contact for those who were alive. FF2F and OS were estimated with 95% confidence intervals (95% CI). Survival curves were generated using the Kaplan-Meier method and compared using the log-rank test, with P≤0.05 defining statistical significance. Cumulative incidences were compared using the Gray test.

Results Patients Eligibility criteria were as follows: biopsy-proven HL (World Health Organization, classic type); either primary refractory or first-relapsed HL; age less than 60 years (age ≤50 years for patients scheduled to receive tandem ASCT). Written informed consent was required before enrollment.

Stratification and treatment In the H96 trial, the intensity of high-dose therapy (single or tandem ASCT) was adapted to risk assessed at the onset of salvage treatment, based on the primary refractory status or the number of risk factors at first relapse, which included relapse less than 12 months after the end of first-line treatment, stage III or IV at relapse, and/or relapse in a previously irradiated site (>30 Gy) after combined-modality therapy. Patients were stratified as follows: the poor-risk group included patients with primary refractory HL or two or more risk factors at relapse; and the intermediate-risk group comprised patients with only one risk factor at relapse. In the poor-risk group, salvage treatment was followed by tandem ASCT. Salvage treatment consisted of two cycles of ifosfamide, etoposide, and doxorubicin (IVA75) or mitoguazone, ifosfamide, vinorelbine, and etoposide (MINE). The first conditioning regimen consisted of cyclophosphamide, carmustine, etoposide, and mitoxantrone (CBVM) or carmustine, etoposide, cytarabine, and melphalan (BEAM). A second conditioning regimen was reserved for patients with no evidence of disease progression at that time. For previously unirradiated patients, it consisted of total-body irradiation (12 Gy in 6 × 2 Gy twice-daily fractions), cytarabine, and melphalan (TAM). For patients who had received prior dose-limiting radiation, the second conditioning regimen was BAM (the same as TAM except that busulfan replaced total body irradiation). After the second ASCT, radiotherapy was optional. In the intermediate-risk group, salvage treatment was followed by single ASCT. Salvage treatment consisted of three IVA50 or MINE cycles and the conditioning regimen was BEAM. After ASCT, radiotherapy was optional.

Follow-up Computed-tomography scans were performed 3 months after the last ASCT, every 6 months until 3 years after the ASCT and yearly thereafter. Response to treatment – complete response (CR), unconfirmed CR (CRu), partial response (PR), stable disease haematologica | 2016; 101(4)

Between 1995 and 2002, 245 patients were enrolled. Their characteristics at diagnosis and at the time of treatment failure/relapse were reported previously.3 Briefly, the median age at inclusion was 32 (range, 11-60) years. The poor-risk group included 150 patients (77 with primary refractory HL and 73 with unfavorable relapse) and the intermediate-risk group comprised 95 patients. The median follow-up is now 10.3 years (10.4 years for the poorrisk group and 10.1 years for the intermediate-risk group).

Outcomes according to risk group Outcomes based on intent-to-treat analysis. For the poor-risk group, the 10-year FF2F and OS were 41% (95% CI, 33% to 49%) and 47% (95% CI, 39% to 55%), respectively (Figure 1). Within this group, no significant difference was observed, respectively, between primary refractory HL and unfavorable relapse for the 10-year FF2F (34% and 49%; P=0.09) and OS (43% and 51%; P=0.26) (Figure 1). For the intermediate-risk group, the 10-year FF2F and OS were 64% (95% CI, 54% to 74%) and 70% (95% CI, 61% to 80%), respectively (Figure 1). Outcomes of patients completing ASCT. In the poor-risk group, for patients completing tandem ASCT (n=105), the 10-year FF2F and OS were 50% (95% CI, 41% to 60%) and 56% (95% CI, 46% to 66%), respectively (Figure 2). Within this group, no significant difference was observed, respectively, between primary refractory HL and unfavorable relapse for the 10-year FF2F (48% and 52%; P=0.76) and OS (58% and 54%; P=0.91) (Figure 2). In the intermediate-risk group, for patients completing ASCT (n=92), the 10-year FF2F and OS were 65% (95% CI, 55% to 75%) and 72% (95% CI, 62% to 81%), respectively (Figure 2).

Outcomes according to the response to salvage treatment The response to salvage treatment was assessable for 243 patients (76 with primary refractory HL, 73 with unfavorable relapse and 94 with intermediate risk). For the poor-risk group, the 10-year FF2F according to each response category was 65% (95% CI, 49% to 80%) for CR/CRu (n=39), 47% (95% CI, 34% to 60%) for PR (n=55), unassessable (follow-up <10 years) for SD (n=24), 475


D. Sibon et al.

and 23% (95% CI, 8% to 37%) for PD (n=31) (Online Supplementary Figure S1). Significant differences in FF2F were found between the CR/CRu and PR groups (P=0.03), PR and SD groups (P=0.002), and PR and PD groups (P=0.001). The 10-year OS was 68% (95% CI, 51% to 84%) for patients with CR/CRu, 58% (95% CI, 45% to 71%) for those with PR, 16% (95% CI, 0 to 31%) for patients with SD, and 26% (95% CI, 10% to 41%) for those with PD (Online Supplementary Figure S1). Concerning OS, no significant differences were observed between the CR/CRu and PR groups (P=0.1), whereas there were statistically significant differences between the CR/CRu and SD groups (P<0.001); the CR/CRu and PD groups (P<0.001); the PR and SD groups (P<0.001); and the PR and PD groups (P<0.001). For the intermediate-risk group, the 10-year FF2F was 67% (95% CI, 55% to 79%) for patients with CR/CRu (n=65) and 64% (95% CI, 44% to 83%) for those with PR (n=26), with no significant difference between the two (P=0.95) (Figure 3). No patients had SD and three had PD. The 10-year OS was 72% (95% CI, 61% to 84%) for patients with CR/CRu and 66% (95% CI, 47%

to 85%) for those with PR (P=0.97) (Online Supplementary Figure S1).

A

C

B

D

Relapse/progression after study inclusion One-hundred patients relapsed/progressed after study inclusion (75 poor-risk and 25 intermediate-risk patients). For the poor-risk group, the 5- and 10-year cumulative incidences of relapse/progression were 48% and 53%, respectively (Figure 3). Within this group, the 5- and 10year cumulative incidences of relapse/progression were 53% and 59% for the 43 primary refractory patients, and 43% and 47% for the 32 patients with unfavorable relapse, respectively (P=0.12) (Figure 3). Among the 43 poor-risk patients who relapsed after completing tandem ASCT, the median interval from second ASCT to relapse was 0.7 (0.2–8) years. For the intermediate-risk group, the 5- and 10-year cumulative incidences of relapse were 25% and 27%, respectively (Figure 3). Among the 23 patients who relapsed after ASCT, the median ASCT-to-relapse interval was 1.5 (0.4–6.5) years. After post-inclusion relapse/progression, 23 patients

Figure 1. Outcomes based on intent-to-treat analysis. (A) Freedom from second failure (FF2F) and (B) overall survival (OS) of patients with first-relapsed or primary refractory Hodgkin lymphoma according to risk group. (C) FF2F and (D) OS of patients with primary refractory or unfavorable relapse (≥2 of the following risk factors at first relapse: relapse <12 months, stage III or IV at relapse, and/or relapse within previously irradiated sites). Median follow-up: 10.3 years.

476

haematologica | 2016; 101(4)


Long-term results of the H96 trial

underwent allogeneic stem cell transplantation (alloSCT), 20 of them after a reduced-intensity conditioning (RICallo) regimen. Overall, from that relapse, the 10-year OS was 30% for the 23 patients or considering only the 20 RIC-allo patients (Figure 3). For the 77 relapsing/progressing patients who did not undergo alloSCT, the 10-year OS was only 5% (19% for the 20 intermediate-risk and 2% for the 57 poor-risk patients (Figure 3). To assess the effect of RIC-allo in this setting better, we conducted a matched analysis using three relevant risk factors: age at relapse after ASCT (â&#x2030;Ľ50 years),5 early relapse after ASCT (<6 months),5 and treatment arm (poor versus intermediate). The 20 RIC-allo patients had undergone at least one previous ASCT, they were all younger than 50 years at relapse and four (all in the poor-risk group) had early relapse after ASCT. They were matched to 20 control patients who relapsed after ASCT, but did not undergo alloSCT. From the post-ASCT relapse, the median OS was 31.7 months for RIC-allo patients versus 38.9 months for matched controls, and the 10-year OS was 30% for RICallo patients versus 20% for matched controls; the survival curves did not differ significantly (P=0.55) (Figure 3).

Death In all, 110 patients died [30 in the intermediate-risk group and 80 in the poor-risk group (35 unfavorable relapses and 45 primary refractory HL)]. Overall, HL was the main cause of death (n=83; 75% of causes of death). However, HL was the cause of 50% (n=15) of deaths in the intermediate-risk group whereas it represented 85% (n=68) of deaths in the poor-risk group. Other deaths resulted from infections (n=5), acute leukemia (n=5), postalloSCT complications (n=5), solid tumors (n=3), nonHodgkin lymphomas (n=2), cardiac toxicity (n=2), or other (n=5). The distribution of causes of death according to risk group is shown in Online Supplementary Figure S2.

Late effects Second primary malignancies (SPM). Sixteen SPM occurred (Table 1). Overall, from inclusion, the 10- and 15-year cumulative incidences of SPM were 8% and 15%, respectively (Figure 4). According to whether the patients were intermediate- or poor-risk, the 10-year cumulative incidences of SPM were 15% and 1.5%, respectively (Online Supplementary Figure S3). Considering only patients who did not relapse after completing ASCT, the 10- and 15-

A

C

B

D

Figure 2. Outcomes of patients completing autologous stem-cell transplantation. (A) Freedom from second failure (FF2F) and (B) overall survival (OS) of patients with first-relapsed or primary refractory Hodgkin lymphoma according to risk group. (C) FF2F and (D) OS of patients with primary refractory or unfavorable relapse.

haematologica | 2016; 101(4)

477


D. Sibon et al.

Discussion

year cumulative incidences of SPM were 9% and 13%, respectively (Figure 4), with 10- and 15-year cumulative incidences of 16% and 24%, respectively, for the 70 intermediate-risk patients and 2% and 2% for the 75 poor-risk patients (Online Supplementary Figure S3). All acute leukemias were fatal, while some patients with nonHodgkin lymphomas or solid tumors achieved long-term survival. Cardiovascular events. These were defined as heart failure requiring treatment, coronary artery disease requiring treatment or death resulting from a cardiac or vascular cause. Thirteen cardiovascular events occurred. Overall, from inclusion, the 10-year cumulative incidence of cardiovascular events was 6.5% (Figure 4) (10% for the intermediate-risk group and 3% for the poor-risk group; Online Supplementary Figure S3). Considering only those patients not relapsing after completing ASCT, the 10-year cumulative incidence of cardiovascular events was 6% (Figure 4) (7% for the intermediate-risk group and 5% for the poorrisk group; Online Supplementary Figure S3).

The long 10.3-year median follow-up of H96-trial patients gave us a unique opportunity to analyze longterm outcomes of a large series of adults with firstrelapsed/refractory HL. The 10-year OS was 70% for the intermediate-risk group and 47% for the poor-risk group. Several studies have shown that single ASCT can provide a cure in roughly 50% of unselected patients.6,7 Thus, the outcome of intermediate-risk patients treated with single ASCT in our trial was better than the outcome of these unselected patients treated with single ASCT. Moreover, the outcome of poor-risk patients treated with tandem ASCT in our trial was similar to the outcome of these unselected patients treated with single ASCT. The results of the SWOG 0410 phase II trial of tandem ASCT in 82 refractory/relapsed HL patients were recently reported: the 2-year progression-free survival (PFS) was 63%, reaching the predicted end-point of at least a 15% improvement based on

A

C

B

D

Figure 3. Cumulative incidences of relapse/progression and overall survival probabilities. Cumulative incidence of relapse/progression after study inclusion (A) according to risk group and (B) for patients with primary refractory or unfavorable relapse. (C) From post-inclusion relapse/progression, overall survival (OS) of patients who underwent allogeneic stem-cell transplantation (alloSCT) or not. (D). From post-autologous stem-cell transplantation relapse, OS after matched analysis between patients who underwent alloSCT using a reduced-intensity conditioning regimen (RIC-allo) and patients who did not undergo alloSCT.

478

haematologica | 2016; 101(4)


Long-term results of the H96 trial

the historical 2-year PFS of 45% in the previous SWOG 9011 study.8,9 Late effects, especially SPM and cardiovascular events, are well-known after first-line treatment of HL.10-16 In contrast, only a few retrospective studies have addressed long-term effects after ASCT for relapsed or refractory HL.17,18 Thus, long-term prospective data are lacking for first-relapsed or refractory HL treated with ASCT. Concerning the cumulative incidence of SPM, data from retrospective studies are difficult to interpret because of differences in median follow-up, selection of patients (all transplanted patients or only patients who survived ≥2 years after ASCT), period of inclusion, number of treatment lines before ASCT, heterogeneity of treatments, conditioning regimens and source of stem-cells.7,17-19 Given this, the cumulative incidences of SPM in retrospective studies ranged from 5.8% to 14.7% at 10 years, and from 8% to 15.3% at 15 years. In the present study, considering patients who did not relapse after completing single ASCT (intermediate-risk group), the 10- and 15-year cumulative incidences of SPM were 16% and 24%, respectively, which are higher than in retrospective studies. These results suggest that special attention to SPM should be recommended. The relatively low cumulative incidence of SPM in the poor-risk patients who did not relapse after completing ASCT is somewhat surprising. Data concerning SPM were collected in the same way in both risk groups without bias. We have no clear explanation for this low incidence, but we cannot rule out an effect due to chance. Our study has some limitations. Since the start of this trial in 1995, changes have taken place in the standard treatment of patients with HL. Fluorodeoxyglucose positron-emission tomography (PET) scanning was not routinely done for disease assessment, and it is possible that PET scanning done before ASCT could have more

accurately classified patients’ responses to salvage treatment. In the retrospective study by Devillier et al., for responders to salvage treatment according to the Cheson 2007 criteria, the PET response at the time of ASCT influenced outcome and identified patients requiring single or tandem ASCT.20 Additionally, because the H96 trial was not a randomized study comparing single versus tandem ASCT, tandem ASCT can only be considered an option for poor-risk patients. Various treatment strategies to improve outcomes after ASCT have been investigated, including PET-adapted approaches,21,22 dose-intensity of salvage treatment,23 radiation therapy after ASCT,24 tandem auto– alloSCT,25 or consolidation with brentuximab vedotin (BV).26 Recently, the AETHERA randomized trial showed that early consolidation with BV versus placebo after ASCT improved 2-year PFS in patients with HL with risk factors for relapse or progression after transplantation (63% versus 51%, respectively; hazard ratio 0,57; P=0.0013).26 Interim analysis of OS showed no significant difference between treatment groups, and, because the crossover of patients in the placebo group to the BV group confounds this survival analysis, whether early BV consolidation can provide a greater survival benefit than BV treatment after progression cannot yet be answered. A direct comparison between AETHERA and H96 is difficult given the differences in median follow-up (2.5 years for AETHERA versus 10.3 years for H96) and inclusion criteria: in AETHERA, patients were included after ASCT and had at least one of the following risk factors: primary refractory HL, relapsed HL with an initial remission duration of less than 12 months, or relapse after 12 or more months with extranodal involvement. In H96, patients were included at the time of relapse/progression, before salvage treatment, and were stratified into two risk groups. Furthermore, in AETHERA, patients had to have had CR, PR, or SD after salvage treatment, whereas in H96, all

Table 1. Characteristics of the 16 second primary malignancies.

Patient 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

SPM

Risk group

N. of ASCT

Post-ASCT HL relapse

ASCT-to-SPM interval (years)

Age at SPM onset (years)

Last follow- up

SPM: cause of death

DLBCL DLBCL BL PTCL CD30+ALK– AML AML AML AML ALL Lung cancer Lung cancer Lung cancer Lung cancer Breast cancer Kidney cancer Oligodendroglioma

Intermediate Intermediate Intermediate Intermediate Intermediate Intermediate Intermediate Intermediate Intermediate Intermediate Intermediate Intermediate Poor Intermediate Poor Intermediate

1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 1

No No No No No No No Yes No No No Yes Yes No No No

8.8 3.1 10.5 4.8 4.5 7 4.8 10.9 1.3 9.9 7.1 6.3 10.8 4.5 6.4 4.1

45 61 67 39 43 35 34 60 44 64 56 52 60 43 46 46

Alive Died Died Alive Died Died Died Died Died Died Died Died Alive Alive Alive Alive

--Yes Yes --Yes Yes Yes Yes Yes Yes Yes Yes ---------

ALK: anaplastic lymphoma kinase; ALL: acute lymphoblastic leukemia; AML: acute myeloid leukemia; ASCT: autologous stem-cell transplantation; BL: Burkitt lymphoma; DLBCL: diffuse large B-cell lymphoma; HL: Hodgkin lymphoma; PTCL: peripheral T-cell lymphoma.

haematologica | 2016; 101(4)

479


D. Sibon et al.

A

C

B

D

Figure 4. Cumulative incidence of second primary malignancies (SPM) and cardiovascular events. Cumulative incidence of SPM (A) for the whole population and (B) for patients who did not relapse after completing autologous stem-cell transplantation (ASCT). Cumulative incidence of cardiovascular events (heart failure requiring treatment, coronary artery disease requiring treatment or death resulting from a cardiac or vascular cause) (C) for the whole population and (D) for patients who did not relapse after completing ASCT.

patients could undergo the first ASCT whatever their response, including PD. Because the definition of primary refractory HL is not equivocal, it is probably more adequate to analyze results of primary refractory patients between AETHERA and H96. In AETHERA, the 2-year PFS (from randomization after ASCT) of primary refractory patients was 60% and 42% in the BV and placebo arms,27 respectively, whereas in H96, the 2-year PFS (from ASCT) of primary refractory patients in CR, PR, or SD after salvage treatment was 67% according to the intentto-treat analysis, and 74% for patients completing tandem ACST (data not shown). In view of the results of AETHERA, early BV consolidation after ASCT may certainly be proposed to intermediate-risk patients but whether BV consolidation after the first ASCT may replace tandem ASCT in poor-risk patients remains an open question. Besides BV, many new drugs, including histone deacetylase inhibitors (mocetinostat and panobinostat),28,29 everolimus,30 bendamustine,31 and PD-1â&#x20AC;&#x201C;blocking antibodies (nivolumab32 and pembrolizumab33), have shown activity in relapsed/refractory HL and could be used either to increase the rate of CR prior to ASCT, or as consolida480

tion therapy after ASCT. Finally, the role of alloSCT in the management of patients who relapse after ASCT remains controversial. In our matched analysis, the 10-year OS did not differ significantly between RIC-allo and no alloSCT patients after post-ASCT relapse. However, the limited number of patients may have prevented the demonstration of a significant benefit of RIC-allo. Considering the 4-year OS (45% for RIC-allo patients and 35% for matched controls), our results are similar to those reported previously.5,34 In conclusion, with long-term follow-up, the risk-adapted strategy remains appropriate, but special attention should be paid to SPM. Integration of PET findings and new drugs such as BV or PD-1â&#x20AC;&#x201C;blocking antibodies may help to refine this strategy, especially the need for a second ASCT in poor-risk patients, and possibly improve outcomes for patients with first-relapsed or refractory HL. Acknowledgments The authors would like to thank Janet Jacobson for editorial assistance. haematologica | 2016; 101(4)


Long-term results of the H96 trial

References 1. Linch DC, Winfield D, Goldstone AH, et al. Dose intensification with autologous bonemarrow transplantation in relapsed and resistant Hodgkin's disease: results of a BNLI randomised trial. Lancet. 1993;341 (8852):1051-1054. 2. Schmitz N, Pfistner B, Sextro M, et al. Aggressive conventional chemotherapy compared with high-dose chemotherapy with autologous haemopoietic stem-cell transplantation for relapsed chemosensitive Hodgkin's disease: a randomised trial. Lancet. 2002;359(9323):2065-2071. 3. Morschhauser F, Brice P, Ferme C, et al. Riskadapted salvage treatment with single or tandem autologous stem-cell transplantation for first relapse/refractory Hodgkin's lymphoma: results of the prospective multicenter H96 trial by the GELA/SFGM study group. J Clin Oncol. 2008;26(36):5980-5987. 4. Cheson BD, Horning SJ, Coiffier B, et al. Report of an international workshop to standardize response criteria for non-Hodgkin's lymphomas. NCI Sponsored International Working Group. J Clin Oncol. 1999;17(4): 1244. 5. Martinez C, Canals C, Sarina B, et al. Identification of prognostic factors predicting outcome in Hodgkin's lymphoma patients relapsing after autologous stem cell transplantation. Ann Oncol. 2013;24(9): 2430-2434. 6. Sureda A, Constans M, Iriondo A, et al. Prognostic factors affecting long-term outcome after stem cell transplantation in Hodgkin's lymphoma autografted after a first relapse. Ann Oncol. 2005;16(4):625-633. 7. Sirohi B, Cunningham D, Powles R, et al. Long-term outcome of autologous stem-cell transplantation in relapsed or refractory Hodgkin's lymphoma. Ann Oncol. 2008; 19(7):1312-1319. 8. Smith EP, Li H, Friedberg JW, et al. SWOG S0410/BMT CTN 0703: A phase II trial of tandem autologous stem cell transplantation (AHCT) for patients with primary progressive or recurrent Hodgkin lymphoma (HL) (ClinicalTrials.gov Identifier: NCT00233987). Blood. 2014;124(21). Abstract 676. 9. Stiff PJ, Unger JM, Forman SJ, et al. The value of augmented preparative regimens combined with an autologous bone marrow transplant for the management of relapsed or refractory Hodgkin disease: a Southwest Oncology Group phase II trial. Biol Blood Marrow Transplant. 2003;9(8):529-539. 10. Ng AK. Review of the cardiac long-term effects of therapy for Hodgkin lymphoma. Br J Haematol. 2011;154(1):23-31. 11. Rugbjerg K, Mellemkjaer L, Boice JD, Kober L, Ewertz M, Olsen JH. Cardiovascular disease in survivors of adolescent and young adult cancer: a Danish cohort study, 19432009. J Natl Cancer Inst. 2014;106(6):dju110. 12. Hodgson DC, Gilbert ES, Dores GM, et al.

haematologica | 2016; 101(4)

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

Long-term solid cancer risk among 5-year survivors of Hodgkin's lymphoma. J Clin Oncol. 2007;25(12):1489-1497. Elkin EB, Klem ML, Gonzales AM, et al. Characteristics and outcomes of breast cancer in women with and without a history of radiation for Hodgkin's lymphoma: a multiinstitutional, matched cohort study. J Clin Oncol. 2011;29(18):2466-2473. Morton LM, Dores GM, Curtis RE, et al. Stomach cancer risk after treatment for Hodgkin lymphoma. J Clin Oncol. 2013;31 (27):3369-3377. Koontz MZ, Horning SJ, Balise R, et al. Risk of therapy-related secondary leukemia in Hodgkin lymphoma: the Stanford University experience over three generations of clinical trials. J Clin Oncol. 2013;31 (5):592-598. Eichenauer DA, Thielen I, Haverkamp H, et al. Therapy-related acute myeloid leukemia and myelodysplastic syndromes in patients with Hodgkin lymphoma: a report from the German Hodgkin Study Group. Blood. 2014;123(11):1658-1664. Goodman KA, Riedel E, Serrano V, Gulati S, Moskowitz CH, Yahalom J. Long-term effects of high-dose chemotherapy and radiation for relapsed and refractory Hodgkin's lymphoma. J Clin Oncol. 2008;26(32):52405247. Forrest DL, Hogge DE, Nevill TJ, et al. High-dose therapy and autologous hematopoietic stem-cell transplantation does not increase the risk of second neoplasms for patients with Hodgkin's lymphoma: a comparison of conventional therapy alone versus conventional therapy followed by autologous hematopoietic stemcell transplantation. J Clin Oncol. 2005;23 (31):7994-8002. Minn AY, Riedel E, Halpern J, et al. Longterm outcomes after high dose therapy and autologous haematopoietic cell rescue for refractory/relapsed Hodgkin lymphoma. Br J Haematol. 2012;159(3):329-339. Devillier R, Coso D, Castagna L, et al. Positron emission tomography response at the time of autologous stem cell transplantation predicts outcome of patients with relapsed and/or refractory Hodgkin's lymphoma responding to prior salvage therapy. Haematologica. 2012;97(7):1073-1079. Moskowitz CH, Matasar MJ, Zelenetz AD, et al. Normalization of pre-ASCT, FDG-PET imaging with second-line, non-cross-resistant, chemotherapy programs improves event-free survival in patients with Hodgkin lymphoma. Blood. 2012;119(7):1665-1670. Thomson KJ, Kayani I, Ardeshna K, et al. A response-adjusted PET-based transplantation strategy in primary resistant and relapsed Hodgkin Lymphoma. Leukemia. 2013;27(6):1419-1422. Josting A, Muller H, Borchmann P, et al. Dose intensity of chemotherapy in patients with relapsed Hodgkin's lymphoma. J Clin Oncol. 2010;28(34):5074-5080. Biswas T, Culakova E, Friedberg JW, et al.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

Involved field radiation therapy following high dose chemotherapy and autologous stem cell transplant benefits local control and survival in refractory or recurrent Hodgkin lymphoma. Radiother Oncol. 2012;103(3):367-372. Castagna L, Crocchiolo R, Giordano L, et al. High-dose melphalan with autologous stem cell support in refractory Hodgkin lymphoma patients as a bridge to second transplant. Bone Marrow Transplant. 2015;50(4): 499-504. Moskowitz CH, Nademanee A, Masszi T, et al. Brentuximab vedotin as consolidation therapy after autologous stem-cell transplantation in patients with Hodgkin's lymphoma at risk of relapse or progression (AETHERA): a randomised, double-blind, placebo-controlled, phase 3 trial. Lancet. 2015; 385(9980):1853-1862. Moskowitz CH, Paszkiewicz-Kozik E, Nadamanee A, et al. Analysis of primaryrefractory Hodgkin lymphoma pts in a randomized, placebo-controlled study of brentuximab vedotin consolidation after autologous stem cell transplant [ICML abstract 120]. Hematol Oncol. 2015;33(suppl 1):165. Younes A, Oki Y, Bociek RG, et al. Mocetinostat for relapsed classical Hodgkin's lymphoma: an open-label, singlearm, phase 2 trial. Lancet Oncol. 2011;12 (13):1222-1228. Younes A, Sureda A, Ben-Yehuda D, et al. Panobinostat in patients with relapsed/refractory Hodgkin's lymphoma after autologous stem-cell transplantation: results of a phase II study. J Clin Oncol. 2012;30(18):2197-2203. Johnston PB, Inwards DJ, Colgan JP, et al. A phase II trial of the oral mTOR inhibitor everolimus in relapsed Hodgkin lymphoma. Am J Hematol. 2010;85(5):320-324. Moskowitz AJ, Hamlin PA, Jr., Perales MA, et al. Phase II study of bendamustine in relapsed and refractory Hodgkin lymphoma. J Clin Oncol. 2013;31(4):456-460. Ansell SM, Lesokhin AM, Borrello I, et al. PD-1 blockade with nivolumab in relapsed or refractory Hodgkin's lymphoma. N Engl J Med. 2015;372(4):311-319. Moskowitz CH, Ribrag V, Michot JM, et al. PD-1 blockade with the monoclonal antibody pembrolizumab (MK-3475) in patients with classical Hodgkin lymphoma after brentuximab vedotin failure: preliminary results from a phase 1b study [ASH abstract 290]. Blood. 2014;124(21):290. Sureda A, Canals C, Arranz R, et al. Allogeneic stem cell transplantation after reduced intensity conditioning in patients with relapsed or refractory Hodgkin's lymphoma. Results of the HDR-ALLO study a prospective clinical trial by the Grupo Espanol de Linfomas/Trasplante de Medula Osea (GEL/TAMO) and the Lymphoma Working Party of the European Group for Blood and Marrow Transplantation. Haematologica. 2012;97 (2):310-317.

481


ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Cell Therapy & Immunotherapy

Ferrata Storti Foundation

Expression of a dominant T-cell receptor can reduce toxicity and enhance tumor protection of allogeneic T-cell therapy

Angelika Holler,1 Mathias Zech,1 Sara Ghorashian,1 Rebecca Pike,1 Alastair Hotblack,1 Pedro Veliça,1 Shao-An Xue,1 Ronjon Chakraverty,1,2 Emma C. Morris,1 and Hans J. Stauss1

Haematologica 2016 Volume 101(4):482-490

Institute of Immunity and Transplantation, UCL Division of Infection and Immunity, University College London, Royal Free Hospital London; and 2Department of Haematology, Cancer Institute, University College London, UK

1

ABSTRACT

D

Correspondence: h.stauss@ucl.ac.uk

Received: July 5, 2015. Accepted: January 13, 2016. Pre-published: January 22, 2016. doi:10.3324/haematol.2015.132712

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/4/482

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

482

ue to the lack of specificity for tumor antigens, allogeneic T-cell therapy is associated with graft-versus-host disease. Enhancing the anti-tumor specificity while reducing the graft-versus-host disease risk of allogeneic T cells has remained a research focus. In this study, we demonstrate that the introduction of ‘dominant’ T-cell receptors into primary murine T cells can suppress the expression of endogenous T-cell receptors in a large proportion of the gene-modified T cells. Adoptive transfer of allogeneic T cells expressing a ‘dominant’ T-cell receptor significantly reduced the graft-versus-host toxicity in recipient mice. Using two bone marrow transplant models, enhanced anti-tumor activity was observed in the presence of reduced graft-versus-host disease. However, although transfer of T-cell receptor gene-modified allogeneic T cells resulted in the elimination of antigen-positive tumor cells and improved the survival of treated mice, it was associated with accumulation of T cells expressing endogenous T-cell receptors and the development of delayed graft-versus-host disease. The in vivo deletion of the engineered T cells, mediated by endogenous mouse mammary tumor virus MTV8 and MTV9, abolished graft-versus-host disease while retaining significant anti-tumor activity of adoptively transferred T cells. Together, this study shows that the in vitro selection of allogeneic T cells expressing high levels of a ‘dominant’ T-cell receptor can lower acute graft-versus-host disease and enhance anti-tumor activity of adoptive cell therapy, while the in vivo outgrowth of T cells expressing endogenous T-cell receptors remains a risk factor for the delayed onset of graft-versus-host disease.

Introduction TCR gene transfer is an attractive strategy to direct T-cell specificity towards defined target antigens.1-6 This approach can readily produce large numbers of antigen-specific T cells for adoptive cell therapy and no longer requires the selection and expansion of rare specificities present in the autologous TCR repertoire. Recent clinical trials have shown the feasibility and efficacy of TCR gene therapy in melanoma and synovial cell carcinoma.7 However, the safety concerns of TCR gene therapy include the mis-pairing of introduced TCR chains with the endogenous chains expressed in the gene-modified T cells. Recent murine model experiments have indicated that TCR mis-pairing can generate auto-reactive T cells that cause toxicity when adoptively transferred into syngeneic hosts.8 It is likely that similar toxicities can occur in humans, although it has not yet been described in clinical trials to date. One of the most effective forms of T-cell therapy in humans is the adoptive transfer of allogeneic donor T cells into patients transplanted with hematopoietic stem haematologica | 2016; 101(4)


TCR transfer to suppress allo-reactivity

cells derived from the same donor, typically as a T-cell replete allogeneic hematopoietic stem cell transplantation (HSCT) or post allo-HSCT donor lymphocyte infusion (DLI).9 In this scenario, allo-reactive donor T cells can efficiently attack and eliminate patient leukemia cells, resulting in cure of the malignant disease. However, since the major and minor histocompatibility antigens recognized by allo-reactive T cells are not selectively expressed in leukemia cells, but also in normal tissues, this can lead to normal tissue damage and result in graft-versus-host disease (GvHD).10 Thus, enhancing the leukemia specificity while reducing the ‘damaging’ allo-reactive potential of donor T-cell therapy has remained the focus of research. A recent study has demonstrated in a fully MHC-mismatched murine transplant model that TCR gene transfer into donor T cells was able to enhance graft-versusleukemia (GvL) and reduce GvHD. In this study, Koestner et al. transferred the OT1-TCR, specific for a peptide epitope of ovalbumin, into donor T cells and achieved 20%-70% reduction of the endogenous TCR repertoire.11 Here, we explored whether ‘dominant’ TCR can mediate complete inhibition of endogenous TCR expression in donor T cells in vivo. Together with other groups, we had previously shown that ‘dominant’ TCR can suppress the expression of ‘non-dominant’ TCR on the surface of genemodified T cells in vitro.12-14 The mechanism of TCR dominance has not yet been fully defined, but it may involve transcriptional and translational regulation as well as the efficiency of the TCR α and b chain pairing and assembly with CD3 chains prior to the migration of the TCR complex from the endoplasmic reticulum to the cell surface. The observation that modifications improving RNA translation into protein and modifications improving TCR α/b chain pairing both enhanced TCR dominance indicated that efficient protein synthesis and efficient formation of α/b heterodimers have a direct impact on TCR dominance.12 Here, we have codon optimized the TCR gene sequences and introduced additional disulphide bonds in the constant domains to generate ‘dominant’ TCR. We tested the ability for these ‘dominant’ TCR constructs to suppress the surface expression of endogenous TCR in polyclonal primary T cells in vitro. We then performed in vivo experiments to test the working hypothesis that the introduction of a ‘dominant’ TCR into allogeneic donor T cells may control graft-versus-host toxicity, while enhancing protection against tumors expressing the target antigen recognized by the ‘dominant’ TCR.

pentamer consisted of pNP366 bound to H2-Db. The influenza A virus NP-derived peptide, pNP366 (ASNENMDAM) is presented by H2-Db MHC Class-I molecules and was obtained from Proimmune.

Vectors The retroviral F5-TCR vector pMP71-TCRα-2A-TCRb was modified from pMX-TCRα-IRES-TCRb (a kind gift from Prof T. Schumacher). The F5-TCR recognizes the influenza A virus NP (NP366-374) peptide in the context of murine H2-Db MHC Class-I. For tumor challenge experiments, we either used the F5-TCR vector or a version of it encoding for a truncated CD19 downstream of an IRES sequence as a marker of transduction. CD19 staining was used to track adoptively transferred DBA/J1 donor T cells, which lacked a congenic marker. The retroviral OTII-TCR vector pMP71TCRα-2A-TCRb was generated in the lab. The OTII-TCR recognizes the OVA (OVA323-339-ISQAVHAAHAEINEAGR) peptide in the context of murine H2-Ab MHC Class-II. The control vector pMP71-iCre-IRES-GFP (inverted Cre-recombinase) was generated in the lab. The retroviral pMP71 vectors containing the F5- or OTII-TCR were modified for optimal gene (and surface TCR) expression by codon optimization and addition of an engineered cysteine bond between the TCR chains (Geneart).

Bone marrow transplantation, T-cell transfer and tumor challenge Recipient mice (Thy1.2) were lethally irradiated (C57BL/6, 11 Gy; BALB/c, 8 Gy X-ray irradiation split into 2 fractions, separated by 24 h) and T-cell depleted bone marrow cells were injected intravenously 4 h later. Vb11 or NP-pentamer sorted donor T cells (Thy1.1) were adoptively transferred via the tail vein the following day. Mock transduced T cells were used as control. On the same day, mice received i.p. injections of C57BL/6 bone marrow derived DCs pulsed with NP peptide. Five doses of 105U IL2 (Chiron) i.p. were administered to recipient mice, with the first dose given on the day of T-cell transfer and subsequently twice daily on the following two days. For tumor challenge experiments, C57BL/6 recipient mice (Thy1.2) were conditioned as described above, but with the addition of subcutaneous inoculation of 106 EL4-NP cells on the day of bone marrow transplantation. NP-pentamer sorted donor T cells, either from DBA/J1 (Figure 4) or BALB/c (Figure 5) origin were transduced with the F5-TCR and adoptively transferred via the tail vein the following day. GFP sorted or mock transduced T cells were used as a control. Tumors were measured with a calliper in two different dimensions (a and b) at different intervals and the growth evaluated applying the following formula: axbx /4.

Methods Mice Female C57BL/6, C57BL/6 (Thy1.1), BALB/c, BALB/c (Thy1.1) and DBA/J1 mice were purchased from Charles Rivers Laboratories or bred in house by UCL Biological Services. All experiments were conducted in accordance with United Kingdom Home Office regulations.

Cell lines, pentamer and peptide Phoenix-Ecotropic (PhEco) adherent packaging cells (Nolan Laboratory) were transiently transfected with retroviral vectors for the generation of supernatant containing the recombinant retrovirus required for infection of target cells. EL4 is a murine lymphoma cell line, EL4-NP a variant, stably expressing the influenza A virus nucleoprotein (NP) (a kind gift from Dr B. Stockinger). The haematologica | 2016; 101(4)

Results Dominant TCR can suppress expression of endogenous TCR In this study, we have used an MHC Class-I restricted TCR (F5-TCR) specific for a peptide epitope of the influenza virus nucleoprotein presented by H2-Db and an MHC Class-II restricted TCR (OTII-TCR) specific for an ovalbumin-derived peptide presented by H2-Ab. Both TCR constructs were codon optimized and contained an additional disulphide bond in the constant domain to improve RNA translation and α/b chain pairing. The mod483


A. Holler et al.

ified F5- and OTII-TCR genes were inserted into the retroviral pMP71 vector for gene transfer into primary murine T cells. In order to test the ability of the two TCR constructs to suppress the cell surface expression of the endogenous TCR chains, we used murine splenocytes and purified the T cells expressing Vb8.1, 8.2 and 8.3 TCR, which represented approximately 16% of the total T cells. This allowed us to use antibodies specific for Vb8.1,2,3 to measure the expression of endogenous TCR, and antibodies specific for the Vb11 and Vb5 chains to assess expression of the introduced F5-TCR and OTII-TCR, respectively. Figure 1 shows the staining profile of purified Vb8.1,2,3 T cells that were mock transduced, or transduced with the retroviral constructs encoding the F5-TCR or the OTIITCR. The majority of freshly transduced T cells expressed high levels of the introduced Vb11 or Vb5 TCR chains and sharply reduced levels of the endogenous Vb8.1,2,3 chains. Approximately 30% of the T cells expressed both the introduced as well as the endogenous TCR chains. Less than 10% of T cells expressed the endogenous TCR only, which most likely represents untransduced T cells as the efficiency of retroviral TCR gene transfer does not usually reach 100%. We observed an inverse correlation between the level of expression of the introduced TCR and the expression levels of the endogenous TCR. For example, the mean fluorescent intensity (MFI) of Vb11 or Vb5 expression in the ‘single-positive’ T cells expressing primarily the introduced TCR was higher than the Vb11 or Vb5 MFI in the ‘double-positive’ T cells expressing both introduced as well as endogenous TCR. In the case of the F5-TCR, the MFI for the introduced b chain in the ‘single’ versus ‘double-positive’ T cells was 16300 versus 11300, and for the OTII-TCR the difference was 10300 versus 7815. Together, these experiments indicated that ‘domi-

nant’ TCR can suppress the cell surface expression of endogenous TCR, and that the suppression was most effective in T cells expressing high levels of the ‘dominant’ TCR.

Reduced toxicity of allo-reactive T cells expressing ‘dominant’ TCR We used an MHC-mismatched bone marrow transplantation model to test whether the introduction of the ‘dominant’ F5-TCR into C57BL/6 T cells reduced the toxicity of donor T-cell infusion into BALB/c mice transplanted with C57BL/6 hematopoietic stem cells. BALB/c mice received myeloablative irradiation and were then transplanted with T-cell depleted C57BL/6 (Thy1.2) bone marrow, followed by adoptive transfer of polyclonal donor T cells (Thy1.1) that were mock transduced or transduced to express the F5-TCR (Figure 2A). A short course of high-dose IL2 was given to promote in vivo T-cell expansion. Following transduction of the donor T cells, we used flow cytometry sorting to purify cells expressing high levels of the introduced Vb11 chain (Figure 2B). These T cells were adoptively transferred into the transplanted mice, which were then monitored using weight loss and survival as two objectively measurable criteria of potential toxicity caused by the infused T cells. After three weeks the animals were sacrificed to analyze the number and the phenotype of the engrafted T cells. Figure 2C shows that the measured weight of mice treated with F5-TCR transduced T cells was similar to the weight of control mice receiving a bone marrow transplant, but no T cells. In contrast, mice treated with mock transduced donor T cells showed substantial weight loss which was most pronounced in the first week after T-cell transfer (Figure 2C). Comparison of the weight score at day 8 showed that

Figure 1. TCR transfer suppresses expression of endogenous TCR chains. BALB/c splenocytes were Vβ8 sorted (representing one endogenous TCR Vb chain family) followed by mock transduction or transduction with the F5- or OTII-TCR and then stained with antibodies against CD3, Vb8 (endogenous TCR), Vb11 (F5-TCR) or Vb5 (OTIITCR) to assess the expression levels of endogenous (endo) and introduced (intro) TCR. Plots show live-gated CD3+ T cells.

484

haematologica | 2016; 101(4)


TCR transfer to suppress allo-reactivity

A

B

C

D

E

F

G

H

Figure 2. Allogeneic T cells expressing introduced TCR display reduced toxicity in MHC-mismatched bone marrow transplantation (BMT) recipients. (A) Allogeneic chimeras were generated by lethal irradiation of recipient BALB/c mice (Thy1.2) followed by reconstitution with T-cell depleted (TCD) C57BL/6 bone marrow (BM). Mice received 3x106 C57BL/6 T cells (Thy1.1) transduced with the F5-TCR sorted to express high levels of the introduced Vb11 chain. Control animals received 3x106 mock transduced C57BL/6 T cells (Thy1.1) or no T cells. To stimulate antigen driven expansion of F5-TCR donor T cells all mice were injected intraperitoneally with 0.5x106 C57BL/6 DCs loaded with NP peptide. 105U of IL2 were administered on the day of T-cell transfer and twice daily for the following two days. (B) Staining profile of F5-TCR transduced T cells before and after Vb11 sorting. The indicated P2 gate representing the top 33% of Vb11-expressing T cells was used for FACS sorting. Purity after sorting is shown. Plots show live-gated Thy1.1+ lymphocytes. (C) Kinetics of weight loss observed in the 3 groups of mice [(mock and F5-TCR (Vb11sorted) T cells: n=13; no T cells: n=12)]. Faint black lines represent individual mice and the bold black line represents the average of weight loss per group. Pooled results of 3 independent experiments are shown. (D) Weight loss on day 8 and day 21 (end point). Symbols represent individual mice, bars show group averages. P value for mock versus F5-TCR (Vb11sorted) transduced T cells; P=0.0003 (weight loss on day 8: n=13 mice per group) and P=0.0057 [(weight loss on day 21, mock transduced T cells: n=8; F5-TCR (Vb11sorted) T cells: n=13)]. Pooled results of 3 independent experiments are shown. (E) Kaplan-Meier survival plot for mice receiving mock or F5-TCR (Vb11sorted) transduced T cells. P=0.015; n=13 mice per group. Pooled results of 3 independent experiments are shown. (F) Absolute number of transferred Thy1.1 donor T cells in the spleens of recipient mice. Symbols represent individual mice, bars show group averages (n=5 mice per group). One representative experiment of 2 is shown. (G) Ex vivo phenotypical analysis of spleens. Splenocytes from mice that had received mock or F5-TCR (Vb11sorted) transduced T cells were stained with antibodies against Thy1.1, CD4, CD8, Vb11 and NP-pentamer. Plots show live-gated Thy1.1+ T cells of one representative mouse per group. (H) Summary of percentages of F5-TCR (Vb11sorted) donor T cells binding Vb11 antibody and NP-pentamer of all mice (n=3).

haematologica | 2016; 101(4)

485


A. Holler et al.

mice treated with mock transduced donor T cells lost significantly more weight than mice treated with F5-TCR transduced T cells (Figure 2D). In fact, 5 out of 13 mice succumbed to severe weight loss or had to be sacrificed in accordance with UK Home Office regulations (Figure 2E). The scoring of the surviving mice at week 3 after T-cell transfer showed that the group treated with mock transduced T cells continued to weigh significantly less than the F5-TCR group (Figure 2D). Although these data indicate that the F5-TCR prevented GvHD-associated weight loss in the first three weeks, our experiments did not test whether F5-TCR transduced T cells might cause long-term pathology such as chronic GvHD. At week 3, all mice were sacrificed to analyze the level of T-cell engraftment and TCR expression. The absolute number of engrafted donor Thy1.1 T cell in the spleen of recipient mice was similar for mock and F5-TCR transduced T cells (Figure 2F). Flow cytometric analysis showed that the majority of F5-TCR transduced T cells continued to express high levels of the introduced Vb11 chain. Surprisingly, staining with H2-Db/NP-peptide pentamer reagents showed that

A

the majority of the Vb11+ donor T cells were unable to bind the pentamer, indicating that they did not express the correctly paired F5-TCR Îą/b heterodimer, although they expressed high levels of the F5-TCR b chain (Figure 2G and H).

Pentamer-negative donor T cells expand preferentially in vivo We explored whether pentamer-negative T cells existed before injection, or whether they arose in vivo from T cells expressing F5-TCR Îą/b heterodimer. Most of the freshly transduced T cells expressing high levels of the Vb11 chain did bind the pentamer, although a minority of T cells were pentamer-negative (Figure 3A). If the pentamer-negative, potentially allo-reactive T cells expanded preferentially after in vivo transfer, then exclusion of these cells before adoptive transfer should avoid the emergence of pentamer-negative T cells. We therefore FACS purified transduced donor T cells to obtain a highly pure Vb11+/pentamer+ population (Figure 3B). These C57BL/6 donor T cells (Thy1.1) were then adoptively transferred into

B

C

Figure 3. Pentamer sorted T cells retain NP-pentamer binding in MHC-mismatched bone marrow transplantation (BMT) recipients. (A) The pre-injection Vb11+ T-cell population contains a small number of NP-pentamer negative cells. C57BL/6 splenocytes (depleted of the endogenous Vb11 T-cell population) were transduced with the F5-TCR and stained with antibodies against CD8 and Vb11 (left panel). The right panel shows the NP-pentamer binding profile of the gated Vb11+ T-cell population. (B) FACS sorting of NP-pentamer binding T cells. C57BL/6 T cells were transduced with the F5-TCR and stained with antibodies against Vb11 and NP-pentamer. The P3 gate was used for FACS sorting. Purity after sorting is shown in the right panel. Plots show live-gated Thy1.1+ lymphocytes. (C) Ex vivo phenotypical analysis of C57BL/6 donor T cells (Thy1.1) after adoptive transfer into BALB/c recipient mice previously transplanted with T-cell depleted C57BL/6 bone marrow. Donor T cells were transduced with the F5-TCR and purified as shown in (B); mock transduced T cells were used as a control. Three weeks after T-cell transfer, mice were sacrificed and splenocytes were stained with antibodies against Thy1.1, CD4, CD8, Vb11 and NP-pentamer. Plots show Vb11 staining (left panels) and NP-pentamer staining (right panels) of live-gated Thy1.1+ donor T cells of mice that received mock transduced T cells or F5-TCR (NP-pentamersorted) T cells.

486

haematologica | 2016; 101(4)


TCR transfer to suppress allo-reactivity

A

B

C

D

E

F

G

Figure 4. TCR transfer enhanced the anti-tumor effects of allogeneic T-cell therapy. (A) Allogeneic chimeras were generated by lethal irradiation of C57BL/6 mice transplanted with allogeneic T-cell depleted bone marrow followed by EL4-NP tumor challenge and allogeneic T-cell therapy. The allogeneic bone marrow and T cells were either of DBA/J1 origin (see Figure 4) or BALB/c origin (see Figure 5). (B) Tumor-bearing mice were treated with 1x106 F5-TCR-CD19 (NP-pentamersorted) bulk T cells or purified CD8+ T cells from DBA/J1 donors. Control mice received no T cells or 1x106 GFP transduced and FACS sorted T cells from DBA/J1 donors. Tumor growth observed in the 4 groups of mice is shown (n=5, except for the CD8+ group n=6). P values on day 11 post T-cell transfer are: GFP control T cells versus bulk F5-TCR-CD19 T cells is non-significant (ns); bone marrow transplantation (BMT) control versus bulk F5-TCR-CD19 T cells (P=0.02) and GFP control T cells versus CD8+ F5-TCR-CD19 T cells (P=0.0002). (C) EL4-NP tumors from mice treated with F5-TCR-CD19 CD8+ T cells were re-isolated at the time point when mice were sacrificed due to GvHD toxicity. They were cultured for four days and then mixed with F5-TCR transduced T cells to analyze IFNÎł responses. Parental EL4-NP cells were used as a positive control. FACS plots show percentage IFNg production of transduced CD8+ T cells. (D) Kaplan-Meier survival plot for mice receiving GFP control T cells, F5TCR-CD19 bulk T cells or no T cells. (E) Clinical graft-versus-host disease (GvHD) severity score of mice treated with F5-TCR-CD19 CD8+ T cells. Mice were weighed and assessed for signs of clinical GvHD 2-3 times a week (see Methods). GvHD score is shown. (F) Ex vivo phenotypical analysis of mice treated with GFP control T cells or F5-TCR-CD19 CD8+ T cells. Splenocytes were stained with antibodies against CD19, CD4, CD8, and NP-pentamer. Plots show the level of pentamer binding of live-gated GFP+ T cells (left) and live-gated CD19+ T cells (right). Combined data of all analyzed mice are shown (G). Data of one representative mouse per group are shown or combined data of all analyzed mice (F5-TCR n=6; GFP T cells n=1).

haematologica | 2016; 101(4)

487


A. Holler et al.

BALB/c recipients previously transplanted with C57BL/6 bone marrow cells (Thy1.2). As before, recipient mice were sacrificed three weeks after T-cell transfer to determine TCR expression. The majority of Thy1.1 donor T cells expressed high levels of Vb11 and bound H2-Db/NPpentamer (Figure 3C). However, some of the re-isolated T cells did not bind pentamer ex vivo (Figure 3C). It is difficult to know whether these cells derived from the small number (<1%) of pentamer-negative cells present prior to transfer (Figure 3B) or whether they lost pentamer binding in vivo.

TCR-transfer improves tumor protection of allogeneic donor T-cell therapy We next explored whether transfer of the â&#x20AC;&#x2DC;dominantâ&#x20AC;&#x2122; F5TCR enhanced the ability of donor T cells to control tumor growth. To address this question, we challenged C57BL/6 mice with syngeneic EL4 lymphoma cells expressing the TCR-recognized NP antigen. Prior to tumor challenge, the C57BL/6 mice (H2b) were transplanted with T-cell depleted bone marrow from MHC-mismatched BALB/c (H2d) or DBA/J1 (H2q) donor mice (Figure 4A). While H2d mice express MHC class-II IE molecules, this is not the case in H2b and H2q mice which lack the gene encoding the alpha chain of the IE alpha/beta heterodimer. Lack of IE molecules results in lack of the presentation of endogenous super-antigens that are encoded by mouse mammary tumor virus MTV8 and MTV9 present in the genome of most inbred mouse strains.15 The presentation of MTV8 and MTV9 by IE leads to the selective reduction

of T cells expressing the Vb11 chain in the natural repertoire of BALB/c mice.16 Since the F5 TCR uses Vb11, this is expected to reduce T-cell engraftment in C57BL/6 mice reconstituted with IE-positive BALB/c bone marrow, while allowing efficient engraftment and persistence in mice reconstituted with IE-negative DBA/J1 bone marrow. We first investigated allogeneic T-cell therapy of EL4-NP tumor in C57BL/6 mice transplanted with DBA/J1 bone marrow. DBA/J1 donor T cells were transduced with an F5-TCR construct containing a CD19 marker (see Methods) to identify transduced T cells in vivo. As control, donor T cells were transduced with GFP. We used flow cytometry to sort donor T cells binding high levels of NP-pentamer (F5-TCR) or for GFP expression (control) prior to adoptive transfer. Compared to mice that did not receive any T cells, the transfer of allogeneic GFP transduced T cells did not reduce tumor growth (Figure 4B). In contrast, treatment with allogeneic bulk T cells transduced with the F5-TCR resulted in substantial reduction in tumor burden. Transfer of purified TCR transduced CD8+ T cells was more effective than transfer of bulk T cells. This is possibly due to our previous observation that CD8+ T cells expressing this TCR recognized approximately 10fold lower concentration of antigen than CD4+ T cells expressing the same TCR.17 However, despite the near elimination of tumors after treatment with DBA/J1 CD8+ T cells, protection was incomplete and tumors relapsed in most animals. We explored whether this was due to immune-editing and tumor escape. The analysis of tumors re-isolated from the mice treated with TCR transduced

A

B

C

Figure 5. Depletion of TCR transduced T cells reduces toxicity and tumor protection. In these experiments, C57BL/6 mice were transplanted with BALB/c bone marrow and treated with TCR transduced BALB/c donor T cells (see Figure 4A). (A) EL4-NP tumor growth in mice receiving no T cells (n=3) or treated with mock (n=5) or F5-TCR transduced bulk T cells (n=7). One representative experiment of 2 is shown. (B) Kaplan-Meier survival plot for mice receiving mock T cells (n=10), F5-TCR T cells (n=11) or no T cells (n=8). Pooled data from 2 independent experiments are shown. (C) Absolute numbers of transferred mock or F5-TCR transduced T cells in the spleen of treated mice, showing selective depletion of Vb11+ F5-TCR T cells.

488

haematologica | 2016; 101(4)


TCR transfer to suppress allo-reactivity

CD8+ T cells revealed that they no longer stimulated antigen-specific cytokine production by responder T cells expressing the F5-TCR (Figure 4C). This indicates that Tcell therapy was effective, and that immune-edited escape variants were responsible for tumor relapse in treated mice. The impressive anti-tumor activity of F5-TCR transduced DBA/J1 T cells was also associated with improved survival of F5-TCR treated mice. Mice that did not receive any T cells or were treated with GFP control T cells died between day 13-19 with progressive tumor growth (Figure 4D). In contrast, mice treated with F5-TCR transduced CD8+ T cells did not die because of tumor growth, but instead had to be sacrificed at day 26 because of development of GvHD as assessed by clinical and histological GvHD severity scoring (Figure 4E and Online Supplementary Figure S1). When mice were sacrificed, we used flow cytometry analysis of splenocytes to detect the transferred DBA/J1 T cells by their expression of GFP (control T cells) or CD19 (TCR transduced T cells). Both, GFP+ and CD19+ T cells were readily detectable in the spleen of treated mice. As expected, the gated GFP+ control T cells did not bind the NP-pentamer, while most of the gated CD19+ T cells were NP-pentamer-positive (Figure 4F). However, the summary data of all mice show that 15%30% of the CD19+ T cells had lost pentamer staining, suggesting that these T cells no longer expressed high levels of the F5-TCR, but instead potentially allo-reactive endogenous or mis-paired TCR that could account for the observed GvHD toxicity (Figure 4G). Finally, we explored whether the toxicity that was observed approximately three weeks after T-cell transfer could be reduced by limiting the in vivo persistence of the F5-TCR transduced T cells. Therefore, we performed tumor protection experiments in C57BL/6 mice reconstituted with BALB/c bone marrow. In these mice, treatment with F5-TCR transduced BALB/c T cells delayed tumor growth when compared to treatment with mock transduced T cells (Figure 5A). The delay in tumor growth was associated with improved survival of mice treated with F5-TCR transduced BALB/c T cells (Figure 5B). In this setting, the mice did not develop any signs of GvHD toxicity but succumbed to progressive tumor growth. When mice treated with control or F5-TCR T cells reached lethal tumor burden, they were sacrificed and analyzed to determine the number of engrafted BALB/c T cells. As expected, control transduced T cells not bearing the F5-TCR were readily detectable in the spleen of treated mice, while the Vb11-positive T cells expressing the F5-TCR were selectively depleted (Figure 5C). The data indicated that the depletion of the F5-TCR expressing T cells prevented any detectable GvHD toxicity, but did not abolish an improvement in tumor protection compared to mock transduced T cells.

Discussion Strategies to reduce graft-versus-host disease and enhance anti-tumor activity of allogeneic T cells remain an important area of research.18 Various approaches have been developed, including the transfer of purified memory T cells19,20 or the in vitro depletion of allo-reactive T cells stimulated with recipient antigen-presenting cells.21-23 Similarly, in vitro stimulation of allo-reactive T cells in the haematologica | 2016; 101(4)

presence of co-stimulation blockade was used to induce T-cell anergy.24 An elegant alternative to the in vitro depletion of allo-reactive T cells is the introduction of ‘suicide genes’ that can be triggered in vivo in patients developing GvHD after T-cell transfer.25-28 Most recently, the disruption of endogenous TCR genes has been achieved using zinc finger nucleases, generating T cells unable to respond to allo-stimulation in vitro.29,30 The introduction of TCR or single chain antibody constructs has been used to equip these TCR-disrupted T cells with new specificities. In this study, we explored a simplified strategy to simultaneously suppress endogenous TCR expression while introducing therapeutic TCR. The described approach is based on the working hypothesis that ‘dominant’ TCR can redirect T-cell specificity towards tumor antigens and also suppress the expression of endogenous TCR that are required for donor T cells to recognize major and minor histocompatibility antigens involved in GvHD. We have taken advantage of previous reports showing that codon optimization and introduction of additional cysteine bonds can improve TCR expression and reduce unwanted mis-pairing.31-33 Using modified F5- and OTIITCR constructs, we found that both were effective in suppressing the expression of endogenous TCR in the majority of transduced primary T cells. We found that in vitro selection of transduced T cells for high-level expression of introduced TCR b chain (using antibodies) or for the correctly paired TCR α/b combination (using pentamer) was required to reduce expression of potentially allo-reactive endogenous TCR. Following in vivo transfer, we observed the selective expansion of T cells that lost pentamer binding, indicating that they expressed endogenous or mis-paired TCR that might be driven by alloantigen recognition in vivo. While pentamer selected T cells remained largely pentamer-positive when they were transferred into tumor-free mice, the transfer into tumor-bearing recipients resulted in the accumulation of pentamer-negative T cells. A major difference between tumor-free and tumor-bearing mice is the absence and presence, respectively, of the F5-TCR recognized target antigen. It is possible that the stimulation via the F5-TCR lowers the triggering threshold for endogenous or mispaired TCR that can recognize allo-antigens. Once triggered, these allo-reactive T cells may preferentially expand in the MHC-mismatched recipient mice. The results described here are similar to the results obtained by Koestner et al., when they showed that allogeneic T cells engineered to express the OT1-TCR displayed increased GvL and reduced GvHD activity.11 In these studies, the engineered T cells contained some untransduced cells and the OT1-TCR achieved only 20%-70% reduction in the level of endogenous TCR expression. In our case, we purified T cells for high expression of the introduced F5-TCR, and this ‘dominant’ TCR was capable of suppressing endogenous TCR expression to undetectable levels. Nevertheless, our results show that, despite the use of purified T-cell populations, the in vivo exposure to allo-antigens provided a strong selection for reduced F5- and increased endogenous TCR expression. When the F5-TCR transduced allogeneic T cells were able to expand and persist in large numbers, they mediated effective control of EL4 tumors expressing the TCR recognized target antigen. Only immune-edited tumor variants that were no longer recognized by the TCR transduced T cells were able to grow progressively in 489


A. Holler et al.

these mice. However, persistence of the TCR transduced allogeneic T cells led to the delayed development of severe GvHD toxicity, which was associated with the enrichment of T cells that no longer expressed the F5TCR α/b combination. The GvHD toxicity was not seen when the in vivo persistence of the F5-TCR transduced T cells was limited. This was achieved by the use of recipient mice presenting the Vb11 depleting MTV8 and MTV9 antigens in the context of MHC class-II IE molecules. In this setting, the F5-TCR was still able to mediate anti-tumor activity, but the lack of T-cell persistence was associated with the lack of GvHD toxicity. These experiments indicated that ‘dominant’ TCR gene transfer can substantially enhance the anti-tumor activity of allogene-

References 13. 1. Schumacher TN. T-cell-receptor gene therapy. Nat Rev Immunol. 2002; 2(7):512-519. 2. Stauss HJ, Morris EC. Immunotherapy with gene-modified T cells: limiting side effects provides new challenges. Gene Ther. 2013;20(11):1029-1032. 3. Rosenberg SA, Restifo NP. Adoptive cell transfer as personalized immunotherapy for human cancer. Science. 2015; 348(6230):6268. 4. Schumacher TN, Schreiber RD. Neoantigens in cancer immunotherapy. Science. 2015;348(6230):62-68. 5. Stanislawski T, Voss RH, Lotz C, et al. Circumventing tolerance to a human MDM2-derived tumor antigen by TCR gene transfer. Nat Immunol. 2001; 2(10):962-970. 6. Xue SA, Gao L, Hart D, et al. Elimination of human leukemia cells in NOD/SCID mice by WT1-TCR gene-transduced human T cells. Blood. 2005;106(9):3062-3067. 7. Robbins PF, Morgan RA, Feldman SA, et al. Tumor regression in patients with metastatic synovial cell sarcoma and melanoma using genetically engineered lymphocytes reactive with NY-ESO-1. J Clin Oncol. 2011;29(7):917-924. 8. Bendle GM, Linnemann C, Hooijkaas AI, et al. Lethal graft-versus-host disease in mouse models of T cell receptor gene therapy. Nat Med. 2010;16(5):565-570. 9. Kolb HJ, Schattenberg A, Goldman JM, et al. Graft-versus-leukemia effect of donor lymphocyte transfusions in marrow grafted patients. Blood. 1995;86(5):2041-2050. 10. Horowitz MM, Gale RP, Sondel PM, et al. Graft-versus-leukemia reactions after bone marrow transplantation. Blood. 1990;75(3):555-562. 11. Koestner W, Hapke M, Herbst J, et al. PD-L1 blockade effectively restores strong graftversus-leukemia effects without graft-versus-host disease after delayed adoptive transfer of T-cell receptor gene-engineered allogeneic CD8+ T cells. Blood. 2011;117(3):1030-1041. 12. Hart DP, Xue SA, Thomas S, et al. Retroviral transfer of a dominant TCR prevents surface expression of a large proportion of the

490

14.

15.

16.

17.

18. 19.

20.

21.

22.

23.

ic T-cell therapy, but that long-term persistence of these T cells leads to the expansion of allo-reactive T cells that no longer express high levels of the ‘dominant’ TCR. Although this relatively simple strategy of using ‘dominant’ TCR can clearly improve tumor immunity, the complete removal of the risk of GvHD requires the complete depletion of endogenous TCR to avoid in vivo expansion of allo-reactive specificities.29,30 Funding The authors would like to thank Bloodwise, Medical Research Council, EU ATTRACT consortium, CRUK Experimental Cancer Medicine Centre, NIHR UCL/UCLH Biomedical Research Centre and NIHR CRN for supporting this work.

endogenous TCR repertoire in human T cells. Gene Ther. 2008;15(8):625-631. Heemskerk MH, Hagedoorn RS, van der Hoorn MA, et al. Efficiency of T-cell receptor expression in dual-specific T cells is controlled by the intrinsic qualities of the TCR chains within the TCR-CD3 complex. Blood. 2007;109(1):235-243. Sommermeyer D, Neudorfer J, Weinhold M, et al. Designer T cells by T cell receptor replacement. Eur J Immunol. 2006;36(11): 3052-3059. Kozak C, Peters G, Pauley R, et al. A standardized nomenclature for endogenous mouse mammary tumor viruses. J Virol. 1987;61(5):1651-1654. Braun MY, Jouvin-Marche E, Marche PN, MacDonald HR, Acha-Orbea H. T cell receptor V beta repertoire in mice lacking endogenous mouse mammary tumor provirus. Eur J Immunol. 1995;25(3):857-862. Morris EC, Tsallios A, Bendle GM, Xue SA, Stauss HJ. A critical role of T cell antigen receptor-transduced MHC class I-restricted helper T cells in tumor protection. Proc Natl Acad Sci USA. 2005; 102(22):7934-7939. Bleakley M, Riddell SR. Molecules and mechanisms of the graft-versus-leukaemia effect. Nat Rev Cancer. 2004; 4(5):371-380. Chen BJ, Cui X, Sempowski GD, Liu C, Chao NJ. Transfer of allogeneic CD62Lmemory T cells without graft-versus-host disease. Blood. 2004;103(4):1534-1541. Foster AE, Marangolo M, Sartor MM, et al. Human CD62L- memory T cells are less responsive to alloantigen stimulation than CD62L+ naive T cells: potential for adoptive immunotherapy and allodepletion. Blood. 2004;104(8):2403-2409. Amrolia PJ, Muccioli-Casadei G, Yvon E, et al. Selective depletion of donor alloreactive T cells without loss of antiviral or antileukemic responses. Blood. 2003; 102(6):2292-2299. Hartwig UF, Nonn M, Khan S, Meyer RG, Huber C, Herr W. Depletion of alloreactive T cells via CD69: implications on antiviral, antileukemic and immunoregulatory T lymphocytes. Bone Marrow Transplant. 2006;37(3):297-305. Wehler TC, Nonn M, Brandt B, et al. Targeting the activation-induced antigen

24.

25.

26.

27.

28. 29.

30.

31.

32.

33.

CD137 can selectively deplete alloreactive T cells from antileukemic and antitumor donor T-cell lines. Blood. 2007; 109(1):365-373. Davies JK, Singh H, Huls H, et al. Combining CD19 redirection and alloanergization to generate tumor-specific human T cells for allogeneic cell therapy of B-cell malignancies. Cancer Res. 2010;70(10): 3915-3924. Bonini C, Ferrari G, Verzeletti S, et al. HSVTK gene transfer into donor lymphocytes for control of allogeneic graft-versus-leukemia. Science. 1997; 276(5319):1719-1724. Introna M, Barbui AM, Bambacioni F, et al. Genetic modification of human T cells with CD20: a strategy to purify and lyse transduced cells with anti-CD20 antibodies. Hum Gene Ther. 2000;11(4):611-620. Sato T, Neschadim A, Konrad M, Fowler DH, Lavie A, Medin JA. Engineered human tmpk/AZT as a novel enzyme/prodrug axis for suicide gene therapy. Mol Ther. 2007;15(5):962-970. Straathof KC, Pule MA, Yotnda P, et al. An inducible caspase 9 safety switch for T-cell therapy. Blood. 2005;105(11):4247-4254. Provasi E, Genovese P, Lombardo A, et al. Editing T cell specificity towards leukemia by zinc finger nucleases and lentiviral gene transfer. Nat Med. 2012; 18(5):807-815. Torikai H, Reik A, Liu PQ, et al. A foundation for universal T-cell based immunotherapy: T cells engineered to express a CD19specific chimeric-antigen-receptor and eliminate expression of endogenous TCR. Blood. 2012;119(24):5697-5705. Cohen CJ, Li YF, El-Gamil M, Robbins PF, Rosenberg SA, Morgan RA. Enhanced antitumor activity of T cells engineered to express T-cell receptors with a second disulfide bond. Cancer Res. 2007; 67(8):38983903. Kuball J, Dossett ML, Wolfl M, et al. Facilitating matched pairing and expression of TCR chains introduced into human T cells. Blood. 2007;109(6):2331-2338. Thomas S, Xue SA, Cesco-Gaspere M, et al. Targeting the Wilms tumor antigen 1 by TCR gene transfer: TCR variants improve tetramer binding but not the function of gene modified human T cells. J Immunol. 2007;179(9):5803-5810.

haematologica | 2016; 101(4)


ARTICLE

Stem Cell Transplantation

High-risk HLA alleles for severe acute graft-versus-host disease and mortality in unrelated donor bone marrow transplantation

Satoko Morishima,1 Koichi Kashiwase,2 Keitaro Matsuo,3 Fumihiro Azuma,2 Toshio Yabe,2 Aiko Sato-Otsubo,4 Seishi Ogawa,4 Takashi Shiina,5 Masahiro Satake,2 Hiroh Saji,6 Shunichi Kato,7 Yoshihisa Kodera,8 Takehiko Sasazuki,9 and Yasuo Morishima10 for the Japan Marrow Donor Program

Department of Hematology, Fujita Health University School of Medicine, Toyoake, Aichi; Japanese Red Cross Kanto-Koshinetsu Block Blood Center, Tokyo; 3Divsion of Molecular Medicine, Aichi Cancer Center Research Institute, Nagoya; 4Department of Pathology and Tumor Biology, Graduate School of Medicine, Kyoto University; 5Division of Basic Medical Science and Molecular Medicine, Department of Molecular Life Science, Tokai University School of Medicine, Isehara, Kanagawa; 6HLA Foundation Laboratory, Kyoto; 7Department of Cell Transplantation and Regenerative Medicine, Tokai University School of Medicine, Isehara, Kanagawa; 8Department of Promotion for Blood and Marrow Transplantation, Aichi Medical University School of Medicine, Nagakute; 9Institute for Advanced Study, Kyushu University, Fukuoka; and 10Division of Epidemiology and Prevention, Aichi Cancer Center Research Institute, Nagoya, Japan 1

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Haematologica 2016 Volume 101(4):491-498

2

ABSTRACT

H

LA molecules play an important role for immunoreactivity in allogeneic hematopoietic stem cell transplantation. To elucidate the effect of specific HLA alleles on acute graft-versus-host disease, we conducted a retrospective analysis using 6967 Japanese patients transplanted with T-cell-replete marrow from an unrelated donor. Using unbiased searches of patient and donor HLA alleles, patient and/or donor HLA-B*51:01 (patient: HR, 1.37, P<0.001; donor: HR, 1.35, P<0.001) and patient HLA-C*14:02 (HR, 1.35, P<0.001) were significantly associated with an increased risk of severe acute graft-versus-host disease. The finding that donor HLA-C*14:02 was not associated with severe acute graft-versus-host disease prompted us to elucidate the relation of these high-risk HLA alleles with patient and donor HLAC allele mismatches. In comparison to HLA-C allele match, patient mismatched HLA-C*14:02 showed the highest risk of severe acute graft-versus-host disease (HR, 3.61, P<0.001) and transplant-related mortality (HR, 2.53, P<0.001) among all patient mismatched HLA-C alleles. Although patient HLA-C*14:02 and donor HLA-C*15:02 mismatch was usually KIR2DL-ligand mismatch in the graft-versus-host direction, the risk of patient mismatched HLA-C*14:02 for severe acute graft-versushost disease was obvious regardless of KIR2DL-ligand matching. The effect of patient and/or donor HLA-B*51:01 on acute graft-versus-host disease was attributed not only to strong linkage disequilibrium of HLAC*14:02 and -B*51:01, but also to the effect of HLA-B*51:01 itself. With regard to clinical implications, patient mismatched HLA-C*14:02 proved to be a potent risk factor for severe acute graft-versus-host disease and mortality, and should be considered a non-permissive HLA-C mismatch in donor selection for unrelated donor hematopoietic stem cell transplantation.

haematologica | 2016; 101(4)

Correspondence: smorishi@fujita-hu.ac.jp

Received: September 21, 2015. Accepted: January 13, 2016. Pre-published: January 14, 2016. doi:10.3324/haematol.2015.136903

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/4/491

Š2016 Ferrata Storti Foundation Material published in Haematologica is covered by copyright. All rights reserved to Ferrata Storti Foundation. Copies of articles are allowed for personal or internal use. A permission in writing by the publisher is required for any other use.

491


S. Morishima et al.

Introduction

Table 1. Characteristics of patients and donors.

Characteristics Hematopoietic stem cell transplantation (HSCT) from an unrelated (UR) donor is now an established curative therapy for hematologic malignancies and other hematologic or immunological disorders. Human leukocyte antigen (HLA) molecules play an important role in the cellular discrimination of ‘self’ and ‘non-self’ in allogeneic HSCT, and the critical role of patient and donor human HLA matching status on graft rejection, graft-versus-host disease (GvHD), and survival after UR-HSCT has been well defined. The risk of transplant-related immunological events may vary according to patient-donor mismatching HLA locus,1-4 mismatching specific HLA epitopes,5,6 and specific mismatch combinations.7 Furthermore, several studies have demonstrated that specific individual HLA genotypes may also affect transplant outcome,8-12 although the results are not always consistent. The human major histocompatibility (MHC) region is the most gene-dense region of the human genome, containing more than 150 protein coding genes at approximately 4 Mb. Many of the HLA gene products have coordinated roles in immune function.13 The MHC region has been recognized as one of the most important genetic regions associated with many kinds of human disease, including autoimmune diseases, infections, and cancers.14 Linkage disequilibrium (LD) is the non-random association of alleles at different loci, and the long-range LD blocks observed in MHC haplotypes15 induce genetic fixity in this region. Our previous study revealed that the three common Japanese HLA haplotypes were highly conserved among unrelated individuals up to 8 Mb, and suggested that these haplotypes might contribute to the occurrence of acute GvHD in HLA-identical UR-HSCT.16 We therefore hypothesized that the immunological response following UR-HSCT depends, not only on patient-donor HLA allele matching, but also on specific individual HLA alleles themselves and their linked HLA region. Here we analyzed the effect of specific HLA alleles on acute GvHD using a large UR-HSCT cohort with extensive HLA-allele data. Grades III-IV acute GvHD was the clinical end point in this study, as this is considered to represent a severe or life-threatening immunological event after allo-HSCT. In unbiased screening of patient and donor HLA alleles, patient and/or donor HLA-B*51:01 and patient HLA-C*14:02 were significantly associated with a higher risk of severe acute GvHD. The effects of these alleles on acute GvHD were attributed to the specific patient mismatched HLA-C allele and its genetically linked HLA-B allele.

Methods Study population A total of 6967 unrelated donor transplant pairs from the Japan Marrow Donor Program (JMDP) database satisfied the following criteria and were included in the analysis: 1) transplanted from a serologically HLA-A, -B, and -DR antigen-matched donor between January 1993 and December 2010; 2) transplantation pairs were re-typed for HLA-A, -B, -C, -DRB1, -DQB1 and -DPB1 alleles; 3) transplanted non-T-cell-depleted marrow without in vivo use of anti-thymocyte globulin (ATG) for GvHD prophylaxis; 4) first transplantation; 5) Japanese ethnicity; and 6) survival for more 492

N.

Median patient age, years (range) 35 (0-73) Median donor age, years (range) 34 (20-58) Year of transplantation 1993-2000 2188 2001-2005 2738 2006-2010 2041 Donor-patient sex Male-male 2806 Female-female 1305 Male-female 1511 Female-male 1345 Disease Acute lymphoblastic leukemia 1653 Acute myeloid leukemia 2281 Myelodysplastic syndrome 721 Chronic myeloid leukemia 913 Aplastic anemia 410 Other 989 Leukemia risk Standard 3013 Advanced 2256 NA 1698 GvHD prophylaxis Cyclosporine-based 2916 Tacrolimus-based 3922 Other 129 Conditioning regimen Myeloablative 5171 Reduced-intensity 1240 Missing 556 HLA allele mismatching in the GvH direction HLA-A 760 HLA-B 353 HLA-C 1965 HLA-DRB1 1281 HLA-DQB1 1526 HLA-DPB1 4635

(31%) (39%) (29%) (40%) (19%) (22%) (19%) (24%) (33%) (10%) (13%) (6%) (14%) (43%) (32%) (24%) (42%) (56%) (2%) (74%) (18%) (8%) (11%) (5%) (28%) (18%) (22%) (67%)

Leukemia risk: standard risk indicates the first or second complete remission (CR) of acute myeloid leukemia; first CR of acute lymphoblastic leukemia, and first chronic phase of chronic myeloid leukemia, refractory anemia, and refractory anemia with ringed sideroblasts. Advanced risk indicates all other leukemias. NA: not applicable. Conditioning regimen: myeloablative regimen indicates conditioning regimens with total-body irradiation >8 Gy, oral busulfan ≥9 mg/kg, intravenous busulfan ≥7.2 mg/kg, or melphalan >140 mg/m2; reduced-intensity, reduced-intensity regimen indicates all other conditioning regimens.

than seven days after transplantation. Patients’ and donors’ characteristics are shown in Table 1. A final clinical survey of the patients was completed by September 2012. Informed consent was obtained from patients and donors in accordance with the Declaration of Helsinki, and approval of the study was obtained from the Institutional Review Board of Aichi Cancer Center, Fujita Health University and the JMDP.

HLA typing and definitions All donor-patient pairs were retrospectively genotyped for HLA-A, -C, -B, -DRB1, -DQB1, and -DPB1 alleles at the field 1 and field 2 level using either or both the polymerase chain reaction sequence-specific oligonucleotide and PCR-Sequencing Based Typing methods, as described elsewhere.4 HLA allele mismatch in the GvH direction among donor-patient pairs was defined as the haematologica | 2016; 101(4)


High-risk HLA allele for acute GvHD Table 2. Effect of HLA-C*14:02, -B*51:01 and their haplotypes on acute graft-versus-host disease (GvHD) and mortality.

HLA allele/haplotype Patient B*51:01 C*14:02 C*14:02 - B*51:01 Donor B*51:01 C*14:02 C*14:02 - B*51:01

N.

(%)

Acute GVHD grades III-IV HR (95% CI) P

HR

Overall mortality (95% CI)

P

1058 860 843

(15) (12) (12)

1.37 1.35 1.37

(1.19-1.59) (1.15-1.58) (1.17-1.60)

<0.001 <0.001 <0.001

1.18 1.18 1.18

(1.07-1.29) (1.07-1.30) (1.07-1.31)

<0.001 0.001 0.001

1079 868 858

(15) (12) (12)

1.35 1.11 1.12

(1.17-1.56) (0.94-1.31) (0.95-1.33)

<0.001 0.212 0.176

1.15 1.08 1.07

(1.05-1.26) (0.97-1.19) (0.91-1.10)

0.001 0.156 0.194

N: number of HLA allele-positive donors or patients; HR: hazard ratio indicates comparison of the specific HLA allele/haplotype-positive group to the -negative group, adjusted for clinical factors and HLA allele matching as listed in Table 1; CI: confidence interval.

patient’s alleles not being shared by the donor. Natural killer cell immunoglobulin-like receptor (KIR) ligand specificity of the HLAC antigen was determined according to the amino acid residues of the HLA-C allele. C1 ligand specificity consists of Asn80 and C2 ligand specificity consists of Lys80. KIR ligand mismatch in the GvH direction was defined as the donor’s KIR ligand for HLA-C not being shared by the patient’s ligand.17

Biostatistical methods The study evaluated the impact of specific HLA alleles with a frequency of more than 5% on the outcome of grades III-IV acute GvHD. Outcomes were compared among patient- and donor-specific HLA allele-positive and -negative groups using multivariable competing risk regression analysis,18 adjusted for clinical factors and HLA allele matching (Table 1). We included separate variables for HLA-A,-C,-B,-DRB1,-DQB1 and -DPB1 allele mismatches in the GvH direction. To analyze the effect of patient mismatched HLA-C allele, those pairs matched for 1 HLA-C allele and mismatched for another HLA-C allele were extracted. The risk of each patient mismatched HLA-C allele on grades III-IV acute GvHD was compared with the HLA-C allele match. The influences of the level of expression of the patient mismatched HLA-C allotype were assessed as described previously.19 The effects of HLA-C allele mismatch combinations were also evaluated using the pairs matched for 1 HLA-C allele and mismatched for another HLA-C allele, and the risk of each HLA-C mismatch combination of grades III-IV acute GvHD was compared with the HLA-C allele match. Multivariable competing risk regression analyses18 were conducted to evaluate the impact on acute GvHD and transplantrelated mortality. A Cox’s proportional hazards regression model was used to evaluate the impact on overall survival (OS).20 A detailed description of the statistical methods is available in the Online Supplementary Appendix.

[CI], 1.19-1.59; P<0.001; donor: HR, 1.35; 95%CI: 1.171.56; P<0.001) and patient HLA-C*14:02 (HR, 1.35; 95%CI: 1.15-1.58; P<0.001) (Table 2 and Online Supplementary Table S1). These HLA alleles were also associated with a higher risk of mortality (patient HLAB*51:01: HR, 1.18; 95%CI: 1.07-1.29; P<0.001; donor HLA-B*51:01: HR, 1.15; 95%CI: 1.05-1.26; P=0.001; patient HLA-C*14:02: HR, 1.18; 95%CI: 1.07-1.30; P=0.001).

HLA-C*14:02 and -B*51:01 were in strong linkage disequilibrium Since patient and/or donor HLA-B*51:01, patient HLAC*14:01 were associated with a higher risk of severe acute GvHD, the linkage between these alleles was examined. Firstly, because almost all patients with HLA-B*51:01 (1053 of 1058) received transplants from donors with HLA-B*51:01, we were unable to determine which patient or donor HLA-B*51:01 contributed to increasing the risk of acute GvHD. Secondly, HLA-B*51:01 demonstrated strong positive LD with HLA-C*14:02 among Japanese.21 In the present analysis, 98% of patients with HLA-C*14:02 (843 of 860) were HLA-B*51:01-positive; therefore, the patient HLA-C*14:02-B*51:01 haplotype showed a similar effect in increasing the risk of grades III-IV acute GvHD (HR, 1.37; 95%CI: 1.17-1.60; P<0.001) and mortality (HR, 1.18; 95%CI: 1.07-1.31; P<0.001) as patient HLA-C*14:02 (Table 2). Among HLA-B*51:01-positive patients (n=1058), 843 patients (80%) had HLA-C*14:02, while 215 patients had HLA-C alleles other than HLA-C*14:02. We analyzed the effect of HLA-B*51:01 on acute GvHD in patients without HLA-C*14:02 to ascertain whether HLA-B*51:01 itself has an increased risk of acute GvHD irrespective of HLA-C*14:02. Patient HLA-B*51:01 was significantly associated with an increased risk of grades III-IV GvHD in the subgroup of excluded patients with HLA-C*14:02 (n=6197) (HR, 1.34; 95%CI: 1.01-1.79; P=0.046) (Table 3).

Results Identification of HLA alleles associated with grades III-IV acute GvHD

Increasing risk effect of HLA-C*14:02 and -B*51:01 on acute GvHD was prominent in HLA-C mismatched transplant

The number of HLA alleles with a frequency more than 5% in each locus was as follows: HLA-A 7, -C 8, -B 8, DRB1 7, -DQB1 8, and -DPB1 5. P<0.00116 was considered statistically significant (Bonferroni correction). Among 43 HLA alleles with a frequency more than 5%, the only alleles significantly associated with an increased risk of grades III-IV acute GvHD were patient and donor HLA-B*51:01 (patient: HR, 1.37; 95% confidence interval

In contrast to the increasing risk effect of patient HLAC*14:02 on grades III-IV acute GvHD, the effect of donor HLA-C*14:02 was not significant (HR, 1.11; 95%CI: 0.941.31; P=0.212) (Table 2). We assumed that the difference in the impact of HLA-C*14:02 on acute GvHD between patients and donors was attributable to HLA-C allele mismatch between patient and donor, and accordingly performed subgroup analysis of patient groups stratified by

haematologica | 2016; 101(4)

493


S. Morishima et al.

Figure 1. Impact of patient mismatched HLA-C allele on grades III-IV acute GvHD. Results of multivariable competing risk regression analysis for the effect of patient mismatched HLA-C allele on grades III-IV acute GvHD. The hazard ratio (HR) of each mismatched HLA-C allele was estimated by comparison to HLA-C match (n=4825). The solid vertical line at 1.69 indicates the HR of overall HLA-C mismatch in the GvH direction. Others contain patient mismatched HLA-C alleles with fewer than 20 patients, as follows: C*01:03, C*03:14, C*03:23, C*05:01, C*07:04, C*12:02, C*12:03 and HLA-C*14:03.

HLA-C matching status (Table 3). The association between patient HLA-C*14:02 and risk of grades III-IV acute GvHD was more prominent in pairs with HLA-C allele mismatch in the GvH direction (n=1965) (HR, 1.49; 95%CI: 1.18-1.87; P<0.001) than pairs with HLA-C match in the GvH direction (n=5002) (HR, 1.23; 95%CI: 0.991.53; P=0.067). Similarly, the effect of HLA-B*51:01 on acute GvHD was more obvious in pairs with HLA-C allele mismatch in the GvH direction (HR, 1.47; 95%CI: 1.201.82; P<0.001) than in matched pairs (HR, 1.26; 95%CI: 1.02-1.55; P=0.029). We also performed subgroup analysis of patient groups stratified by various clinical factors (i.e. GvHD prophylaxis, conditioning regimen, disease, year of transplantation) (Table 3). There were no statistically significant heterogeneities in the effects of patient HLA-B*51:01 and HLAC*14:02 in these subgroups, indicating the absence of interactions between the risk of this HLA allele and these clinical factors.

Effect of patient mismatched HLA-C alleles on acute GvHD The effects of patient HLA-B*51:01 and -C*14:02 on grades III-IV acute GvHD were more prominent in HLAC allele mismatched pairs than matched pairs, which prompted us to examine the relationship between patientdonor HLA-C allele mismatch and the effect of these alleles on acute GvHD. First, we examined the effect of patient mismatched HLA alleles on grades III-IV acute GvHD (Figure 1), as these are HLA targets in the GvH direction. The patients transplanted from donors with matched for 1 HLA-C allele and mismatched for another HLA-C allele (n=1795) were compared with HLA-C matched transplant 494

(n=4825). Mismatched alleles having fewer than 20 pairs were grouped together as “Others”. Median HR of patient mismatched HLA-C alleles was 1.56 compared with HLAC matched transplants. Interestingly, patient mismatched HLA-C*14:02 showed a strikingly high risk of grades III-IV acute GvHD among all patient mismatched HLA-C alleles (HR, 3.61; 95%CI: 2.71-4.82; P<0.001). Cumulative incidences of grades III-IV acute GvHD in patients with patient HLA-C*14:02 mismatch, other patient HLA-C allele mismatch, and HLA-C match were 41.6% (95%CI: 32.6-50.3), 23.0% (95%CI: 21.0-25.0), and 13.6% (95%CI: 12.6-14.6) (Figure 2A). Patient HLA-C*14:02 mismatched transplant also showed the highest risk of transplant-related mortality (HR, 2.53; 95%CI: 1.92-3.34; P<0.001) and overall mortality (HR, 1.91; 95%CI: 1.52-2.42; P<0.001) among all patient mismatched HLA-C alleles (Online Supplementary Figures S1 and S2). Overall survival in patients with patient HLA-C*14:02 mismatch, with other patient HLA-C allele mismatch, and with HLA-C match were 35.5% (95%CI: 26.7-44.3), 46.1% (95%CI: 43.648.6) and 51.4% (95%CI: 49.9-52.9) (Figure 2B). A recent report showed that the expression level of a patient’s mismatched HLA-C allotype affects severe acute GvHD and mortality after unrelated hematopoietic stem cell transplantation.19 We analyzed whether the expression levels of patient mismatched HLA-C allotype influence grades III-IV acute GvHD. In comparison with patient mismatched HLA-C*07, HLA-C*14 was associated with a significantly higher risk of grades III-IV acute GvHD (HR, 2.14; 95%CI: 1.27-3.60; P=0.004) and higher transplant-related mortality (HR, 1.82; 95%CI: 1.15-2.86; P=0.01), whereas there were no significant differences between other HLA-C allotypes and HLA-C*07 (Online Supplementary Table S2). An increasing expression level of patient mismatched HLA-C haematologica | 2016; 101(4)


High-risk HLA allele for acute GvHD

was not associated with either the risk of grades III-IV acute GvHD (P=0.228) or transplant-related mortality (P=0.108) (Online Supplementary Figure 3A and B).

(HR, 2.66; 95%CI: 1.43-4.93; P=0.002), and donor HLAC*01:02 and patient C*14:02 (HR, 4.01; 95%CI: 2.19-7.35; P<0.001).

Impact of HLA-C allele mismatch combinations on acute GvHD

Relation of KIR2DL ligand mismatch and high-risk HLA-C allele mismatch combination

The previous JMDP study showed that specific donorpatient allele mismatch combinations were associated with severe acute GvHD.6 We analyzed the association between the impact of patient HLA-C*14:02 itself and that of donor-patient HLA-C allele mismatch combinations on grades III-IV acute GvHD. There were 33 HLA-C allele mismatch combinations with more than 15 donor-patient pairs, and the mismatch combinations having fewer than 15 pairs were grouped together as â&#x20AC;&#x153;Othersâ&#x20AC;?. HRs of HLAC allele mismatch combinations for grades III-IV acute GvHD in comparison to HLA-C allele match are shown in the Online Supplementary Table S3. Among 33 HLA-C mismatch combinations, there were three combinations having a patient HLA-C allele of -C*14:02, and these three combinations showed higher HRs than the median HR for all mismatch combinations (median HR=1.75), as follows: donor C*15:02 and patient C*14:02 (HR, 4.39; 95%CI: 2.98-6.45; P<0.001), donor C*03:04 and patient C*14:02

As described above, the donor C*15:02 and patient C*14:02 mismatch combination was one of the highest risk HLA-C allele mismatch combinations for grades III-IV acute GvHD. Our previous JMDP study demonstrated that KIR2DL-L mismatch in the GvH direction (KIR2DL-LMM-G) was associated with an increased risk of severe acute GvHD and mortality in T-cell-replete UR-HSCT.22 We examined the impact of patient-donor KIR2DL ligand mismatch on the risk of acute GvHD related to this HLA-C mismatch combination. In our Japanese cohort, the frequencies of KIR2DL-ligand HLA-C1 and -C2 were 93% and 7%, respectively, and the majority (91%) of KIR2DL-LMM-G was the combination of patient HLA-C1/C1 and donor HLA-C1/C2. HLA-C*14:02 belongs to HLA-C1, while HLA-C*15:02 belongs to HLA-C2; therefore, donor HLA-C*15:02 and patient HLA-C*14:02 mismatch combination is usually KIR2DL-L-MM-G. In fact, 92% of pairs (60 of 65) with the donor C*15:02 and patient C*14:02 mis-

A

B

Figure 2. Acute graft-versus-host disease (GvHD) and survival curve by patient mismatched HLA-C. Patients were divided into three groups as follows: patient HLA-C*14:02 mismatch (n=118), other patient HLA-C allele mismatch (n=1677), and HLA-C allele match (4825). (A) Cumulative incidences of grades III-IV and (B) Kaplan-Meier curve of survival are shown. P-value indicates comparison between patient HLA-C*14:02 mismatch and other patient HLA-C mismatch.

haematologica | 2016; 101(4)

495


S. Morishima et al. Table 3. Subgroup analysis of the effect of patient HLA-B*51:01 and -C*14:02 on grades III-IV acute GvHD.

HLA-B*51:01 N N HR (95%CI) (total) (positive) All patients HLA-C mismatch in GvH direction HLA-C matched HLA-C mismatched Assciation with HLA-C*14:02 C*14:02-negative GvHD prophylaxis Cyclosporine-based Tacrolimus-based Conditioning regimen Myeloablative Reduced intensity Disease Leukemia Other diseases Years of transplantation 1993-2000 2001-2005 2006-2010

6967

1058

1.37 (1.19-1.59)

P

Heterogeneity N HR P (positive)

<0.001

HLA-C*14:02 (95%CI) P

Heterogeneity P

860

1.35 (1.15-1.58) <0.001

576 284

1.23 (0.99-1.53) 0.067 1.49 (1.18-1.87) <0.001

0.303 656 402

5002 1965

1.26 (1.02-1.55) 1.47 (1.20-1.82)

0.029 <0.001

6197

215

1.34 (1.01-1.79)

0.046

2916 3922

430 609

1.36 (1.10-1.69) 1.34 (1.09-1.65)

0.005 0.006

5171 1240

797 178

1.39 (1.18-1.63) 1.37 (0.92-2.02)

<0.001 0.119

5568 760

865 193

1.32 (1.13-1.54) 1.70 (1.17-2.47)

<0.001 0.005

2188 2738 2041

330 418 310

1.41 (1.12-1.77) 1.26 (0.99-1.59) 1.54 (1.13-2.10)

0.004 0.058 0.006

0.249

N.A. 0.923

0.623 358 490

1.37 (1.09-1.72) 0.006 1.26 (1.00-1.59) 0.046

649 141

1.38 (1.15-1.65) <0.001 1.47 (0.96-2.26) 0.080

705 155

1.29 (1.09-1.53) 0.004 1.78 (1.20-2.65) 0.004

274 332 254

1.39 (1.08-1.78) 0.009 1.31 (1.02-1.69) 0.034 1.35 (0.96-1.92) 0.089

0.943

0.802

0.270

0.199

0.585

0.957

N (total): total number of patients in subgroup; N (positive): number of HLA-B*51:01- or -C*14:02-positive patients in each subgroup; HR: hazard ratio, adjusted for clinical factors and HLA allele matching as listed in Table 1; CI: confidence interval; NA: not applicable.

Table 4. Effect of patient HLA-C*14:02 mismatch on severe acute graft-versus-host disease (GvHD) and mortality in relation to KIR2DL-L-MM-G

Patient HLA-C*14:02 mismatch (-) (-) (+) (+)

KIR2DL-L -MM-G

N

(-) (+) (-) (+)

1600 222 81 62

Grades III-IV acute GvHD HR (95%CI) P

HR

Overall mortality (95%CI)

P

1.00 1.09 1.67 2.59

1.00 1.21 1.50 1.54

(1.00-1.46) (1.12-2.01) (1.12-2.11)

0.047 0.007 0.008

(0.82-1.46) (1.12-2.50) (1.80-3.72)

0.552 0.012 <0.001

Patients receiving transplants from HLA-C mismatched donors in the GvH direction were divided into four groups according to the presence of patient HLA-C*14:02 mismatch and KIR2DL-ligand mismatch in the GvH direction (KIR2DL-L-MM-G).The risk of grades III-IV acute GvHD and overall mortality were compared to those of the patient HLA-C*14:02 mismatch(-) KIR2DL-L-MM-G(-) group. N: number of pairs; HR: hazard ratio, adjusted for clinical factors and HLA allele matching as listed in Table 1; CI: confidence interval.

match combination were KIR2DL-L-MM-G, and they accounted for approximately 21% of all pairs with KIR2DLL-MM-G (n=284). KIR2DL-L-MM-G among patients transplanted from donors with HLA-C allele mismatch in the GvH direction was associated with a significantly higher risk of grades III-IV acute GvHD (HR, 1.34; 95%CI: 1.051.69; P=0.015). To clarify the effect of patient HLA-C*14:02mismatch on grades III-IV acute GvHD in relation to KIR2DL-L-MM-G, patients transplanted from HLA-C mismatched donors were divided into four groups (Table 4). Both the patient HLA-C*14:02-mismatch(+)KIR-2DL-LMM-G(+) group (n=62) and patient HLA-C*14:02-mismatch(+)KIR-2DL-L-MM-G(-) group (n=81) showed a significantly higher risk of acute GvHD (HR, 2.59; P<0.001 and HR, 1.67; P=0.012, respectively) and overall mortality (HR, 1.54; P=0.008 and HR, 1.50; P=0.007, respectively) than the patient HLA-C*14:02-mismatch(-)KIR-2DL-L-MM-G(-) group (n=1600). On the other hand, the patient HLAC*14:02-mismatch(-)KIR-2DL-L-MM-G(+) group (n=222) showed no significant difference in grades III-IV acute GvHD (HR, 1.09; P=0.552), in spite of the higher trend in 496

overall mortality (HR, 1.21; P=0.047) than in the HLAC*14:02-mismatch(-)KIR-2DL-L-MM-G(-) group. Among 62 pairs in the patient HLA-C*14:02-mismatch(+)KIR-2DLL-MM-G(+) group, 60 pairs had the mismatch combination of patient C*14:02 and donor C*15:02.

Discussion In this study, we examined the influence of patient and donor HLA alleles on severe acute GvHD after UR-HSCT using 6967 unrelated Japanese pairs. Given the importance of GvHD prophylaxis to acute GvHD in allogeneic HSCT, we aimed to elucidate the biological effects of HLA alleles on acute GvHD by selecting patients who underwent transplantation from non-T-cell-depleted bone marrow donors without ATG. We demonstrated that patient and/or donor HLAB*51:01 and patient HLA-C*14:02 were significantly associated with an increased risk of grades III-IV acute GvHD. Our finding that patient HLA-C*14:02 was associated haematologica | 2016; 101(4)


High-risk HLA allele for acute GvHD

with an increased risk of severe acute GvHD, whereas donor HLA-C*14:02 was not, prompted us to investigate further the association of these high-risk HLA-alleles with patient-donor HLA-C allele mismatches. Results showed that patient mismatched HLA-C*14:02 was the key factor in the risk of severe acute GvHD. Petersdorf et al. reported that increasing expression of patient mismatched HLA-C was significantly associated with an increased risk of grades III-IV acute GvHD in UR-HSCT.19 In their report, the expression level of HLA-C*14 was highest among 14 HLA-C allotypes, and was associated with a strikingly high risk of acute GvHD and non-relapse mortality compared to other HLA-C allotypes. In our present study, patient mismatched HLA-C*14:02 also showed the highest risk of severe acute GvHD, transplant-related mortality, and overall mortality among all the patient mismatched HLA-C alleles. Although we did not demonstrate correlations between the expression levels of patient mismatched HLA-C alleles and transplant outcomes in our study, an extraordinarily high risk of patient mismatched HLA-C*14:02 for severe acute GvHD has been seen in other ethnic cohorts, indicating that this particular HLA-C*14:02 could play a very important role in immunoreactivity in allo-HSCT. In our analysis of the association between donor-patient HLA-C allele mismatch combinations and acute GvHD, each of the three mismatch combinations with patient HLA-C*14:02 showed a higher HR than the other HLA-C mismatch combinations. In the previous JMDP study,6 the donor-C*15:02 and patient C*14:02 mismatch was identified as a non-permissive mismatch combination for severe acute GvHD, and the other two mismatches (donorC*03:04 and patient-C*14:02, donor-C*01:02 and patientC*14:02) were also associated with a higher risk of severe acute GvHD. Our present results are, therefore, consistent with those of this previous study. KIR2DL-L-MM-G was reported to be significantly associated with a higher risk of severe acute GvHD than KIR2DL-L-match in the previous JMDP study.22 In our present JMDP study, KIR2DL-L-MM-G was consistently associated with a higher risk of grades III-IV acute GvHD even after restriction of the analysis population to pairs with HLA-C mismatch in the GvH direction. KIR2DL-L mismatching was first reported to be associated with improved survival and reduced relapse without acute GvHD in HLA mismatched T-cell-depleted and CD34+selected haploidentical transplantation.23 Subsequent studies examining the effect of KIR2DL-L mismatch on transplant outcome in UR-HSCT showed inconsistent results, presumably because of differences in patient population, conditioning regimen, GvHD prophylaxis and sample size.24-29 Petersdorf et al. demonstrated that the negative effect of non-relapse mortality was evident in the high expression of patient mismatched HLA-C allotype (C*01 and C*14) compared to low expression (C*03 and C*07) among KIR2DL-L mismatched pairs, but not among KIR2DL-L matched pairs, suggesting that the KIR2DL-L mismatching effect was influenced by the expression levels of patient mismatched HLA-C alleles.19 In our cohort, patient HLA-C*14:02 and donor HLA-C*15:02 mismatch, which was eventually classified as KIR2DL-L-MM-G, showed the highest frequency among mismatch combinations with patient HLA-C*14:02. Furthermore, the negative effect of patient mismatched HLA-C*14:02 on grades III-IV acute GvHD and survival was also evident in KIR2DL-L haematologica | 2016; 101(4)

matched pairs. These findings indicate that patient mismatched HLA-C*14:02 is a critical factor in severe acute GvHD and survival regardless of KIR2DL-L matching status. Therefore, we assume that the higher risk of grades IIIIV acute GvHD for KIR2DL-L-MM-G in the JMDP cohort was attributable to this particular mismatch combination of patient HLA-C*14:02 and donor -C*15:02. A recent study using Japanese UR-HSCT data in the Transplant Registry Management Program demonstrated that the adverse impact of high-risk HLA allele mismatch combinations identified in a previous JMDP study was significant in the early transplant years (1993-2001), but became less significant in later transplant years (after 2002).30 We analyzed the effect of patient HLA-C*14:02 and HLA-C*51:01 on grades III-IV acute GvHD and overall mortality (Online Supplementary Table S4) and the effect of patient mismatched HLA-C on transplant outcomes (Online Supplementary Figure S4) in the patients transplanted from 2002 through 2010. The adverse effect of patient HLA-C*14:02, -B*51:01, and patient mismatched HLAC*14:02 on grades III-IV acute GvHD and mortality were obvious during this period. These results suggest that the biological impact of these HLA alleles on transplant outcomes should be considered even in current transplantation conditions. Since HLA-C*14:02 is strongly linked to HLA-B*51:01 among Japanese, it is possible that the significantly negative effect of patient HLA-B*51:01 on severe acute GvHD was detected as a consequence of the effect of patient mismatched HLA-C*14:02. However, patient HLA-B*51:01 was also associated with a higher risk of severe acute GvHD in subgroups of patients without HLA-C*14:02 and of HLA-C matched pairs. We therefore assume that HLA-B*51:01 itself plays a role in the development or exacerbation of acute GvHD. As for the genetic association between HLA-B51 and autoimmune or inflammatory diseases, an association with Behçet’s disease has been shown in numerous ancestry groups and populations.31 Behçet’s disease is an inflammatory disease characterized by recurrent oral aphthous ulcers and numerous potentially systemic manifestations.32 Although the pathogenic mechanisms related to HLAB51 still need to be clarified, several possible mechanisms for a genetic linkage to HLA-B51 have been suggested. HLA-B*51:01 binds a potentially large pool of peptides with relatively low affinity, and it has been speculated that inefficient tolerance to HLA-B*51:01binding self-peptides might be a predisposing factor for the development of Behçet’s disease.33 Recently, HLAB*51:01 and HLA-C*14:02 were reported to be susceptibility alleles among Japanese for Crohn’s disease, a subtype of inflammatory bowel mucosal disease.34 Since the linkage between HLA-C*14:02 and -B*51:01 in other ethnic groups is weaker than in Japanese,35 further extensive studies are required to elucidate the immunological mechanisms of the genetic association between HLAB*51:01and acute GvHD in other ethnic groups. In conclusion, using unbiased searches of patient and donor HLA alleles in the JMDP database, we found that patient mismatched HLA-C*14:02 was the most potent risk factor of severe acute GvHD and mortality, and that this was eventually the most highly expressed HLA-C allele, as previously reported, and might evoke strong allogeneic immune reactions. Furthermore, an increased risk of patients and/or donor HLA-B*51:01 for severe acute GvHD 497


S. Morishima et al.

was attributed not only to the genetic linkage of HLAC*14:02 and -B*51:01 in the Japanese population, but also to HLA-B*51:01 itself. With regard to clinical implications, patient mismatched HLA-C*14:02 should be considered a non-permissive mismatch in UR-HSCT. Studies should investigate whether consideration of this particular high-risk HLA mismatch can be extended to other graft sources and donors in alloHSCT, such as cord blood and HLA-mismatch-related transplantation. Acknowledgments The authors thank the staff members of the transplantation

References 1. Sasazuki T, Juji T, Morishima Y, et al. Effect of matching of class I HLA alleles on clinical outcome after transplantation of hematopoietic stem cells from an unrelated donor. Japan Marrow Donor Program. N Engl J Med. 1998;339(17):1177-1185. 2. 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. 3. 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. 4. 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. 5. Ferrara GB, Bacigalupo A, Lamparelli T, et al. Bone marrow transplantation from unrelated donors: the impact of mismatches with substitutions at position 116 of the human leukocyte antigen class I heavy chain. Blood. 2001;98(10):3150-3155. 6. Kawase T, Morishima Y, Matsuo K, et al. High-risk HLA allele mismatch combinations responsible for severe acute graft-versus-host disease and implication for its molecular mechanism. Blood. 2007; 110(7):2235-2241. 7. Fernandez-Vina MA, Wang T, Lee SJ, et al. Identification of a permissible HLA mismatch in hematopoietic stem cell transplantation. Blood. 2014;123(8):1270-1278. 8. Martin PJ, Gooley T, Anasetti C, Petersdorf EW, Hansen JA. HLAs and risk of acute graft-vs.-host disease after marrow transplantation from an HLA-identical sibling. Biol Blood Marrow Transplant. 1998; 4(3):128-133. 9. Battiwalla M, Hahn T, Radovic M, et al. Human leukocyte antigen (HLA) DR15 is associated with reduced incidence of acute GVHD in HLA-matched allogeneic transplantation but does not impact chronic GVHD incidence. Blood. 2006;107(5):19701973. 10. Stern M, Passweg J, Tiercy JM, et al. Human leukocyte antigen DR15 is associated with reduced relapse rate and improved survival after human leukocyte antigen-identical sibling hematopoietic stem cell transplantation. Biol Blood Marrow Transplant. 2006;12(11):1169-1175. 11. Battiwalla M, Ellis K, Li P, et al. HLA DR15 antigen status does not impact graft-versus-

498

12.

13.

14. 15.

16.

17.

18. 19.

20. 21.

22.

23.

24.

centers and donor centers, the Japan Marrow Donor Program office, and the Japanese Data Center for Hematopoietic Cell Transplantation for their generous co-operation. Funding This work was supported by grants from Practical Research for Allergic Disease and Immunology (Research on Technology of Medical Transplantation), the Japan Agency for Medical Research and Development, and the Japanese Ministries of Health, Labor and Welfare (H23-Immunology-010 and H26Immunology-106) and Education, Culture, Sports, Science and Technology (MEXT KAKENHI Grant Number 22133011 and JSPS KAKENHI Grant Number 26461455).

host disease or survival in HLA-matched sibling transplantation for hematologic malignancies. Biol Blood Marrow Transplant. 2012;18(8):1302-1308. Khouri IF, Bassett R, Poindexter N, et al. Nonmyeloablative allogeneic stem cell transplantation in relapsed/refractory chronic lymphocytic leukemia: long-term followup, prognostic factors, and effect of human leukocyte histocompatibility antigen subtype on outcome. Cancer. 2011; 117(20):4679-4688. Shiina T, Hosomichi K, Inoko H, Kulski JK. The HLA genomic loci map: expression, interaction, diversity and disease. J Hum Genet. 2009;54(1):15-39. Trowsdale J. The MHC, disease and selection. Immunol Lett. 2011;137(1-2):1-8. Yunis EJ, Larsen CE, Fernandez-Vina M, et al. Inheritable variable sizes of DNA stretches in the human MHC: conserved extended haplotypes and their fragments or blocks. Tissue Antigens. 2003;62(1):1-20. Morishima S, Ogawa S, Matsubara A, et al. Impact of highly conserved HLA haplotype on acute graft-versus-host disease. Blood. 2010;115(23):4664-4670. Ruggeri L, Capanni M, Casucci M, et al. Role of natural killer cell alloreactivity in HLAmismatched hematopoietic stem cell transplantation. Blood. 1999;94(1):333-339. Fine JP GR. A proportional hazards model for subdistribution of a competing risk. J Am Stat Assoc. 1999;94:456-509. Petersdorf EW, Gooley TA, Malkki M, et al. HLA-C expression levels define permissible mismatches in hematopoietic cell transplantation. Blood. 2014;124(26):3996-4003. Cox D. Regression model and life tables. J R Stat Soc B. 1972;34(2):187-200. Ikeda N, Kojima H, Nishikawa M, et al. Determination of HLA-A, -C, -B, -DRB1 allele and haplotype frequency in Japanese population based on family study. Tissue Antigens. 2015;85(4):252-259. Morishima Y, Yabe T, Matsuo K, et al. Effects of HLA allele and killer immunoglobulin-like receptor ligand matching on clinical outcome in leukemia patients undergoing transplantation with T-cell-replete marrow from an unrelated donor. Biol Blood Marrow Transplant. 2007;13(3):315-328. Ruggeri L, Capanni M, Urbani E, et al. Effectiveness of donor natural killer cell alloreactivity in mismatched hematopoietic transplants. Science. 2002;295(5562):20972100. Davies SM, Ruggieri L, DeFor T, et al. Evaluation of KIR ligand incompatibility in mismatched unrelated donor hematopoietic transplants. Killer immunoglobulin-like receptor. Blood. 2002;100(10):3825-3827.

25. Giebel S, Locatelli F, Lamparelli T, et al. Survival advantage with KIR ligand incompatibility in hematopoietic stem cell transplantation from unrelated donors. Blood. 2003;102(3):814-819. 26. Bornhauser M, Schwerdtfeger R, Martin H, Frank KH, Theuser C, Ehninger G. Role of KIR ligand incompatibility in hematopoietic stem cell transplantation using unrelated donors. Blood. 2004;103(7):2860-2861. 27. Schaffer M, Malmberg KJ, Ringden O, Ljunggren HG, Remberger M. Increased infection-related mortality in KIR-ligandmismatched unrelated allogeneic hematopoietic stem-cell transplantation. Transplantation. 2004;78(7):1081-1085. 28. Beelen DW, Ottinger HD, Ferencik S, et al. Genotypic inhibitory killer immunoglobulin-like receptor ligand incompatibility enhances the long-term antileukemic effect of unmodified allogeneic hematopoietic stem cell transplantation in patients with myeloid leukemias. Blood. 2005; 105(6):2594-2600. 29. Farag SS, Bacigalupo A, Eapen M, et al. The effect of KIR ligand incompatibility on the outcome of unrelated donor transplantation: a report from the center for international blood and marrow transplant research, the European blood and marrow transplant registry, and the Dutch registry. Biol Blood Marrow Transplant. 2006; 12(8):876-884. 30. Kanda Y, Kanda J, Atsuta Y, et al. Changes in the clinical impact of high-risk human leukocyte antigen allele mismatch combinations on the outcome of unrelated bone marrow transplantation. Biol Blood Marrow Transplant. 2014;20(4):526-535. 31. Maldini C, Lavalley MP, Cheminant M, de Menthon M, Mahr A. Relationships of HLAB51 or B5 genotype with Behcet's disease clinical characteristics: systematic review and meta-analyses of observational studies. Rheumatology. 2012;51(5):887-900. 32. Direskeneli H. Behcet's disease: infectious aetiology, new autoantigens, and HLA-B51. Ann Rheum Dis. 2001;60(11):996-1002. 33. Gebreselassie D, Spiegel H, Vukmanovic S. Sampling of major histocompatibility complex class I-associated peptidome suggests relatively looser global association of HLAB*5101 with peptides. Hum Immunol. 2006;67(11):894-906. 34. Oryoji D, Hisamatsu T, Tsuchiya K, et al. Associations of HLA class I alleles in Japanese patients with Crohn's disease. Genes Immun. 2015;16(1):54-56. 35. Maiers M, Gragert L, Klitz W. High-resolution HLA alleles and haplotypes in the United States population. Hum Immunol. 2007;68(9):779-788.

haematologica | 2016; 101(4)


ARTICLE

Stem Cell Transplantation

Infused total nucleated cell dose is a better predictor of transplant outcomes than CD34+ cell number in reduced-intensity mobilized peripheral blood allogeneic hematopoietic cell transplantation

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Paul S. Martin,4 Shuli Li,3 Sarah Nikiforow,1,2 Edwin P. Alyea III,1,2 Joseph H. Antin,1,2 Philippe Armand,1,2 Corey S. Cutler,1,2 Vincent T. Ho,1,2 Natasha Kekre,1,2 John Koreth,1,2 C. John Luckey,5 Jerome Ritz,1,2 and Robert J. Soiffer1,2

Division of Hematologic Malignancies, Department of Medical Oncology, Dana-Farber Cancer Institute, Boston, MA; 2Harvard Medical School, Boston, MA; 3Department of Biostatistics and Computational Biology, Dana-Farber Cancer Institute, Boston, MA; 4 Fred Hutchinson Cancer Research Center and University of Washington School of Medicine, Seattle, WA; and 5Department of Pathology, University of Virginia, Charlottesville, VA, USA 1

Haematologica 2016 Volume 101(4):499-505

ABSTRACT

M

obilized peripheral blood is the most common graft source for allogeneic hematopoietic stem cell transplantation following reduced-intensity conditioning. In assessing the effect of donor cell dose and graft composition on major transplant outcomes in the reduced-intensity setting, prior studies focused primarily on CD34+ cell dose and reported conflicting results, especially in relation to survival end-points. While the impact of total nucleated cell dose has been less frequently evaluated, available studies suggest higher total nucleated cell dose is associated with improved survival outcomes in the reduced-intensity setting. In order to further explore the relationship between CD34+ cell dose and total nucleated cell dose on reduced-intensity transplant outcomes, we analyzed the effect of donor graft dose and composition on outcomes of 705 patients with hematologic malignancies who underwent reduced-intensity peripheral blood stem cell transplantation at the Dana Farber Cancer Institute from 2000 to 2010. By multivariable analysis we found that higher total nucleated cell dose (top quartile; ≥10.8 x 1010 cells) was associated with improved overall survival [HR 0.69 (0.54-0.88), P=0.0028] and progression-free survival [HR 0.68 (0.54-0.85), P=0.0006]. Higher total nucleated cell dose was independently associated with decreased relapse [HR 0.66 (0.51-0.85), P=0.0012] and increased incidence of chronic graft-versus-host disease [HR 1.4 (1.121.77), P=0.0032]. In contrast, higher doses of CD34+ cells (top quartile; ≥10.9 x 106/kg) had no significant effect on graft-versus-host disease or survival outcomes. These data suggest total nucleated cell dose is a more relevant prognostic variable for reduced-intensity transplant outcomes than the more commonly studied CD34+ cell dose.

Introduction The impact of graft composition and donor cell dose on allogeneic hematopoietic stem cell transplantation outcomes has been extensively studied in myeloablative conditioning transplantation, and CD34+ cell dose is often utilized as the productrelated variable of choice for outcomes analysis. Given that a cell-based graft-versus-tumor effect is the central therapeutic mechanism in reduced-intensity conditioning (RIC) transplantation, product-related variables may have an even more sighaematologica | 2016; 101(4)

Correspondence: psmartin@fhcrc.org

Received: August 13, 2015. Accepted: January 4, 2016. Pre-published: Juanuary 14, 2016. doi:10.3324/haematol.2015.134841

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/4/499

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

499


P.S. Martin et al.

nificant impact on transplant outcomes in the RIC setting. However, relatively few studies have reported on productrelated outcome data in RIC transplants. These studies have primarily focused on CD34+ cell dose, most reporting that higher CD34+ cell doses are associated with more rapid engraftment with a variable effect on the incidence of graft-versus-host disease (GVHD) and survival outcomes.1-5 Gomez-Almaguer and colleagues reported on 138 recipients of RIC hematopoietic stem cell transplants, showing that higher CD34+ doses correlated with improved overall survival.6 A CIBMTR registry analysis of 1054 RIC hematopoietic stem cell transplant recipients found that lower CD34+ doses correlated with increased transplant-related mortality and decreased overall survival.7 In contrast, Remberger and colleagues reported that higher CD34+ doses were actually associated with higher relapse rates and lower overall survival.8 Relative to CD34+ cell dose, total nucleated cell (TNC) dose has been less frequently studied. However, in the limited number of studies that have examined TNC, a trend toward improved survival outcomes with higher TNC dose, independently of CD34+ dose, has been found.9,10 The current study examines the impact of both TNC and CD34+ cell dose on major transplant outcomes in a cohort of 705 consecutive patients undergoing RIC allogeneic peripheral blood hematopoietic stem cell transplantation.

cells as a graft source. Patients who underwent either in vivo T-cell depletion with antithymocyte globulin or ex-vivo T-cell depletion were excluded from the study. All patients had at least 3 years of follow-up after their transplant. All patients provided consent to the use of protected health data for research, as approved by a common institutional review board of the Dana-Farber/Harvard Cancer Center.

Definitions and end-points Neutrophil engraftment was defined as the first day of an absolute neutrophil count >500/mL on three consecutive measurements. Platelet recovery was defined as the first day of two consecutive measurements of >20,000/mL (unsupported). Acute and chronic GVHD were graded by consensus grading criteria and their cumulative incidence was calculated through day +200 after hematopoietic stem cell transplantation. Chronic GVHD was defined clinically by treating physicians utilizing standard criteria ultimately supplanted by NIH Consensus Conference recommendations.11 Relapse was defined as disease recurrence documented by morphological, histological or radiographic means. Treatmentrelated mortality was defined as death while in continuous remission. Progression-free survival was defined as the time from transplantation to relapse or death from any cause, and overall survival was defined as the time from transplantation to death from any cause.

Statistical methods Methods We analyzed data from 705 consecutive patients with hematologic malignancies undergoing RIC utilizing granulocyte colonystimulating factor-mobilized donor peripheral blood mononuclear

For each cell type of interest, cell dose level was classified into groups using quartile cutoffs (Table 1). The Spearman coefficient for TNC versus CD34+ cells was 0.25 (0.18-0.32) (Figure 1). The effects of cell doses on categorical or continuous outcomes were tested using the Fisher exact test or Wilcoxon rank-sum test. Progression-free survival and overall survival were estimated using

Figure 1. Scatter plot of TNC vs. CD34+ cell dose with Spearman coefficient of 0.25 (0.18-0.32).

Table 1. Quartile summaries.

TNC (1010) TNC/kg (108) CD34+ (108) CD34+/kg (106) CD3+ (1010) CD3+/kg (108) 500

Q1, median (range)

Q2, median (range)

Q3, median (range)

Q4, median (range)

5.66 (2.21, 6.79) 6.81 (2.43, 8.39) 3.44 (0.97, 4.41) 4.07 (0.83, 5.4) 1.32 (0.16, 1.72) 1.58 (0.29, 2.09)

7.74 (6.81, 8.74) 9.56 (8.42, 10.78) 5.34 (4.43, 6.3) 6.56 (5.41, 7.83) 2.05 (1.72, 2.38) 2.52 (2.09, 2.96)

9.71 (8.75, 10.89) 12.23 (10.81, 13.96) 7.45 (6.31, 8.73) 9.24 (7.84, 10.81) 2.76 (2.39, 3.26) 3.52 (2.97, 4.1)

12.7 (10.9, 22.31) 16.41 (13.99, 33.94) 11.31 (8.76, 35.57) 14.26 (10.82, 47.67) 3.94 (3.27, 8.3) 5.02 (4.1, 15.77) haematologica | 2016; 101(4)


Impact of TNC and CD34 cell dose on RIC HSCT outcomes

numbers of CD34+, nucleated (TNC), CD3+, CD4+, and CD8+ donor cells infused as well as cell number per kg of recipient body weight. Outcomes are reported with regard to CD34+ cell dose per kg of actual body weight or total TNC dose, as these have been routine reporting metrics in prior analyses for these graft-associated variables. An exploratory analysis was also performed for total CD34+ dose as well as TNC/kg.

the Kaplan-Meier method. Patients alive without a relapse were censored at the date of last contact. Log-rank tests were used to compare progression-free survival and overall survival between groups and Cox proportional hazards regression models were used to study the effects of cell doses on the survival outcomes while adjusting for other risk factors. Time to GVHD onset was analyzed using early relapse and death without GVHD as competing risks. Time to relapse and time to non-relapse mortality were analyzed using each other as competing risks. In competing risk settings, Gray methods were used to compare between groups and competing risk regressions were used to test the significance of cell dose effects adjusting for other covariates. Other factors adjusted for in the regression setting included patients’ age, donors’ age, GVHD prophylaxis, patient-donor sex-match status, and the disease risk index.12 Analyses were conducted on total

Results Patients The patients’ characteristics are listed in Table 2. The vast majority of patients (99%) received RIC with busulfan and fludarabine (IV busulfan total dose of 3.2 or 6.4

Table 2. Patients’ characteristics by quartiles of TNC and CD34/kg.

Q1

Q2

Age , years(median, range) 55.5(20, 72) 57 (20, 69) Age N (%) N (%) < 50 52 (30) 40 (23) ≥50 124 (70) 136 (77) Patients’ sex Male 87 (49) 100 (57) Female 89 (51) 76 (43) Patient-donor sex Male/male 45 (26) 63 (36) Female/male 47 (27) 36 (20) Male/female 42 (24) 37 (21) Female/female 42 (24) 40 (23) Diagnosis AML 48 (27) 46 (26) ALL 6 (3) 8 (5) CML 6 (3) 3 (2) MDS, MPD, MDS/MPD 27 (16) 30 (18) CLL/SLL/PLL 28 (16) 32 (18) NHL 37 (21) 33 (19) HL 13 (7) 13 (7) MM/PCD 9 (5) 10 (6) Anemia/RCD 2 (1) 1 (1) Patients’ CMV status Negative 77 (44) 90 (51) Positive 95 (54) 85 (48) Acute GVHD prophylaxis Tacrolimus +sirolimus 131 (74) 129 (73) Tacrolimus - sirolimus 42 (24) 43 (24) Type of donor Matched related 79 (45) 64 (36) Matched unrelated 86 (49) 102 (58) Mismatched related 2 (1) 1 (1) Mismatched unrelated 9 (5) 9 (5) Risk status Low risk 43 (24) 37 (21) Intermediate risk 89 (51) 90 (51) High risk 39 (22) 44 (25) Very high risk 5 (3) 5 (3) Donor age, years 41.5 (19, 67) 39.5(13, 67) median (range)

TNC Q3

Q4

P value

Q1

57 (19, 73) N (%) 46 (26) 130 (74)

59 (18, 74) N (%) 34 (19) 142 (81)

0.22

57(20, 72) N (%) 43 (24) 133 (76)

58(19, 72) 57(18, 74) N (%) N (%) 36 (20) 45 (26) 141 (80) 131 (74)

118 (67) 58 (33)

129 (73) 47 (27)

117 (66) 59 (34)

111 (63) 66 (37)

71 (40) 40 (23) 47 (27) 18 (10)

80 (45) 28 (16) 49 (28) 19 (11)

63 (36) 22 (13) 54 (31) 37 (21)

56 (32) 6 (3) 7 (4) 28 (16) 30 (17) 34 (19) 7 (4) 8 (5) 0 (0)

51 (29) 4 (2) 6 (3) 32 (18) 22 (13) 34 (19) 16 (9) 8 (5) 1 (1)

81 (46) 93 (53)

92 (52) 80 (45)

127 (72) 47 (27)

136 (77) 39 (22)

52 (30) 103 (59) 1 (1) 20 (11)

78 (44) 82 (47) 0 (0) 16 (9)

44 (25) 86 (49) 41 (23) 5 (3) 38(17, 66)

40 (23) 84 (48) 46 (26) 6 (3) 43(12, 73)

0.14

CD34/kg Q3

Q4

P value

56(19, 73) N (%) 48 (27) 128 (73)

0.64

110 (63) 66 (38)

97 (55) 79 (45)

0.00023

55 (31) 29 (16) 56 (32) 37 (21)

70 (40) 41 (23) 40 (23) 25 (14)

71 (40) 59 (34) 26 (15) 20 (11)

46 (26) 3 (2) 5 (3) 35 (20) 29 (16) 35 (20) 11 (6) 10 (6) 2 (1)

43 (24) 8 (5) 6 (3) 30 (17) 32 (18) 35 (20) 12 (7) 10 (6) 0 (0)

66 (38) 7 (4) 4 (2) 18 (10) 27 (15) 37 (21) 9 (5) 7 (4) 0 (0)

46 (26) 6 (3) 7 (4) 35 (20) 24 (14) 31 (18) 17 (10) 8 (5) 2 (1)

90 (51) 83 (47)

87 (49) 87 (49)

79 (45) 95 (54)

84 (48) 89 (51)

125 (71) 48 (27)

130 (73) 44 (25)

131 (74) 45 (26)

138 (78) 34 (19)

75 (43) 83 (47) 0 (0) 18 (10)

78 (44) 88 (50) 0 (0) 11 (6)

67 (38) 95 (54) 1 (1) 13 (7)

53 (30) 108 (61) 3 (2) 12 (7)

Q2

0.48

0.00001

0.00006

0.0000013

0.89

0.43

0.37

0.46

0.78

0.35

0.021

0.076

0.99

0.022

0.086 44 (25) 92 (52) 37 (21) 3 (2) 41(12, 67)

48 (27) 33 (19) 86 (49) 99 (56) 38 (21) 38 (22) 5 (3) 6 (3) 43(17, 71) 40(16, 73)

39 (22) 73 (41) 57 (32) 7 (4) 37(17, 67)

0.038

AML: acute myeloid leukemia; ALL: acute lymphocytic leukemia; CML: chronic myeloid leukemia; MDS: myelodysplastic syndrome; MPD: myeloproliferative disease; CLL: chronic lymphocytic leukemia; SLL: small lymphocytic leukemia; PLL: prolymphocytic leukemia; NHL: non-Hodgkin lymphoma; HL: Hodgkin lymphoma; MM: multiple myeloma; PCD: plasma cell dyscrasia; RCD: refractory cytopenia with dysplasia; CMV: cytomegalovirus.

haematologica | 2016; 101(4)

501


P.S. Martin et al.

mg/kg; IV fludarabine total dose of 120 mg/m2). GVHD prophylaxis was mostly tacrolimus based, with or without sirolimus (74% versus 24%, respectively). The median age of the patients was 57 years (range, 18-74 years). Sixty-two percent of the patients were male. Acute myeloid leukemia was the most common diagnosis (29%). The patientsâ&#x20AC;&#x2122; disease risk status was categorized as low (23%), intermediate (50%), high (24%) or very high (3%) based on the disease risk index described by Armand et al.12 Transplants were categorized as matched (8/8 highresolution HLA-A, B, C and DRB1-matched) with their respective recipients; matched related donor (39%), matched unrelated donor (53%), or mismatched (7/8 highresolution HLA-A, B, C and DRB1-matched); mismatched related donor (1%), mismatched unrelated donor (8%). During the first year, all patients received acyclovir as prophylaxis against herpes simplex virus/varicella zoster virus infections, and sulfamethoxazole/trimethoprim or atovaquone as prophylaxis against Pneumocystis jirovecii pneumonia. Patients were monitored for cytomegalovirus reactivation during the first 100 days after transplantation, and pre-emptive therapy with valganciclovir was given if viral reactivation occurred. Granulocyte colony-stimulating factor was routinely used in most patients to accelerate engraftment.

Engraftment TNC dose did not influence the onset or durability of engraftment. In contrast, as expected, time to neutrophil

engraftment was significantly shorter for the highest quartile of CD34+ cells infused/kg compared with the lower three quartiles combined in a competing risk regression (P=0.0004). This analysis censored subjects who did not have a nadir absolute neutrophil count below 500/mL.

Graft-versus-host disease Neither TNC nor CD34+ cell dose affected grades II-IV acute GVHD. However, there was a significant association between TNC dose and the development of chronic GVHD. The 1-year cumulative incidence of chronic GVHD across the four TNC dose quartiles (Q1-Q4) were 37%, 44%, 47%, and 52%, respectively (P=0.017). In a regression model evaluating the quartiles, Q1 was associated with decreased risk [hazard ratio (HR)=0.80, P=0.0004] versus Q4. Dichotomized regression analysis was performed by collapsing the top three quartiles and comparing them to Q1. In this dichotomized analysis, Q1 was associated with a decreased risk of chronic GVHD versus Q2+Q3+Q4 (HR=0.71, P=0.0063) (Figure 2). There was no association between CD34+ cell dose and chronic GVHD.

Relapse The 3-year cumulative incidence of relapse across the four TNC dose groups (Q1-Q4) was 54%, 60%, 54%, and 42%, respectively (P=0.017). In a regression model evaluating the quartiles, each of the first three quartiles was associated with increased risk of relapse compared to Q4

Figure 2. Regression analyses of TNC dose. Dichotomized regression analysis of the top three quartiles (collapsed) vs. quartile one, showing increased cumulative risk of chronic GVHD at 1 year with higher TNC dose (HR=0.71, P=0.0063).

A

B

Figure 3. Regression analyses of TNC dose. (A) Dichotomized regression analysis of the first three quartiles (collapsed) vs. quartile four, showing decreased risk of 3-year cumulative relapse with higher TNC dose (HR=0.66, P=0.0012). (B) Dichotomized regression analysis of the first three quartiles (collapsed) vs. quartile four, showing no significant effect of TNC dose on 3-year cumulative incidence of nonrelapse mortality (NRM) (P=0.43).

502

haematologica | 2016; 101(4)


Impact of TNC and CD34 cell dose on RIC HSCT outcomes

(HR=1.41, 1.63, and 1.53; P=0.029, 0.0013, and 0.0068, respectively). In a dichotomized regression analysis performed by collapsing Q1, Q2 and Q3, Q4 was associated with decreased risk of relapse when compared with Q1+Q2+Q3 (HR=0.66, P=0.0012) (Figure 3A). After adjusting for chronic GVHD as a time-dependent covariate in a Cox model, high TNC dose (Q4 versus Q1+Q2+Q3) was still significantly associated with better progression-free survival (P<0.0001) and lower relapse rate (P<0.0001). CD34+ cell dose did not affect relapse rates. The 3-year cumulative incidence of relapse across the four CD34+/kg dose groups (Q1-Q4) was 48%, 51%, 51%, and 59%, respectively (P=0.20). In dichotomized regression analysis of CD34+ cell dose, no significant effect was found by collapsing Q1, Q2 and Q3 and comparing with Q4 (P=0.15), or by collapsing Q2+Q3+Q4 and comparing with Q1 (P=0.13).

survival, we performed an exploratory analysis on a subset of patients in whom the CD3+ (n=445), CD4+ (n=443) and CD8+ (n=444) fractions of the graft were reliably enumerated. No association between CD3+, CD4+ or CD8+ cell count and these outcomes was identified, either because the analysis was underpowered to demonstrate a significant difference or because other specific cellular components rather than total CD3+, CD4+ or CD8+ cell count influence these outcomes.

Total CD34+ cell dose and total nucleated cells per kilogram Exploratory analysis was performed for total CD34+ dose as well as TNC/kg, demonstrating no significant association between either total CD34+ dose or TNC/kg and the aforementioned transplant outcomes.

Discussion Non-relapse mortality There was no association between TNC dose (Figure 3B) or CD34+ cell infusion numbers and the 3-year cumulative incidence of non-relapse mortality.

Progression-free and overall survival The 3-year progression-free survival across the four TNC dose groups (Q1-Q4) was 37%, 32%, 32%, and 46%, respectively (P=0.029), and the 3-year overall survival was 48%, 49%, 40%, and 58% (P=0.050). In a regression model evaluating the quartiles, each of the first three quartiles was associated with worse progressionfree survival compared with Q4 (HR=1.32, 1.48, and 1.64; P=0.046, 0.0034, and 0.0002, respectively), as well as worse overall survival (HR=1.42, 1.31, and 1.62; P=0.020, 0.069, and 0.0009, respectively). In a dichotomized analysis by collapsing Q1, Q2 and Q3, Q4 was associated with better overall survival and progression-free survival (HR=0.69 and 0.68; P=0.0028 and 0.00067, respectively) (Figure 4A,B). CD34+ cell counts bore no association with 3-year progression-free or overall survival.

Graft T-lymphocyte composition In an effort to elucidate which components of the graft could account for the significant impact of TNC on chronic GVHD, relapse, progression-free survival, and overall

A

The majority of studies analyzing the impact of graft composition on major transplant outcomes in RIC HSCT have focused on the influence of CD34+ cell dose, reporting mixed results with regard to survival outcomes, engraftment and incidence of GVHD. In their study of 86 patients who underwent RIC allogeneic HSCT, PerezSimon and colleagues reported an increased incidence of chronic GVHD and decreased relapse rates with increased CD34+/kg (depending on disease state at transplant), but no significant effect on overall survival or event-free survival.1 Pulsipher and colleagues reported on 160 patients undergoing RIC allogeneic HSCT, describing faster engraftment, reduced transplant-related mortality and better 3-year overall survival in the cohort that received a high CD34+ cell dose (>4.5x106 cells/kg) without there being any associated increase in the incidence of acute or chronic GVHD.4 In an analysis of 1054 RIC transplant recipients reported to the CIBMTR from 2002 to 2011, Torlen and colleagues determined that low CD34+ cell dose (<4x106 CD34+ cells/kg) was associated with increased non-relapse mortality and inferior overall survival.7 In a recent analysis, Remberger and colleagues reported inferior outcomes for patients receiving fewer than 2.5x106 CD34+ cells/kg, but also poorer survival due to increased relapse rates in patients receiving over 11x106

B

Figure 4. Dichotomized regression analyses of TNC dose. (A) Dichotomized regression analysis of first three quartiles (collapsed) vs. quartile four, showing improved 3-year progression-free survival (PFS) with higher TNC dose (HR=0.68, P=0.00067). (B) Dichotomized regression analysis of first three quartiles (collapsed) vs. quartile four, showing improved 3year overall survival (OS) with higher TNC dose (HR=0.69, P=0.0028).

haematologica | 2016; 101(4)

503


P.S. Martin et al.

CD34+ cells/kg, albeit in pooled analyses of both myeloablative conditioning and RIC transplant patients.8 These examples typify the myriad of published analyses that have, thus far, failed to demonstrate reproducible, consensus data regarding the effect of CD34+ cell dose on common RIC transplant outcomes.1-5,13-18 In contrast to CD34+ cell dose, there have been only a limited number of studies that report on the effect of TNC dose on RIC transplant outcomes. Baron and colleagues analyzed graft composition data from 125 RIC patients, showing a non-significant, but appreciable trend (HR=0.08, P=0.06) toward improved overall survival, as well as progression-free survival, with higher TNC dose.9 In their study of 253 patients with acute myeloid leukemia undergoing RIC transplantation, Gorin and colleagues found that CD34+/kg cell dose had no effect on either GVHD or overall survival, but that increased TNC dose was independently associated with more chronic GVHD as well as improved overall survival.10 In the aforementioned study by Remberger and colleagues, TNC dose showed no correlation with survival outcomes, relapse rates, engraftment or GVHD. As discussed for the CD34+ cell dose, albeit with the caveat of far fewer published data, the impact of TNC dose on RIC transplant outcomes remains undefined. To our knowledge, our current single institution study of 705 RIC patients is the largest to date to anlayze the effects product-related variables on transplant outcomes in the RIC setting. We found improved survival data (overall survival and progression-free survival) with higher TNC dose, whereas CD34+ cell dose (total and per kg) did not have a significant effect on these critical outcomes. Based on recent studies demonstrating an association between RIC transplant survival outcomes and absolute lymphocyte count recovery19 as well as donor chimerism,20 we analyzed the effect of TNC dose on these additional prognostic indicators, but failed to demonstrate any significant association. Nevertheless, our findings support prior reports9,10 indicating that TNC dose is potentially a more relevant product-related variable than CD34+ cell dose with regard to RIC transplant outcomes. It is unlikely that the top quartile of TNC enumerated in our cohort (>10.9x1010 cells) represents the ideal cell target across transplant center collection facilities despite its association with lower relapses rates and improved survival. Rather, our experience may suggest that “more is better” when it comes to TNC. An improved survival benefit with higher TNC doses should influence those current RIC transplant protocols that cap the TNC infusion dose. Importantly, our analysis showed no significant association between

References 1. Perez-Simon JA, Diez-Campelo M, Martino R, et al. Impact of CD34+ cell dose on the outcome of patients undergoing reducedintensity-conditioning allogeneic peripheral blood stem cell transplantation. Blood. 2003;102(3):1108-1113. 2. Panse JP, Heimfeld S, Guthrie KA, et al. Allogeneic peripheral blood stem cell graft composition affects early T-cell chimaerism

504

TNC and CD34+ cell dose (Spearman correlation coefficient of 0.24), further demonstrating the mutual exclusivity of these product-related parameters. It remains unclear whether there is an identifiable cell subset within the TNC population that drives this observed survival benefit (e.g. dendritic cells, antigen-presenting cells, natural killer cells, T-regulatory cells, T-effector cells), or if such benefit is engendered by a pleiotropic effect of the immunomodulatory milieu that accompanies larger TNC doses. Various studies analyzing graft T-cell composition have not convincingly demonstrated any particular cellular component as reproducibly having either deleterious or beneficial effects on transplant outcomes.2124 In their study of 63 RIC transplant patients, Cao and colleagues reported improved freedom from progression, overall survival and attainment of full donor T-cell chimerism with higher CD8+ cell doses, with no correlation between CD8+ dose and GVHD and no significant effect of CD3+ or CD4+ subsets on any of these outcomes.21 In contrast, Mohty and colleagues22 reported increased acute GVHD with higher CD8+ doses in their analysis of 57 RIC transplant patients and Collins and colleagues23 reported no significant effect of CD3+, CD4+ or CD8+ lymphocyte subsets on overall survival in their review including 70 RIC transplant patients. Reshef and colleagues reported on T-cell subset data in 200 RIC transplant patients, describing improved survival and reduced relapse rates with increased CD8+ cells, with no significant effect of CD3+, CD4+ or CD34+ subsets on these outcomes (of note, TNC dose was not reported in this study).24 While our subset analysis of CD3+, CD4+ and CD8+ cell doses in approximately 445 RIC transplant patients failed to demonstrate any significant association between these T-lymphocyte subsets and our observed transplant outcomes, these studies both suggest that the CD34-negative fraction of the graft is most critical to transplant outcomes after RIC. These findings support more comprehensive analysis of graft subset populations as a compelling direction for future research. Such an analysis could potentially elucidate a cell subset (or synergistic groups of cell subsets) within the TNC population that most directly confers a survival benefit, allowing for the future possibility of improving RIC transplant outcomes through directed graft manipulation. Acknowledgments The authors would like to thank the Ted and Eileen Pasquarello Research Fund at DFCI and the National Institutes of Health (National Cancer Institute grant 5R01-CA183559) for their support of this work.

and later clinical outcomes after non-myeloablative conditioning. Br J Haematol. 2005;128(5):659-667. 3. Mehta J, Frankfurt O, Altman J, et al. Optimizing the CD34+ cell dose for reduced-intensity allogeneic hematopoietic stem cell transplantation. Leuk Lymphoma. 2009;50(9):1434-1441. 4. Pulsipher MA, Chitphakdithai P, Logan BR, et al. Donor, recipient, and transplant characteristics as risk factors after unrelated donor PBSC transplantation: beneficial

effects of higher CD34+ cell dose. Blood. 2009;14(13):2606-2616. 5. Tsirigotis P, Shapira MY, Or R, et al. The number of infused CD34+ cells does not influence the incidence of GVHD or the outcome of allogeneic PBSC transplantation, using reduced-intensity conditioning and antithymocyte globulin. Bone Marrow Transplant. 2010;45(7):1189-1196. 6. Gómez-Almaguer D, Gómez-Peña Á, JaimePérez JC, et al. Higher doses of CD34+ progenitors are associated with improved over-

haematologica | 2016; 101(4)


Impact of TNC and CD34 cell dose on RIC HSCT outcomes

7.

8.

9.

10.

11.

all survival without increasing GVHD in reduced-intensity conditioning allogeneic transplant recipients with clinically advanced disease. J Clin Apher. 2013;28(5): 349-355. Törlén J, Ringdén O, Le Rademacher J, et al. Low CD34 dose is associated with poor survival after reduced-intensity conditioning allogeneic transplantation for acute myeloid leukemia and myelodysplastic syndrome. Biol Blood Marrow Transplant. 2014;20(9): 1418-1425. Remberger M, Törlén J, Ringdén O, et al. The effect of total nucleated and CD34+ cell dose on outcome after allogeneic hematopoietic stem cell transplantation. Biol Blood Marrow Transplant. 2015;21(5): 889893. Baron F, Maris MB, Storer BE, et al. High doses of transplanted CD34+ cells are associated with rapid T-cell engraftment and lessened risk of graft rejection, but not more graft-versus-host disease after nonmyeloablative conditioning and unrelated hematopoietic cell transplantation. Leukemia. 2005;19(5):822-828. Gorin NC, Labopin M, Boiron JM, et al. Results of genoidentical hemopoietic stem cell transplantation with reduced-intensity conditioning for acute myelocytic leukemia: higher doses of stem cells infused benefit patients receiving transplants in second remission or beyond – the Acute Leukemia Working Party of the European Cooperative Group for Blood and Marrow Transplantation. J Clin Oncol. 2006;24(24): 3959-3966. Filipovich AH, Weisdorf D, Pavletic S, et al. National Institutes of Health consensus development project on criteria for clinical trials in chronic graft-versus-host disease: I.

haematologica | 2016; 101(4)

12.

13.

14.

15.

16.

17.

18.

Diagnosis and staging working group report. Biol Blood Marrow Transplant. 2005;11(12): 945-956. Armand P, Gibson CJ, Cutler C, et al. A disease risk index for patients undergoing allogeneic stem cell transplantation. Blood. 2012;120(4):905-913. Zaucha JM, Gooley T, Bensinger WI, et al. CD34 cell dose in granulocyte colony-stimulating factor-mobilized peripheral blood mononuclear cell grafts affects engraftment kinetics and development of extensive chronic graft-versus-host disease after human leukocyte antigen-identical sibling transplantation. Blood. 2001;98(12):32213227. Mohty M, Bilger K, Jourdan E, et al. Higher doses of CD34+ peripheral blood stem cells are associated with increased mortality from chronic graft-versus-host disease after allogeneic HLA-identical sibling transplantation. Leukemia. 2003;17(5):869-875. Ringden O, Barrett AJ, Zhang MJ, et al. Decreased treatment failure in recipients of HLA-identical bone marrow or peripheral blood stem cell transplants with high CD34 cell doses. Br J Haematol. 2003;121(6):874885. Bittencourt H, Rocha V, Chevret S, et al. Association of CD34 cell dose with hematopoietic recovery, infections, and other outcomes after HLA-identical sibling bone marrow transplantation. Blood. 2002;99(8):2726-2733. Singh AK, Savani BN, Albert PS, Barrett AJ. Efficacy of CD34+ stem cell dose in patients undergoing allogeneic peripheral blood stem cell transplantation after total body irradiation. Biol Blood Marrow Transplant. 2007;13(3):339-344. Gorin NC, Labopin M, Reiffers J, et al. Higher

19.

20.

21.

22.

23.

24.

incidence of relapse in patients with acute myelocytic leukemia infused with higher doses of CD34+ cells from leukapheresis products autografted during the first remission. Blood. 2010;116(17):3157-3162. Kim HT, Armand P, Frederick D, et al. Absolute lymphocyte count recovery after allogeneic hematopoietic stem cell transplantation predicts clinical outcome. Biol Blood Marrow Transplant. 2015;21(5):873880. Koreth J, Kim HT, Nikiforow S, et al. Donor chimerism early after reduced-intensity conditioning hematopoietic stem cell transplantation predicts relapse and survival. Biol Blood Marrow Transplant. 2014;20(10): 1516-1521. Cao TM, Shizuru JA, Wong RM, et al. Engraftment and survival following reducedintensity allogeneic peripheral blood hematopoietic cell transplantation is affected by CD8+ T-cell dose. Blood. 2005;105 (6):2300-2306. Mohty M, Bagattini S, Chabannon C, et al. CD8+ T cell dose affects development of acute graft-vs-host disease following reduced-intensity conditioning allogeneic peripheral blood stem cell transplantation. Exp Hematol. 2004;32(11):1097-1102. Collins NH, Gee AP, Durett AG, et al. The effect of the composition of unrelated donor bone marrow and peripheral blood progenitor cell grafts on transplantation outcomes. Biol Blood Marrow Transplant. 2010;16(2): 253-262. Reshef R, Huffman AP, Gao A, et al. High graft CD8 cell dose predicts improved survival and enables better donor selection in allogeneic stem-cell transplantation with reduced-intensity conditioning. J Clin Oncol. 2015;33(21):2392-2398.

505


ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Stem Cell Transplantation

Ferrata Storti Foundation

Haematologica 2016 Volume 101(5):506-514

Multi-state analysis illustrates treatment success after stem cell transplantation for acute myeloid leukemia followed by donor lymphocyte infusion

Matthias Eefting,1* Liesbeth C. de Wreede,2* Constantijn J.M. Halkes,1 Peter A. von dem Borne,1 Sabina Kersting,1 Erik W.A. Marijt,1 Hendrik Veelken,1 Hein Putter,2 Johannes Schetelig,3 and J.H. Frederik Falkenburg1

Department of Hematology, Leiden University Medical Center, the Netherlands; Department of Medical Statistics and Bioinformatics, Leiden University Medical Center, the Netherlands; and 3Medical Department I, University Hospital Carl Gustav Carus, Technische Universität Dresden & DKMS, German Bone Marrow Donor Center, Germany 1 2

*ME and LCdW contributed equally

ABSTRACT

I Correspondence: l.c.de_wreede@lumc.nl

Received: September 23, 2015. Accepted: January 15, 2016. Pre-published: January 22, 2016. doi:10.3324/haematol.2015.136846

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/4/506

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

506

n the field of hematopoietic stem cell transplantation, the common approach is to focus outcome analyses on time to relapse and death, without assessing the impact of post-transplant interventions. We investigated whether a multi-state model would give insight into the events after transplantation in a cohort of patients who were transplanted using a strategy including scheduled donor lymphocyte infusions. Seventy-eight consecutive patients who underwent myeloablative Tcell depleted allogeneic stem cell transplantation for acute myeloid leukemia or myelodysplastic syndrome were studied. We constructed a multi-state model to analyze the impact of donor lymphocyte infusion and graft-versus-host disease on the probabilities of relapse and nonrelapse mortality over time. Based on this model we introduced a new measure for outcome after transplantation which we called ‘treatment success’: being alive without relapse and immunosuppression for graftversus-host disease. All relevant clinical events were implemented into the multi-state model and were denoted treatment success or failure (either transient or permanent). Both relapse and non-relapse mortality were causes of failure of comparable magnitude. Whereas relapse was the dominant cause of failure from the transplantation state, its rate was reduced after graft-versus-host disease, and especially after donor lymphocyte infusion. The long-term probability of treatment success was approximately 40%. This probability was increased after donor lymphocyte infusion. Our multi-state model helps to interpret the impact of post-transplantation interventions and clinical events on failure and treatment success, thus extracting more information from observational data. Introduction In the field of allogeneic stem cell transplantation (SCT), outcome analyses are mainly focused on time to relapse and death, measured from the moment of transplantation.1 This approach, however, does not take into account the increasing availability and potential of post-SCT interventions. These interventions, in particular donor lymphocyte infusions (DLI), are of utmost interest as part of a treatment strategy in which T-cell depleted (TCD-)SCT is followed by T-cell based therapeutic interventions. T-cell depletion of the graft efficiently prevents the development of severe graft-versus-host disease (GvHD) directly after transplantation, but also haematologica | 2016; 101(4)


Multi-state analysis for outcomes after SCT

Table 1. Patient characteristics.

Variable

Categories

Entire cohort (79 patients) n (%)/Median (range)

Gender

male female years AML MDS intermediate poor very poor CR1 CR2 Cyclophosphamide/TBI Busulphan/Cyclophosphamide bone marrow G-CSF stimulated PB sibling unrelated Alemtuzumab CD34-selection 106/kg grade 0-1 grade 2 grade 3-4 0 1 2 3 4 days no yes

41 (52%) 38 (48%) median 46 years (18-61 years) 70 (89%) 9 (11%; median IPSS 2.0 range 1.0-3.0) 12 (17%) 45 (64%) 13 (19%) 66 (84%) 13 (16%) 77 (97%) 2 (3%) 6 (8%) 73 (92%) 52 (66%; 51 10/10 HLA-match) 27 (34%; 17 10/10 HLA-match) 73 (92%) 6 (8%) median 7.0 (1.1-28.3) 53 (68%) 18 (23%) 7 (9%) 42 (53%) 16 (20%) 9 (11%) 9 (11%) 3 (4%) median 216 (91-820) 24 (65%) 13 (35%; 6 required IS)

Age at SCT Initial diagnosis prior to SCT AML risk group at diagnosis (n=70)

Remission status prior to allo SCT Conditioning regimen Stem cell source Donor type Ex vivo T-cell depletion CD34 cells Acute GvHD after SCT of patients who engrafted (n=78)

Number of prophylactic DLIs per patient

Time to first DLI (n=37) GvHD after first DLI (n=37)

SCT: allogeneic stem cell transplantation; IPSS: international prognostic scoring system; CR1: first complete remission; CR2: second complete remission; TBI: total body irradiation; G-CSF: granulocyte-colony stimulating factor; CD34-selection: MACS-sorted CD34+cell collection; DLI: donor lymphocyte infusion; IS: systemic immunosuppression. Intermediate-risk AML was defined as the combination of WBC <100x109/L at diagnosis, absence of cytogenetic abnormalities (except for -X, -Y), achievement of morphological complete remission after the first remission induction course and transplantation in first complete remission (CR1). Poor-risk AML was defined as either WBC >100x109/L at diagnosis, and/or presence of cytogenetic abnormalities (excluding the favorable cytogenetic abnormalities t(15;17), t(8;21) and inv(16)), and/or FLT3/ITD positivity, and/or absence of morphological complete remission after the first remission induction course and/or transplantation not in CR1.Very poor-risk AML was defined as monosomal karyotype, and/or 3q26 rearrangement, and/or EVI-1 expression at diagnosis. One patient (AML, poor-risk) was not included in the multi-state analysis because of graft rejection.

adversely affects post-transplant anti-tumor and antiinfectious immunity.2-4 Early intervention with DLI after TCD-SCT may prevent the relapse of the malignancy and improve immune reconstitution against pathogens, but is associated with the reintroduction of the risk of GvHD.5,6 However, a long-term delay of DLI in patients without clinical signs of GvHD may increase relapse risk in these patients. At present, the optimal timing of DLI has not been established. There is a growing interest in the analysis of treatment outcome measures other than overall survival and relapsefree survival. Although these are crucial endpoints, they do not concentrate on highly relevant intermediate events or interventions essential to evaluate causes of failure or success after transplantation. To assess the dynamic impact of post-transplant clinical events and interventions, the methodology of multi-state models has been developed.7-9 Multi-state models have several advantages with respect to more widely used standard survival analysis haematologica | 2016; 101(4)

methods.10 The primary advantage is that sequences of events -such as SCT, DLI, GvHD, death - and competing events - such as relapse and death before relapse - can be modeled simultaneously. Thus, the models can deal with different post-transplant sequences and timing of events. In contrast to composite survival outcomes such as failurefree survival11 or GvHD-free survival12 where a patientâ&#x20AC;&#x2122;s subsequent events after the first failure are not considered anymore, multi-state models can be used to assess the role of temporary states such as GvHD in relation to additional interventions after transplantation. These models are essential to evaluate more complex treatment strategies like TCD-SCT with differentially scheduled post-transplant cellular immune interventions. Even though this methodology has been available for over a decade thanks to the work of Klein and colleagues, the number of clinical questions in the field of hematology addressed by means of these models has been very limited thus far.8,13-17 This has partly been caused by a lack of famil507


M. Eefting et al.

Figure 1. Multi-state model. The name of each state reflects the event the occurrence of which makes the patient enter the state; he/she remains in this state until a next event occurs or until the end of follow-up. Patients entered a new state on the day when such an event took place and could not visit the same state twice. The starting state of all patients is 1. SCT. The intermediate states are depicted by numbers 2 to 6. Each arrow indicates a possible transition to an intermediate state in the model. Transition between severe GvHD (Start IS) and initiation of DLI was only possible after cessation of systemic immunosuppression. The absorbing states (in red) are Relapse and NRM. States 1, 3, 4, 6 (in green) indicate treatment success; states 2, 5, 7, 8 (in orange and red) indicate treatment failure. Transition to the absorbing states was possible from any other state in the model; for simplicity these transitions are omitted from the figure.

iarity with the method or its potential, and partly by insufficient data quality: a multi-state model requires reliable follow-up information for the events in the model. Another major reason was the lack of easily accessible software. This has changed in the last years, among others by the development of the package ‘mstate’ in R that enables users to analyze general multi-state models.7,9 In the current study we investigated how multi-state models can be applied to complex phenomena including cellular interventions after transplant. We focused on two topics: (1) the impact of DLI and development of GvHD requiring immuno-suppressive treatment (IS) on the failure probabilities of relapse and non-relapse mortality (NRM) over time, and; (2) the probability of treatment success, which is defined as the absence of disease and GvHD requiring IS over time. We demonstrate how a multi-state model helps to interpret the consequences of the therapeutic interventions after SCT in a cohort of patients with acute myeloid leukemia (AML) and myelodysplastic syndrome (MDS) who were transplanted in our center using a TCD-SCT strategy including scheduled post-transplant DLI.

Methods

included in the analysis. Until June 2007 only patients with mixedchimerism but without early relapse of AML or MDS were eligible for prophylactic DLI at 6 months after SCT. If severe GvHD (overall grade II or higher) was present, DLI infusion was postponed. From June 2007 onwards, patients with poor or very poor-risk AML were eligible for low dose prophylactic DLI at 3 months after SCT (see Table 1 for definition of risk groups). For details on transplantation protocol, engraftment, dosing of DLI and assessment of GvHD and mixed-chimerism, see the Online Supplementary Appendix. The study was approved by Leiden University Medical Center Research Ethics Committee. Informed consent was obtained prior to data collection. Data were analyzed as of December 2012.

Definitions of events after SCT For an accurate and objective determination of the duration of severe GvHD, the time interval between the start date of IS (indicated for GvHD) until cessation of IS (stop IS) was taken. In patients with an unrelated donor receiving prophylactic cyclosporine, the date of the start of additional IS, e.g. prednisone, was taken as the starting point of severe GvHD. Relapse after SCT was defined as an increase of blasts in the bone marrow (BM) to ≥5% by morphology; and/or by the presence of >1% blasts in peripheral blood (PB); and/or by the reappearance of molecular and/or cytogenetic AML markers. NRM was defined as death in continuous complete remission after SCT.

Study protocol Seventy-nine patients who underwent myeloablative allogeneic SCT for AML or MDS in complete remission (CR) in our center between January 2002 and June 2011 were included in this analysis. Two other patients who received CD4+-DLI in the context of an experimental CD4+-DLI study during this time period were not 508

Statistical analysis For all analyses, time was measured from the date of SCT. Probabilities of overall survival with associated 95% confidence intervals (95% CI) were calculated by the Kaplan-Meier method. Overall survival was estimated with SPSS/PASW Statistics 20, haematologica | 2016; 101(4)


Multi-state analysis for outcomes after SCT

Figure 2. Overall survival curves. Kaplan Meier curves illustrating probabilities of overall survival of 12 patients with intermediate-risk AML (solid green line), 58 patients with poor-/very poor-risk AML (dashed red line), and 9 patients with MDS (dotted orange line).

release 20.0.0 (2011). The cumulative incidence of prophylactic DLI was estimated in a competing risks framework, considering relapse and NRM before prophylactic DLI as competing events. The cumulative incidence of DLI was estimated by means of the ‘cmprsk’ library in R. All other outcomes were analyzed by means of a multi-state model (see the Online Supplementary Appendix for an explanation of the methodology). The quantities of interest in the multi-state model were estimated in R, version 3.0.1, with library ‘mstate’.7,9

Results Patient cohort Characteristics of patients and transplantation procedures are presented in Table 1. A rejection was observed in one patient (day 43). The median follow-up of surviving patients of the entire cohort was 63 months (range 20-128 months). The cumulative incidence of DLI was 21% at 6 months and 29% at 9 months (see Online Supplementary Figure S2). At the end of follow-up, 37 patients (47%) had received prophylactic DLI (median 2 prophylactic DLIs per patient; range 1-4). Although very poor-risk patients were scheduled for early DLI, in practice the timing of DLI was not very different for those patients compared to the other patients (median time to DLI for the 5 very poor-risk patients: 6.9 months (range 3.2-9.9 months) for the other patients: 7.7 months (3.0-27.0 months)). Forty-two patients (53% of 79 patients) did not receive DLI (very poor-risk AML n=8, poor-risk AML n=24, intermediaterisk AML n=7, MDS n=3). Reasons for not receiving DLI were rejection (n=1), NRM (n=13), early relapse (n=9), severe GvHD (n=8), full-donor hematopoiesis (n=9, including one patient with poor-risk AML who relapsed at day 270), and logistical problems (n=2; these patients with very poor-risk AML who were transplanted before June 2007 showed mixed-chimerism at 6 months after SCT but relapsed before DLI was initiated (at days 216 and 275, respectively)). The estimated 2-year overall survival calculated from haematologica | 2016; 101(4)

the start day of SCT of the entire cohort of 79 patients was 58% (95% CI 48-68%). The two-year overall survival of patients with intermediate-risk AML, poor-/very poor-risk AML, and MDS calculated from the start day of SCT were 67% (95% CI 40%-94%), 55% (95% CI 42-68%), and 67% (95% CI 35%-98%), respectively (see Figure 2).

The multi-state model A multi-state model was constructed to model the occurrence and impact of relevant events after SCT (see Figure 1). Since no patient received prophylactic DLI before IS had been tapered, the transition between ‘start IS’ and ‘DLI’ was omitted from the model. Only one patient in the entire cohort experienced a rejection (at day 43), which excluded transition to the state DLI, and therefore both this patient and the state ‘Rejection’ were not included in the model. Outcomes of the other 78 patients were analyzed in the multi-state model. None of them were lost to follow-up. The number of events in the dataset and the number of patients in each state at the end of their follow-up time are presented in the Online Supplementary Figure S1.

Outcomes of the multi-state model All transition probabilities from SCT during the first 60 months of the follow-up are presented in the Online Supplementary Figure S3. The proportions of patients suffering relapse and NRM are comparable in size. From 11 months after SCT, a substantial proportion of patients (ca. 30%) remained in the ‘DLI’ state, implying they had received DLI, were still alive without relapse and did not require IS for GvHD after DLI. Correspondingly, we observed that only a minority of patients suffered from severe GvHD after DLI (see ‘Start IS after DLI’, Online Supplementary Figure S3) and that the probability to be in the ‘Start IS’ or ‘Start IS after DLI’ state was zero after 33 months, indicating that IS could be tapered successfully in all patients with an episode of severe GvHD. It cannot be excluded that patients likely to develop GvHD post-DLI failed before DLI could have been admin509


M. Eefting et al.

Figure 3. Transition probabilities to all states from SCT. Transition probabilities derived from the multi-state model of Figure 1. At each point in time, the distance between 2 adjacent curves represents the probability of being in the corresponding state, given that the patient was in the ‘SCT’ state at time 0. Online Supplementary Figure S3 shows these same transition probabilities in separate panels. As before, names of states represent the events with which patients enter them. Since all patients started in the ‘SCT’ state, the probability of being in that state was 1 at time 0 and decreased afterward, because patients could only leave this state. For all other states, the probability to have entered this state at time 0 was 0. For the ‘Relapse’ and ‘NRM’ states, which are absorbing states, the probabilities can only increase over time since patients cannot leave these states anymore. For the 5 intermediate states, probabilities increase and decrease over time. The figure shows that already at 1 year postSCT only a very small minority of patients are still in the SCT state. The probability to receive IS is always low, both before and after DLI, and transient in all cases. For relapse-free patients, from ca. 1 year onward, the dominant state is being alive after DLI.

istered. However, several factors are likely to contribute to the reduced development of GvHD after DLI. The timing of DLI is one factor. DLI was not given to patients with active GvHD, but postponed in these patients (n=9). After DLI, new episodes of IS-requiring GvHD were hardly observed (1 out of 9 patients (11%) with previous ISrequiring GvHD, 5 out of 28 (18%) without previous ISrequiring GvHD). In conclusion, postponing DLI appears to alleviate the induction of severe GvHD, even in patients with GvHD prior to infusion of donor lymphocytes. Thus, the selection of patients who were unlikely to develop GvHD does not appear to have played a major role. Finally, the small probability of remaining in the state SCT in the long run indicates that follow-up without experiencing a further event was highly uncommon, which is in line with the strategy of scheduled prophylactic post-transplant DLI. The wide confidence intervals reflect the small size of the dataset and show that all estimates should be interpreted with caution. Figure 3 gives an overview of the outcomes of the multistate model, by combining all individual transition probabilities as represented in the Online Supplementary Figure S3 in one figure. The area below the relapse curve of this figure represents relapse-free survival and indicates how the patients alive and relapse-free were distributed over the different states at different moments during follow-up.

time-point 7.5 months after SCT was chosen as the start time of these analyses because prophylactic DLI was scheduled at 6 months after SCT, after which an effective immune response is generally seen 5-6 weeks later, adding up to 7.5 months.18 According to the model, patients who had not experienced severe GvHD and had not yet received prophylactic DLI within 7.5 months after SCT (in state ‘SCT’ at 7.5 months) had a predicted probability of relapse of 0.26 (95% CI: 0.07-0.45) within the first 2 years after transplantation. In contrast, patients in the state ‘DLI’ at the same time point had a predicted probability of relapse of 0.07 (95% CI: 0.00-0.21) within 2 years after transplantation. These results suggest that the risk of relapse, which is the major reason for failure from the ‘SCT’ state, is reduced after GvHD, and especially after DLI administration. In patients without any event, and also in patients in the ‘DLI’ state at 7.5 months after SCT, NRM did not play a prominent role during later follow-up. These differences in outcomes cannot be explained by relevant differences in the baseline risk characteristics of the patients in the respective states at 7.5 months, since (very) poor-risk patients were not underrepresented in the ‘DLI’ state compared to the other states (‘SCT’ state: 55% poor AML, 15% very poor AML; ‘Start IS’ state: 45% poor, 9% very poor; ‘DLI’ state: 58% poor, 17% very poor).

Treatment success Predictions from 7.5 months after SCT The multi-state model enables the update of predictions of subsequent failures when information about post-SCT events accumulates over time. This approach can be considered as an extension of landmark modeling in which both start and end time, and starting and end state are variable. Figure 4 shows transition probabilities for patients in different starting states (1 to 4) at 7.5 months after SCT. The 510

The same multi-state model can be used to summarize the outcomes of the transplantation procedure by focusing on the probability of treatment success at different time points after transplantation (see Figure 5). In general, the goal of our treatment strategy was to administer DLI safely to all patients at risk for relapse. Therefore, at each time point, treatment success included patients who experienced relapse-free survival, and did not suffer from severe haematologica | 2016; 101(4)


Multi-state analysis for outcomes after SCT

Starting from SCT (state 1) at 7.5 months

Starting from Start IS (state 2) at 7.5 months

Starting from Stop IS (state 3) at 7.5 months

Starting from DLI (state 4) at 7.5 months

Figure 4. Predictions from 7.5 months after SCT. Estimated probabilities of relapse (green line) and NRM (black line) over time for patients present at 7.5 months since transplantation in states ‘SCT’, ‘Start IS’, ‘Stop IS’, and ‘DLI’, respectively. Results from ‘Start IS after DLI’ and ‘Stop IS after DLI’ are not shown because of the small sample size.

GvHD at that time point, i.e. patients remaining in the state ‘SCT’, or present in the states ‘Stop IS’, ‘DLI’, or ‘Stop IS after DLI’. The probability of treatment success over time was calculated by adding up the estimated timedependent probabilities of being in these states. The longterm estimated probability of treatment success from start was approximately 0.43 (Figure 5A). If a patient survived the first 7.5 months after SCT without subsequent events, this success probability substantially improved to approximately 0.60. Patients who required IS for GvHD at 7.5 months after SCT (Figure 5B) showed a similar success probability of approximately 0.55 at 60 months after SCT. By definition, these patients were not considered to experience treatment success at 7.5 months after SCT, but a part of them acquired treatment success after cessation of IS possibly followed by DLI, corresponding to an increasing curve. Patients who required IS after DLI or failing from NRM or relapse led the curve to decline again. Finally, a comparison with the prospects for patients in the ‘DLI’ state at the same point in time (Figure 5C) suggests that their probability of treatment success was higher: 0.70 at 60 months after SCT.

Discussion This detailed study of a group of AML/MDS patients with a long, complex follow-up illustrates the potential of haematologica | 2016; 101(4)

a multi-state model to assess the risk of relapse and NRM for patients in a TCD-SCT population for which an infusion of donor lymphocytes was an essential intervention within the strategy. The model separates fatal and nonfatal events, thus enabling the summing up of several favorable episodes to estimate the probability of treatment success. The model also enables one to give longterm prognoses adjusted during follow-up which is relevant for clinical decision making. We previously reported the feasibility of a TCD-SCT with sequential DLI strategy in a population of patients with acute lymphoblastic leukemia.19 In the present study, the majority of patients had poor or very poor-risk AML prior to SCT, and thus the estimated 2-year overall survival of 58% of the entire cohort appears favorable. A substantial proportion of patients receiving DLI did not experience an adverse event during later follow-up. This implies that few patients relapsed after DLI, and also that the development of severe GvHD after DLI was rare. These results cannot be explained by a selection for DLI of patients with favorable-risk characteristics at SCT as a comparison of the patients in different states at 7.5 months shows. Predictions from 7.5 months after SCT show that patients who had not experienced any event (i.e., no severe GvHD, DLI or relapse) so far had the highest probability of relapse, whereas NRM did not play a prominent role in these patients. According to these predictions, patients in the ‘DLI’ state at 7.5 months after SCT 511


M. Eefting et al.

A

C

B

Figure 5. Probability of treatment success over time. Treatment success as calculated in the multi-state model of Figure 1 is defined as the probability of being alive without disease or severe GvHD, i.e., the sum of the probabilities of being in either of the states 1, 3, 4 and 6. (A) probability of treatment success for a patient in the ‘SCT’ state at 0 and 7.5 months after SCT, respectively. (B) probability of treatment success for a patient in the ‘Start IS’ state at 7.5 months after SCT. (C) probability of treatment success for a patient in the ‘DLI’ state at 7.5 months after SCT. Outcomes from ‘Start IS’ and ‘DLI’ at 0 months have been omitted since these do not represent a clinically meaningful situation. The curves from ‘SCT’ and ‘DLI’ start at probability 1 since a patient in one of these states by definition experiences treatment success at the moment when the clock starts; on the contrary, the curve from ‘Start IS’ begins at 0 because a patient in that state first has to make a transition to a beneficial state before being considered a treatment success. The steep rise in this curve can be interpreted as a sign of successful quick cessation of IS in the majority of patients who require IS for the treatment of early GvHD. The treatment success probabilities increase and decrease over time depending on the proportions of patients entering and leaving the beneficial states.

had a relatively low relapse risk and NRM, leading to the highest probability of treatment success. Finally, most patients never needed IS, and IS was stopped during later follow-up in the majority of patients who required IS for GvHD at 7.5 months after SCT. The results show the favorable outcomes of a strategy of TCD-SCT with DLI. The vast majority of patients alive and relapse-free were also free from IS. After 1 year, by far the largest part of this group consisted of patients who had received a DLI. Although the decision to administer DLI may be driven by many clinical events, and although the variable timing of the DLI itself is associated with different disease histories hampering causal interpretation, the results show that ultimately being in the DLI state is beneficial. From these different observations it can be interpreted that the prevention of severe GvHD by T-cell depletion contributes to the reduction of NRM, and that the absence of significant GvHD leads to an increased relapse rate that may be decreased by DLI. Apparently, the low probability of relapse after DLI was not outbalanced by the induction of severe GvHD, which is supported by several other studies.20-22 In conclusion, treatment of patients with absent or limited GvHD after TCD-SCT can safely be consolidated by DLI, which appears to reduce the relapse rate. The observation that the relapse risk was decreased after DLI without a high probability of severe GvHD can be explained by a sufficiently long time interval between the myeloablative conditioning regimen and infusion of DLI, allowing the recipient to recover from conditioninginduced inflammation, thus reducing pro-inflammatory 512

conditions at the time of DLI, without fully extinguishing the beneficial effect of the anti-recipient hematopoietic cell directed T-cells.6,21,23-25 Timely infusion of donor lymphocytes will remain an important factor, because the delay of DLI will increase the time to a favorable immune response and is therefore associated with an increased relapse risk.26-28 Our treatment strategy was not part of a randomized trial, but we aimed to administer DLI to all patients with a high-risk of relapse, e.g. patients with high-risk AML or patients with persisting mixed-chimerism, and to all highrisk patients from June 2007 onwards, whose risk may be modulated by DLI.5,22,23,29 The different DLI strategies for different risk groups make the role of timing of DLI even more difficult to assess and complicates the comparison of outcomes of patients with or without DLI. These issues can only be addressed properly in prospective randomized trials.30,31 However, in current transplant practice, randomized trials are performed infrequently due to both the relatively small numbers of patients affected and to patient and physician treatment preferences.32 This lack of randomized trials necessitates alternative analysis methods, in which as many insights as possible are extracted from observational data.17 For instance, a comparison of treatment strategies could be performed by analyzing the data from different centers by the same model. Multi-state models such as the example presented herein offer several advantages over traditional statistical approaches. The models are more flexible than Cox models with time-dependent covariates since there is no proportional hazards assumption considering hazard ratios as haematologica | 2016; 101(4)


Multi-state analysis for outcomes after SCT

constant over time. They are more comprehensive than landmark models where the particular choice of the landmark time point leads to restrictions. Moreover, landmark analyses do not take into account the probability of arriving in the state from where the prediction is made and are not suitable for analyzing sequences of events after the landmark time point. In addition, multi-state models do not only offer insight into the incidence of certain events as yielded by competing risks methods, but also into the probability of remaining in a certain condition over time, thus offering the option to consider new treatment outcomes. With our study, we illustrated the feasibility of assessing outcomes of a treatment strategy by means of a multi-state model.33 Multi-state models have been advocated before as a means to model alternative outcomes after SCT, e.g. current leukemia-free survival, yet these models focused on fewer events and on outcomes since SCT, not using the full potential of updated prediction.8,14,15,34 Other similar models can be constructed to assess several post-SCT interventions or events, such as the timing of CMV reactivation in relationship to the development of GvHD, also in the context of T-cell replete transplantations. Multi-state models can also help to extend and refine new outcomes such as GvHD-free, relapse-free survival after transplant, or fail-

References 1. Iacobelli S, Committee ES. Suggestions on the use of statistical methodologies in studies of the European Group for Blood and Marrow Transplantation. Bone Marrow Transplant. 2013;48 Suppl 1:S1-37. 2. Barge RM, Starrenburg CW, Falkenburg JH, Fibbe WE, Marijt EW, Willemze R. Longterm follow-up of myeloablative allogeneic stem cell transplantation using Campath "in the bag" as T-cell depletion: the Leiden experience. Bone Marrow Transplant. 2006;37 (12):1129-1134. 3. Marmont AM, Horowitz MM, Gale RP, et al. T-cell depletion of HLA-identical transplants in leukemia. Blood. 1991;78(8):21202130. 4. Chakrabarti S, Mackinnon S, Chopra R, et al. High incidence of cytomegalovirus infection after nonmyeloablative stem cell transplantation: potential role of Campath-1H in delaying immune reconstitution. Blood. 2002;99(12):4357-4363. 5. Peggs KS, Thomson K, Hart DP, et al. Doseescalated donor lymphocyte infusions following reduced intensity transplantation: toxicity, chimerism, and disease responses. Blood. 2004;103(4):1548-1556. 6. Stevanovic S, van Bergen CA, van Luxemburg-Heijs SA, et al. HLA class II upregulation during viral infection leads to HLA-DP-directed graft-versus-host disease after CD4+ donor lymphocyte infusion. Blood. 2013;122(11):1963-1973. 7. de Wreede LC, Fiocco M, Putter H. The mstate package for estimation and prediction in non- and semi-parametric multi-state and competing risks models. Comput Methods Programs Biomed. 2010;99(3):261274. 8. Klein JP, Szydlo RM, Craddock C, Goldman

haematologica | 2016; 101(4)

9.

10. 11.

12.

13.

14.

15.

16.

17.

ure-free survival after the initial systemic treatment of chronic graft-versus-host disease.11,35 In conclusion, our treatment strategy to infuse donor lymphocytes in patients without severe GvHD after myeloablative TCD-SCT did not lead to an increase of severe GvHD after DLI, and appeared to decrease relapse risk. Therefore, the probability of treatment success of these patients was relatively high. In the absence of randomized trials, large cohorts with detailed follow-up are required to further specify the role of prophylactic DLI in patients at risk for relapse. In general, multi-state models can increase insight into complex clinical situations by predicting outcomes of patients with different disease states which may change over time. The current study shows the potential of multi-state models in the analysis of new clinically meaningful outcomes. Funding This work was supported by research grant 2008-4263 from the Dutch Cancer Society, Amsterdam, the Netherlands. Acknowledgments The authors would like to thank Illyea Hawke (DKMS, German Bone Marrow Donor Center, Dresden, Germany) for correcting the language of the manuscript.

JM. Estimation of current leukaemia-free survival following donor lymphocyte infusion therapy for patients with leukaemia who relapse after allografting: application of a multistate model. Stat Med. 2000;19(21):3005-3016. de Wreede LC, Fiocco M, Putter H. mstate: An R Package for the Analysis of Competing Risks and Multi-State Models. J Stat Softw. 2011;38(7):1-30. Putter H, Fiocco M, Geskus R. Tutorial in biostatistics: competing risks and multi-state models. Stat Med. 2007;26(11):2389-2430. Inamoto Y, Flowers ME, Sandmaier BM, et al. Failure-free survival after initial systemic treatment of chronic graft-versus-host disease. Blood. 2014;124(8):1363-1371. Bolanos-Meade J, Logan BR, Alousi AM, et al. Phase 3 clinical trial of steroids/mycophenolate mofetil vs steroids/placebo as therapy for acute GVHD: BMT CTN 0802. Blood. 2014;124(22):3221-3227. Logan BR. Review of multistate models in hematopoietic cell transplantation studies. Biol Blood Marrow Transplant. 2013;19(1 Suppl):S84-87. Schmoor C, Schumacher M, Finke J, Beyersmann J. Competing risks and multistate models. Clin Cancer Res. 2013;19(1):12-21. Klein JP, Keiding N, Shu Y, Szydlo RM, Goldman JM. Summary curves for patients transplanted for chronic myeloid leukaemia salvaged by a donor lymphocyte infusion: the current leukaemia-free survival curve. Br J Haematol. 2000;109(1):148-152. Craddock C, Szydlo RM, Klein JP, et al. Estimating leukemia-free survival after allografting for chronic myeloid leukemia: a new method that takes into account patients who relapse and are restored to complete remission. Blood. 2000;96(1):86-90. Brand R, Putter H, van Biezen A, et al.

18.

19.

20.

21. 22.

23.

24.

Comparison of allogeneic stem cell transplantation and non-transplant approaches in elderly patients with advanced myelodysplastic syndrome: optimal statistical approaches and a critical appraisal of clinical results using non-randomized data. PLoS One. 2013;8(10):e74368. Kolb HJ, Schattenberg A, Goldman JM, et al. Graft-versus-leukemia effect of donor lymphocyte transfusions in marrow grafted patients. Blood. 1995;86(5):2041-2050. Eefting M, Halkes CJ, de Wreede LC, et al. Myeloablative T cell-depleted alloSCT with early sequential prophylactic donor lymphocyte infusion is an efficient and safe postremission treatment for adult ALL. Bone Marrow Transplant. 2014;49(2):287-291. El-Cheikh J, Crocchiolo R, Furst S, et al. Donor CD3(+) lymphocyte infusion after reduced intensity conditioning allogeneic stem cell transplantation: single-center experience. Exp Hematol. 2013;41(1):17-27. Mackinnon S. Donor leukocyte infusions. Baillieres Clin Haematol. 1997;10(2):357367. Krishnamurthy P, Potter VT, Barber LD, et al. Outcome of donor lymphocyte infusion after T cell-depleted allogeneic hematopoietic stem cell transplantation for acute myelogenous leukemia and myelodysplastic syndromes. Biol Blood Marrow Transplant. 2013;19(4):562-568. Mapara MY, Kim YM, Wang SP, Bronson R, Sachs DH, Sykes M. Donor lymphocyte infusions mediate superior graft-versusleukemia effects in mixed compared to fully allogeneic chimeras: a critical role for host antigen-presenting cells. Blood. 2002;100(5):1903-1909. Peggs KS, Mackinnon S. Exploiting graft-versus-tumour responses using donor leukocyte infusions. Best Pract Res Clin Haematol. 2001;14(4):723-739.

513


M. Eefting et al. 25. Kolb HJ. Graft-versus-leukemia effects of transplantation and donor lymphocytes. Blood. 2008;112(12):4371-4383. 26. Breems DA, Van Putten WL, Huijgens PC, et al. Prognostic index for adult patients with acute myeloid leukemia in first relapse. J Clin Oncol. 2005;23(9):1969-1978. 27. Schmid C, Labopin M, Nagler A, et al. Donor lymphocyte infusion in the treatment of first hematological relapse after allogeneic stem-cell transplantation in adults with acute myeloid leukemia: a retrospective risk factors analysis and comparison with other strategies by the EBMT Acute Leukemia Working Party. J Clin Oncol. 2007;25(31): 4938-4945. 28. Schmid C, Labopin M, Nagler A, et al. Treatment, risk factors, and outcome of adults with relapsed AML after reduced

514

29.

30.

31. 32.

intensity conditioning for allogeneic stem cell transplantation. Blood. 2012;119(6): 1599-1606. Shaw BE, Byrne JL, Das-Gupta E, Carter GI, Russell NH. The impact of chimerism patterns and predonor leukocyte infusion lymphopenia on survival following T celldepleted reduced intensity conditioned transplants. Biol Blood Marrow Transplant. 2007;13(5):550-559. Gordon D, Taddei-Peters W, Mascette A, Antman M, Kaufmann PG, Lauer MS. Publication of trials funded by the National Heart, Lung, and Blood Institute. N Engl J Med. 2013;369(20):1926-1934. Devereaux PJ, Yusuf S. When it comes to trials, do we get what we pay for? N Engl J Med. 2013;369(20):1962-1963. Cutler CS, Lee SJ, Greenberg P, et al. A deci-

sion analysis of allogeneic bone marrow transplantation for the myelodysplastic syndromes: delayed transplantation for low-risk myelodysplasia is associated with improved outcome. Blood. 2004;104(2):579-585. 33. Klein JP, Shu Y. Multi-state models for bone marrow transplantation studies. Stat Methods Med Res. 2002;11(2):117-139. 34. Socie G, Schmoor C, Bethge WA, et al. Chronic graft-versus-host disease: long-term results from a randomized trial on graft-versus-host disease prophylaxis with or without anti-T-cell globulin ATG-Fresenius. Blood. 2011;117(23):6375-6382. 35. 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.

haematologica | 2016; 101(4)


Haematologica, Volume 101, issue 4