Haematologica, Volume 104, Issue 6

haematologica Journal of the Ferrata Storti Foundation

Editor-in-Chief Luca Malcovati (Pavia)

Managing Director Antonio Majocchi (Pavia)

Associate Editors Omar I. Abdel-Wahab (New York), Hélène Cavé (Paris), Simon Mendez-Ferrer (Cambridge), Pavan Reddy (Ann Arbor), Andreas Rosenwald (Wuerzburg), Monika Engelhardt (Freiburg), Davide Rossi (Bellinzona), Jacob Rowe (Haifa, Jerusalem), Wyndham Wilson (Bethesda), Paul Kyrle (Vienna), Swee Lay Thein (Bethesda), Pieter Sonneveld (Rotterdam)

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

Editorial Board 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); Giampaolo Merlini (Pavia); Anna Rita Migliaccio (New York); Mohamad Mohty (Nantes); Martina Muckenthaler (Heidelberg); Ann Mullally (Boston); Stephen Mulligan (Sydney); German Ott (Stuttgart); Jakob Passweg (Basel); Melanie Percy (Ireland); Rob Pieters (Utrecht); Stefano Pileri (Milan); Miguel Piris (Madrid); Andreas Reiter (Mannheim); Jose-Maria Ribera (Barcelona); Stefano Rivella (New York); Francesco Rodeghiero (Vicenza); Richard Rosenquist (Uppsala); Simon Rule (Plymouth); Claudia Scholl (Heidelberg); Martin Schrappe (Kiel); Radek C. Skoda (Basel); Gérard Socié (Paris); Kostas Stamatopoulos (Thessaloniki); David P. Steensma (Rochester); Martin H. Steinberg (Boston); Ali Taher (Beirut); Evangelos Terpos (Athens); Takanori Teshima (Sapporo); Pieter Van Vlierberghe (Gent); Alessandro M. Vannucchi (Firenze); George Vassiliou (Cambridge); Edo Vellenga (Groningen); Umberto Vitolo (Torino); Guenter Weiss (Innsbruck).

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

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

haematologica Journal of the 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 2019 are as following: Print edition

Institutional Euro 700

Personal Euro 170

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: Press Up, zona Via Cassia Km 36, 300 Zona Ind.le Settevene - 01036 Nepi (VT)

haematologica Journal of the Ferrata Storti Foundation

Table of Contents Volume 104, Issue 6: June 2019 Cover Figure Diffuse bone marrow infiltration by large pleomorphic cells in a young girl with ALK-positive anaplastic large cell lymphoma. Courtesy of Prof. Rosangela Invernizzi.

Editorials 1093

To induce or not to induce: the fight over hepcidin regulation Veena Sangkhae and Elizabeta Nemeth

1096

The -7 chromosomal abnormalities with signs of myelodysplasia in chronic myeloid leukemia as a major red signal Emilie Cayssials and François Guilhot

1098

Another piece of the puzzle added to understand t(4;11) leukemia better Rolf Marschalek

1101

Conditioning intensity and antilymphocyte globulin: towards personalized transplant strategies?

Martin Bornhäuser

1103

Is the mysterious platelet receptor GPV an unsuspected major target for platelet autoantibodies?

Paquita Nurden and Alan T Nurden

Perspective Article 1106

The carrier state for sickle cell disease is not completely harmless Julia Zhe Xu and Swee Lay Thein

Review Articles 1112

Thrombopoietin receptor agonists: ten years later Waleed Ghanima et al.

1124

Rituximab in the treatment of immune thrombocytopenia: what is the role of this agent in 2019?

Elisa Lucchini et al.

Articles Hematopoiesis

1136

CD150high CD4 T cells and CD150high regulatory T cells regulate hematopoietic stem cell quiescence via CD73 Yuichi Hirata et al.

Iron Metabolism & its Disorders

1143

The opposing effects of acute inflammation and iron deficiency anemia on serum hepcidin and iron absorption in young women Nicole U. Stoffel et al.

Chronic Myeloid Leukemia

1150

Poor prognosis of chromosome 7 clonal aberrations in Philadelphia-negative metaphases and relevance of potential underlying myelodysplastic features in chronic myeloid leukemia Audrey Bidet et al.

Acute Myeloid Leukemia

1156

Lysine specific demethylase 1 inactivation enhances differentiation and promotes cytotoxic response when combined with all-trans retinoic acid in acute myeloid leukemia across subtypes Kimberly N. Smitheman et al

Haematologica 2019; vol. 104 no. 6 - June 2019 http://www.haematologica.org/

haematologica Journal of the Ferrata Storti Foundation Acute Myeloid Leukemia

1168

Impact of induction regimen and allogeneic hematopoietic cell transplantation on outcome in younger adults with acute myeloid leukemia with a monosomal karyotype FrĂŠdĂŠric Baron et al.

Acute Lymphoblastic Leukemia

1176

Unraveling the cellular origin and clinical prognostic markers of infant B-cell acute lymphoblastic leukemia using genome-wide analysis Antonio Agraz-Doblas et al.

1189

Enhanced hemato-endothelial specification during human embryonic differentiation through developmental cooperation between AF4-MLL and MLL-AF4 fusions Clara Bueno et al.

Non-Hodgkin Lymphoma

1202

Association of early disease progression and very poor survival in the GALLIUM study in follicular lymphoma: benefit of obinutuzumab in reducing the rate of early progression John F. Seymour et al.

Plasma Cell Disorders

1209

Proteolysis targeting chimeric molecules as therapy for multiple myeloma: efficacy, biomarker and drug combinations Su Lin Lim et al.

Stem Cell Transplantation

1221

Total body irradiation dose escalation decreases risk of progression and graft rejection after hematopoietic cell transplantation for myelodysplastic syndromes or myeloproliferative neoplasms Federico Monaco et al.

1230

Antilymphocyte globulin for matched sibling donor transplantation in patients with myelofibrosis Marie Robin et al.

Platelet Biology & its Disorders

1237

Glycoprotein V is a relevant immune target in patients with immune thrombocytopenia Richard Vollenberg et al.

1244

Downregulation of TREM-like transcript-1 and collagen receptor a2 subunit, two novel RUNX1-targets, contributes to platelet dysfunction in familial platelet disorder with predisposition to acute myelogenous leukemia Ana C. Glembotsky et al.

1256

High-throughput elucidation of thrombus formation reveals sources of platelet function variability Johanna P. van Geffen et al.

Hemostasis

1268

Generation of anti-idiotypic antibodies to detect anti-spacer antibody idiotopes in acute thrombotic thrombocytopenic purpura patients An-Sofie Schelpe et al.

Coagulation & its Disorders

1277

The Khorana score for prediction of venous thromboembolism in cancer patients: a systematic review and meta-analysis Frits I. Mulder et al.

Haematologica 2019; vol. 104 no. 6 - June 2019 http://www.haematologica.org/

haematologica Journal of the Ferrata Storti Foundation

Letters to the Editor Letters are available online only at www.haematologica.org/content/104/6.toc

e230

Hypoxia attenuates inflammation-induced hepcidin synthesis during experimental human endotoxemia Dorien Kiers et al. http://www.haematologica.org/content/104/6/e230

e233

Kinase activity is impaired in neutrophils of sepsis patients Arie J. Hoogendijk et al. http://www.haematologica.org/content/104/6/e233

e236

Clinical correlates, prognostic impact and survival outcomes in chronic myelomonocytic leukemia patients with the JAK2V617F mutation Mrinal M. Patnaik et al. http://www.haematologica.org/content/104/6/e236

e240

FLT3 ligand plasma levels in acute myeloid leukemia Pierre Peterlin et al. http://www.haematologica.org/content/104/6/e240

e244

Durable remissions in TCF3-HLF positive acute lymphoblastic leukemia with blinatumomab and stem cell transplantation Brice Mouttet et al. http://www.haematologica.org/content/104/6/e244

e248

False-negative rates for MYC fluorescence in situ hybridization probes in B-cell neoplasms Rebecca L. King et al. http://www.haematologica.org/content/104/6/e248

e252

Convergence of risk prediction models in follicular lymphoma Anjali Silva et al. http://www.haematologica.org/content/104/6/e252

e256

New molecular and therapeutic insights into canine diffuse large B-cell lymphoma elucidates the role of the dog as a model for human disease Luca Aresu et al. http://www.haematologica.org/content/104/6/e256

e260

Inherited missense variants that affect GFI1B function do not necessarily cause bleeding diatheses Rinske van Oorschot et al. http://www.haematologica.org/content/104/6/e260

e265

Poor performance of D-dimer in excluding venous thromboembolism among patients with lymphoma and leukemia Aiham Qdaisat et al. http://www.haematologica.org/content/104/6/e265

Case Reports Case Reports are available online only at www.haematologica.org/content/104/6.toc

e269

Role of Epstein-Barr virus in transformation of follicular lymphoma to diffuse large B-cell lymphoma: a case report and review of the literature Massimo Granai et al. http://www.haematologica.org/content/104/6/e269

e274

Hereditary platelet function disorder from RASGRP2 gene mutations encoding CalDAG-GEFI identified by whole-exome sequencing in a Korean woman with severe bleeding Jae Won Yun et al. http://www.haematologica.org/content/104/6/e274

e277

Transfer of ADAMTS13 antibody-mediated thrombotic thrombocytopenic purpura via kidney transplantation Lara Zafrani et al. http://www.haematologica.org/content/104/6/e277

Haematologica 2019; vol. 104 no. 6 - June 2019 http://www.haematologica.org/

haematologica Journal of the Ferrata Storti Foundation

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

Ancient Greek

Scientific Latin

Scientific Latin

Modern English

aÂma [haima] = blood a·matow [haimatos] = of blood lÒgow [logos]= reasoning haematologicus (adjective) = related to blood haematologica (adjective, plural and neuter, used as a noun) = hematological subjects The oldest hematology journal, publishing the newest research results.

2017 JCR impact factor = 9.090

EDITORIALS To induce or not to induce: the fight over hepcidin regulation Veena Sangkhae and Elizabeta Nemeth Center for Iron Disorders, David Geffen School of Medicine, University of California, Los Angeles, CA, USA E-mail: NEMETH ELIZABETA - enemeth@mednet.ucla.edu doi:10.3324/haematol.2019.216960

S

ystemic iron homeostasis is co-ordinated by the hepatic hormone hepcidin.1 Hepcidin inhibits iron export through the cellular iron transporter ferroportin, thereby preventing iron absorption and the release of recycled or stored iron into plasma, resulting in decreased plasma iron levels.2 Hepcidin production changes rapidly and over a large dynamic range to ensure the maintenance of iron homeostasis. Hepcidin is suppressed in conditions that require increased iron supply, such as stress erythropoiesis, hypoxia, growth and pregnancy.3,4 Conversely, hepcidin is induced by iron loading to prevent the accumulation of excess iron, or by inflammation as part of the host defense response to infection.5,6 Although the regulation of hepcidin by singular stimuli has been well studied, particularly in animal models, we still do not have an understanding of the complex interplay of opposing signals in regulating hepcidin expression and iron homeostasis in humans.

In this issue of Haematologica, Stoffel et al. report on a prospective study in young women to evaluate the relative contribution of iron deficiency anemia and acute inflammatory stimulus on iron homeostasis7 (Figure 1). The well-controlled study included a total of 46 women: 25 non-anemic and 21 with iron-deficiency anemia (IDA). Compared to their non-anemic counterparts, anemic women had 2 g/dL lower hemoglobin, lower serum iron, transferrin saturation, ferritin and body iron stores, and higher erythropoietin and serum transferrin receptor. The study excluded subjects with confounding factors affecting iron metabolism, including pre-existing inflammation, chronic disease, obesity, pregnancy, or vitamin and/or mineral supplementation for two weeks prior to and during the study. An acute inflammatory stimulus was modeled using an intramuscular injection of an influenza/diphtheria-tetanus-pertussis vaccine in all subjects.

Inflammatory and iron markers were measured at baseline and 8, 24 and 36 hours (h) post vaccination.

The subjects also received test meals containing 57Fe (nonradioactive isotope of iron), that allowed assessment of erythrocyte 57Fe incorporation as a measure of iron absorption.

The first 57Fe meal and the first erythrocyte 57Fe measurement were completed before the inflammatory stimulus (“baseline”). The second 57Fe meal was administered 24 h after the vaccine, at the time of maximal or near-maximal IL6 and hepcidin increase, followed by the second erythrocyte 57 Fe measurement [“post-vaccine” (Figure 1)]. Although the erythrocyte 57Fe measurements were performed 19 days after each ingestion of 57Fe, they should closely reflect iron absorption on the day of the meal consumption for the following reasons. In humans who are not iron-loaded, the majority of the absorbed iron is loaded onto transferrin and is destined for erythropoiesis: ferrokinetics experiments showed that following the ingestion of 59Fe, approximately 82-91% of absorbed radiolabeled iron is detected in erythrocytes after two weeks.8,9 Furthermore, erythrocyte lifespan is around 120 days, much longer than the duration of the haematologica | 2019; 104(6)

Stoffel et al. study. Thus, any confounding effect of 57Fe-red blood cell recycling and hepcidin modulation of the recycled iron flows would have been minimal.

Administration of the vaccine induced systemic inflammation in both cohorts of women, as reflected by an increase in interleukin-6 (IL-6), a major regulator of hepcidin production. Despite this, there was a surprising difference in the hepcidin response. Serum hepcidin increased in the non-anemic group within 24 h after vaccination but was unchanged in the IDA group. IL-6 and hepcidin significantly correlated at 24 h after vaccination only in the non-anemic but not in the IDA group. Serum iron levels mirrored the hepcidin response: in the non-anemic cohort, increased serum hepcidin was associated with decreased serum iron, whereas in the IDA group, no change in serum iron was observed. The authors therefore concluded that during IDA, regulation of hepcidin by iron and/or erythropoietic activity supersedes hepcidin regulation by acute inflammation. Measurement of erythrocyte iron incorporation from 57Fe-labeled test meals provided a valuable insight into iron absorption before and after the acute inflammatory stimulus. Erythrocyte iron incorporation was higher in IDA compared to non-anemic subjects at all time points examined, reflecting increased iron absorption in this group. Interestingly, erythrocyte 57Fe incorporation was not affected by inflammation in either group, despite increased hepcidin in the non-anemic women. As the authors point out, a possible explanation is that enterocytes may be less sensitive to the effect of hepcidin than recycling macrophages.10 Thus, a moderate increase in hepcidin after vaccination in non-anemic women could cause hypoferremia without, at the same time, affecting duodenal 57 Fe absorption, because serum iron concentration is predominantly determined by macrophage iron export.

Interestingly, in the non-anemic group, erythrocyte 57Fe incorporation was inversely correlated with serum hepcidin both at baseline (r=‒0.792; P<0.001) and after vaccination (r=‒0.708; P<0.001). This suggests that, over a broader range of concentrations, hepcidin does modulate iron absorption, but that hepcidin changes after the vaccination were too small to exert an effect on enterocytes.

This study is the first to test the dynamic hierarchical regulation of hepcidin by iron and inflammation in a well-controlled trial in humans, and showed that iron-deficiency anemia exerted a dominant effect over that of acute inflammation in this setting. What is the molecular mechanism that could explain this observation? The hepcidin promoter contains both bone morphogenetic protein (BMP)-response elements (RE) and a STAT3-RE.11 Iron-mediated hepcidin regulation occurs via the BMP-SMAD pathway. It is thought that liver sinusoidal endothelial cells secrete BMP2 and BMP6 in proportion to the liver iron stores;12,13 these ligands then act in a paracrine fashion, and bind BMP receptors and their coreceptor hemojuvelin (HJV) on hepatocytes to induce phosphorylation of SMAD1/5. Phosphorylated SMAD1/5 form a 1093

Editorials

A

B

Fiure 1. The design (A) and results (B) of the prospective study in women by Stoffel et al.7 The study evaluated the relative contribution of iron deficiency anemia and acute inflammatory stimuli on iron homeostasis. h: hours; d: day.

complex with SMAD4, translocate to the hepatocyte nucleus, and bind to the BMP-RE to induce hepcidin expression. Holo-transferrin concentrations, which are sensed by the TfR1/HFE and TfR2 proteins on hepatocytes, are also thought to modulate the same BMP signaling pathway in these cells. Low iron stores and low circulating iron (observed in the IDA group in this study), would result in decreased BMP signaling and a low level of hepcidin transcription. Low hepcidin would then allow for increased iron absorption and mobilization from stores. However, in the presence of infection, increased iron bioavailability becomes a liability, as pathogens also require iron for proliferation and survival. As part of the host defense, hepcidin is induced by infection and inflammation to limit iron availability to pathogens. Hepcidin regulation by infection and inflammation is mediated in large part by IL-6.14 IL-6 binding to its receptor, IL-6RÎą, and co-receptor, gp130, results in phosphorylation of JAK1/2 in hepatocytes, which then phosphorylates STAT3. This then dimerizes and translocates to the nucleus to induce hepcidin expression.

Importantly, the BMP pathway was shown to synergize with STAT3 pathway to induce hepcidin transcription.

1094

Disruption of BMP-RE in a liver cell line impaired hepcidin response to IL-6.15 Studies using mouse models have demonstrated that hepcidin induction in response to inflammation is blunted when hepatic BMP signaling is genetically disrupted.16-20 Absence of HJV or ALK3 prevented induction of hepcidin in vivo following acute inflammatory stimulus (LPS or IL-6).17 Similarly, in Hfe and Tfr2 knockout mice, hepcidin induction in response to LPS was also blunted.16 Although these mouse studies did not model iron-deficiency anemia, such as was seen in the study subjects in Stoffel et al.7 they provided the proof of principle that the BMP pathway plays an important role in hepcidin responsiveness to inflammation.

However, it remains to be determined whether the presumed decrease in BMP-SMAD signaling in the study subjects is caused by iron deficiency, or by anemia and increased erythropoietic activity, or the combination of these factors. Anemia induces erythropoietin (EPO) secretion by the kidneys.21 EPO in turn acts on the bone marrow erythroblasts to induce expression of erythroferrone (ERFE),22 and ERFE is shown to function as a BMP trap to suppress hepcidin.23 Although EPO was elevated in IDA subjects in this study, serum ERFE levels were not meashaematologica | 2019; 104(6)

Editorials

ured but could provide insight into the contribution of anemia to the blunted hepcidin response. It would be interesting to see if iron deficiency alone is sufficient to prevent hepcidin induction following acute inflammatory stimulation. In fact, eight of the 25 women in the nonanemic group were reported to be iron-deficient, but were not analyzed as a subgroup to determine the contribution of anemia versus iron deficiency.

In addition to the convergence of signals onto the hepcidin promoter, another aspect to consider in the regulation of hepcidin is the relative strength and duration of each signal. In this study, iron-deficiency anemia was relatively mild (median hemoglobin of 11.3 g/dL), but likely chronic. The inflammatory signal was moderate and likely transient, with IL-6 increasing approximately 2-3-fold after vaccination compared to baseline. Hepcidin induction was similarly moderate: in the non-anemic group, hepcidin levels increased 2-fold by 24 h compared to baseline. Whether a stronger or more prolonged inflammatory stimulus, such as during an active infection, would over-ride the effect of IDA on hepcidin remains to be determined. Nevertheless, in agreement with Stoffel et al., a cross-sectional study that compared patients with anemia of chronic disease (ACD) to those with IDA or mixed ACD/IDA condition, reported that hepcidin was increased in patients with ACD compared to control subjects, but that in patients with mixed ACD/IDA, despite elevated IL-6, hepcidin levels were comparable to those observed in IDA patients.24 In conclusion, the data obtained from the welldesigned and well-executed prospective study in human subjects by Stoffel et al. support the conclusion that during iron-deficiency anemia, when challenged by moderate but transient acute inflammation, iron acquisition is prioritized over iron restriction. The questions about the molecular mechanism and relative contribution of erythropoietic activity versus iron deficiency in preventing an inflammation-mediated increase in hepcidin still have to be answered. Importantly, this human study pioneers the analysis of the interactions of iron deficiency and inflammation, a subject of great importance for designing and implementing policies to prevent and treat anemia in regions where iron deficiency, infection and inflammation are all too common.

Acknowledgments and disclosures Sources of support: NIH Ruth L. Kirschstein National Research Service Award T32-5T32HL072752-13 (to VS). EN is a shareholder and scientific advisor of Intrinsic LifeSciences.

References 1. Ganz T. Systemic iron homeostasis. Physiol Rev. 2013;93(4):17211741.

2. Nemeth E, Tuttle MS, Powelson J, et al. Hepcidin regulates cellular iron efflux by binding to ferroportin and inducing its internalization.

haematologica | 2019; 104(6)

Science. 2004;306(5704):2090-2093.

3. Sangkhae V, Nemeth E. Regulation of the Iron Homeostatic Hormone Hepcidin. Adv Nutr. 2017;8(1):126-136.

4. Nicolas G, Chauvet C, Viatte L, et al. The gene encoding the iron regulatory peptide hepcidin is regulated by anemia, hypoxia, and inflammation. J Clin Invest. 2002;110(7):1037-1044.

5. Cassat JE, Skaar EP. Iron in infection and immunity. Cell Host Microbe. 2013;13(5):509-519.

6. Arezes J, Jung G, Gabayan V, et al. Hepcidin-induced hypoferremia is a critical host defense mechanism against the siderophilic bacterium Vibrio vulnificus. Cell Host Microbe. 2015;17(1):47-57.

7. Stoffel NU, Lazrak M, Bellitir S, et al. The opposing effects of acute inflammation and iron deficiency anemia on serum hepcidin and iron absorption in young women. Haematologica. 2019;104(6):11431149.

8. Marx JJ, Dinant HJ. Ferrokinetics and red cell iron uptake in old age: evidence for increased liver iron retention? Haematologica.

1982;67(2):161-168.

9. Marx JJ. Normal iron absorption and decreased red cell iron uptake in the aged. Blood. 1979;53(2):204-211.

10. Chaston T, Chung B, Mascarenhas M, et al. Evidence for differential effects of hepcidin in macrophages and intestinal epithelial cells.

Gut. 2008;57(3):374-382.

11. Truksa J, Lee P, Beutler E. Two BMP responsive elements, STAT, and bZIP/HNF4/COUP motifs of the hepcidin promoter are critical for BMP, SMAD1, and HJV responsiveness. Blood. 2009;113(3):688-695.

12. Canali S, Zumbrennen-Bullough KB, Core AB, et al. Endothelial cells produce bone morphogenetic protein 6 required for iron homeostasis in mice. Blood. 2017;129(4):405-414.

13. Koch PS, Olsavszky V, Ulbrich F, et al. Angiocrine Bmp2 signaling in murine liver controls normal iron homeostasis. Blood.

2017;129(4):415-419.

14. Nemeth E, Rivera S, Gabayan V, et al. IL-6 mediates hypoferremia of inflammation by inducing the synthesis of the iron regulatory hormone hepcidin. J Clin Invest. 2004;113(9):1271-1276.

15. Verga Falzacappa MV, Casanovas G, Hentze MW, Muckenthaler MU. A bone morphogenetic protein (BMP)-responsive element in the hepcidin promoter controls HFE2-mediated hepatic hepcidin expression and its response to IL-6 in cultured cells. J Mol Med. 2008;86(5):531-540.

16. Wallace DF, McDonald CJ, Ostini L, Subramaniam VN. Blunted hepcidin response to inflammation in the absence of Hfe and transferrin receptor 2. Blood. 2011;117(10):2960-2966.

17. Fillebeen C, Wilkinson N, Charlebois E, Katsarou A, Wagner J, Pantopoulos K. Hepcidin-mediated hypoferremic response to acute inflammation requires a threshold of Bmp6/Hjv/Smad signaling.

Blood. 2018;132(17):1829-1841.

18. Huang H, Constante M, Layoun A, Santos MM. Contribution of STAT3 and SMAD4 pathways to the regulation of hepcidin by opposing stimuli. Blood. 2009;113(15):3593-3599.

19. Steinbicker AU, Sachidanandan C, Vonner AJ, et al. Inhibition of bone morphogenetic protein signaling attenuates anemia associated with inflammation. Blood. 2011;117(18):4915-4923.

20. Mayeur C, Lohmeyer LK, Leyton P, et al. The type I BMP receptor Alk3 is required for the induction of hepatic hepcidin gene expression by interleukin-6. Blood. 2014;123(14):2261-2268.

21. Haase VH. Regulation of erythropoiesis by hypoxia-inducible factors. Blood Rev. 2013;27(1):41-53.

22. Kautz L, Jung G, Valore EV, Rivella S, Nemeth E, Ganz T. Identification of erythroferrone as an erythroid regulator of iron metabolism. Nature Genet. 2014;46(7):678-684.

23. Arezes J, Foy N, McHugh K, et al. Erythroferrone inhibits the induction of hepcidin by BMP6. Blood. 2018;132(14):1473-1477.

24. Theurl I, Aigner E, Theurl M, et al. Regulation of iron homeostasis in anemia of chronic disease and iron deficiency anemia: diagnostic and therapeutic implications. Blood. 2009;113(21):5277-5286.

1095

Editorials

The -7 chromosomal abnormalities with signs of myelodysplasia in chronic myeloid leukemia as a major red signal Emilie Cayssials and François Guilhot Inserm CIC 1402, University Hospital of Poitiers, CHU de Poitiers, France E-mail: GUILHOT FRANÇOIS - fr.guilhot@wanadoo.fr doi:10.3324/haematol.2019.217034

C

hronic myeloid leukemia (CML) is a leukemic disorder the prognosis of which has been revolutionized by the use of several generations of tyrosine kinase inhibitors (TKI). During the course of treatment, patients can develop chromosomal abnormalities in addition to Philadelphia-positivity (Ph+).1 These Ph+ clonal chromosomal abnormalities (CCA/Ph+) could be considered markers of disease progression. A subgroup of patients has been clearly identified with fewer responses and worse outcome.2 In addition, a few patients can also develop clonal chromosomal abnormalities in Ph-negative (Ph-) cells (CCA/Ph-). Based on small case series and anecdotal reports, the European Leukemia Net (ELN)3 and the National Comprehensive Cancer Network Guidelines4 did not consider that the presence of CCA/Ph- negatively affects the prognosis provided there was no sign of bone marrow (BM) dysplasia, but chromosome -7/del(7q) abnormalities were identified as a red signal. Since 2001, a number of cases with chromosome 7 abnormalities have been reported, some of them harboring signs of myelodysplastic syndrome (MDS) or of acute myeloid leukemia (AML).5-7 The precise risk of MDS or AML with 7/del(7q) abnormalities is unclear and most of the cases have been reported with a short follow up.

In this issue of Haematologica, Bidet et al.8 report on the largest group of 26 CML patients presenting -7/del(7q) abnormalities treated front line with TKI with a median follow up of 6.47 years. These patients achieve lower cumulative incidence of deep molecular response, more frequently present BM signs of dysplasia, and are more frequently switched to second-generation TKI. It is important to point out that CCA/Ph- clones can only be identified when patients with CML achieve a cytogenetic response. Thus, if physicians want to capture these clones, they should propose that patient undergoes BM cytogenetic analysis during the course of the disease, and at least for the first two years. In addition, BM smears are also essential to detect signs of MDS/AML. Clonal additional abnormalities are usually identified as abnormalities present in ≥2 out of 20 metaphases or if the abnormalities are present in one metaphase in ≥2 assessments.

The first case with a deletion of 7q was reported by Gambacorti-Passeri,9 and subsequently more details were provided by several groups on occasional individual cases. Soon after, Andersen et al.10 reported on a patient who developed monosomy 7 after 12 months of imatinib therapy with BM hypoplastic signs and dysplastic features. Although CCA/Ph- abnormalities is a rare observation, the risk of developing MDS or AML has not been precisely defined, and the proportion of patients with chromosome 7 abnormalities is not really known. Thus, it is important to perform studies on a large group of 1096

CML patients developing additional clonal abnormalities.

By 2011, Groves et al.5 had reviewed all cases published since 2002. The study cohort included 53 patients. The majority were in chronic phase but six had accelerated phase at the time of starting TKI. Previous treatment was interferon for 39 patients and only six patients were previously untreated. Of the 53 patients, -7 was the sole abnormality in 29 patients, -7 with +8 as an additional abnormal Ph- clone in 14 patients, and del(7q) in ten patients. The chromosome 7abnormality was present in two or more metaphases in all but three patients. In the nine patients who developed AML, only one patient survived despite intensive chemotherapy or allogeneic stem cell transplantation. Resolution of MDS (refractory anemia with ring sideroblasts) was noted in one patient after discontinuation of TKI. Based on this survey, it was suggested that the risk of a second malignancy with -7 was significantly higher since none of the patients with del(7q) developed MDS or AML. As AML is an important event, the risk of developing AML was analyzed. Factors influencing the onset of AML appeared to be persistence (compared to transience) of a Ph-7clone, particularly 7sole, and a clone size of 50% or more at diagnosis; time to diagnosis of MDS/AML should also be considered.

Transformation to MDS/AML was observed within six months of Ph -7 detection in 15 patients, with only one patient diagnosed at 11 months. Of the 12 patients with follow up of more than six months, none developed a second myeloid malignancy, suggesting that the risk of MDS/AML is higher within the first six months of Ph -7 detection than at later time points (P<0.0001). Karimata et al.11 reported a single case of a 60-year Japanese man who was treated for seven years with various doses of imatinib. He had several episodes of cytopenia, and five years after the start of imatinib, a cytogenetic analysis revealed a complete cytogenetic response but monosomy 7 was detected in 16 out of 20 cells. BM smears showed a refractory anemia with multilineage dysplasia which progressed to AML and death. Two additional cases of monosomy 7 out of 155 CML chronic phase patients were subsequently reported.6 The median time of the first appearance of monosomy 7 was six months. The clone of monosomy 7 persisted for eight years in the patient who achieved sustained major molecular response (MMR) and six months until BM transplantation in the second patient. More recently, the group of Cortes conducted a retrospective analysis of patients treated front line with first-, second-, and third-generation TKI.7 Among the 598 evaluable patients, 108 (18%) had CCA/Ph-. Of these, 4 patients with monosomy 7 had the worst survival.

In this issue of Haematologica, Bodet et al. provide useful additional information on patients who are diagnosed with -7/del(7q) abnormalities during the course of their haematologica | 2019; 104(6)

Editorials

Table 1. Chronic myeloid leukemia cases with -7 del(7q) in Philadelphia-negative cells.

Author, year (ref) Gambacorti-Passeri 20019 Groves 20115

Karimata 201111 Wasilewska 20176 Issa 20177

Bidet 20198

N. of cases /type

N. with other CCA/Ph-

1 Del(7)(q22q23) 53 (including 6 AP) -7 (29) -7 + 8 (14) Del7q (10) 1 -7 2 -7 4 -7

26 -7

TKI

Previous TT

MDS

AML

Onset of CCA/Ph-

Outcome

1 +8

Imatinib

Yes IFN

No

No

6 months

NA

14 +8

Nilotinib Imatinib Dasatinib

Yes IFN

7

9

10 months 2.8-53)

AML alive 1/9 MDS alive 6 1 NA

Imatinib

Yes IFN Imatinib

Yes

Yes

5 years

Died of AML

No

No

6 months

No

1

Yes (50%° of Pts) IFN

50%

2 +8 None

Dasatinib Nilotinib Imatinib Dasatinib Nilotinib Ponatinib

+8 (19.5%)

Imatinib Dasatinib Nilotinib Bosutinib Ponatinib

-8 years MMR -AlloBMT alive 1 9 months 5 years OS 37% (95%CI: 1-80) -1 died of BC -1 died post allo BMT for AML -1 MDS in MR4.5 on bosutinib -1 MR 4.5 on imatinib No 2.08 years (0.8-12.65) Worse EFS and PFS Same OS

N: number; CCA: clonal chromosomal abnormalities; Ph: Philadelphia; TKI: tyrosine kinase inhibitor; TT: therapy; MDS: myelodysplastic syndrome; AML: acute myeloid leukemia; IFN: interferon; NA: not available; AP: accelerated phase; MR: molecular response;MMR: major molecular response; alloBMT: allogeneic bone marrow transplantation; Pts: patients; BC: blast crisis; EFS: event-free survival; PFS: progression-free survival; OS: overall survival.

TKI therapy. Patients were selected from French centers of the French CML group. They screened 102 CML patients with CCA/Ph-, of these 26 had an abnormality of chromosome 7 [-7/del(7q)]. A control group of 11 MDS patients, four with 7 [-7/del(7q)], was obtained for comparison. All karyotypes were reviewed by the Groupe Francophone de Cytogenetique Hematologique (GFCH) members and then classified according to the International System for Human Cytogenetic Nomenclature (ISCN 2016), molecular monitoring being performed according to the ELN recommendations. Importantly, BM smears were assessed by morphological central review in 48 cases.

Morphological dysplasia was considered significant when it was observed in 10% or more cells in any hematopoietic lineage with or without excess blasts (>5%). Underlying MDS was documented both by centralized morphological analysis of BM smears and by sequencing a targeted panel of 27 genes frequently altered in MDS and AML performed by a next-generation sequencing (NGS) assay. Chromosome 7 abnormalities were isolated in 13 out of the 26, and associated with +8 in 19.5% of cases. Patients who developed -7/del(7q) CCA/Ph- were significantly younger than other CCA/Phabnormalities (mean, 48 vs. 55 years old; P=0.035) and mostly benefit second-generation TKI as first line of treatment. Twelve out of the 26 patients who developed -7/del(7q) CCA/Ph- were in complete cytogenetic response at time of detection. Median follow up since haematologica | 2019; 104(6)

CCA/Ph- detection was 5.35 (1-14) years for -7/del(7q) and 7.41 (0-15) years for others CCA/Ph- (Pâ&#x2030;Ľ0.05). MDS signs were more frequent in -7/del(7q) CCA/Ph- patients as compared to other types of CCA/Ph-. Molecular response (MR) was less frequently observed in these patients: 33% (n=4 out of 12) of the -7/del(7q) patients with MDS signs achieved MMR or better versus 75% (9 out of 12) of these patients without MDS signs. Using NGS, MDS morphological features were significantly associated with the presence of mutation. Among the 12 patients with -7/del(7q), 4 presented MDS morphological signs and 3 were mutated. Cumulative incidence of MR4.5 was lower in patients with -7/del(7q). Although type of CCA/Ph- did not impact on overall survival, landmark analysis at three years after first TKI initiation revealed a strong negative impact of -7/del(7q) CCA/Phon event-free survival and progression-free survival.

The study of the French Group reports on the largest cohort with a long follow up (median 6.47 years) of CML patients with -7/del(7q) abnormalities.8 These patients are younger, achieved lower cumulative incidence of MR4.5, and are more frequently associated with dysplastic features. One patient with -7 CCA/Ph- presented EZH2 mutations and then progressed to advanced CML phase; however, the role of this mutation can only be speculated as it was at low variant frequency.

Considering all the rare cases of -7/del(7q) CCA/Phabnormalities reported in the literature (Table 1), it 1097

Editorials

appears that such abnormalities are mostly observed during the first two years of TKI therapy. While there is a significant risk of a second myeloid malignancy in patients with -7 CCA/Ph-, less than half of these patients will develop MDS/AML. The aggregate data provide evidence in support of the commonly held view that preemptive therapeutic strategies are not justified in all patients with detectable -7 CCA/Ph-. Nevertheless, once a diagnosis of AML is confirmed in these patients, intensive treatment strategies, including allogeneic BM transplantation, are ineffective in most patients. One may speculate on the role of TKI in the mechanism of MDS development and the presence of -7/del(7q) CCA/Phabnormalities. The mutagenic effect of TKI on hematopoietic stem cells is not yet fully understood.

However, it has been reported that a gastrointestinal stromal tumor patient developed MDS with monosomy 7 during imatinib treatment, suggesting that imatinib plays a direct role in causing MDS.12 The routine monitoring of CML patients is currently molecular assessment of the response. However, cytogenetic analysis is still relevant and should be performed with a BM smear certainly in cases of cytopenia during TKI therapy. Signs of dysplasia with -7 CCA/Ph- cells should be considered as a red signal and a switch to alternative treatment be discussed.

References 1. Guilhot F. Cytogenetics in CML: more important than you think.

Blood. 2016;127(22):2661-2662.

2. Wang W, Cortes JE, Tang G, et al. Risk stratification of chromosomal abnormalities in chronic myelogenous leukemia in the era of tyrosine kinase inhibitor therapy. Blood. 2016;127(22):2742-2750.

3. Baccarani M, Deininger MW, Rosti G, et al. European LeukemiaNet recommendations for the management of chronic myeloid leukemia: 2013. Blood. 2013;122(6):872-884.

4. National Comprehensive Cancer Network (NCCN). Chronic myeloid leukemia (version 1.2019). Available at: https://www.nccn.org.

5. Groves MJ, Sales M, Baker L, Griffiths M, Pratt N, Tauro S. Factors influencing a second myeloid malignancy in patients with Philadelphia-negative -7 or del(7q) clones during tyrosine kinase inhibitor therapy for chronic myeloid leukemia. Cancer Genet.

2011;204(1):39-44.

6. Wasilewska EM, Panasiuk B, Gniot M, et al. Clonal chromosomal aberrations in Philadelphia negative cells such as monosomy 7 and trisomy 8 may persist for years with no impact on the long-term outcome in patients with chronic myeloid leukemia. Cancer Genet.

2017;216-217:1-9.

7. Issa GC, Kantarjian HM, Gonzalez GN, et al. Clonal chromosomal abnormalities appearing in Philadelphia chromosome-negative metaphases during CML treatment. Blood. 2017;130(19):2084-2091.

8. Bidet A, Dulucq S, Smol T, et al. Poor prognosis of chromosome 7 clonal aberrations in Philadelphia-negative metaphases and relevance of potential underlying myelodysplastic features in chronic myeloid leukaemia. Hematologica. 2019;104(6):1150-1155.

9. Gambacorti-Passeri C, Giudici G, le Coutre P, et al. Non-random chromosomal abnormalities in Ph-negative bone marrow (BM) cells from CML patients achieving major cytogenetic responses (MCR) with STI571 (GleevecTM). Blood. 2001;98:257b (Abstr. 4762).

10. Andersen MK, Pedersen-Bjergaard J, Kjeldsen L, Dufva IH, BrøndumNielsen K. Clonal Ph-negative hematopoiesis in CML after therapy with imatinib mesylate is frequently characterized by trisomy 8.

Leukemia. 2002;16(7):1390-1393.

11. Karimata K, Masuko M, Ushiki T, et al. Myelodysplastic Syndrome with Ph Negative Monosomy 7 Chromosome following Transient Bone Marrow Dysplasia during Imatinib Treatment for Chronic Myeloid Leukemia. Intern Med. 2011;50(5):481-485.

12. Pitini V, Arrigo C, Sauta MG, Altavilla G. Myelodysplastic syndrome appearing during imatinib mesylate therapy in a patient with GIST. Leuk Res. 2009;33(9):143-144.

Another piece of the puzzle added to understand t(4;11) leukemia better Rolf Marschalek Institute of Pharmaceutical Biology / Diagnostic Center of Acute Leukemia, University of Frankfurt, Frankfurt/Main, Germany; E-mail: ROLF MARSCHALEK - Rolf.Marschalek@em.uni-frankfurt.de doi:10.3324/haematol.2018.213397

T

he story about t(4;11) leukemia, involving the MLL/KMT2A gene from chromosome 11q23.3 and the AF4/AFF1 gene from chromosome 4q21, is still a mystery. The study by Agraz-Doblas et al., published in this issue of Haematologica, adds some new and important information regarding the mysterious pathomechanism.1 Agraz-Doblas et al. showed, for the first time, that the therapeutic outlook of patients with expression of both reciprocal MLL fusions, MLL-AF4 and AF4-MLL, is promising, but only 50% of the investigated patients seem to have this favorable condition; patients expressing only the MLL-AF4 allele have an event-free survival of 10% and an overall survival of 30%. Moreover, only leukemic cells expressing both fusion alleles display the typical HOXA signature.

The fact that t(4;11) patients can be divided into two subgroups on the basis of HOXA transcription was first 1098

recognized by Trentin et al. in 2009,2 and later confirmed

by Stam et al. and Kang et al. in 2010 and 2012, respectively.3,4 The missing HOXA transcription was correlated with overexpression of either IRX1 or IRX22,4 and a 3-fold higher relapse rate.3,4 Experimental overexpression of IRX1 revealed an interesting mechanism because it resulted in EGR1-3 expression.5 EGR1 and EGR2 both control the p21CIP1 gene and, thus, shut down the cell cycle and may even induce cellular quiescence, a known mechanism of resistance to treatment. CDK6 counteracts the actions of EGR proteins.6 The second mechanism involves the IRX proteins, which are able to turn on HOXB4, a known stem cell marker of hematopoietic cells that activates factors such as TAL1, GATA factors, TGFB1, etc. Thus, expression of MLL-AF4 alone - with upregulated IRX proteins but without HOXA expression - may provoke treatment resistance or a stem cell-like mechanism which is haematologica | 2019; 104(6)

Editorials

not possible when AF4-MLL is present. This could be a rational explanation for the observed clinical behavior of both groups of patients.

Another explanation could lie in the recent findings from Yokoyamaâ&#x20AC;&#x2122;s laboratory.7,8 Okuda et al. elegantly showed that one of the functions of the pSer domain of the AF4 protein9 - which is fused to the N-terminus of MLL in the MLL-AF4 fusion protein - is recruitment of the SL1 complex. The SL1 complex is usually bound to RNA Pol I, which is present in the nucleolus and required for the transcription of rRNA genes. SL1 is artificially recruited to MLL-AF4, but not to native MLL or AF4 complexes.10 This powers up MLL-AF4 leading to a strong increase in gene transcription. The simple presence of the MLLAF4 fusion protein causes a condition of severe stress, because it compromises protein biosynthesis, and cells may therefore easily display a phenotype of growth arrest or senescence.11 This is probably one of the reasons why it is so complicated to generate a true MLL-AF4 mouse model, and why so many laboratories have failed so far. With the exception of a recent study in which a hybrid between human MLL and mouse Af4 was used,12 no-one had been able to develop a satisfactory model

with only the human MLL-AF4 fusion. Since the sequences of the pSer domains of human and mouse AF4 differ slightly, it is very plausible that the human/mouse chimeric MLL-Af4 is unable to attract the SL1 component and thus does not impair protein biosynthesis. This needs to be tested in future experiments.

So, what is the precise role of AF4-MLL? AF4-MLL has been shown to strongly enhance gene transcription by overwriting the transcriptional elongation control.10,13 This massive increase in gene transcription (3- to 4-fold more mRNA) may help its molecular counterpart, MLLAF4, to set the programming of its target genes, even under conditions of nucleolar stress (see Figure 1). In addition, we have shown by ATAC sequencing in two independent cell lines that AF4-MLL strongly activates chromatin in a very short time frame (unpublished data from our laboratory). The expression of AF4-MLL for 48 h was sufficient to open up the chromatin of all chromosomes apart from the centromeric regions and to massively increase gene transcription. Thus, the presence of AF4MLL would allow the expression of any gene of interest, and increases the plasticity of the tumor cells. According to the data shown by Agraz-Doblas et al. this results in

Figure 1. Proposed fuctions of both t(4;11) fusion proteins. MLL-AF4 binds to Menin1/LEDGF and SL1 to target gene promotors and strongly activate gene transcription, while AF4-MLL overwrites transcriptional elongation control and strongly activates chromatin within a very short time window. Patients who express both fusions diplay HOXA gene signatures and have a better outcome, while patients who either do not express AF4-MLL or actively repress it, usually activate the homeobox proteins IRX1/2. Expression of both proteins has been correlated with a worse outcome (IRX1/2 strongly activate EGR1-3 and HOXB4).

haematologica | 2019; 104(6)

1099

Editorials

cells that are more vulnerable to chemotherapy because the presence of AF4-MLL was associated with a much better treatment outcome.1 Lastly, expression of AF4-MLL alone was shown to be necessary and sufficient to cause acute leukemia.14 Our group used a low expression retroviral vector backbone (PIDE vector) to express MLL-AF4, AF4-MLL or both in LSK cells purified from C57BL/6 mice. Empty vector or MLL-AF4 alone did not result in the development of leukemia, while AF4-MLL or the expression of both fusion genes resulted in full-blown pro-B-acute lymphoblastic leukemia and mixed-lineage leukemia. The latency was 9 months and the penetrance was only 35%.

However, this could be attributed to a low infection rate with about 1/1,000 cells for the AF4-MLL and 1/10,000 cells for the MLL-AF4 construct because these constructs were oversized for in vitro packaging (11.3 kb for MLLAF4 and 13.3 kb for AF4-MLL). Therefore, an estimated 200 cells in 200,000 non-tranduced cells were transplanted into primary mice, which nevertheless caused a disease outbreak (MLL-AF4 with only 20 cells did not work).

It is noteworthy that all "leukemic cells" subsequently tested positive for the transcription and integrity of the appropriate transgenes, while the investigated white blood cells of mice who did not develop leukemia remained negative in reverse transcriptase and genomic polymerase chain reaction experiments. This indicates that the leukemia-negative mice either never received or lost the cells carrying the corresponding transgene (negative selection of MLL-AF4 alone).

We, therefore, assume that, in humans, AF4-MLL and MLL-AF4 are both necessary, but AF4-MLL could presumably be shut off after "preparing the ground" for MLL-AF4, and that this process of shutting down AF4MLL makes the leukemic disease even more aggressive (positive selection). This explains in part the molecular situation diagnosed in human patients with leukemia, regardless of whether they are infants or adults. It would be of interest to compare primary diagnostic material with relapsed material from the same patient, and determine whether AF4-MLL expression is lost in the relapse, in order to have another argument in favor of the above mentioned hypothesis.

1100

The study by Agraz-Doblas et al. adds another, important piece to the puzzle of the molecular mechanism of t(4;11) leukemia.1 It is to be hoped that the precise mechanism of this disease can be understood soon, because the full picture is needed in order to develop new drugs that can really help patients with t(4;11) leukemia.

References 1. Agraz-Doblas A, Bueno C,2 Bashford-Rogers R, et al. Unraveling the cellular origin and clinical prognostic markers of infant B-cell acute lymphoblastic leukemia using genome-wide analysis.

Haematologica 2019;104(6):1176-1188.

2. Trentin L, Giordan M, Dingermann T, et al. Two independent gene signatures in pediatric t(4;11) acute lymphoblastic leukemia patients.

Eur J Haematol. 2009;83(5):406-419.

3. Stam RW, Schneider P, Hagelstein JA, et al. Gene expression profiling-based dissection of MLL translocated and MLL germline acute lymphoblastic leukemia in infants. Blood. 2010;115(14):2835-2844.

4. Kang H, Wilson CS, Harvey RC, et al. Gene expression profiles predictive of outcome and age in infant acute lymphoblastic leukemia: a Children's Oncology Group study. Blood. 2012;119(8):1872-1881.

5. Kühn A, Löscher D, Marschalek R. The IRX1/HOXA connection: insights into a novel t(4;11)- specific cancer mechanism. Oncotarget.

2016;7(23):35341-35352.

6. Guzman ML. CDK6 is a regulator of stem cells "Egr" to wake up.

Blood. 2015;125(1):7-9.

7. Okuda H, Kanai A, Ito S, Matsui H, Yokoyama A. AF4 uses the SL1 components of RNAP1 machinery to initiate MLL fusion- and AEPdependent transcription. Nat Commun. 2015;6:8869.

8. Okuda H, Takahashi S, Takaori-Kondo A, Yokoyama A. TBP loading

by AF4 through SL1 is the major rate-limiting step in MLL fusiondependent transcription. Cell Cycle. 2016;15(20):2712-2722.

9. Nilson I, Reichel M, Ennas MG, et al. Exon/intron structure of the human AF-4 gene, a member of the AF-4/LAF-4/FMR-2 gene family coding for a nuclear protein with structural alterations in acute leukaemia. Br J Haematol. 1997;98(1):157-169.

10. Benedikt A, Baltruschat S, Scholz B, et al. The leukemogenic AF4MLL fusion protein causes P-TEFb kinase activation and altered epigenetic signatures. Leukemia. 2011;25(1):135-144.

11. Caslini C, Serna A, Rossi V, Introna M, Biondi A. Modulation of cell cycle by graded expression of MLL-AF4 fusion oncoprotein.

Leukemia. 2004;18(6):1064-1071.

12. Lin S, Luo RT, Ptasinska A, et al. Instructive role of MLL-fusion proteins revealed by a model of t(4;11) pro-B acute lymphoblastic leukemia. Cancer Cell. 2016;30(5):737-749.

13. Mück F, Bracharz S, Marschalek R. DDX6 transfers P-TEFb kinase to the AF4/AF4N (AFF1) super elongation complex. Am J Blood Res. 2016;6(3):28-45.

14. Bursen A, Schwabe K, Rüster B, et al. The AF4-MLL fusion protein is capable of inducing ALL in mice without requirement of MLL-AF4.

Blood. 2010;115(17):3570-3579.

haematologica | 2019; 104(6)

Editorials

Conditioning intensity and antilymphocyte globulin: towards personalized transplant strategies?

Martin Bornhäuser Department of Internal Medicine I, University Hospital Carl Gustav Carus, TU Dresden and National Center for Tumor Disease (NCT), Germany E-mail: MARTIN BORNHÄUSER - martin.bornhaeuser@uniklinikum-dresden.de doi:10.3324/haematol.2019.216952

I

n the last two decades allogeneic hematopoietic cell transplantation (HCT) has been used with increasing frequencies in hematologic malignancies with curative intent. The increased understanding of immune tolerance and allogeneic antileukemic immune reactivity has led several investigators to develop optimized conditioning protocols and new strategies to manipulate the effector cells either within the graft or in vivo.

More specifically, the development of minimal intensity or so called “nonmyeloablative” conditioning regimens paved the way towards the application of allogeneic HCT in older patients and all of those who probably would not tolerate classical intensity conditioning.1,2 The most frequently used protocol, spear-headed by investigators from the Fred Hutchinson Cancer Research Center, was based on a single dose of 200 cGy of total-body irradiation (TBI) and a few doses of fludarabine followed by pharmacological immunosuppression.3 After a wave of fascinating reports on the feasibility and efficacy of this “revolutionary” approach, some studies revealed that patients with a high risk of either disease recurrence or non-engraftment did not fare too well with this strategy.4 At the other end of the spectrum, randomized comparisons suggested that less toxic

but still intensive conditioning with 800 cGy of fractionated TBI combined with fludarabine did not increase the rate of disease recurrence at the same time as significantly reducing extramedullary toxicity compared to 1200 cGy TBI with high-dose cyclophosphamide.5 Along these lines, Monaca and co-workers,6 again from Seattle, have convincingly demonstrated in this issue of Haematologica that subtle dose increases of unfractionated TBI can significantly decrease the failure rate after allogeneic HCT. Most interestingly, applying a differential dose escalation strategy, they identified the optimal TBI dose for patients with high-risk myelodysplastic syndromes and chronic myelomonocytic leukemia (450 cGy) and patients with low-risk myelodysplastic syndromes and myeloproliferative neoplasms (300 cGy). Intermediate doses of unfractionated TBI have been successfully used by other colleagues in diseases such as chronic myeloid leukemia.7 The current optimization of the nonmyeloablative protocol developed by the investigators in Seattle clearly shows that differential doses of TBI complement the most frequently applied protocols based on alkylating agents combined with purine analogs.

As stated above, the fine tuning of the allogeneic immune response by T-cell depletion or pharmacological means rep-

Figure 1. A personally tailored approach to hematopoietic cell transplantation. TBI; total body irradiation; CTX: chemotherapy; GvHD: graft-versus-host disease; HCT: hematopoietic cell transplantation.

haematologica | 2019; 104(6)

1101

Editorials

resents the second important pillar in the development of optimized transplantation protocols. Specifically, the use of antihuman T lymphocyte globulin (ATG), a polyvalent preparation generated in rabbits against Jurkat T cells, was introduced several decades ago. Antithymocyte gobulin, derived from rabbits or horses against human thymocytes, has predominantly been used for the treatment of severe aplastic anemia but also within conditioning for allogeneic HCT in selected protocols, mainly for non-malignant indications.8 The pivotal trial testing ATG in the setting of unrelated donors and intensive conditioning suggested a significant reduction in the incidence of chronic graft-versus-host disease without an increase in the risk of relapse.9 While this trial tested a dose of 60 mg/kg ATG (day -3 to -1), a similar finding was made after allogeneic HCT from matched siblings applying 30 mg/kg again within intensive conditioning.10 Both trials included patients with acute leukemia and myelodysplastic syndromes. As a note of caution, a second controlled trial performed in the unrelated donor setting did not confirm the initial observations, suggesting a potential interaction between the choice of conditioning regimen and the chosen ATG regimen.11 So far, no randomized trial had formally tested ATG after reduced intensity conditioning in transplants for a specific indication. Robin and coworkers now present the results of a retrospective analysis performed in patients with myelofibrosis receiving reduced intensity conditioning.12 Interestingly, ATG reduced the cumulative incidence of acute graft-versus-host disease while it did not affect the rate of chronic graft-versus-host disease. Although the authors speculate that this may be due to the use of reduced intensity conditioning, there might be various other factors that could lead to differential effects in this specific group of patients, most of whom still have relatively high lymphocyte counts and organomegaly which can influence the pharmacodynamics of ATG.13 Although both trials address different aspects in the effort to optimize outcomes after allogeneic HCT, they clearly demonstrate that patient and disease characteristics but also graft source and choice of donor can significantly affect the outcome of treatment. Integrating the mentioned factors, the current toolbox offers conditioning protocols of almost all intensities including sequential regimens14 as well as various options of modulating allogeneic immune responses such as by graft manipulation or in vivo T-cell depletion.

Taking advantage of all available knowledge gained by artificial intelligence and large databases15 may enable design of the ideal preparative regimen and post-grafting immunosuppression for a given patient and in a specific immunogenetic setting thereby making allogeneic HCT a tailored approach comparable to that of several other current strategies in modern oncology (Figure 1).

1102

References 1. Giralt S, Estey E, Albitar M, et al. Engraftment of allogeneic hematopoietic progenitor cells with purine analog-containing chemotherapy: harnessing graft-versus-leukemia without myeloablative therapy. Blood.

1997;89(12):4531–4536.

2. Slavin S, Nagler A, Naparstek E, et al. Nonmyeloablative stem cell transplantation and cell therapy as an alternative to conventional bone marrow transplantation with lethal cytoreduction for the treatment of malignant and nonmalignant hematologic diseases. Blood.

1998;91(3):756–763.

3. McSweeney PA, Niederwieser D, Shizuru JA, et al. Hematopoietic cell transplantation in older patients with hematologic malignancies: replacing high-dose cytotoxic therapy with graft-versus-tumor effects.

Blood. 2001;97(11):3390–3400.

4. Sorror ML, Sandmaier BM, Storer BE, et al. Long-term outcomes among older patients following nonmyeloablative conditioning and allogeneic hematopoietic cell transplantation for advanced hematologic malignancies. JAMA. 2011;306(17):1874–1883.

5. Bornhauser M, Kienast J, Trenschel R, et al. Reduced-intensity conditioning versus standard conditioning before allogeneic haemopoietic cell transplantation in patients with acute myeloid leukaemia in first complete remission: a prospective, open-label randomised phase 3 trial.

Lancet Oncol. 2012;13(10):1035–1044.

6. Monaco F, Scott BL, Chauncey TR, et al. Total body irradiation dose escalation decreases risk of progression and graft rejection after hematopoietic cell transplantation for myelodysplastic syndromes or myeloproliferative neoplasms. Haematologica. 2019;104(6):1221-1229 7. Adkins DR, DiPersio JF. Total body irradiation before an allogeneic stem cell transplantation: is there a magic dose? Curr Opin Hematol.

2008;15(6):555–560.

8. Bacigalupo A. Antilymphocyte/thymocyte globulin for graft versus host disease prophylaxis: efficacy and side effects. Bone Marrow Transplant. 2005;35(3):225–231.

9. Finke J, Bethge WA, Schmoor C, et al. Standard graft-versus-host disease prophylaxis with or without anti-T-cell globulin in haematopoietic cell transplantation from matched unrelated donors: a randomised, open-label, multicentre phase 3 trial. Lancet Oncol. 2009;10(9):855– 864.

10. Kröger N, Solano C, Wolschke C, et al. Antilymphocyte Globulin for Prevention of chronic graft-versus-host disease. N Engl J Med. 2016;374(1):43–53.

11. Soiffer RJ, Kim HT, McGuirk J, et al. Prospective, randomized, doubleblind, phase III clinical trial of anti-T-lymphocyte globulin to assess impact on chronic graft-versus-host disease-free survival in patients undergoing HLA-matched unrelated myeloablative hematopoietic cell transplantation. J Clin Oncol. 2017;35(36):4003–4011.

12. Robin M, Chevret S, Koster L, et al. Antilymphocyte globulin for matched sibling donor transplantation in patients with myelofibrosis.

Haematologica. 2019;104(6):1230-1236.

13. Kennedy VE, Chen H, Savani BN, et al. Optimizing antithymocyte globulin dosing for unrelated donor allogeneic hematopoietic cell transplantation based on recipient absolute lymphocyte count. Biol Blood Marrow Transplant. 2018;24(1):150–155.

14. Schmid, C, Schleuning, M, Schwerdtfeger R, et al. Long-term survival in refractory acute myeloid leukemia after sequential treatment with chemotherapy and reduced-intensity conditioning for allogeneic stem cell transplantation. Blood. 2006;108(3):1092–1099.

15. Shouval R, Bondi O, Mishan H, Shimoni A, Unger R, Nagler A. Application of machine learning algorithms for clinical predictive modeling: a data-mining approach in SCT. Bone Marrow Transplant.

2014;49(3):332–337.

haematologica | 2019; 104(6)

Editorials

Is the mysterious platelet receptor GPV an unsuspected major target for platelet autoantibodies?

Paquita Nurden and Alan T Nurden Institut Hospitalo-Universitaire LIRYC, Pessac, France E-mail: PAQUITA NURDEN - paquita.nurden@gmail.com doi:10.3324/haematol.2019.2018.214908

T

o confirm an origin of immune thrombocytopenia (ITP), serum autoantibodies are routinely tested for by enzyme-linked immunosorbent assay (ELISA), or monoclonal antibody immobilization of platelet antigens (MAIPA), with the integrin alpha IIb-beta 3 (αIIbb3) and glycoprotein (GP)Ib-IX receptors on platelets as major targets.1 A range of commercially available kits also allows the detection of human antibodies to these and other platelet targets. In spite of this, confirmation of an immune process for an acquired thrombocytopenia is often difficult to establish, as many potential antigenic targets occur on less well-characterized or minor surface components of platelets that are not tested for. Over the past 20 years, familial thrombocytopenia, often linked with increased platelet size, has no longer been systematically considered as ITP.2 The recognition of a possible genetic origin, even in the absence of a family history, has been facilitated by the tremendous progress linked to the development of next generation sequencing procedures.3 This revolution has reinforced the need for more positive criteria to diagnose ITP and to be able to elaborate the best strategy for their treatment.

In this issue of Haematologica, Vollenberg et al.4 identify GPV, a little understood constituent of the GPIb-IX-V complex, as a frequent target for autoantibodies in ITP. Firstly, they developed methods to improve the detection of platelet-bound and free autoantibodies, and also to evaluate their pathological consequences, using a phagocytosis assay or a NOD/SCID mouse model. In particular, they applied surface plasmon resonance (SPR) technology using recombinant His-tagged GPV. With SPR, they were able to detect lower avidity autoantibodies in the sera of patients compared to indirect MAIPA. Using a direct MAIPA, they found platelet-bound anti-GPV autoantibodies either alone (2.9%) or more often associated with antibodies to GPIb-IX and/or to αIIbb3 in 61.8% of 343 positive samples in a series of 1140 ITP patients. Free antibodies to GPV in the sera were found in 66.6% of the 45 patients found positive by indirect MAIPA, which is more frequent than those found for αIIbb3. When sera from 222 patients positive for platelet-bound GPV were tested using SPR, 88 (39.6%) now tested positively, showing a higher specificity due to the ability of the SPR approach to detect lower avidity antibodies. It should be noted, however, that low avidity antibodies to αIIbb3 and GPIbIX were not evaluated in serum. Blocking with recombinant GPV confirmed the specificity for serum autoantibodies to GPV in MAIPA. Vollenberg et al.4 clearly demonstrated the role of antiGPV autoantibodies in mediating platelet clearance. First, they showed that both high- and low-avidity anti-GPV were able to mediate platelet uptake in a phagocytosis haematologica | 2019; 104(6)

assay using human macrophages. They then performed an in vivo experiment using a NOD/SCID mouse model with transfused human platelets. Their model was validated using a murine monoclonal antibody (MoAb) SW16 against human GPV injected at two concentrations that induced comparable clearance to a murine MoAb (SZ21) specific for αIIbb3. Next, the IgG fraction of sera from patients containing exclusively anti-GPV autoantibodies of low- or high-avidity were administered.

Interestingly, the human anti-GPV increased platelet clearance, although to a lower extent than was seen with SW16; results were similar for the high- or low-avidity IgG fractions and platelet survival was increased after adsorption of the sera with recombinant GPV. It should be noted, however, that the extent to which platelet GPV was saturated was not evaluated. In conclusion, as for classical autoantibodies in ITP, anti-GPV are able to cause thrombocytopenia.

The mechanism by which platelets are cleared for patients with anti-GPV antibodies is worthy of further investigation. Early biosynthesis and assembly of the GPIb-IX complex includes N-glycosylation of the individual subunits and extensive O-glycosylation of the extracellular mucin-like domain of GPIbα.5 GPV (approx. 85 kDa) is non-covalently associated by way of transmembrane interactions with GPIbα and its association with GPIb-IX is required for full expression at the platelet surface.6 Classically, the GPIb-IX-V complex is lacking or dysfunctional in patients with inherited biallelic BernardSoulier syndrome (BSS) where bleeding occurs through the adhesion defect caused by the absence of GPIbα, the platelet receptor for von Willebrand factor (VWF).7 It has 50% expression in monoallelic forms of BSS linked to macrothrombocytopenia alone.2,7 In contrast to the other subunits, the specific absence of GPV in mice (GPV-/-) did not result in BSS, platelet and MK morphology were normal, and there was abundant GPIb-IX expression.8 Classically, mutations in GPV do not give rise to human BSS.7 In ITP, pathways to remove platelets from the circulation include Fc-receptor-mediated clearance by mononuclear macrophages primarily, although not exclusively, located in the spleen.9 Vollenberg et al.4 clearly show that antibodies to GPV can bring about phagocytosis by macrophages. But are there other possible mechanisms?

Platelet activation in platelets incubated with monoclonal antibodies (MoAbs) to GPIb (but not with antibodies to αIIbb3) or in ITP patients with anti-GPIb autoantibodies can be followed by neuraminidase translocation to the platelet surface.10 Removal of terminal sialic acid residues from the O-linked oligosaccharides of the mucin-like domain of GPIbα results in severe thrombocytopenia due 1103

Editorials

Figure 1. GPV, a major target for autoantibodies in immune thrombocytopenia (ITP). (Left) Representation of a platelet showing the two principal receptor targets for platelet autoantibodies. Vollenberg et al.,4 while confirming the abundance of autoantibodies detected by direct MAIPA in a series of 343 positive samples from among 1140 tested ITP patients, also showed a significant presence of antibodies to GPV. (Middle) The GPIb-IX-V complex showing the presence of anti-GPV antibody and highlighting binding or cleavage sites for thrombin and metalloproteases ADAM10 and ADAM17. Vollenberg et al.4 raise questions as to the consequences of bound autoantibodies on GPV structure, the cleavage of released soluble forms, and a potential desialylation of the subunits on platelet survival in ITP. (Right) The detection of both high avidity autoantibodies to GPV by MAIPA and additional low avidity antibodies by surface plasmon resonance (SPR) technology is intriguing, as both of these appear to contribute to the diagnosis of ITP.

to platelet clearance in the liver via aMb2 on Kupffer cells or Ashwell-Morell receptors (AMR) on hepatocytes.10 For GPV, loss of sialic acid occurs concomitantly to GPIba after cold stored platelets are rewarmed, a process also followed by metalloprotease-induced shedding of these receptors.11 It is now important to determine whether thrombocytopenia observed with anti-GPV antibodies results from a Fc-dependent clearance or from an Fc-independent mechanism implying liver receptors.

Neuraminidases are present in platelet a-granules, but it is not known how they are translocated to the platelet surface or how they come into contact with sialic acid residues that are extended far from the surface. Is their action on the same platelet or after platelet-to-platelet contact? Not all anti-GPIb antibodies are capable of activating the GPIb-IX complex; significantly, some directed against the ligand binding domain induce the juxtamembrane mechanosensory domain (MSD) of GPIba to be unfolded and Fc-independent platelet clearance by a mechanosensory mechanism.12 It will be interesting to verify whether anti-GPV autoantibodies share some of these properties or will modify the sialic acid expression of GPIba, or indeed of GPV itself. Heterogeneity in the response may be predicted, as infusion of rat MoAbs to murine GPV had no effect in a mouse model, whereas MoAbs to GPIba resulted in thrombocytopenia and megakaryocyte abnormalities.13 Along with the target on GPV, it is difficult to speculate as to the relevance of high and low avidity anti-GPV antibodies.

The GPV ectodomain contains: i) a cleavage site for thrombin, leading to loss of the bulk of the extracellular domain of GPV from the platelet surface and the generation of soluble GPV (sGPV); and ii) a second cleavage site 1104

for endogeneous metalloproteinases and, in particular, ADAM17, which is the major sheddase for GPV on the platelet surface (Figure 1).14 Whether shedding of GPV is a contributor to or a consequence of immune clearance has still not been determined. The presence of elevated levels of sGPV in models of thrombosis suggests that it may be a marker of thrombotic activity.15 The role of sGPV in the circulation remains unknown, as does the possible role of natural antibodies in clearing the soluble form.

So far, the physiological role and function, if any, of GPV remains elusive. GPV-/- mouse platelets are morphologically indistinguishable from wild-type platelets, but two reports show an increased platelet sensitivity to low doses of thrombin in the absence of GPV, suggesting an anti-thrombotic role for GPV.9,16,17 In contrast, platelet GPV reportedly binds to collagen and participates in platelet adhesion and aggregation.18 Mice lacking GPV have mildly reduced tail bleeding times and, depending on the severity of the injury, display slightly accelerated thrombus formation.19 So the fascinating questions as to the potential value of GPV as an anti-thrombotic target and the possible modulating roles of sGPV remain. Are these observations sufficient to predict a modification of platelet function when anti-GPV antibodies are present?

Only future studies will be able to provide an answer.

In their report, Vollenberg et al.4 focus on the prevalence of auto anti-GPV antibodies and their contribution to platelet clearance. They clearly establish the importance of including this target in the diagnosis of ITP, and confirm the findings of previously published literature and of a recent short report by Porcelijn et al.20 GPV should certainly form part of antibody testing kits, but more studies are required to evaluate their clinical relehaematologica | 2019; 104(6)

Editorials

vance. However, in spite of this, the manuscript of Vollenberg et al.4 opens new fields of investigation concerning the evaluation of the mechanisms involved in immune thrombocytopenia and the bleeding tendency of the patients affected.

References 1. Kiefel V, Santoso S, Kaufmann E, Mueller-Eckhardt C. Autoantibodies against platelet glycoprotein Ib/IX: a frequent finding in autoimmune thrombocytopenic purpura. Br J Haematol.

1991;79(2):256-262.

2. Noris P, Balduini CL. Inherited thrombocytopenias in the era of personalized medicine. Haematologica. 2015;100(2):145-148.

3. Bastida JM, Lozano ML, Benito R, et al. Introducing high-throughput sequencing into mainstream genetic diagnostic practice in inherited platelet disorders. Haematologica. 2018;103(1):148-162.

4. Vollenberg, R, Jouni R, Norris PAA, et al. Glycoprotein V is a relevant immune target in patients with immune thrombocytopenia.

Haematologica. 2019;104(6):1237-1243.

5. Ulsemer P, Strassel C, Baas MJ, et al. Biosynthesis and intracellular post-translational processing of normal and mutant platelet glycoprotein GPIb-IX. Biochem J. 2001;358(Pt 2):295-303.

6. Li CQ, Dong JF, Lanza F, Sanan DA, Sae-Tung G, Lopez JA. Expression of platelet glycoprotein (GP) V in heterologous cells and evidence for its association with GPIbalpha in forming a GPIb-IX-V complex on the cell surface. J Biol Chem. 1995;270(27):16302-16307.

7. Savoia A, Pastore A, DE Rocco D, et al. Clinical and genetic aspects of Bernard-Soulier syndrome: searching for genotype/phenotype correlations. Haematologica. 2011;96(3):417-423.

8. Poujol C, Ramakrishnan V, DeGuzman F, Nurden AT, Phillips DR, Nurden P. Ultrastructural analysis of megakaryocytes in GPV knockout mice. Thromb Haemost. 2000;84(2):312-318.

haematologica | 2019; 104(6)

9. McMillan R. The pathogenesis of chronic immune thrombocytopenic purpura. Semin Hematol. 2007;44(4 Suppl 5):S3-S11.

10. Li J, van der Wal DE, Zhu G, et al. Desialylation is a mechanism of Fc-independent platelet clearance and a therapeutic target in immune thrombocytopenia. Nat Commun. 2015;6:7737.

11. Jansen AJ, Josefsson EC, Rumjantseva V, et al. Desialylation accelerates platelet clearance after refrigeration and initiates GPIbÎą metalloproteinase-mediated cleavage in mice. Blood. 2012;119(5):12631273.

12. Quach ME, Dragovich MA, Chen W, et al. Fc-independent immune thrombocytopenia via mechanomolecular signaling in platelets.

Blood. 2018;131(7):787-796.

13. Poujol C, Bergmeier W, Nurden P, et al. Effect of infusing rat monoclonal antibodies to the murine GPIb-IX-V complex on platelet and megakaryocyte morphology in mice. Platelets. 2003;14(1):35-45.

14. Bender M, Stegner D, Nieswandt B. Model systems for platelet receptor shedding. Platelets. 2017;28(4):325-332.

15. Ravanat C, Freund M, Mangin P, et al. GPV is a marker of in vivo platelet activation--study in a rat thrombosis model. Thromb Haemost. 2000;83(2):327-333.

16. Kahn ML, Diacovo TG, Bainton DF, Lanza F, Trejo J, Coughlin SR. Glycoprotein V-deficient platelets have undiminished thrombin responsiveness and do not exhibit a Bernard-Soulier phenotype.

Blood. 1999; 94(12):4112-4121.

17. Ramakrishnan V, Reeves PS, DeGuzman F, et al. Increased thrombin responsiveness in platelets from mice lacking glycoprotein V. Proc Natl Acad Sci USA 1999;96(23):13336-13341.

18. Moog S, Mangin P, Lenain N, et al. Platelet glycoprotein V binds to collagen and participates in platelet adhesion and aggregation. Blood.

2001;98(4):1038-1046.

19. Nonne C, Hechler B, Cazenave JP, et al. Reassessment of in vivo thrombus formation in glycoprotein V deficient mice backcrossed on a C57Bl/6 strain. J Thromb Haemost. 2008;6(1):210-212.

20. Porcelijn L, Huiskes E, Oldert G, et al. Detection of platelet autoantibodies to identify immune thrombocytopenia: state of the art. Br J Haematol. 2018;182(3):423-426.

1105

PERSPECTIVE ARTICLE Ferrata Storti Foundation

The carrier state for sickle cell disease is not completely harmless Julia Zhe Xu and Swee Lay Thein

Sickle Cell Branch, National Heart Lung and Blood Institute, National Institutes of Health, Bethesda, MA, USA

Introduction

Haematologica 2019 Volume 104(6):1106-1111

Sickle cell disease (SCD) is a clinical syndrome caused by the presence of hemoglobin S (HbS), in which glutamic acid in position 6 of the b chain of hemoglobin is substituted by valine (bGlu6Val). It is generally recognized as an autosomal recessive disorder, in that individuals who have inherited one copy of the HbS allele and one normal HbA allele (i.e. have HbAS or sickle cell trait, SCT), are typically asymptomatic and spared the serious complications associated with possessing two copies of the mutant allele (i.e. HbSS). It is estimated that 300 million people (~5% of the world’s population) carry the HbS allele, and nearly 5.5 million births are affected annually.1 Individuals with SCD die prematurely, but the life expectancy of individuals with SCT is similar to that of people without the trait.2 However, early literature attributed a number of possible disease associations to this heterozygous state, partly as a result of unreliable diagnostic laboratory testing for hemoglobinopathies and partly due to questionable conclusions drawn from uncontrolled observational studies, individual case reports, and small case series.3,4 Nevertheless, while some of the associations historically attributed to SCT are unfounded, recent meta-analyses found high-quality evidence that SCT is indeed a risk factor for a handful of complications common to SCD.2

Exercise in individuals with sickle cell trait Correspondence: SWEE LAY THEIN sl.thein@nih.gov Received: February 1, 2019.

Accepted: April 29, 2019.

Pre-published: May 16, 2019.

doi:10.3324/haematol.2018.206060 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/104/6/1106 ©2019 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

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

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

1106

Sudden death is the most feared complication of SCT. Despite a lack of clear evidence, early concerns about SCT-related sudden death led to the initial adoption of universal SCT screening in the 1970s for all United States (US) Armed Forces recruits and mandatory occupational restrictions for those recruits found to have SCT. The complication of sudden death was reinforced in a study in 1987 which evaluated deaths among two million US military recruits during training; African-American recruits with SCT had a 28-fold increase in relative risk [95% confidence interval (CI): 9-100] of sudden unexplained death compared to those without SCT.5 Although the relative risk was seemingly very high, the study was limited by the small absolute number of deaths in each group and its inability to examine separately subsets of sudden unexplained death (cardiac, exertional heat stroke, heat stress, and rhabdomyolysis), which may not share the same disease mechanism. Various other studies have examined physiological responses to exercise – e.g. aerobic metabolism, energy expenditure, maximal oxygen consumption, and maximal exercise performance – in subjects with HbAS and have found no difference compared to those in subjects with normal HbAA,6-9 even when intravascular sickling in subjects with SCT is observed. Additionally, a recent longitudinal analysis of a large cohort found no association of SCT with fitness, or with the development of hypertension, diabetes, and metabolic syndrome.7 On the civilian side, the sudden death of a college football player during training in 2006 led to a lawsuit whose settlement prompted the National Collegiate Athlete Association (NCAA) to adopt universal SCT screening for incoming Division I athletes in 2010, a policy that was later extended to Division II and III athletes.10 A retrospective analysis of athletes in the NCAA also found a 37-fold increase of exertional death in Division I football players with SCT compared to those without, but again the absolute risk was low.11 A more recent retrospective study of 47,944 black soldiers showed that with adoption of universal preventive measures, the risk of sudden death attributed to SCT appears to be completely mitigated.12 An outcome of these events and reports is the recommendation that preventive measures should be universally adopted to protect all soldiers and athletes, rather than singling out individuals with SCT with mandatory screening practices.4 Indeed, there are a handful of haematologica | 2019; 104(6)

Sickle cell trait is not completely harmless

studies showing that hydration and progressive exercise training may reverse various hematologic abnormalities observed during exertion in SCT, such as increased blood viscosity, red blood cell rigidity, oxidative stress, endothelial activation, and red blood cell sickling.12-17 Nonetheless, soldiers with SCT appear to be at higher risk of developing exertional rhabdomyolysis;12 the risk is modest (hazard ratio, 1.54; 95% CI: 1.12-2.12), under conditions of extreme exertion, as also supported by a recent systematic review.2

Renal manifestations of sickle cell trait The renal medulla with its unique hyperosmolar and acidotic environment, together with the low oxygen tension and low-flow state in medullary vasa recta, create optimal conditions for HbS polymerization,3 leading to a reduced number of vasa recta and loss of its normal vascular architecture.18 Hyposthenuria (a defect in concentrating urine) results, and in turn may predispose individuals to dehydration, postulated to be a contributory factor to the exertional rhabdomyolysis associated with SCT. Among the renal manifestations associated with SCT, the most common are hematuria and proteinuria. In one large-scale study, hematuria was found to be twice as common in hospitalized African-American patients with HbAS than in those with normal hemoglobin.19 Urographic evaluation in a separate case series revealed that ~50% of hematuria cases in individuals with SCT were related to renal papillary necrosis.1,20 A much rarer cause of hematuria is renal medullary carcinoma, an aggressive malignancy found almost exclusively in young black patients with SCT.21 The tumor is hypothesized to arise from the distal collecting duct epithelium as a result of abnormal proliferation stimulated by chronic ischemia.22 The prognosis of patients with renal medullary carcinoma is poor, with a typical median survival of <1 year, and response to traditional chemotherapy is limited.23 It has been proposed that individuals with SCT presenting with new-onset hematuria should undergo urological evaluation.24 The chronic microvascular damage in the renal medulla also predisposes individuals with SCT to proteinuria and

chronic kidney disease.2 A large study combining data from several African-American population-based prospective cohorts showed that the presence of SCT imparts a 1.86-fold higher odds of albuminuria (95% CI: 1.49-2.31).25 SCT is also a recognized risk factor for chronic kidney disease, for which a 1.5- to 2-fold increased risk is attributed to SCT.25,26 It is important to note that several co-morbid conditions such as type 2 diabetes and hypertension, as well as co-inherited genetic risk factors could influence the risk of chronic kidney disease, which is a potential explanation for the lack of association of SCT and chronic kidney disease in smaller cohort studies involving different ethnic groups.27,28 Additionally, SCT may increase the risk of proteinuria and retinopathy in individuals with diabetes.2,29,30 A common genetic risk modifier of renal disease in the African-American population is APOL1; inheritance of the G1 and G2 risk alleles is believed to account for much of the excess risk of chronic kidney disease and end-stage renal disease in individuals of African ancestry overall.31 APOL1 risk alleles have also been associated with proteinuria in patients with SCD,32,33 but so far, no genetic interactions between SCT and APOL1 risk alleles have been observed.25,26,34 The evidence for SCT itself being a risk factor for the development of end-stage renal disease remains inconclusive.

Another well-known genetic modifier of disease severity in SCD is Îą-thalassemia trait, co-inherited in ~35% of individuals of African descent. Co-inherited Îą-thalassemia reduces intracellular HbS concentration, a key determinant of polymerization kinetics. The protective effect of Îą-thalassemia on anemia and chronic kidney disease in individuals with SCT has been demonstrated in a cohort from the Jackson Heart Study.35

Other complications related to sickle cell trait Despite numerous reported associations with SCT, few complications are supported by strong evidence (Figure 1).

One final strong association is that of venous thromboembolic disease and pulmonary embolism in particular.2 In a study of 65,154 hospitalized African-Americans, HbS carriers had a slightly higher relative risk of pulmonary embolism, but the study was limited by the lack of radio-

Figure 1. The strength of association of sickle cell trait with various complications reported in the literature.

haematologica | 2019; 104(6)

1107

J.Z. Xu et al.

Figure 2. Aids to unraveling the molecular basis of sickle complications in individuals with purported sickle cell “trait”. Individuals with purported sickle cell trait suffering complications of sickle cell disease may have an unrecognized rare genetic alteration or co-inherited red cell disorder and present a diagnostic challenge.

The first step is to obtain a detailed clinical and family history, and to perform family studies if family members are available. Detailed hematologic evaluation, including hemoglobin electrophoresis, high-performance liquid chromatography (HPLC), examination of the peripheral blood smear, and qualitative sickle solubility assay, are essential. The hematologic data should always be interpreted in conjunction with genetic data. HbA in excess of HbS validates heterozygosity for the bS allele and, if found, should prompt investigation into whether the bS allele could be dominantly inherited with a double mutation. The patient could also have co-inherited other genetic variants [e.g. pyruvate kinase (PK) deficiency] that increase the likelihood of HbS polymerization. If HbS is in excess of HbA, and the inheritance pattern from parents is consistent with HbAS, somatic mosaicism should be considered. If hemoglobin electrophoresis or HPLC shows only HbS, genetic testing shows heterozygosity for HbS, and only one parent has HbAS, one should consider the possibility of a “functional homozygote” with the trans b gene structurally intact but functionally inactivated, such as can be seen in deletion of the trans upstream b locus control region.

logical confirmation.19 Subsequent large cohort studies have identified SCT as a risk factor for venous thromboembolism36 as well as pulmonary embolism, but not deep vein thrombosis.36-38 One study suggested that the risk of venous thromboembolism attributable to SCT among blacks is higher than the risk attributable to the prothrombin G20210A mutation among whites.36 However, the reason that HbS might predispose a subject to pulmonary embolism over deep vein thrombosis is unknown and merits further investigation.

A number of other reported associations – e.g. splenic infarction, pregnancy complications, acute chest syndrome, retinopathy and traumatic hyphema – are backed by at times significant anecdotal evidence and have been reviewed in recent publications.1,2,4 This perspective article is limited in its ability to explore all reported associations in depth, but it is important to note that interpretation of these associations may not be straightforward. For example, the occurrence of splenic infarction in individuals with SCT has been documented in a number of case reports, often but not always in men exposed to high altitudes.39-41 However, as discussed in the next section, detailed genetic testing was not performed in these cases, so it is not possible to draw the conclusion that the splenic infarction was a complication of SCT alone.

SCT has also been reported to be associated with a variety of maternal and fetal complications during preg1108

nancy or the puerperium, but the existing evidence is conflicting.1,4 Some of these associations are concerning, such as potentially higher rates of pre-eclampsia, maternal infections, and fetal loss.42,43 However, in a prospective Nigerian cohort, pregnant women with SCT did not experience more morbidity than women with HbAA, and in fact, had fewer attacks of malaria during pregnancy.44 In yet another study the observed risk of venous thromboembolism was higher in pregnant subjects with SCT (relative risk, 1.6; 95% CI: 0.5-5.5),45 but the magnitude of the difference did not suggest that the risk of venous thromboembolism was increased above that associated with SCT alone.

A recent analysis found moderate-quality evidence for a null association between SCT and low pediatric height and weight, as well as between SCT and heart failure/cardiomyopathy and stroke.2 In a field in which many questions remain unanswered, it is crucial to recognize and promote such null findings in order to prevent unnecessary concern, unfounded stigmatization, and psychological harm to carriers of HbS.

Molecular factors worsening the HbAS phenotype The key event in the pathophysiology of SCD is polymerization of the deoxygenated HbS, which under cerhaematologica | 2019; 104(6)

Sickle cell trait is not completely harmless

tain conditions can be irreversible, leading to distortion of the erythrocyte and loss of deformability, ultimately causing vaso-occlusion in the microvasculature and hemolysis. The tendency for HbS to polymerize is highly dependent on the hemoglobin composition in the erythrocyte – mainly the concentration of intracellular HbS, as well as the concentration and type of non S hemoglobin. HbA and HbF present in the cell reduce the concentration of HbS, but HbF additionally has an inhibitory effect on HbS polymerization. Individuals with SCT, who have HbA in excess of HbS (30-40% of the total hemoglobin), do not typically suffer from the effects of sickling and are asymptomatic. Apart from the intracellular hemoglobin composition, other factors that influence the likelihood of HbS polymerization include oxygen saturation, intracellular pH and 2,3-diphosphoglycerate levels.

Environmental and co-inherited genetic factors can change these parameters, modulating the kinetics of HbS polymerization in individuals with SCT. Clinically symptomatic HbAS individuals could be considered to fall under four genetic categories: (i) co-inheritance of HbS with a genetic modifier; (ii) dominant forms of HbS alleles; (iii) apparent heterozygosity for HbS; and (iv) non-Mendelian inheritance of HbS.

Co-inheritance of HbS with a genetic modifier Compound heterozygosity for HbS and genetic variants causing non-hemoglobin red cell (i.e. membrane and enzyme) disorders, while uncommon, can have important modifying effects on the clinical outcome of SCT. Two cases of HbS carriers experiencing typical SCD complications of chronic hemolytic anemia, recurrent acute pain, recalcitrant leg ulceration and end-stage renal disease, have been reported.46,47 Each of the probands had coinherited mutations in the PKLR gene, causing a deficiency of pyruvate kinase protein. Pyruvate kinase is a key enzyme in the final step of glycolysis; it converts phosphoenolpyruvate to pyruvate, generating 50% of the total red cell ATP that is essential for metabolism of the red blood cell. Of particular relevance to SCD, a reduction in pyruvate kinase activity also leads to accumulation of the upstream enzyme substrates, including 2,3-diphosphoglycerate which decreases oxygen affinity, favoring polymerization of deoxy-HbS. In one case, fresh blood was available for in-vitro functional assays that supported the role of 2,3-diphosphoglycerate as a key factor in HbS polymerization.46 Another example is the co-inheritance of hereditary spherocytosis, implicated as the cause of splenic infarction in a handful of SCT cases.48 The presence of hereditary spherocytosis is believed to increase mean corpuscular hemoglobin concentration and intracellular HbS concentration, thereby increasing the propensity to HbS polymerization. Along the same lines, glucose-6-phosphate dehydrogenase deficiency, a common red cell enzyme disorder in individuals of African ancestry, might be expected to increase hemolysis and modify the clinical phenotype of SCD and SCT. However, there is a lack of evidence addressing co-inheritance of SCT and glucose-6phosphate dehydrogenase deficiency, and even the evidence for SCD is mixed, with no consistent differences in red cell indices, degree of anemia, hemolysis, frequency of vaso-occlusive complications, or stroke risk in subjects who have co-inherited glucose-6-phosphate dehydrogenase deficiency.49-51 These conflicting reports could be haematologica | 2019; 104(6)

related to limitations of the methodologies used for the enzyme assays, or the panel of glucose-6-phosphate dehydrogenase variants genotyped. These reports serve as an important caveat when examining past anecdotal evidence of complications in SCT, as in-depth genetic investigation is often lacking in reported cases of symptomatic SCT.

Dominant forms of HbS alleles Rare HbS alleles that behave dominantly have been reported. One example is that found in a 19-month old girl, identified as having SCT through newborn screening; she developed a vaso-occlusive crisis with splenic sequestration during a flight and required splenectomy.52 Subsequent DNA sequencing analysis confirmed that the baby had inherited a maternal normal HbA allele but had acquired a new mutation, bLeu68Phe, cis to the paternal HbS allele. bLeu68Phe has previously been identified as Hb Rockford, a known Hb variant with reduced affinity for oxygen. The double mutant, named Hb Jamaica Plain (JP, bGlu6Val,Leu68Phe), has similar electrophoretic motility as HbS, hence it was missed at the newborn screening. The bLeu68Phe mutation in Hb JP causes it to desaturate easily at lower oxygen tension, thus polymerizing more readily than typical HbS, converting it to a dominant mutation.52 Two other Hb variants with double mutations, HbS Antilles (bGlu6Val, Val23Ile) and HbS-Sao Paolo (bGlu6Val, Lys65Glu), also promote polymerization through reduced oxygen affinity in heterozygotes.53,54 In addition, HbS Antilles has lower solubility and HbS-Sao Paolo forms more stable polymers than HbS, potentially further enhancing irreversible polymerization and red blood cell sickling. In the case of HbS Antilles, family studies led to the identification of 24 HbA/S Antilles individuals in the proband’s family, many of whom had recurrent sickle pain crises, chronic hemolytic anemia, and splenomegaly,54 a phenotype similar to that of HbSC disease.

Apparent heterozygosity for HbS

Compound heterozygotes of bS and very mild b-thalassemia mutations (bS/b++-thalassemia) can appear as having SCT, but the giveaway here is an excess of HbS over HbA. A 38-year old, previously healthy man, presented with a 6-month history of worsening pruritis, jaundice and ascites. Extensive work-up for causes of liver disease was negative, but hemoglobin electrophoresis showed 49.6% HbS and 41.3% HbA. The patient had not received any blood transfusion. While this result could easily be misinterpreted as HbS trait, given the slight increase of HbS over HbA, DNA and family studies were pursued. He was revealed as a compound heterozygote for bS and a novel, very mild b-thalassemia mutation (b IVS2-844 C→A) that was transmitted to both of his sons, and the liver pathology was ascribed as sickle-related.55 Another diagnostic conundrum is the discrepant findings of HbSS on hemoglobin electrophoresis but HbAS on genetic testing. Two such cases have been reported, where the individuals presented with typical hematologic and clinical phenotypes of SCD.56 In both instances, DNA testing showed that the individual possessed one HbA and one HbS allele, but expression of the HbA allele was abolished by a deletion of the upstream b locus control region, resulting in sole expression of the HbS allele and, thereby, a functionally homozygous HbS phenotype.56 1109

J.Z. Xu et al.

Non-Mendelian inheritance of HbS Homozygosity due to uniparental disomy of chromosome 11 is another rare genetic defect. Two individuals were reported to have inherited HbS from one parent and normal HbA from the other parent but presented later in life with phenotypic SCD. In-depth DNA analysis revealed that post-zygotic mitotic recombination had occurred, leading to mosaic segmental isodisomy.57,58 These individuals had dual populations of HbSS and HbAS erythroid progenitors and peripheral red blood cells, with HbSS erythrocytes accounting for the chronic hemolytic phenotype of SCD. These cases demonstrate that the diagnosis of sickle trait can be nuanced and deserves further workup if the individual has a phenotype of SCD. In such diagnostically challenging scenarios, family studies are extremely useful, and detailed hemoglobin evaluation is essential (Figure 2).

A relative excess of HbA over HbS with no other Hb variant present indicates that only one copy of abnormal

References 1. Tsaras G, Owusu-Ansah A, Boateng FO, Amoateng-Adjepong Y. Complications associated with sickle cell trait: a brief narrative review. Am J Med. 2009;122(6):507-512.

2. Naik RP, Smith-Whitley K, Hassell KL, et al.

Clinical outcomes associated with sickle cell trait: a systematic review. Ann Intern Med. 2018;169(9):619-627.

3. Key NS, Derebail VK. Sickle-cell trait: novel clinical significance. Hematology Am Soc Hematol Educ Program. 2010;2010:418-422.

4. Naik RP, Haywood C Jr. Sickle cell trait diagnosis: clinical and social implications.

Hematology Am Soc Hematol Educ Program. 2015;2015:160-167.

5. Kark JA, Posey DM, Schumacher HR, Ruehle CJ. Sickle-cell trait as a risk factor for sudden death in physical training. N Engl J Med. 1987;317(13):781-787.

6. Connes P, Reid H, Hardy-Dessources MD, Morrison E, Hue O. Physiological responses of sickle cell trait carriers during exercise.

Sports Med. 2008;38(11):931-946.

7. Liem RI, Chan C, Vu TT, et al. Association among sickle cell trait, fitness, and cardiovascular risk factors in CARDIA. Blood.

2017;129(6):723-728.

8. Martin TW, Weisman IM, Zeballos RJ, Stephenson SR. Exercise and hypoxia increase sickling in venous blood from an exercising limb in individuals with sickle cell trait. Am J Med. 1989;87(1):48-56.

9. Monchanin G, Connes P, Wouassi D, et al.

Hemorheology, sickle cell trait, and alphathalassemia in athletes: effects of exercise.

Med Sci Sports Exerc. 2005;37(7):1086-1092.

10. Ferrari R, Parker LS, Grubs RE, Krishnamurti L. Sickle cell trait screening of collegiate athletes: ethical reasons for program reform. J Genet Couns. 2015;24(6):873-877.

11. Harmon KG, Drezner JA, Klossner D, Asif IM. Sickle cell trait associated with a RR of death of 37 times in National Collegiate Athletic Association football athletes: a database with 2 million athlete-years as the denominator. Br J Sports Med. 2012;46(5): 325-330.

12. Nelson DA, Deuster PA, Carter R, 3rd, Hill OT, Wolcott VL, Kurina LM. Sickle cell trait,

1110

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

allele is present. If phenotypic and genetic workup both show HbAS, a search for modifiers that promote HbS polymerization should be undertaken.

Conclusion While HbAS represents an asymptomatic carrier state, clinical and epidemiological studies have shown that SCT is certainly not an entirely harmless condition. The presence of HbS in SCT may contribute to specific disease processes, particularly under extreme conditions that promote HbS polymerization. Additionally, individuals with HbAS can present with complications typical of the SCD phenotype. Such cases of unusually severe HbAS can pose a diagnostic challenge, but elucidating their molecular basis provides further insight into the pathophysiology of SCD and help to identify genetic risk modifiers in SCT.

rhabdomyolysis, and mortality among U.S. srmy soldiers. N Engl J Med. 2016;375 (5):435-442.

Aufradet E, Monchanin G, Oyonno-Engelle S, et al. Habitual physical activity and endothelial activation in sickle cell trait carriers. Med Sci Sports Exerc. 2010;42(11): 1987-1994.

Bergeron MF, Cannon JG, Hall EL, Kutlar A. Erythrocyte sickling during exercise and thermal stress. Clin J Sport Med. 2004;14(6):354-356.

Chirico EN, Martin C, Faes C, et al. Exercise training blunts oxidative stress in sickle cell trait carriers. J Appl Physiol. 2012;112 (9):1445-1453.

Diaw M, Samb A, Diop S, et al. Effects of hydration and water deprivation on blood viscosity during a soccer game in sickle cell trait carriers. Br J Sports Med. 2014;48 (4):326-331.

Tripette J, Loko G, Samb A, et al. Effects of hydration and dehydration on blood rheology in sickle cell trait carriers during exercise.

Am J Physiol Heart Circ Physiol. 2010;299 (3):H908-914.

Nath KA, Hebbel RP. Sickle cell disease: renal manifestations and mechanisms. Nat Rev Nephrol. 2015;11(3):161-171.

Heller P, Best WR, Nelson RB, Becktel J. Clinical implications of sickle-cell trait and glucose-6-phosphate dehydrogenase deficiency in hospitalized black male patients. N Engl J Med. 1979;300(18):1001-1005.

Eckert DE, Jonutis AJ, Davidson AJ. The incidence and manifestations of urographic papillary abnormalities in patients with S hemoglobinopathies. Radiology. 1974;113 (1):59-63.

Alvarez O, Rodriguez MM, Jordan L, Sarnaik S. Renal medullary carcinoma and sickle cell trait: a systematic review. Pediatr Blood Cancer. 2015;62(10):1694-1699.

Kiryluk K, Jadoon A, Gupta M, Radhakrishnan J. Sickle cell trait and gross hematuria. Kidney Int. 2007;71(7):706-710.

Watanabe IC, Billis A, Guimaraes MS, et al.

Renal medullary carcinoma: report of seven cases from Brazil. Mod Pathol. 2007;20 (9):914-920.

Pecker LH, Naik RP. The current state of

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

sickle-cell trait: implications for reproductive and genetic counseling. Blood. 2018 Nov 28.

[Epub ahead of print] Naik RP, Derebail VK, Grams ME, et al.

Association of sickle cell trait with chronic kidney disease and albuminuria in African Americans. JAMA. 2014;312(20):2115-2125.

Naik RP, Irvin MR, Judd S, et al. Sickle cell trait and the risk of ESRD in blacks. J Am Soc Nephrol. 2017;28(7):2180-2187.

Dueker ND, Della-Morte D, Rundek T, Sacco RL, Blanton SH. Sickle cell trait and renal function in Hispanics in the United States: the Northern Manhattan Study. Ethn Dis. 2017;27(1):11-14.

Mukendi K, Lepira FB, Makulo JR, Sumaili KE, Kayembe PK, Nseka MN. Sickle cell trait is not associated with chronic kidney disease in adult Congolese patients: a clinicbased, cross-sectional study. Cardiovasc J Afr. 2015;26(3):125-129.

Ajayi AA, Kolawole BA. Sickle cell trait and gender influence type 2 diabetic complications in African patients. Eur J Intern Med. 2004;15(5):312-315.

Lonsdorfer A, Comoe L, Yapo AE, Lonsdorfer J. Proteinuria in sickle cell trait and disease: an electrophoretic analysis. Clin Chim Acta. 1989;181(3):239-247.

Genovese G, Friedman DJ, Ross MD, et al.

Association of trypanolytic ApoL1 variants with kidney disease in African Americans.

Science. 2010;329(5993):841-845.

Ashley-Koch AE, Okocha EC, Garrett ME, et al. MYH9 and APOL1 are both associated with sickle cell disease nephropathy. Br J Haematol. 2011;155(3):386-394.

Saraf SL, Shah BN, Zhang X, et al. APOL1, alpha-thalassemia, and BCL11A variants as a genetic risk profile for progression of chronic kidney disease in sickle cell anemia.

Haematologica. 2017;102(1):e1-e6.

Hicks PJ, Langefeld CD, Lu L, et al. Sickle cell trait is not independently associated with susceptibility to end-stage renal disease in African Americans. Kidney Int. 2011;80(12): 1339-1343.

Raffield LM, Ulirsch JC, Naik RP, et al.

Common alpha-globin variants modify hematologic and other clinical phenotypes in sickle cell trait and disease. PLoS Genet.

haematologica | 2019; 104(6)

Sickle cell trait is not completely harmless

2018;14(3):e1007293.

36. Austin H, Key NS, Benson JM, et al. Sickle cell trait and the risk of venous thromboembolism among blacks. Blood. 2007;110(3): 908-912.

37. Bucknor MD, Goo JS, Coppolino ML. The risk of potential thromboembolic, renal and cardiac complications of sickle cell trait.

Hemoglobin. 2014;38(1):28-32.

38. Folsom AR, Tang W, Roetker NS, et al.

Prospective study of sickle cell trait and venous thromboembolism incidence. J Thromb Haemost. 2015;13(1):2-9.

39. Sheikha A. Splenic syndrome in patients at high altitude with unrecognized sickle cell trait: splenectomy is often unnecessary. Can J Surg. 2005;48(5):377-381.

40. Seegars MB, Brett AS. Splenic infarction associated with sickle cell trait at low altitude. Hematology. 2015;20(10):607-609.

41. Goodman J, Hassell K, Irwin D, Witkowski EH, Nuss R. The splenic syndrome in individuals with sickle cell trait. High Alt Med Biol. 2014;15(4):468-471.

42. Larrabee KD, Monga M. Women with sickle cell trait are at increased risk for preeclampsia. Am J Obstet Gynecol. 1997;177(2):425428.

43. Hamdi IM, Karri KS, Ghani EA. Pregnancy outcome in women with sickle cell trait.

Saudi Med J. 2002;23(12):1455-1457.

44. Adeyemi AB, Adediran IA, Kuti O, Owolabi AT, Durosimi MA. Outcome of pregnancy in a population of Nigerian women with sickle cell trait. J Obstet Gynaecol.

haematologica | 2019; 104(6)

2006;26(2):133-137.

45. Porter B, Key NS, Jauk VC, Adam S, Biggio J, Tita A. Impact of sickle hemoglobinopathies on pregnancy-related venous thromboembolism. Am J Perinatol. 2014;31(9):805-809.

46. Cohen-Solal M, Prehu C, Wajcman H, et al.

A new sickle cell disease phenotype associating Hb S trait, severe pyruvate kinase deficiency (PK Conakry), and an alpha2 globin gene variant (Hb Conakry). Br J Haematol.

1998;103(4):950-956.

47. Alli N, Coetzee M, Louw V, et al. Sickle cell disease in a carrier with pyruvate kinase deficiency. Hematology. 2008;13(6):369-372.

48. Ustun C, Kutlar F, Holley L, Seigler M, Burgess R, Kutlar A. Interaction of sickle cell trait with hereditary spherocytosis: splenic infarcts and sequestration. Acta Haematol.

2003;109(1):46-49.

49. Bernaudin F, Arnaud C, Kamdem A, et al.

Biological impact of alpha genes, beta haplotypes, and G6PD activity in sickle cell anemia at baseline and with hydroxyurea.

Blood Adv. 2018;2(6):626-637.

50. Benkerrou M, Alberti C, Couque N, et al.

Impact of glucose-6-phosphate dehydrogenase deficiency on sickle cell anaemia expression in infancy and early childhood: a prospective study. Br J Haematol.

2013;163(5):646-654.

51. Bouanga JC, Mouele R, Prehu C, Wajcman H, Feingold J, Galacteros F. Glucose-6-phosphate dehydrogenase deficiency and homozygous sickle cell disease in Congo.

Hum Hered. 1998;48(4):192-197.

52. Geva A, Clark JJ, Zhang Y, Popowicz A, Manning JM, Neufeld EJ. Hemoglobin Jamaica Plain--a sickling hemoglobin with reduced oxygen affinity. N Engl J Med. 2004;351(15):1532-1538.

53. Jorge SE, Petruk AA, Kimura EM, et al. Hb SSao Paulo: a new sickling hemoglobin with stable polymers and decreased oxygen affinity. Arch Biochem Biophys. 2012;519(1):2331.

54. Monplaisir N, Merault G, Poyart C, et al.

Hemoglobin S antilles: A variant with lower solubility than hemoglobin S and producing sickle cell disease in heterozygotes. Proc Natl Acad Sci U S A. 1986;83(24):9363-9367.

55. Cross TJ, Berry PA, Akbar N, Wendon J, Thein SL, Harrison PM. Sickle liver diseasean unusual presentation in a compound heterozygote for HbS and a novel beta-thalassemia mutation. Am J Hematol. 2007;82 (9):852-854.

56. Koenig SC, Becirevic E, Hellberg MS, et al.

Sickle cell disease caused by heterozygosity for Hb S and novel LCR deletion: report of two patients. Am J Hematol. 2009;84(9):603606.

57. Swensen JJ, Agarwal AM, Esquilin JM, et al.

Sickle cell disease due to uniparental disomy in a child who inherited sickle cell trait.

Blood. 2010;116(15):2822-2825.

58. Vinatier I, Martin X, Costa JM, Bazin A, Giraudier S, Joly P. A late onset sickle cell disease reveals a mosaic segmental uniparental isodisomy of chromosome 11p15.

Blood Cells Mol Dis. 2015;54(1):53-55.

1111

REVIEW ARTICLE Ferrata Storti Foundation

Thrombopoietin receptor agonists: ten years later

Waleed Ghanima,1,2 Nichola Cooper,3 Francesco Rodeghiero,4 Bertrand Godeau5 and James B. Bussel6

Departments of Medicine, Hematology-Oncology and Research, Østfold Hospital Trust, Norway; 2Department of Hematology, Institute of Clinical Medicine, University of Oslo, Norway; 3Department of Medicine, Hammersmith Hospital, Imperial College, London, UK; 4 Hematology Project Foundation and Department of Cell Therapy and Hematology, S. Bortolo Hospital, Vicenza, Italy; 5Department of Internal Medicine, Henri Mondor University Hospital, Assistance Publique-Hopitaux de Paris, UPEC, Créteil, France and 6Department of Pediatrics, Weill Cornell Medicine, New York, NY, USA 1

Haematologica 2018 Volume 104(6):1112-1123

ABSTRACT

T

Correspondence: WALEED GHANIMA waleed.ghanima@SO-HF.no Received: November 28, 2018.

Accepted: March 11, 2019.

Pre-published: May 9, 2019.

doi:10.3324/haematol.2018.212845 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/104/6/1112 ©2019 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

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

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

1112

The two thrombopoietin receptor agonists (TPO-RA), eltrombopag and romiplostim, were licensed in the US for treatment of immune thrombocytopenia (ITP) in 2008 and, since then, their use has progressively increased around the world; they are currently used in more than 100 countries. The six largest randomized controlled trials conducted in ITP have used one of these two agents. All studies have demonstrated a platelet response rate between 50-90%, depending on the criteria used, with good safety and tolerability. TPO-RA were shown to be effective in reducing bleeding and the need for concomitant or rescue medication.

Many other investigations of their mechanism of effect, prospective and retrospective trials, and studies focusing on toxicity have been performed widening our knowledge of these two agents. Initial concerns on issues such as myelofibrosis have not been confirmed. Only a small number of patients develop moderate-severe reticulin fibrosis and/or collagen fibrosis; however, these are usually reversed after discontinuation of TPO-RA. Studies indicate, however, that TPO-RA may increase the risk of venous thromboembolism. Both TPO-RA are currently approved in patients with chronic ITP aged >1-year who are refractory to at least one other treatment. Eltrombopag has acquired two additional indications: severe aplastic anemia refractory to first-line treatment and hepatitis C patients undergoing treatment with interferon-ribavirin. Despite these wide-ranging studies, important questions still need to be answered. This summary review on TPO-RA will summarize what is known regarding efficacy in ITP, evaluate safety concerns in more depth, and focus on the questions that remain.

Introduction Over the last 20 years, and before the regular availability of thrombopoietin receptor agonists (TPO-RA), the most commonly used second-line treatments for patients with immune thrombocytopenia (ITP) were splenectomy and rituximab.

Both options have the potential to provide a cure. However, long-term responses are not completely satisfactory (60% after splenectomy and only 20% 2-5 year long-term responses after rituximab).1,2 Adverse events following these interventions are also significant, if uncommon: post-operative morbidity and increased risk of infections and thromboembolism (TE) after splenectomy, and very rare cases of progressive multifocal leukoencephalopathy (PML) and slight increased infectious rates after rituximab.3 The two TPO-RA, romiplostim and eltrombopag, represent a completely different approach to ITP; they both have a very good chance of supporting the platelet count with undemanding daily or weekly treatment. Their goal is to support the patient’s platelet count until adequate levels are achieved and treatment is no haematologica | 2019; 104(6)

Thrombopoietin receptor agonists in ITP

longer required. The TPO-RA were licensed in the US for the treatment of ITP in 2008, and, since then, their use has progressively increased around the world; they are currently used in more than 100 countries. Their introduction heralded a paradigm shift in the treatment of ITP. They are now widely used and many hematologists are well-acquainted with them. This is the 10-year anniversary of their licensure in the US for ITP and it seems appropriate to review the state of the art of these agents: what is known about their mechanism of effect, efficacy, and toxicity, and what remains to be learned, including an exploration of other clinical situations in which they might be useful.

Mechanism of action Romiplostim and eltrombopag both bind to the thrombopoietin (TPO) receptor, causing conformational change in the TPO receptor, activation of the JAK2/STAT5 pathway, and a resulting increased megakaryocyte progenitor proliferation and increased platelet production.4,5

However, there are some differences between the two agents (Figure 1). Romiplostim is a peptibody that binds directly and competitively at the TPO binding site, whereas eltrombopag is a small molecule which binds at a transmembrane site. There are also differences in the activation of other signaling pathways in megakaryocytes (MK) such as STAT3, ERK and AKT (Table 1).6-8 Furthermore, romiplostim mostly stimulates mature precursors, while eltrombopag appears to act earlier in the pathway, stimulating MK precursor cells and MK differentiation.4,6 In addition to differences in TPO-receptor activation, eltrombopag also has off-target effects. For example, eltrombopag chelates both extra- and intra-cellular calcium and iron and can shuttle iron out of cells.9 The ironchelating action of eltrombopag causes anti-proliferative effects on leukemic cells lines,10 and a TPO-independent effect on stimulating stem cells and MK precursors in vivo.

These differences may explain why some patients respond to one agent and not the other,11,12 and why treatment with both agents can be useful in very refractory patients.

Although the prime mechanism of action of the TPO-

Figure 1. Cellular mechanisms of action of thrombopoietin (TPO) and of thrombopoietin receptor agonists (TPO-RA). Binding of the ligand (TPO/TPO-RA) to the cMPL receptor on the megakaryocyte causes conformational change in the receptor, resulting in downstream activation of the various signaling pathways including JAK2/STAT5, PI3K/AKT, ERK, ultimately resulting in increased platelet production. Various pathways can be activated by the different substances (see also Table 1).

GRB2: growth factor receptor-binding protein 2; JAK: Janus kinase; MAPK: mitogen-activated protein kinase; P: phosphorylation; RAF: rapidly accelerated fibrosarcoma kinase; RAS: rat sarcoma GTPase; SHC: Src homology collagen protein; STAT: signal transducer and activator of transcription; PI3K: phosphatidylinositol 3kinases; ERK: extracellular-signal-regulated kinase.

haematologica | 2019; 104(6)

1113

W. Ghanima et al.

RA is thought to be due to increased platelet production, both TPO-RA have also been described to have immunomodulatory effects, with increased regulatory Tand B-cell effects in patients on TPO-RA.13 This effect has been suggested to be mediated by TGF-B, a major cytokine involved in T-regulatory (Treg) cell development, and found in abundance in MK and platelets.14,15 Alternatively, TPO-RA may also affect antigen processing and presentation by MK.16 Whether these potential immunomodulatory effects result in the treatment-free durable responses reported with both TPO-RA has not yet been understood.

Efficacy of romiplostim and eltrombopag Platelet response The response rate of these agents depends on the definition of “response.” If a consistent “durable” platelet count response is required, then the response rate may be 40-60%. This type of response, with platelet counts consistently higher than 50x109/L without bleeding and/or need for rescue therapy, is a realistic goal for patients with ITP. If, however, a “response” is a single platelet count over 50x109/L during a finite period of time, then the response rate is closer to 60-90%.17-23 Table 2 summarizes the effect of the two agents in randomized controlled trials (RCT) performed in adult ITP patients. The rate for a durable response in the pivotal trials for romiplostim was around 60% in non-splenectomized patients but lower in previously splenectomized patients. Splenectomized patients treated with eltrombopag in the Randomized Placebo-Controlled Idiopathic Thrombocytopenic Purpura (RAISE) and Eltrombopag Extended Dosing (EXTEND) studies also had lower response rates than non-splenectomized patients. While splenectomized patients and those with platelet counts <15x109/L respond less well to both romiplostim18 and eltrombopag,24,25 there is still good evidence of the effect of these agents in these patients.

A recent meta-analysis, which included 1126 patients from 13 RCT performed in eight adult and five pediatric ITP populations, showed that TPO-RA significantly increased platelet response by 3-fold and durable response rates by almost 8-fold as compared to placebo or Standard of Care (SoC).26 In adult studies, a 3-fold increase in response [Risk Ratio (RR): 3.1, 95% Confidence Interval (CI): 2.0-5.0] and 7.5-fold increase in durable response (RR: 7.4, 95%CI: 3.2-17.1) was seen.26 Therefore, while this is a very effective approach to treatment of ITP, not all patients will have a clinically meaningful response to TPO-RA, whereas some have to discontinue TPO-RA because of the lack of response.20,24,27 In the EXTEND study, of the 302 patients enrolled, 55% withdrew from the study due to adverse events (14%), patient decision (13%), lack of efficacy (11%), or other reasons,20,24,27 whereas, in the long-term romiplostim study, 31% of the 292 enrolled patients discontinued because of patient decision (27%), adverse events (12%), alternative therapy (12%), or for other reasons.20,24,27

Reduction in bleeding and concomitant medications The meta-analysis showed that TPO-RA significantly reduced incidences of any or severe bleeding events (RR: 0.8, 95%CI: 0.7-0.9; RR: 0.5, 95%CI: 0.3-0.99, respectively).26 Especially with eltrombopag, there were substantial reductions in any or severe bleeding events in treated patients compared with controls (RR: 0.7, 95%CI: 0.5-0.9; and RR: 0.3, 95%CI: 0.1-1.0, respectively). In parallel with reduced bleeding episodes, pooled results of eight studies indicated a significant reduction in the need for rescue medications in the TPO-RA groups compared with control groups (RR: 0.5, 95%CI: 0.4-0.6).26 Treatment studies with both agents have also demonstrated an ability to reduce or stop concomitant medications (RR: 1.8, 95%CI: 1.1-3.0).

Health-related quality of life and thrombopoietin treatment Health-related quality of life (HRQoL) was studied in many of the RCT and extension studies conducted with

Table 1. Characteristics and down-stream effect of eltrombopag, romiplostim and endogenous thrombopoietin.

Structure and discovery

Binding site

Effect on endogenous thrombopoietin Demonstrated down-stream effect on various signal pathways*

Effect on MK Off-target effect

Eltrombopag

Romiplostim

Small molecule discovered by screening libraries of small molecules that stimulate the TPO receptor down-stream pathways Binds the transmembrane and juxtamembrane domains of the TPO receptor No displacement of TPO, may be additive

Peptibody developed by screening peptide libraries for sequences that can stimulate the TPO receptor

JAK2/STAT5 PI3K/Akt ERK

Thrombopoietin

Binds to the extracytoplasmic domain of the TPO receptor in same manner as TPO Can displace TPO from its receptor

Acts at TPO binding site

JAK2/STAT5 PI3K/Akt ERK STAT3

JAK2/STAT5 PI3K/Akt ERK STAT3 MAPK, STAT1 All stages

Earlier MK (including CD41-) and late MK Mature MK (CD41+ CD61+) Iron and Ca chelation

TPO: thrombopoietin; MK: megakaryocytes; Ca: calcium. *Not all pathways have been fully explored.

1114

haematologica | 2019; 104(6)

Thrombopoietin receptor agonists in ITP

TPO-RA using different generic and disease-specific questionnaires. Unquestionably, ITP has a major negative impact on HRQoL.28,29 In general, short-term treatment with TPO-RA does not seem to affect HRQoL,19,22 while long-term studies with both agents show improvements in HRQoL.17,25 In the

open-label RCT comparing romiplostim to SoC, clinically significant improvement in seven scales of the Immune Thrombocytopenic Purpura Patient Assessment Questionnaire (ITP-PAQ) was observed in both treatment arms at 52 weeks compared with baseline.17 However, the romiplostim group, and in particular the responders, had

Table 2. Summary of the randomized controlled trials performed in adult immune thrombocytopenia patients with romiplostim or eltrombopag.

Comparator Study arm duration

Definition of the primary end point for efficacy

Romiplostim

Results Primary end point

Any increase in plate count

Bleeding

Overall platelet response rate was noted in 88% of non-splenectomized and 79% of splenectomized patients given romiplostim compared with 14% of non-splenectomized and 0% splenectomized patients given placebo; P<0.0001 Between weeks 2 and 52 a 71-92% platelet response in the romiplostim group and 26-51% in the SoC

Significant bleeding events were reported in 12% of the patients in the placebo group and 7% in the romiplostim group

Bussel21

Placebo

6w

Proportion of patients achieving platelet count 50-450x109/L

Kuter18

Placebo

24 w

Proportion of patients achieving platelet count ≥50x109/L during ≥ 6 of the last 8 weeks of treatment

Kuter17

SoC

52 w

Incidences of treatment failure and splenectomy

11% in romiplostim arm vs. 30% in SoC arm (OR 0.31, 95% CI: 0.15-0.61; P<0.001)

Number of weeks with platelet response, defined as a platelet count ≥50x109/L (not including the 4 weeks after administration of rescue medication)

Weekly responses occurred for a median of 11 weeks with romiplostim vs. 0 weeks with placebo; P<0.0001

28-81% response Not reported depending on the dose vs. 11% in the placebo arm; P<0.001 59% in eltrombopag 16% Patients in the in placebo arms, eltrombopag group also OR 9.61, 95%CI: 3.31-27.86; had significantly greater P<0.0001 odds of responding at any point during the 6-week treatment period than did those in the placebo arm; OR 8.79, 95%CI: 3.54-21.86; P<0.0001 79% vs. 28% 38% of patients receiving OR 8·2, 95%CI: 3·59-18·73; eltrombopag vs. 7% placebo P<0·0001 responded at 75% or more of assessments, OR 10·53, 95%CI: 3.48-31.91; P<0·0001 60% in eltrombopag-treated patients; 0% in placebotreated patients

Shirasugi 88

Placebo

75% had platelet counts that reached or exceeded the targeted range vs. 25% in the placebo group 38% of splenectomized patients given romiplostim vs. 0% given placebo; P=0.0013, and 61% of non-splenectomized given romplostim vs. 5%; P<0.0001

Romiplostim group had significantly lower adjusted incidences of overall bleeding events (P=0.001) and bleeding events of grade 3 or higher (P=0.02) vs.the SoC

95% of romiplostim-treated patients achieved platelet responses

Eltrombopag Bussel22

Placebo

6w

Proportion of patients achieving platelet counts 50x109/L ≥ at day 43

Bussel19

Placebo

6w

Proportion of patients achieving platelet counts 50x109/L ≥ at day 43

Cheng25

Placebo

6m

OR of response defined as a platelet count of 50-400x109/L

Tomiyama89

Placebo

6w

Platelet count of ≥50x109/L at week 6 of the 6-week cycle

w: week; m:month; SoC: Standard of Care; OR: Odds Ratio; CI: Confidence Interval.

haematologica | 2019; 104(6)

1115

W. Ghanima et al.

significantly greater improvements, although the magnitude of the effect was of uncertain clinical benefit.30 In the RAISE study, HRQoL was significantly improved in the eltrombopag arm only, in five of the eight SF-36 domains at week 26 compared to baseline.25 In the EXTEND trial, all the HRQoL instruments used had positive mean changes from baseline over time. The improvements from baseline persisted through five years of treatment.31 This study found positive and clinically-meaningful mean changes from baseline in all HRQoL scores.

Practical issues related to use of thrombopoietin receptor agonists Indication and dosage The current label for both TPO-RA in Europe and in the US is patients aged ≥1 year with chronic ITP who are refractory to at least one other treatment (e.g. corticosteroids, immunoglobulins).

Initial dosing with eltrombopag starts at 50 mg daily, unless the patient is East Asian in whom a lower dose should initially be used. If a response is not seen in two weeks, the dose is increased to 75 mg daily, the maximum dose licensed for ITP. In the RAISE study of 197 adults, approximately equal numbers of patients were on 50 and 75 mg daily after six months of treatment. With romiplostim, the package insert recommends 1 μg/kg/week and increasing by 1 μg/kg/week until a response is achieved; however, this approach would take nine weeks to achieve the maximum dose of 10 μg/kg/week. A more practical schema would be to start at 3 μg/kg/week, particularly if a rapid response is needed, or one full vial of 250 μg, and increasing weekly to 5, 7, and then 10 μg/kg/week until a response is achieved. Median dose of romiplostim in adults is 3-5 μg/kg/week.27,32 In Europe, approximately one-third of the patients self-administer romiplostim subcutaneously.32 In the US, self-administration is still not licensed; however, this appears likely to be allowed in the near future.

In the pediatric studies, many children needed the maximum dose of 10 μg/kg/week of romiplostim and 75 mg of eltrombopag, which corresponds to 3-6 mg/kg as compared to 0.5-1 mg/kg for adults.23,33-35

Choice of agent The two TPO-RA have comparable overall efficacy.

Eltrombopag is given orally while romiplostim is dosed as a weekly subcutaneous injection. However, eltrombopag must be given on an empty stomach; in particular, it should be taken four hours after and two hours before food containing cations, e.g. iron, calcium, milk or other dairy products. In the US, different criteria for medical insurance are used for the two agents, which may impact on the decision to adopt one treatment or the other depending on which is likely to be approved first. If patients have absorption problems or transaminitis, it may be prudent to use romiplostim. If patients do not have stable platelet counts, or if they do not want to come to the clinic every week for injections, then eltrombopag may be better.

Dealing with non-responders: switching or combination Approximately one-third of the patients discontinue TPO-RA because of lack of response.36 If one TPO-RA does not work, switching to the other TPO-RA has been seen to be surprisingly effective. In a study of 46 patients 1116

who switched from one agent to another, 80% of the patients who failed to respond to eltrombopag eventually responded to romiplostim, and 46% of patients who did not respond to romiplostim responded to eltrombopag.37 These results were confirmed in a more recent retrospective study in which 106 patients underwent switching with 60% achieving response with either agent after switching.38 Switching can also be an effective policy in case of severe platelet-count fluctuations or side-effects.3739 Finally, stopping one agent before starting the other is not essential, unless adverse effects are the indication to switch (W Ghanima et al., personal observation, 2019).

The addition of a small dose of steroid (2.5-5 mg prednisolone) to a TPO-RA may have a good effect in some patients and can be tried in non-responding patients (W Ghanima et al., personal observation, 2019).

Treatment-free durable responses after discontinuation of thrombopoietin receptor agonists Approximately 10-30% of patients taking a TPO-RA will be able to discontinue their TPO-RA and maintain response after discontinuation.40 In one study of 75 adults with ITP of <6 months duration treated with romiplostim for ≤12 months, 32% were able to discontinue the medication and to obtain treatment-free durable responses (platelet counts >50x109/L) lasting at least six months.41 Higher mean platelet count (138x109/L) for the first two months was associated with remission. However, lasting treatment-free response has also been reported in chronic patients. In a retrospective study, 10% of 260 patients treated with eltrombopag maintained acceptable platelet counts after discontinuation of the drug.42 In another small retrospective study of 54 patients who were treated with TPO-RA for at least five years, TPO-RA were discontinued in 20 out of 28 patients who achieved a complete response. Of these, eight patients showed a sustained response for a median of 13 months (range 5-27 months).40 However, it is still not known how TPO-RA induce longlasting off-treatment responses, although it is unlikely that this is simply due to a selection of patients who would eventually remit.41 Durable responses have even been observed in patients with long-lasting disease. Potential mechanisms include: restored immune tolerance by increased exposure to platelet autoantigens, thereby reducing platelet antibodies through increased presence of MK and platelets,41,43 or through improvement of Treg function, which in turn could restore immune tolerance to platelets.13 Predicting who will achieve a durable response and how to discontinue TPO-RA is challenging. We recommend tapering in a patient who achieves and maintains a stable platelet count over 50-100x109/L for at least 3-6 months, particularly if using low doses of a TPO-RA and achieving a normal, stable platelet count for some months. One way to taper treatment would involve gradually decreasing and/or increasing the interval between doses until the platelet count remains <30x109/L or it is possible to discontinue treatment.

Safety and tolerability of thrombopoietin receptor agonists Ten years after their availability, TPO-RA have been proven to be well-tolerated. The long follow up of the haematologica | 2019; 104(6)

Thrombopoietin receptor agonists in ITP

patients included in the pivotal studies and â&#x20AC;&#x153;real-lifeâ&#x20AC;? reports generally give reassuring data. Many of the initial theoretical concerns, such as uncontrolled stem cell proliferations and myelofibrosis with TPO-RA, have not materialized. However, there are emerging reports of adverse events, such as an increased incidence of thrombosis, which remains unexplained.

In the EXTEND study, evaluating long-term safety and efficacy of eltrombopag in 302 adults ITP patients treated with eltrombopag for a mean duration of >2 years, important adverse events were rare and did not increase with treatment duration over one year.24 Fourteen percent of patients on eltrombopag stopped treatment because of adverse events. Although plasma levels and exposure have been shown to be much higher in Asian populations,44 tolerability of eltrombopag in the Chinese population appears similar to that observed in Caucasian populations.

The good tolerance of eltrombopag observed in pivotal studies has been confirmed in the Spanish eltrombopag registry including 220 ITP adults.45 In a pooled analysis from 13 completed studies of romi-

plostim including 1111 patients, exposure-adjusted rates of adverse events were lower in the romiplostim group than in the placebo/SoC group.46 These data were confirmed in another registry study.36 Bone marrow reticulin deposition and TE events are associated with the TPO-RA drug class. However, the safety profiles of TPO-RA do not fully overlap and specific adverse events, i.e. cataract and transaminitis, are more frequently seen with eltrombopag. Others, such as development of neutralizing antibodies, are mainly observed with romiplostim, as is pain after administration. This absence of overlapping toxicity encourages switching when a TPO-RA is stopped because of an adverse event that is not due to class effect.

Bone marrow fibrosis Early concerns were raised regarding the possible induction of bone marrow fibrosis because of sustained stimulation of megakaryopoiesis by TPO-RA, as seen in animal studies.47 Table 3 summarizes results of the published trials showing that, in most patients, grade of fibro-

Table 3. Summary of studies determining the grade of bone marrow fibrosis in patients treated with thrombopoietin receptor agonists.

Author

Study design

N

Agent

Staging

Ghanima

Retrospective, single center. BM biopsies were performed at different intervals.

66

R, E

ECS

Brynes51

Prospective, open-label multicenter study 117 of patients who were included in the EXTEND trial and received eltrombopag were followed up with annual BM biopsies

E

ECS

Brynes49

Prospective, open-label multicenter.

Biopsy specimens were collected at baseline (before treatment) and after 1 and 2 years of treatment

162

E

ECS

Janssens50

Prospective, open-label multicenter.

Bone marrow biopsies were scheduled after 1, 2, or 3 years of romiplostim or earlier if patients discontinued or failed to achieve/maintain a response to romiplostim.

131

R

Bauermeister

48

Results After a median treatment duration of 29 months (range: 16-47), 22% of the BM biopsies were graded as MF-0, 59% as MF-1 and 18% as MF-2.

The proportion of MF-0 decreased from 67% in the pre-treatment biopsies to 22% first on-treatment biopsies, which were largely MF-1.

Two or more biopsies were available in 32 patients; comparing the first to the last on-treatment bone marrows, the grade of fibrosis increased in 11 cases, remained the same in 15 and decreased in 6.

209 on-treatment biopsies collected from 115 patients were re-evaluated. Median duration of treatment was 45 months (range: 2-73m).

98% of patients had findings of MF-0 or MF-1 in any given year over the 5-year study period. Five biopsies from 3 patients (2%) were reported as MF-2 or MF-3 at 25 months; collagen was present in these 5 specimens.

Of 18 patients with 3 biopsies, 8 patients remained at MF-0 over the treatment period and 5 had an increase of one grade. The remaining 5 patients showed a decrease of one grade when compared with the grade from the first on-treatment biopsies.

Median time on treatment was 104 weeks (range: 2.4-113). At 1 year (n=127), 69% had a grade of MF-0, 28% had MF-1, 2% had MF-2, and 2% had MF-3.

Compared with baseline, 79 out of 93 patients (85%) had MF-0 at 2 years, 9 (10%) had a 1-grade increase, 2 (2%) had a 1-grade decrease, 1 (1%) remained MF-1, and none had 2- or 3-grade increases. Five out of 127 patients (4%) at 1 year and 1 out of 93 (1%) at 2 years had collagen deposition.

The median (Q1, Q3) duration of treatment was in cohort 1, 147 (17-156) weeks; cohort 2, 155 (66, 156) weeks; and cohort 3, 155 (66-156) weeks.

9 of 131 (6.9 %) included in the 3 cohorts had increases of â&#x2030;Ľ2 grades (cohort 1: 0/34; cohort 2: 2/39; cohort 3: 7/58), including 2 with collagen.

R: romiplostim; E: eltrombopag; ECS: European Consensus Staging; BM: bone marrow; MF: marrow fibrosis.

haematologica | 2019; 104(6)

1117

W. Ghanima et al.

sis did not change during treatment with TPO-RA, while a slight, non-progressive reticulin fibrosis (MF-1 or Baumeister <2) was observed in 10-50% of patients.48,49 In one study, a moderate increase in reticulin fibrosis (MF-2) was observed in 18% at median time of treatment of 2.5 years,48 whereas in three other studies, reticulin fibrosis progressed by >2 grades or developed ≥MF-2 during the study periods in less than 10%.49-51 Severe grades of reticulin fibrosis (MF-3 and/or collagen fibrosis) were extremely rare in all studies.48-51 In general, it does not seem that TPO-RA induce substantial fibrosis or changes in number or morphology of peripheral blood cells. Both reticulin and collagen fibrosis regressed in most patients after discontinuation of TPO-RA; in a few patients, fibrosis regressed despite continuing therapy.48,49 There is no consensus for patients on TPO-RA as to whether or how to monitor bone marrow (BM) fibrosis.

At the moment, hardly any centers perform routine BM biopsy in TPO-RA treated patients. However, if a biopsy is performed and severe reticulin (MF3) or collagen is discovered, then it is recommended that TPO-RA be discontinued. With moderately increased fibrosis, e.g. MF 2, a patient may continue TPO-RA but may need a repeat biopsy in six months. Older age and splenectomy could be associated with higher grades of BM fibrosis; fibrosis was not associated with type, dose or duration of treatment.20,27,48

Risk of clonal evolution and malignancy The TPO-receptor is expressed in many hematopoietic

cells, including early stem cells.52 Sustained stimulation of the hematopoietic cells raised concerns regarding potential clonal evolution associated with prolonged use of TPO-RA. Based on clinical trials, safety databases and ten years of clinical experience, there are no indications that TPO-RA induce neoplastic changes in ITP patients. A safety analysis of more than 1000 patients treated with romiplostim showed that rates of hematologic and non-hematologic malignancies were comparable between the romiplostim group and the placebo/SoC.53 In the EXTEND study, ten (3%) patients reported malignancies diagnosed during the 6-year study.24 In one ITP study, routine BM flow cytometry and cytogenetic studies were performed and no karyotypic or immunophenotypic changes indicative of monoclonality were detected.48 In myelodysplastic syndromes (MDS), the risk of leukemogenesis had been a matter of concern.54-56 In aplastic anemia, comparison of natural history between eltrombopag-treated patients and those receiving immunosuppression alone showed no difference in incidence of malignancy, although those treated with eltrombopag tended to develop malignancy sooner.57

Thromboembolism In early trials with TPO-RA, sporadic thromboembolic events (TEE) gave impetus to extensive epidemiological studies exploring the association between thrombosis and ITP and the role of TPO-RA. The incidence of TEE in patients with chronic ITP not exposed to TPO-RA was compared with age- and sex-matched non-ITP control

Table 4. Incidence of thromboembolism with romiplostim and eltrombopag in long-term studies or in pooled analyses in adults.

Ref.

Description

Patients with history N. of patients Mean or risk factors for with TEEs exposure TE excluded time (years)

Romiplostim Kuter27§ Long-term In most cases, investigation but not invariably on 291 pts who completed a previous randomized romiplostim study* (Aug 2004 - Jan 2010) Rodeghiero53 Pooled analysis of In most cases, 13 clinical trials^ but not invariably (Oct 2003 and June 2009) Eltrombopag Wong24 Long-term investigation Yes on 302 out of 371 patients completing previous trials fed into this study (June 2006-July 2015)

19/291 (6.5%)

39/653 (5.9%)

19/302 (6.3%)

Total n. of TEEs Arterial

Venous

Others

Arterial 16

25 Venous 9

Other 0

Arterial 26

69 Venous 40

Other 3

Arterial 14

24 Venous 10

Other 0

2.11

1.41

2.37 (median, range 2 days-8.8 yrs)

First TEE rate per 100 pt-years

All TEE rate per 100 pt-years Arterial Venous

3.1

4.1 Arterial Venous 2.6 1.5

4.2

7.5 Arterial Venous 2.8 4.3

2.9#

3.4 Arterial Venous 2.0 1.4

From an interim analysis (Aug 2004-July 2007) describing 142 patients for a mean exposure time of 69 weeks, 12 thromboembolism (TE) events were found in seven patients, with an incidence rate of TE of 6.4/100 patient-years (pt-yrs) (3.7/100 pt-yrs after censoring at the first event).20 *Analysis includes patients completing previous studies conducted in the US, Europe, Canada and Australia (described in studies 2, 3, 4 of Online Supplementary Table S1 and n. 12 and 13 of Online Supplementary Table S2) who were subsequently enrolled in an open label extension study. Only thrombotic events occurring after enrolling in the extension were considered. ^Analysis includes all 291 patients of Kuter et al., 201327 and of an additional 362 patients, including: Japanese patients, described in n. 5 of Table A and n. 14 of Table B; 22 children enrolled in a double-blind study,23 and patients directly enrolled in different open-label studies (a large expanded access study, now available as a full paper;90 a Japanese extension study;88 a long-term biopsy study).47 Details of studies can be found in Rodeghiero et al.53 The same studies with a later cut off (June 2011) were also analyzed by Cines et al.46 also including preliminary data in abstract form of Janssens et al.50 The analysis of Cines et al. provided a similar thrombotic events rate of 5.5/100 pt-yrs both in the romiplostim (n. 994, 1520 pts/yrs) and in the placebo/Standard of Care (n. 138, 110 pts/yrs) patients. # 2.7 if two additional patients who only had TEE off-study (see Table 3) are excluded.

N/n: number; TEE: thromboembolic events. §

1118

haematologica | 2019; 104(6)

Thrombopoietin receptor agonists in ITP

populations.58-62 The annualized incidence was 0.41-0.67 for venous thromboembolism (VTE) and 0.96-1.15 for arterial thrombosis (AT), whereas the control populations had 0.28-0.42 and 0.67-0.91, respectively, showing a slightly but statistically significantly higher risk of VTE and possibly AT in ITP patients.63 Thromboembolic events in the long-term studies and pooled analyses are summarized in Table 4. Online Supplementary Table S1 describes thromboembolic events in phase I-II and in randomized, placebo-controlled studies, while Online Supplementary Table S2 refers to single arm trials. As shown in Online Supplementary Table S1, the exposure time to TPO-RA was generally short and ranged from a few weeks to â&#x2030;¤6 months (or 1 year in a single study17). Overall, there have been 15 events out of 415 patients exposed to romiplostim (3.6%) versus 4 events in 202 controls (2%), and 5 events in the 391 patients exposed to eltrombopag (1.3%) versus none out of 155 controls. Conversely, more consistent and significant data could be derived from the large long-term studies or pooled analyses mainly based on the long-term extension studies, fed by patients who had completed previous trials and in a single-arm study investigating a large number of patients. These studies included greater numbers of patients and report on longer treatment exposure.24,27,50,53 Notably, patients with history of or important risk factors for thrombosis were excluded upfront from the RCT, and patients experiencing TEE during their previous study were excluded from long-term extension studies, resulting in a generally smaller thrombosis risk population in the long-term studies. Despite that, a relatively large number of TEE occurred in the long term-studies.

The incidence per 100 patient-years (censoring after first TEE) ranged from 3.1 to 4.2 with romiplostim and was 2.9 in the single eltrombopag study. Without censoring after first event, the incidence ranged from 4.1 to 7.5 with romiplostim and 3.4 with eltrombopag.63 In a pooled analysis of romiplostim studies, an incidence rate per 100 patientyears of 5.5 was reported for both patients exposed to romiplostim or to placebo/SoC.46 Unfortunately, the low number and short exposure of placebo/SoC make the figures in non-exposed patients unreliable.

Thrombotic events have also been reported in pediatric trials. In a multicenter retrospective study on 79 children with ITP treated with eltrombopag, romiplostim or both, two cases of pulmonary embolism were reported.64 The randomized controlled trials did not identify any TEE, and overall TEE incidence is clearly lower than in adults.23,33-35,65 The TEE events were neither associated with thrombocytosis nor with a higher dose of TPO-RA. At least 3050% of cases occurred in patients with lower than normal platelet counts. In general, TEE events tended to happen in the first year of treatment, creating a trend towards lower incidence figures with more prolonged exposure time.

Among arterial events, cerebrovascular (stroke) and myocardial (infarction) were predominant and seen more in patients >70 years of age. However, <20% of TEE resulted in permanent disability and only three deaths could be attributed to thrombosis. In a pooled analysis, the annualized risk of thromboembolism in splenectomized patients (6.3) was not significantly higher than non-splenectomized patients (4.3).66 The pathogenic mechanisms responsible for the increased thrombotic risk linked to TPO-RA have not yet been identified.67 The expected findings that TPO-RA haematologica | 2019; 104(6)

lower the threshold of platelet activation have not been demonstrated.68,69 In general, ITP per se seems to be a procoagulant condition, as indicated by an increase in the various coagulation activation markers, including D-dimer, prothrombin fragment F1+2 and thrombin generation, and in the antifibrinolytic marker plasminogen activator inhibitor-1 (PAI-1) compared to controls.70,71 No further increase in the coagulation activation markers has been observed after the initiation of TPO-RA.70,71 However, a recent study reported increased PAI-1 levels in patients treated with TPO-RA, possibly leading to the formation of a more fibrinolysis-resistant clot; the study also showed increased microparticle-associated phosphatidylserine procoagulant activity.72 Moreover, levels of soluble Pselectin and basal exposure of P-selectin in quiescent platelets were significantly increased in TPO-RA treated patients compared to pretreatment levels or to untreated patients; however, the significance of these findings is still not known.70,72 In summary, although they have not been substantiated in properly designed trials, the annualized thrombosis rates in adults appear to be 2-3 times higher (annualized incidence rate of TEE of 4-7%) with TPO-RA treatment than in an ITP population not treated with TPO-RA, and even higher if compared to non-ITP control populations.63 On the other hand, most available data on the risk of thrombosis are based on retrospective and registry studies, which probably underestimate the risk of thrombosis in the ITP population. The patient's individual risk profile should be considered when initiating treatment with a TPO-RA to evaluate if the expected reduction in bleeding risk outweighs the risk of thrombotisis. Comorbidities more prevalent in ITP should be considered and/or investigated; these include previous thromboembolism, splenectomy, presence of antiphospholipid antibodies, and concomitant medications like estroprogestinic preparations. Efforts should be made to correct modifiable risk factors, and thrombo-prophylaxis is recommended for surgery, provided the patient has a safe platelet count.67 Furthermore, antiplatelet agents or even anticoagulation could be considered in patients at high risk of thrombosis once platelet counts reach >50x109/L after initiation of TPO-RA.

Rebound thrombocytopenia In most patients receiving TPO-RA, platelet counts return to pre-therapy baseline values on discontinuation of therapy; however, in up to 10% of patients, platelet counts temporarily drop below pretreatment levels after discontinuation of TPO-RA.18 Endogenous TPO activity, which is regulated by platelet mass, may be suppressed while platelet and MK levels are elevated on TPO-RA and may not rapidly re-equilibrate when TPO-RA are abruptly discontinued. However, the REPEAT study, which involved intermittent administration of eltrombopag, provided reassuring data.73 Nonetheless, when TPO-RA treatment is discontinued, tapering is preferred to immediate withdrawal.

Fluctuating platelet counts Substantial fluctuations in platelet count on stable treatment doses may occur and can be difficult to manage. They are more common with romiplostim than eltrombopag, possibly due to the longer dosing intervals and inconsistent delivery with subcutaneous administration.36-38 Some 1119

W. Ghanima et al.

patients experiencing such platelet fluctuations with romiplostim can be stabilized by switching to eltrombopag.74

Adverse events mainly associated with eltrombopag

tion of neutralizing antibodies did not automatically result in loss of response.66 However, neutralizing antibodies were detected in a group of four patients treated with romiplostim who lost response.75 Current estimates, based on very limited data, suggest that these occur at a rate of up to 1%, more frequently in children than adults.

Monitoring could be yearly and when response is lost.

Cataract Treatment-related cataracts were observed in juvenile rodents on eltrombopag and were dose and time dependent. Cataracts have been reported with both eltrombopag and romiplostim. Given multiple confounding risk factors, e.g. steroid use, older age, smoking, no clinical study has unequivocally demonstrated this suspected risk with TPO-RA. In a 6-month study, the incidence of cataract in patients treated with eltrombopag was similar to placebo.25 In the open-label EXTEND study, cataracts developed in 28 patients (9%) in up to eight years of treatment.

In 16 (5%), it was considered a severe adverse event, which led to withdrawal of eltrombopag in four (1.3%) In the Pediatric Patients with patients.24 Thrombocytopenia from Idiopathic Thrombocytopenic Purpura (PETIT2) study, two children developed cataracts, raising serious concern.33 The analysis of up to 1000 patients treated with romiplostim for ITP reported 37 events of cataracts, but only one case in patients with placebo or SoC, suggesting cataracts, given the big difference in exposure, might also be associated with romiplostim.46 An alternative option to routine ophthalmological evaluation for all patients on eltrombopag is to reserve ophthalmic examination for patients with one or more risk factors.

Transaminitis Hepatocyte degeneration, associated with increased serum liver enzymes, was observed in animals at doses that were associated with morbidity and mortality. In humans, development of transaminitis occurs in up to 10%, especially on eltrombopag.22,25 Bilirubin elevations are also possible but involve mainly non-conjugated bilirubin (not indicative of serious liver injury). Transaminitis is mostly asymptomatic and reversible with dose interruption, reduction or discontinuation; only 3% of children and adults were unable to tolerate eltrombopag in large studies.24,25 Transaminitis occurs more in the first year of treatment, which justifies regular monitoring of liver enzymes at more frequent intervals particularly during the first years.

Adverse events mainly associated with romiplostim Risk of antibody development Romiplostim is a chimeric fusion protein produced by genetic engineering. Neutralizing antibodies directed against romiplostim have been reported. In contrast, because small molecules do not typically elicit an immune response, development of neutralizing antibodies is not considered to be a risk for eltrombopag. In an analysis of up to 1000 patients treated with romiplostim, neutralizing antibodies to romiplostim were reported in only six patients; importantly, none cross-reacted with endogenous TPO. All had a platelet response, and detec1120

Other indications for thrombopoietin receptor agonists Treatment of severe aplastic anemia (SAA) with eltrombopag yielded multilineage clinical responses in certain patients with refractory severe aplastic anemia.11 Consequently, eltrombopag has been approved for use in patients with SAA failing immunosuppression who are not eligible for transplantation. A recent study has shown benefit when eltrombopag was used upfront together with immunosuppression, with more than one-third of the patients achieving a complete response by six months.57 Thrombocytopenia is a common complication of liver disease, and eltrombopag was licensed to support the platelet count in patients with hepatitis C undergoing treatment with interferon and ribavirin. However, improvements in current hepatitis C therapy have meant that interferon is no longer used.76 Studies into the use of TPO-RA in MDS are no longer actively pursued, perhaps not so much because of the risk of induction of leukemia, as because of failure to provide evidence of survival benefit.54,55 The major indication under study is in solid tumor chemotherapy. Studies are now available that demonstrate promising results. Schedule and dosing in solid tumor chemotherapy can be better maintained, but translating this into a survival advantage still has to be clearly demonstrated.77 Treatment of inherited thrombocytopenias with eltrombopag has been studied in MYH9-related disorders and Wiskott-Aldrich syndrome, showing platelet response in both conditions.78-80

Other thrombopoietic agents Two current studies with avatrombopag and lusutrombopag, both of which were recently approved for procedures in thrombocytopenic patients with liver disease in the US, were careful to verify adequate pre-procedure portal flow and use only several days of TPO-RA prior to and during the procedure to avoid risks of thrombosis, especially that of the portal vein. Avatrombopag, an oral small molecule, apparently binds to the TPO-R similarly to eltrombopag, but does not have any dietary limitations.81 It was shown to be effective in a phase II study of ITP,81 and also in a recently published phase III trial, which confirmed the superiority of avatrombopag (5-40 mg daily) over placebo with regard to acute and durable platelet response in patients with chronic ITP. Headache was the most frequent side effect.82 In view of the two RCT in ITP and the approved indication in liver disease, we expect that avatrombopag will be licensed for ITP. A recombinant human thrombopoietin (rhTPO) has been licensed in China for many years in adults and children with ITP. Studies have also recently been performed in pregnant women, all with good results.83,84 One concern is the uncertainty regarding development of antibodies to the TPO agent which, unlike those seen with romiplostim, might cross-react with endogenous TPO and create lasting haematologica | 2019; 104(6)

Thrombopoietin receptor agonists in ITP

and substantial thrombocytopenia in affected recipients.

Importantly, neither romiplostim nor eltrombopag are recommended to be used in pregnancy; however, there are limited case reports in which the use of these agents in pregnant women with difficult ITP appeared to be safe.

Future treatments for immune thrombocytopenia and the role of thrombopoietin receptor agonists The two TPO-RA licensed for use in ITP are now both licensed for use after one year from diagnosis after failure of corticosteroids, further consolidating their position as the mainstay for second-line therapy in ITP. However, many other agents are currently under development, at various stages of clinical testing, or are being considered for registration. Among these, fostamatinib, an inhibitor of spleen tyrosine kinase (syk) has an overall response rate of almost 50% with continued treatment, and an 18% rate of stable responses in heavily pretreated ITP patients.85 This agent was licensed in the US in 2018 for treatment of

References 1. Rodeghiero F. A critical appraisal of the evidence for the role of splenectomy in adults and children with ITP. Br J Haematol.

2018;181(2):183-195.

2. Patel VL, Mahevas M, Lee SY, et al.

Outcomes 5 years after response to rituximab therapy in children and adults with immune thrombocytopenia.

Blood.

2012;119(25):5989-5995.

3. Ghanima W, Godeau B, Cines DB, Bussel JB. How I treat immune thrombocytopenia: the choice between splenectomy or a medical therapy as a second-line treatment. Blood.

2012;120(5):960-969.

4. Erickson-Miller CL, Delorme E, Tian SS, et al. Preclinical activity of eltrombopag (SB497115), an oral, nonpeptide thrombopoietin receptor agonist. Stem Cells.

2009;27(2):424-430.

5. Broudy VC, Lin NL. AMG531 stimulates megakaryopoiesis in vitro by binding to Mpl. Cytokine. 2004;25(2):52-60.

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

2009;114(18):3899-3908.

7. Sun H, Tsai Y, Nowak I, Liesveld J, Chen Y. Eltrombopag, a thrombopoietin receptor agonist, enhances human umbilical cord blood hematopoietic stem/primitive progenitor cell expansion and promotes multilineage hematopoiesis. Stem Cell Res. 2012;9(2):77-86.

8. Di Buduo CA, Currao M, Pecci A, Kaplan DL, Balduini CL, Balduini A. Revealing eltrombopag's promotion of human megakaryopoiesis through AKT/ERKdependent pathway activation.

Haematologica. 2016;101(12):1479-1488.

9. Vlachodimitropoulou E, Chen YL, Garbowski M, et al. Eltrombopag: a powerful chelator of cellular or extracellular iron(III) alone or combined with a second

haematologica | 2019; 104(6)

chronic ITP in adults. Several blockers of FcRn have entered trials in adults with persistent and chronic ITP; at least two (rozanolixizumab and ARGX-117) have completed phase II studies.86 The mechanism is a dramatic increase in IgG turnover as a result of inhibition of IgG recycling; not only â&#x20AC;&#x153;normalâ&#x20AC;? but also IgG autoantibody levels decrease markedly.87 Preliminary results are encouraging but efficacy and toxicity need to be better defined in phase III studies. It remains to be seen how these agents will influence the future role of TPO-RA.

Conclusion Romiplostim and eltrombopag are well tolerated and effective therapies for ITP with acceptable toxicity. Both agents increase the platelet count in up to three-quarters of patients. Ten years after their introduction, available evidence from short- and long-term and registry-based studies confirm the general safety of chronic long-term use of these medications, as well as persistent efficacy in most patients.

chelator. Blood. 2017;130(17):1923-1933.

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

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

2012;120(2):386-394.

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

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

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

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

2013;121(7):1240-1242.

13. Bao W, Bussel JB, Heck S, et al. Improved regulatory T-cell activity in patients with chronic immune thrombocytopenia treated with thrombopoietic agents. Blood.

2010;116(22):4639-4645.

14. Wan YY, Flavell RA. 'Yin-Yang' functions of transforming growth factor-beta and T regulatory cells in immune regulation. Immunol Rev. 2007;220:199-213.

15. Schifferli A, Kuhne T. Thrombopoietin receptor agonists: a new immune modulatory strategy in immune thrombocytopenia?

Semin Hematol. 2016;53 Suppl 1:S31-4.

16. Zufferey A, Kapur R, Semple JW. Pathogenesis and Therapeutic Mechanisms in Immune Thrombocytopenia (ITP). J Clin Med. 2017;6(2).

17. Kuter DJ, Rummel M, Boccia R, et al.

Romiplostim or standard of care in patients with immune thrombocytopenia. N Engl J Med. 2010;363(20):1889-1899.

18. Kuter DJ, Bussel JB, Lyons RM, et al. Efficacy of romiplostim in patients with chronic immune thrombocytopenic purpura: a double-blind randomised controlled trial.

Lancet. 2008;371(9610):395-403.

19. Bussel JB, Provan D, Shamsi T, et al. Effect of eltrombopag on platelet counts and bleeding during treatment of chronic idiopathic thrombocytopenic purpura: a randomised, double-blind, placebo-controlled trial.

Lancet. 2009;373(9664):641-648.

20. Bussel JB, Kuter DJ, Pullarkat V, Lyons RM, Guo M, Nichol JL. Safety and efficacy of

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

long-term treatment with romiplostim in thrombocytopenic patients with chronic ITP. Blood. 2009;113(10):2161-2171.

Bussel JB, Kuter DJ, George JN, et al. AMG 531, a thrombopoiesis-stimulating protein, for chronic ITP. N Engl J Med. 2006;355(16):1672-1681.

Bussel JB, Cheng G, Saleh MN, et al.

Eltrombopag for the treatment of chronic idiopathic thrombocytopenic purpura. N Engl J Med. 2007;357(22):2237-2247.

Bussel JB, Buchanan GR, Nugent DJ, et al. A randomized, double-blind study of romiplostim to determine its safety and efficacy in children with immune thrombocytopenia. Blood. 2011;118(1):28-36.

Wong RSM, Saleh MN, Khelif A, et al.

Safety and efficacy of long-term treatment of chronic/persistent ITP with eltrombopag: final results of the EXTEND study. Blood.

2017;130(23):2527-2536.

Cheng G, Saleh MN, Marcher C, et al.

Eltrombopag for management of chronic immune thrombocytopenia (RAISE): a 6month, randomised, phase 3 study. Lancet.

2011;377:393-402.

Wang L, Gao Z, Chen XP, et al. Efficacy and safety of thrombopoietin receptor agonists in patients with primary immune thrombocytopenia: A systematic review and metaanalysis. Sci Rep. 2016;6:39003.

Kuter DJ, Bussel JB, Newland A, et al. Longterm treatment with romiplostim in patients with chronic immune thrombocytopenia: safety and efficacy. Br J Haematol. 2013;161 (3):411-423.

McMillan R, Bussel JB, George JN, Lalla D, Nichol JL. Self-reported health-related quality of life in adults with chronic immune thrombocytopenic purpura. Am J Hematol.

2008;83(2):150-154.

Mathias SD, Gao SK, Miller KL, et al. Impact of chronic Immune Thrombocytopenic Purpura (ITP) on health-related quality of life: a conceptual model starting with the patient perspective. Health Qual Life Outcomes. 2008;6:13.

Kuter DJ, Mathias SD, Rummel M, et al.

Health-related quality of life in nonsplenec-

1121

W. Ghanima et al.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

1122

tomized immune thrombocytopenia patients receiving romiplostim or medical standard of care. Am J Hematol. 2012;87(5): 558-561.

Saleh MN, Bussel JB, Khelif A, et al.

Improvements in Patient Health-Related Quality of Life (HRQoL) with Clinical Efficacy in Patients Treated with Eltrombopag: Final Results from the LongTerm, Open-Label Extend Study. Blood.

2016;128(22):3742.

Steurer M, Quittet P, Papadaki HA, et al. A large observational study of patients with primary immune thrombocytopenia receiving romiplostim in European clinical practice. Eur J Haematol. 2017;98(2):112-120.

Grainger JD, Locatelli F, Chotsampancharoen T, et al. Eltrombopag for children with chronic immune thrombocytopenia (PETIT2): a randomised, multicentre, placebo-controlled trial. Lancet. 2015;386 (10004):1649-1658.

Bussel JB, de Miguel PG, Despotovic JM, et al. Eltrombopag for the treatment of children with persistent and chronic immune thrombocytopenia (PETIT): a randomised, multicentre, placebo-controlled study.

Lancet Haematol. 2015;2(8):e315-325.

Tarantino MD, Bussel JB, Blanchette VS, et al. Romiplostim in children with immune thrombocytopenia: a phase 3, randomised, double-blind, placebo-controlled study.

Lancet. 2016;388(10039):45-54.

Khellaf M, Michel M, Quittet P, et al.

Romiplostim safety and efficacy for immune thrombocytopenia in clinical practice: 2-year results of 72 adults in a romiplostim compassionate-use program. Blood.

2011;118(16):4338-4345.

Khellaf M, Viallard JF, Hamidou M, et al. A retrospective pilot evaluation of switching thrombopoietic receptor-agonists in immune thrombocytopenia.

Haematologica. 2013;98(6):881-887.

Cantoni S, Carpenedo M, Mazzucconi MG, et al. Alternate use of thrombopoietin receptor agonists in adult primary immune thrombocytopenia patients: A retrospective collaborative survey from Italian hematology centers. Am J Hematol. 2018;93(1):5864.

Kuter DJ, Macahilig C, Grotzinger KM, et al.

Treatment patterns and clinical outcomes in patients with chronic immune thrombocytopenia (ITP) switched to eltrombopag or romiplostim. Int J Hematol. 2015;101(3): 255-263.

Mahevas M, Fain O, Ebbo M, et al. The temporary use of thrombopoietin-receptor agonists may induce a prolonged remission in adult chronic immune thrombocytopenia.

Results of a French observational study. Br J Haematol. 2014;165(6):865-869.

Newland A, Godeau B, Priego V, et al.

Remission and platelet responses with romiplostim in primary immune thrombocytopenia: final results from a phase 2 study. Br J Haematol. 2016;172(2):262-273.

Gonzalez-Lopez TJ, Pascual C, AlvarezRoman MT, et al. Successful discontinuation of eltrombopag after complete remission in patients with primary immune thrombocytopenia. Am J Hematol. 2015;90(3):E40-43.

Ghadaki B, Nazi I, Kelton JG, Arnold DM. Sustained remissions of immune thrombocytopenia associated with the use of thrombopoietin receptor agonists. Transfusion.

2013;53(11):2807-2812.

Gibiansky E, Zhang J, Williams D, Wang Z, Ouellet D. Population pharmacokinetics of eltrombopag in healthy subjects and

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

57.

58.

patients with chronic idiopathic thrombocytopenic purpura. J Clin Pharmacol.

2011;51(6):842-856.

Gonzalez-Lopez TJ, Alvarez-Roman MT, Pascual C, et al. Use of eltrombopag for secondary immune thrombocytopenia in clinical practice. Br J Haematol. 2017;178(6):959970.

Cines DB, Gernsheimer T, Wasser J, et al.

Integrated analysis of long-term safety in patients with chronic immune thrombocytopaenia (ITP) treated with the thrombopoietin (TPO) receptor agonist romiplostim. Int J Hematol. 2015;102(3):259-270.

Kuter DJ, Mufti GJ, Bain BJ, Hasserjian RP, Davis W, Rutstein M. Evaluation of bone marrow reticulin formation in chronic immune thrombocytopenia patients treated with romiplostim. Blood. 2009;114(18): 3748-3756.

Ghanima W, Geyer JT, Lee CS, et al. Bone marrow fibrosis in 66 Immune Thrombocytopenia patients treated with thrombopoietin receptor agonists: a single center long-term follow-up. Haematologica.

2014;99(5):937-944.

Brynes RK, Wong RS, Thein MM, et al. A 2Year, Longitudinal, Prospective Study of the Effects of Eltrombopag on Bone Marrow in Patients with Chronic Immune Thrombocytopenia.

Acta Haematol.

2017;137(2):66-72.

Janssens A, Rodeghiero F, Anderson D, et al.

Changes in bone marrow morphology in adults receiving romiplostim for the treatment of thrombocytopenia associated with primary immune thrombocytopenia. Ann Hematol. 2016;95(7):1077-1087.

Brynes RK, Orazi A, Theodore D, et al.

Evaluation of bone marrow reticulin in patients with chronic immune thrombocytopenia treated with eltrombopag: Data from the EXTEND study. Am J Hematol.

2015;90(7):598-601.

Imbach P, Crowther M. Thrombopoietinreceptor agonists for primary immune thrombocytopenia. N Engl J Med. 2011;365 (8):734-741.

Rodeghiero F, Stasi R, Giagounidis A, et al.

Long-term safety and tolerability of romiplostim in patients with primary immune thrombocytopenia: a pooled analysis of 13 clinical trials. Eur J Haematol. 2013;91(5): 423-436.

Mittelman M, Platzbecker U, Afanasyev B, et al. Eltrombopag for advanced myelodysplastic syndromes or acute myeloid leukaemia and severe thrombocytopenia (ASPIRE): a randomised, placebo-controlled, phase 2 trial.

Lancet Haematol.

2018;5(1):e34-e43.

Fenaux P, Muus P, Kantarjian H, et al.

Romiplostim monotherapy in thrombocytopenic patients with myelodysplastic syndromes: long-term safety and efficacy. Br J Haematol. 2017;178(6):906-913.

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

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

2015;2(10):e417-426.

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

Severinsen MT, Engebjerg MC, Farkas DK, et al. Risk of venous thromboembolism in

59.

60.

61.

62.

63.

64.

65.

66.

67.

68.

69.

70.

71.

72.

73.

patients with primary chronic immune thrombocytopenia: a Danish populationbased cohort study. Br J Haematol.

2011;152(3):360-362.

Sarpatwari A, Bennett D, Logie JW, et al.

Thromboembolic events among adult patients with primary immune thrombocytopenia in the United Kingdom General Practice Research Database. Haematologica.

2010;95(7):1167-1175.

Norgaard M, Severinsen MT, Lund Maegbaek M, Jensen AO, Cha S, Sorensen HT. Risk of arterial thrombosis in patients with primary chronic immune thrombocytopenia: a Danish population-based cohort study. Br J Haematol. 2012;159(1):109-111.

Norgaard M, Cetin K, Maegbaek ML, et al.

Risk of arterial thrombotic and venous thromboembolic events in patients with primary chronic immune thrombocytopenia: a Scandinavian population-based cohort study. Br J Haematol. 2016;174(4):639-642.

Enger C, Bennett D, Forssen U, Fogarty PF, McAfee AT. Comorbidities in patients with persistent or chronic immune thrombocytopenia. Int J Hematol. 2010;92(2):289-295.

Rodeghiero F. Is ITP a thrombophilic disorder? Am J Hematol. 2016;91(1):39-45.

Ramaswamy K, Hsieh L, Leven E, Thompson MV, Nugent D, Bussel JB. Thrombopoietic agents for the treatment of persistent and chronic immune thrombocytopenia in children. J Pediatr. 2014;165(3): 600-6055.e4.

Elalfy MS, Abdelmaksoud AA, Eltonbary KY. Romiplostim in children with chronic refractory ITP: randomized placebo controlled study. Ann Hematol. 2011;90(11): 1341-1344.

Cines DB, Wasser J, Rodeghiero F, et al.

Safety and efficacy of romiplostim in splenectomized and nonsplenectomized patients with primary immune thrombocytopenia. Haematologica. 2017;102(8):13421351.

Rodeghiero F. ITP and thrombosis: an intriguing association.

Blood Adv. 2017;1(24):2280.

Psaila B, Bussel JB, Linden MD, et al. In vivo effects of eltrombopag on platelet function in immune thrombocytopenia: no evidence of platelet activation. Blood. 2012;119(17): 4066-4072.

Haselboeck J, Kaider A, Pabinger I, Panzer S. Function of eltrombopag-induced platelets compared to platelets from control patients with immune thrombocytopenia. Thromb Haemost. 2013;109(4):676-683.

Garabet L, Ghanima W, Monceyron Jonassen C, et al. Effect of thrombopoietin receptor agonists on markers of coagulation and P-selectin in patients with immune thrombocytopenia. Platelets. 2017:1-7.

Alvarez Roman MT, Fernandez Bello I, Arias-Salgado EG, et al. Effects of thrombopoietin receptor agonists on procoagulant state in patients with immune thrombocytopenia. Thromb Haemost. 2014;112(1):6572.

Justo Sanz R, Monzon Manzano E, Fernandez Bello I, et al. Platelet Apoptosis and PAI-1 are Involved in the Pro-Coagulant State of Immune Thrombocytopaenia Patients Treated with Thrombopoietin Receptor Agonists. Thromb Haemost. 2019 Feb 11. [Epub ahead of print] Bussel JB, Saleh MN, Vasey SY, Mayer B, Arning M, Stone NL. Repeated short-term use of eltrombopag in patients with chronic immune thrombocytopenia (ITP). Br J Haematol. 2013;160(4):538-546.

haematologica | 2019; 104(6)

Thrombopoietin receptor agonists in ITP 74. Gonzalez-Porras JR, Mingot-Castellano ME, Andrade MM, et al. Use of eltrombopag after romiplostim in primary immune thrombocytopenia. Br J Haematol. 2015;169 (1):111-116.

75. Carpenedo M, Cantoni S, Coccini V, Pogliani EM, Cairoli R. Response loss and development of neutralizing antibodies during long-term treatment with romiplostim in patients with immune thrombocytopenia: a case series. Eur J Haematol. 2016;97(1):101103.

76. Afdhal NH, Dusheiko GM, Giannini EG, et al. Eltrombopag increases platelet numbers in thrombocytopenic patients with HCV infection and cirrhosis, allowing for effective antiviral therapy.

Gastroenterology.

2014;146(2):442-452.e1.

77. Zhang X, Chuai Y, Nie W, Wang A, Dai G. Thrombopoietin receptor agonists for prevention and treatment of chemotherapyinduced thrombocytopenia in patients with solid tumours. Cochrane Database Syst Rev. 2017;11:CD012035.

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

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

79. Gerrits AJ, Leven EA, Frelinger AL 3rd, et al.

Effects of eltrombopag on platelet count and

haematologica | 2019; 104(6)

80.

81.

82.

83.

84.

85.

platelet activation in Wiskott-Aldrich syndrome/X-linked thrombocytopenia. Blood.

2015;126(11):1367-1378.

Rodeghiero F, Pecci A, Balduini CL. Thrombopoietin receptor agonists in hereditary thrombocytopenias. J Thromb Haemost. 2018;16(9):1700-1710.

Bussel JB, Kuter DJ, Aledort LM, et al. A randomized trial of avatrombopag, an investigational thrombopoietin-receptor agonist, in persistent and chronic immune thrombocytopenia. Blood. 2014;123(25):3887-3894.

Jurczak W, Chojnowski K, Mayer J, et al.

Phase 3 randomised study of avatrombopag, a novel thrombopoietin receptor agonist for the treatment of chronic immune thrombocytopenia. Br J Haematol. 2018;183(3):479490.

Kong Z, Qin P, Xiao S, et al. A novel recombinant human thrombopoietin therapy for the management of immune thrombocytopenia in pregnancy. Blood. 2017;130(9): 1097-1103.

Huang Y, Liu X, Xue F, et al. [Efficacy and safety of rhTPO in the treatment of pediatric primary immune thrombocytopenia].

Zhonghua Xue Ye Xue Za Zhi. 2015;36(6): 511-514.

Bussel J, Arnold DM, Grossbard E, et al.

Fostamatinib for the treatment of adult persistent and chronic immune thrombocy-

86.

87.

88.

89.

90.

topenia: Results of two phase 3, randomized, placebo-controlled trials. Am J Hematol. 2018;93(7):921-930.

Robak T, Jarque I, Musteata V, et al. Phase II, Multiple-Dose Study of Anti-FcRn Antibody, Rozanolixizumab (UCB7665), in Patients with Primary Immune Thrombocytopenia: Interim Analysis.

Blood. 2017;130(Suppl 1):15.

Kiessling P, Lledo-Garcia R, Watanabe S, et al. The FcRn inhibitor rozanolixizumab reduces human serum IgG concentration: A randomized phase 1 study. Sci Transl Med. 2017;9(414).

Shirasugi Y, Ando K, Miyazaki K, et al.

Romiplostim for the treatment of chronic immune thrombocytopenia in adult Japanese patients: a double-blind, randomized Phase III clinical trial. Int J Hematol.

2011;94(1):71-80.

Tomiyama Y, Miyakawa Y, Okamoto S, et al. A lower starting dose of eltrombopag is efficacious in Japanese patients with previously treated chronic immune thrombocytopenia. J Thromb Haemost. 2012;10(5):799806.

Janssens A, Tarantino M, Bird RJ, et al.

Romiplostim Treatment in Adults with Immune Thrombocytopenia of Varying Duration and Severity. Acta Haematol.

2015;134(4):215-228.

1123

REVIEW ARTICLE Ferrata Storti Foundation

Rituximab in the treatment of immune thrombocytopenia: what is the role of this agent in 2019?

Elisa Lucchini,1 Francesco Zaja1 and James Bussel2

SC Ematologia, Azienda Sanitaria Universitaria Integrata Trieste, Italy and 2Weill Cornell Medicine, New York, NY, USA

1

Haematologica 2019 Volume 104(6):1124-1135

ABSTRACT

T

he use of rituximab for the treatment of immune thrombocytopenia was greeted enthusiastically: it led to up to 60% response rates, making it, nearly 20 years ago, the main alternative to splenectomy, with far fewer side effects. However, long-term follow-up data showed that only 20-30% of patients maintained the remission. No significant changes have been registered using different dose schedules and timing of administration, while the combination with other drugs seemed promising. Higher response rates have been observed in young women before the chronic phase, but apart from that, other clinical factors or biomarkers predictive of response are still lacking. In this review we examine the historical and current role of rituximab in the management of immune thrombocytopenia, 20 years after its first use for the treatment of autoimmune diseases.

Introduction Correspondence: FRANCESCO ZAJA francesco.zaja@asuits.sanita.fvg.it Received: February 28, 2019.

Accepted: May 9, 2019.

Pre-published: May 24, 2019. doi:10.3324/haematol.2019.218883 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/104/6/1124 Š2019 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

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

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

Primary immune thrombocytopenia (ITP) is an autoimmune bleeding disorder due to a variable combination of increased platelet destruction and impaired platelet production, as a consequence of defects in central and/or peripheral immune tolerance which allow the escape of autoreactive lymphocytes.1â&#x20AC;&#x201C;3 B cells have a well-established role in the pathogenesis of the disease, as the source of antibodies directed against platelet-surface glycoproteins.4â&#x20AC;&#x201C;6 Rituximab, a monoclonal antibody directed against CD20, a membrane glycoprotein expressed on the surface of B cells, was introduced for the treatment of Bcell lymphomas towards the end of the 1980s.7 Binding to an antigen that is only expressed on mature B cells, rituximab leads to a fast and deep, but reversible B-cell depletion.8 The transience of the B-cell depletion and the low toxicity profile represented the rationale for its use in the treatment of autoimmune conditions, especially those in which B-cell activity was considered the main pathogenic mechanism, such as ITP. Many studies have been carried out in this field: in monotherapy, with different dose schedules and in combination with other drugs, proving its efficacy, although some differences exist across certain studies.

Rituximab has also been explored in a number of other autoimmune auto-antibody-mediated diseases such as systemic lupus erythematosus,9 rheumatoid arthritis,10 autoimmune hemolytic anemia,11 type II mixed cryoglobulinemia,12 myasthenia gravis,13 multiple sclerosis,14 thrombotic thrombocytopenic purpura,15 Sjogren syndrome,16 pemphigus17 and others. Despite these extensive investigations, autoimmune conditions for which rituximab is licensed by the Food and Drug Administration and the European Medicines Agency are rheumatoid arthritis and ANCA-associated vasculitis.

In this review, we discuss the development and current role of rituximab in the management of ITP.

Pathophysiology of immune thrombocytopenia The milestone role of autoantibodies in the pathogenesis of ITP was first reported in 1951 by Harrington et al., who showed that the infusion of plasma from ITP

1124

haematologica | 2019; 104(6)

The role of rituximab in ITP

patients into normal controls caused thrombocytopenia, thus imputing the cause of the disease to a plasma-derived factor.4 This “factor” was subsequently identified as an IgG anti-platelet antibody, directed against platelet glycoprotein (GP) IIb/IIIa and/or the GPIb-IX-V complex.5 Very rarely, antibodies against GPIa-IIa or GPIV can be found (5%).18 Antibody-opsonized platelets are then recognized through the Fcγ-receptors by macrophages in the spleen, liver and bone marrow, phagocytized and prematurely destroyed.19 Other mechanisms through which antibodies can mediate platelet destruction are complement deposition with intravascular lysis and induction of platelet apoptosis.20–23 Plasma from patients with ITP also inhibits megakaryocyte growth and function in the bone marrow.6,24 The Ashwell-Morell receptors in hepatocytes have been invoked as a further pathogenic mechanism, because they physiologically remove desialylated, “old” platelets from the circulation. Anti-GPIb/IX autoantibodies are thought to enhance the desialylation of GPIb, increasing hepatic clearance of platelets.25,26 Many abnormalities have been shown in T cells of patients with ITP: an altered Th1/Th2 balance, with an increased number of Th1 T-helper cells27,28 and a decrease in the number and function of regulatory T cells.29,30 The abnormal activation of cytotoxic CD8+ T cells may also have a role in the pathogenesis of ITP, contributing to both platelet destruction and impaired platelet production.31 Circulating thrombopoietin levels in ITP are not increased proportionally to the level of thrombocytopenia, and are usually normal or only slightly increased.32

Early history of rituximab In the late 1980s, the idea of using monoclonal antibodies that recognize tumor-associated antigens for the treatment of hematologic malignancies became reality, and rituximab became a well-tolerated and highly effective option initially used for patients with multi-refractory lymphoproliferative diseases.7 CD20, a transmembrane glycoprotein expressed on the surface of normal and malignant B cells, appeared ideal for targeted therapy, because it does not shed from the cell surface and is not internalized upon antibody binding.33 CD20 is expressed from early pre-B to mature B lymphocytes, but is not expressed on hematopoietic stem cells, plasma cells or other cells of the body.34 Rituximab is a type 1 IgG1-κ human-mouse chimeric monoclonal antibody directed against CD20, which acts through three mechanisms: complement-dependent cytotoxicity, antibody-dependent cellular cytotoxicity and induction of direct apoptosis of the target cell.35 The first report of a case in which rituximab was used for the treatment of an autoimmune disease was published in 1998, when a patient with a cold agglutinin disease and a small IgM paraprotein was successfully treated with four weekly infusions of rituximab.36 A few years later, in 2001, the first report of the successful use of rituximab for the treatment of ITP associated with a low-grade non-Hodgkin lymphoma was published.37 Since then, rituximab has been widely used for the treatment of autoimmune manifestations associated with lymphoproliferative disorders, and, because it is a therapy borrowed from lymphomas, the dose schedule of 375 mg/m2 weekly for 4 haematologica | 2019; 104(6)

weeks also became the “standard dose” of rituximab in autoimmune diseases.

Since CD20 is also expressed in normal B cells, there was a strong rationale for using rituximab to deplete pathological antibodies in autoimmune diseases. One of the first cases, published in 2000, was that of a young man with refractory myasthenia gravis who, after four standard doses of rituximab, experienced a complete clinical response with disappearance of anti-acetylcholine receptor antibodies, and he did not relapse.13 This case was an impressive proof of principle that transient B-cell depletion secondary to rituximab treatment may positively modulate the immune system, inhibit the production of autoantibodies, and cause long-term clinical improvement.

Rituximab in immune thrombocytopenia In 2001, Stasi et al. reported the results of the first prospective study in which 25 patients with chronic ITP were treated with four weekly infusions of rituximab at a dose of 375 mg/m2. The overall response rate (ORR) was 52%, with 28% sustained responses. No clinical or laboratory parameters were found to predict treatment response, but they noticed that women and younger patients had a better chance of response, and that, in some patients who relapsed, retreatment was effective.38 Subsequent studies showed that the overall initial response to rituximab used as second- or further-line of therapy ranges between 52% and 73%, with the complete response (CR) rate ranging between 20% and 54%39-48, 56, 58, 68 (Table 1). In a systematic review including 313 adult patients with chronic ITP, a response rate of 62.5% and a CR rate of 46.3% were found, with a median duration of response of 10.5 months.42 Three different patterns of response can be distinguished: a first group of patients respond rapidly, within the first month; in the second group, the platelet count starts increasing after 3-4 weeks, and a CR is achieved within 8 weeks after treatment; in the third group of patients, the platelet count increases very slowly, only reaching normal values 3 months after therapy.46 Only two randomized, placebo-controlled studies have been performed. One very small pilot randomized trial compared rituximab with placebo as second-line therapy.

No difference in terms of treatment failure (65.6% vs.

80.8%)47 was found between the two groups. Overall platelet count response (i.e. platelet count ≥30x109/L) was achieved by 62.5% of the patients in the rituximab group and 73.1% in the placebo group at 6 months.

In the much larger RITP trial,48 the rate of treatment failure was not significantly different between patients given rituximab or placebo (58% vs. 69%), and the ORR was 81% in the rituximab group compared to 73% in the placebo group. Improvements in platelet counts were seen up to week 72 in the rituximab arm.

Rituximab doses “Standard dose” rituximab results in a marked reduction of malignant and non-malignant B cells in peripheral blood and bone marrow.49 Since the total mass of B cells is much smaller in patients with ITP than in patients with lymphoma, it was not clear whether a lower dose of rituximab or a different schedule could be equally effective.

1125

E. Lucchini et al.

Potential advantages of the lower dose include avoidance of severe side effects, steroid-sparing effects, the greater possibility of administering repeated courses, and decreased cost. Zaja et al. investigated the efficacy of rituximab given at a dose of 100 mg (“low dose”) weekly for 4 weeks in 48 ITP patients.50 In an indirect, non-randomized comparison, both the initial response (ORR 60.5% and CR 39.5%) and the duration of response (12- and 24-month cumulative relapse-free survival rates of 61% and 45%, respectively) were moderately lower in this group than in patients treated with “standard dose”. The time to response was also longer than that observed with the “standard dose”. This may be due to the fact that the depth of B-cell depletion reached in peripheral blood might not correlate with the depletion in other organs, and the 100 mg dose is probably not enough.

A recent UK study retrospectively compared 113 patients who received “standard dose” rituximab to 169 who received the “low dose”. They found that the low dose was not significantly different from the standard dose with regards to ORR (at 2 months, 56% vs. 59%; at 6 months, 62% vs. 64%), time to maximum platelet count (77 vs. 74 days), time to next treatment (4.6 vs. 4.3 months) and duration of response.51 Some groups explored a fixed dose of rituximab, 1000 mg given twice on days 1 and 15, which is the dose schedule licensed for rheumatoid arthritis. Khellaf et al., in a prospective registry including 248 patients, compared the “standard dose” to the “rheumatoid arthritis-like” regimen. They did not find any difference in terms of initial or long-term response between the two groups.41 In a multicenter, single-arm study (R-ITP1000 study), Tran et al.

found that the “rheumatoid arthritis-like” fixed dose led to an ORR of 44% at week 8 in patients with relapsed/refractory ITP.52 A multicenter, randomized, phase II Dutch trial compared three rituximab dosing schemes in 156 patients with relapsed or refractory ITP: “standard dose” rituximab, two weekly 375 mg/m2 doses and two weekly 750 mg/m2 doses. Response rates were similar within the three arms (63%, 59% and 61%, respectively), with a relapsefree survival of 72% at 1 year and 58% at 2 years.53 The results of the most relevant studies with different dosing schedules of rituximab are summarized in Table 2.

Factors predictive of response Over the years, multiple factors have been investigated with the aim of predicting response to treatment. Several studies highlighted the correlation of age and gender with outcome. In the very first study, Stasi had already pointed out that women and young patients had better responses.38 This finding was subsequently extended by Bussel et al.,54 who showed that women of child-bearing age whose duration of ITP was less than 24 months had a long-term response comparable with that obtained after splenectomy (60% long-term treatment-free responses). Another study with the same treatment schedule pointed out that adolescent females with an ITP duration of less than 12 months had the longest duration of response.55 Similar results were also reported with rituximab alone: young (<40 years) women had a significantly higher probability of achieving a response (73%), a complete response (56%), and as well as a better long-term response (47% after 72 months) compared with the other groups.56 It must be noted, however,

Table 1. Most relevant studies with rituximab administered at a standard dose of 375 mg/m2 weekly for 4 weeks in patients with immune thrombocytopenia.

Author

Number of Median age, patients years (range)

F:M

Stasi 200138 Cooper 2004£, 46

25 57

46 (22-74) 46 (21-79)

64%:36% 68%:32%

Zaja 2006£, 58

37

NA

NA

Medeot 2008£, 68

26

55 (18-76)

81%:19%

Godeau 2008£, 39 60 Cervinek 2012£, 43 114

48 (18-84) 55 (21-89)

67%:33% 53%:47%

Patel 2012£, 40 Arnold 2012£, 47 Mahevas 2013£, 45

72 33 61

39 (18-78) 40 (30-59) 52 (34-70)

65%:35% 58%:42% 64%:36%

Khellaf 2014$, 41 173 Ghanima 2015$, 48 55

51 (21-71) 46 (27-61)

64%:36% 73%:27%

Marangon 2017£, 56 103

46 (15-82)

59%:41%

ITP phase Chronic Chronic

ORR

6 months

52% (20% CR)° 54% (32% CR)° 73% (54% CR)°

NA NA

12 months 24 months 36 months NA NA

Median ITP NA NA duration: 34.5 (1-264) months Median ITP 69% NA 55% duration 34.5 (54% CR)° (4 - 264) months Chronic NA NA 40% (30% CR)° 16% ND; 84% chronic NA 72% (48% CR)* ORR 69% or persistent phase (45% CR) Most chronic phase 57%° NA 38% 50% ND NA 62.5% (53% CR)* NA 18% ND; 26% 54% NA 36% (28% CR) persistent phase; (32% CR) 56% chronic phase at 3 months* 56% chronic 62%* 80% 62% 33% ND; 24% 73% 60% 45% persistent phase; (51% CR)* 44% chronic phase Median ITP duration 55% NA NA 20 (1 - 403) months (36%CR)*

Last follow-up

NA NA

NA NA

NA 32% at a median FU of 72 weeks (18 months) 40.5% at a median FU of 25 months (3-55)

NA

NA

45%

41%

35% at a median FU of 57 months

33% NA

NA NA

NA NA

31% NA NA

21% at 5 years FU NA NA

50% NA

NA NA 31% (26% CR) median FU of 36 months 38% NA

42%

40%

21% at 96 months

NA 24% at 78 weeks FU

F:M: female to male ratio; ITP: immune thrombocytopenia; ORR: overall response rate; CR: complete response; NA: not available; FU: follow-up; ND: newly diagnosed. £Considering as denominator all the patients.$Considering as denominator only patients who responded to rituximab. *response = platelet count ≥30x109/L. °response = platelet count ≥50 x 109/L

1126

haematologica | 2019; 104(6)

The role of rituximab in ITP

that in several other studies the predictive role of age and gender could not be confirmed.40,41,43,46,57 A disease duration of less than 12 months has also been frequently related to better outcomes,41,43,54,58 as well as the achievement of complete remission,39,40,46,54,56 while the influence of a previous splenectomy is not completely clear.40,46 The role of antiplatelet autoantibodies (APA) as factors predictive of response is controversial. Differently from other autoimmune diseases, APA in ITP are neither very specific nor sensitive, and for this reason APA are not currently recommended as a diagnostic test for ITP.59,60 A reduction of APA levels has been associated with an increase in platelet count,61 and the presence of platelet-bound antibodies has been associated with a better response to rituximab.62 The persistence of autoantibodies in non-responders may suggest that rituximab had not removed the longlived, antibody-producing plasma cells. There is also a small cohort of patients who respond despite undetectable antibodies: in those cases, either laboratory assays are not able to identify the antibody, or the response comes from the elimination of B-cell-mediated activation of T cells.61 A more recent study did not show any correlation between the presence of APA and the response to ritux-

imab, but found that rituximab resulted in a significant reduction of anti-GPIIb/IIIa but not anti-GPIb/IX levels.63 In an unconfirmed study, patients with anti-GPIb/IX were shown to have a lesser response to intravenous immunoglobulins and steroids.64,65 In patients treated with rituximab, the presence of antibodies against GPIIb/IIIa led to a higher response rate than that in patients without anti-GPIIb/IIIa, while the presence of anti-GPIb/IX did not significantly influence the outcome.66 This is potentially due to the different modes of action of the autoantibodies: anti-GPIIb/IIIa antibodies induce platelet destruction by Fc-dependent phagocytosis, while the action of antiGPIb/IX may be FcR-independent67 and instead increase the hepatic clearance of desialylated platelets.

In summary, the role of APA in response to rituximab remains unclear and may depend on the laboratory doing the testing, the phase of the disease, and which tests are performed.

Long-term outcome of rituximab treatment of immune thrombocytopenia Only a small proportion of patients maintain a longterm remission after rituximab: in a prospective French

Table 2. Most relevant studies with different dose schedules of rituximab.

Author

Rituximab dose

Mahevas 2013£,45

375 mg/m2 weekly x 4

61

1000 mg on days 1 and 15 375 mg/m2 weekly x 4 1000 mg 2 weeks apart 1000 mg on days 1 and 15

Khellaf 2014$,41

Tran 201452

Number Median age, F:M of patients years (range)

ITP phase

Early response

52 (34-70) 64%:36%

ND 18%; P 26%; C 56%

54% (32% CR) at 3 months*

NA

36% (28% CR)

NA

NA

46

55 (34-76) 67%:33%

54% (32% CR) at 3 months*

NA

50% (41% CR)

NA

NA

173

51 (21-71) 64%:36%

ND 17%; P 26%; C 57% 56% chronic

62%*

80%

62%

50%

38%

31% (26% CR) Median FU 36 months 48% (41% CR) Median FU 20.5 months NA

72

53 (33-73) 65%:35%

66% chronic

61%*

90%

70%

43%

/

NA

108

49 (19-85) 57%:43%

Median ITP 43.5% (CR 9.3%) NA duration at week 8° 24.8 months (1.2 - 470) Median ITP 60.5% (CR 39.5%)° NA duration 16 months (2 - 451) Median ITP 52% (28% CR)* NA duration 24 months (2-324) Median ITP 66% (50% CR)* NA duration 31 months (3-264) Median ITP NA 61.7% duration (35% CR)* 12.9 months (3.1-62) median ITP NA 64.1% duration (42% CR)* 12 months (2.9-43.5)

29%£ 66%$

NA

NA

NA

61%

45%

40%

NA

32%

28%

23%

23% at 48 months

50%

45%

40%

35% at 48 months

60%

53%

50%

NA

60%

51%

40%

NA

Zaja 2010$,50

100 mg weekly x 4

48

41 (16-74) 62%:38%

Zaja 2012£,44

100 mg weekly x 4

25

43 (14-74)

NA

375 mg/m2 weekly x 4 32

51 (16-80)

NA

Gracie 2018$,51

100 mg weekly x 4

169

57 (30-70) 57%:43%

375 mg/m2 weekly x 4 113

59 (35-72) 42%:58%

6 months 12 months 24 months 36 months Last follow-up

F:M: female to male ratio; ITP: immune thrombocytopenia; ND: newly diagnosed; P: persistent phase ITP; C: chronic phase ITP; NA: not available; CR: complete response; FU: follow-up. $Response rates calculated considering as denominator only patients who responded to rituximab. £Response rates calculated considering as denominator all treated patients. *response = platelet count ≥30x109/L ° response = platelet count ≥50x109/L

haematologica | 2019; 104(6)

1127

E. Lucchini et al.

study in 248 adult patients, Khellaf et al. found that after a median follow-up of 24 months, 39% of patients were still responding.41 In other studies a long-term response, at approximately 2 years after initial treatment, was observed in about 40% of patients.39,56,68 However, the only study with a 3-5 year follow up showed that the response was maintained after 5 years in only 21% of adults and 26% of children treated with rituximab.40 Given that a long-term response greater than 20-30% would be desirable, several strategies have been implemented in order to augment response rates and sustained remissions.

Anticipated use of rituximab in patients with immune thrombocytopenia Although not supported by randomized, controlled trials, some studies pointed out that better outcomes can be achieved if rituximab is administered in an early stage of the disease.46,58 Following this observation, two studies69,70 compared the combination of “standard dose” rituximab and dexamethasone (40 mg/day for 4 days) to dexamethasone alone as first-line therapy in adult patients with ITP. In both studies the combination led to higher sustained response rates (63%69 vs. 36% and 58%70 vs. 37%), compared to those achived with dexamethasone alone.

Combination treatment to improve long-term outcome with rituximab Since rituximab affects almost exclusively B cells, without directly affecting the activity of other cells of the immune system (in particular T cells and plasma cells), the combination with other drugs with different modes of action looked appealing. The addition of 28 mg/m2 dexamethasone (as an anti-plasma cell treatment), given for three 4-day cycles at 2-week intervals, to “standard-dose” rituximab was explored in a cohort of 67 patients (41 adults and 26 children) with ITP, of whom only five were treatment-naïve. This combination led to a 75% initial response rate and an almost 50% estimated long-term cure rate at 5 years.54 However, the good long-term responses were seen almost exclusively in women of child-bearing age within 1 year of diagnosis.

The combination of high-dose dexamethasone, lowdose rituximab and cyclosporine (TT4) administered over 1 month was tested in 20 ITP patients (including 7 with newly diagnosed ITP and 5 with secondary ITP), with the aim of targeting, in addition to B cells, also plasma cells and T cells. The response rate at 6 months was 60% and among responders, the relapse-free survival rate was 92% at 12 months and 76% at 24 months. The treatment was well tolerated.71 Two Chinese studies explored the combination of lowdose rituximab with recombinant human thrombopoietin.

The first study enrolled 14 patients with refractory ITP, who had an ORR of 93%.72 The second was a randomized, open-label study in which the combination was compared with rituximab 100 mg weekly for 4 weeks in patients with relapsed or refractory ITP. The group treated with the combination had a substantially shorter time to response (7 vs. 28 days).73 The long-term response rate (79.2% vs. 71.1%) was not significantly different between the two groups.

Gómez-Almaguer et al. recently published the results of a single-center, pilot study conducted to assess the safety and efficacy of the combination of eltrombopag, low-dose 1128

rituximab and dexamethasone in 13 newly diagnosed ITP patients. The ORR was 100%, with 92% CR rate and a relapse-free survival rate of almost 80% at 12 months.74 The results of the most relevant studies of the use of rituximab in combination with dexamethasone and other drugs in ITP are summarized in Table 3a and 3b, respectively.

Efficacy of rituximab retreatment Limited data are available concerning retreatment with rituximab. In a retrospective study, Hasan et al. explored the response to retreatment with standard-dose rituximab in patients with chronic ITP: 80% of the retreated patients responded again to standard-dose rituximab.75 Zaja et al. showed that even low-dose rituximab can be effective in patients who previously responded to the standard dose of the drug, although only three patients received a second course of rituximab.50 Khellaf et al. reported that most of 11 patients retreated with rituximab responded again.41 In other studies, very small numbers of patients were retreated with rituximab, in most of cases with good responses.38,56,68,76–78 Rai et al., in 17 patients with autoimmune cytopenias (including 11 cases of ITP) who previously responded, but then relapsed after a standard course of rituximab, explored the use of rituximab maintenance: a single 375 mg/m2 infusion every 4 months for a total of 2 years: 88% of patients achieved a CR, with a mean duration of response of 48 months.79

B-cell recovery after rituximab The depletion of circulating B cells after rituximab administration is rapid (within 1 week) and deep, with Bcell counts remaining low in the peripheral blood for at least 6-12 months.8 The repopulating pool is dominated by immature B cells, while memory B cells recover after 2 years.80 A similar depth of peripheral B-cell depletion has also been observed with low-dose rituximab.50 In the vast majority of patients who achieve a complete remission, the recovery of B cells is not associated with disease relapse,46 while in non-responders or in patients who relapse, B cells tend to reappear sooner in the peripheral blood and increase to higher levels.40,46 Rituximab also induces nearly complete depletion of splenic B cells, and in patients who do not respond the recovery of B cells in the spleen is faster.81

Effects of rituximab on T cells In rituximab responders, all the immune-system abnormalities seen in patients with active disease tend to revert towards normal: in the peripheral blood, there is restoration of the Th1/Th2 and Tc1/Tc2 ratios, a decreased expression of Fas ligand and Bcl-2 mRNA, an increased expression of Bax mRNA and an increased number of regulatory T cells82,83. The numbers of interleukin-10-producing B cells (regulatory B cells) and interleukin-6-producing B cells are also normalized after treatment.84 In contrast, these abnormalities are still detectable in the spleens of non-responders: reduced regulatory T cells, an increased Th1/T-regulatory cell ratio,81 and persistence of the Tc1 polarization with CD8+ cytotoxic T cells displaying the phenotype of effector memory T cells with a restricted T-cell receptor repertoire.85 These findings suggest that the action of rituximab is not limited to B cells and humoral immunity, but that rituximab also affects celhaematologica | 2019; 104(6)

The role of rituximab in ITP

mal activation of T cells, whose activity is not switched off by B-cell depletion. Finally, if the patient does not actually have ITP this would likely explain non-response as well.

lular immunity. In a mouse model of T-cell-mediated ITP, B-cell depletion resulted in a significantly decreased proliferation of splenic CD8+ T cells in vitro, which correlated with an in vivo normalization of platelet counts.86 However, T-cell remodulation in responders seems not to be a specific effect of rituximab since similar changes have also been observed in patients who respond to dexamethasone or thrombopoietin receptor agonists.87–89 It is possible that other biological factors concur to this effect.

In particular, it has been proposed that the platelet count increase itself, either directly or via release of transforming growth factor-b, may cause a positive immune-modulating effect.90 Why a proportion of patients do not respond to rituximab is unknown. This could be caused by either the persistence of long-lived, antibody-secreting plasma cells in the spleen81,91 and/or in the bone marrow or by the abnor-

Adverse events Rituximab is generally a well-tolerated therapy, and adverse events are usually mild and easily manageable.

The major concerns derive from the possible induction of hypogammaglobulinemia, and the increased risk of certain infections.

Infusion reactions after the first administration of rituximab are experienced by a variable proportion of patients, ranging from nearly 60% in the first trials46 to 15% in the more recent reports,41 and are related to immediate cytokine release. They are usually easily manageable, and

Table 3a. Studies with rituximab plus dexamethasone.

Author

Treatment

Zaja 201069

Number Median age, of patients years (range) F:M

375 mg/m2 x 4 weekly + 1 cycle dexa (40 mg/day for 4 days) Gudbrandsdottir 375 mg/m2 201370 x 4 weekly + up to 6 cycles dexa (40 mg/day for 4 days every 1 to 4 weeks) Bussel 375 mg/m2 x 54 2014 4 weekly + 3 cycles dexa (28 mg/m2/day for 4 days) Chapin 201655 375 mg/m2 x 4 weekly + 3 cycles dexa (28 mg/m2/day for 4 days)

ITP phase

Early response

6 months 12 months 24 months 36 months

49

49 (33-65) 55%:45%

Previously untreated

37%° at week 4

63%

50%

47%

43%

62

51 (36-63) 58%:42%

ND

NA

57%°

53%

NA

NA

41

36 (18-64) 53%:46% Median ITP 88% at duration: 16 week 8° (1-286) months

82%

65%

58%

50%

NA

NA

NA

NA

49

37

55%:45% Duration of ITP: range 0 – 258 months

NA

Last follow-up 43% 30-months estimated probability of duration of response NA

47% estimated sustained response at 64 months FU 33.3% sustained response at 72 months FU°

F:M: female to male ratio; ITP: immune thrombocytopenia; dexa: dexamethasone; ND: newly diagnosed; NA: not available; FU: follow-up; yrs: years. Response rates are calculated using as denominator all the patients treated with rituximab.°response = platelet count ≥50x109/L

Table 3b. Studies with rituximab in combination with drugs other than dexamethasone.

Author

Treatment

Zhou 2015£,73

Rituximab 100 mg weekly for 4 weeks + rhTPO 300 μg/kg/day for 14 days Rituximab 100 mg weekly for 4 weeks

Choi 2015$,71

Number Median age, of patients years (range) 77

42 (13-82)

38

42.5 (12-68)

Rituximab 100 mg weekly for 4 weeks 20 + dexa 40 mg days 1-4 + cyclosporine 2.5-3 mg/kg/day (days 1-28) Li 2015£,72 Rituximab 100 mg weekly for 4 weeks 14 + rhTPO 300 μg/kg/day for 14 days Gomez-Almaguer Eltrombopag 50 mg/day, days 1-28 13 2017£,74 + dexa 40 mg days 1-4 + rituximab 100 mg weekly for 4 weeks

NA

F:M

ITP phase

ORR

65%:35%

Median ITP 93% (50% CR)* duration 12.5 months (3-72) 66%:34% Median ITP 93% (50% CR)* duration 11 months (3-65) 55%:45% ND or Persistent NA ITP: 7; Chronic ITP: 13

52 (18-76)

93%:7%

Unknown

40 (16-61)

62%:38%

ND

6 months 12 months 24 months 67.2%

24.6%

NA

55.6%

18.5%

NA

60%*

92%

76%

93% (50% CR)* 71% (40% CR)

NA

NA

100% (92% CR)*

80%

70%

NA

F:M: female to male ratio; ITP: immune thrombocytopenia; ORR: overall response rate; rhTPO: recombinant human thrombopoietin; CR: complete response; NA: not available; dexa: dexamethasone; ND: newly diagnosed. $Response rates calculated considering as denominator only patients who responded to rituximab. £Response rates calculated considering as denominator all treated patients. *response = platelet count ≥30x109/L; °response = platelet count ≥50x109/L.

haematologica | 2019; 104(6)

1129

E. Lucchini et al.

serious adverse reactions are exceptional, especially if prednisone is included in premedication.

Serum sickness is a much rarer adverse reaction, characterized by fever, rash, polyarthralgia or arthritis, proteinuria, hematuria, elevated inflammatory markers and decreased complement, which usually arises 10-14 days after treatment.92 It is the result of immune activation against the chimeric mouse-human drug, with the formation and deposition of immune complexes, and consequent activation of the complement cascade.93 This type III delayed hypersensitivity reaction to rituximab has been reported more commonly in autoimmune disorders than in hematologic malignancies, but overall it remains a very rare occurrence in adults, with less than 50 cases having been reported.92

Hypogammaglobulinemia Since rituximab does not affect pre-existing long-lived plasma cells, there are not significant changes in the IgG or IgM levels in patients treated with one “standard dose” course.38,46–48 However, patients treated with multiple courses are more likely to develop hypogammaglobulinemia, which usually recovers spontaneously in a few months.94 Since dexamethasone also affects plasma cells, 10-20% of patients treated with the combination may experience a marked hypogammaglobulinemia, which usually recovers within 1 year.50,54,70 Some reports suggest that a pre-existing hypogammaglobulinemia can be aggravated by rituximab, which can in some cases trigger or accelerate the development of a real common variable immunodeficiency (CVID).95,96 It is worthwhile remembering that ITP may be the initial manifestation of CVID and can precede its diagnosis by years.97 It is therefore important, in ITP patients treated with rituximab, to monitor immunoglobulin levels before and after therapy. In addition to monitoring and possibly even temporarily giving intravenous immunoglobulin replacement, prior assessment of genetic markers and vaccine responses98,99 may be useful.

The efficacy and safety of standard-dose rituximab in patients with CVID-associated ITP or autoimmune hemolytic anemia was assessed in a multicenter, retrospective, French study. The ORR was 85%, with a 74% CR rate and a sustained response rate of 60% at a median follow up of 39 months; severe infections occurred in 24% of patients, four of whom were not on immunoglobulin replacement therapy. The authors concluded that rituximab is highly effective and relatively safe in the management of CVID-associated immune cytopenias, and that immunoglobulin replacement is strongly recommended in this cohort of patients.100

lived plasma cells, thus impairing the immune response to vaccines. Vaccines against encapsulated bacteria (Streptococcus pneumoniae, meningococci and Haemophilus influenzae) should be administered before rituximab, considering that patients, especially non-responding ones, may require a subsequent splenectomy.103 Reactivation of hepatitis B virus (HBV) is a well-recognized complication of immunosuppressive therapy.

Patients who are HBcAb positive (likely occult carriers) should receive antiviral prophylaxis with lamivudine, while patients with HBsAg or HBV-DNA positivity (active carriers and inactive carriers) should be referred to a hepatologist and treated with entecavir or tenofovir.104 A rare but life-threatening infection that has been linked to rituximab is progressive multifocal leukoencephalopathy (PML), caused by the activation of Jakob-Creutzfeld virus and its spread to the central nervous system.105 This complication, although rare, has been almost only reported in patients with lymphoproliferative disorders, and only a very few cases have been observed in patients with autoimmune diseases, especially in systemic lupus erythematosus.106,107 Furthermore, patients who develop PML are usually profoundly immunocompromised from combination chemotherapy, and rituximab is often just one of the many drugs received.108,109 There is only one well-studied case of PML in ITP and the course was unusual with the PML occurring more than 3 years after exposure to rituximab.

The occurrence of late-onset (>4 weeks after treatment) neutropenia has been described in patients treated with rituximab for both malignant110 and non-malignant conditions,111,112 including a few ITP patients.40 Most cases appear to be self-limiting and resolve without issue,113 and according to some authors retreatment with rituximab is safe.86

Malignancies Immunosuppression secondary to rituximab could increase the risk of second primary malignancies. In large follow-up studies, rituximab has not been shown to increase the risk of cancer in patients with rheumatoid arthritis,114 among non-Hodgkin lymphoma patients115 and in patients with ANCA-associated vasculitis.116 No malignancies were reported in ITP by Arnold et al. in a meta-analysis including more than 300 patients,42 or in other studies,40,41,58 including combination studies with dexamethasone.54,69,70 In other reports nearly 3-4% of patients developed a second primary malignancy after having received rituximab for ITP, but the low percentage and the heterogeneity of the neoplasia led the authors to conclude that a causative relationship with rituximab could not be assessed.39,44,56

Infections An increased risk of infections after rituximab therapy is generally uncommon and more frequently observed in severely immunocompromised patients.101 Chugh et al., in a meta-analysis including five trials and 463 ITP patients treated with rituximab, did not find an increased risk of infection.102 Khellaf et al. concluded that the risk of infections was acceptable (cumulative incidence of 2.3 infections per 100 patient-years at a median follow up of 24 months), with the most severe infections occurring in adults older than 70 years of age, who suffered from severe comorbidities.41 Rituximab therapy reduces the genesis of new long1130

Use of rituximab in children Following the development of rituximab for adults with ITP, studies soon migrated to pediatric patients (Table 4).

The first large series was reported in 2005 and included 24 chronic ITP patients who were refractory to or relapsed after previous treatments. The ORR to “standard-dose” rituximab was 78%, with a 63% CR rate and an overall sustained response rate of 37%.117 The first prospective phase I/II study of rituximab in children and adolescents with chronic ITP included 36 haematologica | 2019; 104(6)

The role of rituximab in ITP

Table 4. Rituximab in children.

Author

Rituximab dose Number Median age (mg/m2) of patients years (range)

Wang 2005117 Taube 2005121 Bennett 2006118 & Mueller 2009119 Parodi 2009123 124

Grace 2012 Patel 201240

Matsubara 2014125

Oved 2017122

F:M

ITP phase

ORR (CR)

12 months

24 months

36 months

Last follow-up

78%* (63%) 59%* (32%) 31%°

NA

NA

NA

NA

NA

NA

NA

NA

31%

NA

NA

NA

NA NA 33%

NA NA NA

60% NA NA

NA NA 26% at 5 years

27%

18%

NA

14% at 5 years

NA

NA

NA

27% at a median

375 mg/m2 x 4 weekly 375 mg/m2

24

12 (2-19)

58%:42%

Chronic

22

5.8 (2.5-15.2)

64%:36%

Chronic

375 mg/m2 x 4 weekly 375 mg/m2 x 4 weekly NA 375 mg/m2 x 4 weekly

36

11.2 (2.6-18.3)

42%:58%

Chronic

49 80 66

7.4 (0.7-17.6) 7.5 (4.9-12) 12 (2-17)

67%: 33% 56%:44% 61%:39%

375 mg/m2 x 4 weekly

22

4.2 (0.25-11.6)

Rituximab 375 mg/m2 x 4 weekly + dexamethasone 28 mg/m2 days 1-4; 15-18; 29-32

33

11 (1-17)

77% chronic 69%° (53%) Chronic 64%° Median 57%° duration of ITP 20 months (1 month – 9 years) 50%:50% Median 50%* duration of (41%) ITP 18.5 months (2 – 120 months) 58%:42% Persistent/ 48%° chronic ITP

FU of 35 months (range 10-58)

F:M: female to male ratio; ITP: immune thrombocytopenia; ORR: overall response rate; CR: complete response; NA: not available; FU: follow-up; Response rates are calculated using as denominator all the patients treated with rituximab. *response = platelet count ≥30x109/L; °response = platelet count ≥50 x 109/L; §sustained platelet count over 50x109/L in 4 consecutive weeks.

patients with severe refractory ITP or Evans syndrome treated with “standard-dose” rituximab. After a follow up of 1 year, 31% of them maintained a platelet count >50x109/L.118,119 A systematic review including 14 studies with a total of 323 pediatric ITP patients reported a pooled response rate of 68%, and a pooled CR rate of 39%, with a median duration of response of 12.8 months.120 “Low-dose” rituximab has also been tested in children with ITP: Taube et al. explored the efficacy of a single dose of rituximab (375 mg/m2) in 22 patients with chronic ITP; the ORR was 59%, with a 27% CR rate, and 36% of patients maintained a long-term remission (median duration of remission 13.5 months; range 2-16 months).121 Oved et al. explored the addition of three 4-day cycles of dexamethasone (28 mg/m2) to “standard-dose” rituximab in 33 children with persistent/chronic ITP. The ORR was close to 50%, and 62% of the responders maintained the remission for a median of 35.5 months.122 Factors predictive of response were also sought in children. Bennett et al.118 found a weak association between response and Evans syndrome, female sex and black race.

Parodi and colleagues performed a retrospective study including 49 children (77% with chronic ITP) treated with “standard-dose” rituximab (ORR 69%, 60% relapse-free survival at 36 months), and found a significantly higher probability of relapse-free survival in males aged >14 years and females aged >12 years (88.9% vs. 56.7%), in patients who achieved a CR (70.2% vs. 25%) and in patients who achieved the response within 20 days of treatment (73.7% vs. 22%).123 Among 80 pediatric patients with chronic ITP treated with rituximab (ORR 64%), Grace et al. found a higher haematologica | 2019; 104(6)

response rate in patients who had previously responded to steroids (87.5% vs. 47.9%) and in those with secondary ITP (89.5% vs. 55.7%).124 In a study by Oved et al.,122 female adolescents with ITP lasting less than 24 months had a higher sustained remission rate (47%) than that of either the entire group (27%) or the male patients (7%).

The drug is usually well tolerated also in children, with a toxicity profile superimposable to that of adult patients in terms of immediate and long-term toxicity.117,118,125,40,123 Significant hypogammaglobulinemia was observed in 15% of patients treated with the combination of rituximab and dexamethasone,122 although this adverse event was transient.

The reported rate of serum sickness was higher in pediatric patients than in adults, particularly in the first series, occurring in up to 10% of the former.117,118 Clinical manifestations ranged from severe to mild. Laboratory confirmation can be sought by checking levels of C3 and C4; of note, certain pediatric patients may have a congenitally low C4 level (which may predispose to ITP) and C4 levels should, therefore, be checked prior to rituximab treatment.

Discussion Twenty years after the first use of rituximab in ITP, published studies show an ORR of nearly 60% and a CR rate of 50%. Long-term remissions occur in 20-30% of patients, with slightly different outcomes possibly related to different selection of populations of patients.

In the interpretation of the results it is worth noting that 1131

E. Lucchini et al.

heterogeneous criteria for response were adopted across different studies (according to older or updated criteria); moreover, older studies included mostly chronic, plurirefractory patients, while in the more recent studies patients with newly diagnosed/persistent ITP were also included. Finally, the ability to compare studies is often only partial because of different follow-up periods.

The data from placebo-controlled studies seem discouraging, but some considerations should be made: in the experimental treatment and placebo arms, patients were allowed to continue corticosteroid therapy, which could have biased the results: in the placebo arms the ORR were 67% (39% CR)48 and 73% (46% CR).47 Probably a more meaningful observation is that the median time to relapse in patients who achieved an overall response was 36 weeks in the rituximab group and 7 weeks in the placebo group.48 The main criticism that only 20% to 30% of patients achieve long-term remission for more than 3-5 years deserves further consideration. The main “error” was probably the mistaken belief that rituximab could represent the medical substitute of splenectomy: a single treatment administered once in a lifetime that could definitively cure many or even most patients with ITP. In some groups of patients, it is worth considering that even a sustained response of 12-18 months can have a significant, positive effect on a patient’s quality of life: during this period of time, they do not have to take any ITP medication and can avoid frequent hospital checks.

Furthermore, several clinical studies found that in young women a course of rituximab administered before the chronic phase (ITP duration <12 months) can lead to response rates and at least mid-term remission rates comparable to those obtained with splenectomy. This finding suggests that at least in some selected patients the treatment outcome may be much better.

Retreatment with rituximab is effective in most cases, especially in patients who maintained the preceding remission for more than 12 months.

References 1. Rodeghiero F, Stasi R, Gernsheimer T, et al.

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

Blood. 2009;113(11):2386–2393.

2. Audia S, Mahévas M, Samson M, Godeau B, et al. Pathogenesis of immune thrombocytopenia. Autoimmunity Rev. 2017;16(6): 620–632.

3. Cooper N, Bussel J. The pathogenesis of immune thrombocytopaenic purpura. Br J Haematol. 2006;133(4):364–374.

4. Harrington WJ, Minnich V, Hollingsworth JW, et al. Demonstration of a thrombocytopenic factor in the blood of patients with thrombocytopenic purpura. J Lab Clin Med. 1951;38(1):1-10.

5. McMillan R, Tani P, Millard F, et al. Plateletassociated and plasma anti-glycoprotein autoantibodies in chronic ITP. Blood.

1987;70(4):1040-1045.

6. McMillan R, Wang L, Tomer A, et al.

Suppression of in vitro megakaryocyte production by antiplatelet autoantibodies from

1132

7.

8.

9.

10.

11.

12.

Biomarkers predictive of response, including the controversial role of anti-platelet antibody testing, are still lacking and further studies are needed for their establishment and subsequent application in clinical practice.

The efficacy and safety of combinations of treatment with rituximab and other agents, such as thromboietinreceptor agonists, immunosuppressive agents, agents targeting plasma cells or yet others, need to be better evaluated and proven in comparison studies with rituximab monotherapy.

As far as concerns different rituximab doses, comparisons between the “standard dose” and “rheumatoid arthritis dose” did not reveal different results in terms of efficacy and toxicity. Low-dose rituximab could be equally effective, although prospective studies comparing the standard dose and low doses are lacking.

Rituximab is a well-tolerated drug; serious side effects are extremely rare and life-threatening infectious complications are usually only seen in patients with other concomitant causes of immunodeficiency.

Conclusions In conclusion, based on what has been published in the last 20 years, it is still difficult to give clear indications on when, to whom and how rituximab should be administered. In 2019, the choice of rituximab over other treatment options has to be weighed considering the individual patient’s features and expectations, the disease’s characteristics, the availability of the drug and the single center’s experience.

The authors of this review think that rituximab still represents a valuable therapeutic option for patients with ITP and, based on current knowledge, believe that it should be considered especially (although not exclusively) at an early stage of the disease, as second- or third-line therapy, in young patients (particularly young women) and in patients treated with curative purposes.

adult patients with chronic ITP. Blood.

2004;103(4):1364-1369.

Press OW, Appelbaum F, Ledbetter JA, et al.

Monoclonal antibody 1F5 (anti-CD20) serotherapy of human B cell lymphomas.

Blood. 1987;69(2):584-591.

Roll P, Palanichamy A, Kneitz C, et al.

Regeneration of B cell subsets after transient B cell depletion using anti-CD20 antibodies in rheumatoid arthritis. Arthritis Rheum.

2006;54(8):2377-2386.

Leandro MJ, Edwards JC, Cambridge G, et al. An open study of B lymphocyte depletion in systemic lupus erythematosus.

Arthritis Rheum. 2002;46(10):2673-2677.

Leandro MJ, Edwards JC, Cambridge G, Clinical outcome in 22 patients with rheumatoid arthritis treated with B lymphocyte depletion. Ann Rheum Dis. 2002;61(10):883-888.

Zecca M, Nobili B, Ramenghi U, et al.

Rituximab for the treatment of refractory autoimmune hemolytic anemia in children.

Blood. 2003;101(10):3857-3861.

Zaja F, De Vita S, Mazzaro C, et al. Efficacy and safety of rituximab in type II mixed cryoglobulinemia. Blood. 2003;101(10):

3827-3834.

13. Zaja F, Russo D, Fuga G, et al. Rituximab for myasthenia gravis developing after bone marrow transplant. Neurology. 2000;55(7): 1062-1063.

14. Monson NL, Cravens PD, Frohman EM, et al. Effect of rituximab on the peripheral blood and cerebrospinal fluid B cells in patients with primary progressive multiple sclerosis. Arch Neurol. 2005;62(2):258-264.

15. Scully M, Cohen H, Cavenagh J, et al.

Remission in acute refractory and relapsing thrombotic thrombocytopenic purpura following rituximab is associated with a reduction in IgG antibodies to ADAMTS-13. Br J Haematol. 2007;136(3):451-461.

16. Somer BG, Tsai DE, Downs L, et al.

Improvement in Sjögren’s syndrome following therapy with rituximab for marginal zone lymphoma. Arthritis Rheum. 2003;49 (3):394-398.

17. Salopek TG, Logsetty S, Tredget EE. AntiCD20 chimeric monoclonal antibody (rituximab) for the treatment of recalcitrant, lifethreatening pemphigus vulgaris with implications in the pathogenesis of the disorder. J Am Acad Dermatol. 2002;47(5):785-788.

haematologica | 2019; 104(6)

The role of rituximab in ITP 18. Beer JH, Rabaglio M, Berchtold P, et al.

Autoantibodies against the platelet glycoproteins (GP) IIb/IIIa, Ia/IIa, and IV and partial deficiency in GPIV in a patient with a bleeding disorder and a defective platelet collagen interaction. Blood. 1993;82(3):820828.

19. Kuwana M, Okazaki Y, Kaburaki J, et al.

Spleen is a primary site for activation of platelet-reactive T and B cells in patients with immune thrombocytopenic purpura. J Immunol. 2002;168(7):3675-3682.

20. Kayser W, Mueller-Eckhardt C, Bhakdi S, et al. Platelet-associated complement C3 in thrombocytopenic states. Br J Haematol.

1983;54(3):353-363.

21. Peerschke EI, Andemariam B, Yin W, Bussel JB. Complement activation on platelets correlates with a decrease in circulating immature platelets in patients with immune thrombocytopenic purpura. Br J Haematol.

2010;148(4):638-645.

22. Zhu X, Zhang J, Wang Q, et al. Diminished expression of 2-GPI is associated with a reduced ability to mitigate complement activation in anti-GPIIb/IIIa-mediated immune thrombocytopenia. Ann Hematol. 2018;97 (4):641-654.

23. Unterberger U, Eichelberger B, Ulz A, Panzer S. Antibodies against complementregulatory proteins on platelets in immune thrombocytopenia. Platelets. 2017;28(4): 409-413.

24. Chang M, Nakagawa PA, Williams SA, et al.

Immune thrombocytopenic purpura (ITP) plasma and purified ITP monoclonal autoantibodies inhibit megakaryocytopoiesis in vitro. Blood. 2003;102(3):887-895.

25. Grozovsky R, Begonja AJ, Liu K, et al. The Ashwell-Morell receptor regulates hepatic thrombopoietin production via JAK2-STAT3 signaling. Nat Med. 2015;21(1):47-54.

26. Jansen AJ, Peng J, Zhao HG, Hou M, Ni H. Sialidase inhibition to increase platelet counts: A new treatment option for thrombocytopenia.

Am J Hematol.

2015;90(5):E94-95.

27. Semple JW, Milev Y, Cosgrave D, et al.

Differences in serum cytokine levels in acute and chronic autoimmune thrombocytopenic purpura: relationship to platelet phenotype and antiplatelet T-cell reactivity. Blood.

1996;87(19):4245-4254.

28. Panitsas FP, Theodoropoulou M, Kouraklis A, et al. Adult chronic idiopathic thrombocytopenic purpura (ITP) is the manifestation of a type-1 polarized immune response.

Blood. 2004;103(7):2645-2647.

29. Liu B, Zhao H, Poon MC, et al. Abnormality of CD4(+)CD25(+) regulatory T cells in idiopathic thrombocytopenic purpura. Eur J Haematol. 2007;78(2):139-143.

30. Yu J, Heck S, Patel V, et al. Defective circulating CD25 regulatory T cells in patients with chronic immune thrombocytopenic purpura. Blood. 2008;112(4):1325-1328.

31. Olsson B, Andersson PO, Jernas M, et al. Tcell-mediated cytotoxicity toward platelets in chronic idiopathic thrombocytopenic purpura. Nat Med. 2003;9(9):1123-1124.

32. Kosugi S, Kurata Y, Tomiyama Y, et al.

Circulating thrombopoietin level in chronic immune thrombocytopenic purpura. Br J Haematol. 1996;93(3):704-706.

33. Einfeld DA, Brown JP, Valentine MA, Clark EA, Ledbetter JA. Molecular cloning of the human B cell CD20 receptor predicts a hydrophobic protein with multiple transmembrane domains. EMBO J. 1988;7(3): 711-717.

34. Nadler LM, Ritz J, Hardy R, et al. A unique cell surface antigen identifying lymphoid

haematologica | 2019; 104(6)

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

malignancies of B cell origin. J Clin Invest.

1981;67(1):134-140.

Weiner G. J. Rituximab: mechanism of action. Semin Hematol. 2010;47(2):115-123.

Lee EJ, Kueck B. Rituxan in the treatment of cold agglutinin disease. Blood. 1998;92(9): 3490-3491.

Shvidel L, Klepfish A, Berrebi A. Successful treatment with rituximab for relapsing immune thrombocytopenic purpura (ITP) associated with low-grade non-Hodgkin’s lymphoma. Am J Hematol. 2001;67(3):213214.

Stasi R, Pagano A, Stipa E, Amadori S. Rituximab chimeric anti-CD20 monoclonal antibody treatment for adults with chronic idiopathic thrombocytopenic purpura.

Blood. 2001;98(4):952-957.

Godeau B, Porcher R, Fain O, et al.

Rituximab efficacy and safety in adult splenectomy candidates with chronic immune thrombocytopenic purpura: results of a prospective multicenter phase 2 study.

Blood. 2008;112(4):999-1004.

Patel VL, Mahévas M, Lee SY, et al.

Outcomes 5 years after response to rituximab therapy in children and adults with immune thrombocytopenia. Blood. 2012; 119(25):5989-5995.

Khellaf M, Charles-Nelson A, Fain O, et al.

Safety and efficacy of rituximab in adult immune thrombocytopenia: results from a prospective registry including 248 patients.

Blood. 2014;124(22):3228-3236.

Arnold DM, Dentali F, Crowther MA, et al.

Systematic review: efficacy and safety of rituximab for adults with idiopathic thrombocytopenic purpura. Ann Intern Med. 2007;146(1):25-33.

Cervinek L, Cerna O, Caniga M, et al.

Efficacy of rituximab in primary immune thrombocytopenia: an analysis of adult pretreated patients from everyday hematological practice. Int J Hematol. 2012;96(5):594599.

Zaja F, Volpetti S, Chiozzotto M, et al. Longterm follow-up analysis after rituximab salvage therapy in adult patients with immune thrombocytopenia. Am J Hematol. 2012;87 (9):886-889.

Mahévas M, Ebbo M, Audia S, et al. Efficacy and safety of rituximab given at 1,000 mg on days 1 and 15 compared to the standard regimen to treat adult immune thrombocytopenia. Am J Hematol. 2013;88(10):858-861.

Cooper N, Stasi R, Cunningham-Rundles S, et al. The efficacy and safety of B-cell depletion with anti-CD20 monoclonal antibody in adults with chronic immune thrombocytopenic purpura. Br J Haematol. 2004;125 (2):232-239.

Arnold DM, Heddle NM, Carruthers J, et al.

A pilot randomized trial of adjuvant rituximab or placebo for nonsplenectomized patients with immune thrombocytopenia.

Blood. 2012;119(6):1356-1362.

Ghanima W, Khelif A, Waage A, et al.

Rituximab as second-line treatment for adult immune thrombocytopenia (the RITP trial): a multicentre, randomised, double-blind, placebo-controlled trial. Lancet. 2015;385 (9978):1653-1661.

Maloney DG, Grillo-Lopez AJ, White CA, et al. IDEC-C2B8 (Rituximab) anti-CD20 monoclonal antibody therapy in patients with relapsed low-grade non-Hodgkin’s lymphoma. Blood. 1997;90(6):2188-2195.

Zaja F, Vianelli N, Volpetti S, et al. Low-dose rituximab in adult patients with primary immune thrombocytopenia.

Eur J Haematol. 2010;85(4):329-334.

Gracie C, Zaidi A, Doobaree U, et al.

52.

53.

54.

55.

56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

Comparsion of standard and low-dose rituximab in primary immune thrombocytopenia (ITP): data from the UK ITP registry. EHA Abstr. 2018.

Tran H, Brighton T, Grigg A, et al. A multicentre, single-arm, open-label study evaluating the safety and efficacy of fixed dose rituximab in patients with refractory, relapsed or chronic idiopathic thrombocytopenic purpura (R-ITP1000 study). Br J Haematol.

2014;167(2):243-251.

Zwaginga JJ, Van Der Holt B, Te Boekhorst PA, et al. Multi-center randomized open label phase II trial on three rituximab dosing schemes in immune thrombocytopenia patients. Haematologica. 2015;100(3):e9092.

Bussel JB, Lee CS, Seery C, et al. Rituximab and three dexamethasone cycles provide responses similar to splenectomy in women and those with immune thrombocytopenia of less than two years duration.

Haematologica. 2014;99(7):1264-1271.

Chapin J, Lee CS, Zhang H, et al. Gender and duration of disease differentiate responses to rituximab–dexamethasone therapy in adults with immune thrombocytopenia. Am J Hematol. (2016);91(9):907911.

Marangon M, Vianelli N, Palandri F, et al.

Rituximab in immune thrombocytopenia: gender, age, and response as predictors of long-term response. Eur J Haematol.

2017;98(4):371-377.

Peñalver FJ, Jimenez-Yuste V, Almagro M, et al. Rituximab in the management of chronic immune thrombocytopenic purpura: an effective and safe therapeutic alternative in refractory patients.

Ann Hematol.

2006;85(6):400-406.

Zaja F, Vianelli N, Battista M, et al. Earlier administration of rituximab allows higher rate of long-lasting response in adult patients with autoimmune thrombocytopenia. Exp Hematol. 2006;34(5):571-572.

Provan D, Stasi R, Newland AC, et al.

International consensus report on the investigation and management of primary immune thrombocytopenia. Blood. 2010; 115(2):168-186.

Neunert C, Lim W, Crowther M, et al. The American Society of Hematology 2011 evidence-based practice guideline for immune thrombocytopenia. Blood. 2011;117(16): 4190-4207.

Cooper N, Stasi R, Cunningham-Rundles S, et al. Platelet-associated antibodies, cellular immunity and FCGR3a genotype influence the response to rituximab in immune thrombocytopenia. Br J Haematol. 2012;158 (4):539-547.

Porcelijn L, Huiskes E, Schipperus M, et al.

Lack of detectable platelet autoantibodies is correlated with nonresponsiveness to rituximab treatment in ITP patients. Blood.

2017;129(25):3389-3391.

Arnold DM, Vrbensky JR, Karim N, et al.

The effect of rituximab on anti-platelet autoantibody levels in patients with immune thrombocytopenia. Br J Haematol.

2017;178(2):302-307.

Peng J, Ma SH, Liu J, et al. Association of autoantibody specificity and response to intravenous immunoglobulin G therapy in immune thrombocytopenia: a multicenter cohort study. J Thromb Haemost. 2014;12 (4):497-504.

Zeng Q, Zhu L, Tao L, et al. Relative efficacy of steroid therapy in immune thrombocytopenia mediated by anti-platelet GPIIbIIIa versus GPIb antibodies. Am J Hematol.

2012;87(2):206-208.

1133

E. Lucchini et al.

66. Feng R, Liu X, Zhao Y, et al. GPIIb/IIIa autoantibody predicts better rituximab response in ITP. Br J Haematol. 2018;182 (2):305-307.

67. Nieswandt B, Bergmeier W, Rackebrandt K, Gessner JE, Zirngibl H. Identification of critical antigen-specific mechanisms in the development of immune thrombocytopenic purpura in mice. Blood. 2000;96(7):25202527.

68. Medeot M, Zaja F, Vianelli N, et al.

Rituximab therapy in adult patients with relapsed or refractory immune thrombocytopenic purpura : long-term follow-up results. Eur J Haematol. 2008;81(3):165-169.

69. Zaja F, Baccarani M, Mazza P, et al.

Dexamethasone plus rituximab yields higher sustained response rates than dexamethasone monotherapy in adults with primary immune thrombocytopenia. Blood. 2010; 115 (14):2755-2762.

70. Gudbrandsdottir S, Birgens HS, Frederiksen H, et al. Rituximab and dexamethasone vs dexamethasone monotherapy in newly diagnosed patients with primary immune thrombocytopenia. Blood. 2013;121(11): 1976-1981.

71. Choi PY, Roncolato F, Badoux X, et al. A novel triple therapy for ITP using high-dose dexamethasone, low-dose rituximab, and cyclosporine (TT4). Blood. 2015;126(4):500503.

72. Li Y, Wang YY, Fei HR, Wang L, Yuan CL. Efficacy of low-dose rituximab in combination with recombinant human thrombopoietin in treating ITP. Eur Rev Med Pharmacol Sci. 2015;19(9):1583-1588.

73. Zhou H, Xu M, Qin P, et al. A multicenter randomized open-label study of rituximab plus rhTPO vs rituximab in corticosteroidresistant or relapsed ITP. Blood.

2015;125(10):1541-1547.

74. Gómez-Almaguer D, Colugna-Pedraza PR, Gómez-De León A, et al. Eltrombopag, lowdose rituximab, and dexamethasone combination as frontline treatment of newly diagnosed immune thrombocytopaenia. Br J Haematol. 2019;184(2):288-290.

75. Hasan A, Michel M, Patel V, et al. Repeated courses of rituximab in chronic ITP: three different regimens. Am J Hematol. 2009;84 (10):661-665.

76. Giagounidis AA, Anhuf J, Schneider P, et al.

Treatment of relapsed idiopathic thrombocytopenic purpura with the anti-CD20 monoclonal antibody rituximab: a pilot study.

Eur J Haematol. 2002;69(2):95-100.

77. Reboursiere E, Fouques H, Maigne G, et al.

Rituximab salvage therapy in adults with immune thrombocytopenia: retrospective study on efficacy and safety profiles. Int J Hematol. (2016);104(1):85-91.

78. Garcia-Chavez J, Majluf-Cruz A, MontielCervantes L, Esparza MG, Vela-Ojeda J, Rituximab therapy for chonic and refractory immune thrombocytopenic purpura: a longterm follow-up analysis. Ann Hematol.

2007;86(12):871-877.

79. Rai MP, Lee E, Bussel JB. Rituximab (Ritux) Maintenance infusions in immune thrombocytopenia ( ITP ) to prolong remission following relapse after initial ritux induction but responded to a second course. Blood.

2018;132(Suppl 1):3753.

80. Leandro MJ, Cambridge G, Ehrenstein MR, Edwards JC. Reconstitution of peripheral blood B cells after depletion with rituximab in patients with rheumatoid arthritis.

Arthritis Rheum. 2006;54(2):613-620.

81. Audia S, Samson M, Janikashvili N, et al.

Immunologic effects of rituximab on the human spleen in immune thrombocytope-

1134

nia. Blood. 2011;118(16):4394-43400.

82. Stasi R, Del Poeta G, Stipa E, et al. Response to B-cell-depleting therapy with rituximab reverts the abnormalities of T-cell subsets in patients with idiopathic thrombocytopenic purpura. Blood. 2007;110(8):2924-2930.

83. Stasi R, Cooper N, Del Poeta G, et al.

Analysis of regulatory T-cell changes in patients with idiopathic thrombocytopenic purpura receiving B cell depleting therapy with rituximab. Blood. 2008;112(4):11471150.

84. Gudbrandsdottir S, Brimnes M, Køllgaard T, Hasselbalch HC, Nielsen CH. Effects of rituximab and dexamethasone on regulatory and proinflammatory B-cell subsets in patients with primary immune thrombocytopenia. Eur J Haematol. 2018;100(1):45-52.

85. Audia S, Samson M, Mahévas M, et al.

Preferential splenic CD8+T-cell activation in rituximab-nonresponder patients with immune thrombocytopenia.

Blood.

2013;122(14):2477-2486.

86. Guo L, Kapur R, Aslam R, et al. CD20+B-cell depletion therapy suppresses murine CD8+T-cell-mediated immune thrombocytopenia. Blood. 2016;127(6):735-738.

87. Bao W, Bussel JB, Heck S, et al. Improved regulatory T-cell activity in patients with chronic immune thrombocytopenia treated with thrombopoietic agents. Blood.

2010;116(22):4639-4645.

88. Guo C, Chu X, Shi Y, et al. Correction of Th1-dominant cytokine profiles by highdose dexamethasone in patients with chronic idiopathic thrombocytopenic purpura. J Clin Immunol. 2007;27(6):557-562.

89. Li J, Wang Z, Hu S, Zhao X, Cao L. Correction of abnormal T cell subsets by high-dose dexamethasone in patients with chronic idiopathic thrombocytopenic purpura. Immunol Lett. 2013;154(1-2):42-48.

90. Schifferli A, Kühne T. Thrombopoietin receptor agonists: a new immune modulatory strategy in immune thrombocytopenia?

Semin Hematol. 2016;53 (Suppl 1):S31-34.

91. Mahévas M, Patin P, Huetz F, et al. B cell depletion in immune thrombocytopenia reveals splenic long-lived plasma cells. J Clin Invest. 2013;123(1):432-442.

92. Karmacharya P, Poudel DR, Pathak R, et al.

Rituximab-induced serum sickness: a systematic review. Semin Arthritis Rheum.

2015;45(3):334-340.

93. Kumar A, Khamkar K, Gopal H. Serum sickness and severe angioedema following rituximab therapy in RA. Int J Rheum Dis. 2012;15(1):e6-7.

94. Edwards JC, Cambridge G, Leandro MJ. B cell depletion therapy in rheumatic disease.

Best Pract Res Clin Rheumatol. 2006;20 (5):915-928.

95. Diwakar L, Gorrie S, Richter A, et al. Does rituximab aggravate pre-existing hypogammaglobulinaemia? J Clin Pathol. 2010;63(3): 275-277.

96. Levy R, Mahévas M, Galicier L, et al.

Profound symptomatic hypogammaglobulinemia: a rare late complication after rituximab treatment for immune thrombocytopenia. Report of 3 cases and systematic review of the literature. Autoimmun Rev. 2014;13(10):1055-1063.

97. Michel M, Chanet V, Galicier L, et al.

Autoimmune thrombocytopenic purpura and common variable immunodeficiency: analysis of 21 cases and review of the literature. Medicine (Baltimore) (2004);83(4):254263.

98. Cooper N, Arnold D. M. The effect of rituximab on humoral and cell mediated immunity and infection in the treatment of

autoimmune diseases. Br J Haematol.

2010;149(1):3-13.

99. Kado R, Sanders G, McCune WJ. Diagnostic and therapeutic considerations in patients with hypogammaglobulinemia after rituximab therapy. Curr Opin Rheumatol.

2017;29(3):228-233.

100. Gobert D, Bussel JB, Cunningham-Rundles C, et al. Efficacy and safety of rituximab in common variable immunodeficiency-associated immune cytopenias: a retrospective multicentre study on 33 patients. Br J Haematol. 2011;155(4):498-508.

101. Calabrese LH, Molloy ES, Huang D, Ransohoff RM. Progressive multifocal leukoencephalopathy in rheumatic diseases: evolving clinical and pathologic patterns of disease. Arthritis Rheum. 2007;56(7):21162128.

102. Chugh S, Darvish-Kazem S, Lim W, et al.

Rituximab plus standard of care for treatment of primary immune thrombocytopenia: a systematic review and meta-analysis.

Lancet Haematol. 2015;2(2):e75-81.

103. Nazi I, Kelton JG, Larché M, et al. The effect of rituximab on vaccine responses in patients with immune thrombocytopenia.

Blood. 2013;122(11):1946-1953.

104. Cho CH, Hwang WL, Cheng SB, Lee TY, Teng CL. Hepatitis B reactivation induced by rituximab maintenance therapy for lymphoma. Ann Hematol. 2011;90(1):111-112.

105. Sabath BF, Major EO. Traffic of JC virus from sites of initial infection to the brain: the path to progressive multifocal leukoencephalopathy. J Infect Dis. 2002;186 (Suppl 2):S180-186.

106. Bohra C, Sokol L, Dalia S. Progressive multifocal leukoencephalopathy and monoclonal antibodies: a review. Cancer Control.

2017;24(4):1073274817729901 107. Henegar CE, Eudy AM, Kharat V, Hill DD, Bennett D, Haight B. Progressive multifocal leukoencephalopathy in patients with systemic lupus erythematosus: a systematic literature review. Lupus. 2016;25(6):617-626.

108. Carson KR, Evens AM, Richey EA, et al.

Progressive multifocal leukoencephalopathy after rituximab therapy in HIV-negative patients: a report of 57 cases from the Research on Adverse Drug Events and Reports project. Blood. 2009;113(20):48344840.

109. Berger JR, Malik V, Lacey S, Brunetta P, Lehane PB. Progressive multifocal leukoencephalopathy in rituximab-treated rheumatic diseases: a rare event. J Neurovirol.

2018;24(3):323-331.

110. Grant C, Wilson WH, Dunleavy K. Neutropenia associated with rituximab therapy. Curr Opin Hematol. 2011;18(1):49-54.

111. Monaco WE, Jones JD, Rigby WF. Rituximab associated late-onset neutropenia-a rheumatology case series and review of the literature. Clin Rheumatol. 2016;35(10):24572462.

112. Knight A, Sundstrom Y, Borjesson O, et al.

Late-onset neutropenia after rituximab in ANCA-associated vasculitis. Scand J Rheumatol. 2016;45(5):404-407.

113. Moore DC. Drug-Induced Neutropenia: A focus on rituximab-induced late-onset neutropenia. P T. 2016;41(12):765-768.

114. Van Vollenhoven RF, Fleischmann RM, Furst DE, Lacey S, Lehane PB. Longterm safety of rituximab: final report of the rheumatoid arthritis global clinical trial program over 11 years. J Rheumatol. 2015;42(10):1761-1766.

115. Fleury I, Chevret S, Pfreundschuh M, et al.

Rituximab and risk of second primary malignancies in patients with non-Hodgkin lymphoma: a systematic review and meta-

haematologica | 2019; 104(6)

The role of rituximab in ITP analysis. Ann Oncol. 2016;27(3):390-397.

116. Van Daalen EE, Rizzo R, Kronbichler A, et al. Effect of rituximab on malignancy risk in patients with ANCA-associated vasculitis.

Ann Rheum Dis. 2017;76(6):1064-1069.

117. Wang J, Wiley JM, Luddy R, Greenberg J, Feuerstein MA, Bussel JB. Chronic immune thrombocytopenic purpura in children: assessment of rituximab treatment. J Pediatr.

2005;146(2):217-221.

118. Bennett CM, Rogers ZR, Kinnamon DD, et al. Prospective phase 1 / 2 study of rituximab in childhood and adolescent chronic immune thrombocytopenic purpura. Blood.

2006;107(7):2639-2642.

119. Mueller BU, Bennett CM, Feldman HA, et al.

One year follow-up of children and adoles-

haematologica | 2019; 104(6)

cents with chronic immune thrombocytopenic purpura (ITP) treated with rituximab. Pediatr Blood Cancer. 2009;52(2):259262.

120. Liang Y, Zhang L, Gao J, Hu D, Ai Y. Rituximab for children with immune thrombocytopenia : a systematic review. PLoS One. 2012;7(5):e36698.

121. Taube T, Schmid H, Reinhard H, von Stackelberg A, Overberg US. Effect of a single dose of rituximab in chronic immune thrombocytopenic purpura in childhood.

Haematologica. 2005;90(2):281–283.

122. Oved JH, Lee CSY, Bussel JB. Treatment of children with persistent and chronic idiopathic thrombocytopenic purpura: 4 infusions of rituximab and three 4-day cycles of dexam-

ethasone. J Pediatr. 2017;191:225–231.

123. Parodi E, Rivetti E, Amendola G, et al. Longterm follow-up analysis after rituximab therapy in children with refractory symptomatic ITP : identification of factors predictive of a sustained response. Br J Haematol. 2009;144 (4):552–558.

124. Grace RF, Bennett CM, Ritchey AK, et al.

Response to steroids predicts response to rituximab in pediatric chronic immune thrombocytopenia. Pediatr Blood Cancer.

2012;58(2):221-225.

125. Matsubara K, Takahashi Y, Hayakawa A, et al. Long-term follow-up of children with refractory immune thrombocytopenia treated with rituximab. Int J Hematol. 2014;99 (4):429-436.

1135

ARTICLE Ferrata Storti Foundation

Hematopoiesis

CD150high CD4 T cells and CD150high regulatory T cells regulate hematopoietic stem cell quiescence via CD73 Yuichi Hirata,1,2,3,4 Miwako Kakiuchi,1,2,3 Simon C Robson5 and Joji Fujisaki1,2,6*

Columbia Center for Translational Immunology, Columbia University College of Physicians and Surgeons, New York, NY, USA; 2Columbia Stem Cell Initiative, Columbia University College of Physicians and Surgeons, New York, NY, USA; 3Department of Medicine, Columbia University College of Physicians and Surgeons, New York, NY, USA; 4MSD K.K., Tokyo, Japan; 5 Department of Medicine, Liver Center and Transplantation Institute, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA and 6Department of Pediatrics, Division of Hematology and Oncology, Columbia University College of Physicians and Surgeons, New York, NY, USA 1

Haematologica 2019 Volume 104(6):1136-1142

ABSTRACT

V

Correspondence: JOJI FUJISAKI jf2819@cumc.columbia.edu Received: May 23, 2018.

Accepted: December 10, 2018.

Pre-published: December 13, 2018. doi:10.3324/haematol.2018.198283 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/104/6/1136

arious extrinsic signals tightly control hematopoietic stem cell quiescence. Our recent study showed that hematopoietic stem cells are regulated by a special FoxP3+ regulatory T-cell population with high expression of a hematopoietic stem cell marker, CD150. Extracellular adenosine generated via a cell-surface ectoenzyme CD39 on CD150high regulatory T cells maintained hematopoietic stem cell quiescence. It remains unclear how conventional T cells and the other cell-surface ectoenzyme, CD73, contribute to regulation of hematopoietic stem cells. This work shows that CD150high regulatory T cells as well as unique CD150high CD4+ conventional T cells regulate hematopoietic stem cells via CD73. Global CD73 deletion increased the numbers of hematopoietic stem cells, cycling stem cell frequencies, and levels of reactive oxygen species in hematopoietic stem cells. In vivo antioxidant treatment inhibited the increase of hematopoietic stem cells in CD73 knockout mice, suggesting that CD73 maintains stem cell quiescence by preventing oxidative stress. High levels of CD73 expression were frequently found on CD150high regulatory T cells and CD150high FoxP3- CD4+ T cells within the bone marrow. Transfer of these CD150high regulatory T cells and CD150high CD4+ conventional T cells abolished the increase of hematopoietic stem cells in CD73 knockout mice.

In addition, the increase of stem cells in CD73 knockout mice was also inhibited by pharmacological activation of adenosine receptor 2A which is highly expressed by hematopoietic stem cells. Taken together, these results suggest that CD73 of CD150high regulatory T cells and CD150high CD4+ conventional T cells protects hematopoietic stem cells from oxidative stress, maintaining stem cell quiescence via adenosine receptor 2A.

Introduction Š2019 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

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

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

1136

The bone marrow (BM) microenvironment provides various cues to regulate hematopoietic stem cell (HSC) quiescence, self-renewal, and multilineage differentiation,1-4 and to protect HSC from various stresses, such as oxidative stress5 and toxic substances.6 Different mesenchymal subsets and megakaryocytes form a specialized regulatory zone for HSC residence, called the niche, within the BM.1-4 It is thought that tight control of HSC quiescence and function helps to prevent HSC exhaustion and genetic mutation. Due to a growing demand for clinical BM transplantation, understanding how the BM microenvironment regulates HSC remains important.

Our recent study demonstrated that HSC were regulated by a unique population of regulatory T cells (Treg) with high expression of a HSC marker, CD150.7 These CD150high Treg frequently localized adjacent to HSC.7 Treg-mediated HSC regulation depended on a cell-surface ectoenzyme, CD39, which was highly haematologica | 2019; 104(6)

CD73 regulates HSC quiescence

expressed by CD150high Treg.7 CD39 is known to convert extracellular adenosine triphosphate (ATP) and adenosine diphosphate (ADP) into adenosine monophosphate (AMP) which is further hydrolyzed by the other cell-surface ectoenzyme, CD73, into extracellular adenosine, a purine nucleotide with various tissue-protective effects.8 The results of our study using conditional knockout (KO) of CD39 in Treg suggested that extracellular adenosine generated via CD39 on Treg protected HSC from oxidative stress, maintaining HSC quiescence.7 It remains unclear how the other cell-surface ectoenzyme, CD73, contributes to HSC regulation and which BM cell populations regulate HSC via CD73. In addition, while the BM serves as a reservoir of memory T cells,9,10 little is known about the role of these conventional T cells in HSC regulation.

This work identified unique CD150highCD4+FoxP3- conventional T cells (nonTreg) which highly expressed CD73 and CD39, like CD150high Treg. Our observations in CD73 KO mice into which these CD150high T-cell populations were transferred suggest that CD73 of CD150high CD4+ nonTreg and CD150high Treg maintain HSC quiescence and abundance.

Flow cytometric analysis

Methods

Stromal cell analysis

Animals C57BL/6J mice, SJL mice, BALB/c mice, CD73 KO mice, FoxP3YFP mice, and Lep-cre mice (Jackson Laboratory, Bar Harbor, ME, USA) were housed in a specific pathogen-free environment.

CD39-flox mice were kindly provided by Dr. Simon C. Robson (Harvard Medical School). Seven-week old CD73 KO mice were analyzed. The mice were sacrificed by CO2 inhalation and cervical dislocation. Studies were conducted with approval from Institutional Review Boards and Animal Care and Use Committees at Columbia University.

BM cells were isolated by crushing tibiae and femora. Following treatment with a red blood cell lysis buffer (Biolegend), the cell suspension (2x106 cells) was plated onto 96-well plates and incubated with culture media containing 2 μM CellROX Deep Red (Invitrogen) for 30 min. Flow cytometry was performed using an LSRII (BD Biosciences), LSRFortessa (BD Biosciences), or FACSCanto (BD Biosciences) cytometer followed by analysis using FlowJo software (Tree Star Inc.).

Colony-forming assay BM cells (2.0x104 cells/each well) were plated in six-well plates (Corning, NY, USA) containing 1 mL MethoCultTM (M3234, Stemcell Technologies Inc.) supplemented with 1% penicillin/streptomycin (Gibco), stem cell factor (50 ng/mL), interleukin-3 (15 ng/mL), interleukin-6 (20 ng/mL), and granulocytemacrophage colony-stimulating factor (15 ng/mL). Colonies were maintained at 37°C in humidified incubators. Colony formation was scored on day 10.

Flow cytometry following intracellular staining of FoxP3 Intracellular FoxP3 staining was performed according to the manufacturer’s protocol (eBioscience).

For analysis of stromal cells, long bones were gently crushed using Hanks balanced saline solution to harvest the BM cells.

Whole bone marrow was digested with collagenase IV (200 U/mL) and DNase I (200 U/mL) at 37°C for 30 min. Following treatment with a red blood cell lysis buffer (Biolegend), the cell suspension (2x106 cells) was plated onto 96-well plates and then stained with antibodies. Anti-CD140a (APA5), anti-CD140b (APB5), anti-CD45 (30F-11), anti-CD31 (390) and anti-Ter119 antibodies (all from Biolegend) were used to stain perivascular stromal cells.

T-cell transfer assay Antibodies and reagents We used FITC-conjugated Lineage monoclonal antibodies (B220, Mac1, GR-1, CD2, CD3a, CD8a, CD4, CD19 and Ter119), APC-780-conjugated cKit monoclonal antibodies, PECy7- or APCconjugated CD39 monoclonal antibodies, FITC-conjugated FoxP3 monoclonal antibodies, PE-conjugated Ki67 monoclonal antibodies (all purchased from eBioscience), APC/Cy7-conjugated NK1.1 monoclonal antibodies, BV510-conjugated CD3 monoclonal antibodies, PE/Cy7-, or BV605-conjugated CD4 monoclonal antibodies, Alexa700- or Pacific blue-conjugated CD48 monoclonal antibodies, PerCP/Cy5.5-, APC-, or BV605-conjugated CD73 monoclonal antibodies, PE- or PE/Cy7-conjugated CD150 monoclonal antibodies (all from Biolegend), and BV605-conjugated Sca-1 monoclonal antibodies (from BD Pharmingen or Biolegend).

N-acetyl-L-cysteine (A9165) was purchased from SigmaAldrich.

Competitive reconstitution assay SJL (CD45.1) mice were irradiated at 475 cGy twice (950 cGy in total) at least 2 h apart. Two hours after the last irradiation, donor BM cells (CD45.2), together with competitor BM cells (SJL) (3 x 105/each), were injected into the tail veins of SJL recipients.

Peripheral blood samples were analyzed periodically. Red blood cells were lysed with an ammonium chloride potassium buffer.

The antibodies used to analyze donor chimerism were antiCD45.1, anti-CD45.2, anti-GR1, anti-CD11b, anti-B220, and antiTCR-b (all from Biolegend). haematologica | 2019; 104(6)

HSC numbers were determined in CD73 KO mice 7 days after intravenous injection of CD150high BM Treg, CD150low BM Treg, CD150high BM nonTreg, or CD150low BM nonTreg (30,000 cells/mouse). Data were pooled from three independent experiments (4-10 mice/group).

Adenosine 2A receptor agonist treatment CD73 KO mice were given PSB0777, a potent adenosine 2A receptor (A2AR) agonist, daily for 7 consecutive days (25 μg/mouse, intraperitoneally). Total HSC numbers in one tibia and one femur were analyzed 1 day after the final injection.

Statistics Statistical analyses were performed with GraphPad Prism software (version 6.0). Statistical significance was determined using a two-tailed t-test or one-way analysis of variance (ANOVA) with a Bonferroni post-test correction. P values less than 0.05 were considered to be statistically significant. All data are presented as mean ± standard deviation (SD).

Results CD73 deletion increased hematopoietic stem cell pool size The effect of global CD73 deletion on hematopoiesis was first analyzed. CD73 KO mice showed increases in 1137

Y. Hirata et al.

diated B6 SJL mice (CD45.1), together with competitor SJL BM (CD45.1; 3 x 105 cells/mouse). CD45.2 donor blood chimerism from CD73 KO BM cells remained significantly higher than that from control BM cells for 6 months after transplantation, suggesting that CD73 deletion increased functional HSC and HSPC frequencies (Figure 1D). Additionally, no myeloid skewing was observed in donor hematopoietic cells derived from CD73KO BM (Online Supplementary Figure S1E). Taken together, these results indicate that CD73 maintains quiescence and pool size of HSPC and HSC.

BM cellularity (Figure 1A). CD73 deletion significantly increased the frequencies of cycling cKit+Sca1+Linhematopoietic stem and progenitor cells (HSPC) and CD150+CD48-cKit+Sca1+Lin- HSC, as well as numbers of HSPC and HSC (Figure 1A-C, Online Supplementary Figure S1A,B and Online Supplementary Table S1). Consistently, the numbers of colonies formed following in vitro culture of BM cells isolated from CD73 KO mice were significantly higher than those from wild-type mice (Online Supplementary Figure S1C). The numbers of other BM cell populations were not significantly altered by CD73 deletion (Online Supplementary Figure S1D). To assess the size of the pool of functional HSC and HSPC, we performed competitive BM transplantation assays to evaluate the reconstituting potential of BM cells. BM cells of CD73 KO mice or control wildtype mice (B6 CD45.2; 3 x 105 cells/mouse) were intravenously injected into lethally-irra-

CD73 maintains hematopoietic stem cell pool size by preventing oxidative stress CD73 KO mice showed slight but significant increases in the levels of reactive oxygen species (ROS) in HSPC and HSC but not in Lin+ cells (Figure 2A, Online Supplementary

A

B

C

D

Figure 1. CD73 deletion increased hematopoietic stem cell pool size. (A) Bone marrow (BM) cellularity of CD73 knockout (KO) mice. We analyzed the numbers of total BM cells isolated from one tibia and one femur by crushing. Data from three independent experiments with nine mice/group were pooled. Data are presented as mean ± SD and analyzed by a two-tailed t-test. (B) Flow cytometric analysis of frequencies of Ki67- cells among hematopoietic stem and progenitor cells (HSPC: cKit+Sca1+Lin-) and hematopoietic stem cells (HSC: CD150+CD48-cKit+Sca1+Lin-) in CD73 KO mice. The results were reproducible in two independent experiments (7 mice/group total). A representative figure from one independent experiment is shown here. Data are presented as mean ± SD and analyzed by a two-tailed t-test.

(C) Flow cytometric analysis of HSPC and HSC frequencies and numbers in CD73 KO mice. The results were reproducible in three independent experiments (3 mice/group in each experiment; refer to Online Supplementary Table S1). A representative figure is shown here. Data are presented as mean ± SD and analyzed by a two-tailed t-test. (D) Donor chimerism in the peripheral blood of lethally irradiated mice that received wildtype SJL bone marrow (BM) cells (CD45.1) together with BM cells of CD73 KO or control mice (CD45.2). BM cells from each donor mouse (7 control mice and 9 CD73 KO mice) were transplanted into one recipient. The results were pooled from two independent experiments. Data are presented as mean ± SD and analyzed by a two-tailed t-test. wt: wildtype: M: months.

1138

haematologica | 2019; 104(6)

CD73 regulates HSC quiescence

Figure S2A), suggesting that CD73 prevents oxidative stress against HSC and HSPC but not against mature cells.

To test whether CD73 maintains HSC quiescence in a ROS-dependent manner, we used treatment with an antioxidant, N-acetylcysteine (NAC), which reversed the increases in HSC numbers and reconstituting potential of BM cells in CD73 KO mice (Figure 2B,C). These results indicate that CD73 maintains HSC quiescence by preventing oxidative stress.

This ROS-mediated expansion of HSC pool size in CD73 KO mice was consistently observed in our recently reported study, using two models: (i) mice with conditional deletion of CD39 in Treg; and (ii) mice with reduction of BM Treg achieved by CXCR4 deletion in Treg.7 In contrast, some previous studies showed that increased ROS levels in HSC led to loss of HSC quiescence, and HSC exhaustion.11-13 HSC proliferation in our models is likely explained by moderate increases in ROS levels (1.2- to 1.5fold) compared to greater increases in other models (3- to 5-fold).11-13 Indeed, some studies showed that a moderate increase in ROS levels induced HSC proliferation.14,15 Moreover, the peripheral blood of CD73 KO mice showed a non-significant trend toward myeloid skewing (Online Supplementary Figure S2B), which may also reflect a moderate increase in ROS levels in HSC.

A

CD73high cells were frequently found in CD150high regulatory T cells and CD150highCD4+ non-regulatory T cells To identify cell populations which play important roles in CD73-mediated HSC regulation, flow cytometric analysis was performed to measure CD73 and CD39 expression levels on hematopoietic cell populations within the BM. Intermediate to high expression of CD39 was found on the following hematopoietic cell populations: HSC; HSPC; CD11b+Gr1int cells; CD11b+Gr1high cells; B220+ B cells; CD4+FoxP3- cells (CD4+ nonTreg); CD4+FoxP3+ Treg; CD8+ T cells; CD4+CD3+NK1.1+ NKT cells; and NK cells (Online Supplementary Figure S3A). In contrast, high levels of CD73 expression were mainly observed within CD4+ T cells (Treg, CD4+ nonTreg) (Figure 3A,B). While CD8+ T cells, CD11b+Gr1high cells, and CD4+ NKT cells showed intermediate levels of CD73 expression, CD73 was not expressed by HSC, B cells, or CD11b+Gr1int myeloid cells (Figure 3A,B).

As previously reported,7 CD73high cells among Treg were predominantly CD150high, showing equivalent levels of expression of CD150 as those of HSC (Figure 3B-D).

These CD150high Treg also highly expressed CD39 (Figure 3D, Online Supplementary Figure S3A). Notably, there were also CD150high fractions among BM CD4+ nonTreg, which

C

B

Figure 2. CD73 maintains hematopoietic stem cell quiescence and pool size by preventing oxidative stress. (A) Flow cytometric analysis of reactive oxygen species (ROS) levels in hematopoietic stem and progenitor cells (HSPC: cKit+Sca1+Lin-) and hematopoietic stem cells (HSC: CD150+CD48-cKit+Sca1+Lin-) in CD73 knockout (KO) or control B6 mice. The results were reproducible in two independent experiments (7 mice/group total). A representative figure from one independent experiment is shown here. Data are presented as mean ± SD and were analyzed by a two-tailed t-test. (B) Flow cytometric analysis of HSPC and HSC numbers in CD73 KO mice treated with N-acetylcysteine (NAC). CD73 KO or B6 control mice were given NAC (100 mg/kg) or mock treatment (daily; subcutaneously) for 4 weeks prior to bone marrow (BM) isolation. The results were reproducible in two independent experiments (7 mice/group total). Data are presented as mean ± SD and were analyzed by one-way analysis of variance (ANOVA) with a Bonferroni post-test. (C) Donor chimerism in the peripheral blood of lethally irradiated SJL mice that were administered wildtype SJL BM cells (CD45.1) together with BM cells of NAC-treated CD73 KO or control B6 mice (CD45.2) 5 months after transplantation. CD73 KO or control mice received NAC (100 mg/kg) or mock treatment (daily; subcutaneously) for 4 weeks prior to isolation. BM cells from each donor mouse were transplanted into one or two recipients. Experiments included three donors and 11-12 recipients/group. Data are presented as mean ± SD and were analyzed by one-way ANOVA with a Bonferroni post-test. MFI: mean fluorescence intensity; wt: wildtype.

haematologica | 2019; 104(6)

1139

Y. Hirata et al.

highly expressed CD39 and CD73 as compared to the rest of the CD4+ nonTreg (CD150neg-low) (Figure 3B-D, Online Supplementary Figure S3B). CD150high populations comprised 20% of CD4+ nonTreg and 40% of Treg (Figure 3E).

CD150high nonTreg frequently showed a CD44highCD62Llow effector memory phenotype as compared to the rest of the CD4+ nonTreg (CD150neg-low) (Figure 3F; Online Supplementary Figure S3C), which is consistent with our previous observations in CD150high Treg.7 The frequencies of CD39high and CD73high cells among CD150high nonTreg were equivalent to those among CD150high Treg, and high-

er than those in BM CD150low Treg, BM CD150neg-lowCD4+ nonTreg, and lymph node CD4+ nonTreg (Figure 3B,D, Online Supplementary Figure S3A,D).

CD39 and CD73 expression among CD45- mesenchymal cells was further analyzed following division of the CD45- cells into the following three populations; CD45CD31+ vasculature; CD45-CD31-CD140a+CD140b+ cells; cells (Online and CD45-CD31-CD140a-CD140bSupplementary Figure S3E). Previous studies suggested that the former two populations were putative cellular constituents of the HSC niche.2,3 CD45-CD140a+CD140b+ cells

A

B

C

D

E

F

G

H

Figure 3. CD39 and CD73 expression levels in various bone marrow cell populations. (A) Flow cytometric analysis of CD73 expression levels in different bone marrow (BM) cell populations: regulatory T cells (Treg: CD4+CD3+NK1.1-FoxP3YFP+; non-regulatory T cells (nonTreg: CD4+CD3+NK1.1-FoxP3YFP-); hematopoietic stem cells (HSC: CD150+CD48-cKit+Sca1+Lin-); hematopoietic stem and progenitor cells (HSPC: cKit+Sca1+Lin-); CD8 cells (CD8+CD3+NK1.1-); natural killer cells (NK: NK1.1+CD3-); CD4NKT cells (NK1.1+CD3+CD4+ NKT cells); and B cells (B220+ cells). (B) Frequencies of CD73high and CD73int cells in different BM cell populations with the thresholds of CD73 expression levels determined as shown in Figure 3A, D, G and H: CD150H Treg (CD150high Treg); CD150L Treg (CD150low Treg); CD150H nonTreg (CD150high nonTreg); and CD150N-L nonTreg (CD150neg-low nonTreg). Thresholds of CD150 expression levels were determined as in Figure 3E. CD140a/b+: CD45CD140a+CD140b+CD31-. CD140a/b-: CD45-CD140a-CD140b-CD31-. CD31+: CD45-CD31-CD140a-CD140b-. The results were reproducible in two independent experiments (n=6). A representative figure is shown here. Data are presented as mean Âą SD. (C) Representative histograms showing levels of CD150 expression on HSC, Treg, and CD4+ nonTreg. (D) Representative histograms showing levels of CD39 or CD73 expression on CD150high Treg, CD150low Treg, CD150high nonTreg, and CD150neglow nonTreg. (E-F) Representative flow cytometric dot plots of BM CD4+CD3+NK1.1- T cells including both Treg and nonTreg (E) and BM CD4+ nonTreg (F). (G-H) Representative histograms showing levels of CD39 and CD73 expression of CD150 high Treg, CD150low Treg, CD140a+CD140b+CD45-CD31- cells, CD140a-CD140b-CD45CD31- cells (G), and CD31+CD45- cells (H). Iso ctrl Ab: isotype control antibody.

1140

haematologica | 2019; 104(6)

CD73 regulates HSC quiescence

were shown to overlap exclusively with leptin receptorpositive (lepr+) perivascular niche cells.2,3 CD39 and CD73 expression was observed on CD31+ vasculature and CD45CD31-CD140a+CD140b+ cells, but not on CD45-CD31CD140a-CD140b- cells (Figure 3G,H, Online Supplementary Figure S3F). The frequencies of CD39high and CD73high cells within CD31+ vasculature and CD45-CD140a+CD140b+ cells were comparable to those in CD150low Treg, and lower than those in CD150high Treg and CD150high nonTreg (Figure 3B,G-H, Online Supplementary Figure S3A). Taken together, these observations indicate that, while various BM cell populations expressed CD39 and/or CD73, CD73high cells were frequently found within CD150high Treg and CD150high nonTreg.

observations that transfer of CD150high Treg reversed the increase of HSC in mice with conditional deletion of CD39 in Treg.7 These results suggest that CD150high Treg and CD150high nonTreg contribute largely to CD73-mediated HSC regulation.

The downstream signaling of CD73 in HSC regulation was further assessed. HSC showed higher levels of expression A2AR than various other BM cell populations (Figure 4B). In vivo A2AR agonist treatment significantly reversed the increase of HSC in CD73 KO mice (Figure 4C). These observations suggest that A2AR signaling plays an important role in CD73-mediated HSC regulation.

The increase of hematopoietic stem cells in CD73 knockout mice was reversed by transfer of CD150high regulatory T cells and CD150high non-regulatory T cells and by in vivo adenosine receptor agonist treatment

Discussion

To assess the role of CD150high Treg and CD150high CD4+ nonTreg in CD73-mediated HSC regulation, we analyzed how transfer of these T-cell populations influences HSC. Transfer of CD150high Treg and of CD150high CD4+ nonTreg significantly reversed the increase of HSC in CD73 KO mice relative to control wildtype mice. In contrast, the size of the HSC pool was not significantly altered by transfer of CD150neg-low nonTreg or CD150low Treg (Figure 4A). These observations are consistent with our previous

A

This study identified CD73 and CD150high nonTreg as important regulators of HSC quiescence and abundance in the adult BM. Global CD73 deletion increased ROS levels in HSC, and HSC numbers. This increase of HSC was reversed by anti-oxidant treatment, suggesting that CD73 maintains HSC quiescence by preventing oxidative stress.

Because HSC did not express CD73 but CD39, CD73mediated HSC regulation is driven by the microenvironment. While various BM cells showed intermediate levels of CD73 expression, CD73high cells were frequently found within unique CD150high Treg and CD150high nonTreg.

C

B

Figure 4. CD73 of CD150high regulatory T cells and CD150high non-regulatory T cells regulates hematopoietic stem cells via adenosine 2A receptors. (A) Numbers of hematopoietic stem cells (HSC) in CD73 knockout (KO) mice 7 days after intravenous injection of CD150high bone marrow (BM) regulatory T cells (Treg), CD150low BM Treg, CD150high BM nonTreg, or CD150low BM nonTreg (30,000 cells/mouse). Data were pooled from three independent experiments (total 4-10 mice/group) and analyzed by one-way analysis of variance (ANOVA) with the Bonferroni multiple comparison test. (B) Levels of adenosine 2A receptor (A2AR) expression in different BM cell populations. MFI: mean fluorescence intensity. Four mice were used in the experiment. (C) HSC numbers in CD73 KO mice that received daily intraperitoneal injections of PSB0777 (25 Îźg/mouse) for 7 days. The total HSC numbers in one tibia and one femur were analyzed 1 day after the final injection. The results from two independent experiments were pooled (total 4-10 mice/group).

haematologica | 2019; 104(6)

1141

Y. Hirata et al.

Transfer of these CD150high Treg and CD150high nonTreg, but not of CD150low Treg or CD150neg-low nonTreg, reversed the increase of HSC in CD73 KO mice. Additionally, pharmacological activation of A2AR, highly expressed by HSC, reversed the increase of HSC in CD73 KO mice.

Taken together, these results suggest that CD73 of CD150high Treg and CD150high nonTreg regulates HSC quiescence and abundance via A2AR. To the best of our knowledge, this is the first study showing the role of conventional T cells in HSC regulation. This work is complemented by our recent study7 showing that CD39 on CD150high Treg played a critical role in maintaining HSC quiescence. As both CD150high Treg and CD150high nonTreg frequently displayed an effector memory T-cell phenotype,7 the observations of our current and previous studies7 suggest that BM CD4 memory T cells and memory Treg coordinate each other to generate extracellular adenosine via CD39 and CD73, maintaining HSC quiescence. As the BM is known to be a site to which memory T cells frequently home and in which they are maintained,9,10 memory T cells and Treg generated following infection may play important roles in protecting BM HSC from oxidative and inflammatory stresses, controlling hematopoiesis. As CD150high Treg frequently localized adjacent to HSC, a future histological analysis is warranted to identify the spatial distribution of CD150high nonTreg (CD3+CD4+NK1.1-FoxP3-) with respect to HSC (CD150+CD48-CD41-Lin-), although such a study is technically challenging because of the requirement of multiple colors.

Our study does not rule out the possibility that HSC are regulated by other adenosine receptors or by P2 receptors that bind ATP metabolized by CD39. HSC expression of CD39, but not of CD73, may reflect the possibility that tight control of ATP/adenosine ratios is required for the maintenance of HSC quiescence. Indeed, a previous study showed that global P2YR deletion abrogated the radioresistance of HSC.16 However, under normal conditions,

References 1. Kunisaki Y, Bruns I, Scheiermann C, et al.

Arteriolar niches maintain haematopoietic stem cell quiescence. Nature. 2013;502 (7473):637-643.

2. Ding L, Saunders TL, Enikolopov G, Morrison SJ. Endothelial and perivascular cells maintain haematopoietic stem cells.

Nature. 2012;481(7382):457-462.

3. Ding L, Morrison SJ. Haematopoietic stem cells and early lymphoid progenitors occupy distinct bone marrow niches. Nature.

2013;495(7440):231-235.

4. Itkin T, Gur-Cohen S, Spencer JA, et al.

Distinct bone marrow blood vessels differentially regulate haematopoiesis. Nature.

2016;532(7599):323-328.

5. Jang YY, Sharkis SJ. A low level of reactive oxygen species selects for primitive hematopoietic stem cells that may reside in the low-oxygenic niche. Blood. 2007;110(8): 3056-3063.

1142

P2YR KO mice did not show significant alteration of HSC numbers,16 suggesting that the observed phenotypes in CD73 KO mice and FoxP3cre CD39fl/wt mice under normal conditions were not attributable to P2YR. CD150high nonTreg and CD150high Treg are likely to generate adenosine in concert with various CD39+ or CD73+ BM cell populations, including HSC and the following two niche constituents: CD31+ vasculature and CD140a+CD140b+CD45- mesenchymal cells which exclusively overlap with lepr+ perivascular cells.2,3 Nevertheless, these two niche constituents are unlikely to be the major source of adenosine, because the frequencies of CD39high and CD73high cells in these mesenchymal cells were comparable to those in CD150low Treg and transfer of these latter cells did not alter HSC numbers in CD73 KO mice.

Indeed, our additional study using leprcre CD39fl/wt mice showed that conditional deletion of CD39 in lepr+ cells did not alter HSC number or reconstituting potential of BM cells (Online Supplementary Figure S4A-C). This observation further supports the important role of CD150high Treg and CD150high nonTreg in adenosine-mediated HSC regulation.

In summary, this work showed that CD150high Treg and CD150high nonTreg maintain HSC quiescence via CD73.

An examination of the roles of adenosine and memory T cells in human hematopoiesis and transplantation is warranted.

Acknowledgments This work was supported by NIH NHLBI R01HL129506 (JF), an ASH Junior Faculty Scholar Award (JF), a Schaefer Research Scholar Award (JF) and an Uehara Memorial Foundation Research Fellowship Award (YH). Research reported in this publication was performed in the CCTI Flow Cytometry Core, supported in part by the Office of the Director, National Institutes of Health under award S10OD020056. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

6. Arai F, Hirao A, Ohmura M, et al.

Tie2/angiopoietin-1 signaling regulates hematopoietic stem cell quiescence in the bone marrow niche. Cell. 2004;118(2):149161.

7. Hirata Y, Furuhashi K, Ishii H, et al.

CD150(high) bone marrow tregs maintain hematopoietic stem cell quiescence and immune privilege via adenosine. Cell Stem Cell. 2018;22(3): 445-453.

8. Jacobson KA, Gao ZG. Adenosine receptors as therapeutic targets. Nat Rev Drug Discov.

2006;5(3):247-264.

9. Mazo IB, Honczarenko M, Leung H, et al.

Bone marrow is a major reservoir and site of recruitment for central memory CD8+ T cells. Immunity. 2005;22(2):259-270.

10. Di Rosa F, Gebhardt T. Bone marrow T cells and the integrated functions of recirculating and tissue-resident memory T cells. Front Immunol. 2016;7:51.

11. Abbas HA, Maccio DR, Coskun S, et al.

Mdm2 is required for survival of hematopoietic stem cells/progenitors via dampening of

12.

13.

14.

15.

16.

ROS-induced p53 activity. Cell Stem Cell.

2010;7(5):606-617.

Miyamoto K, Araki KY, Naka K, et al.

Foxo3a is essential for maintenance of the hematopoietic stem cell pool. Cell Stem Cell. 2007;1(1):101-112.

Liu J, Cao L, Chen J, et al. Bmi1 regulates mitochondrial function and the DNA damage response pathway.

Nature.

2009;459(7245):387-392.

Juntilla MM, Patil VD, Calamito M, et al.

AKT1 and AKT2 maintain hematopoietic stem cell function by regulating reactive oxygen species. Blood. 2010;115(20):40304038.

Lewandowski D, Barroca V, Duconge F, et al. In vivo cellular imaging pinpoints the role of reactive oxygen species in the early steps of adult hematopoietic reconstitution.

Blood. 2010;115(3):443-452.

Cho J, Yusuf R, Kook S, et al. Purinergic P2Y(1)(4) receptor modulates stress-induced hematopoietic stem/progenitor cell senescence. J Clin Invest. 2014;124(7):3159-3171.

haematologica | 2019; 104(6)

ARTICLE

Iron Metabolism & its Disorders

The opposing effects of acute inflammation and iron deficiency anemia on serum hepcidin and iron absorption in young women

Ferrata Storti Foundation

Nicole U. Stoffel,1 Meryem Lazrak,2 Souhaila Bellitir,2 Nissrine El Mir,2 Asmaa El Hamdouchi,2 Amina Barkat,3 Christophe Zeder,1 Diego Moretti,1 Hassan Aguenaou2 and Michael B. Zimmermann1

ETH Zürich, Laboratory of Human Nutrition, Institute of Food Nutrition and Health, Department of Health Science and Technology, Zürich, Switzerland; 2Ibn Tofaïl UniversityCNESTEN, Joint Research Unit in Nutrition and Food, RDC-Nutrition AFRA/IAEA, RabatKénitra, Morocco and 3Mohamed V University, Unit of Research on Nutrition and Health of Mother and Nutrition, Faculty of Medicine and Pharmacy, Rabat, Morocco 1

Haematologica 2019 Volume 104(6):1143-1149

ABSTRACT

H

epatic hepcidin synthesis is stimulated by inflammation but inhibited during iron deficiency anemia (IDA). In humans, the relative strength of these opposing signals on serum hepcidin and the net effect on iron absorption and systemic iron recycling is uncertain. In this prospective, 45-day study, in young women (n=46; age 18-49 years) with or without IDA, we compared iron and inflammation markers, serum hepcidin and erythrocyte iron incorporation from 57Fe-labeled test meals, before and 8, 24 and 36 hours (h) after influenza/DPT vaccination as an acute inflammatory stimulus. Compared to baseline, at 24-36 h after vaccination: 1) interleukin-6 increased 2-3-fold in both groups (P<0.001); 2) serum hepcidin increased >2-fold in the non-anemic group (P<0.001), but did not significantly change in the IDA group; 3) serum iron decreased in the non-anemic group (P<0.05) but did not change in the IDA group; and 4) erythrocyte iron incorporation did not change in either of the two groups, but was approximately 2-fold higher in the IDA group both before and after vaccination (P<0.001). In this study, mild acute inflammation did not increase serum hepcidin in women with IDA, suggesting low iron status and erythropoietic drive offset the inflammatory stimulus on hepcidin expression.

In non-anemic women, inflammation increased serum hepcidin and produced mild hypoferremia, but did not reduce dietary iron absorption, suggesting iron-recycling macrophages are more sensitive than the enterocyte to high serum hepcidin during inflammation. The study was registered as a prospective observational trial at clinicaltrials.gov identifier: 02175888. The study was funded by the International Atomic Energy Agency.

Introduction The hepatic hormone hepcidin regulates serum iron concentrations and iron homeostasis.1 Hepcidin synthesis is stimulated by high liver iron stores, high plasma iron concentrations and inflammation, and is inhibited by hypoxia and erythropoietic drive.1 In conditions such as the iron-loading anemias, the anemia of chronic disease, and acute infection during iron deficiency anemia (IDA), several of these factors simultaneously generate opposing signals on hepcidin expression. The relative strength of these opposing signals to regulate serum hepcidin (SHep) in humans is uncertain.

Iron-deficient mice up-regulate hepcidin expression when injected with lipopolysaccharide (LPS) to induce inflammation.2 Other animal studies suggest erythroid demand for iron is a stronger regulator of hepcidin expression than inflammation.3-5 Most human studies examining opposing stimuli on SHep have been descriptive; some have suggested that iron status and/or erythropoiesis are the major regulators of SHep,4,6,7 others suggest inflammation is associated with haematologica | 2019; 104(6)

Correspondence: NICOLE U. STOFFEL nicole.stoffel@hest.ethz.ch Received: October 8, 2018.

Accepted: January 2, 2019.

Pre-published: January 10, 2019. doi:10.3324/haematol.2018.208645 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/104/6/1143 ©2019 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

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

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

1143

N.U. Stoffel et al. high SHep even in anemic subjects.8,9 Nearly all experimental data are derived from cells or mice, where often very strong stimuli are applied (e.g. injection of LPS,2,3,5 phlebotomy,4 severe iron deficiency2,5). In humans, there is a lack of experimental data describing the effect of milder, physiological changes in these opposing stimuli on SHep and the net effects on iron recycling and iron absorption, but these interactions are common and relevant in many disorders.

Therefore, the objective of this prospective study was to assess changes in iron markers, SHep, iron absorption and erythrocyte iron incorporation during acute inflammation, in both non-anemic women and women with IDA. Previous human studies reported the effects of an acute inflammatory stimulus [e.g. infusions of LPS10 or interleukin-6 (IL-6)11] on SHep, but they did not assess effects on iron absorption, or how anemia or iron status may modulate this response. As the inflammatory stimulus in this study, we used vaccination, a practical, safe and standardized model for the study of mild-to-moderate inflammation in humans.12,13 Our hypotheses were: 1) in nonanemic women, vaccination would induce acute inflammation and increase Shep. This would decrease iron absorption and produce hypoferremia; and 2) in contrast, in women with IDA, vaccination would induce acute inflammation, but would not increase SHep or affect serum iron or iron absorption. We used stable iron isotope techniques to quantify erythrocyte iron incorporation of dietary iron before and after vaccination.

Methods Study subjects We recruited women from the staff of the University Hospital Ibn Sina in Rabat, Morocco. Detailed inclusion criteria are described in the Online Supplementary Appendix. In this prospective, 45-day study, in women (n=46, age 18-49 years) with IDA or without anemia, we compared iron and inflammation markers and SHep before and 8, 24 and 36 hours (h) after influenza/diphtheria-tetanus-pertussis (DTP) vaccination and erythrocyte iron incorporation from 57Fe-labeled test meals, before and 24 h after the vaccination as an acute inflammatory stimulus (Figure 1). The study was approved by the ethics committees of the ETH Zurich, Zurich, Switzerland and the University Mohammed V, Rabat, Morocco. All participants gave informed written consent.

On study day 1, an afternoon baseline blood sample was taken.

On study day 2, after an overnight fast, a baseline morning blood sample was taken and we administered a test meal containing 6 mg labeled 57Fe as ethylenediaminetetraacetic acid ferric sodium salt (NaFeEDTA), added to a standardized test meal, given with bottled water, as described in the Online Supplementary Appendix.

Blood samples were taken in the afternoon on day 2 as well as the next morning (day 3). After a 19-day isotope incorporation period, on study day 22, a blood sample was taken to measure erythrocyte iron incorporation; this blood sample also served as the new baseline afternoon sample for the second absorption study. In the morning of day 23, a morning blood sample was taken. Then, all subjects received the trivalent Influenza Virus Vaccine Vaxigrip (Sanofi Pasteur, Lyon, France) and the DTP Virus Vaccine Dultavax (Sanofi Pasteur) given intramuscularly. Blood samples were taken at 8 h, 24 h and 36 h after vaccination. At 24 h after vaccination, on study day 24, an identical labeled test meal was administered, as described above. The final blood sample was taken on day 45.

We assessed total and fractional iron absorption (FIA) by measur1144

ing the amount of stable isotopic tracers incorporated in red blood cells 19 days after administration of the labeled test meals.14-16 Hemoglobin (Hb), iron- and inflammatory biomarkers were measured as described in the Online Supplementary Appendix.

Assuming a standard deviation (SD) of 0.20 on differences in log transformed erythrocyte iron incorporation from previous ETH studies, a type I error rate of 5% and 80% power, we expected to detect a difference in FIA of 35% within groups with a sample size of 20 subjects per group. Assuming a drop-out rate of 20%, we enrolled 50 women (25 anemic and 25 non-anemic women).

Statistical analysis We performed the statistical analyses using SPSS (IBM SPSS statistics, v.22), as described in detail in the Online Supplementary Appendix. We used linear mixed effect model analysis to assess the effect of the group (anemic vs. non-anemic) and treatment (vaccination) on different variables. Group and treatment were defined as fixed effects, participants as random intercept effects using a variance component structure matrix. Regression analyses were performed with SHep, FIA and serum iron as dependent variables.

Pearson and Spearman correlations were applied. For withingroup effects, dependent sample t-tests or related samples nonparametric tests were used. P<0.05 was considered significant.

Results We began recruiting on 1st September 2017, and from September 2017 to February 2018 we enrolled 50 women (28 non-anemic and 22 with IDA) into the study. We completed the study on 29th March 2018. Six women in the non-anemic group left the study because they no longer wanted to participate in the study: three before the 19 day blood sample (when we measured erythrocyte iron incorporation from the first test meal) and three after the 19 day blood sample. In the IDA group, one subject left the study before the 19 day blood sample because she no longer wanted to participate in the study (Figure 1). Data from the three non-anemic women who left the study after the 19 day blood sample were included in the analytical models, resulting in a total of n=46 women (21 with IDA and 25 non-anemic).

Baseline characteristics of the subjects by group are shown in Table 1. There were no significant betweengroup differences in age, Body Mass Index or markers of inflammation, and no subject had increased markers of inflammation. There were significant between-group differences in hemoglobin (Hb), serum ferritin (SF), soluble transferrin receptor (sTfR), body iron stores (BIS), Shep, and erythropoietin (EPO) (all P<0.001). Eight of the women in the non-anemic group were iron-deficient, as defined by SF<15 Îźg/L.17 There was a significant vaccination effect (P<0.001) on IL-6, but no significant group effect or group-vaccination interaction (Table 2). There were no significant betweengroup differences in IL-6 measured at 4:00 pm the day before and at 8:00 am just before vaccination or at 8 h, 24 h, and 36 h after vaccination (Figure 2A). Median interquartile range (IQR) IL-6 (pg/mL) significantly increased in both groups comparing baseline to 24 h after vaccination (P<0.001): in the non-anemic group from 1.12 (0.92-1.66) to 3.37 (2.88-4.32), and in the IDA group from 1.53 (1.41-1.81) to 3.14 (2.48-4.33) (Table 2).

There was a significant group (P<0.001) and vaccination (P<0.001) effect on SHep, but no significant group-vaccihaematologica | 2019; 104(6)

Iron, absorption, hepcidin, anemia, inflammation

nation interaction (Table 2). There were significant between-group differences in SHep measured at 4:00 pm and at 8:00 am before vaccination and at 8 h, 24 h, and 36 h after vaccination (all P<0.05) (Figure 2B). In the non-anemic group, median IQR SHep (nM) significantly increased from 1.60 (0.93-2.86) before to 3.56 (1.04-5.53) at 24 h after vaccination (P<0.001). In contrast, in the IDA group, vaccination did not induce a significant increase in SHep: median (IQR) SHep (nM) was 0.45 (0.23-0.61) before and 0.45 (0.32-1.17) 24 h after vaccination (Table 2).

There was a significant group effect (P<0.001) and group-vaccination interaction (P<0.001) on serum iron and transferrin saturation (TSAT), but no significant vaccination effect (Table 3). There were significant betweengroup differences in serum iron and TSAT at baseline (P<0.001 for both) and at 24 h after vaccination (P<0.05 for both). In the non-anemic group, geometric mean [-Standard Deviation (SD), +SD] serum iron (μg/mL) significantly decreased from 0.91 (0.71, 1.17) before to 0.78

(0.59, 1.02) at 24 h after vaccination (P<0.05) (Figure 3A).

Geometric mean (-SD, +SD) TSAT (%) decreased from 19.19 (14.69, 25.07) before to 16.23 (12.39, 21.26) at 24 h after vaccination (P=0.066). In the IDA group, vaccination had no significant effect on either serum iron or TSAT. There was a significant group effect on FIA (erythrocyte iron incorporation) (P<0.001), but no significant vaccination effect or group-vaccination interaction (Table 2).

Comparing erythrocyte iron incorporation before and after vaccination, there were no significant differences within either of the two groups (Figure 3B). However, between groups, erythrocyte iron incorporation was significantly higher (by approx. 2-fold) in women with IDA both before and after vaccination (Table 2).

Iron and inflammation indicators at 24 h after vaccination are shown in Table 3. There was a significant group effect on SF, sTfR, BIS, EPO (P<0.001), but none on CRP or AGP. There was a significant vaccination effect on sTfR (P<0.01), CRP (P<0.05) and AGP (P<0.001); all of these

Figure 1. Study design. IDA: iron deficiency anemia; DTP: diphtheria-tetanus-pertussis.

Table 1. Baseline characteristics of the women (n=46) in the iron deficiency anemia group and the non-anemic group.

Age, years* Body Mass Index, kg/m2** Hemoglobin, g/dL* Serum ferritin, μg/L*** Serum transferrin receptor, mg/L*** Body iron stores, mg/kg BW** Serum iron, μg/mL*** Total iron binding capacity, μg/mL* Transferrin saturation, %*** Erythropoietin, mIU/mL*** C-reactive protein, mg/L*** Alpha-1-acid glycoprotein, g/L***

IDA (n=21)

Non-anemic (n=25)

23 (21-27) 22.9±2.5 11·3 (10.7-11.6)¶ 6·3 (3.9, 10.3)¶ 8·33 (6.08, 11.43)¶ -2.5±2.7¶ 0·51 (0·33, 0.78)¶ 4·79 (4·49-5.06) 10·78 (6·89, 16.89)¶ 20·37 (11.48, 36.17)¶ 0·70 (0.18, 2.76) 0·71 (0.51, 0.98)

24 (23-30) 22.3 ± 3.1 13·3 (12.6-13.9) 23·8 (10.7, 52.9) 5·99 (4.60, 7.79) 3.5±3.5 0.91 (0.71, 1.17) 4.81 (4.41-5.01) 19.19 (14·69, 25.07) 8.61 (5.40, 13.74) 0.59 (0.13, 2.59) 0.72 (0.49, 1.05)

BW: body weight; IDA: iron deficiency anemia. *Median [interquartile range (IQR)]. **Mean±Standard Deviation (SD). ***Geometric mean (-SD, +SD). ¶Different between groups (P<0.001).

haematologica | 2019; 104(6)

1145

N.U. Stoffel et al.

variables were lower after vaccination, and the decrease in sTfR in the IDA group reached borderline significance (P=0.063). There was a significant group-vaccination interaction on SF (P<0.001) and BIS (P<0.01).

There was no correlation between IL-6 and SHep at baseline in either group. At 24 h after vaccination, IL-6 and SHep significantly correlated in the non-anemic group

(r=0.426; P<0.05), but did not in the IDA group. In the regressions including all subjects, sTfR, Hb, EPO and IL-6 explained 55% and 62% of the variation of SHep at baseline and at 24 h after vaccination, respectively. At baseline in all subjects, the only significant predictors of SHep were EPO (b=-0.570; P<0.01) and Hb (b=0.372; P<0.05). In contrast, at 24 h after vaccination, the only significant predic-

Table 2. Serum interleukin-6 (IL-6), serum hepcidin and erythrocyte iron incorporation (iron absorption and utilization), at baseline and 24 hours after vaccination, in the iron deficiency anemia group and the non-anemic group.

IL-6, pg/mL IDA Non-anemic Serum hepcidin, nM IDA Non-anemic Erythrocyte iron incorporation, % IDA Non-anemic

Baseline (n=46)

24h after vaccination (n=43)

Group effect

P Vaccination effect

Group-vaccination effect

1.53 (1.41 - 1.81) 1.12 (0.92 - 1.66)

3.14 (2.48 - 4.33)° 3.37 (2.88 - 4.32)°

0.212

<0.001

0.349

0.45 (0.23-0.61)* 1.60 (0.93 - 2.86)

0.45 (0.32 - 1.17)* 3.56 (1.04 - 5.53)°

<0.001

<0.001

0.810

<0.001

0.396

0.629

36.15 (26.08 - 39.35)* 33.09 (28.84 - 38.74)* 16.66 (9.33 - 24.05) 15.89 (11.86 - 24.71)

h: hours; n: number; IDA: iron deficient anemia. Medians [interquartile range (IQR)]. *Different between groups (P<0.001). °Within group, different from baseline (P<0.001).

Analyzed by linear mixed models with Bonferroni-corrected multiple comparisons.

Table 3. Iron and inflammatory variables 24 hours after vaccination, in the iron deficiency anemia group and the non-anemic group.

24h after vaccination (n=43) Serum ferritin, µg/L* IDA 10.89 (6.86, 17.30)ⱡ Non-anemic 14.98 (7.31, 30.70)ⱡ Serum transferrin receptor, mg/L* IDA 6.60 (4.73, 9.22)$ Non-anemic 5.32 (3.98, 7.11) Body iron stores, mg/kg BW** IDA 0·63 ± 2·28$ⱡ Non-anemic 2·26 ± 3·30 Serum iron, μg/mL* IDA 0·59 (0.38, 0.93)$ Non-anemic 0·78 (0·59, 1·02)# Total iron binding capacity, μg/mL*** IDA 4.89 (4.69-5.13) Non-anemic 4.71 (4.47-5.16) Transferrin saturation, %* IDA 12.01 (7.60, 18.97)$ Non-anemic 16.23 (12.39, 21.26) Erythropoietin, mIU/mL* IDA 21.95 (12.62, 38.17)¶ Non-anemic 9.59 (6.23, 14.78) C-reactive protein, mg/L* IDA 0.27 (0.10, 0.69)ⱡ Non-anemic 0.45 (0.16, 1.26) Alpha-1-acid glycoprotein, g/L* IDA 0.45 (0.33, 0.61)° Non-anemic 0.52 (0.38, 0.72)ⱡ

Group effect

P Vaccination effect

Group-vaccination effect

<0.001

0.716

<0.001

<0.001

<0.01

0·357

<0.001

0.175

<0·01

<0.001

0.971

<0·01

0.284

0.116

0·827

<0.001

0.637

<0·05

<0.001

0.057

0·876

0.499

<0.05

0.138

0.248

<0.001

0.293

h: hours; n: number; IDA: iron deficient anemia. *Geometric mean [-Standard Deviation (SD), +SD]. **Mean±SD. ***Median [interquartile range (IQR)]. ¶Different between groups (P<0.001). $Different between groups (P<0.05). °Within group, different from baseline (P<0.001) (baseline values shown in Table 1). ⱡWithin group, different from baseline (P<0.01) (baseline values shown in Table 1). #Within group, different from baseline (P<0.05) (baseline values shown in Table 1). Analyzed by linear mixed models with Bonferroni-corrected multiple comparisons.

1146

haematologica | 2019; 104(6)

Iron, absorption, hepcidin, anemia, inflammation

tors of SHep were Hb (b= 0.520; P<0.01) and IL-6 (b= 0.508; P<0.001). In the non-anemic group, at baseline, erythrocyte iron incorporation significantly correlated with SF (r=-0.822; P<0.001), SHep (r=-0.792; P<0.001), EPO (r=0.631; P<0.01) and Hb (r=-0.475; P<0.05); at 24 h after vaccination, incorporation significantly correlated with SHep (r=-0.708; P<0.001), Hb (r=-0.563; P<0.01) and serum iron (r=0.516; P<0.05). In the IDA group, at baseline and at 24 h after vaccination, erythrocyte iron incorporation was not significantly correlated with Hb, iron biomarkers, SHep or EPO. In the regression analysis including all subjects, sTfR, Hb, EPO and IL-6 explained 53% and 50% of the variation of erythrocyte iron incorporation at baseline and at 24 h after vaccination, respectively. In all subjects, the only significant predictor of erythrocyte iron incorporation was Hb, both at baseline (b= -0.410; P<0.05) and 24 h after vaccination (b= -0.659; P<0.01). In the regressions including all subjects, SHep, Hb, EPO and IL-6 explained 61% and 38% of the variation of serum iron at baseline and at 24 h after vaccination, respectively. At baseline in all subjects, significant predictors of serum iron were Hb (b=0.549; P<0.01) and EPO (b=-0.333; P=0.05). In contrast, at 24 h after vaccination, significant predictors of

A

serum iron were Hb (b= 0.864; P<0.01), SHep (b= -0.449; P<0.05) and IL-6 (b= 0.347; P<0.05).

Discussion Our main findings are that at 24-36 h after vaccination: 1) there was a comparable 2-3-fold increase in serum IL-6 in both groups (P<0.001); 2) there was a significant >2-fold increase in SHep in the non-anemic group (P<0.001), but no significant change in SHep in the IDA group; 3) serum iron decreased only in the non-anemic group (P<0.05); and 4) there was no significant change in erythrocyte iron incorporation in either of the two groups; incorporation was approximately 2-fold higher in the IDA group both before and after vaccination (P<0.001).

Previous experimental human studies have examined the hepcidin response to inflammation.10,11 In healthy adults given an infusion of IL-6 (iron status was not reported but subjects were presumably non-anemic), after 2 h urinary hepcidin increased 7.5-fold, while serum iron and TSAT decreased by 33-34%.11 In healthy adults injected with LPS (iron status was not reported but subjects were presumably

B

Figure 2. Interleukin-6 (IL-6) and hepcidin response to influenza/diphtheriatetanus-pertussis (DTP) vaccination. (A) Serum IL-6 and (B) serum hepcidin concentrations before and at 8, 24 and 36 hours (h) after vaccination in the iron deficiency anemia (IDA) group (n=21) and the non-anemic group (n=22). BL: baseline.

A

B

Figure 3. Effects of influenza/diphtheriatetanus-pertussis (DTP) vaccination on serum iron and erythrocyte iron incorporation. (A) Serum iron and (B) erythrocyte iron incorporation (iron absorption and utilization) at baseline (n=46) and at 24 hours (h) after vaccination (n=43) in the iron deficiency anemia (IDA) group and the non-anemic group.

haematologica | 2019; 104(6)

1147

N.U. Stoffel et al.

non-anemic), serum IL-6 increased within 3 h after injection, and urinary hepcidin peaked at 6 h, followed by significant hypoferremia.10 In a study in Gambian newborns, routine immunizations at birth did not affect serum IL-6 or SHep at 72-96 h post vaccination, but the inflammationhepcidin axis was already activated, likely due to the birth process.18 However, these studies did not compare responses between anemic and non-anemic subjects or measure effects on iron absorption.

In our study, vaccination induced a rapid and sustained inflammatory response reflected in an approximately 2- to 3-fold increase in IL-6 apparent at 8 h after vaccination and persisting at 36 h in both the IDA and non-anemic groups (Figure 2A). Despite this, in the women with IDA, SHep did not significantly increase (Figure 2A and B). Several factors likely contributed to this effect. High circulating diferric transferrin and high liver iron stores increase hepatic hepcidin synthesis via stimulation of the bone morphogenic protein [sons of mothers against decapentaplegic (BMPSMAD) pathway].1 EPO, the main driver of erythropoiesis, stimulates a hepcidin-suppressing factor synthesized in the bone marrow; this factor may be erythroferrone. In a recent study in mice, erythroferrone suppressed hepcidin by inhibiting hepatic BMP/SMAD signaling through BMP5, BMP6, and BMP7.19 However, the role of erythroferrone during IDA remains uncertain.20 In the IDA group, BMPSMAD signaling was likely suppressed by low TSAT, depleted liver iron stores, and high erythropoietic drive, as indicated by high EPO and sTfR concentrations (Table 1).

Our data suggest that moderate IL-6 stimulation of the Janus kinase/ signal transducer and activator of transcription (JAK/STAT) and BMP-SMAD pathway was unable to overcome this suppression, and SHep remained low. In contrast, in the non-anemic group, BMP-SMAD signaling was not suppressed, and, as a result, IL-6 activation resulted in a rapid increase in SHep. These data suggest that, in mild IDA, low iron status and erythropoietic drive can keep SHep low even in the face of an acute inflammatory stimulus. In addition to the suppression of erythropoiesis by iron restriction through hepcidin, cytokines may directly affect erythropoiesis.21 In both the IDA and the non-anemic groups, erythropoiesis appeared to be mildly suppressed 24 h after vaccination, as indicated by a vaccination effect to decrease sTfR and increase EPO (Table 3).

In animal studies, inflammation is a strong inducer of hepcidin, but its effects can be blunted by iron deficiency and/or enhanced erythropoiesis.3-5 In mice, erythropoietic drive down-regulated hepcidin even during inflammation induced by LPS injection.3 Ferroportin transcription in macrophages may be attenuated by inflammation independent of hepcidin,22,23 but activation of Nrf2 reverses this attenuation,24 suggesting that iron may dominate over inflammatory stimuli. Conversely, other studies suggest erythroid and inflammatory regulators dominate over iron stores: iron-deficient mice injected with LPS up-regulated hepcidin expression,2 while iron-loaded mice with experimentally induced anemia down-regulated hepcidin expression.25 Studies in humans with the anemia of chronic inflammation (ACD) and/or IDA suggested erythroid demand for iron is a stronger regulator of hepcidin expression than mild inflammation: SHep was similar between ACD/IDA and IDA, but was higher in ACD; duodenal ferroportin expression was inversely related to SHep and SF, but not to IL-6; and IL-6 levels were similar between ACD (with high 1148

SHep) and ACD/IDA subjects (with low Shep).4 In African children, Hb and SF were positively associated with hepcidin while IL-6 levels were not.26 In another study in African children, erythropoietic drive (sTfR) was a much stronger negative predictor of SHep than inflammation.27 The differing results of these studies suggest that either inflammation or IDA can be the dominant factor regulating hepcidin, depending on the varying strengths of the opposing stimuli.

In our study, there was a significant and sustained increase in SHep after vaccination in the non-anemic group, with median SHep (nM) more than twice baseline values at 8, 24 and 36 h after vaccination. We anticipated this would decrease erythrocyte iron incorporation because in our previous studies, acute SHep increases of similar magnitude and duration (approx. 1-2 nM) in non-anemic women after the administration of high oral iron doses reduced iron absorption by approximately 40%.28,29 Contrary to our hypothesis, in the present study, there was no significant change in erythrocyte iron incorporation from a labeled test meal given at 24 h after vaccination, at the peak of the SHep increase, although the women developed mild hypoferremia. Several mechanisms may explain this effect.

Hepcidin promotes rapid degradation of ferroportin in liver cells and macrophages reducing iron recycling and serum iron,30 but enterocyte ferroportin may be less sensitive to acute changes in hepcidin.31-34 Another potential explanation why high SHep did not effect iron absorption in our study is that inflammation also induced hypoferremia, and the reduction in circulating diferric transferrin was sensed by the enterocyte through TfR1, leading to changes within the enterocyte35,36 that induced divalent metal transporter 1 (DMT-1) and ferroportin expression. This may have offset ferroportin degradation by SHep, and allowed continued iron export from the enterocyte. This is supported by our findings that, at 24 h after vaccination, serum iron was a significant predictor of erythrocyte iron incorporation in the non-anemic group, but not in the IDA group.

The strengths of this study are: 1) we prospectively studied the effects of a standardized (and safe) inflammatory stimulus in both non-anemic women and women with IDA; 2) our subjects were young women who were otherwise healthy and free of potential confounding comorbidities; 3) we precisely quantified erythrocyte iron incorporation (absorption and utilization) using iron stable isotopic labels. Limitation of the study are: 1) we included women in the IDA group who were only mildly anemic, and we induced only a moderate acute inflammatory state. More severe, chronic inflammation and/or anemia may have resulted in differing effects; 2) using the stable iron isotope method we measured erythrocyte iron incorporation, which reflects both iron absorption by enterocytes and iron utilization for production of erythrocytes in the bone marrow, and we could not differentiate between these.

Finally, we were unable to distinguish whether the lack of SHep increase in response to the inflammatory stimulus in the women with IDA was due to the effect of erythropoietic drive, iron deficiency or both.

To our knowledge, this is the first human experimental study showing that erythrocyte iron incorporation (absorption and utilization of dietary iron) is not reduced in non-anemic subjects by an acute increase in SHep that induces hypoferremia. This finding suggests the enterocyte may be less sensitive to the effect of acute changes in SHep than macrophage iron recycling. This pattern of regulation haematologica | 2019; 104(6)

Iron, absorption, hepcidin, anemia, inflammation

by hepcidin is consistent with the relative contributions of these pathways to the maintenance of body iron homeostasis. Also, to our knowledge, this is the first human experimental study showing that SHep does not increase in anemic subjects after a mild acute inflammatory stimulus. This suggests that, in IDA, iron homeostasis prioritizes correction of iron deficiency, rather than iron withholding, during mild acute infection/inflammation. Our findings provide new insights into how the relative and opposing stimuli that effect hepcidin expression combine to determine net circulating hepcidin, iron absorption and iron homeostasis in young women, and may be clinically rele-

References 1. Sangkhae V, Nemeth E. Regulation of the Iron Homeostatic Hormone Hepcidin. Adv Nutr.

2017;8(1):126-136.

2. Constante M, Jiang W, Wang D, Raymond VA, Bilodeau M, Santos MM. Distinct requirements for Hfe in basal and induced hepcidin levels in iron overload and inflammation. Am J Physiol Gastrointest Liver Physiol.

2006;291(2):G229-237.

3. Huang H, Constante M, Layoun A, Santos MM. Contribution of STAT3 and SMAD4 pathways to the regulation of hepcidin by opposing stimuli. Blood. 2009; 113(15):35933599.

4. Theurl I, Aigner E, Theurl M, et al. Regulation of iron homeostasis in anemia of chronic disease and iron deficiency anemia: diagnostic and therapeutic implications. Blood.

2009;113(21):5277-5286.

5. Darshan D, Frazer DM, Wilkins SJ, Anderson GJ. Severe iron deficiency blunts the response of the iron regulatory gene Hamp and proinflammatory cytokines to lipopolysaccharide. Haematologica. 2010; 95(10):1660-1667.

6. Kearney SL, Nemeth E, Neufeld EJ, Thapa D, Ganz T, Weinstein DA, et al. Urinary hepcidin in congenital chronic anemias. Pediatr Blood Cancer. 2007;48(1):57-63.

7. Gardenghi S, Marongiu MF, Ramos P, et al.

Ineffective erythropoiesis in beta-thalassemia is characterized by increased iron absorption mediated by down-regulation of hepcidin and up-regulation of ferroportin.

Blood.

2007;109(11):5027-5035.

8. Prentice AM, Doherty CP, Abrams SA, et al.

Hepcidin is the major predictor of erythrocyte iron incorporation in anemic African children.

Blood. 2012;119(8):1922-1928.

9. Hella J, Cercamondi CI, Mhimbira F, et al.

Anemia in tuberculosis cases and household controls from Tanzania: Contribution of disease, coinfections, and the role of hepcidin.

PloS One. 2018;13(4):e0195985.

10. Kemna E, Pickkers P, Nemeth E, van der Hoeven H, Swinkels D. Time-course analysis of hepcidin, serum iron, and plasma cytokine levels in humans injected with LPS. Blood.

2005;106(5):1864-1866.

11. Nemeth E, Rivera S, Gabayan V, et al. IL-6 mediates hypoferremia of inflammation by inducing the synthesis of the iron regulatory hormone hepcidin.

J Clin Invest.

2004;113(9):1271-1276.

haematologica | 2019; 104(6)

vant given that young women are a main target group for iron supplements and fortification to reduce iron deficiency and anemia.

Acknowledgments The authors thank all the women who participated in the study, the nursing staff and the other members of the RDCNutrition.

Funding They also thank the International Atomic Energy Agency for funding and technical assistance (MOR6022).

12. van der Beek MT, Visser LG, de Maat MPM. Yellow fever vaccination as a model to study the response to stimulation of the inflammation system. Vasc Pharmacol. 2002;39(3):117121.

13. Tsai MY, Hanson NQ, Straka RJ, et al. Effect of influenza vaccine on markers of inflammation and lipid profile. J Lab Clin Med. 2005;145 (6):323-327.

14. Hotz K, Krayenbuehl PA, Walczyk T. Mobilization of storage iron is reflected in the iron isotopic composition of blood in humans.

J Biol Inorg Chem. 2012;17(2):301-309.

15. Brown E, Hopper J Jr, Hodges JL Jr, Bradley B, Wennesland R, Yamauchi H. Red cell, plasma, and blood volume in the healthy women measured by radiochromium cell-labeling and hematocrit. J Clin Invest. 1962;41:2182-2190.

16. Hosain F, Marsaglia G, Noyes W, Finch CA. The nature of internal iron exchange in man.

Trans Assoc Am Physicians. 1962;75:59-63.

17. WHO. Nutritional anaemias: tool for effective prevention and control. 2017. (Available from: https://www.who.int/nutrition/publications/micronutrients/anaemias-tools-prevention-control/en/Last accessed 8 April 2019.

18. Prentice S, Jallow MW, Prentice AM, Group MR-IN. The effect of BCG on iron metabolism in the early neonatal period: A controlled trial in Gambian neonates. Vaccine.

2015;33(26):2963-2967.

19. Arezes J, Foy N, McHugh K, et al.

Erythroferrone inhibits the induction of hepcidin by BMP6. Blood. 2018; 4;132(14):14731477 20. Muckenthaler MU, Rivella S, Hentze MW, Galy B. A Red Carpet for Iron Metabolism.

Cell. 2017;168(3):344-361.

21. Nemeth E, Ganz T. Anemia of inflammation.

Hematol Oncol Clin North Am. 2014;28(4):671-681, vi.

22. Guida C, Altamura S, Klein FA, et al. A novel inflammatory pathway mediating rapid hepcidin-independent hypoferremia. Blood.

2015;125(14):2265-2275.

23. Deschemin JC, Vaulont S. Role of Hepcidin in the Setting of Hypoferremia during Acute Inflammation. Am J Hematol. 2013; 88(5):E132-E133.

24. Harada N, Kanayama M, Maruyama A, et al.

Nrf2 regulates ferroportin 1-mediated iron efflux and counteracts lipopolysaccharideinduced ferroportin 1 mRNA suppression in macrophages. Arch Biochem Biophys.

2011;508(1):101-109.

25. Nicolas G, Chauvet C, Viatte L, et al. The gene encoding the iron regulatory peptide hepcidin is regulated by anemia, hypoxia, and inflammation. J Clin Invest. 2002; 110(7):1037-1044.

26. Cherian S, Forbes DA, Cook AG, et al. An Insight into the Relationships between Hepcidin, Anemia, Infections and Inflammatory Cytokines in Pediatric Refugees: A Cross-Sectional Study. PloS One. 2008;3(12):e4030.

27. Atkinson SH, Armitage AE, Khandwala S, et al. Combinatorial effects of malaria season, iron deficiency, and inflammation determine plasma hepcidin concentration in African children. Blood. 2014; 123(21):3221-3229.

28. Moretti D, Goede JS, Zeder C, et al. Oral iron supplements increase hepcidin and decrease iron absorption from daily or twice-daily doses in iron-depleted young women. Blood.

2015;126(17):1981-1989.

29. Stoffel NU, Cercamondi CI, Brittenham G, et al. Iron absorption from oral iron supplements given on consecutive versus alternate days and as single morning doses versus twice-daily split dosing in iron-depleted women: two open-label, randomised controlled trials.

Lancet Haematol. 2017; 4(11):e524-e533.

30. Knutson MD, Oukka M, Koss LM, Aydemir F, Wessling-Resnick M. Iron release from macrophages after erythrophagocytosis is upregulated by ferroportin 1 overexpression and down-regulated by hepcidin. Proc Natl Acad Sci U S A. 2005;102(5):1324-1328.

31. Laftah AH, Ramesh B, Simpson RJ, et al. Effect of hepcidin on intestinal iron absorption in mice. Blood. 2004;103(10):3940-3944.

32. Chaston T, Chung B, Mascarenhas M, et al.

Evidence for differential effects of hepcidin in macrophages and intestinal epithelial cells.

Gut. 2008;57(3):374-382.

33. Drakesmith H, Nemeth E, Ganz T. Ironing out Ferroportin. Cell Metab. 2015; 22(5):777-787.

34. Mena NP, Esparza A, Tapia V, Valdes P, Nunez MT. Hepcidin inhibits apical iron uptake in intestinal cells. Am J Physiol Gastrointest Liver Physiol. 2008; 294(1):G192-198.

35. Galy B, Ferring-Appel D, Becker C, et al. Iron regulatory proteins control a mucosal block to intestinal iron absorption. Cell Rep. 2013;3(3):844-857.

36. Taylor M, Qu AJ, Anderson ER, et al.

Hypoxia-Inducible Factor-2 alpha Mediates the Adaptive Increase of Intestinal Ferroportin During Iron Deficiency in Mice.

Gastroenterology. 2011;140(7):2044-2055.

1149

ARTICLE Ferrata Storti Foundation

Haematologica 2019 Volume 104(6):1150-1155

Chronic Myeloid Leukemia

Poor prognosis of chromosome 7 clonal aberrations in Philadelphia-negative metaphases and relevance of potential underlying myelodysplastic features in chronic myeloid leukemia

Audrey Bidet,1 Stéphanie Dulucq,1 Thomas Smol,2,3 Alice Marceau-Renaut,4,5 Stéphane Morisset,6 Valérie Coiteux,7 Marie-Pierre Noël-Walter,7 FranckEmmanuel Nicolini,6,8 Isabelle Tigaud,9 Isabelle Luquet,10 Stéphanie Struski,10 Baptiste Gaillard,11 Dominique Penther,12 Sylvie Tondeur,13 Nathalie Nadal,14 Eric Hermet,15 Lauren Véronèse,16 Delphine Réa,17 Carine Gervais,18 Olivier Theisen,19 Christine Terré,20 Pascale Cony-Makhoul,21 Christine Lefebvre,22 Jean-Baptiste Gaillard,23 Isabelle Radford24 Anne-Laure Vervaeke,1 Carole Barin,25 Elise Chapiro,26 Florence Nguyen-Khac,26 Gabriel Etienne,27 Claude Preudhomme,3,4,5 François Xavier Mahon27 and Catherine Roche-Lestienne2,3,5 on behalf of the Groupe Francophone de Cytogénétique Hématologique (GFCH) and the French Intergroup of Chronic Myeloid Leukemia (Fi-LMC).

Laboratoire d’Hématologie, CHU Bordeaux; 2Institut de Génétique Médicale, Hôpital Jeanne de Flandre, CHU Lille; 3Centre de Recherche Jean-Pierre Aubert, UMR-S 1172, Université de Lille; 4Institut d’Hématologie, Centre de Biologie Pathologie Génétique, CHU Lille; 5Inserm, UMR-S 1172, Lille; 6Département d’Hématologie, Centre Léon Bérard, Lyon; 7 Service des Maladies du Sang, Hôpital Claude Huriez, CHU Lille; 8Inserm U1052, Centre de Recherche en Cancérologie, Centre Léon Bérard, Lyon; 9Laboratoire de Cytogénétique et de Biologie Moléculaire, Service d'Hématologie Biologique - CBPAS, GHS - Hospices Civils de Lyon, Pierre-Bénite Cedex, France; 10Laboratoire d’Hématologie, Plateau Technique Hématologie-Oncologie, Institut Universitaire du Cancer de Tolouse Oncopole; 11 Laboratoire Central d’Hématologie, Hôpital Robert Debré, Reims; 12Laboratoire de Génétique Oncologique, Centre de Lutte Contre le Cancer Henri Becquerel, Rouen; 13 Laboratoire d’Hématologie–Cytogénétique, CHU Saint-Etienne, Hôpital Nord, SaintEtienne Cedex 2; 14Laboratoire de Génétique Chromosomique et Moléculaire, Plateau Technique de Biologie, CHU de Dijon; 15Service d’Hématologie Clinique, CHU Estaing, Clermont-Ferrand; 16Laboratoire de Cytogénétique, CHU Estaing, Clermont-Ferrand; 17 Service Clinique des Maladies du Sang, Hôpital St Louis, Paris; 18Laboratoire Régional de Cytogénétique Hématologique d’Alsace, CHU de Haute Pierre, Strasbourg Cedex; 19 Laboratoire de Cytogénétique Hématologique, Plateau Technique Hôtel Dieu, Nantes; 20 Laboratoire de Cytogénétique du Centre Hospitalier Valence, Le Chesnay; 21Service d’Hématologie, Centre Hospitalier Annecy-Genevois, Epagny Metz-Tessy; 22Unité de Génétique des Hémopathies, Institut de Biologie et Pathologie, CHU Grenoble Alpes, Grenoble Cedex 9; 23Unité de Génétique Médicale et Cytogénétique, CHU de Nîmes; 24 Laboratoire de Cytogénétique, Hôpital Necker – Enfants Malades, Paris; 25Laboratoire de Cytogénétique Onco-Hématologie, Hôpital Bretonneau, Tours; 26Service d’Hématologie Biologique, Groupe Hospitalier Pitié-Salpêtrière, Assistance Publique des Hôpitaux de Paris et Sorbonne Université, Paris and 27Département d’Hématologie, Institut Bergonié, Bordeaux, France.

1

Correspondence: CATHERINE ROCHE-LESTIENNE catherine.roche@chru-lille.fr Received: October 8, 2018.

Accepted: December 18, 2018.

Pre-published: December 20, 2018. doi:10.3324/haematol.2018.208801 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/104/6/1150 ©2019 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

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

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

1150

ABSTRACT

C

lonal chromosome abnormalities in Philadelphia-negative cells could concern chronic myeloid leukemia patients treated by tyrosine kinase inhibitors. The European LeukemiaNet distinguishes -7/del(7q) abnormalities as a “warning”. However, the impact of clonal chromosome abnormalities, and specifically those of -7/del(7q), in Philadelphia-negative cells on clinical outcomes is unclear and based on case-reports showing morphological dysplasia and increased risk of acute myeloid leukemia, suggesting the coexistence of chronic myeloid leukemia and high-risk myelodysplastic syndrome. The aim of this study was to determine whether the impact of -7/del(7q) clonal chromosome abnormalities in Philadelphia-negative cells on the clinical outcome is different from that of other types of abnormalities, and we argue for an underlying associated high-risk myelodysplastic syndrome. Among 102 chronic myeloid haematologica | 2019; 104(6)

Prognosis of -7/del(7q) in Ph- cells and underlying MDS

leukemia patients with clonal chromosome abnormalities in Philadelphia-negative cells with more than a median of 6 years of follow up, patients with -7/del(7q) more frequently had signs of dysplasia, a lower cumulative incidence of deep molecular response and often needed further treatment lines, with the consequent impact on event-free and progression-free survival. Morphological features of dysplasia are associated with myelodysplastic syndrome/acute myeloid leukemia mutations and compromise the optimal response to tyrosine kinase inhibitors, irrespectively of the type of clonal chromosome abnormalities in Philadelphianegative cells. However, mutation patterns determined by next-generation sequencing could not clearly explain the underlying high-risk disease. We hereby confirm the pejorative prognostic value of -7/del(7q) clonal chromosome abnormalities in Philadelphia-negative cells and suggest that myelodysplastic features constitute a warning signal that response to tyrosine kinase inhibitors may be less than optimal.

Introduction Concurrently with the BCR-ABL1 fusion gene resulting from t(9;22)(q34;q11) in Philadelphia-positive (Ph+) cells, clonal chromosome abnormalities (CCA) can be present at the time of diagnosis of chronic myeloid leukemia (CML) or emerge during therapy. CCA in Ph+ cells (CCA/Ph+) are well known; they are associated with clonal cytogenetic evolution and failure of tyrosine kinase inhibitor (TKI) therapy.1 However, CCA may also occur in Philadelphia-negative cells (CCA/Ph-). According to reported series, CCA/Phcould be present in 2% to 17% of CML patients treated with TKI. These differences in frequencies could be partly explained by taking into account (or not) the loss of chromosome Y and transitory abnormalities.2–7 While the frequency of CCA/Ph- varies greatly from study to study, other characteristics seem more reproducible, such as the median age at onset (between 49 and 58 years), the median time of the first appearance during TKI therapy (between 10 and 17 months) and the type of CCA/Ph- which is, according to their frequencies: trisomy 8 (+8), monosomy 7/deletion 7q [-7/del(7q)], loss of the Y chromosome (-Y), deletion 20q (20q) and others. Despite very limited information on the occurrence of this phenomenon among patients treated with second-generation TKI, the incidence and type of abnormalities after nilotinib or dasatinib treatment seem to be similar to those reported in patients after imatinib therapy.3,7,8 Controversies still exist regarding the emergence of CCA/Ph-, not only on the time of appearance (before or after treatment), but also on the potential impact of the type of TKI or high doses of TKI. For Kovitz et al.9 the appearance of CCA was not observed in the Ph- cells prior to therapy with interferon or imatinib. This pleads for a supporting action of the treatment or a selective pressure on progenitor cells. Conversely, HUMARA-polymerase chain reaction and deep mutational screening of Ph- cells could argue for the existence of clonal Ph- hematopoiesis, either before or after CML therapy.10,11 It could then be speculated that a clonal Ph- state could precede the acquisition of BCRABL1, as evidence of a two-step model of CML. Another hypothesis is that Ph+ or Ph- concomitant clones arise from genetically unstable progenitors, with the Ph- clone revealed at the time of TKI-induced remission of CML. Most CCA/Ph- are similar to those commonly associated with myelodysplastic syndromes (MDS) or secondary acute myeloid leukemia (AML), raising many questions about their association with myelodysplasia and the risk of transformation to MDS or AML. In 2013, the European LeukemiaNet (ELN) recommendations,1 -7/del(7q) abnormalities are considered as a “warning” compared to other haematologica | 2019; 104(6)

CCA/Ph-. This is an innovation compared to the 2009 ELN recommendations. This singling out of -7/del(7q) abnormalities was based on some case reports indicating MDS features and an increased risk of AML development among patients with such abnormalities, not present with other CCA/Ph- in the absence of dysplasia.2,9,12–14 Indeed Deninger et al. reported 17 patients treated with imatinib after failure of interferon therapy who had CCA/Ph- and developed MDS; eight of them had -7.2 In a meta-analysis, Groves et al. found that 32% of 16 CCA/Ph- patients with -7 had transformation to MDS/AML.12 However, the prognostic impact of CCA/Ph-, and specifically those of chromosome 7, on clinical outcomes is unclear as only limited studies have had a sufficiently long follow up.4 We, therefore, conducted a large, retrospective French multicenter study of CCA/Ph- CML patients with a prolonged follow up to: (i) evaluate the frequency of abnormalities of chromosome 7 among CCA/Ph-, especially in the era of second-generation TKI; (ii) determine their impact on clinical outcomes; and (iii) assess the existence of an underlying associated highrisk MDS by systematic bone marrow morphological review and next-generation sequencing (NGS) using a MDS/AML panel.

Methods Patients The databases of French institutions were screened, leading to the identification of 102 CML patients with a CCA/Ph- at diagnosis or during the course of treatment with one or more lines of TKI, and 11 MDS patients without a history of CML who had an abnormal karyotype who formed the control group. Among these patients, a chromosome 7 abnormality [-7/del(7q)] was detected in 26 CML cases and in four MDS cases. Patients in blast phase CML at diagnosis were excluded.

The patients’ clinical and biological data were obtained from medical records. The study was performed in accordance with the Declaration of Helsinki, and was approved by a local investigational review board (CPP DC 2015/133).

Procedures Cytogenetic response was determined using standard procedures. Clonality was defined by two or more metaphases presenting the same abnormality, or three metaphases presenting the same monosomy except for -Y because the rate of this monosomy could be age-related. The potential involvement of -Y in the leukemic process was only established for patients with more than 75% –Y metaphases. All karyotypes were reviewed by Groupe Francophone de Cytogénétique Hématologique members and then classified according to the 2016 International System for Human 1151

A. Bidet et al.

Cytogenetic Nomenclature. Molecular monitoring was performed according to the ELN recommendations.1 A morphological central review was used to screen for myelodysplastic features at the time of CCA/Ph- emergence in 48 cases. Morphological dysplasia was considered significant when it was observed for 10% or more cells in any hematopoietic lineage with or without excess of blasts (>5%). Erythroid lineage dysplasia criteria include nuclear and cytoplasmic abnormalities (multinuclearity, laminated cytoplasm, macroerythroblasts). Dysgranulopoiesis also includes hypogranular or hypergranular precursors and/or neutrophils and/or lack of nuclear segmentation. Micromegakaryocytes, multinuclear or hypolobulated megakaryocytes were the main abnormalities observed in the megakaryocyte lineage. Patients were stratified according to the presence or absence of chromosome 7 abnormalities, whether isolated or not, leading to -7/del(7q) CCA/Ph- identified by conventional cytogenetics. In the case of CCA/Ph- detection after the diagnosis of CML, time of emergence was retrospectively evaluated on prior samples by fluorescence in situ hybridization when possible. Complex karyotypes (≥3 anomalies) affected only five of the 102 patients; since this precluded statistically meaningful analyses, these abnormalities were not considered as a separate category. Underlying MDS was documented both by centralized morphological analysis of bone marrow smears and by NGS for a targeted panel of 27 genes frequently altered in MDS and AML (ASXL1, CBL, CEBPA, DNMT3A, ETV6, EZH2, FLT3, IDH1, IDH2, JAK2, KIT, KRAS, MPL, NPM1, NRAS, PHF6, PTPN11, RIT1, RUNX1, SETBP1, SF3B1, SRSF2, TET2, TP53, U2AF1, WT1, ZRSR2). NGS data obtained for 45 CML patients were compared with a MDS control group of patients with (n=4) or without (n=7) cytogenetic chromosome 7 anomalies. NGS was performed as previously described15 on follow-up bone marrow samples at complete cytogenetic response and preferably at best molecular response, to limit the risk of residual Ph+ clone contamination.

The quality of response to treatment was classified by the best cytogenetic or molecular response obtained at any point after one, two, three or more lines of treatments into no cytogenetic response, partial cytogenetic response, complete cytogenetic response, major molecular response and deep molecular response ≤0.0032% (MR4.5). The response criteria were according to the standard ELN definitions.1

Statistical analysis Bivariate analyses were performed to compare -7/del(7q versus other CCA/Ph-. Quantitative variables were described by their mean and standard deviation in both groups and compared by a Student t test if the distribution was normal or with their median, range and quartiles along with a non-parametric Mann-Whitney test when the distribution was not normal. Cumulative incidences from diagnosis or initiation of first-line TKI treatment until the achievement of the considered response were calculated with death without response as a competing risk. Gray tests were performed to compare cumulative incidence curves. Analyses from time of CCA/Ph- emergence were not performed because detection time may be delayed and underestimated for some patients depending on the clinical context and requirement of additional cytogenetic controls despite complete cytogenetic response.

However, the effect of type of CCA/Ph- on response, progression, or late events was considered by landmark analyses 3 years after the initiation of TKI therapy. Probabilities of overall survival, progression-free survival and event-free survival since initiation of first-line TKI therapy were calculated and illustrated by the Kaplan-Meier method: until death at any time and for any reason (overall survival); until death or progression to accelerated phase or blastic transformation (progression-free survival); and until intoler1152

ance, loss of response, resistance, treatment switch, progression, or death (event-free survival). Survival curve comparisons, including landmark analyses, were performed using log-rank tests. The level of statistical significance was set at 5%. All the statistical analyses were performed with R V3.2.3.

Results Clinical and cytogenetic presentation of patients with chronic myeloid leukemia The patients’ baseline characteristics are summarized in Online Supplementary Table S1). Twenty-six out of the 102 (25.5%) patients had CCA/Ph- affecting chromosome 7, which was an isolated abnormality in 13 out of the 26 cases (50%) and associated with +8 in 19.5% cases. In one patient with -7 CCA/Ph-, +8 was detected in a separate clone. Among other CCA/Ph- cases, +8 was the most frequent other abnormality, affecting 31 out of the 46 patients (67.4%); 13.7% (n=14) patients had -Y CCA/Ph-. Patients who developed -7/del(7q) CCA/Ph- were significantly younger (mean 48 vs. 55 years old; P=0.035) and mostly benefited from second-generation TKI as first-line treatment (mainly dasatinib: 15.4% vs. 2.6% cases in the other CCA/Ph group; P=0.027). Other baseline characteristics were similar between the -7/del(7q) CCA/Ph- cases and the other CCA/Ph- cases. The median follow up from the diagnosis of CML was 6.47 (range, 1-19) years. The median time of CCA/Ph- detection after starting TKI therapy was 2.08 (range, -0.8 to 12.65) years for -7/del(7q) CCA/Ph- and 1.02 (range, -3.32 to 11.02) years for other CCA/Ph-. The -7 CCA/Ph- was present at diagnosis in only one case, as determined by retrospective fluorescence in situ hybridization analysis. Twelve out of the 26 patients who developed 7/del(7q) CCA/Ph- were in complete cytogenetic response at the time the abnormality was detected. The median follow up since CCA/Ph- detection was 5.35 (range, 1-14) years for -7/del(7q) and 7.41 (range, 0-15) years for other CCA/Ph- (P≥0.05). The trend approached statistical significance for the number of treatment lines required after the first 3 years of follow up, as patients with -7/del(7q) CCA/Ph- more frequently needed two or more treatment lines (30% vs. 13.10% of patients with other CCA/Ph-; P=0.049, chi-square).

Biological analysis of underlying associated myelodysplastic syndrome The presence of MDS features at the time of CCA/Ph-, determined by morphological analysis of the bone marrow, was evaluated in 48 patients and appeared to alter the quality of response to treatment, as only 9.4% of patients with MDS signs (24 patients) reached a major molecular response or better versus 47% of patients without MDS signs (P<0.0001, chi-square). Morphological MDS features were more frequent among the -7/del7q CCA/Ph- group [50% of patients with -7/del(7q) CCA/Phvs. 20% with other CCA/Ph-], and only 33% (4 out of 12) of the -7/del(7q) patients with MDS signs achieved a major molecular response or better compared to 75% (9 out of 12) of these patients without MDS signs. For patients with other types of CCA/Ph-, only six out of 49 patients (18.8%) with an optimal response had morphological features of MDS (P<0.00001, chi-square). NGS was performed for 45 CML patients with available DNA at the time of detection of CCA/Ph- and 11 MDS patients as a haematologica | 2019; 104(6)

Prognosis of -7/del(7q) in Ph- cells and underlying MDS

Figure 1. Landscape of mutations detected by next-generation sequencing in patients with chronic myeloid leukemia and myelodysplastic syndromes. Columns (patients) are first sorted by the type of clonal chromosome abnormalities in Philadelphia chromosome-negative cells (CCA/Ph-), then by disease [chronic myeloid leukemia in green, myelodysplastic syndrome (MDS) in yellow, advanced disease in orange) and by the presence of morphological features of MDS (in blue characters) from the youngest to the oldest. Pink cells indicate somatic mutations that were detected.

control group (Figure 1). At least one mutation was identified in 90% (10 out of 11 patients) of the MDS group versus 37% (16 out of 43 patients) of the CCA/Ph- CML group. The proportion of patients with mutations in the MDS group was consistent with that reported in the literature16 and the MDS/AML mutations detected were significantly associated with the MDS group versus CCA/PhCML patients (P<0.0011, chi-square). In CML cases, 13 out of the 24 patients with morphological signs of MDS had DNA available for NGS. No significant association was found between NGS-detected mutations and patientsâ&#x20AC;&#x2122; age or type of CCA/Ph- but MDS morphological features were significantly associated with the presence of mutations as nine (68.75%) of the patients with morphological signs of MDS had mutations versus four (31.25%) without signs (P<0.00001, chi-square). This result supports the presence of an underlying MDS disease in CML patients with observed morphological signs of dysplasia.

Among the 12 patients with -7/del(7q), four had morphological signs of MDS and three had mutations. Two of them had some mutations in a gene known to be implicated in disease progression: one patient had an undescribed EZH2 c.826C>T:p.Gln276* alteration with a low variant frequency (1%), and the other had an isolated TP53 c.709A>C:p.Met273Leu mutation (variant frequency, 71%) but reached MR4.5 after first-line TKI treatment with dasatinib. The other two have an isolated mutation of ASXL1 or STAG2. Of note, EZH2 on chromosome 7 was also mutated in two other patients with +8 CCA/Ph: one with EZH2 c.2079T>A: p.Asn693Lys (variant frequency, 17%) with a poor response to second-line TKI and the other with EZH2 c.826C>T:p.Gln276* (variant frequency, 1%) with a poor response after third-line TKI.

Outcomes and survival The cumulative incidence of MR4.5 is illustrated in Figure 2. The median time to MR4.5 after starting TKI therapy haematologica | 2019; 104(6)

A

B

Figure 2. Cumulative incidence of deep molecular response following the first line of tyrosine kinase inhibitor therapy. (A) Cumulative incidence of deep molecular response (MR4.5) following initiation of the first line of tyrosine kinase inhibitor (TKI) therapy according to the type of clonal chromosome abnormalities in Philadelphia chromosome-negative cells (CCA/Ph-). (B) Cumulative incidence of MR4.5 according to the type of CCA/Ph- by landmark analysis 36 months after initiation of the first line of TKI therapy.

1153

A. Bidet et al.

was 2.3 (range, 0.66-10.5) years for -7/del(7q) CCA/Phpatients versus 3.55 (range, 0.5-11.77) years for patients with other CCA/Ph- (P=NS, exact Mann-Whitney test) (Figure 2A). However, landmark analysis after 3 years of therapy revealed the adverse effect of -7/del(7q) CCA/Ph- (P=0.04) (Figure 2B) on the cumulative incidence of MR4.5. The type of CCA/Ph- did not have an impact on overall survival (P=0.717; data not shown). The overall rates of progressionfree survival at 3, 5 and 10 years were 94.74% [95% confidence interval (95% CI): 90.35-99.35], 86.28% (95% CI: 78.94-94.30), and 78.60% (95% CI: 69.12-89.39), respectively. The overall rates of event-free survival at 3, 5 and 10 years were 67.36% (95% CI: 58.63-77.39), 56.32% (95% CI: 46.72-67.90), and 43.61% (95% CI: 33.34-57.06), respectively (data not shown). Landmark analyses at 3 years after starting first TKI therapy revealed a strong negative impact of -7/del(7q) CCA/Ph- on both event-free and progression-free survival (P=0.015 and P=0.022, respectively, log-rank test), as shown in Figure 3.

A

B

Discussion To our knowledge, this is the largest study of CML patients with CCA/Ph-, with 26 (25.4%) out of 102 patients having -7/del(7q) abnormalities in a consistent, long follow up [median 6.47 (range, 1-11) years]. Our cohort required particular care as 37% of them needed at least two switches of TKI prior to achieving their optimal response. Patients who developed -7/del(7q) CCA/Ph- were significantly younger (means 48 years old) and mostly benefited from second- and/or third-generation TKI, with a significantly lower cumulative incidence of MR4.5 at the landmark analysis 3 years after starting TKI therapy (P=0.04) than other patients. Almost half of the patients (46%) with 7/del(7q) CCA/Ph- were in complete cytogenetic remission at the time of detection of the CCA/Ph-. Because the quality of cytogenetic responses significantly affects the prognosis, the high proportion of good cytogenetic responders in this group strengthens the involvement of chromosome 7 abnormalities in our prognostic results. Furthermore, additional cytogenetic abnormalities in Ph+ cells (which are considered as clonal evolution with greater risk of transformation) cannot be suspected to be involved in prognosis in our study because only one out of 102 patients had an atypical t(7;14)(p21;q12) translocation in Ph+ cells at diagnosis.

Associated dysplastic features, identified by morphological analysis of bone marrow, seemed to be a more frequent event in the -7/del(7q) group, and had a significant negative impact on quality of response and outcomes as compared to that in other types of CCA/Ph- cases. In addition, irrespectively of the type of CCA/Ph-, patients with morphological MDS signs had significantly lower rates of major molecular response (P<0.0001). Likewise, 78% of the patients who did not have a complete cytogenetic response were those who had morphological features of MDS. This is in accordance with the findings of a preceding study by Deininger et al.2 who reported that patients with isolated CCA/Ph- without morphological evidence of dysplasia do not require special management. In a study by TerrĂŠ et al., CCA/Ph- were not associated with myelodysplasia and did not impair the cytogenetic response to imatinib.6 However, although this was the largest published series of CCA/Phcases to date, only two out of 28 evaluated patients had dysgranulopoiesis, the median follow up was short (only 24 1154

Figure 3. Landmark survival analyses according to the type of clonal chromosome abnormalities in Philadelphia chromosome-negative cells 36 months after initiation of the first line of tyrosine kinase inhibitor therapy. (A) Eventfree survival. (B) Progression-free survival. TKI: tyrosine kinase inhibitor.

months after imatinib initiation), and the majority of patients in this study had been previously treated.

Even though the overall survival rates were not statistically significantly different between the two groups, landmark analyses of the impact of -7/del(7q) CCA/Ph- after 3 years of starting TKI treatment revealed lower event-free and progression-free survival rates (P=0.015 and P=0.022, respectively). Recently, Issa et al.4 reported the prognostic relevance of CCA/Ph- in 58 patients with a median follow-up of 7.6 years. Among them, only four patients had -7/del(7q) type abnormalities. The disease progressed in three of these four patients, transforming to either MDS or CML blast crisis, without an obvious link with MDS signs in this small cohort. Excluding â&#x20AC;&#x201C;Y CCA/Ph-, a decreased survival was observed for patients with CCA/Ph- and seemed to be related to MDS or CML transformation. A specific -7/del(7q)associated risk could not be determined in their study. In another recent study, Wasilewska et al. found that -7 and +8 CCA/Ph- did not have an impact on the long-term outcome of CML patients, but survival was not investigated in this study and the series was limited to five observations.17 The occurrence of Ph+ and CCA/Ph- in the same patient could suggest the presence of two concurrent, distinct hematologic disorders: CML and MDS. To investigate the biological features of such underlying diseases, we screened for mutations in frequently altered genes in MDS or AML

by NGS at the time of best response to therapy. Mutations were detected more frequently in our MDS control group of patients than in the CML group, and were also mostly found haematologica | 2019; 104(6)

Prognosis of -7/del(7q) in Ph- cells and underlying MDS

in CCA/Ph- CML patients with features of MDS (P=0.006).

Mutations were not associated with age and there was no preferentially significant association with -7/del(7q), despite a higher rate of MDS features in this subgroup. As already reported in CML,18–20 ASXL1 is one of the most frequently mutated genes. Although somatic mutation or decreased expression of EZH2 at 7q36.1 plays a role in cancer, the molecular mechanisms responsible for the poor prognosis of chromosome 7 alterations in hematologic diseases remains undefined. In this study, three patients had EZH2 mutations.

One of them had -7 CCA/Ph-, thus suggesting a bi-allelic alteration. This patient further progressed to an advanced phase of CML, but the direct role of the EZH2 mutation in this case could not be determined because the frequency of the variant was low (1%) and it was associated with other mutations in ASXL1, CBL, SETBP1 and SRSF2. The two other patients with EZH2 mutations had +8 CCA/Ph-, associated in one case with an ASLX1 mutation, and neither of them had further disease evolution. It has been described that NGS screening of CML patients in major molecular response revealed a higher rate of other gene mutations in patients with CCA/Ph- than in patients without CCA/Ph-.21 These latter mutational events may therefore be independent, and could result from genetic instability.

In conclusion, in this study we were able to demonstrate the negative prognostic impact of -7/del(7q) CCA/Ph- on the cumulative incidences of MR4.5, event-free survival and

References 8.

1. Baccarani M, Deininger MW, Rosti G, et al.

European LeukemiaNet recommendations for the management of chronic myeloid leukemia: 2013. Blood. 2013;122(6):872– 884.

2. Deininger MWN, Cortes J, Paquette R, et al. The prognosis for patients with chronic myeloid leukemia who have clonal cytogenetic abnormalities in Philadelphia chromosome-negative cells.

Cancer.

2007;110(7): 1509–1519.

3. De Melo VAS, Milojkovic D, Khorashad JS, et al. Philadelphia-negative clonal hematopoiesis is a significant feature of dasatinib therapy for chronic myeloid leukemia. Blood. 2007;110(8):3086–3087.

4. Issa GC, Kantarjian HM, Gonzalez GN, et al. Clonal chromosomal abnormalities appearing in Philadelphia chromosomenegative metaphases during CML treatment. Blood. 2017;130(19):2084–2091.

5. Lee S-E, Choi SY, Bang J-H, et al. The longterm clinical implications of clonal chromosomal abnormalities in newly diagnosed chronic phase chronic myeloid leukemia patients treated with imatinib mesylate.

Cancer Genet.

2012;205(11):563–571.

6. Terre C, Eclache V, Rousselot P, et al.

Report of 34 patients with clonal chromosomal abnormalities in Philadelphia-negative cells during imatinib treatment of Philadelphia-positive chronic myeloid leukemia. Leukemia. 2004;18(8):1340– 1346.

7. Wang H, Jin J, Wang Y, Huang X, Huang J. Clonal chromosomal abnormalities in Philadelphia-negative cells in chronic myeloid leukemia patients treated with

haematologica | 2019; 104(6)

9.

10.

11.

12.

13.

14.

progression-free survival in a cohort of 102 CML patients with CCA/Ph- with a median follow up of more than 6 years. Morphological features of MDS represent a negative prognostic factor for optimal response to treatment, irrespectively of the type of CCA/Ph-, even though they are more frequent in the -7/del(7q) CCA/Ph- subgroup. NGS screening for MDS/AML mutations failed to clearly explain the underlying high risk of myelodysplastic disease in the -7/del(7q) CCA/Ph- group, but the presence of mutations in genes of our MDS/AML panel was statistically associated with morphological features of MDS on bone marrow smears. Further analyses will help to determine whether morphological signs of MDS in the bone marrow and/or MDS mutations at the time of emergence of CCA/Ph- represent a warning signal in these particular CML patients.

Funding This work was funded with the support of the Fi-LMC group, which received partial contributions from Bristol Myers Squibb, Novartis and Incyte Pharma.

Acknowledgments The authors would like to thank Laurence Meyer for her morphological bone marrow partial review, and Olivier Nibourel, Jean-Michel Cayuela, and Sandrine Hayette for their BCR-ABL1 molecular monitoring data.

nilotinib used in first-line therapy. Ann Hematol. 2013;92(12):1625–1632.

Baldazzi C, Luatti S, Marzocchi G, et al.

Emergence of clonal chromosomal abnormalities in Philadelphia negative hematopoiesis in chronic myeloid leukemia patients treated with nilotinib after failure of imatinib therapy. Leuk Res. 2009;33 (12):e218-220.

Kovitz C, Kantarjian H, Garcia-Manero G, Abruzzo LV, Cortes J. Myelodysplastic syndromes and acute leukemia developing after imatinib mesylate therapy for chronic myeloid leukemia. Blood. 2006;108(8): 2811–2813.

Paquette RL, Nicoll J, Chalukya M, et al.

Clonal hematopoiesis in Philadelphia chromosome-negative bone marrow cells of chronic myeloid leukemia patients receiving dasatinib. Leuk Res. 2010;34(6):708– 713.

Schmidt M, Rinke J, Schäfer V, et al.

Molecular-defined clonal evolution in patients with chronic myeloid leukemia independent of the BCR-ABL status.

Leukemia. 2014;28(12):2292–2299.

Groves MJ, Sales M, Baker L, Griffiths M, Pratt N, Tauro S. Factors influencing a second myeloid malignancy in patients with Philadelphia-negative -7 or del(7q) clones during tyrosine kinase inhibitor therapy for chronic myeloid leukemia. Cancer Genet. 2011;204(1):39–44.

Jabbour E, Cortes JE, Kantarjian HM. Suboptimal response to or failure of imatinib treatment for chronic myeloid leukemia: what is the optimal strategy?

Mayo Clin Proc. 2009;84(2):161–169.

Zeidan A, Kakati S, Anderson B, Barcos M, Wetzler M. Monosomy 7 in t(9;22)-negative cells during nilotinib therapy in an imatinib-resistant chronic myeloid

15.

16.

17.

18.

19.

20.

21.

leukemia case. Cancer Genet Cytogenet.

2007;176(2):169–171.

Renneville A, Attias P, Thomas X, et al.

Genetic analysis of therapy-related myeloid neoplasms occurring after intensive treatment for acute promyelocytic leukemia.

Leukemia. 2018;32(9):2066–2069.

Papaemmanuil E, Gerstung M, Malcovati L, et al. Clinical and biological implications of driver mutations in myelodysplastic syndromes. Blood. 2013;122(22):3616– 3627; quiz 3699.

Wasilewska EM, Panasiuk B, Gniot M, et al. Clonal chromosomal aberrations in Philadelphia negative cells such as monosomy 7 and trisomy 8 may persist for years with no impact on the long term outcome in patients with chronic myeloid leukemia.

Cancer Genet. 2017;216–217:1–9.

Boultwood J, Perry J, Zaman R, et al. Highdensity single nucleotide polymorphism array analysis and ASXL1 gene mutation screening in chronic myeloid leukemia during disease progression. Leukemia.

2010;24(6):1139–1145.

Roche-Lestienne C, Marceau A, Labis E, et al. Mutation analysis of TET2, IDH1, IDH2 and ASXL1 in chronic myeloid leukemia.

Leukemia. 2011;25(10):1661–1664.

Togasaki E, Takeda J, Yoshida K, et al.

Frequent somatic mutations in epigenetic regulators in newly diagnosed chronic myeloid leukemia. Blood Cancer J. 2017;7 (4):e559.

Schnittger S, Meggendorfer M, Nadarajah N, et al. In CML patients with good response to TKIs other gene mutations are frequently (37%) present in addition to Philadelphia negative, cytogenetically aberrant clones but are rare (4%) in cases with MMR and normal karyotype. Blood.

2014;124(21):3126.

1155

ARTICLE Ferrata Storti Foundation

Haematologica 2018 Volume 104(6):1156-1167

Acute Myeloid Leukemia

Lysine specific demethylase 1 inactivation enhances differentiation and promotes cytotoxic response when combined with all-trans retinoic acid in acute myeloid leukemia across subtypes

Kimberly N. Smitheman,1 Tesa M. Severson,2 Satyajit R. Rajapurkar,1 Michael T. McCabe,1 Natalie Karpinich,1 James Foley,1 Melissa B. Pappalardi,1 Ashley Hughes,3 Wendy Halsey,3 Elizabeth Thomas,3 Christopher Traini,3 Kelly E. Federowicz,1 Jenny Laraio,1 Fredrick Mobegi,2 Geraldine Ferron-Brady,4 Rabinder K. Prinjha,1 Christopher L. Carpenter,1 Ryan G. Kruger,1 Lodewyk Wessels2,5 and Helai P. Mohammad1

Epigenetics Discovery Performance Unit, Oncology R&D, GlaxoSmithKline, Collegeville, PA, USA; 2Division of Molecular Carcinogenesis, Oncode Institute, the Netherlands Cancer Institute, Amsterdam, the Netherlands; 3Target Sciences, GlaxoSmithKline, Collegeville, PA, USA; 4Clinical Pharmacology and Modeling Sciences, GlaxoSmithKline, Collegeville, PA, USA and 5Faculty of EEMCS, Delft University of Technology, the Netherlands 1

ABSTRACT

L

Correspondence: HELAI P.MOHAMMAD helai.x.mohammad@gsk.com Received: June 4, 2018.

Accepted: November 30, 2018.

Pre-published: December 4, 2018.

ysine specific demethylase 1 (LSD1) is a histone modifying enzyme that suppresses gene expression through demethylation of lysine 4 on histone H3. The anti-tumor activity of GSK2879552 and GSK-LSD1, potent, selective irreversible inactivators of LSD1, has previously been described. Inhibition of LSD1 results in a cytostatic growth inhibitory effect in a range of acute myeloid leukemia cell lines. To enhance the therapeutic potential of LSD1 inhibition in this disease setting, a combination of LSD1 inhibition and all-trans retinoic acid was explored. All-trans retinoic acid is currently approved for use in acute promyelocytic leukemia in which it promotes differentiation of abnormal blast cells into normal white blood cells.

Combined treatment with all-trans retinoic acid and GSK2879552 results in synergistic effects on cell proliferation, markers of differentiation, and, most importantly, cytotoxicity. Ultimately the combination potential for LSD1 inhibition and ATRA will require validation in acute myeloid leukemia patients, and clinical studies to assess this are currently underway.

doi:10.3324/haematol.2018.199190

Introduction Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/104/6/1156 Š2019 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

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

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

1156

Acute myelocytic leukemia (AML) is characterized by excessive growth of hematopoietic progenitor cells that reach varying stages of differentiation depending on the subtype. With the exception of acute promyelocytic leukemia (APL) few patients with AML are cured, despite treatment that includes high-dose induction and consolidation therapy and even, for some, bone marrow transplant.1 The disease is classified using the French-American-British (FAB) classification that divides AML into eight subtypes (M0 to M7) based on the differentiation status of the tumor cells as well as the cell type from which the cancer arises. The World Health Organization (WHO) further distinguishes AML types by also considering somatic genetic alterations.2 For most subtypes, first-line treatment consists of chemotherapy followed, in some instances, with hematopoietic stem cell transplant (HSCT).3 Due to the intensity of HSCT treatment, this approach is often only recommended for younger patients or those deemed fit enough to tolerate it.

Even among the younger patient population, the 5-year overall survival is only approximately 40%.3 For patients over the age of 60, only approximately 20% survive;4 therefore, more effective second-line treatment options are needed.

Lysine specific demethylase 1 (LSD1) is a histone-modifying enzyme that is a member of the monoamine oxidase family.5 LSD1 has been shown to suppress haematologica | 2019; 104(6)

Combined LSD1 and ATRA in AML

gene expression through demethylation of mono and dimethyl groups present on lysine 4 of histone H3.6 LSD1 is a critical regulator of hematopoiesis, in part, through interaction with the transcription factors GFI-1 and GFI1b. This LSD1-containing complex regulates expression of key myeloid differentiation genes and ultimately controls hematopoietic progenitor cell differentiation.7 LSD1 is frequently over-expressed in human cancers, including AML, and knockdown of LSD1 has been shown to inhibit the growth of AML cells.1,8-10 These data have spurred interest in LSD1 as a potential target for treatment of AML. As previously reported, potent, selective, irreversible inactivators of LSD1 have been developed, and among the cancer cell lines evaluated, these show selective anti-proliferative activity in SCLC and AML cell lines.9,11-13 Preclinical data such as these have led to the clinical development of LSD1 inhibitors in relapsed, refractory AML patients.

To build upon the therapeutic potential of LSD1 inhibition in AML, rational combination hypotheses and combinations with standard of care agents were considered.

All-trans retinoic acid (ATRA) is used clinically to treat acute promyelocytic leukemia (APL), a subtype of AML, and has been shown to be hugely successful, achieving curative effects in this disease subtype.14 ATRA triggers the transcription factor retinoic acid receptor alpha (RARα) to bind to retinoic acid response elements found in the genome and initiate transcription of target genes, including those important for differentiation.15 APL is characterized by a PML-RARα fusion that inactivates RARα by preventing it from its normal binding and thus locking the tumor in an undifferentiated state. ATRA degrades this fusion, allowing RARα to activate its target genes, leading to differentiation and apoptosis of the cancer cells.16,17 Many clinical trials have attempted to extend the use of ATRA into non-APL AML, but unfortunately these have demonstrated very little success.18 Since the discovery of LSD1 and the characterization of its role in hematopoiesis, there has been speculation as to the possibility of combining an inhibitor of LSD1 with ATRA. One report demonstrated that combination of ATRA with knockdown of LSD1 or tranylcypromine, a nonselective monoamine oxidase inhibitor with weak LSD1 inhibitory activity, leads to transcriptional activation of many RAR target genes that normally lack methylation of H3K4me2 at their promoters.19,20 This combination also had more robust anti-leukemic activity than either treatment alone in the model tested.19 The current report demonstrates the synergistic activity of a combination of a selective, potent inhibitor of LSD1, GSK2879552, with ATRA, and characterizes the mechanism associated with this combination. As a single agent, LSD1 inhibition promotes differentiation of AML cell lines and synergistic differentiation activity is observed when used in combination with ATRA across AML subtypes. This combination also enhances LSD1 inhibitormediated growth inhibition of AML cell lines and primary patient samples. Most importantly, treatment of AML cell lines with an LSD1 inhibitor and ATRA results in synergistic cytotoxicity and caspase-mediated cell death. Collectively, these data suggest that this combination may result in greater efficacy and may be more likely to translate to response in AML patients than LSD1 inhibitor monotherapy. Moreover, this combination may expand the therapeutic potential of ATRA, a proven therhaematologica | 2019; 104(6)

apy, to non-APL AML patients. Clinical studies are currently underway to test this hypothesis in relapsed, refractory AML patients (clinicaltrials.gov identifiers: 02273102, 02717884, 02842827).

Methods Cell lines and human biological samples Cell lines were obtained from the American Type Culture Collection (ATCC) or the Deutsche Sammlung von Mikroorganismen und Zellbulturen (DSMZ). Human biological samples were sourced ethically and informed consent was obtained for their use in research. The use of human tissue samples was reviewed and approved by the GSK Research & Development Compliance (RDC) Human Biological Sample Use Committee.

Compounds GSK2879552 and GSK-LSD1 (GlaxoSmithKline), ATRA (Sigma), and Bortezomib (Chemietek) were prepared in 100% dimethyl sulfoxide (DMSO) (Sigma).

Cell proliferation Proliferation assays were conducted as previously described.11 The growth-death index (GDI) value was determined as the percentage of cells relative to end-of-assay vehicle and assay start, where the number of cells in the vehicle is set at 100% and the number of cells at the time of compound addition (T0) is set to 0%.

A minimum of two biological replicates were evaluated for each assay. Bliss Independence analysis was performed, and a synergy score was determined.9,21

Mechanistic and phenotypic assays A minimum of two biological replicates were evaluated for each assay. Dose response curves were generated using a 4-parameter equation (XLfit, IDBS). All kits were used according to the manufacturer’s recommendations. Cell cycle phase distribution was determined by flow cytometry using the Cycletest™ PLUS DNA Reagent Kit (BD Biosciences) and data were analyzed using FlowJo software (TreeStar Inc.). Proliferation was assessed using BrdU Cell Proliferation Assay Kit (Cell Signaling Technology) with a 4-hour pulse. Values were expressed as percent of vehicle.

Caspase activity was measured using Caspase-Glo 3/7 (Promega).

Luminescence values were normalized to CTG (Promega) for cell number. Peak activation was determined for MOLM-13, OCIAML3, MV-4-11, THP-1, SIG-M5, HL-60, and Kasumi-1 to be days 4, 6, 6, 3, 4, 5, and 5, respectively. Values were expressed as a fold change relative to vehicle. Gene expression was evaluated using real time-quantitative polymerase chain reaction (RT-qPCR), as previously described.11 Superoxide anion production was measured using LumiMax Superoxide Anion Detection Kit (Agilent Technologies). Values were expressed as fold increase over vehicle. Morphology was visually assessed after staining with MayGrunwald and Giemsa (Sigma).

Flow cytometry Cells were stained with surface marker antibodies (CD11b, 555388 or 557754; CD86, 555657; CD71, 555537) and 7-AAD obtained from BD Biosciences. Percent positive for each marker was determined relative to isotype control from cells that were 7AAD negative. Additionally, cells were stained with the above cocktail including annexin V (BD Biosciences) to evaluate Annexin V positivity. A minimum of two biological replicates were evaluated for each assay.

1157

K.N. Smitheman et al.

A

B

C

D

E

F

G

H

I

Figure 1. Inactivation of LSD1 inhibits the growth of acute myeloid leukemia (AML) cells by releasing myeloid differentiation block. (A) Percent maximum inhibition (bars) ± Standard Error and EC50 (circles) for 20 human AML cell lines treated with a titration of GSK2879552 for ten days. (B) Dose response of incorporation by BrdU Cell Proliferation Assay Kit (Cell Signaling Technology) of MOLM-13 cells treated with a titration of GSK-LSD1 for six days. (C) Fold caspase 3/7 activation (± Standard Error) relative to vehicle control of AML cell lines treated with 1000 nM GSK2879552 for 1-6 days or 100 nM bortezomib for three days. (D) Dose response of ITGAM (CD11b, left) and CD86 (right) gene expression on MOLM-13 cells treated with a titration of GSK2879552 for one day. (E) Representative flow cytometric histograms of CD11b protein expression on THP-1 cells (left) and CD86 protein expression on MOLM-13 cells (right) treated with GSK2879552 for one day. Concentrations of GSK2879552 from top to bottom are 0, 7, 47, and 3000 nM. (F) The change (treatment-vehicle) in the percentage of cells positive for CD86 (blue) and CD11b (red) protein expression (± Standard Error) for ten AML cell lines treated with 500 nM GSK2879552 (or 1000 nM GSK2879552*) for three days. † Significance (P<0.05) between dimethyl sulfoxide (DMSO) and GSK552 treatment. (G) Peak superoxide anion production (± Standard Error) of four AML cell lines treated with 1 μM GSK2879552 for 4, 7, 11, and 14 days. (H) Peak superoxide anion production of SKM-1 cells treated with a titration of GSK2879552 for seven days. (I) Dose responses of AML blast colony formation of 14 patient samples treated with a titration of GSK2879552. APL: acute promyelocytic leukemia.

1158

haematologica | 2019; 104(6)

Combined LSD1 and ATRA in AML

Results

Primary acute myeloid leukemia blast colony Clonogenic progenitors from primary AML patients' samples were assessed in a semi-solid methylcellulose-based medium (ColonyGel™, ReachBio) containing recombinant human (rh) SCF (50 ng/mL), rhIL-3 (10 ng/mL), rhGM-CSF (10 ng/mL), and rhEpo (3 U/mL). Each concentration of compound was tested with three technical replicates. Values were expressed as percent of vehicle and graphed as dose response curves (XLfit, IDBS). All known patients' sample characteristics are provided in Online Supplementary Table S1A.

RNA-sequencing Methods for RNA-sequencing can be found in the Online Supplementary Appendix.

Lysine specific demethylase 1 inhibition reduces growth and promotes differentiation of acute myeloid leukemia cells To investigate the role of LSD1 activity on the growth of AML cells, AML cell lines were treated with LSD1 inhibitors. As shown previously, GSK2879552 and a structurally similar compound, GSK-LSD1, are both potent, selective inactivators of LSD1.11 GSK2879552 was analyzed for its ability to inhibit tumor cell growth in a 10day proliferation assay using CellTiter-Glo (Figure 1A).

The concentration required to achieve half of the maximal observed effect (EC50) as well as the maximum percent

A

B

C

D

Figure 2. Sensitivity of acute myeloid leukemia (AML) cells to GSK2879552 increases with addition of all-trans retinoic acid (ATRA). (A) Dose response curves of MOLM-13 (left) and OCI-AML3 (right) cells treated with a titration of GSK2879552 ± 1, 10, 100, or 1000 nM ATRA for four days. Values were graphed relative to the appropriate ATRA alone concentration for each curve. (B) Percent maximum inhibition (blue bars) ± Standard Error and IC50 values (red circles) ± Standard Error for MOLM-13 (left) and OCI-AML3 (right) cells from (A). *Significance (P<0.05) between dimethyl sulfoxide (DMSO) and ATRA treatments for maximum inhibition; †significance (P<0.05) between dimethyl sulfoxide (DMSO) and ATRA treatments for IC50. (C) Synergy scores (± Standard Error) for seven AML cell lines treated with a titration of GSK2879552 ± 1, 10, 100, or 1000 nM ATRA for four days. (D) Synergy scores (± Standard Error) for seven AML cell lines treated with a titration of GSK2879552 ± 1, 10, 100, or 1000 nM ATRA for six days.

haematologica | 2019; 104(6)

1159

K.N. Smitheman et al.

A

C

B

D

E

Figure 3. RNA sequencing reveals differentiation gene signature with combination (COMBO). (A) Venn diagram depicting the overlap of significantly differentially expressed genes in the specific treatments [GSK2879552 + all-trans retinoic acid (ATRA) combination relative to dimethyl sulfoxide (DMSO) (light blue), ATRA relative to DMSO (yellow), and GSK2879552 relative to DMSO (blue)] for MOLM-13 (top) and OCI-AML3 (bottom) cell lines at day 2 (left) and day 4 (right). (B) Corresponding gene expression heatmaps depicting the union of genes from day 2 and day 4 Venn diagrams (from A) for MOLM-13 (top) and OCI-AML3 (bottom) cell lines. Values are row-mean centered, log2 normalized, gene expression data shown high (yellow) to low (blue). (C) Venn diagrams depicting the overlap of significantly differentially expressed genes for the GSK2879552 + ATRA combination relative to DMSO in each of the four cell lines, OCIAML-3, MOLM-13, MV-4-11, and THP-1. Data for different time points are shown, day 2 (top, left), day 4 (top, right), and union of day 2 and day 4 (bottom). Red value indicates overlap of significant genes in all four cell lines for both days. (D) Gene expression heatmaps of overlapping combination versus DMSO genes in intersect (from C) for day 2 (top) and day 4 (bottom). Values are row-mean centered, log2 normalized, gene expression data shown high (yellow) to low (blue). (E) Barplots showing the top 10 KEGG pathways found to be significantly enriched from genes in (C) in intersect (day 2, day 4 and union of day 2 and day 4 tested separately). Bars show the â&#x20AC;&#x201C;log((False Discovery Rate P-value) on the x-axis and the MSigDB KEGG pathways identified as enriched in at least one timepoint on the y-axis. Red bars indicate pathways that were also identified as enriched in all four cell lines across both time points using an alternative method (Online Supplementary Figure S3D).

1160

haematologica | 2019; 104(6)

Combined LSD1 and ATRA in AML

inhibition were determined for each cell line. GSK2879552 inhibited cell growth with a potent average EC50 value of 137±30 nM across 20 cell lines. The average maximum inhibition was 62±5% suggesting complete cytostasis is not achieved by ten days with LSD1 inhibition alone. By comparing 6-day to 10-day proliferation assay results (Online Supplementary Figure S1A and B), it is evident that an increase in maximum inhibition can be achieved with

A

longer treatment, indicating that the biological consequences of LSD1 inhibition, including growth effects, may require time to be fully revealed. Several studies were conducted to determine if the lag in inhibition of cancer cell line growth is due to a delayed effect on all cells or involves some cell death. Cell cycle evaluation of MOLM13 cells treated with a titration of GSK-LSD1 for six days revealed an increase in cells in G1 phase with a concomi-

B

C

D

Figure 4. The addition of all-trans retinoic acid (ATRA) enhances GSK2879552 induced differentiation effects on acute myeloid leukemia (AML) cells. (A) Representative histograms of CD11b surface marker expression on MOLM-13 cells (left) and OCI-AML3 cells (right) treated with dimethyl sulfoxide (DMSO) (green), 1000 nM GSK2879552 (orange), 100 nM ATRA (blue), or 1000 nM GSK2879552 plus 100 nM ATRA (red) for two days. (B) Dose response curves of CD11b surface marker expression on MOLM-13 cells (left) and OCI-AML3 cells (right) treated with a titration of GSK2879552 + 0, 0.1, 1, 10, 100, or 1000 nM ATRA for two days.

(C) Percent CD11b+ expression (blue bars, ±Standard Error) on MOLM-13 cells (top) and OCI-AML3 cells (bottom) treated with the indicated concentrations of GSK2879552 and ATRA for two days. The value representing the predicted additive effect at each combination concentration is indicated with a red circle. (D) Shading represents percent CD11b+ expression relative to additive threshold on AML cells treated with the indicated concentrations of GSK2879552 and ATRA for two days. Dark green shading indicates >10% above additive threshold, light green indicates >5-10% above additive threshold, and light brown indicates 0-5% above additive threshold.

haematologica | 2019; 104(6)

1161

K.N. Smitheman et al.

tant decrease in S phase (Online Supplementary Figure S1C).

No appreciable accumulation of cells in sub-G1 was observed. To confirm the cell cycle findings, a BrdU assay that measures actively dividing cells by incorporation of BrdU into DNA during S-phase was performed (Figure 1B). MOLM-13 cells treated for six days had a dose

A

responsive decrease in BrdU signal (EC50 = 1.9±0.9 nM).

Finally, induction of caspase 3/7 was assessed to determine if cells were activating the apoptosis pathway. Seven AML cell lines were treated for up to six days with GSK2879552. In six of seven AML cell lines there was no appreciable induction of caspase 3/7 at any time point

B

C

Figure 5. The combination of GSK2879552 and all-trans retinoic acid (ATRA) promotes caspase-mediated cell death. (A) GDI values (±Standard Error) of MOLM13 (left) and OCI-AML3 (right) cells treated with 1000 nM GSK2879552, 100 nM ATRA, or 1000 nM GSK2879552 + 100 nM ATRA (COMBO) for 1 to 6 days. † Significance (P<0.05) between GSK552 and COMBO. *Significance (P<0.05) between GSK552 and COMBO as well as between ATRA and combination (COMBO).

(B) Fold caspase 3/7 cleavage values from day of peak induction (blue bars, ±Standard Error) for MOLM-13 cells (top) and OCI-AML3 cells (bottom) treated with the indicated concentrations of GSK2879552 and ATRA. The value representing the additive effect at each combination concentration is indicated with a red circle. (C) Shading represents fold caspase 3/7 cleavage values relative to threshold from day of peak induction of AML cells treated with the indicated concentrations of GSK2879552 and ATRA. Dark green shading indicates >10-fold above additive threshold, light green indicates 5- to 10-fold above additive threshold, light brown indicates 1- to 5-fold above additive threshold.

1162

haematologica | 2019; 104(6)

Combined LSD1 and ATRA in AML

(Figure 1C). These data indicate that the growth inhibitory effects of LSD1 inhibitors are due to a slowing of cell division and not significant killing of tumor cells.

Lysine specific demethylase 1 is a known regulator of normal hematopoiesis whereby it maintains hematopoietic stem and progenitor cell populations in a quiescent state through GFI1-mediated transcriptional repression of HoxA9 and Meis1.22 Additionally, several reports have shown that LSD1 inhibition promotes differentiation of leukemic cells.10 Knockdown of LSD1 leads to an increase in markers of differentiation, and this effect has been recapitulated using small molecule inhibitors of LSD1.1,12 To determine if growth inhibition achieved with GSK2879552 occurs in association with cellular differentiation, MOLM-13 cells were treated for one day with a titration of GSK2879552, and expression of myeloid differentiation marker genes, ITGAM (CD11b) and CD86, was measured by RT-PCR (Figure 1D). Higher expression of CD11b is indicative of differentiation from an immature myelo/monoblast to a more mature myeloid cell, and CD86 is co-expressed with CD80 on mature macrophages and dendritic cells where it is poised to interact with CD4+ T cells.23,24 GSK2879552 increased expression of CD11b and CD86 genes with average EC50 values of 31±1 nM and 28±6 nM, respectively. To determine if these increases in gene expression translate to increased protein levels, cell surface expression of CD11b and CD86 were measured in THP-1 and MOLM-13 cells by flow cytometry (Figure 1E and Online Supplementary Figure S1D and E). The percent of cells positive for each surface marker was determined by gating against an isotype control. Cells treated with a titration of GSK2879552 for one day showed a dosedependent increase in protein expression with average EC50 values of 23±4 nM and 44±4 nM for THP-1 and MOLM-13 cells, respectively.

To determine if the effect of GSK2879552 on differentiation markers occurs across cell lines that represent a range of AML subtypes, a panel of cell lines was evaluated by flow cytometry after treatment with either GSK2879552 or GSK-LSD1 (Figure 1F). Thirteen of 16 AML cell lines tested had at least a 10% increase in CD86 after three days of treatment, while 10 of 16 had at least a 10% increase in CD11b. Increases for both markers occurred in all five subtypes tested. Overall, the lowest increases in CD11b and CD86 occurred in the M6 subtype-derived AML cell lines (Online Supplementary Figure S1F). This is not surprising given that M6 AMLs are erythroleukemias, and differentiation of this subtype, similar to hematopoietic progenitor cells that undergo erythopoiesis to become mature red blood cells, may not involve increases in expression of CD11b or CD86.22 To better understand the effects on M6 subtype AML cells, another marker of hematopoietic progenitor cells, transferrin receptor (CD71), was evaluated.25 CD71 decreases as progenitor cells mature and, indeed, five of six M6 subtype AML cell lines showed at least a 10% decrease in expression of CD71 after three days of treatment with GSK2879552 (Online Supplementary Figure S1F).

One characteristic of active macrophages and neutrophils is their ability to produce superoxide anion in response to stimulation;26,27 therefore, the ability of GSK2879552 to induce superoxide anion production was evaluated as a functional readout of differentiation. Four AML cell lines were treated with GSK2879552 for 4, 7, 11, and 14 days. Upon stimulation with PMA, all four cell haematologica | 2019; 104(6)

lines released higher levels of superoxide anion when pretreated with LSD1 inhibitor (range: 3- to 249-fold over DMSO treated cells) (Figure 1G). SKM-1 cells were treated with a titration of GSK2879552 for seven days, and upon PMA stimulation, superoxide anion production was induced with an EC50 of 222±103 nM (Figure 1H).

To determine if the anti-proliferative effect of LSD1 inhibition in AML cell lines translates to primary AML, patient-derived samples that had not been subjected to long-term cell culture were evaluated using a blast colony formation assay. The number of blast colony forming units was enumerated after treatment with a titration of GSK2879552 in 14 AML patient samples. Twelve of the 14 samples were sensitive, defined as ≥30% inhibition, to GSK2879552 (Figure 1I). The majority of samples showed reduced colony growth by less than 50% while two samples reached complete reduction in blast colonies at the highest dose tested. These data suggest that the antitumor activity of GSK2879552 is not limited to cell lines and extends to AML patient-derived cells.

The combination of LSD1 inhibitor with all-trans retinoic acid enhances anti-tumor activity Given that the effects of LSD1 inhibition alone on AML cells are largely cytostatic and, therefore, may not be sufficient to combat the disease clinically, efforts were made to explore combinations that could improve outcome.

Previous reports suggested the combination potential of LSD1 inhibition and ATRA in non-APL AML.19,12 ATRA is used clinically for the treatment of APL and works by degrading the PML-RARα fusion, resulting in differentiation of blast cells and elimination of disease.16 Schenk et al. have shown enhanced reduction of cell growth when tranylcypromine, a much less potent, non-selective inhibitor of LSD1, is combined with ATRA.19,20 Therefore, studies were undertaken to determine if a similar combination effect could be observed with the selective LSD1 inhibitor, GSK2879552. Seven AML cell lines were treated with a titration of GSK2879552 + 0, 1, 10, 100, and 1000 nM ATRA and growth was monitored for up to six days.

The maximum inhibition achieved with the combination increased as the ATRA concentration increased in five of seven cell lines with a corresponding decrease in the IC50 value in four of seven cell lines on day 4 (Figure 2A and B and Online Supplementary Figure S2A). An additional cell line showed a decrease in IC50 value without a concomitant increase in maximum inhibition. Together these data indicate that the combination of GSK2879552 and ATRA results in enhanced growth inhibition of AML cell lines.

To determine whether the effect observed was synergistic, a Bliss Independence model was used (Online Supplementary Table S2A and B).21 A value ≥100 indicates synergy by calculating a single score that summarizes the Bliss Independence values obtained.9 Synergistic growth inhibition is evident by day 6 in all seven AML cell lines evaluated upon treatment with GSK2879552 combined with ATRA (Figure 2C-D).

Combination of GSK2879552 with all-trans retinoic acid leads to enrichment of differentiation associated gene signatures in acute myeloid leukemia cells To elucidate the mechanism by which GSK2879552 and ATRA combine to effect tumor cell growth, an RNA sequencing study to examine differential gene expression was conducted in six AML cell lines treated with vehicle, 1163

K.N. Smitheman et al.

GSK2879552, ATRA, or the combination of GSK2879552 plus ATRA at two different time points. For most cell lines and time points, a smaller number of gene expression changes are found in the single treatments compared to the combination (Figure 3A and B and Online Supplementary Figure S3A and B). In addition, a large set of genes is found exclusively differentially expressed in combination versus DMSO, indicating that there is an effect

from the combination of treatments not present in the single agent treatments (Figure 3A and B and Online Supplementary Figure S3A and B). The overlap of significantly differentially expressed genes under combination treatment for each time point in four cell lines (OCIAML3, MOLM-13, MV4-11, and THP-1) is shown in Figure 3C (top). This indicates that a portion of genes are found in all four cell lines at each time point (intersect).

A

B

C

D

E

F

Figure 6. Acute myeloid leukemia (AML) blast colony forming ability is impaired in AML patient-derived samples treated with a combination of LSD1 inhibitor plus all-trans retinoic acid (ATRA). (A) Dose response of AML blast colony formation of patient sample AML0007 treated with a titration of GSK2879552 Âą 100 nM ATRA or a titration of ATRA. Three replicate wells per concentration were tested. (B) Maximum percent of inhibition achieved for patient sample AML0007 treated with 100 nM ATRA or calculated from the GSK2879552 dose response curves + 100 nM ATRA. (C) IC50 values of the GSK2879552 dose response curves Âą 100 nM ATRA for patient sample AML0007. (D-F) Same as (A-C) for patient-derived sample 4031113SH.

1164

haematologica | 2019; 104(6)

Combined LSD1 and ATRA in AML

The union of the time points across all four cell lines shows that 593 genes are altered in all four cell lines in response to the combination, defining an ATRA + LSD1i combination signature (Figure 3C, bottom). The log2 transformed gene expression of the intersect genes for each time point (Figure 3C, top) is depicted in a heatmap that shows the row-mean centered data clustered hierarchically (Figure 3D). A subset of genes was validated by qRTPCR (Online Supplementary Figure S3C). Upon further examination of the combination treatment genes found in both time points (Figure 3C, 593 genes indicated in red) using gene set enrichment analyses, strong enrichment in KEGG pathways include hematopoietic cell lineage and cell adhesion molecule (CAMs) pathways (Figure 3E). This finding was validated using an additional methodology (Online Supplementary Figure S3D) as well as in a subset of cell lines assessed individually (Online Supplementary Figure S3E). CAMs are involved in immune cell function and, therefore, their expression is consistent with a more differentiated state. Together with the enriched expression of genes in the hematopoietic cell lineage pathway, enrichment in CAMs through the combined treatment of GSK2879552 and ATRA suggests enhanced differentiation of AML cells.

Gene set enrichment analyses revealed additional pathways in the combination that are associated with immune cell function. The mTOR, focal adhesion, and lysosome pathways are involved in autophagy, an immune mechanism for clearing intra-cellular proteins and organelles usually in response to cell stress (Figure 3E).28 Additionally, the systemic lupus erythematosus (SLE) pathway is highly enriched in the combination treatment. Given that SLE is an autoimmune disease, this finding is consistent with an immune-related functional effect.29 In conjunction with enrichment of hematopoietic cell lineage and CAMs signatures, perturbation of these additional pathways may provide further evidence of a molecular shift towards a gain of immune cell function.

The addition of all-trans retinoic acid enhances GSK2879552 induced differentiation effects on acute myeloid leukemia cells Given that the GSK2879552 and ATRA combination affects hematopoiesis and cell adhesion molecule pathways, studies to confirm these results by measuring differentiation in AML cell lines were performed. Myeloid differentiation cell surface marker expression was measured by flow cytometry on 4 AML cell lines treated for two days with GSK2879552. Representative histograms of CD11b expression on MOLM-13 and OCI-AML3 cells treated with DMSO, 1000 nM GSK2879552, 100 nM ATRA, or the combination of GSK2879552 and ATRA are shown (Figure 4A). In both cell lines, an increase in CD11b expression is evident with the combination. Cell lines were treated with a titration of GSK2879552 plus varying concentrations of ATRA. Percent of cells positive for CD11b were gated relative to isotype control and the values were plotted as dose response curves (Figure 4C). The increase in CD11b was dose responsive both with GSK2879552 and as increasing concentrations of ATRA were combined with GSK2879552 (Online Supplementary Table S3A). To determine whether the impact on CD11b expression with the combination is synergistic, an additivity threshold was calculated for each combination by adding together the increases obtained with each single haematologica | 2019; 104(6)

agent and was compared to the observed values achieved with the combination (Figure 4D and Online Supplementary Figure S4A). MOLM-13 and SIG-M5 cells treated with as little as 1 nM ATRA combined with GSK2879552 showed synergistic increases in CD11b. Synergistic effects were seen in OCI-AML3 cells with as little as 10 nM ATRA. THP-1 cells had a very robust response to GSK2879552 alone, reaching near 100% positive with as little as 12 nM GSK2879552, precluding assessment of the combination effect (Online Supplementary Figure S4A).

The difference between the additive threshold and the actual percent of CD11b expression achieved was determined for each cell line (Figure 4B). Synergistic increases in CD11b expression were observed in three of four cell lines treated with the combination of ATRA and GSK2879552.

Because GSK2879552 alone increased CD11b expression on nearly the entire population of THP-1 cells, median fluorescence intensity (MFI) values were determined to quantify shifts in CD11b expression (Online Supplementary Figure S4B and C). Fold changes in CD11b MFI were determined relative to DMSO-treated control and additivity threshold values were calculated by adding together the fold increases obtained with each single agent. Actual increases in CD11b MFI were synergistic with as little as 1 nM ATRA combined with GSK2879552.

A hallmark of functional differentiation involves the appearance of morphological changes.30 To investigate the ability of the combination to elicit an altered cell appearance, MOLM-13 cells were treated for three days with DMSO, GSK-LSD1, ATRA, or the combination of GSKLSD1 plus ATRA. Morphology was assessed by May Grunwald/Giemsa stain (Online Supplementary Figure S4D). The combination treatment resulted in the presence of lobular nuclei and granulocytic cytoplasm; both are characteristics of differentiated myeloid cells. These changes were largely absent from the cells treated with a single agent. These observations suggest that morphological differentiation represents a functional consequence of the gene expression and surface marker changes in AML cells caused by treatment with LSD1 inhibitor plus ATRA.

The combination of GSK2879552 and all-trans retinoic acid promotes caspase-mediated cell death In addition to inhibiting the growth of cells (Figure 2), a decrease in cell number relative to the number of cells at the start of the assay was observed in the combination (Figure 5A). By day 3, in MOLM-13 cells, the combination treatment produced a negative growth death index (GDI) value that continued to decrease with additional duration of treatment. Neither ATRA nor GSK2879552 alone achieved negative GDI values. By day 5, in OCI-AML3 cells, GSK2879552 plus 1000 nM ATRA achieved a GDI value of -50% compared to 3% with 1000 nM ATRA alone. A decrease in GDI value with the combination was observed in four of seven cell lines tested (Figure 5A and Online Supplementary Figure S5). These results indicate that, in addition to the improved inhibition of cell growth, AML cell death can be achieved with the combination of GSK2879552 and ATRA, and this cell death occurs at a time point after the observed increases in CD11b expression.

To confirm that the combination kills AML cells, and better understand the mechanism associated with this cytotoxic response, AML cells were treated with a dose titration of GSK2879552 plus and minus 1, 10, 100, and 1165

K.N. Smitheman et al.

1000 nM ATRA for up to six days and caspase 3/7 induction was measured. Similar to the method used in analysis of the cell surface marker data (see above), on the day of peak caspase induction an additivity threshold was determined for each combination by adding the fold caspase induction values obtained with each individual treatment condition. By comparing the observed fold induction achieved to the additive threshold, synergistic caspase activation is revealed with the combination at ATRA concentrations as low as 100 nM in MOLM-13 and OCIAML3 cells and as low as 10 nM ATRA in SIG-M5, MV-411, and HL-60 cells (Figure 5C and Online Supplementary Figure S5B). Additionally, annexin V staining was performed on MOLM-13 cells treated for three days with GSK-LSD1Âą1000 nM ATRA. Annexin V positive cells increased with the combination to a greater extent than each compound alone, providing supporting evidence for the cell death caused by the combination (Online Supplementary Figure S5C). Importantly, the cells that are positive for Annexin V are also positive for CD11b, demonstrating that the cells that differentiate also undergo apoptosis (Online Supplementary Figure S5D).

Using an approach similar to that utilized in the cell surface marker assessment, the difference between the additive threshold of caspase activation and the observed value was determined for each cell line and a heatmap was generated (Figure 5B). The combination of GSK2879552 plus ATRA induced caspase 3/7 in five out of seven AML cell lines tested. Importantly, the caspase induction in the combination reached levels greater than 10-fold above the additive threshold in all five of the positive cell lines at 100 nM ATRA, a concentration that corresponds to clinically achieved exposures.31,32

Acute myeloid leukemia blast colony forming ability is impaired in patient-derived samples treated with a combination of GSK2879552 plus all-trans retinoic acid Translating a combination effect beyond cell lines is important in understanding the potential for success in clinical trials. To this end, we assessed GSK2879552 combined with 100 nM ATRA in blast cells derived from BM of nine different AML patients (Online Supplementary Supplementary Table S1A). Dose response curves for GSK2879552, ATRA, and the combination of GSK2879552 plus 100 nM ATRA are shown for two patients (Figure 6A and D). For patient AML0007, the maximum inhibition of blast colony formation achieved with GSK2879552 was 39% while 100 nM ATRA had no effect on colony count. The combination of GSK2879552 and ATRA reached almost complete inhibition, a >50% decrease in AML blast colony forming ability (Figure 6B).

Similarly, the IC50 improved >2-fold from 105 nM with GSK2879552 to 43 nM with the combination (Figure 6C).

A second patient-derived sample, 4031113SH, had a modest improvement in maximum inhibition (54% inhibition with GSK2879552, 4% inhibition with ATRA, 69% inhibition with the combination); however, the IC50 decreased >10 fold from 150 nM with GSK2879552 to 11 nM with the combination (Figure 6E and F). Overall, a combination effect was observed in seven out of nine patient samples analyzed in at least one of the two parameters measured (Online Supplementary Figure S6A and B). An increase in maximum inhibition of at least 15% with the combination was achieved in five out of eight samples and of at 1166

least 30% in three out of eight samples. One sample achieved 100% inhibition with GSK2879552 alone so an improvement in inhibition was not possible with the combination in this sample. A decrease in IC50 of at least 2-fold with the combination was achieved in six out of nine samples and of at least 5-fold in two out of nine samples. Together, these data suggest that the combination effects are not limited to cell lines and that a lower dose of LSD1 inhibitor than is maximally efficacious as single agent may be sufficient to elicit the combination response.

Discussion Acute myeloid leukemia is a deadly cancer characterized by accumulation of immature myeloid cells in the BM that undergo replication in an uncontrolled manner.

Treatments such as chemotherapy focus on de-bulking the tumor by killing fast-growing cells. Unfortunately, it has been demonstrated that even after treatment some refractory cells remain that can resume growth resulting in disease relapse.33 A recent study has shown that inhibition of LSD1 can affect the leukemia-initiating population suggesting a potential clinical opportunity for LSD1 inhibitors in combination with de-bulking agents, even if these are not obviously beneficial as monotherapy.13 Another treatment option that has been more than 90% effective in APL involves the use of ATRA. ATRA has been shown to promote differentiation of myeloid cells and, therefore, the cells no longer possess replication potential.16 Applying this methodology more broadly to AML has thus far been unsuccessful because ATRA is unable to elicit a differentiation phenotype in non-APL AML.18 Interestingly, a new differentiation therapy, Enasidenib, was recently approved in the USA for the treatment of IDH2 mutant AML, demonstrating the potential of differentiation therapies in the non-APL AML setting.34 The results presented here indicate that features consistent with differentiation in AML cells can be achieved by inhibition of LSD1. Changes in cell surface expression of myeloid maturation markers is evident in all AML subtypes tested. LSD1 inhibition also increases superoxide anion production, indicative of a more granulocytic phenotype, in AML cells. These results combined with the decrease in blast cells of primary AML patient samples treated with GSK2879552 suggest the therapeutic opportunity for LSD1 inhibition in treating non-APL AML. The addition of ATRA synergistically enhanced the differentiation effect achieved with GSK2879552 through increased expression of myeloid cell surface marker CD11b. This is achieved at lower GSK2879552 concentrations than required when AML cells are treated without ATRA thus potentially widening the therapeutic index of LSD1 inhibition. Importantly this effect is observed across AML subtypes and is not exclusive to APL. Gene expression analysis revealed enrichment of the hematopoietic and CAM pathways further indicating that a primary mechanism in association with this combination involves differentiation. Pushing AML cells to differentiate with a combination of GSK2879552 and ATRA may offer profound implications for the success of AML treatment options for patients beyond APL or those with IDH2 mutations.

While enhancing differentiation provides an attractive rationale for pursuing the combination of LSD1 inhibitors with ATRA in the clinic, the observation that the combihaematologica | 2019; 104(6)

Combined LSD1 and ATRA in AML

nation promotes cytotoxic response provides an additional rationale to explore this combination in AML patients.

Treatment of non-APL AML cell lines with GSK2879552 and ATRA results in synergistic cleavage of caspases 3 and 7, the executioner caspases.35 The synergistic cell death occurs after the observed changes in cell surface marker expression and may suggest that as cells reach terminal differentiation, this process is a signal to the cells to undergo apoptosis in a manner similar to the life cycle of normal immune cells.36 Given that the gene set enrichment analysis revealed that the lysosome and mTOR pathways are also affected with the combination, another possibility is

References 1. Fang J, Ying H, Mao T, et al. Upregulation of CD11b and CD86 through LSD1 inhibition promotes myeloid differentiation and suppresses cell proliferation in human monocytic leukemia cells.

Oncotarget.

2017;8(49):85085-85101.

2. Arber DA, Orazi A, Hasserjian R, et al. The 2016 revision to the World Health Organization classification of myeloid neoplasms and acute leukemia. Blood. 2016; 127(20):2391-2405.

3. Dombret H, Gardin C. An update of current treatments for adult acute myeloid leukemia. Blood. 2016;127(1):53-61.

4. Wouters BJ, Delwel R. Epigenetics and approaches to targeted epigenetic therapy in acute myeloid leukemia. Blood. 2016; 127(1):42-52.

5. Zhou C, Wu F, Lu L, et al. Structure activity relationship and modeling studies of inhibitors of lysine specific demethylase 1.

PLoS One. 2017;12(2):e0170301.

6. Shi YJ, Matson C, Lan F, Iwase S, Baba T, Shi Y. Regulation of LSD1 histone demethylase activity by its associated factors. Mol Cell.

2005;19(6):857-864.

7. Saleque S, Kim J, Rooke HM, Orkin SH. Epigenetic regulation of hematopoietic differentiation by Gfi-1 and Gfi-1b is mediated by the cofactors CoREST and LSD1. Mol Cell. 2007;27(4):562-572.

8. Harris WJ, Huang X, Lynch JT, et al. The histone demethylase KDM1A sustains the oncogenic potential of MLL-AF9 leukemia stem cells. Cancer Cell. 2012;21(4):473-487.

9. McGrath JP, Williamson KE, Balasubramanian S, et al. Pharmacological Inhibition of the Histone Lysine Demethylase KDM1A Suppresses the Growth of Multiple Acute Myeloid Leukemia Subtypes.

Cancer Res. 2016;76(7):1975-1988.

10. Lynch JT, Harris WJ, Somervaille TC. LSD1 inhibition: a therapeutic strategy in cancer?

Expert Opin Ther Targets. 2012;16(12):12391249.

11. Mohammad HP, Smitheman KN, Kamat CD, et al. A DNA Hypomethylation Signature Predicts Antitumor Activity of LSD1 Inhibitors in SCLC. Cancer Cell.

2015;28(1):57-69.

12. Maes T, Mascaro C, Tirapu I, et al. ORY-

haematologica | 2019; 104(6)

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

that the differentiation triggers an autophagic response that leads to a switch to apoptosis.28 Importantly, applying the combination to primary patient samples ex vivo successfully illustrates that a beneficial combination effect is not restricted to cultured cell lines. The broad response of cell lines and primary AML samples to the combination does preclude the ability to identify a biomarker preclinically. Pretreatment biopsy assessment will, therefore, be required to determine which patients will most likely respond to the combination. Taken together, these results highlight the possibility of achieving a cure in AML through treatment with LSD1 inhibitors and ATRA.

1001, a Potent and Selective Covalent KDM1A Inhibitor, for the Treatment of Acute Leukemia. Cancer Cell. 2018; 33(3):495-511 e412.

Cusan M, Cai SF, Mohammad HP, et al.

LSD1 inhibition exerts its antileukemic effect by recommissioning PU.1- and C/EBPalpha-dependent enhancers in AML. Blood. 2018;131(15):1730-1742. de The H, Pandolfi PP, Chen Z. Acute Promyelocytic Leukemia: A Paradigm for Oncoprotein-Targeted Cure. Cancer Cell.

2017;32(5):552-560.

Al Tanoury Z, Piskunov A, Rochette-Egly C. Vitamin A and retinoid signaling: genomic and nongenomic effects. J Lipid Res. 2013;54(7):1761-1775.

Melnick A, Licht JD. Deconstructing a disease: RARalpha, its fusion partners, and their roles in the pathogenesis of acute promyelocytic leukemia. Blood. 1999; 93(10):3167-3215.

Zhou GB, Zhao WL, Wang ZY, Chen SJ, Chen Z. Retinoic acid and arsenic for treating acute promyelocytic leukemia. PLoS Med. 2005;2(1):e12.

Johnson DE, Redner RL. An ATRActive future for differentiation therapy in AML. Blood Rev. 2015;29(4):263-268.

Schenk T, Chen WC, Gollner S, et al.

Inhibition of the LSD1 (KDM1A) demethylase reactivates the all-trans-retinoic acid differentiation pathway in acute myeloid leukemia. Nat Med. 2012;18(4):605-611.

Kaniskan HU, Martini ML, Jin J. Inhibitors of Protein Methyltransferases and Demethylases. Chem Rev. 2018; 118(3):9891068.

Foucquier J, Guedj M. Analysis of drug combinations: current methodological landscape. Pharmacol Res Perspect. 2015 3(3):e00149.

Sprussel A, Schulte JH, Weber S, et al.

Lysine-specific demethylase 1 restricts hematopoietic progenitor proliferation and is essential for terminal differentiation.

Leukemia. 2012;26(9):2039-2051. van Lochem EG, van der Velden VH, Wind HK, te Marvelde JG, Westerdaal NA, van Dongen JJ. Immunophenotypic differentiation patterns of normal hematopoiesis in human bone marrow: reference patterns for age-related changes and disease-induced shifts. Cytometry B Clin Cytom.

2004;60(1):1-13.

24. Chen L, Flies DB. Molecular mechanisms of T cell co-stimulation and co-inhibition. Nat Rev Immunol. 2013;13(4):227-242.

25. Marsee DK, Pinkus GS, Yu H. CD71 (transferrin receptor): an effective marker for erythroid precursors in bone marrow biopsy specimens. Am J Clin Pathol. 2010; 134(3):429-435.

26. Song HO, Ryu JS. Superoxide anion production by human neutrophils activated by Trichomonas vaginalis. Korean J Parasitol.

2013;51(4):479-484.

27. Finkel TH, Pabst MJ, Suzuki H, et al. Priming of neutrophils and macrophages for enhanced release of superoxide anion by the calcium ionophore ionomycin. Implications for regulation of the respiratory burst. J Biol Chem. 1987;262(26):12589-12596.

28. Marino G, Niso-Santano M, Baehrecke EH, Kroemer G. Self-consumption: the interplay of autophagy and apoptosis. Nat Rev Mol Cell Biol. 2014;15(2):81-94.

29. Danchenko N, Satia JA, Anthony MS. Epidemiology of systemic lupus erythematosus: a comparison of worldwide disease burden. Lupus. 2006;15(5):308-318.

30. Grassi L, Pourfarzad F, Ullrich S, et al.

Dynamics of Transcription Regulation in Human Bone Marrow Myeloid Differentiation to Mature Blood Neutrophils. Cell Rep. 2018;24(10):27842794.

31. Regazzi MB, Iacona I, Gervasutti C, Lazzarino M, Toma S. Clinical pharmacokinetics of tretinoin. Clin Pharmacokinet.

1997;32(5):382-402.

32. Russo D, Regazzi M, Sacchi S, et al. All-trans retinoic acid (ATRA) in patients with chronic myeloid leukemia in the chronic phase.

Leukemia. 1998;12(4):449-454.

33. Hourigan CS, Gale RP, Gormley NJ, Ossenkoppele GJ, Walter RB. Measurable residual disease testing in acute myeloid leukaemia. Leukemia. 2017;31(7):14821490.

34. Kim ES. Enasidenib: First Global Approval.

Drugs. 2017;77(15):1705-1711.

35. Koff JL, Ramachandiran S, Bernal-Mizrachi L. A time to kill: targeting apoptosis in cancer. Int J Mol Sci. 2015;16(2):2942-2955.

36. Nagata S. Apoptosis and clearance of apoptotic cells. Annu Rev Immunol. 2018; 36:489-517.

1167

ARTICLE Ferrata Storti Foundation

Haematologica 2019 Volume 104(6):1168-1175

Acute Myeloid Leukemia

Impact of induction regimen and allogeneic hematopoietic cell transplantation on outcome in younger adults with acute myeloid leukemia with a monosomal karyotype

Frédéric Baron,1 Marian Stevens-Kroef,2 Michal Kicinski,3 Giovanna Meloni,4 Petra Muus,2,5 Jean-Pierre Marie,6 Constantijn J.M. Halkes,7 Xavier Thomas,8 Radovan Vrhovac,9 Francesco Albano,10 François Lefrère Sr.,11 Simona Sica,12 Marco Mancini,4 Adriano Venditti,13 Anne Hagemeijer,14 Joop H. Jansen,2 Sergio Amadori,13 Theo de Witte,2 Roelof Willemze7 and Stefan Suciu3

Groupe Interdisciplinaire de Génoprotéomique Appliquée (GIGA), Laboratory of Hematology, University of Liege, Belgium; 2Radboud University Medical Center, Nijmegen, the Netherlands; 3EORTC Headquarters, Brussels, Belgium; 4Department of Hematology, Sapienza University, Rome, Italy; 5King’s College Hospital, London, UK; 6Department of Hematology, Saint Antoine Hospital, Paris, France; 7Leiden University Medical Center, the Netherlands; 8CHU Lyon, France; 9University Hospital Center Zagreb, Croatia; 10University of Bari, Italy; 11Necker Hospital, Paris, France; 12Università Cattolica Sacro Cuore, Roma, Italy; 13 University Tor Vergata, Roma, Italy and 14University of Leuven, Belgium 1

ABSTRACT

M

Correspondence: FRÉDÉRIC BARON f.baron@ulg.ac.be Received: August 17, 2018.

Accepted: November 29, 2018.

Pre-published: December 6, 2018. doi:10.3324/haematol.2018.204826 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/104/6/1168 ©2019 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

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

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

1168

onosomal karyotype confers a poor prognosis in patients with acute myeloid leukemia. Here, we determined the impact of the type of remission-induction chemotherapy and the impact of having a donor in younger acute myeloid leukemia patients with a monosomal karyotype included in two phase III trials. In the first trial patients were randomized to receive either daunorubicin, mitoxantrone, or idarubicin in addition to standard-dose cytarabine and etoposide for induction chemotherapy. In the second trial patients were randomized to standarddose cytarabine or high-dose cytarabine induction, both with daunorubicin and etoposide. In both trials, patients who achieved a complete remission with or without complete hematologic recovery underwent allogeneic hematopoietic stem cell transplantation if they had a donor; otherwise, they underwent autologous transplantation. In comparison to patients with intermediate-risk cytogenetics without a monosomal karyotype (n=1,584) and with adverse cytogenetics without a monosomal karyotype (n=218), patients with a monosomal karyotype (n=188) were more likely not to achieve a complete remission with or without count recovery [odds ratio=2.85, 95% confidence interval (95%, CI): 2.10-3.88] and had shorter overall survival [hazard ratio, (HR)=2.44, 95% CI: 2.08-2.88]. There was no impact of the type of anthracycline or of the dose of cytarabine on outcomes in patients with a monosomal karyotype. Among monosomal karyotype patients who achieved a complete remission with or without count recovery, HLA-identical related donor availability was associated with longer survival from complete remission with or without count recovery (HR=0.59, 95% CI: 0.37-0.95). ClinicalTrials.gov identifiers: AML-10: NCT00002549; AML-12: NCT00004128.

Introduction The prognosis of young adult patients with intermediate/high-risk acute myelogenous leukemia (AML) remains unsatisfactory. With current remission induction chemotherapy, 15-40% of such patients fail to achieve a complete remission (CR), and only 30-50% of them remain alive for more than 5 years.1-4 Approximately 55% of AML patients have at least one chromosomal abnormalihaematologica | 2019; 104(6)

MK in EORTC/GIMEMA AML-10&12

ty that can be detected by conventional cytogenetics.5 In younger AML patients, the karyotype of leukemic cells has remained one of the main prognostic factors.5,6 Specifically, patients can be classified into a favorable, intermediate or adverse group.5,6 This classification not only provides information on prognosis, but also influences the choice of post-remission treatment.5,7 In 2008, Breems et al. identified a group of cytogenetic abnormalities, termed monosomal karyotype (MK), which was associated with a particularly poor prognosis.8 MK is defined as the presence of two or more autosomal monosomies, or a single monosomy in the presence of structural abnormalities.8,9 The aim of this study was to determine the impact of the type of remission-induction chemotherapy and allogeneic hematopoietic stem cell transplantation (HSCT) in younger AML patients with a MK, using the data from the European Organization for Research and Treatment on Cancer/Gruppo Italiano Malattie Ematologiche dell'Adulto (EORTC/GIMEMA) AML-10 and AML-12 phase III multicenter trials.

Methods Study design In the EORTC Leukemia Group/GIMEMA AML-10 trial,2 patients 15-60 years old were randomized to receive either daunorubicin (50 mg/m2), mitoxantrone (12 mg/m2), or idarubicin (10 mg/m2) on days 1, 3 and 5 in addition to standard-dose cytarabine (25 mg/m2 bolus followed by 100 mg/m2 given as a continuous infusion daily for 10 days) and etoposide (100 mg/m2 on days 1-5) for induction chemotherapy.

In the EORTC Leukemia Group/GIMEMA AML-12 trial,1 patients 15-60 years old were randomized between standard-dose cytarabine induction: daunorubicin (50 mg/m2 per day on days 1, 3, and 5) plus etoposide (50 mg/m2 per day on days 1-5) plus 10 days of cytarabine (100 mg/m2 per day as a continuous intravenous infusion) and high-dose cytarabine induction: daunorubicin (50 mg/m2 on days 1, 3, and 5) plus etoposide (50 mg/m2 per day on days 1-5) plus cytarabine (3,000 mg/m2 every 12 h as a 3 h intravenous infusion on days 1, 3, 5, and 7).

In both trials, a second cycle of induction was administered to patients who achieved a partial response. Patients who achieved a CR or a CR with incomplete blood count recovery (CRi) after one or two courses of induction chemotherapy received consolidation chemotherapy with the same anthracycline as in the induction course plus intermediate-dose cytarabine (500 mg/m2 every 12 h as a 2 h intravenous infusion on days 1-6). Younger patients (<46 years in AML-10 and <50-60 years in AML-12) were then scheduled to undergo allogeneic HSCT in first CR/CRi if they had an HLA-identical family donor (in both trials) or if they had an unrelated donor and needed two induction courses to achieve a CR/CRi or had chromosome abnormalities involving 3q, 5, 7, 11q23, t(6;9), t(9;22) or complex abnormalities (in the AML-12 trial). Patients without a donor were scheduled to undergo autologous HSCT in first CR/CRi.

Ethics approval and consent to participation This is a retrospective analysis limited to data from patients included in phase III multicenter prospective trials (either the EORTC/GIMEMA AML-10 or the EORTC/GIMEMA AML-12).

Both prospective phase III trials were approved by the internal review boards of EORTC and GIMEMA and the ethical committee of each participating institution, and were conducted in accorhaematologica | 2019; 104(6)

dance with the Declaration of Helsinki. All patients signed the respective informed consent form.

Cytogenetic assessment Cytogenetic examinations were performed at diagnosis.

Cytogenetic data were centrally reviewed.2 For the current analysis, cytogenetics were centrally re-reviewed, described according to International System for Cytogenetic Nomenclature (ISCN)10 and classified using the refined UK Medical Research Council (MRC) classification.6 MK was defined as the presence of two or more autosomal monosomies or a single monosomy in the presence of structural abnormalities, as introduced by Breems et al.8

Statistical analyses The duration of overall survival (OS) was calculated from the date of randomization until death. The Kaplan-Meier method was used to estimate the OS rates.11 Confidence intervals for the 5-year OS rates were obtained using the normal approximation of the distribution of log[-log(survival)] and the Greenwood variance formula.12 The confidence interval of the median OS from CR/CRi was estimated based on the Brookmeyer and Crowley method.13 Log-rank tests and Cox models were used to compare OS between groups.14 Logistic regression was used to assess associations with CR/CRi achievement after induction. The analysis was stratified (in the case of survival analysis) or adjusted (in the case of logistic regression) by protocol when it included data from two trials. Multivariate Cox and logistic regression models were performed to assess the associations of MK and adverse MRC risk group with outcomes, adjusting for known prognostic factors. A Fisher exact test was used to investigate the association between two categorical variables. All reported P-values are two-sided. SAS 9.4 software (SAS Institute Inc. Cary, NC, USA) was used for the statistical analyses.

Results Patients In the AML-10 trial, 2,157 patients were randomized to receive daunorubicin, mitoxantrone or idarubicin. The current analyses were performed in a subgroup of 911 patients for whom cytogenetic data were available and who did not have t(8;21), inv(16) or t(15;17); 696 of them were classified in the MRC intermediate cytogenetic risk group and 215 in the adverse group. Out of the 911 patients, 93 had a MK (5 with intermediate-risk cytogenetics and 88 with adverse risk) (Table 1, Figure 1). In the AML-12 trial, 1,942 patients were randomized between high-dose cytarabine or standard-dose cytarabine. The current analyses were performed in a subgroup of 1,079 patients for whom cytogenetic data were available and who did not have t(8;21), inv(16) or t(15;17); 896 of them were classified in the MRC intermediate cytogenetic risk group and 182 in the adverse-risk group (information was missing for 1 patient). Out of these 1,079 patients, 95 had a MK (4 with intermediate-risk cytogenetics and 91 with adverse-risk) (Table 1, Figure 1). The patientsâ&#x20AC;&#x2122; median follow up was 10.8 years in the whole population, 16.6 years among those in the AML-10 study and 9.9 years among AML-12 patients.

Monosomal karyotype is an independent poor prognostic factor in young acute myeloid leukemia patients The impact of MK on AML outcomes was assessed by comparing outcomes of patients without MK and without 1169

F. Baron et al.

adverse-risk cytogenetics (NotAdvMK–, n=1,584), those without MK but with adverse-risk cytogenetics (AdvMK–, n=218) and those with a MK (MK+, n=188).

Achievement of complete response with or without blood count recovery Patients with AdvMK– [odds ratio (OR)=1.80, 95% confidence interval (95% CI): 1.33-2.42] or MK+ (OR=3.09, 95% CI: 2.26-4.22) had a higher probability of not reaching a CR/CRi after induction compared to NotAdvMK– patients (Table 2). CR/CRi was achieved in 76%, 63% and 50% of NotAdvMK–, AdvMK– and MK+ patients, respectively. Comparing MK+ to MK– patients (NotAdvMK– or AdvMK–), the odds of not achieving a CR/CRi were almost three times higher (OR=2.85, 95% CI: 2.10-3.88) for MK+ patients. The probability of not achieving a CR/CRi was also significantly higher in MK+ than in AdvMK– patients (OR=1.72, 95% CI: 1.15-2.57).

In a multivariate logistic regression model including age, WHO Performance Status, and white blood cell count, AdvMK– (OR 1.91, 95% CI: 1.41-2.59) and MK+ (OR 3.34, 95% CI: 2.42-4.59) were associated with higher probabilities of not achieving a CR/CRi compared to NotAdvMK– (Table 2).

(56%), 14 out of 28 patients (50%) in the mitoxantrone arm and 13 out of 31 (42%) patients in the idarubicin arm (P=0.54). The 5-year OS rates were 13.0% (95% CI: 4.127.1%) in daunorubicin patients, 6.7% (95% CI: 1.219.2%) in mitoxantrone patients, and 11.7% (95% CI: 3.126.6%) in idarubicin patients (Figure 2C). The 5-year OS rates from CR/CRi were 17.6% (95% CI: 4.3-38.3%) in daunorubicin patients, 11.5% (95% CI: 0.9-37.5%) in idarubicin patients and 14.3% (95% CI: 2.3-36.6%) in mitoxantrone patients (logrank P=0.53).

No benefit of high-dose cytarabine in patients with monosomal karyotype Response data after induction were available for 93 out of 95 MK+ and 978 out of 984 MK– patients from the AML12 trial. MK (present vs. absent) was of predictive importance for the effect of high-dose cytarabine on the probability of reaching CR/CRi after induction (interaction test P-value: 0.01). MK– patients randomized to the high-dose cytarabine arm were more likely to reach CR/CRi com-

Table 1. Patients’ characteristics.

Study

Overall survival The 5-year OS rates were 39.1%, 24.1% and 7.2% in the NotAdvMK–, AdvMK– and MK+ patients, respectively (Figure 2A). The estimated hazard ratios comparing AdvMK– and MK+ patients to NotAdvMK– patients were 1.48 and 2.58, respectively (Table 3). Comparing MK+ to MK– (NotAdvMK– or AdvMK–) and AdvMK– patients, the estimates of the unadjusted hazard ratio were 2.44 (95% CI: 2.08-2.88) and 1.74 (95% CI: 1.41-2.15), respectively.

In a multivariate Cox model, in comparison to NotAdvMK– patients, those with AdvMK– (HR 1.51, 95% CI: 1.28-1.77) or MK+ (HR 2.71, 95% CI: 2.29-3.20) had a shorter OS (Table 3).

Overall survival from complete remission with or without hematologic recovery The 5-year OS rates from CR/CRi were 48.5%, 35.5% and 11.4% in NotAdvMK–, AdvMK– and MK+ patients, respectively (Figure 2B). The estimated hazard ratios comparing AdvMK– and MK+ patients to NotAdvMK– patients were 1.50 and 2.87, respectively (Table 4). Comparing MK+ to MK– (NotAdvMK– or AdvMK–) and AdvMK– patients, the estimates of the hazard ratio were 2.73 (95% CI: 2.17-3.45) and 1.91 (95% CI: 1.42-2.57), respectively.

In a multivariate Cox model, in comparison to NotAdvMK–, AdvMK– (HR 1.52, 95% CI: 1.23-1.88) and MK+ (HR 2.95, 95% CI: 2.32-3.74) were associated with shorter OS from CR/CRi (Table 4). In a sensitivity analysis, we modified the multivariate model by additionally stratifying it by donor availability. The results of this analysis were similar to those of the main analysis (HR=1.59, 95% CI: 1.28-1.98 for AdvMK– versus NotAdvMK– and HR=3.00, 95% CI: 2.35-3.82, for MK+ versus NotAdvMK–).

No impact of the type of anthracycline on outcomes in patients with a monosomal karyotype Response data were available for 91 out of 93 MK+ patients from the AML-10 trial. CR/CRi was reached after induction by 18 out of 32 patients in the daunorubicin arm 1170

NotAdvMK-

N. of patients 1,584 MRC cytogenetic risk group, N. of patients (%) Adverse 0 Intermediate 1583 (100) Missing 1 Trial AML-10 691 Daunorubicin, n. of patients (%) 221 (32.0) Idarubicin, n. of patients (%) 226 (32.7) Mitoxantrone, n. of patients (%) 244 (35.3) AML-12 893 SDAC, n. of patients (%) 466 (52.2) HiDAC, n. of patients (%) 427 (47.8) Male / Female, n.

792 / 790 Age (years), n. of patients (%) 15-25 155 (10) 26-45 663 (42) 46-60 766 (48) WHO performance status, n. (%) 0 679 (43) 1 672 (42) 2-4 226 (14) Missing 7 (0) 9 WBC x10 /L at diagnosis, n. of patients (%) < 25 892 (56) 25-99.9 496 (31) ≥100 195 (12) Missing 1 (0) N. of patients with CR/CRi 1194 after induction Donor, n. among pts with CR/CRi (%) No 710 (59.5) Yes 412 (34.5) Missing 72 (6.0)

AdvMK-

MK+

218

188

218 (100) 0 0

179 (5) 9 (95) 0

127 29 (22.8) 56 (44.1) 42 (33.1) 91 42 (46.2) 49 (53.8) 107 / 111

93 32 (34.4) 31 (33.3) 30 (32.3) 95 47 (49.5) 48 (50.5) 109 / 79

28 (13) 90 (41) 100 (46)

15 (8) 71 (38) 102 (54)

84 (39) 104 (48) 30 (14)

62 (33) 100 (53) 26 (14)

147 (67) 51 (23) 20 (9) 0 (0) 137

140 (75) 38 (20) 10 (5) 0 (0) 92

5 (54.7) 58 (42.3) 4 (2.9)

55 (59.8) 35 (38.0) 2 (2.2)

NotAdvMK–: not adverse cytogenetic excluding a monosomal karyotype; AdvMK–: adverse cytogenetic excluding a monosomal karyotype; MK+: monosomal karyotype; SDAC, standard-dose cytarabine; HiDAC: high-dose cytarabine; WHO: World Health Organization; WBC: white blood cell count; CR: complete remission; CRi, complete remission with incomplete hematologic recovery.

haematologica | 2019; 104(6)

MK in EORTC/GIMEMA AML-10&12

pared to those randomized to the standard-dose cytarabine arm (OR=1.51, 95% CI: 1.12-2.04; CR/CRi rate 80% vs.

73%, respectively). Among MK+ patients, 19 (1 without hematologic recovery and 2 with missing hematologic recovery data) out of 46 patients from the high-dose cytarabine arm and 28 (5 without hematologic recovery and 1 with missing hematologic recovery data) out of 47 patients in the standard-dose cytarabine arm reached CR/CRi. In other words, in MK+ patients the trend was even in a different direction (OR=0.48, 95% CI: 0.21-1.09; CR/CRi rate 41% in the high-dose cytarabine arm vs. 60% in the standard-dose cytarabine arm, P=0.080). Excluding

six patients without hematologic recovery, the CR rate was 39% in the high-dose cytarabine arm versus 49% in the standard-dose cytarabine arm. Furthermore, among 38 patients with reported CR with a full hematologic recovery after induction, six (including 2 high-dose cytarabine and 4 standard-dose cytarabine patients) were reported not to have had a full hematologic recovery after consolidation.

Interestingly, no benefit of high-dose cytarabine on OS was observed in the subgroup of MK+ patients (HR=1.03, 95% CI: 0.68-1.57; P=0.88) (Figure 2D). The estimate of the hazard ratio for OS after CR/CRi among MK+ patients was 0.82 (95% CI: 0.44-1.53; P=0.53).

Figure 1. Flow chart of the patients included in the current analyses. MK: monosomal karyotype (MK-: without MK; MK+: with MK); OS: overall survival; CR/Cri: complete remission/complete remission with incomplete blood count recovery; Allo-HSCT: allogeneic hematopoietic stem cell transplantation; Auto-HSCT: autologous hematopoietic stem cell transplantation.

haematologica | 2019; 104(6)

1171

F. Baron et al.

Having an HLA-identical related donor improved overall survival from complete remission in patients with a monosomal karyotype A total of 92 MK+ patients achieved a CR/CRi on study.

Among them, 35 patients had and 55 did not have an HLA-identical related donor (information was missing for 2 patients). Among those with an HLA related donor, 25 patients (71%) received an allogeneic HSCT, including 20 (57%) who were transplanted in first CR/CRi (Figure 1). In addition, 23 patients (including 22 without an HLAmatched related donor) underwent autologous HSCT in first CR/CRi.. The median time between CR/CRi and HSCT in first CR/CRi was 91 days for patients who received an autologous transplant (range, 42-167), 76 days for patients who received an allogeneic HSCT (range, 46171), 91 days for patients without a donor (range: 42.0 171.0) and 71 days for patients with a donor (range, 46.0 148.0). OS from CR/CRi was longer in patients with a donor than in those without, such that the 5-year OS rates following CR/CRi were 24.1% (95% CI: 11.4-39.3%) and 3.8% (95% CI: 0.7-11.5%), respectively (HR=0.59, 95% CI: 0.37-0.95) (Figure 3A).

Given the clinical importance of the observations described just above, we performed several sensitivity analyses to assess the impact of allogeneic HSCT on outcomes among MK+ patients. The exclusion of patients with a CRi had little impact on the estimated hazard ratio (HR=0.60, 95% CI: 0.37-0.98). The advantage of having an HLA-identical related donor remained present (HR=0.61, 95% CI: 0.38-0.99) after adjusting for patient’s age (< or ≥ 45 years old).

Furthermore, the size of the estimated treatment effect confirmed the benefit of allogeneic HSCT when the posttransplant survival of patients who received an allogeneic graft was compared to that of patients who received an autologous graft (HR=0.54, 95% CI: 0.26-1.12) (Figure 3B).

Second, in a Cox model including allogeneic HSCT (modeled as a time-varying covariate) and age, and stratified by protocol, allogeneic HSCT was associated with a longer survival from CR/CRi (HR=0.61, 95% CI: 0.36-1.02).

Discussion As demonstrated in this study, several other studies have established that the presence of a MK is associated with a particularly poor outcome in younger AML patients.8,15-17 However, the number of studies regarding the best remission-induction regimens for MK+ AML patients as well as the impact of allogeneic HSCT on the outcomes of such patients have been the focus of only a few studies.16,18 In order to investigate these important issues, we used the data concerning the MK+ patients included in the EORTC/GIMEMA AML-10 and AML-12 phase III multicenter trials. Several observations were made.

First, our study confirms the poor prognosis associated with MK+ in younger AML patients. Specifically, MK+ was associated with a lower probability of achieving a CR/CRi, a shorter OS and a shorter OS from CR/CRi, whether taking other prognostic factors into account or not. This is in concordance with prior observations that the CR/CRi rate for adult MK+ patients ranged from 1443% and the OS rate from 9-18%.8,15-17 In addition, we investigated whether the type of anthra1172

cycline given during remission induction affected the outcomes of MK+ patients among patients included in the AML-10 trial. Unfortunately, none of the assessed anthracyclines was associated with better outcomes among MK+ patients. These findings are consistent with the results of prior phase III trials showing that increasing the dose of daunorubicin was beneficial mainly in patients with favorable or intermediate-risk cytogenetics.19,20 In the first analysis of the AML-12 trial, we observed that induction with high-dose cytarabine increased the proportion of patients achieving a CR/CRi and prolonged OS in patients younger than 46 years of age.1 This benefit was also observed in patients with adverse cytogenetic abnormalities and/or FLT3-internal tandem duplication (ITD) mutations, as well as in those with secondary AML. Here we found no evidence of a benefit of high-dose cytarabine in the subgroup of patients with a MK. There was even a suggestion of a lower incidence of CR/CRi in MK+ patients randomized to the high-dose cytarabine arm. This finding could be explained by previous studies showing that up to 80% of MK+ patients have a mutation in the TP53 gene.21 Patients with TP53 mutation in their tumor cells are resistant to high-dose cytarabine, as has been demonstrated in patients with mantle cell lymphoma.22 Unfortunately, the TP53 gene mutation was not evaluated in the present study.

Finally, we investigated whether patients with a donor had an OS benefit in comparison to those without a donor. As previously demonstrated by us and by other groups of investigators, among MK– AML patients with intermediate or unfavorable karyotype, the presence of a

Table 2. Association between monosomal karyotype and Medical Research Council adverse-risk group and achievement of complete remission with or without hematologic recovery after induction.

Covariate

OR of no CR/CRi

95% CI

P value

Unadjusted analysis^ Cytogenetic group NotAdvMKAdvMKMK+

<0.001 1 1.80 1.33 - 2.42 3.09 2.26 - 4.22 Multivariate analysis*

Cytogenetic group NotAdvMK1 AdvMK1.91 MK+ 3.34 Age (years) 15-25 1 26-45 1.35 46-60 1.61 WHO Performance Status at baseline 0 1 1 1.07 2-4 1.83 WBC at diagnosis (x109/L) <25 1 ≥ 25 and < 100 1.21 ≥ 100 1.79

<0.001 1.41 - 2.59 2.42 - 4.59 0.028 0.92 - 1.99 1.10 - 2.36 <0.001 0.85 - 1.34 1.35 - 2.48 0.001 0.96 - 1.52 1.31 - 2.46

OR: odds ratio; 95% CI: 95% confidence interval; NotAdvMK–: not adverse cytogenetic excluding a monosomal karyotype; AdvMK–: adverse cytogenetic excluding a monosomal karyotype; MK+: monosomal karyotype; WHO: World Health Organization; WBC: white blood cell count. ^Obtained with a logistic regression model including protocol (AML-10 vs. AML-12) and cytogenetic group. *Obtained with a logistic regression model including protocol and all covariates presented in the Table.

haematologica | 2019; 104(6)

MK in EORTC/GIMEMA AML-10&12

donor for an allogeneic HSCT prolonged OS.7,23 Interestingly, we made similar findings in patients with MK+, suggesting a positive impact of allogeneic HSCT in this population of patients with a generally poor outcome.

To our knowledge, our study is the first to investigate the impact of allogeneic HSCT in MK+ patients using a donor versus no donor comparison (which is considered as the gold-standard technique to address this question when donor availability is prospectively collected). These results were further confirmed in a Cox model handling allogeneic HSCT as a time-varying covariate and by comparing patients who received an allogeneic transplant from an HLA-matched related donor with those who received an autologous transplant. The three approaches, which are based on different assumptions, consistently indicated a positive impact of allogeneic HSCT. Among 20 patients who underwent allogeneic HSCT from an HLA-matched related donor, two were still alive and in follow-up 10 years after reaching CR, indicating a curative potential of the treatment. Importantly, our results are concordant with prior results from Kayser et al., who observed that allo-

A

C

geneic HSCT prolonged OS in younger (18-60 years of age) MK+ AML patients (using Mantel-Byar analysis)13, as well as with a recent publication by Cornelissen et al., who demonstrated better leukemia-free survival in younger MK+ AML patients offered allogeneic HSCT (using a timedependent Cox analysis).18 This is also concordant with previous studies showing that, although the prognosis of MK+ patients remains worse than that of MKâ&#x20AC;&#x201C; patients after transplantation, they have a 3-4 year OS probability ranging from 25% to 34%.24,25 Interestingly, a study by the Acute Leukemia Working Party of the European Society for Blood and Marrow Transplantation observed relatively comparable relapse incidence (HR=1.3, 95% CI: 0.7-2.6) and OS (HR=0.9, 95% CI: 0.5-1.6) in MK+ patients given grafts after reduced-intensity or myeloablative conditioning.25 This suggests that cure of MK+ patients after allogeneic HSCT might depend on immune-mediated graft-versustumor effects rather than on the intensity of the conditioning regimen.26 In summary, this retrospective analysis of two large prospective phase III trials confirmed the poor outcome of

B

D

Figure 2. Impact of cytogenetic risk on overall survival and impact of randomization in patients with a monosomal karyotype. (A) Overall survival (OS) according to cytogenetic risk group. (B) OS from complete remission/complete remission with incomplete blood count recovery according to cytogenetic risk group. (C) OS according to randomized induction therapy in the AML-10 trial among patients with a monosomal karyotype. (D) OS according to randomized induction therapy in the AML-12 trial among patients with a monosomal karyotype. MK: monosomal karyotype; 95% CI: 95% confidence interval; NotAdvMKâ&#x20AC;&#x201C;: not adverse cytogenetics excluding a monosomal karyotype; AdvMKâ&#x20AC;&#x201C;: adverse cytogenetics excluding a monosomal karyotype; MK+: monosomal karyotype.; DNR: daunorubicin; IDA: idarubicin; MTX: Methotrexate.

haematologica | 2019; 104(6)

1173

F. Baron et al.

Table 3. Association between monosomal karyotype and Medical Research Council adverse-risk group and overall survival.

Covariate

HR

P value

95% CI

Table 4. Association between monosomal karyotype and Medical Research Council adverse-risk group and overall survival from complete remission with or without hematologic recovery.

Covariate

Unadjusted analysis^ Cytogenetic group NotAdvMKAdvMKMK+

<0.001 1 1.48 1.26-1.74 2.58 2.19-3.04 Multivariate analysis*

Cytogenetic group NotAdvMK1 AdvMK1.51 MK+ 2.71 Age (years) 15-25 1 26-45 1.05 46-60 1.44 WHO Performance Status at baseline 0 1 1 1.09 2-4 1.76 WBC at diagnosis (x109/L) <25 1 ≥ 25 and < 100 1.08 ≥ 100 1.42

<0.001 1.28-1.77 2.29-3.20 <0.001 0.86-1.29 1.18-1.75 <0.001 0.97-1.23 1.50-2.07 <0.001 0.95-1.22 1.20-1.68

HR: hazard ratio; 95% CI: 95% confidence interval; NotAdvMK : not adverse cytogenetic excluding a monosomal karyotype; AdvMK–: adverse cytogenetic excluding a monosomal karyotype; MK+: monosomal karyotype; WHO: World Health Organization; WBC: white blood cell count. ^Obtained with a logistic regression model including protocol (AML-10 vs. AML-12) and cytogenetic group. *Obtained with a logistic regression model including protocol and all covariates presented in the Table.

–

HR

95% CI

P value

Unadjusted analysis^ Cytogenetic group NotAdvMKAdvMKMK+

<0.001 1 1.50 1.22-1.85 2.87 2.27-3.63 Multivariate analysis*

Cytogenetic group NotAdvMK1 AdvMK1.52 MK+ 2.95 Age (years) 15-25 1 26-45 1.05 46-60 1.46 WHO Performance Status at baseline 0 1 1 1.06 2-4 1.43 WBC at diagnosis (x109/L) <25 1 ≥ 25 and < 100 1.00 ≥ 100 1.30

<0.001 1.23-1.88 2.32-3.74 <0.001 0.82-1.35 1.15-1.86 0.004 0.92-1.23 1.16-1.78 0.069 0.85-1.17 1.03-1.64

HR: hazard ratio; 95% CI: 95% confidence interval; NotAdvMK–: not adverse cytogenetic excluding a monosomal karyotype; AdvMK–: adverse cytogenetic excluding a monosomal karyotype; MK+: monosomal karyotype; WHO: World Health Organization; WBC: white blood cell count. ^Obtained with a logistic regression model including protocol (AML-10 vs. AML-12) and cytogenetic group. *Obtained with a logistic regression model including protocol and all covariates presented in the Table.

A

B

1174

Figure 3. Impact of allogeneic hematopoietic stem cell transplantation in patients with a monosomal karyotype. (A) Overall survival from complete remission/complete remission with incomplete blood count recovery according to donor availability among patients with a monosomal karyotype. (B) Overall survival from hematopoietic stem cell transplantation by type of transplant among patients with a monosomal karyotype. HR: hazard ratio; 95% CI: 95% confidence interval; CR/Cri: complete remission/complete remission with incomplete blood count recovery; HSCT: hematopoietic stem cell transplantation; Auto-; autologous; RD allo-: related donor allogeneic.

haematologica | 2019; 104(6)

MK in EORTC/GIMEMA AML-10&12

MK+ patients. We found no evidence that the outcome was affected by the type of remission-induction chemotherapy. Donor availability was associated with prolonged survival among patients who reached a CR/CRi, suggesting a positive effect of allogeneic HSCT in this population of patients. These findings highlight the need for further prospective studies assessing the best strategy to bring MK+ patients to an allogeneic HSCT. In addition, efforts should be made to prevent post-transplant relapse and decrease transplant-related mortality in

References 9.

1. Willemze R, Suciu S, Meloni G, et al. Highdose cytarabine in induction treatment improves the outcome of adult patients younger than age 46 years with acute myeloid leukemia: results of the EORTCGIMEMA AML-12 Trial. J Clin Oncol.

2014;32(3):219-228.

2. Mandelli F, Vignetti M, Suciu S, et al.

Daunorubicin versus mitoxantrone versus idarubicin as induction and consolidation chemotherapy for adults with acute myeloid leukemia: the EORTC and GIMEMA groups study AML-10. J Clin Oncol. 2009;27(32):5397-5403.

3. Burnett AK, Russell NH, Hills RK, et al.

Optimization of chemotherapy for younger patients with acute myeloid leukemia: results of the Medical Research Council AML15 trial. J Clin Oncol. 2013;31(27):33603368.

4. Walter RB, Othus M, Burnett AK, et al.

Resistance prediction in AML: analysis of 4601 patients from MRC/NCRI, HOVON/SAKK, SWOG and MD Anderson Cancer Center. Leukemia. 2015;29(2):312320.

5. Dohner H, Estey EH, Amadori S, et al.

Diagnosis and management of acute myeloid leukemia in adults: recommendations from an international expert panel, on behalf of the European LeukemiaNet. Blood.

2010;115(3):453-474.

6. Grimwade D, Hills RK, Moorman AV, et al.

Refinement of cytogenetic classification in acute myeloid leukemia: determination of prognostic significance of rare recurring chromosomal abnormalities among 5876 younger adult patients treated in the United Kingdom Medical Research Council trials.

Blood. 2010;116(3):354-365.

7. Suciu S, Mandelli F, de Witte T, et al.

Allogeneic compared with autologous stem cell transplantation in the treatment of patients younger than 46 years with acute myeloid leukemia (AML) in first complete remission (CR1): an intention-to-treat analysis of the EORTC/GIMEMAAML-10 trial.

Blood. 2003;102(4):1232-1240.

8. Breems DA, van Putten WL, De Greef GE, et al. Monosomal karyotype in acute myeloid leukemia: a better indicator of poor progno-

haematologica | 2019; 104(6)

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

MK+ AML patients. One possible strategy to do this could be post-transplantation prophylactic administration of hypomethylating agents, given their anti-leukemic activity in MK+ patients27 and their ability to prevent graft-versus-host disease.28 Funding This publication was supported by a donation from the "Fondation Contre le Cancer" from "Belgium" through the EORTC Cancer Research Fund.

sis than a complex karyotype. J Clin Oncol.

2008;26(29):4791-4797.

Cornelissen JJ, Versluis J, Passweg JR, et al.

Comparative therapeutic value of postremission approaches in patients with acute myeloid leukemia aged 40-60 years.

Leukemia. 2014;29(5):1041-1050.

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

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

Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc. 1958;53:457-481.

Greenwood M. The natural duration of cancer. Rep Public Health Med Subj 1926;33:126.

Kalbfleisch JD, Prentice RL. The statistical analysis of failure time data. John Wiley, 2002.

Cox DR. Regression models and life tables (with discussion). J R Stat Soc Series B. 1972;34(2):187-220.

Medeiros BC, Othus M, Fang M, Roulston D, Appelbaum FR. Prognostic impact of monosomal karyotype in young adult and elderly acute myeloid leukemia: the Southwest Oncology Group (SWOG) experience. Blood. 2010;116(13):2224-2228.

Kayser S, Zucknick M, Dohner K, et al.

Monosomal karyotype in adult acute myeloid leukemia: prognostic impact and outcome after different treatment strategies.

Blood. 2012;119(2):551-558.

Yanada M, Kurosawa S, Yamaguchi T, et al.

Prognosis of acute myeloid leukemia harboring monosomal karyotype in patients treated with or without allogeneic hematopoietic cell transplantation after achieving complete remission. Haematologica. 2012;97(6):915918.

Cornelissen JJ, Breems D, van Putten WL, et al. Comparative analysis of the value of allogeneic hematopoietic stem-cell transplantation in acute myeloid leukemia with monosomal karyotype versus other cytogenetic risk categories. J Clin Oncol. 2012;30(17): 2140-2146.

Fernandez HF, Sun Z, Yao X, et al.

Anthracycline dose intensification in acute myeloid leukemia. N Engl J Med. 2009;361(13):1249-1259.

Lowenberg B, Ossenkoppele GJ, van Putten W, et al. High-dose daunorubicin in older patients with acute myeloid leukemia. N

Engl J Med. 2009;361(13):1235-1248.

21. Rucker FG, Schlenk RF, Bullinger L, et al.

TP53 alterations in acute myeloid leukemia with complex karyotype correlate with specific copy number alterations, monosomal karyotype, and dismal outcome. Blood.

2012;119(9):2114-2121.

22. Delfau-Larue MH, Klapper W, Berger F, et al.

High-dose cytarabine does not overcome the adverse prognostic value of CDKN2A and TP53 deletions in mantle cell lymphoma. Blood. 2015;126(5):604-611.

23. Cornelissen JJ, van Putten WL, Verdonck LF, et al. Results of a HOVON/SAKK donor versus no-donor analysis of myeloablative HLA-identical sibling stem cell transplantation in first remission acute myeloid leukemia in young and middle-aged adults: benefits for whom? Blood. 2007;109(9): 3658-3666.

24. Fang M, Storer B, Estey E, et al. Outcome of patients with acute myeloid leukemia with monosomal karyotype who undergo hematopoietic cell transplantation. Blood.

2011;118(6):1490-1494.

25. Poire X, Labopin M, Cornelissen JJ, et al.

Outcome of conditioning intensity in acute myeloid leukemia with monosomal karyotype in patients over 45 year-old: a study from the Acute Leukemia Working Party (ALWP) of the European Group of Blood and Marrow Transplantation (EBMT). Am J Hematol. 2015;90(8):719-724.

26. Baron F, Labopin M, Niederwieser D, et al.

Impact of graft-versus-host disease after reduced-intensity conditioning allogeneic stem cell transplantation for acute myeloid leukemia: a report from the Acute Leukemia Working Party of the European group for blood and marrow transplantation.

Leukemia. 2012;26(12):2462-2468.

27. Lubbert M, Suciu S, Hagemeijer A, et al.

Decitabine improves progression-free survival in older high-risk MDS patients with multiple autosomal monosomies: results of a subgroup analysis of the randomized phase III study 06011 of the EORTC Leukemia Cooperative Group and German MDS Study Group. Ann Hematol.

2016;95(2):191-199.

28. Fransolet G, Ehx G, Somja J, et al.

Azacytidine mitigates experimental sclerodermic chronic graft-versus-host disease. J Hematol Oncol. 2016;9(1):53.

1175

ARTICLE Ferrata Storti Foundation

Haematologica 2019 Volume 104(6):1176-1188

Acute Lymphoblastic Leukemia

Unraveling the cellular origin and clinical prognostic markers of infant B-cell acute lymphoblastic leukemia using genome-wide analysis

Antonio Agraz-Doblas,1,2 Clara Bueno,2# Rachael Bashford-Rogers,3# Anindita Roy,4,# Pauline Schneider,5 Michela Bardini,6 Paola Ballerini,7 Gianni Cazzaniga,6 Thaidy Moreno,1 Carlos Revilla,1 Marta Gut,8,9 Maria G. Valsecchi,10 Irene Roberts,4,11 Rob Pieters,5 Paola De Lorenzo,10 Ignacio Varela,1,$,* Pablo Menendez2,12,13,$,* and Ronald W. Stam5

Instituto de Biomedicina y Biotecnología de Cantabria (IBBTEC), Universidad de CantabriaCSIC, Santander, Spain; 2Josep Carreras Leukemia Research Institute-Campus Clinic, Department of Biomedicine, School of Medicine, University of Barcelona, Spain; 3 Department of Medicine, University of Cambridge, Cambridge Biomedical Campus, UK; 4 Department of Paediatrics, University of Oxford, UK; 5Princess Maxima Center for Pediatric Oncology, Utrecht, the Netherlands; 6Centro Ricerca Tettamanti, Department of Pediatrics, University of Milano Bicocca, Fondazione MBBM, Monza, Italy; 7Pediatric Hematology, A. Trousseau Hospital, Paris, France; 8CNAG-CRG, Center for Genomic Regulation, Barcelona, Spain; 9Universitat Pompeu Fabra, Barcelona, Spain; 10Interfant Trial Data Center, University of Milano-Bicocca, Monza, Italy; 11MRC Molecular Haematology Unit, MRC Weatherall Institute of Molecular Medicine, University of Oxford, UK; 12Instituciò Catalana de Recerca i Estudis Avançats (ICREA), Barcelona, Spain and 13Centro de Investigación Biomédica en Red de Cáncer (CIBERONC), ISCIII, Barcelona, Spain 1

These authors contributed equally to this work.

#

These senior authors contributed equally to this work.

$

ABSTRACT

Correspondence: PABLO MENÉNDEZ pmenendez@carrerasresearch.org IGNACIO VARELA Ignacio.varela@unican.es Received: September 7, 2018.

Accepted: December 20, 2018.

Pre-published: January 24, 2019. doi:10.3324/haematol.2018.206375 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/104/6/1176 ©2019 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

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

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

1176

B

-cell acute lymphoblastic leukemia is the commonest childhood cancer.

In infants, B-cell acute lymphoblastic leukemia remains fatal, especially in patients with t(4;11), present in ~80% of cases. The pathogenesis of t(4;11)/KMT2A-AFF1+ (MLL-AF4+) infant B-cell acute lymphoblastic leukemia remains difficult to model, and the pathogenic contribution in cancer of the reciprocal fusions resulting from derivative translocated-chromosomes remains obscure. Here, “multi-layered” genome-wide analyses and validation were performed on a total of 124 de novo cases of infant B-cell acute lymphoblastic leukemia uniformly diagnosed and treated according to the Interfant 99/06 protocol. These patients showed the most silent mutational landscape reported so far for any sequenced pediatric cancer. Recurrent mutations were exclusively found in K-RAS and N-RAS, were subclonal and were frequently lost at relapse, despite a larger number of non-recurrent/non-silent mutations. Unlike non-MLL-rearranged B-cell acute lymphoblastic leukemias, B-cell receptor repertoire analysis revealed minor, non-expanded B-cell clones in t(4;11)+ infant B-cell acute lymphoblastic leukemia, and RNAsequencing showed transcriptomic similarities between t(4;11)+ infant B-cell acute lymphoblastic leukemias and the most immature human fetal liver hematopoietic stem and progenitor cells, confirming a “pre-VDJ” fetal cellular origin for both t(4;11) and RASmut. The reciprocal fusion AF4-MLL was expressed in only 45% (19/43) of the t(4;11)+ patients, and HOXA cluster genes are exclusively expressed in AF4-MLL-expressing patients. Importantly, AF4-MLL/HOXA-expressing patients had a significantly better 4-year eventfree survival (62.4% vs. 11.7%, P=0.001), and overall survival (73.7 vs. 25.2%, P=0.016). AF4-MLL expression retained its prognostic significance when analyzed in a Cox model adjusting for risk stratification according to the Interfant-06 protocol based on age at diagnosis, white blood cell count and response to prednisone. This study has clinical implications for disease outcome and diagnostic risk-stratification of t(4;11)+ infant B-cell acute lymphoblastic leukemia. haematologica | 2019; 104(6)

Cellular origin and clinical prognostic markers of infant MLLr B-ALL

Introduction

Methods

B-cell precursor acute lymphoblastic leukemia (BCPALL) is the most frequent cancer in children.1 Current 5year survival rates in pediatric BCP-ALL approach 90%.

However, BCP-ALL in infants (iBCP-ALL; <1 year of age) remains clinically challenging with an aggressive early clinical presentation in uniquely vulnerable hosts.2 Approximately 80% of iBCP-ALL are diagnosed with chromosomal rearrangements involving the mixed-lineage leukemia (KMTA2, also called MLL) gene, located on 11q23,3–5 which confers a dismal prognosis especially in patients carrying the t(4;11)/KMT2A-AFF1+ (MLLAF4+).6–8 MLL is a H3K4 histone methyltransferase required for normal hematopoiesis and HOX gene expression.9,10 Leukemia transformation by MLL fusions requires the recruitment of the H3K79 histone methyltransferase Dot1L to the MLL transcriptional complex.11,12 Indeed, an H3K79 methylation profile defines both mouse and human t(4;11)/MLL-AF4+ BCP-ALL.13 Importantly, MLL rearrangements (MLLr) occur prenatally during embryonic/fetal hematopoiesis, and the concordance rate for iBCP-ALL in identical twins with a monochorionic placenta is close to 100%.14–17 This, coupled to the extremely short latency, suggests that MLL fusions might be sufficient for leukemogenesis.4 Accordingly, genome-wide studies using both single nucleotide polymorphism arrays and whole-genome sequencing revealed that MLLr iBCP-ALL has a very low frequency of somatic mutations with the predominant clone carrying ~1.3 non-silent mutations and one copy number alteration.18– 20 Although these studies were performed at low coverage sequencing they reinforce the concept that MLLr iBCP-ALL requires few additional mutations to induce full transformation. In contrast, MLL-AF4-induced leukemogenesis has proven difficult to model.4,9 With the exception of a recent work by Lin et al.21,22 who fused human MLL to murine Af4, creating an artificial leukemogenic human-mouse chimeric fusion, current murine and humanized models of MLL-AF4+ BCP-ALL do not faithfully recapitulate the disease pathogenesis/phenotype, suggesting that MLL-AF4 per se is insufficient to initiate leukemogenesis.23–28 The few mutations and copy number alterations present in MLLr iBCP-ALL seem subclonal and not always retained at relapse.20 Intratumor heterogeneity drives clonal evolution in response to microenvironmental cues and cytotoxic treatment and therefore recurrent mutations at diagnosis and relapse may be found in minor but clinically relevant subclones.29 Here we aimed to address the clinical relevance of subclonal mutations and gene expression signatures in a large cohort of iBCP-ALL. To do this, we performed deeper exome sequencing along with whole-genome DNA- and RNA-sequencing on a large cohort of 50 MLLr and non-MLL iBCP-ALL patients uniformly treated and followed up according to an Interfant treatment protocol.30 Similarly to Anderson et al.,20 we report a silent mutational landscape in iBCP-ALL irrespective of the MLL rearrangement/status. However, strikingly, our genome-wide DNA and RNA analyses revealed new, clinically relevant information about disease outcome and cell-of-origin for t(4;11) and RAS mutations.

Patients

haematologica | 2019; 104(6)

Bone marrow or peripheral blood samples from 124 infants (<12 months old) diagnosed with either pro-B or pre-B-cell ALL were used in this study. The discovery cohort of patients was composed of 42 de novo cases: 27 with the t(4;11) encoding for MLL-AF4, five with the t(9;11) encoding for KMT2A-MLLT3 (MLL-AF9) and ten without MLLr (non-MLL B-other BCP-ALL without numerical or structural chromosomal abnormalities reported at diagnosis). Additionally, for eight MLL-AF4+ iBCP-ALL patients matched diagnostic-relapse samples were available allowing for longitudinal studies. MLL rearrangements were confirmed by fluorescence in-situ hybridization.31,32 For validation, an additional cohort of patients, comprising 43 MLL-AF4+, 11 MLL-AF9+, and 28 non-MLL iBCP-ALL cases, was used. All patients were enrolled in the Interfant99 treatment study. Bone marrow samples were collected at Erasmus MC-Sophia Children’s Hospital (Rotterdam, the Netherlands), Armand Trousseau Hospital (Paris, France), and San Gerardo Pediatric Hospital (Monza, Italy).

Complete remission bone marrow samples were available for all patients. The clinical and genetic features of the patients are presented in Online Supplementary Table S1. As a control for the RNAsequencing studies, CD34+CD19+ healthy B-cell progenitors were purified by fluorescence-activated cell sorting (FACS) from 22week old human fetal livers (FL) as previously described.32 FL hematopoietic stem and progenitor cells (HSPC) were processed and FACS-purified from second trimester human FL as previously described.33 Briefly, cells were processed and stained for flow cytometry with up to ten fluorophore-conjugated monoclonal antibodies [antibodies (clone): CD34PECy7 (8G12), CD45RA FITC (HI100), CD19APC (HIB19), CD123PE (9F5), CD90 PECy5 (5E10), CD38 Pacific blue (HIT2), lineage cocktail APC (CD2 (RPA-2.10)/CD3 (OKT3)/CD14 (61D3)/ CD16(CB16)/ CD19 (HIB19)/CD56 (TULY56)/CD235a (HIR2)]. FACS was performed using a BD FACSAria II (Becton Dickinson). Gates were set with unstained and fluorescence minus one controls, on viable cells.

Data were analyzed using FlowJo software (Tree Star). Gating strategies are as described in the results section. The study was approved by the Barcelona Clinic Hospital (2013/8529) and Hammersmith and Queen Charlotte’s Hospital (04/Q0406/145) research ethics committees.

Statistical analysis For quantitative variables, a one-tailed t-test was used to identify significant differences between groups. For qualitative variables, a Fisher exact test was used in order to identify significant differences between groups of patients. Software for analysis of mutations and gene expression have their own statistical models explained in detailed in the references. Where multiple tests were performed the significance is shown corrected for multiple testing.

Mutation allele frequency evolution was plotted with the R package distribution Fishplot. Patterns Fisher exact test was used to assess the association between clinical characteristics and presence of RAS mutations or AF4-MLL expression. Event-free survival was defined as time from diagnosis to first event, i.e. resistance, relapse, death from any cause, or second malignant neoplasm. Observation periods were censored at the time of last contact when no events were reported. Event-free survival curves were estimated with the Kaplan-Meier method and standard errors (SE) were calculated according to Greenwood. Differences in event-free survival and overall survival between groups were compared with the log-rank test. Analysis of the prognostic relevance of AF4-MLL/HOXA expression in combination with risk

1177

A. Agraz-Doblas et al.

stratification according to the Interfant-06 protocol (based on age at diagnosis, white blood cell count and response to prednisone) was performed with the Cox model and the Wald test. All tests were two-sided. Analyses were performed using SAS 9.2.

DNA, RNA and B-cell receptor (VDJ) repertoire genome-wide analyses and data analysis Preparation and analysis of all DNA and RNA genome-wide high-throughput sequencing is detailed in the Online Supplementary Methods, Online Supplementary Figure S1 and Online Supplementary Table S2.

Results At diagnosis infant B-cell precursor acute lymphoblastic leukemia shows a silent mutational landscape irrespective of MLL gene status Whole-exome sequencing and whole-genome sequencing analyses showed a silent mutational landscape in the three iBCP-ALL subtypes studied here: MLL-AF4+, MLLAF9+ and non-MLL (n=42 patients, Online Supplementary Table S1). Our study revealed an average of one genomic rearrangement and 2.5 non-silent single nucleotide variants, a 2-fold higher number than that reported by

Andersson et al.,20 likely reflecting the 3-fold larger sequencing coverage (Figure 1A, Online Supplementary Figure S1 and Online Supplementary Table S3). All mutations found at diagnosis were validated using orthogonal methods. This mutational frequency is the lowest described for any other pediatric tumor type according to recent reports34 (Online Supplementary Figure S2). Intriguingly, one third of the mutations validated showed a mutant allele frequency (MAF) <20% indicating that iBCP-ALL contains genetically different intratumoral subclones despite its genomic stability, likely explaining the higher mutational load than that reported by Andersson et al.20 (Figure 1A and Online Supplementary Table S3). Despite the paucity of mutations, ~80% of the validated protein-coding mutations (90/116) are predicted to produce deleterious effects on the protein (Online Supplementary Figure S3A) which might support a strong selective pressure in iBCP-ALL. To gain insights into the molecular mechanisms underlying the accumulation of mutations, we analyzed the enrichment of specific mutational signatures as described by Alexandrov et al.35 In the MLL-AF4+ iBCP-ALL subgroup we identified a significant enrichment of signature 1 characterized by the accumulation of C>T/G>A transitions, linked to a spontaneous deamination of 5-methylcytosine (Online Supplementary Figure S3B,C).35 This mutational sig-

A

B

Figure 1. Somatic mutations detected by whole-exome sequencing in the discovery cohort of infant B-cell precursor acute lymphoblastic leukemia. (A) Total number of mutations identified in each individual patient. The total number of non-synonymous mutations (yellow area, right Y axis) and mutant allele frequency (MAF) for each mutation (individual dots, left Y axis) are represented. (B) Oncodrive software identified the PI3K-RAS pathway as the only recurrently mutated pathway in infant B-cell precursor acute lymphoblastic leukemia. The distribution of mutations in genes of the PI3K-RAS pathway is shown for all patients within the three iBCPALL subgroups: [total 42 patients: 27 t(4;11)+, 5 t(9;11)+ and 10 MLLwt].

1178

haematologica | 2019; 104(6)

Cellular origin and clinical prognostic markers of infant MLLr B-ALL

nature has also been described in other pediatric tumors, suggesting that iBCP-ALL is not subjected to a specific mutational signature.

We also determined the molecular breakpoint of all MLLr at the base-pair level. In t(4;11)/MLL-AF4+ iBCPALL, the AF4 breakpoints were almost invariably localized within intron 3 whereas MLL breakpoints were found between introns 9 and 11 (Online Supplementary Table S4).3 We found whole-genome sequencing reads compatible with an AF4-MLL reciprocal rearrangement in all samples (Online Supplementary Figure S4). AF4-MLL genomic breakpoints were validated by polymerase chain reaction capillary sequencing and they were located nearby MLL-AF4 breakpoints, confirming a reciprocal chromosomal translocation.

RAS-PI3K is the only recurrently mutated pathway in infant B-cell precursor acute lymphoblastic leukemia with NRAS mutations being significantly more frequent in t(4;11)+ patients Despite the low number of mutations found per sample, 38% of the sequenced iBCP-ALL patients displayed activating/gain-of-function mutations in either KRAS or NRAS. Additional mutations in other genes members of the RAS-PI3K pathway such as FGFR4, JAK2, PTPN11, SETD2, or FLT3 were also identified (Figure 1B). To further validate the unique recurrence of KRAS and NRAS mutations, we performed targeted sequencing of these mutations in a large, additional validation cohort of infant patients (n=82) and confirmed that 34% of the iBCP-ALL cases carry mutations in either KRAS or NRAS36 (Figure 2A). Interestingly, the overall frequency of RAS mutations differed slightly between the different cytogenetic subgroups of iBCP-ALL, with the MLL-AF4+ subgroup showing the highest frequency (42%) and the MLL-AF9+ subgroup the lowest (19%). This difference was basically attributed to the frequency of NRAS mutations, which was 6-fold more common in the MLL-AF4+ subgroup

(32% vs. 6%, Fisher exact test P=0.01) (Figure 2A,B).

Surprisingly, we observed that many iBCP-ALL patients had mutations in both KRAS and NRAS, or more than one (different) mutation in the same gene (Figure 2A,C). To further analyze the biological contribution of KRAS and NRAS mutations, we calculated the MAF of individual mutations and observed that the majority of patients who had a single RAS mutation (either KRAS or NRAS) had MAF scores between 0.20 and 0.45, suggesting that the mutation is present in a major leukemic subclone (P=0.0025). By contrast, those patients harboring two or more RAS mutations displayed MAF scores between 1% and 20%, compatible with these RAS mutations being in distinct and smaller leukemic subclones. We then analyzed the impact of RAS mutations on disease outcome and found no clinical correlation of RAS mutations with either clinical outcome (overall survival, event-free survival, central nervous system infiltration) or diagnostic parameters (gender, age, percentage of blasts and white blood cells) (Online Supplementary Figure S5).

Evidence of clone selection and genomic instability at relapse Paired diagnostic-relapse samples were available for eight MLL-AF4+ iBCP-ALL patients, permitting longitudinal studies. Whole-exome sequencing revealed an 8-fold increase in the number of somatic non-synonymous mutations at relapse (19.5 mutations/patient, range:1-434, paired t-test P=0.03) (Figure 3A,B and Online Supplementary Table S3). We performed orthogonal validation for 160 random mutations, and 90% and 75% of mutations with MAF >15% and <15%, respectively, were confirmed (data not shown). Similarly to diagnosis, the majority of the somatic mutations found at relapse had MAF commonly <30%, suggesting the existence of multiple leukemic subclones (Figure 3A). Importantly, none of the new de novo somatic mutations found at relapse was found in more than one patient, likely reflecting an intrinsic genomic

A

B

C

Figure 2. Frequent somatic mutations in RAS genes in both the discovery and validation cohorts. (A) Specific KRAS and NRAS mutations recurrently found in each patient by high coverage targeted sequencing. (B) Proportions of patients with mutations in KRAS (brown), NRAS (yellow) or both (gray) within the three infant B-cell precursor acute lymphoblastic leukemia subgroups. (C) Mutant allele frequency of KRAS (brown squares) or NRAS (yellow circles) mutations in each individual patient. Discovery cohort, n=42; validation cohort, n=82.

haematologica | 2019; 104(6)

1179

A. Agraz-Doblas et al.

instability of leukemic clones surviving induction/consolidation chemotherapy. This is further reflected by a significant enrichment of signature 6 associated with defective DNA mismatch repair, including a higher number of small indels, observed in MLL-AF4+ patients at relapse (Online Supplementary Figure S6A).

To delineate the evolutionary clonal structure from diag-

nosis to relapse, we performed high-coverage targeted sequencing on the identified mutations in paired diagnostic-remission-relapse samples.37 Importantly, the main leukemic clone at relapse was always present at diagnosis although in some cases with a very low MAF, suggesting a chemotherapy-induced clonal pressure selecting for resistant/adapted leukemic subclones (Figure 3C).

A

B

C

D

1180

Figure 3. Clonal evolution and genomic instability at relapse. (A) Total number of mutations identified for each patient in paired diagnosticrelapse samples. Total number of non-synonymous mutations (yellow area, right Y axis) and mutant allele frequency (MAF) for each mutation (individual dots, left Y axis) are represented for paired diagnostic and relapsed (R) samples. (B) Circos plot representation of the total number of mutations identified at diagnosis and relapse for a representative patient (MA4_17). Genomic rearrangements are represented by lines connecting both breakpoints. Copy number alterations (blue=gains, red=losses) are represented by the outer gray circle.

Somatic mutations (both single nucleotide variants and indels) are depicted in the center of the circle and the affected gene is indicated. (C) Graphic representation of clonal evolution in paired diagnostic (DX)relapsed (RL) samples. The number of unique somatic mutations called at diagnosis (orange), relapse (yellow) or shared between DX and REL (red) are indicated. Bigger gene names indicate higher MAF for the mutations shared at DX and REL. (D) Dynamics of RASmutated clones identified as MAF in matched DX-Remission-REL trios (n=8).

haematologica | 2019; 104(6)

Cellular origin and clinical prognostic markers of infant MLLr B-ALL

Interestingly, we found a correlation between the number of mutations and time to relapse in MLL-AF4+ patients, with a trend towards a higher mutational load in patients with late relapses (Online Supplementary Figure S6B). We next analyzed the clonal evolution of RAS-mutated leukemic clones at relapse. We found that the contribution

A

of the RAS mutations varied among patients: one-third of the iBCP-ALL patients had RAS-mutated clones at relapse (MA4_20 and MA4_22 increased the size of the RASmutated initial clone and in MA4_14 a de novo RAS mutation emerged), whereas it was lost in two-thirds of the patients (MA4_17, MA4_18, MA4_23, MA4_24) (Figure

B

C

D

E

haematologica | 2019; 104(6)

Figure 4. Transcriptional signature of infant B-cell precursor acute lymphoblastic leukemia samples. (A) Heatmap representing FLT3, PROM1, MEIS1 and HOXA gene expression according to the infant B-cell precursor acute lymphoblastic leukemia (iBCP-ALL) cytogenetic group and RAS mutations. (B) Top panel: heatmap showing HOXA cluster gene expression according to the expression of the reciprocal fusion AF4-MLL. Bottom panel: quantitative polymerase chain reaction validating high expression of HOXA cluster genes in t(4;11) iBCP-ALL patients expressing AF4-MLL. (C,D) Four-year event-free survival (C) and overall survival (D) Kaplan-Meier curves for t(4;11) iBCPALL patients according to AF4-MLL expression, n=43 t(4;11)+ patients.

(E) Heatmap representation of selected genes for the signaling pathways most significantly deregulated. Right panels represent positive pathway enrichment called by gene set enrichment analysis software. Total 42 patients: 27 t(4;11)+, 5 t(9;11)+ and 10 MLLwt.

1181

A. Agraz-Doblas et al.

3D). This indicates that infants with MLL-AF4+ BCP-ALL relapse irrespective of the status of KRAS and NRAS. Thus, subclones carrying KRAS mutations do not exert an advantage over non-mutated clones, despite representing a recurrent genetic insult at diagnosis. Hence, this would argue against a leukemia-initiating role for RAS mutations.38 Alternatively, RAS mutations might indeed be leukemogenic drivers, but the treatment-induced genetic instability observed at relapse may compensate de novo RAS mutations, acting as new leukemia drivers cooperating with MLL-AF4 during relapse.

HOXA cluster genes are only expressed in t(4;11)+ patients expressing the reciprocal fusion AF4-MLL which determines clinical outcome To gain insights into the mechanisms underlying leukemogenesis in these mutationally silent MLLr and MLL germline iBCP-ALL patients, we performed RNAsequencing in the discovery cohort of patients (n=42) using FL-derived CD34+CD19+ healthy B-cell progenitors as controls, as these cells most likely represent the healthy counterparts of the leukemic blast stalled at the pro/pre-Bcell differentiation stage. We first surveyed the expression of the genes previously reported to be specific to either MLLr iBCP-ALL or specifically to MLL-AF4+ iBCP-ALL39.

RNA-sequencing profiling confirmed that these genes segregate patients according to the molecular subtype, MLLAF4+, MLL-AF9+ and MLL germline (Online Supplementary Figure S7). We also observed, at diagnosis, a strong upregulation of the MLL target genes FLT3,40 MEIS1, PROM1 and HOXA genes in many of our MLLr iBCP-ALL samples but not in MLL germline samples (t-test, P<0.05) (Figure 4A), thus validating our RNA-sequencing approach.

Strikingly, the reciprocal AF4-MLL fusion gene was discernibly expressed in 19/43 (45%) of the t(4;11)+ iBCPALL samples, and its expression was always maintained at relapse (data not shown). We then compared the genes differentially expressed between AF4-MLL-expressing and non-expressing t(4;11)+ patients and found a striking positive correlation between the expression of the HOXA gene cluster and overexpression of the reciprocal AF4-MLL fusion (t-test, P=0.002) (Figure 4B). These AF4MLL/HOXA-expressing patients (n=19) had a significantly better prognosis than those lacking AF4-MLL/HOXA expression (n=24). Four-year event-free and overall survival rates were 62.4% (SE, 11.3%) versus 11.7% (SE, 10.2%) (P=0.001) (Figure 4C), and 73.7% (SE, 10.1%) versus 25.2% (SE, 10.3%) (P=0.016) (Figure 4D), respectively.

When â&#x20AC;&#x153;AF4-MLL expressionâ&#x20AC;? was analyzed in a Cox model adjusting for risk stratification (medium risk or high risk according to the Interfant-06 protocol based on age at diagnosis, white blood cell count and response to prednisone), it retained its prognostic significance with a hazard ratio for patients lacking AF4-MLL expression of 3.42 [95% confidence interval (95% CI): 1.35-8.63; P=0.01) compared to those expressing AF4-MLL/HOXA, while risk group was not significant (HR for high risk vs. medium risk, 1.34; 95% CI: 0.59-3.03; P=0.49). This is the first study showing that AF4-MLL overexpression correlates very well with transcriptional deregulation of the HOXA gene cluster in iBCP-ALL and that the co-expression of AF4-MLL and HOXA gene cluster identifies a subgroup of t(4;11)+ iBCP-ALL with a very more favorable clinical outcome.

We next explored new molecular pathways involved in 1182

the pathogenesis of iBCP-ALL, by performing an unbiased transcriptional analysis of the RNA-sequencing data from the iBCP-ALL patients. We found deregulated expression of a total of 3,905 genes, of which 2,575 (66%) were upregulated and 1,330 (34%) downregulated as compared with those of healthy FL-derived B-cell progenitors, illustrating the global transcriptional activation nature of MLL fusions (Online Supplementary Figure S8).25,41 Furthermore, a significant upregulation of genes involved in the control of cell growth, including the CDK inhibitors P21, P16, P19, P27 and components of the transforming growth factor-b pathway such as TGFB1, SMAD and ACVR1B, was observed in iBCP-ALL (Figure 4D and Online Supplementary Figure S9). By contrast, iBCP-ALL showed a robust downregulation of genes involved in DNA integrity checkpoints such as CHEK1, CHEK2, ATM, ATR and RAD17, and in double-strand break repair genes including ERCC4, BRCA1, POLA1 and RAD51 (Figure 4E and Online Supplementary Figure S9). These transcriptional changes were validated by quantitative reverse transcriptase polymerase chain reaction in ten patients per group (Online Supplementary Figure S10). Deregulation of DNA integrity checkpoints and double-strand break repair genes may well contribute to the genomic instability observed at relapse, and might explain the enrichment in C>T/G>A transitions, associated with the spontaneous deamination of 5-methylcytosine (Online Supplementary Figures S3 and S6).

By using FL-derived normal B-cell progenitors as controls, differences between leukemic blasts and their normal counterparts could be identified but this does not allow the definition of transcriptomic differences within the iBCP-ALL cytogenetic groups. We, therefore, analyzed the RNA-sequencing data comparing the genes differentially expressed in MLL-AF4+ versus MLL-AF9+ and MLLwildtype iBCP-ALL patients, without considering normal B-cell progenitors as controls. A gene ontology analysis (gene set enrichment analysis, GSEA) performed with the genes differentially expressed revealed that MLL-AF4+ patients show, as compared to both MLL-AF9+ and MLLwildtype patients, a significant upregulation of genes associated with cellular catabolism, coupled to a significant downregulation of negative regulators of the PI3-MAPK pathway, as well as of genes involved in lymphoid differentiation and RNApol II transcriptional regulation (Figure 5). This suggests, respectively, a metabolic change in MLLAF4+ cells towards rapid energy generation while reinforcing the basal hyperactivation of the PI3-MAPK pathway

by RAS mutations (Figures 1 and 2), a poorly differentiated cellular origin of t(4;11), and an impairment of the normal function of AF4, a key component of the RNApol II transcription complex.

Deep-sequencing analysis of B-cell receptor repertoires suggests a hematopoietic stem cell/early pre-VDJ progenitor as the cell-of-origin for t(4;11) and RAS mutations We next analyzed BCR repertoires to gain insights into the immunoglobulin heavy chain (IgH) rearrangement clonal composition of paired diagnostic-relapse samples from t(4;11)+ iBCP-ALL (4 pairs). BCR are generated through DNA recombination during B-cell differentiation and represent unique markers for each B-cell clone.

Because the BCR sequence provides a molecular tag for each B-cell clone, high-throughput sequencing of BCR haematologica | 2019; 104(6)

Cellular origin and clinical prognostic markers of infant MLLr B-ALL

provides a detailed analysis of B-cell population dynamics and clone tracking.42,43 BCR sequencing was therefore performed to address whether t(4;11)+ iBCP-ALL cells expressed fully rearranged BCR from which increased levels of B-cell clonal expansion may be observed and to determine whether there are detectable levels of B-cell clonal persistence over time indicative of B-cell clonal survival. BCR sequencing was performed on t(4;11)/MLLAF4+ iBCP-ALL peripheral blood samples (blasts >98%) using a polymerase chain reaction-based method37 with additional incorporation of unique molecular barcodes, allowing for accurate quantitation of relative B-cell clone frequency. After BCR sequence filtering, each sample yielded between 1,583-46,863 BCR (1,213-38,426 unique BCR) (Online Supplementary Table S5).

We first delineated the relative clonality in these patients, and found that the BCR repertoires from t(4;11)+ patients did not exhibit significantly expanded VDJrearranged B-cell clones (Figure 6A) either at diagnosis or relapse compared to healthy peripheral blood samples

(Figure 6B). This is in contrast to non-MLL BCP-ALL patients (n=5) including three patients with t(1;19)/TCF3PBX1 (EF2-PBX1), one with t(12;21)/ETV6-RUNX1 (TELAML1) and one with t(9;22)/BCR-ABL1, which were all found to be significantly clonal, with large B-cell clones comprising ~3-40% of total BCR (Figure 6B,C).37 Given the persistence of both t(4;11) and RAS mutations in MLLAF4+ iBCP-ALL, the lack of B-cell clonal expansion or persistence supports the model that t(4;11)/MLL-AF4+ iBCPALL malignant cells are developmentally stalled at the proB stage, and that the cellular origin of such genomic drivers has to be pre-VDJ stem/progenitor cells.

Finally, in order to understand whether the fetal cell-oforigin in iBCP-ALL lies upstream of committed B progenitors, we compared the transcriptome of iBCP-ALL blasts (n=42) with that of highly purified human FL HSPC populations (3-7 for each population) (Figure 7A,B and Online Supplementary Table S6) by RNA-sequencing. In keeping with the results of the BCR analysis, our principal component analysis revealed a gene expression signature for

Figure 5. Specific transcriptional differences between MLL-AF4+ and MLL-AF9+ or MLLwt infant Bcell precursor acute lymphoblastic leukemia patients. Here, FLderived CD34+CD19+ progenitors were not included as normalizers in the analysis in order to avoid potential bias. Gene set enrichment analysis (GSEA) was performed with the genes differentially expressed between MLLAF4+ patients and MLL-AF9+ or MLLwt patients. MLL-AF4+ infant B-cell precursor acute lymphoblastic leukemia (iBCP-ALL) patients showed a significant overexpression of genes associated with cellular catabolism, coupled to a significant downregulation of negative regulators of the PI3-MAPK pathway, as well as of genes involved in lymphoid differentiation and RNApol II transcriptional regulation as compared to both MLL-AF9+ and MLLwt iBCP-ALL patients. The bottom panels represent positive pathway enrichment called by GSEA software. Total 42 patients: 27 t(4;11)+, 5 t(9;11)+ and 10 MLLwt.

haematologica | 2019; 104(6)

1183

A. Agraz-Doblas et al.

primitive Lin-CD34+CD38-CD19- FL HSPC (hematopoietic stem cells, multipotent progenitors, and lymphoid-primed multipotent progenitors, which lie upstream of B progenitors) very similar to t(4;11)+ iBCP-ALL, while FL-committed B progenitors clustered as a transcriptionally different entity (Figure 7C).

Discussion We set out to perform multi-layered sequencing on a large cohort iBCP-ALL patients, all enrolled in the international, collaborative Interfant treatment protocol. The fact that all patients were identically treated provides legitimacy and confidence in potential correlations of clinical value. Our study revealed an average of 2.5 non-silent single nucleotide variants, a 2-fold higher number than that reported by Andersson et al.,20 likely reflecting the 3-fold larger sequencing coverage. This silent mutational landscape, even in non-MLL iBCP-ALL, likely reflects the very young age of these patients, reinforcing the notion that infant cancer is a developmental disease with not enough time to develop somatic mutations. We also found the only recurrent, but subclonal, mutations occur in the KRAS and NRAS genes (gain-of-function mutations), although the frequency of subclonal NRAS mutations is

A

C

1184

significantly higher in t(4;11)+ patients. In line with our previous work we found no recurrent mutations in the FLT3 gene.40 Analysis of clonal evolution of RAS-mutated clones from diagnosis to relapse revealed that one-third of the patients still carry RAS mutations at relapse, whereas the other two-thirds of patients who relapse have lost the diagnostic RAS mutation. This is in accordance with recently published data by Trentin et al.,36 and suggests that the therapy is able to eliminate the RAS-mutated clone in some patients, while in other patients the RAS mutation seems to confer chemoresistance, allowing these clones to evade treatment.44 Intriguingly, ~25% of the patients carry more than one RAS-mutated clone at diagnosis, indicating a selection bias towards mutations in the RAS genes, or activated RAS pathways during leukemic transformation. From this perspective, the occurrence of patients carrying multiple distinct clones with activated RAS pathways may point to convergent evolution of clones capable of controlling the proliferation rate.

However, arguing against this is the substantial representation of patients not carrying RAS mutations at all.

Hence, the role of RAS mutations in t(4;11)+ iBCP-ALL remains obscure, and the available data suggest that RAS pathway mutations are unlikely leukemia-initiating lesions. Indeed, Tamai et al.45 showed that leukemogenesis

B

Figure 6. Analysis of B-cell receptor repertoires suggest a hematopoietic stem cell/early pre-VDJ progenitor as the cellof-origin for t(4;11)/MLL-AF4+ infant B-cell precursor acute lymphoblastic leukemia. (A) Cloud-plots of B-cell receptor (BCR) repertoires from two representative t(4;11)+ infant B-cell precursor acute lymphoblastic leukemia (iBCP-ALL) patients depicting the existence of many minor non-expanded B-cell clones either at diagnosis or relapse. Each vertex represents a unique BCR sequence, and the relative vertex size is proportional to the number of identical reads. (B) Largest BCR clone size in t(4;11)+ iBCP-ALL, healthy individuals and nont(4;11)+ pediatric BCP-ALL. (C) Cloud-plots of BCR repertoires of representative t(1;19)/E2APBX1+, t(12;21)/TEL-AML1+ and patients t(9;22)/BCR-ABL+ showing high clonality of B-cell clones. The samples from the iBCP-ALL patients who were BCR-sequenced were four MLLAF4+ diagnostic-relapse pairs, three E2A-PBX1+ samples, one TEL-AML1+ sample and one BCR-ABL+ sample.

haematologica | 2019; 104(6)

Cellular origin and clinical prognostic markers of infant MLLr B-ALL

of transgenic mice expressing human MLL-AF4 could be significantly accelerated by KRAS mutations. However, although activated KRAS did cooperate with MLL-AF4 in human cord blood-derived CD34+ HSPC to promote extramedullary infiltration and central nervous system infiltration it failed to initiate leukemia in engrafted mice.27 Importantly, we report a lack of correlation between RAS

status and parameters associated with diagnosis or disease outcome such as overall survival, event-free survival, central nervous system infiltration, gender, percentage of blasts and white blood cells and age, further supporting the concept that RAS mutations are not leukemia-initiating/propagating lesions.

Clearly, this brings us back to the central question of

A

B

C

Figure 7. Comparison of the transcriptome of human fetal CD34+ hematopoietic stem and progenitor cell populations to infant B-cell precursor acute lymphoblastic leukemia. (A) Schematic representation of B-cell development in human fetal liver (FL) showing immunophenotypic definitions for hematopoietic stem cells (HSC), multipotent progenitors (MPP), lymphoidprimed multipotent progenitors (LMPP), committed B progenitors (CBP) and B cells. The onset and expected patterns of IgH rearrangements59,60 are depicted as red arrows. (B) Sorting strategy for FL hematopoietic stem and progenitor (HSPC) populations by fluorescence-activated cell sorting. The sorting gates for each population are shown in representative flow plots on the left. The purity of the sorted populations is depicted on the right demonstrating >95% purity. (Lin, Lineage cocktail). (C) Principal component analysis of gene expression of infant B-cell precursor acute lymphoblastic leukemia (iBCP-ALL) samples (n=42) and FL HSPC populations (n=3-7) using the top 1,000 variably expressed genes, as determined by RNA-sequencing. FL HSPC as in (A); MAF4, MLL-AF4+ iBCP-ALL; MA9, MLL-AF9+ iBCP-ALL; MLLwt, MLL wildtype iBCP-ALL.

haematologica | 2019; 104(6)

1185

A. Agraz-Doblas et al.

whether or not MLL-AF4 by itself is sufficient to initiate BCP-ALL in humans. The silent mutational landscape observed in this study and by others20 certainly votes in favor of MLL-AF4+ iBCP-ALL being initiated by a single “big-bang” transformation hit, probably in a short-lived but highly proliferative prenatal B-cell progenitor.4 This hypothesis is supported by recent work by Lin et al., who indeed demonstrated that enforced expression of a fusion transcript consisting of human MLL and murine Af4 in cord blood-derived CD34+ HPSC is sufficient to induce pro-B ALL in xenografted immunodeficient mice.21,22 Yet, similar results using a human MLL-AF4 transcript remain to be established.

Although MLL-AF4 by itself may be sufficient to induce BCP-ALL without significant contributions from cooperative genetic lesions, the contribution of the MLL-AF4 and RAS mutations to leukemogenesis should take into account the nature of both the fetal target cell for transformation and the leukemia-initiating cell, according to the increasingly accepted stochastic stem cell model of BALL.46,47 Here, we employed high-throughput BCRsequencing of the IgH locus to delineate the dynamics of clonality of B-cell populations in paired diagnosis-relapse samples of t(4;11)/MLL-AF4+ iBCP-ALL. While pediatric patients with E2A-PBX1+, TEL-AML1+ and BCR-ABL1+ BALL all had significantly clonal disease, with a major VDJ rearranged B-cell IgH clone accounting for up to 40% of all BCR, infants with MLL-AF4+ BCP-ALL exhibited a BCR repertoire composed of thousands of minor, non-expanded VDJ rearranged IgH B-cell clones. Because MLL fusions are clonal and RAS mutations are found in clones of relative big size, this suggests that MLL fusions with or without RAS mutations are likely to originate in primitive fetal progenitors that have a germline or an incompletely rearranged (DJ) IgH locus.48 Indeed, an unsupervised comparison of the transcriptome of FL HSPC populations and iBCP-ALL blasts suggests that while the gene expression of primitive FL HSPC (Lin-CD38-CD34+CD19- populations) is similar to that of iBCP-ALL, FL B progenitors (CD34+CD19+) are transcriptionally distinct. Our data elegantly reinforces previous fluorescence in-situ hybridization findings suggesting that a primitive “pre-VDJ” stem/progenitor cell (perhaps CD34+CD19-) may represent the cell in which both t(4;11) and RAS mutations arise.14,31,49 Cooperative leukemogenic events in iBCP-ALL may need to be sought beyond genetic insults; for instance, epigenetic and transcriptomic deregulation. MLL-AF4 might only induce BCP-ALL in cells that meet certain epigenetic and transcriptomic make-up criteria, either influenced by microenvironmental cues, or characteristic of the cell-oforigin.31 Indeed, lesions such as RAS mutations may contribute to disease pathogenesis only against certain intrinsic epigenetic or transcriptomic backgrounds present in the cell in which the MLL translocations occurred50,51. This is supported by the limited impact of RAS mutations in transcriptomic signatures associated with leukemia origin, development and pathogenesis, although this is likely due to the subclonal nature of RAS mutations.38 However, in line with the reported contribution of RAS mutations to extramedullary infiltration of MLLr BCP-ALL blasts,27 RAS-mutated patients displayed a transcriptomic signature associated with migration.

The functional and molecular contribution of the reciprocal fusion genes resulting from the derivative translocated chromosomes remains obscure in cancer. The AF41186

MLL genomic fusion was previously detected in 80-85% of t(4;11)+ patients.5,52 Our “multi-layered omics” approach allowed for the exact characterization of the t(4;11) molecular DNA/RNA break points and the identification of those patients expressing the reciprocal AF4MLL fusion. We now report that the AF4-MLL reciprocal fusion is expressed in only 50% of t(4;11)+ iBCP-ALL patients. Strikingly, there was a previously unrecognized and very significant positive correlation between the upregulation of the HOXA gene cluster and the expression of AF4-MLL. Of note, a recent study showed that approximately half of t(4;11)+ patients do not have an activated HOXA signature.44,53,54 Furthermore, in the recent MLL-Af4-induced B-ALL xenograft model MLLAf4 failed to bind to HOXA genes and therefore HOXA gene expression was not upregulated.21 This is experimentally supported by chromatin immunoprecipitationsequencing analysis performed in human embryonic stem cells transduced with MLL-AF4, AF4-MLL or both showing a significant enrichment of H3K79 methylated regions specifically associated with HOX-A cluster genes in double fusion-expressing hematopoietic derivatives, establishing a functional and molecular cooperation between MLL-AF4 and AF4-MLL fusions during human hematopoietic development (data not shown). Strikingly, AF4-MLL-expressing patients had a 5-fold longer eventfree survival and a 3-fold longer overall survival compared to t(4;11)+ iBCP-ALL patients lacking AF4-MLL expression, which is in line with previous reports suggesting that high HOXA gene expression is associated with improved survival and lower risk of relapse.22,39 Because the expression of AF4-MLL is not analyzed in routine molecular diagnosis, our “multi-layered omics” approach was critical to unraveling the association between AF4MLL and HOXA expression, thus identifying a novel subgroup of t(4;11)+ iBCP-ALL with better clinical outcome.

It is very important for routine diagnostic and clinical practice that when the expression of AF4-MLL was evaluated in a Cox model adjusting for risk stratification (medium risk or high risk according to the Interfant-06 protocol), it retained its prognostic significance.

Mechanistically, AF4-MLL contains the SET domain disrupted from its "specification domain", the N-terminal portion of MLL, which binds to MEN1 and LEDGF thus shaping the gene-targeting module of the MLL gene.

When AF4-MLL is expressed, the N-terminal portion is substituted by the AF4 N-terminus (AF4N) which is the crucial domain for binding to and strongly activating RNA polymerase II (RNAP II) for transcriptional elongation.

Thus, expression of A4M-MLL may induce robust RNAP II-dependent gene transcription by overwriting the elongation control process in a dominant fashion.55–58 We hypothesize that a likely function of AF4-MLL could be to prepare the ground for MLL-AF4 or other transcription factors to skew normal and leukemic hematopoietic cell fate decisions. This also explains why MLL-AF4, but not AF4-MLL, seems to be necessary in 100% of patients.

Despite being a developmental cancer, iBCP-ALL patients did not show reactivation of pluripotent or embryonic-like gene expression signatures as revealed by RNA-sequencing. Additional research is required to decipher the nature of the insults initiating MLLr iBCP-ALL, as so far we can only speculate on the data currently available. Whole-genome pyrosequencing will likely provide unique insights into the DNA methylome landscape of haematologica | 2019; 104(6)

Cellular origin and clinical prognostic markers of infant MLLr B-ALL

this mutationally silent iBCP-ALL. This study has clinical implications in the diagnostic risk-stratification of t(4;11)+ iBCP-ALL. Acknowledgments We would like to thank the Santander Supercomputing Service for IT support. The human fetal material was provided by the Joint MRC/Wellcome Trust (grant# MR/R006237/1) Human Developmental Biology Resource (http://hdbr.org). This work was supported by the European Research Council (CoG-2014646903 to PM; and StG-2014-637904 to IV), the Spanish Ministry of Economy and Competitiveness (SAF-SAF201343065 to PM and SAF2016-76758-R to IV), the Asociación

References 16.

1. Pui C-H, Evans WE. A 50-year journey to cure childhood acute lymphoblastic leukemia. Semin Hematol. 2013;50(3):185– 196.

2. Pui CH, Mullighan CG, Evans WE, Relling MV. Pediatric acute lymphoblastic leukemia: where are we going and how do we get there? Blood. 2012;120(6):1165–1174.

3. Meyer C, Burmeister T, Gröger D, et al. The MLL recombinome of acute leukemias in 2017. Leukemia. 2018;32(2):273–284.

4. Sanjuan-Pla A, Bueno C, Prieto C, et al.

Revisiting the biology of infant t(4;11)/MLLAF4+ B-cell acute lymphoblastic leukemia.

Blood. 2015;126(25):2676–2685.

5. Marschalek R. Mechanisms of leukemogenesis by MLL fusion proteins. Br J Haematol.

2011;152(2):141–154.

6. Ribeiro RC, Pui CH. Prognostic factors in childhood acute lymphoblastic leukemia.

Hematol Pathol. 1993;7(3):121–142.

7. Pieters R, Schrappe M, De Lorenzo P, 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.

8. Biondi A, Cimino G, Pieters R, Pui C-H. Biological and therapeutic aspects of infant leukemia. Blood. 2000;96(1):24–33.

9. Milne TA. Mouse models of MLL leukemia: recapitulating the human disease. Blood.

2017;129(16):2217–2223.

10. Nakamura T, Mori T, Tada S, et al. ALL-1 is a histone methyltransferase that assembles a supercomplex of proteins involved in transcriptional regulation. Mol Cell. 2002;10(5): 1119–1128.

11. Chen CW, Armstrong SA. Targeting DOT1L and HOX gene expression in MLLrearranged leukemia and beyond. Exp Hematol. 2015;43(8):673–684.

12. McLean CM, Karemaker ID, van Leeuwen F. The emerging roles of DOT1L in leukemia and normal development. Leukemia.

2014;28(11):2131–2138.

13. Krivtsov AV, Feng Z, Lemieux ME, et al.

H3K79 methylation profiles define murine and human MLL-AF4 leukemias. Cancer.

2009;14(5):355–368.

14. Greaves MF, Maia AT, Wiemels JL, Ford AM. Leukemia in twins: lessons in natural history. Blood. 2003;102(7):2321–2333.

15. Ford AM, Ridge SA, Cabrera ME, et al. In utero rearrangements in the trithorax-related

haematologica | 2019; 104(6)

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

Española Contra el Cáncer (AECC-CI-2015), FERO Foundation, and the ISCIII (PI14-01191) to CB. PM/IV also acknowledge financial support from The Obra Social La CaixaFundaciò Josep Carreras, The Inocente Inocente Foundation, Fundación Ramón Areces and The Generalitat de Catalunya (SGR330). IR was supported by a Programme Grant from Bloodwise (LLR 13001) and by the Oxford NIHR Biomedical Centre based at Oxford University Hospitals NHS Trust and University of Oxford. PM is an investigator of the Spanish Cell Therapy cooperative network (TERCEL). AR was supported by a Clinician Scientist Fellowship from Bloodwise (14041). This work was motivated by our patients and it honors the vital example given to us by the family of AMC.

oncogene in infant leukaemias. Nature.

1993;363(6427):358–360.

Gale KB, Ford AM, Repp R, et al.

Backtracking leukemia to birth: identification of clonotypic gene fusion sequences in neonatal blood spots. Proc Natl Acad Sci U S A. 1997;94(25):13950–13954.

Bueno C, Montes R, Catalina P, Rodríguez R, Menendez P. Insights into the cellular origin and etiology of the infant pro-B acute lymphoblastic leukemia with MLL-AF4 rearrangement. Leukemia. 2011;25(3):400– 410.

Bardini M, Galbiati M, Lettieri A, et al.

Implementation of array based wholegenome high-resolution technologies confirms the absence of secondary copy-number alterations in MLL-AF4-positive infant ALL patients. Leukemia. 2011;25(1):175–178.

Dobbins SE, Sherborne AL, Ma YP, et al. The silent mutational landscape of infant MLLAF4 pro-B acute lymphoblastic leukemia.

Genes Chromosom Cancer. 2013;52(10): 954–960.

Andersson AK, Ma J, Wang J, et al. The landscape of somatic mutations in infant MLLrearranged acute lymphoblastic leukemias.

Nat Genet. 2015;47(4):330–337.

Lin S, Luo RT, Ptasinska A, et al. Instructive role of MLL-fusion proteins revealed by a model of t(4;11) pro-B acute lymphoblastic leukemia. Cancer Cell. 2016;30(5):737–749.

Lin S, Luo RT, Shrestha M, Thirman MJ, Mulloy JC. The full transforming capacity of MLL-Af4 is interlinked with lymphoid lineage commitment. Blood. 2017;130(7):903– 907.

Yamamoto H. Successful sustained engraftment after reduced-intensity umbilical cord blood transplantation for adult patients with severe aplastic anemia. Blood. 2011;116(26): 6123–6132.

Montes R, Ayllón V, Prieto C, et al. Ligandindependent FLT3 activation does not cooperate with MLL-AF4 to immortalize/transform cord blood CD34+ cells. Leukemia.

2014;28(3):666–674.

Bueno C, Montes R, Melen GJ, et al. A human ESC model for MLL-AF4 leukemic fusion gene reveals an impaired early hematopoietic-endothelial specification.

Cell Res. 2012;22(6):986–1002.

Bueno C, Ayllón V, Montes R, et al. FLT3 activation cooperates with MLL-AF4 fusion protein to abrogate the hematopoietic specification of human ESCs. Blood.

2013;121(19):3867–3878.

Prieto C, Stam RWRW, Agraz-Doblas A, et

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

al. Activated KRAS cooperates with MLLAF4 to promote extramedullary engraftment and migration of cord blood CD34+ HSPC but is insufficient to initiate leukemia.

Cancer Res. 2016;76(8):2478–2489.

Krivtsov A V., Armstrong SA. MLL translocations, histone modifications and leukaemia stem-cell development. Nat Rev Cancer. 2007;7(11):823–833.

Varela I, Menendez P, Sanjuan-Pla A. Intratumoral heterogeneity and clonal evolution in blood malignancies and solid tumors.

Oncotarget. 2017;8(39):66742–66746.

Driessen E, Lorenzo P, M C, et al. Outcome of congenital acute lymphoblastic leukemia treated on the Interfant-99 protocol. Blood.

2009;114(18):3764–3768.

Menendez P, Catalina P, Rodríguez R, et al.

Bone marrow mesenchymal stem cells from infants with MLL-AF4+ acute leukemia harbor and express the MLL-AF4 fusion gene. J Exp Med. 2009;206(13):3131–3141.

Muñoz-López A, Romero-Moya D, Prieto C, et al. Development refractoriness of MLLrearranged human B cell acute leukemias to reprogramming into pluripotency. Stem Cell Reports. 2016;7 (4):602–618.

Roy A, Cowan G, Mead AJ, et al.

Perturbation of fetal liver hematopoietic stem and progenitor cell development by trisomy 21. Proc Natl Acad Sci U S A. 2012;109(43):17579–17584.

Gröbner SN, Worst BC, Weischenfeldt J, et al. The landscape of genomic alterations across childhood cancers. Nature. 2018;555 (7696):321-327.

Alexandrov LB, Nik-Zainal S, Wedge DC, et al. Signatures of mutational processes in human cancer. Nature. 2013;500(7463):415– 421.

Trentin L, Bresolin S, Giarin E, et al.

Deciphering KRAS and NRAS mutated clone dynamics in MLL-AF4 paediatric leukaemia by ultra deep sequencing analysis. Sci Rep. 2016;6:34449.

Bashford-Rogers RJM, Nicolaou KA, Bartram J, et al. Eye on the B-ALL: B-cell receptor repertoires reveal persistence of numerous B-lymphoblastic leukemia subclones from diagnosis to relapse. Leukemia.

2016;30(12):2312–2321.

Prelle C, Bursen A, Dingermann T, Marschalek R. Secondary mutations in t(4;11) leukemia patients. Leukemia.

2013;27(6):1425–1427.

Stam RW, Schneider P, Hagelstein JAP, et al.

Gene expression profiling-based dissection of MLL translocated and MLL germline

1187

A. Agraz-Doblas et al.

40.

41.

42.

43.

44.

45.

46.

1188

acute lymphoblastic leukemia in infants.

Blood. 2010;115(14):2835–2844.

Chillón MC, Gómez-Casares MT, LópezJorge CE, et al. Prognostic significance of FLT3 mutational status and expression levels in MLL-AF4 and MLL-germline acute lymphoblastic leukemia. Leukemia. 2012;26(11): 2360–2366.

Boelens JJ, Aldenhoven M, Purtill D, et al.

Outcomes of transplantation using various hematopoietic cell sources in children with Hurler syndrome after myeloablative conditioning Key Points. Blood. 2013;121(10): 3981–3987.

Bashford-Rogers RJM, Palser AL, Huntly BJ, et al. Network properties derived from deep sequencing of human B-cell receptor repertoires delineate B-cell populations. Genome Res. 2013;23(11):1874–1884.

Bruggemann M, Schrauder A, Raff T, et al.

Standardized MRD quantification in European all trials: Proceedings of the Second International Symposium on MRD assessment in Kiel, Germany, 18-20 September 2008. Leukemia. 2010;24(3):521– 535.

Driessen EMC, van Roon EHJ, SpijkersHagelstein JAP, et al. Frequencies and prognostic impact of RAS mutations in MLLrearranged acute lymphoblastic leukemia in infants. Haematologica. 2013;98(6):937– 944.

Tamai H, Miyake K, Takatori M, et al.

Activated K-Ras protein accelerates human MLL/AF4-induced leukemo-lymphomogenicity in a transgenic mouse model.

Leukemia. 2011;25(5):888–891.

Elder A, Bomken S, Wilson I, et al. Abundant

47.

48.

49.

50.

51.

52.

53.

and equipotent founder cells establish and maintain acute lymphoblastic leukaemia.

Leukemia. 2017;31(12):2577–2586.

Prieto C, López-Millán B, Roca-Ho H, et al.

NG2 antigen is involved in leukemia invasiveness and central nervous system infiltration in MLL-rearranged infant B-ALL. Leukemia. 2018;32(3):633–644.

Jansen MWJC, Corral L, van der Velden VHJ, et al. Immunobiological diversity in infant acute lymphoblastic leukemia is related to the occurrence and type of MLL gene rearrangement. Leukemia. 2007;21(4):633– 641.

Hotfilder M, Röttgers S, Rosemann A, et al.

Leukemic stem cells in childhood high-risk ALL/t(9;22) and t(4;11) are present in primitive lymphoid-restricted CD34+CD19-cells.

Cancer Res. 2005;65(4):1442–1449.

Bergmann AK, Castellano G, Alten J, et al.

DNA methylation profiling of pediatric Bcell lymphoblastic leukemia with KMT2A rearrangement identifies hypomethylation at enhancer sites. Pediatr Blood Cancer.

2017;64(3):e26251.

Malouf C, Ottersbach K. The fetal liver lymphoid-primed multipotent progenitor provides the prerequisites for the initiation of t(4;11) MLL-AF4 infant leukemia.

Haematologica. 2018 Jun 14. [Epub ahead of print] Kowarz E, Burmeister T, Lo Nigro L, et al.

Complex MLL rearrangements in t(4;11) leukemia patients with absent AF4MLL fusion allele. Leukemia. 2007;21(6):1232– 1238.

Trentin L, Giordan M, Dingermann T, Basso G, Te Kronnie G, Marschalek R. Two inde-

54.

55.

56.

57.

58.

59.

60.

pendent gene signatures in pediatric t(4;11) acute lymphoblastic leukemia patients. Eur J Haematol. 2009;83(5):406–419.

Kühn A, Löscher D, Marschalek R. The IRX1/HOXA connection: insights into a novel t(4;11)- specific cancer mechanism.

Oncotarget. 2016;7(23):35341–35352.

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.

Benedikt A, Baltruschat S, Scholz B, et al.

The leukemogenic AF4-MLL fusion protein causes P-TEFb kinase activation and altered epigenetic signatures. Leukemia. 2011;25(1): 135–144.

Scholz B, Kowarz E, Rössler T, Ahmad K, Steinhilber D, Marschalek R. AF4 and AF4N protein complexes: recruitment of P-TEFb kinase, their interactome and potential functions. Am J Blood Res. 2015;5(1):10–24.

Mück F, Bracharz S, Marschalek R. DDX6 transfers P-TEFb kinase to the AF4/AF4N (AFF1) super elongation complex. Am J Blood Res. 2016;6(3):28–45. van Zelm MC, van der Burg M, de Ridder D, et al. Ig gene rearrangement steps are initiated in early human precursor B cell subsets and correlate with specific transcription factor expression. J Immunol. 2005;175(9): 5912–5922.

Böiers C, Richardson SE, Laycock E, et al. A human IPS model implicates embryonic Bmyeloid fate restriction as developmental susceptibility to B acute lymphoblastic leukemia-associated ETV6-RUNX1. Dev Cell. 2018;44(3):362–377.e7.

haematologica | 2019; 104(6)

ARTICLE

Acute Lymphoblastic Leukemia

Enhanced hemato-endothelial specification during human embryonic differentiation through developmental cooperation between AF4-MLL and MLL-AF4 fusions

Clara Bueno,1,2 Fernando J Calero-Nieto,3 Xiaonan Wang,3 Rafael Valdés-Mas,4 Francisco Gutiérrez-Agüera,1 Heleia Roca-Ho,1 Veronica Ayllon,5 Pedro J Real,5 David Arambilet,6 Lluis Espinosa,6,2 Raul Torres-Ruiz,1 Antonio Agraz-Doblas,1,7 Ignacio Varela,7 Jasper de Boer,8 Anna Bigas,6,2 Bertie Gottgens,3 Rolf Marschalek9 and Pablo Menendez1,2,10

Josep Carreras Leukemia Research Institute and Department of Biomedicine, School of Medicine, University of Barcelona, Spain; 2Centro de Investigación Biomédica en Red de Cáncer (CIBER-ONC), ISCIII, Barcelona, Spain; 3Department of Hematology, Cambridge Institute for Medical Research and Wellcome Trust-Medical Research Council Cambridge Stem Cell Institute, University of Cambridge, UK; 4Dreamgenics S.L. Oviedo. Spain; 5GENyO, Centre for Genomics and Oncological Research, Pfizer/University of Granada/Andalusian Regional Government and University of Granada, Department of Biochemistry and Molecular Biology, Granada, Spain; 6 Programa de Cáncer, Instituto Hospital del Mar de Investigaciones Médicas. Barcelona.

Spain; 7Instituto de Biomedicina y Biotecnología de Cantabria (CSIC-UC-Sodercan), Departamento de Biología Molecular, Universidad de Cantabria, Santander, Spain; 8 Cancer Section, UCL Great Ormond Street Institute of Child Health, London, UK; 9 Institute of Pharmaceutical Biology, Goethe-University, Frankfurt, Germany and 10 Instituciò Catalana de Recerca i Estudis Avançats (ICREA), Barcelona, Spain 1

Ferrata Storti Foundation

Haematologica 2019 Volume 104(6):1189-1201

ABSTRACT

Correspondence:

T

he t(4;11)(q21;q23) translocation is associated with high-risk infant pro-B-cell acute lymphoblastic leukemia and arises prenatally during embryonic/fetal hematopoiesis. The developmental/pathogenic contribution of the t(4;11)-resulting MLL-AF4 (MA4) and AF4-MLL (A4M) fusions remains unclear; MA4 is always expressed in patients with t(4;11)+ B-cell acute lymphoblastic leukemia, but the reciprocal fusion A4M is expressed in only half of the patients. Because prenatal leukemogenesis manifests as impaired early hematopoietic differentiation, we took advantage of well-established human embryonic stem cell-based hematopoietic differentiation models to study whether the A4M fusion cooperates with MA4 during early human hematopoietic development. Co-expression of A4M and MA4 strongly promoted the emergence of hemato-endothelial precursors, both endothelial- and hemogenic-primed. Double fusionexpressing hemato-endothelial precursors specified into significantly higher numbers of both hematopoietic and endothelial-committed cells, irrespective of the differentiation protocol used and without hijacking survival/proliferation. Functional analysis of differentially expressed genes and differentially enriched H3K79me3 genomic regions by RNA-sequencing and H3K79me3 chromatin immunoprecipitation-sequencing, respectively, confirmed a hematopoietic/endothelial cell differentiation signature in double fusion-expressing hemato-endothelial precursors. Importantly, chromatin immunoprecipitation-sequencing analysis revealed a significant enrichment of H3K79 methylated regions specifically associated with HOX-A cluster genes in double fusion-expressing differentiating hematopoietic cells.

Overall, these results establish a functional and molecular cooperation between MA4 and A4M fusions during human hematopoietic development. haematologica | 2019; 104(6)

CLARA BUENO cbueno@carrerasresearch.org PABLO MENENDEZ pmenendez@carrerasresearch.org Received: July 18, 2018.

Accepted: January 21, 2019.

Pre-published: January 24, 2019. doi:10.3324/haematol.2018.202044 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/104/6/1189 ©2019 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

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

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

1189

C. Bueno et al.

Introduction The mixed-lineage leukemia (MLL) gene encodes for an H3K4 histone methyltransferase important in hematopoietic development.1 The human MLL gene is frequently rearranged in acute leukemia and typically confers a dismal outcome.2,3 Of particular interest is the t(4;11)(q21;q23) translocation, which encodes the fusion proteins MLL-AF4 (MA4) and AF4-MLL (A4M), and is associated with infant B-cell acute lymphoblastic leukemia (B-ALL). This t(4;11)+ infant leukemia is characterized by a very brief latency, raising the question of how it evolves so quickly.4 Moreover, the exceptionally high concordance rate of t(4;11)+ B-ALL in monozygotic twin infants5,6 suggests that all the necessary (epi)genetic events required for leukemogenesis are accomplished prenatally, during embryonic/fetal hematopoietic development.7 However, our understanding of t(4;11)-mediated developmental effects is limited due, at least in part, to the variety of phenotypes and long latency observed in currently available t(4;11) mouse models.2,8-17 These different phenotypes likely result from targeting a cell in the wrong developmental stage, or not addressing the impact of secondary hits, leaving open questions about the developmental impact of the t(4;11) translocation during early human development.

The functional and molecular contribution of the reciprocal fusion genes resulting from the derivative translocated chromosomes remains obscure in cancer. The MA4 fusion is always expressed in t(4;11)+ B-ALL patients, whereas the reciprocal fusion A4M is expressed in only half of the patients.18-20 Importantly, t(4;11)+ cell lines display addiction to MA4 but not to A4M,21,22 and although A4M was not sufficient to initiate leukemia in cord bloodderived CD34+ cells,23 it was nevertheless capable of initiating B-ALL in mice without the requirement of MA4, indicating that it contributes to t(4;11)-driven leukemogenesis.11,24,25 Strikingly, a very recent clinical study has unraveled an independent prognostic value for MA4 expression in t(4;11)+ infant B-ALL, thus adding a new piece to the puzzle.19 Thus, the developmental/pathogenic contribution of the t(4;11)– resulting reciprocal fusion A4M remains enigmatic.

Human embryonic stem cells (hESC) represent a powerful tool for modeling different developmental aspects of human disease that cannot otherwise be addressed by analyses of patients’ samples or mouse models.7,26,27 Given that prenatal leukemogenesis manifests as impaired early hematopoietic differentiation, modeling hematopoietic differentiation in hESC may represent a promising in vitro approach to study the onset of hematopoiesis and the mechanisms underlying early human hematopoietic development.7 During hESC differentiation, a primitive population of CD45– hemato-endothelial precursors (HEP) arises and further differentiates into CD45+ hematopoietic and mature endothelial cells.28-30 Beyond its pathogenic role in acute leukemias, the MLL gene has also been implicated in endothelial cell maturation,31 and endothelial dysfunction was recently linked to disease outcomes in childhood leukemias.32 We previously reported that MA4 favors the emergence of endothelial-primed HEP but not hemogenic HEP from hESC.10 Here, we took advantage of well-established hESC-based differentiation systems to study whether the A4M fusion cooperates with MA4 during early human hematopoietic and endothelial develop1190

ment. We report a functional and molecular cooperation between MA4 and A4M fusions, which results in enhanced hemato-endothelial output during human embryonic development.

Methods Vector construction and lentiviral transduction The cDNA for MA4 and A4M were subcloned into the pRRLEF1α-PGK-NEO vector.11,16 Both fusions have been described previously (Online Supplementary Figure S1A).11,23 We used the following lentivectors containing either neomycin or dTo for cell selection: pRRL-EF1α-PGK-NEO (empty vector; EV), pRRL-EF1αMA4-PGK-NEO (MA4) and pRRL-EF1α-A4M-PGK-dTo (A4M).

VSV-G-pseudotyped lentiviral particles were generated in 293T cells using standard transfection protocols and concentrated by ultracentrifugation.33 hESC were infected overnight with concentrated EV or MA4 lentivirus plus 8 μg/mL polybrene. Viral supernatants were washed away the next day, and EV- and MA4-transduced hESC were then selected with G418 (50-100 μg/mL) for 3 weeks. For dual transduction of MA4 and A4M fusions, G418resistant MA4-expressing hESC were infected with A4M-expressing viruses. EV/G418-selected hESC were also transduced with A4M alone. Transgene expression was confirmed for all the genotypes (Figure 1).

Human embryonic stem cell culture and characterization of transgenic human embryonic stem cell lines hESC (AND1 cell line) were maintained undifferentiated on a layer of irradiated human mesenchymal stem cells in complete knockout Dulbecco modified Eagle medium containing 20% knockout serum replacement and 8 ng/mL basic fibroblast growth factor.34,35 The medium was changed daily, and cells were passaged weekly by dissociation with 1:1 collagenase IV:dispase. Cultures were visualized daily by phase contrast microscopy. Approval for hESC work was obtained from the Spanish National Embryo Ethical Committee. The pluripotency of transgenic hESC was characterized by flow cytometry using antibodies against SSEA-3, SSEA-4 TRA-1-60 and TRA-1-81 (BD Biosciences).36 Expression of the pluripotency-associated transcription factors OCT4, NANOG, SOX2, CRIPTO, and DNMT3B as well as transgene expression (MA4 and A4M) were analyzed by quantitative real-time polymerase chain reaction (PCR) (Online Supplementary Table S1 shows the primers and PCR conditions used).23,37,38

Hematopoietic differentiation from human embryonic stem cells by embryoid body formation Undifferentiated hESC were treated with collagenase IV:dispase for 1 h at 37ºC. To examine embryoid body (EB) formation, cells were transferred to low-attachment plates and incubated overnight in differentiation medium (knockout Dulbecco modified Eagle medium supplemented with 20% fetal bovine serum, 1% non-essential amino acids, 1 mmol/L L-glutamine, and 0.1 mmol/L b-mercaptoethanol). The medium was changed the next day to the same differentiation medium supplemented with the following hematopoietic cytokines: 300 ng/mL stem cell factor, 300 ng/mL Flt3L, 10 ng/mL interleukin-3, 10 ng/mL interleukin-6, 50 ng/mL granulocyte - colony-stimulating factor and 25 ng/mL bone morphogenetic protein-4 (all from R&D).9,29,39-41 EB were dissociated at different time points during development using collagenase B and enzyme-free Cell Dissociation Buffer (Invitrogen).

Dissociated cells were stained with anti-CD34-PE, anti-CD31FITC, anti-CD45-APC or anti-CD34-PE-Cy7, CD31-BV510, antihaematologica | 2019; 104(6)

AF4-MLL in human embryonic hematopoietic development

glycophorin A, anti-CD43-FITC, anti-CD45-APC antibodies (all from BD Biosciences) and 7-actinomycin D, and analyzed using a FACS Canto flow cytometer.9,29,39-41 Colony-forming unit assays were performed on days 10 and 15 of EB differentiation by plating 60x104 EB cells onto serum-free methylcellulose H4435 (Stem Cell Technologies). Colonies were scored after 12 days.9,29,42-44

Cells were then suspended in propidium iodide-containing buffer and acquired and analyzed on a FACS Canto-II using Modfit LT4.0 software, discriminating between quiescent cells (G0/G1), cycling cells (S-phase) and G2/M cells.45,46 Apoptosis was assessed with an Annexin-V Apoptosis Detection kit (BD Biosciences).16

Human embryonic stem cell-OP9 co-cultures Cell cycle and apoptosis analysis For cell cycle analysis of hESC-derived HEP and CD45+ cells, day 15 EB were dissociated and harvested cells were fixed overnight in 70% ice-cold ethanol. Cells were then washed in phosphate-buffered saline and incubated with anti-CD31-FITC, anti-CD34-PE-Cy7 and anti-CD45-APC antibodies for 15 min.

A

hESC-OP9 co-cultures were performed as described elsewhere.47,48 OP9 stroma was prepared by plating OP9 cells in gelatin-coated dishes, and allowing them to overgrow as a monolayer. hESC were prepared as a suspension of small aggregates using collagenase IV:dispase. One-tenth of this suspension was plated on top of the 8-day overgrown OP9 stroma. Media were replaced on

B

C

D

Figure 1. Characterization of transgenic human embryonic stem cells expressing the reciprocal fusion A4M together with MA4. (A) RNA-sequencing and quantitative real-time polymerase chain reaction (qRT-PCR) validation revealed that ~45% (11/25) of the patients with t(4;11)+ B-cell precursor acute lymphoblastic leukemia do not express the reciprocal fusion A4M.18 (B) Left, Phase-contrast morphology of representative colonies from each transgenic human embryonic stem cell (hESC) line. Right, Reverse transcriptase polymerase chain reaction analysis (RT-PCR) confirming expression of both fusions in undifferentiated hESC. (C) qRT-PCR expression of the pluripotency genes OCT4, SOX2, NANOG, CRIPTO, and DNMT3B. (D) Representative FACS data confirming expression of the pluripotency surface markers SSEA-3, SSEA-4, TRA-160, and TRA-1-81. BCP-ALL: B-cell precursor acute lymphoblastic leukemia; RNA-seq: RNAsequencing; pos: positive; neg: negative; EV: empty vector; C+: positive control.

haematologica | 2019; 104(6)

1191

C. Bueno et al.

the next day and one-half volume media changes were performed every other day thereafter. Hematopoietic differentiation was assessed by flow cytometry at day 9 of co-culture. Accordingly, hESC-OP9 co-cultures were treated with collagenase IV/TrypLE and cells were dissociated and filtered through a 70-μm strainer.

Cell suspensions were stained with anti-mouse CD29-FITC and anti-human CD34-PE and CD45-APC antibodies. The proportion of HEP (CD34+CD31+CD45-) and total blood cells (CD45+) were analyzed within the CD29- hESC-derived cell population.

Hemogenic and endothelial HEP were distinguished based on CD34 and CD43 expression.40

Culture of FACS-isolated hemato-endothelial precursors in MS5 stroma or liquid culture Day 9 human hESC-OP9 co-cultures were dissociated as above and both CD45+ cells and HEP were analyzed. FACS-purified HEP (CD29-CD34+CD31+CD45-) were plated onto MS5 stroma or in liquid culture for 30 or 16 days, respectively, in differentiation medium with hematopoietic cytokines (50 ng/mL stem cell factor, 50 ng/mL Flt3L, 10 ng/mL interleukin-3, 20 ng/mL interleukin-7).

The medium was changed every 7 days, and the emergence of CD45+ hematopoietic cells was analyzed by FACS.

Endothelial differentiation of hemato-endothelial precursors HEP (2×104) from day 9 human hESC-OP9 co-cultures were seeded onto 0.1% gelatin-coated plates in complete EGM-2 medium with microvasculature supplements (Lonza) for 7 days. Cells were then fixed, permeabilized and stained with rabbit antihuman VE-cadherin (Cayman), mouse anti-human endothelial nitric oxide synthase (BD Biosciences), and mouse anti-human von Willebrand factor (DAKO) followed by Alexa 488-conjugated anti-rabbit or Cy3-conjugated anti-mouse (Jackson Immunoresearch) antibodies. Nuclei were counterstained with 4′,6-diamidino-2-phenylindole. Images were obtained using an inverted fluorescence microscope. Day 7 differentiating cells were trypsinized and cell suspensions were stained with anti-human CD31-FITC and CD144-PerCP-Cy5.5 antibodies.

Mouse transplantation and analysis of hematopoietic-endothelial engraftment NOD/LtSz-scid IL-2Rγ−/− (NSG) mice were housed under sterile conditions. The Animal Care Committee approved all mouse protocols. Briefly, cord blood-derived CD34+ hematopoietic stem and progenitor cells (3×104 cells) or cells from day 15 EB (5×105 cells) were transplanted into the bone marrow as described previously.49 Animal health was monitored throughout the entire experiment.

Mice were killed 10 weeks after transplantation and cell suspensions were analyzed by FACS for human chimerism using antiHLA-ABC-FITC, anti-CD31-PE, CD144-PerCP-Cy5.5, and antiCD45-APC antibodies.

RNA- and chromatin immunoprecipitation-sequencing Details of the RNA- and chromatin immunoprecipitationsequencing and analysis are provided in the Online Supplementary Methods.

Statistical analysis All data are expressed as mean ± standard error of mean.

Statistical comparisons were performed using GraphPad Prism software with the nonparametric Mann-Whitney test, two-tailed P-value (with 95% confidence interval). Statistical significance was defined as a P-value <0.05.

1192

Results Co-expression of A4M and MA4 does not hijack pluripotency We showed very recently that only 45% of t(4;11)+ BALL patients express the reciprocal fusion A4M, whereas MA4 is consistently expressed in all t(4;11)+ B-ALL patients (Figure 1A).18-20 Here, we generated transgenic hESC lines expressing “MA4 alone”, “A4M alone” or MA4+A4M (double fusion), (Figure 1B and Online Supplementary Figure S1B). EV (control)- and MA4-hESC were established by G418 selection.9 G418-resistant EV- or MA4-expressing hESC were then transduced with A4M/dTo-expressing lentiviruses and greater than 90% transduction efficiency was achieved. Transgenic hESC lines were maintained for more than 50 passages and retained hESC-like morphology (Figure 1B, left), transgene expression (Figure 1B, right), and expression of pluripotency-associated transcription factors (Figure 1C) and surface markers (Figure 1D). All hESC genotypes formed teratomas in NSG mice (data not shown).9,50 Thus, (co-)expression of A4M and/or MA4 is compatible with hESC pluripotency.

A4M and MA4 co-operate to promote emergence of hemato-endothelial precursors and enhance blood production Hematopoietic differentiation was assessed using two distinct and well-established differentiation systems: EB formation43,47 (Figure 2) and OP9 co-culture47,48 (Figure 3).

During differentiation, a population of primitive HEP arises, which is responsible for further hematopoietic and endothelial commitment10,30 (Figures 2A and 3A). We investigated whether co-expression of A4M and MA4 affects hESC-derived hematopoiesis by analyzing the emergence of HEP during EB development in hESC individually expressing the single fusions or the double fusion.

We observed a pronounced (~5- to 10-fold; P<0.05) increase in HEP at days 7 and 10 of development in EB expressing the double A4M and MA4 fusion over those expressing single fusions (Figure 2B, upper-left panel). We next assessed whether co-expression of A4M and MA4 influences subsequent hematopoietic commitment of HEP. The kinetics of emergence and output of both total CD45+ hematopoietic cells and CD45+CD34+ hematopoietic progenitors was faster (EB day 10) from double fusion-expressing hESC than from equivalent single fusion-expressing cells, achieving a 2- to 3-fold higher hematopoietic output by day 15 of EB development (Figure 2B). Furthermore, double fusion-expressing HEP massively accelerated (EB day 10) the emergence of clonogenic hematopoietic progenitors as compared to single fusion-expressing HEP (Figure 2B, bottom-right panel).

According to our previous work, if the kinetics of human EB differentiation is extended, allowing for a continuum of HEP-to-blood transition, MA4-expressing human EB display enhanced HEP production coupled to impaired blood output (EB day 20) (Online Supplementary Figure S2A) and clonogenic potential (EB day 15) (Figure 2B). We confirmed stable expression of ectopic MA4 and A4M upon EB differentiation, supporting the link between genotype and phenotype (Figure 2C).

We also investigated hematopoietic differentiation using the OP9 differentiation system (Figure 3A,B), and by plating FACS-sorted HEP in either hematopoietic liquid haematologica | 2019; 104(6)

AF4-MLL in human embryonic hematopoietic development

culture (Figure 3C) or onto MS5 feeders (Figure 3D). After 10 days on OP9 stroma, double fusion-expressing hESC yielded a 10-fold higher number of CD45+ hematopoietic cells than did single fusion-expressing hESC (Figure 3B).

Moreover, when HEP were FACS-sorted from day 9 OP9

co-cultures and allowed to differentiate into CD45+ blood cells, the yield of CD45+ cells was up to 60-fold higher in double fusion-expressing HEP than in single fusionexpressing HEP (Figure 3C,D). Encouraged by these results, we next investigated whether ectopic expression

A

B

C

D

Figure 2. A4M cooperates with MA4 to accelerate human embryonic stem cell/erythroid body specification towards hemato-endothelial precursors and subsequent hematopoietic differentiation. (A) Schematic of erythroid body hematopoietic differentiation of human embryonic stem cells (hESC) and end-point analyses.

(B) Upper left, specification into hemato-endothelial precursors (HEP; CD31+CD34+CD45-) is accelerated in double fusion-expressing hESC. Subsequent differentiation of HEP into hematopoietic progenitors (upper right) and mature CD45+ blood cells (bottom left) is enhanced in double fusion-expressing HEP. Bottom right, Colony-forming unit read-out and scoring (pie charts) confirming accelerated and enhanced hematopoietic progenitor potential from double fusion-expressing blood derivatives. (C) Reverse transcriptase polymerase chain reaction analysis confirming stable expression of MA4 and A4M upon erythroid body differentiation. (D) Neither MA4- nor double fusion-expressing blood derivatives display in vivo hematopoietic engraftment potential in irradiated NSG mice. Data are presented as mean Âą standard error of mean from at least three independent experiments. *P<0.05. EB: erythroid body; bFGF: basic fibroblast growth factor; BMP4: bone morphogenetic protein-4; SCF: stem cell factor; IL: interleukin; G-CSF: granulocyte colony-stimulating factor; CFU: colony-forming unit; EV: empty vector; IBMT; intra-bone marrow transplantation; C+: positive control; CB: cord blood.

haematologica | 2019; 104(6)

1193

C. Bueno et al.

of both A4M and MA4 confers in vivo engraftment capacity to hESC-derived hematopoietic derivatives. To do this, we transplanted 5×105 hESC hematopoietic derivatives from each genotype into myeloablated NSG mice,4,43,47 finding that, despite regulating hematopoietic development in vitro, double fusion-expression did not confer in vivo engraftment to hESC hematopoietic derivatives (Figure 2D).

The increased hematopoietic output of double fusionexpressing hESC might be the consequence of transgenemediated proliferation/survival of the emerging HEP or CD45+ cells. To address this, we analyzed cell cycle distribution (Online Supplementary Figure S2B) and apoptosis (Online Supplementary Figure S2C) within both HEP and the CD45+ cell population. No differences in the proportions of either cycling HEP or CD45+ cells were detected between genotypes (25–36% for HEP and 35–41% for CD45+ cells) (Online Supplementary Figure S2B). Apoptotic levels were similarly low in the different genotypes of HEP (6–8%) and CD45+ cells (5–7%) (Online Supplementary Figure S2C). Collectively, these results show that A4M cooperates with MA4 to induce HEP specification and blood commitment, without hijacking proliferation or survival of HEP.

A4M and MA4 cooperate to enhance endothelial cell fate from hemato-endothelial precursors We next addressed the developmental impact of A4M in endothelial maturation from HEP.10,47 We hypothesized that co-expression of A4M and MA4 in HEP may concomitantly promote subsequent endothelial and hematopoietic commitment or skew the hematoendothelial commitment in favor of hematopoiesis. To test this, we analyzed the ability of HEP to differentiate

A

into mature endothelial cells. OP9-hESC co-cultures were dissociated on day 9 of development and HEP were FACSsorted and cultured for 1 week in endothelial-promoting conditions (Figure 4A). The expression of the mature endothelial markers VE-cadherin (CD144), von Willebrand factor, endothelial nitric oxide synthase and CD31 was then analyzed. Irrespective of the genotype, HEP cultured in endothelial conditions attached, became spindle-shaped, and formed VE-cadherin+ endothelial-like structures co-expressing endothelial nitric oxide synthase, von Willebrand factor and CD31 (Figure 4B,C, top panel).

However, double fusion-expressing HEP were more prone to differentiate into mature endothelial cells than were single fusion-expressing HEP. Accordingly, they yielded a 20-fold higher number of VE-cadherin+ endothelial-like structures (Figure 4C, top panel) and CD144+CD31+ endothelial cells (Figure 4C, bottom panel). Interestingly, endothelial cells (HLA.ABC+CD31+CD34+CD144+CD45CD43-) were found in the bone marrow of mice transplanted with double fusion-expressing hESC blood derivatives at levels ~4-fold higher than those in mice transplanted with single fusion-expressing cells (Figure 4D).

Within CD34+CD31+CD45- HEP, two subpopulations of phenotypically and functionally distinct HEP can be distinguished based on the expression of CD34 and CD43: hemogenic HEP (CD34low/+CD43+CD45-) and endothelial HEP (CD34++CD43-CD45-) (Figure 5A).40,48,51 We thus analyzed the contribution of both HEP populations to the superior hematopoietic and endothelial differentiation observed in double fusion-expressing HEP. Co-expression of A4M and MA4, but not single fusions, robustly enhanced the emergence of both endothelial and hemogenic HEP (Figure 5B,C). The identity of hemogenic and endothelial HEP was confirmed by the specific

C

D

B

Figure 3. Co-expression of MA4 and A4M enhances hematopoietic differentiation of human embryonic stem cells in OP9 co-culture. (A) Experimental design of OP9-based human embryonic stem cell (hESC) differentiation towards hemato-endothelial precursors (HEP) and further hematopoietic commitment of HEP maintained in either liquid culture for 16 days or in MS5 co-culture for 30 days. (B) Frequency of total CD45+ blood cells after 9 days in OP9 co-culture. (C,D) CD45CD31+CD34+ HEP were FACS-purified at day 9 of OP9 co-culture and allowed to differentiate into CD45+ cells in liquid culture (C) or in MS5 co-culture (D). Data are represented as the mean ± standard error of mean from independent experiments. bFGF: basic fibroblast growth factor; MTG: monothioglycerol; SCF: stem cell factor; IL: interleukin; G-CSF: granulocyte colony-stimulating factor; EV: empty vector.

1194

haematologica | 2019; 104(6)

AF4-MLL in human embryonic hematopoietic development

expression of early hematopoietic and endothelial master genes (Figure 5D). Thus, A4M cooperates with MA4 to promote hematopoietic and endothelial cell fate.

Genome-wide transcriptomic and H3K79 methylation profiles support the developmental cooperation between A4M and MA4 To identify patterns of gene expression that might provide a molecular explanation of the functional cooperation between A4M and MA4 in hematopoietic specification,

we performed RNA-sequencing analysis on FACS-purified EV-, MA4-, A4M- and double fusion-expressing HEP from day 15 EB. Figure 6A shows a heatmap representation of the hierarchical clustering of the 335 genes differentially expressed between the four genotypes (Online Supplementary Table S2). There is a clear transcriptomic transition towards a hematopoietic/endothelial gene signature from EV-HEP to double fusion-expressing HEP. Single fusion-expressing HEP clustered interspersed between EV and double-fusion HEP. The biological func-

A

B

C

D

Figure 4. Enhanced endothelial cell fate from hemato-endothelial precursors co-expressing MA4 and A4M. (A) Scheme of hemato-endothelial precursor (HEP) endothelial differentiation and phenotypic characterization. (B) FACS-sorted HEP from day 9 human ESC-OP9 co-cultures were cultured in EGM2 medium for 5 days and analyzed by immunofluorescence for VE-cadherin, endothelial nitric oxide synthase and von Willebrand factor. (C) Top, Endothelial-like structures were identified and quantified based on VE-cadherin staining (white dotted-lined areas in B, top panel). Bottom, Frequency of CD45-CD31+CD144+ endothelial cells quantified by flow cytometry. (D) In vivo endothelial engraftment potential (HLA.ABC+CD31+CD144+CD45-) analyzed in bone marrow of NSG mice 8 weeks after transplantation of HEP. Data are presented as mean Âą standard error of mean from five independent experiments. *P<0.05. bFGF: basic fibroblast growth factor; EV: empty vector; CB: cord blood.

haematologica | 2019; 104(6)

1195

C. Bueno et al.

tions affected by genes differentially expressed in MA4-, A4M- and double fusion-expressing HEP relative to EV were classified by Ingenuity Pathway Analysis software;47,52 the top significantly enriched functional categories included “hematological system development and function”, “cancer” and “hematological disease” (Figure 6B). Statistical (-logp-value) power shows distant effects of MA4 and A4M; however, co-expression of both fusions seems to establish a molecular balance/developmental cooperation in promoting blood-endothelial specification from hPSC. Strikingly, biofunctions specifically associated with hematologic lymphoid malignancies (not with non-hematologic cancer) were predicted to be activated (positive z-score) exclusively in double fusion-expressing HEP, further suggesting a molecular cooperation between MA4 and A4M in development and infant leukemia.

(Figure 6C).

The C-terminal partners of MLL fusions normally interact with the histone methyltransferase DOT1L, which is the sole histone methyltransferase catalyzing histone 3 lysine 79 (H3K79) methylation, a chromatin modification widely associated with the dysregulated expression of HOX-A cluster genes in MLL leukemias.13,53 We thus performed genome-wide chromatin immunoprecipitation-

sequencing analysis of the H3K79 trimethylation (H3K79me3) profiles in control, MA4-, A4M- and double fusion-expressing hESC-derived blood derivatives (Figure 7, Online Supplementary Figure S3A, Online Supplementary Table S3). In line with the RNA-sequencing data, functional analysis of the differential H3K79me3 peaks specific for double fusion-expressing cells revealed significant gene ontology functional categories associated with ”definitive hematopoiesis”, “myeloid and erythroid differentiation/homeostasis” and “endothelial cell development” (Figure 7A, Online Supplementary Figure S3B).

This further supports the developmental co-operation between A4M and MA4 in promoting hemato-endothelial specification.

Finally, we analyzed the H3K79me3 profiles at genomic loci of the classical MLL target genes reported by Guenther et al.54 Non-HOX-A classical MLL targets such as RUNX1, LMO2, ADMA10, and KDM6A showed a slight although non-significant, enrichment of H3K79me3 in MA4-expressing hESC, validating our chromatin immunoprecipitation-sequencing approach (Figure 7B).

However, enrichment of H3K79me3 in HOX-A cluster genes was observed exclusively in A4M-expressing cells although it was statistically significant only in double

A

B

D

Figure 5. Co-expression of MA4 and A4M significantly enhances the emergence of both endothelial and hemogenic hemato-endothelial precursors. (A) Representative flow cytometry analysis of hemato-endothelial precursors (HEP) with hemogenic (CD45CD31+CD43+CD34dim/+) and endothelial (CD45-CD31+CD43-CD34++) potential.

(B,C) A4M co-operates with MA4 to boost the emergence of both endothelial (B) and hemogenic (C) HEP. Data are presented as the mean ± standard error of mean from three independent experiments. (D) Expression of RUNX1c and Ve-Cad in hemogenic and endothelial HEP. *P<0.05. EB: erythroid body; EV: empty vector.

C

1196

haematologica | 2019; 104(6)

AF4-MLL in human embryonic hematopoietic development

fusion-expressing differentiating hESC (false discovery rate <0.1) (Figure 7C). As such, HOX-A genes were upregulated in double fusion-expressing hematopoietic clonogenic progenitors (Online Supplementary Figure S3C). No differential enrichment of the repressive H3K4me2/3 mark was observed in either HOX-A or non-HOX genes in double fusion-expressing cells (Online Supplementary Figure

S4). Collectively, these data suggest that the deregulated expression of HOX-A genes in MLL leukemias may be imposed by the reciprocal A4M fusion through H3K79 methyltransferase activity. In support of this, a recent RNA-sequencing study performed in 42 infants with t(4;11)+ B-ALL enrolled in the Interfant treatment protocol, revealed that 45% of t(4;11)+ patients express the A4M

A

B

C

Figure 6. Transcriptional transition towards a hematopoietic/endothelial gene signature in double fusion-expressing hemato-endothelial precursors. (A) Heatmap representation of hierarchical clustering of genes differentially expressed between empty vector (EV)-, single fusions- and double fusion-expressing hemato-endothelial precursors (HEP). Each column represents a technical replicate from three independent experiments. (B,C) Statistically significant functional categories (B) and cancer/leukemia-associated biofunctions (C) identified using Ingenuity Pathway Analysis on genes differentially expressed in single fusions-, and double fusionexpressing HEP relative to EV. They are ranked by z-score. Functional categories associated with “hematological system development and function” and “cardiovascular system development” are shown in bold. All significant biofunctions are associated with blood cell differentiation, homeostasis and migration/movement.

haematologica | 2019; 104(6)

1197

C. Bueno et al.

fusion, and that HOX-A cluster genes are exclusively expressed in this AF4-MLL-expressing subgroup of t(4;11)+ patients, who do in fact have a significantly more favorable clinical outcome.19

From etiological and pathogenic standpoints, infant cancer is distinct from adult cancer and should be studied as a developmental disease.4,7,16 A biologically and clinically intriguing infant cancer is the t(4;11)+ B-ALL, which is associated with a dismal outcome.4,21,23 Evidence in support of its prenatal origin comes from studies in monozygotic twins and archived blood spots, providing compelling evidence of a single prenatal cell as the origin for t(4;11),5 and also from recent genome-wide studies demonstrating that this infant leukemia has one of the lowest frequencies of somatic mutations of any sequenced cancer.55 The stable genome of these patients suggests that in infant developmental cancer, one â&#x20AC;&#x153;bighitâ&#x20AC;? might be sufficient to cause overt disease, supporting a key contribution of the prenatal cell-of-origin during a critical developmental window of stem cell vulnerability in leukemogenesis. However, despite its aggres-

siveness and short latency, our current understanding about its etiology, pathogenesis and cellular origin is still limited.2,4,14,16,52 Importantly, a recently developed xenograft model which represents the most bona fide model for t(4;11)+ B-ALL so far, has revealed the instructive role of MLL-Af4 in cord blood-derived CD34+ cells.14 Studies using primary cells from t(4;11)+ B-ALL patients are incapable of addressing the developmental genesis of the hematopoietic system. Recent data suggest that fetal liver lymphoid-primed multipotent progenitors may provide the developmental prerequisites for the initiation of t(4;11)+/MLL-AF4 infant leukemia.56 Because leukemogenesis manifests as a blockage or altered cell differentiation, the hematopoietic differentiation of hESC may represent a promising in vitro model for studying the onset of hematopoiesis and the earliest events leading to the specification of the hematopoietic cells.36 Previous studies have addressed the oncogenic role of leukemic fusion genes in hESC-derived hematopoiesis.57-59 We previously explored the developmental impact of the prenatal fusion MA4 in hESC hemato-endothelial development,10 and found that MA4 expression promotes the emergence of endothelialprimed HEP and further endothelial commitment, but hijacks the specification of hemogenic-primed HEP, impairing hematopoietic output.10

A

B

Discussion

C

Figure 7. H3K79 methylation profiles at genomic loci of MLL targets in MA4-, A4M- and double fusion-expressing human embryonic stem cell-derived blood derivatives. (A) Gene ontology enrichment of differential H3K79me3 peaks specific for double fusion-expressing cells. (B,C) Representative profiles for chromatin immunoprecipitation-sequencing using anti-H3K79me3 antibody at genomic regions of typical non-HOXA (B) and HOXA MLL targets (C).

1198

haematologica | 2019; 104(6)

AF4-MLL in human embryonic hematopoietic development

The functional and molecular contribution of the reciprocal fusion genes resulting from the derivative translocated chromosomes remains obscure in cancer.18,60 The MA4 fusion is always expressed in t(4;11)+ B-ALL patients, whereas the reciprocal fusion A4M is expressed in only 45-50% of the patients.18-20,60 Here, we took advantage of well-established hESC-based differentiation systems to study whether the A4M fusion cooperates with MA4 during early human hematopoietic and endothelial development. Co-expression of A4M and MA4 strongly promoted the emergence of both endothelial-primed and hemogenic-primed HEP. Moreover, the double fusionexpressing HEP specified into significantly higher numbers of both hematopoietic and endothelial cells, irrespective of the in vitro differentiation protocol used and without affecting survival or proliferation, indicating a functional (developmental) cooperation between MA4 and A4M fusions during human hematopoietic development.

This notion was confirmed by genome-wide transcriptomic analysis of differentiating HEP. These developmental biology studies support previous evidence suggesting that A4M contributes to the pathobiology of t(4;11)+ BALL. Accordingly, Bursen et al.11 reported that A4M-transduced murine hematopoietic stem cells developed pro-BALL, whereas co-transduction with MA4 and A4M resulted in mixed lineage leukemia. Moreover, studies from Milneâ&#x20AC;&#x2122;s laboratory demonstrated that RUNX1 is directly activated by MA4 and the RUNX1 protein interacts with the A4M protein, suggesting a mechanism of cooperation between the two fusion genes at the molecular level.25 In the embryo, definitive hematopoiesis cannot occur in the absence of endothelial cell development. Definitive hematopoietic stem cells in both mice and humans emerge from the hemogenic endothelium by a process known as endothelial-to-hematopoietic transition.61 Hematopoietic differentiation of hESC occurs through the generation of HEP, from which both endothelial and hematopoietic cells then originate. Here, co-expression of A4M and MA4 in HEP concomitantly promoted endothelial and hematopoietic commitment rather than skewing the hemato-endothelial commitment in favor of one lineage over the other. This finding is important because beyond its pathogenic role in acute leukemias, the MLL gene is implicated in endothelial cell maturation, and endothelial dysfunction was recently linked to disease outcome in childhood leukemias.31,32 Furthermore, other leukemia fusion oncogenes as well as lymphoma-specific genetic aberrations have been found in endothelial cells from patients with chronic myeloid leukemia and B-cell lymphoma,10,43,44 suggesting that endothelial cells may be part of the neoplastic clone. In addition, bone marrowderived mesenchymal stem cells from infant t(4;11)+ B-cell ALL were recently found to harbor and express the t(4;11) translocation.33 The existence of a common embryonic precursor for mesenchymal stem cells and endothelial cells has been recently demonstrated by hESC-directed differentiation.10,45 The finding of such a common embryonic precursor, and the presence of t(4;11) in both leukemic blasts and bone marrow mesenchymal stem cells of infant patients, suggests that the t(4;11) translocation arises and has a developmental impact on early pre-hematopoietic precursors. As a technical caveat, it is important to emphasize that MA4 and A4M were sequentially transduced to allow for antibiotic selection and homogeneously-transduced hESC cultures. However, in double fusion-expresshaematologica | 2019; 104(6)

ing differentiating blood cells, MA4 was never individually expressed in the absence of A4M. When hematopoietic differentiation of hESC was induced by EB formation both fusions were readily co-expressed, ruling out a biased functional phenotype driven by the sequential expression of each transgene in distinct developmental windows.

Mechanistically, a putative function of A4M is to activate chromatin, rendering a chromatin landscape similar to that present during stem cell development. It is currently unknown how A4M is able to reprogram chromatin, but it does contain the SET domain disrupted from its "specification domain", the N-terminal portion of MLL, which binds to MEN1 and LEDGF, shaping the gene targeting module of the MLL gene. When A4M is expressed, the N-terminal portion is substituted by the AF4 N-terminus, which is the crucial domain (AF4N) that binds to and strongly activates RNA polymerase II (RNAP II) for transcriptional elongation. Overexpression of either AF4, AF4N or the fusion protein A4M induces robust RNAP IIdependent gene transcription by overwriting the elongation control process in a dominant fashion.62-64 Since gene transcription per se and in particular â&#x20AC;&#x153;sterileâ&#x20AC;? transcription is a powerful mechanism for chromatin activation, A4M could potentiate MA4 to skew normal and leukemic hematopoietic cell fate decisions. This also explains why MA4 has a more prominent role in the disease than the reciprocal A4M. If A4M functioned to initiate this process only by itself, then it would become obsolete after fulfilling the "chromatin opening job". However, transcription factors such as MA4, RUNX1 or others then establish the transcriptional program leading to leukemogenesis. This is reflected in the enhanced hematopoietic potential of double fusion-expressing hESC and the enriched H3K79me3 activation mark in HOX-A cluster genes exclusively when MA4 and A4M are co-expressed. Thus, A4M prepares other transcription factors to become oncoproteins.

Molecularly, C-terminal-partners of MLL fusions (AF4, AF9, ENL) interact with DOT1L, which is the sole histone methyltransferase catalyzing H3K79 methylation, a chromatin modification widely associated with the dysregulated expression of the HOX-A gene cluster in MLLrearranged leukemias.13,53 Here, chromatin immunoprecipitation-sequencing analysis of differentially enriched H3K79me3 genomic regions confirmed a hematopoietic/endothelial cell differentiation signature in double fusion-expressing HEP, and revealed a significant enrichment of H3K79 methylated regions specifically associated with HOX-A cluster MLL target genes (but not with non-HOX-A MLL targets) in double fusion-expressing differentiating hematopoietic cells. This is in line with the recently found significant positive correlation between the upregulation of the HOX-A gene cluster and the expression of A4M in primary t(4;11)+ infant B-cell ALL samples, and with previous studies identifying that approximately one-half of t(4;11)+ patients do not have an activated HOX-A signature.20,65,66 This may explain why MA4 failed recently to bind to HOX-A genes to regulate HOXA gene expression.14 Collectively, MA4 and A4M might cooperate through a complex molecular interaction to control HOX-A gene regulation.25 In conclusion, we describe a functional and molecular cooperation between MA4 and A4M fusions during human hematopoietic development, and demonstrate how hESC-based hematopoietic differentiation represents a promising system to explore the developmental impact of the chimeric 1199

C. Bueno et al.

proteins resulting from chromosomal translocations, which remains obscure in human leukemia.

Acknowledgments Financial support for this work was obtained from the European Research Council (CoG-2014-646903 and PoC-2018-811220) and the Generalitat de Catalunya (SGR330 and PERIS 20172019) to PM, the Spanish Ministry of Economy and Competitiveness (SAF2016-80481-R and SAF2016-76758-R) to PM and IV, the Spanish Association against Cancer (AECCCI-2015) and Fero Foudation to CB, the Health Institute Carlos

References 1. Milne TA. Mouse models of MLL leukemia: recapitulating the human disease. Blood.

2017;129(16):2217-2223.

2. Montes R, Ayllon V, Prieto C, et al. Ligandindependent FLT3 activation does not cooperate with MLL-AF4 to immortalize/transform cord blood CD34+ cells. Leukemia.

2014;28(3):666-674.

3. Stam RW, den Boer ML, Schneider P, et al.

Targeting FLT3 in primary MLL-generearranged infant acute lymphoblastic leukemia. Blood. 2005;106(7):2484-2490.

4. Sanjuan-Pla A, Bueno C, Prieto C, et al.

Revisiting the biology of infant t(4;11)/MLLAF4+ B-cell acute lymphoblastic leukemia.

Blood. 2015;126(25):2676-2685.

5. Ford AM, Ridge SA, Cabrera ME, et al. In utero rearrangements in the trithorax-related oncogene in infant leukaemias. Nature.

1993;363(6427):358-360.

6. Greaves M. Infection, immune responses and the aetiology of childhood leukaemia.

Nat Rev Cancer. 2006;6(3):193-203.

7. Bueno C, Montes R, Catalina P, Rodriguez R, Menendez P. Insights into the cellular origin and etiology of the infant pro-B acute lymphoblastic leukemia with MLL-AF4 rearrangement. Leukemia. 2011;25(3):400410.

8. Barrett NA, Malouf C, Kapeni C, et al. MllAF4 confers enhanced self-renewal and lymphoid potential during a restricted window in development. Cell Rep. 2016;16(4):10391054.

9. Bueno C, Ayllon V, Montes R, et al. FLT3 activation cooperates with MLL-AF4 fusion protein to abrogate the hematopoietic specification of human ESCs. Blood.

2013;121(19):3867-3878, S3861-3863.

10. Bueno C, Montes R, Melen GJ, et al. A human ESC model for MLL-AF4 leukemic fusion gene reveals an impaired early hematopoietic-endothelial specification.

Cell Res. 2012;22(6):986-1002.

11. Bursen A, Schwabe K, Ruster B, et al. The AF4.MLL fusion protein is capable of inducing ALL in mice without requirement of MLL.AF4. Blood. 2010;115(17):3570-3579.

12. Chen W, Li Q, Hudson WA, et al. A murine Mll-AF4 knock-in model results in lymphoid and myeloid deregulation and hematologic malignancy. Blood. 2006;108(2):669-677.

13. Krivtsov AV, Feng Z, Lemieux ME, et al.

H3K79 methylation profiles define murine and human MLL-AF4 leukemias. Cancer Cell. 2008;14(5):355-368.

14. Lin S, Luo RT, Ptasinska A, et al. Instructive role of MLL-fusion proteins revealed by a model of t(4;11) pro-B acute lymphoblastic

1200

III (ISCIII/FEDER, PI17/01028 and PI17/01028) to CB and PJR, the NIHR GOSH BRC and Great Ormond Street Hospital Children's Charity to J.dB, and Bloodwise and Cancer Research UK to BG. RM and PM were also supported by the Deutsche José Carreras Leukämie Stiftung. PM also acknowledges financial support from the Obra Social La Caixa-Fundaciò Josep Carreras. R-T-R is supported by a fellowship from the Spanish Association of Cancer Research (AECC). RV-M is supported by a Torres Quevedo fellowship from the Spanish Ministry of Science and Innovation (PTQ-16-08623). P.M is an investigator of the Spanish Cell Therapy cooperative network (TERCEL).

leukemia. Cancer Cell. 2016;30(5):737-749.

15. Metzler M, Forster A, Pannell R, et al. A conditional model of MLL-AF4 B-cell tumourigenesis using invertor technology.

Oncogene. 2006;25(22):3093-3103.

16. Montes R, Ayllon V, Gutierrez-Aranda I, et al. Enforced expression of MLL-AF4 fusion in cord blood CD34+ cells enhances the hematopoietic repopulating cell function and clonogenic potential but is not sufficient to initiate leukemia. Blood. 2011;117(18): 4746-4758.

17. Tamai H, Miyake K, Takatori M, et al.

Activated K-Ras protein accelerates human MLL/AF4-induced leukemo-lymphomogenicity in a transgenic mouse model.

Leukemia. 2011;25(5):888-891.

18. Kowarz E, Burmeister T, Lo Nigro L, et al.

Complex MLL rearrangements in t(4;11) leukemia patients with absent AF4.MLL fusion allele. Leukemia. 2007;21(6):12321238.

19. Agraz-Doblas A. Bueno C, Bashford-Rogers R., et al. Unraveling the cellular origin and clinical prognostic markers of infant B-cell acute lymphoblastic leukemia using genome-wide analysis. Haematologica.

2019;104(6):1176-1188 20. Trentin L, Giordan M, Dingermann T, et al.

Two independent gene signatures in pediatric t(4;11) acute lymphoblastic leukemia patients. Eur J Haematol. 2009;83(5):406419.

21. Kumar AR, Yao Q, Li Q, Sam TA, Kersey JH. t(4;11) leukemias display addiction to MLLAF4 but not to AF4-MLL. Leuk Res. 2011;35(3):305-309.

22. Sanders DS, Muntean AG, Hess JL. Significance of AF4-MLL reciprocal fusion in t(4;11) leukemias? Leuk Res. 2011;35(3):299300.

23. Prieto C, Marschalek R, Kuhn A, et al. The AF4-MLL fusion transiently augments multilineage hematopoietic engraftment but is not sufficient to initiate leukemia in cord blood CD34(+) cells. Oncotarget. 2017;8 (47):81936-81941.

24. Rego EM, Pandolfi PP. Reciprocal products of chromosomal translocations in human cancer pathogenesis: key players or innocent bystanders?

Trends Mol Med. 2002;8(8):396-405.

25. 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.

26. Menendez P, Bueno C, Wang L. Human embryonic stem cells: a journey beyond cell replacement therapies. Cytotherapy. 2006;8 (6):530-541.

27. Romero-Moya D, Santos-Ocana C, Castano J, et al. Genetic rescue of mitochondrial and skeletal muscle impairment in an induced pluripotent stem cells model of coenzyme Q10 deficiency. Stem Cells. 2017;35(7): 1687-1703.

28. Menendez P, Vargas A, Bueno C, et al.

Quantitative analysis of bcl-2 expression in normal and leukemic human B-cell differentiation. Leukemia. 2004;18(3):491-498.

29. Ramos-Mejia V, Melen GJ, Sanchez L, et al.

Nodal/activin signaling predicts human pluripotent stem cell lines prone to differentiate toward the hematopoietic lineage. Mol Ther. 2010;18(12):2173-2181.

30. Wang L, Li L, Shojaei F, et al. Endothelial and hematopoietic cell fate of human embryonic stem cells originates from primitive endothelium with hemangioblastic properties. Immunity. 2004;21(1):31-41.

31. Diehl F, Rossig L, Zeiher AM, Dimmeler S, Urbich C. The histone methyltransferase MLL is an upstream regulator of endothelialcell sprout formation. Blood. 2007;109(4): 1472-1478.

32. Hatzipantelis ES, Athanassiou-Metaxa M, Gombakis N, et al. Thrombomodulin and von Willebrand factor: relation to endothelial dysfunction and disease outcome in children with acute lymphoblastic leukemia.

Acta Haematol. 2011;125(3):130-135.

33. Menendez P, Catalina P, Rodriguez R, et al.

Bone marrow mesenchymal stem cells from infants with MLL-AF4+ acute leukemia harbor and express the MLL-AF4 fusion gene. J Exp Med. 2009;206(13):3131-3141.

34. Ramos-Mejia V, Fernandez AF, Ayllon V, et al. Maintenance of human embryonic stem cells in mesenchymal stem cell-conditioned media augments hematopoietic specification. Stem Cells Dev. 2012;21(9):1549-1558.

35. Sanchez L, Gutierrez-Aranda I, Ligero G, et al. Maintenance of human embryonic stem cells in media conditioned by human mesenchymal stem cells obviates the requirement of exogenous basic fibroblast growth factor supplementation. Tissue Eng Part C Methods. 2012;18(5):387-396.

36. Bueno C, Montes R, Martin L, et al. NG2 antigen is expressed in CD34+ HPCs and plasmacytoid dendritic cell precursors: is NG2 expression in leukemia dependent on the target cell where leukemogenesis is triggered? Leukemia. 2008;22(8):1475-1478.

37. Castano J, Menendez P, Bruzos-Cidon C, et al. Fast and efficient neural conversion of human hematopoietic cells. Stem Cell Reports. 2014;3(6):1118-1131.

38. Munoz-Lopez A, Romero-Moya D, Prieto C, et al. Development refractoriness of MLLrearranged human B cell acute leukemias to

haematologica | 2019; 104(6)

AF4-MLL in human embryonic hematopoietic development

39.

40.

41.

42.

43.

44.

45.

46.

47.

reprogramming into pluripotency. Stem Cell Reports. 2016;7(4): 602-618.

Bueno C, Sardina JL, Di Stefano B, et al.

Reprogramming human B cells into induced pluripotent stem cells and its enhancement

by C/EBPalpha. Leukemia. 2016;30(3):674682.

Giorgetti A, Castano J, Bueno C, et al.

Proinflammatory signals are insufficient to drive definitive hematopoietic specification of human HSCs in vitro. Exp Hematol.

2017;45:85-93 e82.

Menendez P, Wang L, Chadwick K, Li L, Bhatia M. Retroviral transduction of hematopoietic cells differentiated from human embryonic stem cell-derived CD45(neg)PFV hemogenic precursors. Mol Ther. 2004;10(6):1109-1120.

Bueno C, Roldan M, Anguita E, et al. Bone marrow mesenchymal stem cells from patients with aplastic anemia maintain functional and immune properties and do not contribute to the pathogenesis of the disease. Haematologica. 2014;99(7):1168-1175.

Ramos-Mejia V, Navarro-Montero O, Ayllon V, et al. HOXA9 promotes hematopoietic commitment of human embryonic stem cells. Blood. 2014;124 (20):3065-3075.

Toscano MG, Navarro-Montero O, Ayllon V, et al. SCL/TAL1-mediated transcriptional network enhances megakaryocytic specification of human embryonic stem cells. Mol Ther. 2015;23(1):158-170.

Bueno C, Montes R, Menendez P. The ROCK inhibitor Y-27632 negatively affects the expansion/survival of both fresh and cryopreserved cord blood-derived CD34+ hematopoietic progenitor cells: Y-27632 negatively affects the expansion/survival of CD34+HSPCs. Stem Cell Rev. 2010;6(2): 215-223.

Rubio R, Garcia-Castro J, Gutierrez-Aranda I, et al. Deficiency in p53 but not retinoblastoma induces the transformation of mesenchymal stem cells in vitro and initiates leiomyosarcoma in vivo. Cancer Res. 2010;70(10):4185-4194.

Ayllon V, Bueno C, Ramos-Mejia V, et al.

The Notch ligand DLL4 specifically marks

haematologica | 2019; 104(6)

48.

49.

50.

51.

52.

53.

54.

55.

56.

human hematoendothelial progenitors and regulates their hematopoietic fate.

Leukemia. 2015;29(8):1741-1753.

Vodyanik MA, Slukvin, II. Hematoendothelial differentiation of human embryonic stem cells. Curr Protoc Cell Biol. 2007;Chapter 23:Unit 23 26.

Bueno C, Montes R, de la Cueva T, Gutierrez-Aranda I, Menendez P. Intra-bone marrow transplantation of human CD34(+) cells into NOD/LtSz-scid IL-2rgamma(null) mice permits multilineage engraftment without previous irradiation. Cytotherapy.

2010;12(1):45-49.

Gutierrez-Aranda I, Ramos-Mejia V, Bueno C, et al. Human induced pluripotent stem cells develop teratoma more efficiently and faster than human embryonic stem cells regardless the site of injection. Stem Cells.

2010;28(9):1568-1570.

Vodyanik MA, Bork JA, Thomson JA, Slukvin, II. Human embryonic stem cellderived CD34+ cells: efficient production in the coculture with OP9 stromal cells and analysis of lymphohematopoietic potential.

Blood. 2005;105(2):617-626.

Prieto C, Stam RW, Agraz-Doblas A, et al.

Activated KRAS cooperates with MLL-AF4 to promoteextramedullary engraftment and migration of cord blood CD34+ HSPC but is insufficient to initiate leukemia. Cancer Res. 2016;76(8):2478-2489.

Deshpande AJ, Deshpande A, Sinha AU, et al. AF10 regulates progressive H3K79 methylation and HOX gene expression in diverse AML subtypes. Cancer Cell.

2014;26(6):896-908.

Guenther MG, Lawton LN, Rozovskaia T, et al. Aberrant chromatin at genes encoding stem cell regulators in human mixed-lineage leukemia. Genes Dev. 2008;22(24):34033408.

Andersson AK, Ma J, Wang J, et al. The landscape of somatic mutations in infant MLLrearranged acute lymphoblastic leukemias.

Nat Genet. 2015;47(4):330-337.

Malouf C, Ottersbach K. The fetal liver lymphoid-primed multipotent progenitor provides the prerequisites for the initiation of t(4;11) MLL-AF4 infant leukemia.

57.

58.

59.

60.

61.

62.

63.

64.

65.

66.

Haematologica. 2018 Jun 14. [Epub ahead of print] Peters DG, Klucher KM, Perlingeiro RC, et al. Autocrine and paracrine effects of an EScell derived, BCR/ABL-transformed hematopoietic cell line that induces leukemia in mice. Oncogene. 2001;20(21): 2636-2646.

Ji J, Risueno RM, Hong S, et al. Brief report: ectopic expression of NUP98-HOXA10 augments erythroid differentiation of human embryonic stem cells. Stem Cells. 2011; 29(4):736-741.

Tan YT, Ye L, Xie F, et al. Respecifying human iPSC-derived blood cells into highly engraftable hematopoietic stem and progenitor cells with a single factor. Proc Natl Acad Sci U S A. 2018;115(9):2180-2185.

Marschalek R. Mechanisms of leukemogenesis by MLL fusion proteins. Br J Haematol.

2011;152(2):141-154.

Boisset JC, van Cappellen W, Andrieu-Soler C, et al. In vivo imaging of haematopoietic cells emerging from the mouse aortic endothelium. Nature. 2010;464(7285):116120.

Ahmad K, Scholz B, Capelo R, et al. AF4 and AF4-MLL mediate transcriptional elongation of 5-lipoxygenase mRNA by 1, 25-dihydroxyvitamin D3. Oncotarget. 2015;6(28):2578425800.

Benedikt A, Baltruschat S, Scholz B, et al.

The leukemogenic AF4-MLL fusion protein causes P-TEFb kinase activation and altered epigenetic signatures. Leukemia. 2011;25 (1):135-144.

Muck F, Bracharz S, Marschalek R. DDX6 transfers P-TEFb kinase to the AF4/AF4N (AFF1) super elongation complex. Am J Blood Res. 2016;6(3):28-45.

Driessen EM, van Roon EH, SpijkersHagelstein JA, et al. Frequencies and prognostic impact of RAS mutations in MLLrearranged acute lymphoblastic leukemia in infants. Haematologica. 2013;98(6):937944.

Kuhn A, Loscher D, Marschalek R. The IRX1/HOXA connection: insights into a novel t(4;11)- specific cancer mechanism.

Oncotarget. 2016;7(23):35341-35352.

1201

ARTICLE Ferrata Storti Foundation

Haematologica 2019 Volume 104(6):1202-1208

Non-Hodgkin Lymphoma

Association of early disease progression and very poor survival in the GALLIUM study in follicular lymphoma: benefit of obinutuzumab in reducing the rate of early progression John F. Seymour,1 Robert Marcus,2 Andrew Davies,3 Eve Gallop-Evans,4 Andrew Grigg,5 Andrew Haynes,6 Michael Herold,7 Thomas Illmer,8 Herman Nilsson-Ehle,9 Martin Sökler,10 Ulrich Dünzinger,11 Tina Nielsen,12 Aino Launonen12 and Wolfgang Hiddemann13

Peter MacCallum Cancer Centre, Royal Melbourne Hospital and University of Melbourne, Victoria, Australia; 2Kings College Hospital, London, UK; 3Cancer Research UK Centre, University of Southampton, UK; 4Velindre Cancer Centre, Cardiff, UK; 5Austin Hospital, Melbourne, Victoria, Australia; 6Nottingham University Hospitals NHS Trust, UK; 7 HELIOS-Klinikum Erfurt, Germany; 8BAG Freiberg-Richter, Jacobasch, Illmer and Wolf, Dresden, Germany; 9Section of Hematology and Coagulation, Department of Medicine, Sahlgrenska University Hospital, Gothenburg, Sweden; 10Eberhard-Karls-University Tübingen, Germany; 11Roche Pharma AG, Grenzach-Wyhlen, Germany; 12F. Hoffmann-La Roche Ltd., Basel, Switzerland and 13Department of Medicine III, Ludwig-MaximiliansUniversity, Munich, Germany 1

ABSTRACT

W

Correspondence: JOHN SEYMOUR john.seymour@petermac.org Received: October 14, 2018.

Accepted: December 17, 2018.

Pre-published: December 20, 2018. doi:10.3324/haematol.2018.209015 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/104/6/1202 ©2019 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

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

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

1202

e evaluated early disease progression and its impact on overall survival (OS) in previously untreated follicular lymphoma patients in GALLIUM (clinicaltrials.gov identifier: 01332968), and investigated the effect on early disease progression of the two randomization arms: obinutuzumab-based versus rituximab-based immunochemotherapy. Cause-specific Cox regression was used to estimate the effect of treatment on the risk of disease progression or death due to disease progression within 24 months of randomization and to analyze OS in patients with or without disease progression after 24 months.

Mortality in both groups was analyzed 6, 12, and 18 months post randomization (median follow up, 41 months). Fewer early disease progression events occurred in obinutuzumab (57 out of 601) versus rituximab (98 out of 601) immunochemotherapy patients, with an average risk reduction of 46.0% (95%CI: 25.0-61.1%; cumulative incidence rate 10.1% vs. 17.4%).

At a median post-progression follow up of 22.6 months, risk of mortality increased markedly following a progression event [HR of time-varying progression status, 25.5 (95%CI: 16.2-40.3)]. Mortality risk was higher the earlier patients progressed within the first 24 months. Age-adjusted HR for OS after 24 months in surviving patients with disease progression versus those without was 12.2 (95%CI: 5.6-26.5). Post-progression survival was similar by treatment arm. In conclusion, obinutuzumab plus chemotherapy was associated with a marked reduction in the rate of early disease progression events relative to rituximab plus chemotherapy. Early disease progression in patients with follicular lymphoma was associated with poor prognosis, with mortality risk higher after earlier progression. Survival post progression did not seem to be influenced by treatment arm.

Introduction Despite the favorable outcomes for patients with previously untreated follicular lymphoma (FL) that are now achievable with the current standard treatment of rituximab (R) plus chemotherapy (R-chemo) followed by R maintenance, 20-35% of patients still have progressive disease (PD), relapse, or death within two years.1-3 haematologica | 2019; 104(6)

Obinutuzumab reduces POD24 risk in FL patients

Patients with early PD have a much poorer survival than those with later disease progression.4,5 The glycoengineered anti-CD20 antibody obinutuzumab (GA101; G) has reduced complement-dependent cytotoxicity compared with R, but its direct cell-killing ability, as well as its antibody-dependent cell-mediated cytotoxicity and antibody-dependent cellular phagocytosis, are stronger than those of R.6,7 In the recently reported GALLIUM study (clinicaltrials.gov identifier: 01332968) in previously untreated patients with advanced FL, investigatorassessed progression-free survival (PFS) was significantly improved with G-based immunochemotherapy (Gchemo) relative to R-chemo [hazard ratio (HR), 0.68; 95% confidence interval (CI): 0.54-0.87; P=0.002].8 The aims of this exploratory analysis of patients with FL in GALLIUM were to evaluate early PD and its impact on overall survival (OS), and to investigate the impact of G-chemo versus R-chemo treatment on the incidence of early PD.

6, 12, 18, and 24 months post randomization) was estimated using Kaplan-Meier (KM) methods. These estimates allow a descriptive, non-randomized comparison of OS in patients with PD who were still alive versus those with no PD at the landmark.

The 24-month landmark analysis was carried out using a Cox regression model, with the same co-variates and stratification factors as the time-dependent model, but with POD24 status as a time-fixed covariate. The risk of a PFS event within 24 months of randomization was also estimated, and the 24-month KM rates were compared across treatment arms.

Post-progression mortality rates in POD24 patients were estimated for all patients and for each treatment arm separately, and for patients who progressed between 0-6, >6-12, >12-18, and >18-24 months from randomization. NoPOD24 mortality rates (overall and by treatment arm) were also calculated.

Results Patients’ characteristics

Methods GALLIUM was an open-label, randomized, parallel-group study in which patients with previously untreated grade 1-3a FL were randomized to receive induction therapy with G or R plus chemotherapy [cyclophosphamide, doxorubicin, vincristine and prednisone (CHOP), cyclophosphamide, vincristine and prednisone (CVP) or bendamustine] for 6 or 8 cycles, depending on the selected chemotherapy, followed in responders by two years of maintenance with the same antibody. GALLIUM was conducted in accordance with the International Conference on Harmonisation Guidelines for Good Clinical Practice. The protocol was approved by the ethics committees of all participating centers. All patients provided written informed consent. Patient selection, study methods and treatment are described in detail in the Online Supplementary Methods, and elsewhere.9 The principal end point in the current exploratory analysis was PD or death due to PD within 24 months of randomization (“POD24”); “noPOD24” was defined as patients who had completed 24 months of follow up with no PD or death due to PD during that time. Details of analyses to assess PD or suspected transformation are described in the Online Supplementary Appendix. Deaths from causes other than PD during the 24-month window in patients without prior PD, e.g. fatal adverse events or deaths from unrelated causes, contributed as events in noPOD24 mortality, but patients with these events are not included in the description of baseline characteristics for noPOD24 patients.

Analyses were based on investigator assessments and were undertaken using SAS® 9.4 software.

For all randomized patients, the effect of POD24 status on OS was quantified using a time-varying Cox regression model with POD24 status as a time-dependent co-variate. Other co-variates were allocated study treatment and age at randomization; chemotherapy regimen and Follicular Lymphoma International Prognostic Index (FLIPI) score were stratification factors. In this model (and for the calculation of crude death rates), all patients were considered POD24-free until transition to the POD24 group after a POD24 event, or until the end of survival follow up. The effect of the study treatment on the risk of POD24 was estimated by cause-specific Cox regression accounting for POD24 events and censoring for non-PD related deaths within the first 24 months. Analysis was stratified by chemotherapy regimen and FLIPI score.

Overall survival from 4 specified time points (“landmarks”; at haematologica | 2019; 104(6)

Of 1202 patients with FL randomized in GALLIUM (Online Supplementary Figure S1), 1071 were followed for response until at least 24 months post randomization and evaluated for POD24 status. For the remaining 131 patients, follow up was incomplete (mostly due to study discontinuation): 23 of these patients died from non-PD related causes without prior PD. POD24 events occurred in a total of 155 patients (12.9%), of whom 41 died within 24 months of randomization; 916 patients had no POD24 event after at least 24 months of follow up (Online Supplementary Table S1). The data cutoff for this analysis was September 10, 2016, giving a median follow up of 41 months.

Baseline characteristics for POD24 and noPOD24 patients showed some notable differences (Table 1).

Compared with noPOD24 patients, POD24 patients were more likely to have the following features: male sex, stage IV disease, grade 3a histology, high FLIPI risk, elevated serum lactate dehydrogenase (LDH), and bulky disease.

Relatively more POD24 patients than noPOD24 patients were treated with CVP and fewer were treated with bendamustine. Baseline characteristics for those patients who progressed within six months of randomization are shown in Online Supplementary Table S2.

Progressive disease or death due to progressive disease and progression-free survival by treatment arm Progressive disease or death due to progressive disease at 24-months post randomization events occurred in 57 of 601 patients (9.5%) treated with G-chemo and 98 of 601 patients (16.3%) treated with R-chemo; cumulative incidence rates were 10.1% (95%CI: 8-12%) and 17.4% (95%CI: 14-20%), respectively. The average HR-based reduction in the risk of a POD24 event with G-chemo relative to R-chemo was 46.0% (95%CI: 25.0-61.1%; P=0.0003, Gray’s test for equality) (Figure 1). The risk of a PFS event in the 24 months after randomization was 18.9% (95%CI: 15.9-22.4%) for R-chemo and 12.5% (95%CI: 10.1-15.6%) for G-chemo; the relative risk reduction for PFS events was 33.9% (95%CI: 12.849.8%) (Figure 1). In the noPOD24 group, with a median subsequent follow up of 24 months, 88 out of 916 patients had POD events after the 24-month time point [G-chemo, 40 out of 478 (8.4%); R-chemo, 48 out of 438 (11.0%)].

1203

J.F. Seymour et al.

Post-progression survival and mortality by treatment arm

Landmark analysis of overall survival and mortality

With a median follow up of 41 months, 56 out of 155 POD24 patients have died: 40 due to PD (G-chemo, n=15; R-chemo, n=25) and 16 for other reasons (n=4 and 12, respectively) (Online Supplementary Table S3). In addition, 16 out of 916 noPOD24 patients have died: one related to PD (allocated to the G-chemo arm but no study treatment received) and 15 for other reasons (G-chemo, n=9; Rchemo, n=6) (Online Supplementary Table S3). Post-progression survival in POD24 patients was similar in the two treatment arms at a median post-progression follow up of 22.6 months (Figure 2 and Online Supplementary Table S4).

In 110 patients with POD24 events who were still alive at the 24-month landmark (4 censored patients excluded), the OS rate at two years after the landmark was 82.4% (95%CI: 74.2-91.3%), compared with 98.2% (95%CI: 97.1-99.2%) in 916 noPOD24 patients; the age-adjusted HR for OS in this 2-year window was 12.2 (95%CI: 5.6-

Survival and mortality by chemotherapy regimen, Follicular Lymphoma International Prognostic Index risk category, and treatment arm Fewer POD24 events occurred in G-chemo versus R-chemo patients across all three chemotherapy regimen groups; in each treatment arm, the highest rate of events was observed in those treated with CVP. POD24 events were more common in patients with a high FLIPI score when compared with those patients with an intermediate or low FLIPI score (Online Supplementary Tables S5 and S6).

Histological transformation Histologically confirmed transformation to more aggressive lymphoma was observed in 35 patients: 13 (2.2%) in the G-chemo arm and 22 (3.7%) in the R-chemo arm, predominantly in the first 24 months (n=30). In POD24 patients, transformation occurred in almost identical proportions of G-chemo- and R-chemo-treated patients [11 out of 57 (19.3%) and 19 out of 98 (19.4%) patients, respectively]. Of 30 patients (19.4%) with transformation in the first 24 months, 16 died during study follow up (n=8 in each treatment arm). Of the 88 patients in the noPOD24 group with PD events after the 24-month time point, only 5 (5.7%) had transformations [G-chemo, 2 of 40 (5.0%); R-chemo, 3 of 48 (6.3%)]. Cumulative incidence data and lymphoma types at transformation are presented in the Online Supplementary Results, Online Supplementary Table S7 and Online Supplementary Figure S2.

New anti-lymphoma treatment New anti-lymphoma treatment (NALT) of any type was received by 100 of 155 patients (64.5%) with a POD event in the first 24 months after randomization (G-chemo, n=37; R-chemo, n=63) and by 24 of 88 patients (27.3%) with a POD event later than the 24-month time point (G-chemo, n=11; R-chemo, n=13) (Online Supplementary Table S8). A cumulative incidence plot of time from PD to NALT for those with PD before and after 24 months showed that 49.7% (95%CI: 41.3-57.5%) of POD24 patients started NALT within three months of PD, compared with 19.4% (95%CI: 11.2-29.3%) of those who progressed after 24 months (Online Supplementary Figure S3).

Details of the type of NALT received after PD according to the timing of POD are provided in the Online Supplementary Results and Online Supplementary Table S8.

Time-dependent mortality analysis Time-dependent regression analysis showed that, at any given time during study follow up, the risk of death was considerably greater following a POD24 event than in those without a POD24 event (stratified HR, 25.5; 95%CI: 16.2-40.3%) (Online Supplementary Table S9).

1204

Table 1. Baseline disease and patients’ characteristics in POD24 and noPOD24 subgroups with at least 24 months’ response follow up.

Characteristica

POD event within No POD event first 24 months within first (n=155) 24 months (n=916)

Age, years 58 (31-84) 59 (23-85) Aged ≥65 years 49 (31.6) 280 (30.6) Male 87 (56.1) 415 (45.3) Ann Arbor stage at diagnosis (n=1064) I or IIb 11/154 (7.1) 77/910 (8.5) III 46/154 (29.9) 327/910 (35.9) IV 97/154 (63.0) 506/910 (55.6) Bone marrow involvement (n=1063) 80/155 (51.6) 468/908 (51.5) Grade of FL 1 57 (36.8) 353 (38.5) 2 63 (40.6) 437 (47.7) 3a 35 (22.6) 117 (12.8) Other (incl. missing) 0 9 (1.0) Time from diagnosis to randomization, months (n=1068) 1.35 (0.0-83.5) 1.45 (0.1-129.3) (n=154) (n=914) FLIPI Low (0-1) 25 (16.1) 192 (21.0) Intermediate (2) 46 (29.7) 354 (38.6) High (≥3) 84 (54.2) 370 (40.4) FLIPI-2 (n=1043) Low (0) 6/153 (3.9) 93/890 (10.4) Intermediate (1-2) 68/153 (44.4) 452/890 (50.8) High (≥3) 79/153 (51.6) 345/890 (38.8) Elevated serum lactate dehydrogenase >ULN 79/154 (51.3) 248/912 (27.2) >1.5 × ULN 26/154 (16.9) 51/912 (5.6) >2 × ULN 10/154 (6.5) 18/912 (2.0) Investigator-assessed SPD 6531 5128.5 of ≤ 6 target lesions, mm2 (514.0-80,720.0) (235.2-62,725.0) Bulky disease ≥7 cm 79/155 (51.0) 392/914 (42.9) (n=1069) Randomized antibody treatment Obinutuzumab 57 (36.8) 478 (52.2) Rituximab 98 (63.2) 438 (47.8) Allocated chemotherapy treatment Bendamustine 76 (49.0) 526 (57.4) CHOP 55 (35.5) 305 (33.3) CVP 24 (15.5) 85 (9.3) n: number; CHOP: cyclophosphamide, doxorubicin, vincristine+prednisone; CVP: cyclophosphamide, vincristine+prednisone; FL: follicular lymphoma; FLIPI: Follicular Lymphoma International Prognostic Index; POD: progressive disease or death due to progressive disease; POD24: progressive disease or death due to progressive disease events in the 24 months after randomization; SPD: sum of products of 2 longest perpendicular dimensions for ≤6 target lesions; ULN: upper limit of normal. aData are n (%), n/N (%), or median (range). b18 patients were randomized to study treatment after being assessed as stage II or above by the investigators, meeting study eligibility criteria, but were reassessed as stage I after medical review. Staging data were missing for 7 patients (POD24, n=1; noPOD24, n=6).

haematologica | 2019; 104(6)

Obinutuzumab reduces POD24 risk in FL patients

26.5%) (Figure 3D). KM survival plots after the 6-, 12-, and 18-month landmarks showed that the risk of death in patients whose disease progressed was higher the earlier progression occurred (Figure 3A-D). This is also shown by the corresponding mortality rates after progression during the selected time windows (0-6, >6-12, >12-18, and >1824 months from randomization) (Table 2); rates during each time window were similar for G-chemo- and Rchemo-treated patients (Online Supplementary Table S10).

Discussion This analysis of updated data from the GALLIUM study confirmed the findings of prior studies4,10,11 showing that early disease progression in patients with FL is associated with a poor prognosis. The proportion of POD24 patients surviving at two years after the 24-month landmark was 82.4% (although 26% of POD24 patients had died before the 2-year landmark), whereas for the noPOD24 group,

the proportion surviving at this time point was 98.2%.

Irrespective of treatment arm, patients who had PD in the 24 months after randomization were also much more likely to receive additional lymphoma therapy within three months of progression than patients who had PD later. By all of these measures, we confirm the adverse outlook associated with a POD24 event.

Treatment with G-chemo was associated with a marked reduction in the rate of POD24 events relative to R-chemo and these data therefore support the superiority of the G-chemo regimen seen in the GALLIUM primary analysis.9 Notably, there was little difference in the proportion of patients in each treatment arm with PD after the 24-month time point, indicating that, to date, the protective effect of G-chemo against early progression has not been counterbalanced by any increased risk of later progression. With small numbers of deaths in both arms and comparatively short follow up (approximately 3.5 years), it should be noted that, despite a reduced risk of POD24 with G-chemo, we have not yet seen an impact on

Figure 1. Progression-free survival (PFS) and progressive disease or death due to progressive disease (POD) events in the 24 months after randomization (POD24) by treatment arm.

The table below the graph shows the number of PFS and POD events occurring in the 24 months post-randomization, along with the risk of these events. n: number; PD: progressive disease; CI: confidence interval; HR: hazard ratio; G-chemo: obinutuzumab plus chemotherapy; R-chemo: rituximab plus chemotherapy. aAll 155 patients had PD. bAt 24 months after randomization, deaths from any cause had occurred in 26 (Gchemo) and 38 (R-chemo) patients.

Table 2. Post-progression mortality rates, stratified by time of progression, in the POD24 group (n=155).

Time of progression, months

n

Deaths

0-6 >6-12 >12-18 >18-24

22 58 46 29

18 27 9 2

Reason for death PD Other 16 19 5 0

2 8 4 2

Patientâ&#x20AC;&#x201C;years at risk

Deaths per 100 patientâ&#x20AC;&#x201C;years (95%CI)

22.2 111.6 102.2 52.5

81.0 (51.2-100) 24.2 (16.6-35.3) 8.8 (4.6-16.9) 3.8 (1.0-15.2)

n: number; CI: confidence interval; PD: progressive disease; POD24: progressive disease or death due to progressive disease events in the 24 months after randomization.

haematologica | 2019; 104(6)

1205

J.F. Seymour et al.

OS in the GALLIUM study. However, this is not unexpected with current follow up given the available salvage therapies.

Our results suggest that POD24 events are more common in patients treated with CVP compared with patients treated with CHOP or bendamustine. However, it is difficult to draw firm conclusions from these results, as patient numbers within the subgroups are low. Furthermore, statistical comparisons across chemotherapy regimens were not performed, as chemotherapy allocation was not randomized; rather the chemotherapy treatment was selected by the investigator, with all patients at each individual center receiving the same regimen.

As previously reported, fewer patients with FL in the Gchemo arm than in the R-chemo arm started NALT during GALLIUM (estimated proportions without NALT at three years: 87% and 81%, respectively).9 However, the current analysis shows that patients who progressed early (within 24 months of randomization) were much more likely to receive NALT within three months of progression than patients who had PD later, irrespective of treatment arm.

Transformation to a more aggressive lymphoma at the time of PD occurred in just under 20% of those patients with POD24 events, with no discernible difference in these proportions by treatment arm. However, transformations appear to be much less frequent in patients with later progression; at the current data cutoff, only 5.7% of patients in the noPOD24 group experienced transformation.

The increase in mortality risk associated with POD24 in the current analysis (HR from stratified Cox regression analysis, 25.5) was much larger than in a recent pooled analysis of POD24 in 5453 patients from 13 clinical trials (HR for OS, 5.24)5 and the National LymphoCare Study (NLCS) analysis by Casulo et al.,4 which reported an HR of 8.1 in an exploratory analysis that used the same method as in the current analysis, i.e. POD24 as a time-varying covariate and adjusting for FLIPI. Importantly, the follow-up duration in our analysis was shorter than in the NLCS (medians, 3.4 and 7 years, respectively); as the follow-up

duration in GALLIUM increases, the HR value is likely to fall, as we showed that the risk of death is higher for patients with earlier progression. However, survival data for the first two years of follow up in our study look remarkably similar to the NLCS analysis. In GALLIUM, the proportion of POD24 patients surviving at two years after progression was 66% (95%CI: 58.3-73.9%) whereas for the noPOD24 group, the proportion surviving at two years after the 24-month landmark was 98.2% (95%CI: 97.1-99.2); in NLCS, the corresponding proportions were 68% (95%CI: 58.2-76.3) and 97% (95%CI: 94.6-98.1), respectively.4 Our findings also suggest that the earlier progression occurs, the greater the subsequent mortality risk. This subset of patients with FL in whom the disease follows a relatively aggressive clinical course may benefit from alternative initial therapies. Therefore, identifying these patients is crucial. The baseline factor in the current analysis most strongly associated with risk of POD24 was serum LDH level, in line with previous studies that have identified LDH as a prognostic factor for progression in patients with FL.12,13 Other factors associated with greater POD risk were FLIPI risk group, disease stage and histological grade, male sex, and treatment with CVP chemotherapy; although the extent of difference in POD risk accounted for by these was not as great as the influence of anti-CD20 antibody treatment.

Promising prospective identification of patients with a high risk of early PD events has been achieved by using the m7-FLIPI, a clinicogenetic risk model incorporating the impact of mutations in the genes EZH2, ARID1A, MEF2B, EP300, FOXO1, CREBBP, and CARD11.14 The m7-FLIPI had higher predictive accuracy for POD24 than either the FLIPI or another specific clinicogenetic risk model, the POD24 Prognostic Index (POD24-PI), in an analysis of 258 patients from two independent cohorts.10 A new 23-gene expression model was also reported to predict early progression.15 Additionally, TP53 mutation status has been evaluated as a prognostic factor in patients with FL and found to be significantly correlated with shorter PFS, OS,

Figure 2. Overall survival in POD24 patients post-progression by treatment arm. Shaded sections of lines show 95% Hall-Wellner confidence bands for the period during which patients died. G-chemo: obinutuzumab plus chemotherapy; R-chemo: rituximab plus chemotherapy; N: number. POD24: progressive disease or death due to progressive disease events in the 24 months after randomization.

1206

haematologica | 2019; 104(6)

Obinutuzumab reduces POD24 risk in FL patients

and disease transformation.16,17 Interrogation of the GALLIUM dataset using m7-FLIPI and other potentially predictive indices is ongoing.

In conclusion, our results are consistent with previous studies that found a particularly poor outlook for patients with FL who suffer PD within 24 months of starting immunochemotherapy, and also further refine this obser-

vation by showing that there is a gradation of prognosis even within the POD24 subset, with earlier events predicting more adverse outcomes. We observed that patients receiving G-based treatment have a much lower POD24 rate than those receiving R-based treatment; this supports the contention that more effective first-line treatment that reduces the POD24 rate may yield long-term benefit. Our

A

B

C

D

Figure 3. Landmark overall survival (OS) analysis, comparing patients with progressive disease or death due to progressive disease (POD) before the landmark and patients with noPOD. (A) 6-month, (B) 12-month, (C) 18-month, and (D) 24-month landmarks. The shaded sections of lines show 95% Hall-Wellner confidence bands for the period during which patients died. The table below the graph shows 2-year OS estimates (with 95%CI) at each landmark. N: number; CI: confidence interval; HR: hazard ratio.

haematologica | 2019; 104(6)

1207

J.F. Seymour et al.

analysis also adds to the weight of data supporting the use of POD24 rate as an efficacy end point for future clinical trials in patients with previously untreated FL. Acknowledgments The authors would like to thank all of the patients who

References 1. Marcus R, Imrie K, Belch A, et al. CVP chemotherapy plus rituximab compared with CVP as first-line treatment for advanced follicular lymphoma. Blood.

2005;105(4):1417-1423.

2. Press OW, Unger JM, Rimsza LM, et al.

Phase III randomized intergroup trial of CHOP plus rituximab compared with CHOP chemotherapy plus (131)iodine-tositumomab for previously untreated follicular non-Hodgkin lymphoma: SWOG S0016. J Clin Oncol. 2013;31(3):314-320.

3. Rummel MJ, Niederle N, Maschmeyer G, et al. Bendamustine plus rituximab versus CHOP plus rituximab as first-line treatment for patients with indolent and mantle-cell lymphomas: an open-label, multicentre, randomised, phase 3 non-inferiority trial.

Lancet. 2013;381(9873):1203-1210.

4. Casulo C, Byrtek M, Dawson KL, et al. Early relapse of follicular lymphoma after rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisone defines patients at high risk for death: an analysis from the National LymphoCare Study. J Clin Oncol.

2015;33(23):2516-2522.

5. Casulo C, Le-Rademacher J, Dixon J, et al.

Validation of POD24 as a robust early clinical endpoint of poor survival in follicular lymphoma: results from the Follicular Lymphoma Analysis of Surrogacy Hypothesis (FLASH) investigation using individual data from 5,453 patients on 13

1208

participated in the GALLIUM study, and acknowledge all of the study investigators and their staff as well as the GALLIUM study team. Third-party medical writing assistance, under the direction of John Seymour, was provided by Roger Nutter, Scott Malkin and Helen Cathro of Gardiner-Caldwell Communications, and was funded by F. Hoffmann-La Roche Ltd.

clinical trials. Blood. 2017;130(Suppl 1):412.

6. Herter S, Herting F, Mindigl O, et al.

Preclinical activity of the type II CD20 antibody GA101 (obinutuzumab) compared with rituximab and ofatumumab in vitro and in xenograft models. Mol Cancer Ther.

2013;12(10):2031-2042.

7. Mössner E, Brünker P, Moser S, et al.

Increasing the efficacy of CD20 antibody therapy through the engineering of a new type II anti-CD20 antibody with enhanced direct and immune effector cell-mediated Bcell cytotoxicity. Blood. 2010; 115(22):43934402.

8. Hiddemann W, Barbui AM, Canales MA, et al. Immunochemotherapy with obinutuzumab or rituximab for previously untreated follicular lymphoma in the GALLUM study: influence of chemotherapy on efficacy and safety. J Clin Oncol. 2018; 36(23):2395-2404.

9. Marcus R, Davies A, Ando K, et al.

Obinutuzumab for the first-line treatment of follicular lymphoma. N Engl J Med. 2017;377(14):1331-1344.

10. Jurinovic V, Kridel R, Staiger AM, et al.

Clinicogenetic risk models predict early progression of follicular lymphoma after firstline immunochemotherapy.

Blood.

2016;128(8):1112-1120.

11. Maurer MJ, Bachy E, Ghesquières H, et al.

Early event status informs subsequent outcome in newly diagnosed follicular lymphoma. Am J Hematol. 2016;91(11):10961101.

12. Montoto S, Lopez-Guillermo A, Ferrer A, et al. Survival after progression in patients with follicular lymphoma: analysis of prognostic factors. Ann Oncol. 2002;13(4):525-530.

13. Murakami S, Kato H, Yamamoto K, et al.

Combination of high serum lactate dehydrogenase levels at the time of first diagnosis and progression predicts early lymphoma related death in patients with follicular lymphoma receiving R-CHOP regimen. Blood.

2014;124(21):2972.

14. Pastore A, Jurinovic V, Kridel R, et al.

Integration of gene mutations in risk prognostication for patients receiving first-line immunochemotherapy for follicular lymphoma: a retrospective analysis of a prospective clinical trial and validation in a population-based registry. Lancet Oncol.

2015;16(9):1111-1122.

15. Huet S, Tesson B, Jais JP, et al. A geneexpression profiling score for prediction of outcome in patients with follicular lymphoma: a retrospective training and validation analysis in three international cohorts.

Lancet Oncol. 2018;19(4):549-561.

16. O’Shea D, O’Riain C, Taylor C, et al. The presence of TP53 mutation at diagnosis of follicular lymphoma identifies a high-risk group of patients with shortened time to disease progression and poorer overall survival. Blood. 2008;112(8):3126-3129.

17. Kridel R, Chan FC, Mottok A, et al.

Histological transformation and progression in follicular lymphoma: a clonal evolution study. PLoS Med. 2016;13(12):e1002197.

haematologica | 2019; 104(6)

ARTICLE

Plasma Cell Disorders

Proteolysis targeting chimeric molecules as therapy for multiple myeloma: efficacy, biomarker and drug combinations

Ferrata Storti Foundation

Su Lin Lim,1* Alisa Damnernsawad,2* Pavithra Shyamsunder,3 Wee Joo Chng,3 Bing Chen Han,1 Liang Xu,3 Jian Pan,1 Dakle Pushkar Pravin,3 Serhan Alkan,1 Jeffrey W. Tyner2** H. Phillip Koeffler1,3**

Cedars Sinai Medical Center, Los Angeles, CA, USA; 2Division of Hematology and Medical Oncology, Oregon Health and Science University Knight Cancer Institute, Portland, OR, USA and 3Cancer Science Institute of Singapore, National University of Singapore, Singapore 1

Haematologica 2019 Volume 104(6):1209-1220

*SLL and AD contributed equally to this work. **JWT and HPK contributed equally to this work.

ABSTRACT

P

roteolysis targeting chimeric molecule ARV 825 causes ubiquitination of bromodomains resulting in their efficient degradation by proteasome activity. Bromodomain degradation down-regulates MYC transcription contributing to growth inhibition of various human cancers. We examined the therapeutic potential of ARV 825 against multiple myeloma (MM) cells both in vitro and in vivo. In a dose-dependent manner, ARV 825 inhibited proliferation of 13 human MM cell lines and three fresh patient samples, and was associated with cell cycle arrest and apoptosis. ARV 825 rapidly and efficiently degraded BRD 2 and BRD 4. Sensitivity of MM cells to ARV 825 was positively correlated with cereblon levels. RNA sequencing analysis showed important genes such as CCR1, RGS, MYB and MYC were down-regulated by ARV 825. A total of 170 small molecule inhibitors were screened for synergy with ARV 825. Combination of ARV 825 with inhibitor of either dual PI3K/mTOR, CRM1, VEGFR, PDGFRÎą/b, FLT3, IGF-1R, protein kinase C, CBP-EP300 or JAK1/2 showed synergistic activity. Importantly, ARV 825 significantly inhibited the growth of MM xenografts and improved mice survival. Taken together, our results, in conjunction with recently published findings, provide a rationale for investigating the efficacy of ARV 825 for MM, use of cereblon as a biomarker for therapy of MM patients, and the combination of ARV 825 with small molecule inhibitors to improve the outcome of MM patients.

Correspondence: SU LIN LIM sulin_lim_86@hotmail.com Received: July 11, 2018.

Accepted: January 2, 2019.

Pre-published: January 3, 2019. doi:10.3324/haematol.2018.201483

Introduction Multiple myeloma (MM) is characterized by neoplastic proliferation of clonal plasma cells producing a monoclonal immunoglobulin. It accounts for more than 17% of hematologic malignancies in the US.1 Over the past decade, newly introduced therapeutic regimens (e.g. proteasome inhibitor and immunomodulatory drugs) have significantly improved treatment outcome and survival of MM. Nevertheless, most of these patients eventually relapse, underlining the need for new therapeutic approaches. Agents with novel mechanism of action such as monoclonal antibodies (e.g. daratumumab, elotuzumab), histone deacetylase inhibitors, kinesin spindle protein inhibitors, and cereblon modulator iberdomide are under ongoing investigation for treating MM. Other than that, chimeric antigen receptor T (CAR-T) cells directed against B-cell maturation antigen (BCMA) have shown promising tumor cell reduction in MM.2 The search for novel agents is rapidly expanding and this, together with identification of novel combinations, should help revolutionize treatment of this disease.

Bromodomains (BRD) 2, 3 and 4, and T are members of the bromodomain extraterminal domain (BET) family facilitating transcriptional activation by RNA polymerase II.3 BRD 2, 3 and 4 bind to acetylated chromatin promoting progression from G1 to haematologica | 2019; 104(6)

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/104/6/1209 Š2019 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

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

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

1209

S.L. Lim et al.

Cell Titer-Glo luminescent cell viability assay

S phase of the cell cycle by direct interaction with positive transcription elongation factor complex b.4 BRD 4 is often located in super-enhancer regions associated with key genes (e.g. MYC, IGLL5, IRF4, PRDM1/BLIMP-1, and XBP1). These super-enhancer driven genes are also important in MM biology, playing key roles in controlling cell proliferation.5 The BET inhibitor JQ1 has potent anti-MM activity in vitro and in vivo,6 but its reversible binding to BRD proteins causes incomplete transcriptional repression of MYC and other oncoproteins.7 It also does not induce apoptosis in MM. ARV 825 (Arvinas Inc., New Haven, CT, USA) is a hetero-bifunctional molecule composed of a BRD 4 binding moiety (OTX015) joined to pomalidomide. The latter binds to an E3 ubiquitin ligase, cereblon (CRBN) and OTX 015 brings the complex to the BRD molecules. These drugs are called PROTAC (proteolysis targeting chimeric molecules) causing ubiquitination of BRD, resulting in rapid and efficient degradation by proteasome activity..9,10 PROTAC have potent activity against lymphoma, leukemia, and prostate cancers.7,9-11 Their activity on myeloma models has also been described. dBET1 (composed of JQ1 joined to thalidomide) promoted degradation of BRD 4 in an MM cell line (MM1S)12 and a recent publication showed that BET targeted PROTAC (ARV 825 and ARV 763) has anti-myeloma activity associated with decreasing MYC and Akt/mTOR.13 The authors also showed that ARV 825 was active against primary myeloma cells both in vitro and in vivo, and could overcome drug resistance in MM cells.13 In this study, we demonstrate that ARV 825 (8.5-500 nM) inhibited cell proliferation of 13 human MM cell lines and three fresh myeloma samples in vitro. In addition, the drug induced apoptosis, cell cycle arrest in vitro, and had growth inhibitory activity against MM cells in vivo. This PROTAC inhibited growth of MM cells resistant to either glucocorticoids or bortezomib, as well as those with t(4:14) translocation and FGFR3 and MMSET overexpression (poor prognosis). We identified prominent levels of CRBN as a biomarker of responsiveness to the drug.

Those MM cells resistant to ARV 825 were sensitive to another PROTAC (MZ1) relying on a different E3 ligase (VHL). We also examined 170 drugs [US Food and Drug Administration (FDA)-approved or in clinical trial] for their ability to enhance the cell inhibitory activity of ARV 825. In depth analysis showed synergy of ARV 825 with either LY3023414 (dual PI3K/mTOR inhibitors), selinexor (CRM1 inhibitor), cediranib (VEGFR inhibitor), crenolanib (PDGFRα/b and FLT3 inhibitor), GSK 1904529A (IGF-1R inhibitor), motesanib (VEGFR1/2/3 inhibitor), gilteritinib (FLT3/AXL inhibitor), LY333531 (protein kinase C inhibitor), IGC003 (CBP-EP300 inhibitor), or ruxolitinib (JAK1/2 inhibitor).

For in vitro and in vivo experiments, the statistical significance of difference between two groups used two-tailed Student t-test and two-way ANOVA. Significant differences between experimental groups in comparison to controls are shown: *P<0.01; **P<0.001; ***P<0.0001. Means±Standard Deviation (SD) are shown.

All animal care and experimental procedures in this study complied with the protocol approved by the Institutional Animal Care and Use Committee at Cedars Sinai Medical Center, Los Angeles, CA, USA. For detailed information on the materials and methods used see the Online Supplementary Appendix.

Methods

Results

Cell culture

ARV 825 significantly inhibits cellular proliferation and clonogenic growth of multiple myeloma cells

All cell lines were cultured and maintained in RPMI1640 containing 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin (Invitrogen, Carslbad, CA, USA) at 37°C with 5% CO2. The 8226 LR5 cells were maintained in 10 nM melphalan, the 8226 P100V cells were cultured with 100 nM bortezomib for two days every two weeks. Short random repeat (STR) analysis was carried out on all cell lines used in this study.

1210

Three primary MM patient samples obtained from their bone marrow (BM) were plated in 96-well plates with different concentrations of ARV 825 [dimethyl sulfoxide (DMSO) as a vehicle].

After 48 hours (h), cell viability was determined using the CellTiter-Glo® luminescent cell viability kit according to the manufacturer's instructions. The luminescence was measured by a luminometer (GloMax®-Multi Detection System Madison, WI, USA). All experiments were repeated at least three times. The means with standard deviations were shown.

Lentivirus production, gene knockdown and overexpression of CRBN shRNA targeting CRBN in pLKO.1 lentiviral vector (Sequence: CCGGGCCCACGAATAGTTGTCATTTCTCGAGAAATGACAACTATTCGTGGGCTTTTTG) and pLX304-CRBN-V5 vector (PMID: 29764999) were a kind gift from Dr. X. Liang (Cancer Science Institute, Singapore). Luciferase vector was purchased from Addgene (plasmid #17477). Recombinant lentiviral vector and packaging vector (pCMV-dR8.9 and pMD2.G-VSVG) were co-transfected into 293 FT cells using polyethylenimine (PEI) according to the manufacturer’s instructions. Virus supernatants were harvested at 48 h and 72 h after transfection, and placed through a 0.45 µm filter. KMS11 and KMS28BM cells (1x106 per well) were seeded in 6-well plates. Cells were transduced with lentivirus vectors in the presence of 8 μg/mL polybrene (SigmaAldrich) for 24 h. Stable cell lines were selected with either puromycin or blasticidin.

In vivo xenografts To access the in vivo activity of ARV 825, KMS11 expressing luciferase (KMS11LUC) were injected into the lateral tail vein of SCID-Beige mice (n=9) versus diluent control mice (n=9). Mice were monitored for 14 days and images were taken with a Xenogen IVIS spectrum camera (PerkinElmer, MA, USA) to document engraftment before treatment was initiated. After 14 days, mice were treated with either vehicle alone (5% Kolliphor® HS15) or 5 mg/kg of ARV 825 (daily intraperitoneal injection).

Tumor burden in each treatment group was photographed weekly with a Xenogen camera and the overall survival (OS) monitored.

Statistical analysis

Structures of ARV 825 and MZ1 are shown in Figure 1A. The ARV 825 is composed of the BET inhibitor OTX 015 conjugated to the ligand for cereblon E3 ligase.

Another PROTAC (MZ1) is composed of the BET inhibitor JQ1 conjugated to the ligand for VHL E3 ligase.

ARV 825 was tested in a dose-dependent manner against haematologica | 2019; 104(6)

Proteolysis targeting chimeric molecules in MM

A

B

C

D

E

F

Figure 1. ARV 825 anti-proliferative activities: MTT and clonogenic growth of multiple myeloma (MM) cells. (A) Structures of the two proteolysis targeting chimeric molecules (PROTAC) used in this study: ARV 825 and MZ1. (B) Growth inhibition of KMS11 and KMS28BM cells treated with pomalidomide (1 μM-20 μM), OTX015 (1 μM-10 μM), and ARV 825 (1 nM-1000 nM) for 72 hours (h). Results are Mean±Standard Deviation (SD); three experiments were carried out in triplicate. (C) MM cells were treated with ARV 825 (1 nM-500 nM, 72 h). Growth inhibition was measured by MTT assay. Results are Mean±SD; n=3. IC50s are shown in Online Supplementary Table S1. (D) ARV 825 decreased clonogenic growth of KMS11 and KMS28BM cells. Mean±SD of two independent experiments carried out in triplicates. Student t-test, **P≤0.001; ***P≤0.0001. (E) MM cells treated with PROTAC MZ1 (VHL E3 ligase fused to JQ1; 1-1000 nM, 72 h, and measured by MTT assay). Results represent Mean±SD of three experiments carried out in triplicate. IC50s are shown in Online Supplementary Table S2. (F) Primary patient samples (Patient 1, Patient 2, Patient 3) treated with ARV 825 (10-300 nM, 72 h). Growth inhibition measured by luminescence cell viability assay. Results are Mean±SD; n=3.

haematologica | 2019; 104(6)

1211

S.L. Lim et al.

A

B

C

D

Figure 2. CRBN mRNA levels of multiple myeloma (MM) cells and their correlation with sensitivity to ARV 825. (A) CRBN expression in tumor cell lines (Broad Cancer Cell Line Encyclopedia). (B) Positive correlation [correlation coefficient (R)=0.6] between CRBN mRNA expression in eight parental MM cell lines (2-D(Ct) x 105), and their sensitivity to ARV 825, n=3; P<0.001) (Left). Relative CRBN mRNA expression of isogenic lenalidomide resistant (MM1S res and KMS11 res) and isogenic wildtype (MM1S and KMS11) myeloma cells, and comparison of their sensitivity to increasing concentrations of ARV 825 [growth inhibition (MTT) of the cells]. (Right, top) MM1S and MM1S res cells and (right, bottom) shows KMS11 and KMS11 res cells, and their IC50s to ARV 825. (C and D) Levels of CRBN mRNA expression (left) and protein (right, top) from KMS11 and KMS28BM cells, and their viability in ARV 825 (right, bottom) after (C) overexpression and (D) shRNA silencing. OE: overexpression. *P≤0.01; **P≤0.001; ***P≤0.0001.

1212

haematologica | 2019; 104(6)

Proteolysis targeting chimeric molecules in MM

A

B

C

D

Figure 3. ARV 825 rapidly degrades BRD 2 and BRD 4 in KMS11 and KMS28BM multiple myeloma (MM) cells. (A) Immunoblotting of BRD 2, BRD 3 and BRD 4 in KMS11 and KMS28BM cells, cultured for different times with ARV 825: 10 nM (KMS11) and 100 nM (KMS28BM). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH): internal control. (B) Levels of BRD 4 and MYC mRNA expression (left, top and bottom) and protein expression (right) using wild-type KMS11, KMS28BM and their respective cells over-expressing CRBN, cultured with either 10 nM (KMS11) or 100 nM (KMS28BM) of ARV 825 for 4 h. GAPDH: internal control. (C) Protein expression of IKZF 1/3 (Ikaros/Aiolos), BRD 4 and MYC after treatment with ARV 825 (10 nM, KMS11; 100 nM, KMS28BM), MZ1 (100 nM, KMS11; 70 nM, KMS28BM), or combinations of the two PROTAC with either 10 ÂľM pomalidomide (POM) or 5 nM bortezomib (BORT) for 12 h and examined by immunoblotting (GAPDH: internal control). (D) Combination Index plot of ARV 825 with either bortezomib or pomalidomide on KMS11 cells.

haematologica | 2019; 104(6)

1213

S.L. Lim et al.

a panel of 13 human MM cell lines (KMS11, MM1R, KMS12BM, H929, KMS18, 8226 LR5, MM1S, KMS11 res, U266, 8226, KMS28BM, 8226 P100V, MM1S res) using an in vitro proliferation assay (MTT, 72 h). Cell lines included melphalan resistant (8226 LR5), steroid resistant (MM1R), bortezomib resistant (8226 P100V), and lenalidomide resistant (KMS11 res and MM1S res) cell lines. Their cytogenetics varied; some were associated with a poor prognosis [e.g. t(4:14): KMS11, KMS28BM, H929; t(14:16): MM1S, 8226]. ARV 825 was more potent than either OTX 015 or pomalidomide alone against both KMS11 (IC50 for ARV 825, OTX 015 and pomalidomide: 9 nM, 130 nM, >1000 nM, respectively) and KMS28BM cells (IC50 for ARV 825, OTX 015 and pomalidomide: 137 nM, 240 nM, >1000nM, respectively) (Figure 1B). All MM cell lines were sensitive to ARV 825 with an IC50 ranging from 8 nM to 500 nM except for MM1S res cells (>1000 nM) (Figure 1C). MM1S res cells (resistant to lenalidomide) had significantly reduced CRBN levels (Online Supplementary Figure S1A and B). The MM1S res cells had a 40-fold reduction in expression of CRBN compared to parental MM1S cell line due to deletion of one allele of the CRBN gene and a point mutation on the second allele.14 Therefore, because of lack of wild-type CRBN, they had loss of wild-type CRBN expression (western blot, Online Supplementary Figure S1B) associated with a resistance to ARV 825.

Likewise, KMS11 res cells (resistant to lenalidomide) also have a structural deletion of CRBN with reduced CRBN expression15 (Online Supplementary Figure S1A and B) and were 8.3-fold more resistant to ARV 825 compared to its parental cells. Overexpression of wild-type CRBN in MM1S res cells rescued this resistant cell line and increased its sensitivity to ARV 825 (IC50=800 nM) (Online Supplementary Figure S1C). KMS11 cells are the most sensitive, with an IC50 of 8.5 nM; in contrast, KMS28BM is a relatively more resistant cell line (IC50=137 nM) (Online Supplementary Table S1). ARV 825 decreased clonogenic growth of KMS11 and KMS28BM in a dose-dependent manner (Figure 1D).

The other PROTAC (MZ1) significantly suppressed growth of the lenalidomide resistant cells (MM1S res and KMS11 res) as well as some of the other MM cells.

However, KMS18, U266, 8226, 8226 LR5 and 8226 P100V were relatively resistant to MZ1 (Figure 1E). IC50s are shown in Online Supplementary Table S2.

The bortezomib-resistant cell line (8226 P100V) is also relatively resistant to ARV 825 (IC50=500 nM).

Consistently, the BRD 4 degradation in 8226 P100V cells after treatment with different doses of ARV 825 (50 nM, 100 nM, 200 nM) only modestly decreased when compared to its parental strain (8226). 8226 P100V has up-regulated mRNA expression of multidrug resistance-associated protein 1 (MRP-1, encoded by the ABCC1 gene) compared to its parental strain (8226) (Online Supplementary Figure S1D).

ARV 825 inhibit growth of primary myeloma cell samples The effect of ARV 825 (10 nM to 300 nM) on three MM patient samples was evaluated. The PROTAC significantly inhibited growth of these MM patient samples which include one multiple relapsed patient (Patient 3). 10 nM of ARV 825 inhibited the growth of MM Patients 1, 2 and 3 by 82%, 93%, and 64%, respectively (Figure 1F).

1214

A

B

Figure 4. ARV 825 induced cell cycle arrest and apoptosis of multiple myeloma (MM) cells. (A) Cell cycle: KMS11 and KMS28BM MM cells were treated for 48 hours (h) with either ARV 825 (2.5-10 nM or 25-50 nM, respectively, 48 h) or diluent control [dimethyl sulfoxide (DMSO)], stained with propidium iodide (PI) and analyzed by flow cytometry. Histograms showed proportion of cells in different phases of cell cycle. Representative of three independent experiments. (B) Apoptosis: KMS11 and KMS28BM cells treated with of ARV 825 (10-40 nM or 100-400 nM, respectively, for 48 h), stained with annexin V-FITC and PI, and analyzed by flow cytometry. Histograms represent percentage of apoptotic cells.

Mean±Standard Deviation of three independent experiments. *P≤0.01; **P≤0.001; ***P≤0.0001 for ARV 825 versus control.

haematologica | 2019; 104(6)

Proteolysis targeting chimeric molecules in MM

Levels of CRBN mRNA as a potential biomarker of sensitivity of multiple myeloma cells to inhibition

by ARV 825 CRBN expression tends to be higher in hematologic malignancies, including MM, compared to solid tumors (Figure 2A). Levels of CRBN mRNA expression across different MM cell lines positively correlated with sensitivity of the MM cells to ARV 825 (Figure 2B, left). Isogenic

lenalidomide resistant cells (MM1S res and KMS11 res cells) had reduced CRBN levels and were more resistant to ARV 825 compared to their parental cells (Figure 2B, right).

Over-expressed or silenced CRBN in both KMS11 and KMS28BM MM cells correlated with ability of ARV 825 to either enhance (Figure 2C) or to decrease (Figure 2D) the inhibition of cell growth, respectively.

A

B

C

D

Figure 5. Combination Index (CI) plot of ARV 825 with small molecule inhibitors. Synergistic growth inhibition (MTT assay) of KMS11 and KMS28BM multiple myeloma cells in the presence of ARV 825 and a small molecule inhibitor. CI defines interaction between ARV 825 and small molecule inhibitor as plotted against a fraction of cell viability. CI<1, CI=1, and CI>1 represent synergism, additive, and antagonism of the two compounds, respectively. Values of CI analysis are shown in Online Supplementary Table S3.

E

haematologica | 2019; 104(6)

1215

S.L. Lim et al.

ARV 825 degrades bromodomain extraterminal domain proteins We evaluated the ability of ARV 825 to degrade BRD 2, BRD 3 and BRD 4 proteins after KMS11 and KMS28BM cells were treated with 10 nM and 100 nM ARV 825, respectively, for either 2, 8, 24 or 48 h. In addition, after

treatment for 48 h, ARV 825 was washed out and BRD 2, BRD 3 and BRD 4 proteins were examined at 6 h, 24 h and 48 h post wash-out (Figure 3A). Lysates were analyzed by western blot. ARV 825 significantly degraded BRD 2 and BRD 4 by 2 h, but the effect on levels of BRD 3 protein was minimal. After wash-out in KMS11, BRD 2 returned

A

B

C

D

E

1216

Figure 6. Combination Index (CI) plot of MZ1 with small molecule inhibitors. Synergistic growth inhibition of KMS11 and KMS28BM multiple myeloma cells in the presence of MZ1 and small molecule inhibitors as measured by MTT assay. CI defines interaction between MZ1 and small molecule inhibitor as plotted against a fraction of cell viability.

CI<1, CI=1, and CI>1 represent synergism, additive, and antagonism of the two compounds, respectively. Values of CI analysis are shown in Online Supplementary Table S4.

haematologica | 2019; 104(6)

Proteolysis targeting chimeric molecules in MM

to baseline by 24 h and BRD 4 was still less than control at 48 h. For KMS28BM, levels returned to baseline for BRD 2 and BRD 4 by 6 and 24 h, respectively.

ARV 825 PROTAC: CRBN expression levels correlated with levels of degradation of bromodomains 4 and MYC We examined whether CRBN expression correlated with effect of ARV 825-mediated degradation of BRD 4 and MYC levels. The KMS11 and KMS28BM cells+over-

A

expressed CRBN were treated with 10 nM and 100 nM ARV 825, respectively, for 4 h. Levels of BRD 4 and MYC RNA and protein were measured. CRBN over-expressed KMS11 and KMS28BM cells markedly decreased their RNA and protein levels of BRD 4 and MYC compared to their wild-type cells after ARV 825 treatment (Figure 3B) (KMS11 cells showed a greater decrease in BRD 4 and MYC compared to KMS28BM cells exposed to ARV 825).

Figure 7. RNA sequencing to obtain profile of gene expression. (A) Heatmaps show top 20 down- and up-regulated genes upon ARV 825 treatment [20 nM ARV 825, 8 hours (h)] of KMS11 MM cells. Results of two replicates from each group are shown. (B) Level of mRNA of nine genes were validated by quantitative realt-time polymerase chain reaction. (These genes were chosen because their expression levels were altered as viewed by RNA-sequencing). Expression of each gene was normalized to b-actin as a reference (control value converted to the value of 1). *P≤0.01; **P≤0.001; ***P≤0.0001.

B

haematologica | 2019; 104(6)

1217

S.L. Lim et al.

Degradation of IKZF 1/3 after exposure to ARV 825 The effect of ARV 825 on IKZF 1/3 degradation was examined using KMS11 and KMS28BM cells [12 h ARV 825 (10 nM, KMS11; 100 nM, KMS28BM), MZ1 (100 nM, KMS11; 70 nM, KMS28BM), pomalidomide (10 μM), bortezomib (5 nM) and combination of PROTAC with either pomalidomide and bortezomib] ARV 825, but not MZ1, degraded IKZF 1/3 (Ikaros/Aiolos), although the degradation was not as significant as was pomalidomide (10 μM) alone. In contrast, BRD 4 was prominently degraded by both PROTAC (Figure 3C).

Bortezomib and pomalidomide antagonized the activity of ARV 825 Pomalidomide reversed the BRD 4 degradation induced

by ARV 825, but not by MZ1 (Figure 3C). Pomalidomide was antagonistic to ARV 825 (Figure 3D). Pomalidomide competed with ARV 825 for the binding to CRBN; but, as expected, MZ1 activity was not influenced by pomalidomide. In contrast, the proteasome inhibitor (bortezomib) antagonized the ability of PROTAC to degrade BRD 4 (Figure 3C and D), indicating the need for an intact proteasome function for PROTAC action. These findings are consistent with a previous study by Zhang et al.13

A

B

P=0.005

C

Figure 8. ARV 825 inhibits multiple myeloma (MM) samples in vivo. (A) Whole-body bioluminescence images of SCID-beige mice after intravenous injection with KMS11LUC cells (14 days) followed by treatment with ARV 825 (5 mg/kg intraperitoneally daily for 28 days) versus vehicle control. (B) Tumor burden as measured by bioluminescence in SCID-beige mice after intravenous injection with KMS11LUC cells (graphic display). Data represent Mean±Standard Deviation (n=9 per group). (C) Survival curves (Kaplan-Meier) of immunodeficient mice who received human MM. The mice who received ARV 825 statistically lived longer (log-rank test, P<0.001).

1218

haematologica | 2019; 104(6)

Proteolysis targeting chimeric molecules in MM

ARV 825 induces cell cycle arrest and apoptosis in multiple myeloma cells Cell cycle analysis of MM cells was performed in the presence of various concentrations of ARV 825 for 48 h compared to control cells. The drug increased the G1 phase and decreased the S and G2/M phases in MM cells (Figure 4A). Flow cytometric analysis of KMS11 and KMS28BM MM cells showed a marked dose-dependent increase in the percentage of apoptotic cells (apoptotic cells defined as Annexin V+ and PI+) after treatment with various concentrations of ARV 825 for 48 h (Figure 4B). For example, ARV 825 (10, 20 and 40 nM) led to 10%, 30% and 50% of apoptotic KMS11 cells, respectively. ARV 825 at 100, 200 and 400 nM produced 26%, 31% and 34%, respectively, of apoptotic KMS28BM cells.

Small molecule inhibitors which are synergistic with ARV 825 We performed high-throughput small-molecule inhibitor screen (panel of 170 drugs, FDA-approved or in clinical trial) to identify novel anti-MM compounds that may have synergistic activity with ARV 825. IC50s were determined for each compound, both alone and in combination with ARV 825. ARV 825 relatively sensitive KMS11 and relatively resistant KMS28BM MM cells were examined. Of the 170 tested drugs, 60 are shown in Online Supplementary Figure S3A. Combination of ARV 825 and cediranib (VEGFR inhibitor), crenolanib (PDGFRα/b and FLT3 inhibitor), GSK 1904529A (IGF-1R inhibitor), motesanib (VEGFR1/2/3 inhibitor), and LY3023414 (dual PI3K/mTOR inhibitor) produced synergistic growth inhibitory activity against both KMS11 and KMS28BM. Additional confirmation of synergistic effect of these five promising small molecules on 8226 cells was performed [Combination Index (CI) <1] (Figure 5). Selinexor (CRM1 inhibitor), gilteritinib (FLT3/AXL inhibitor), LY333531 (PKCb1 and 2 inhibitor), IGC003 (CBP/EP300 inhibitor), ruxolitinib (JAK inhibitor) produced synergistic growth inhibitory activity against either KMS11 or KMS28BM cells (Online Supplementary Figure S3B). The CI analysis of 10 of these synergistic small molecule inhibitors with ARV 825 is shown in Online Supplementary Table S3.

We also performed combination of MZ1 with promising small molecules (cediranib, Crenolanib, GSK 1904529A, motesanib and LY3023414). Each combination showed synergistic activity with MZ1 inhibiting growth of both KMS11 and KMS28BM MM cells. However, combination of MZ1 with LY3023414 has only synergistic effect against KMS11 but not KMS28BM cells (Figure 6).

The CI analysis of these promising small molecule inhibitors with MZ1 is shown in Online Supplementary Table S4. A list of inhibitors is provided in Online Supplementary Table S5.

Transcriptome analysis showed MYC is significantly down-regulated by ARV 825 in multiple myeloma cells We examined the effect of ARV 825 on mRNA expression of MM cells (KMS11) by RNA sequencing. Heatmaps (Figure 7A) displayed the top 20 down-regulated and upregulated transcripts of KMS11 MM cells following treatment of the cells with 20 nM ARV 825 for 8 h. ARV 825 markedly down-regulated CCR1, RGS1, MYB, and MYC. RNA sequencing data were further verified using quantitative RT-PCR for 9 selected genes (FJX1, ZNF8, SSTR3, CCR1, MYB, NRROS, MYC, RGS1 and DOK4) in KMS11 haematologica | 2019; 104(6)

(Figure 7B) and KMS28BM cells (Online Supplementary Figure S2A). Furthermore, Gene-Set Enrichment Analysis (GSEA) indicated the robust downregulation of functionally-defined MYC targets following ARV 825 treatment of KMS11 MM cells (Online Supplementary Figure S2B and C).

Primers for qRT-PCR are listed in Online Supplementary Table S6.

ARV 825 inhibited multiple myeloma growth in vivo Anti-proliferative effect of ARV 825 was examined in vivo against MM xenografts growing in SCID-Beige mice.

Two weeks after injection, the MM cells were easily observed by bioluminescence imaging, after which, mice (n=9 per group) were randomly assigned to receive either ARV 825 (5 mg/kg) dissolved in 200 μL of vehicle daily intraperitoneally or 200 μL of vehicle alone. ARV 825 significantly slowed tumor growth in experimental mice compared to control mice receiving vehicle, as measured by bioluminescence (Figure 8A and B) at days 7, 14, 21 and 28. Importantly, ARV 825 treatment significantly prolonged the murine OS compared to vehicle-treated mice (Figure 8C). ARV 825 treated mice maintained normal activity and insignificant weight loss compared to diluent control mice (Online Supplementary Figure S4A). The IC50 of ARV 825 using normal mouse BM cells (2x105 cells/well) is 500 nM (Online Supplementary Figure S4B).

Discussion Despite the major advances in the treatment of MM made over the last decade, disease management still remains challenging as most patients either do not achieve a complete remission or eventually relapse. Bortezomib and lenalidomide have become a part of standard management. Auto-transplants are often also given; nevertheless, patients are rarely cured. New targeted therapeutic strategies are needed. Next generation BET inhibitor ARV 825 degrades bromodomains. We found BRD 2 and BRD 4 were profoundly depleted, consistent with previous reports of PROTAC for other malignancies.7,9,10,16 Previous studies have shown a fusion of JQ1 and thalidomide (dBET6) has significant potency against MM.17 We also showed that ARV 825 leads to significant growth inhibition of myeloma cells in liquid culture, clonogenic assay and, most importantly, in a xenograft model. Flow cytometric analysis showed that ARV 825 induced apoptosis and G0/G1 cell cycle arrest of these cells in vitro.

We demonstrated that both ARV 825 and MZ1 have promising activity against MM cells. ARV 825 induced degradation of BET proteins via CRBN E3 ligase.

Importantly, we found a positive correlation of intracellular levels of CRBN and their sensitivity to ARV 825. CRBN expression is prominent in hematologic malignancies, including MM. Response to immunomodulatory drugs is clinically correlated with expression of CRBN.18 Loss of function of CRBN causes resistance to dBET6 by perturbing dBET-mediated BRD 4 degradation.17 We postulate that levels of CRBN will serve as a predictive biomarker for cellular responsiveness to ARV 825. Indeed, two pairs of isogenic cells, one of each pair resistant to lenalidomide (KMS11 res and MM1S res), had very low expression of CRBN. Genetically either silencing or over-expressing CRBN decreased and increased, respectively, the sensitivity of an MM cell line to growth inhibition by ARV 825. In 1219

S.L. Lim et al.

stark contrast, their sensitivity to MZ1, which uses VHL E3 ligase, remained unchanged after either forced expression or silencing CRBN. Taken together, data suggest that ARV 825 may be most potent when this PROTAC is given to patients whose MM cells express CRBN. The sensitivity of MM patients to ARV 825 increased as the CRBN expression increased. In contrast, MZ1 could be a promising therapeutic drug for lenalidomide/pomalidomideresistant MM. Our ARV 825 data are consistent with a recent study which demonstrated that pomalidomide competed with ARV 825 for binding to CRBN.13 The authors further showed that ARV 825 relied on an intact proteasome pathway with proteasome inhibitors (carfilzomib or bortezomib) antagonizing the effects of ARV 825.13 This is consistent with our findings.

Based on high throughput screening of small-molecule inhibitors, we identified novel compounds that have synergistic activity with ARV 825 against MM cell lines (KMS11 and KMS28BM), including those against either dual PI3K/mTOR, VEGFR, PDGFRÎą/b, and FLT3 and IGF1R. Previous studies showed the inhibition of PI3K and BET blocked reactivation of PI3K signaling in diverse cancer models.19 Interestingly, a previous study reported that co-treatment of ARV 825 with ruxolitinib synergistically inhibited growth of secondary acute myeloid leukemia.9 We also observed synergism of this combination against KMS28BM MM cells, but not KMS11 MM cells (Online Supplementary Figure S2B).

References 1. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2018. CA Cancer J Clin. 2018; 68(1):7-30.

2. Cohen AD, Garfall AL, Stadtmauer EA, et al. B-Cell Maturation Antigen (BCMA)Specific Chimeric Antigen Receptor T Cells (CART-BCMA) for Multiple Myeloma (MM): Initial Safety and Efficacy from a Phase I Study. Blood. 2016;128(22):1147.

3. Shu S, Polyak K. BET Bromodomain Proteins as Cancer Therapeutic Targets.

Cold Spring Harb Symp Quant Biol.

2016;81:123-129.

4. Mochizuki K, Nishiyama A, Jang MK, et al.

The Bromodomain Protein Brd4 Stimulates G 1 Gene Transcription and Promotes Progression to S Phase. J Biol Chem.

2008;283(14):9040-9048.

5. LovĂŠn J, Hoke HA, Lin CY, et al. Selective inhibition of tumor oncogenes by disruption of super-enhancers. Cell. 2013; 153(2):320-334.

6. Delmore JE, Issa GC, Lemieux ME, et al.

BET bromodomain inhibition as a therapeutic strategy to target c-Myc. Cell.

2011;146(6):904-917.

7. Lu J, Qian Y, Altieri M, et al. Hijacking the E3 Ubiquitin Ligase Cereblon to Efficiently Target BRD4. Chem Biol. 2015;22(6):755763.

8. Buckley DL, Crews CM. Small-Molecule

1220

9.

10.

11.

12.

13.

14.

RNA sequencing and gene set enrichment analysis demonstrated that MYC expression is significantly downregulated after treatment with ARV 825 (8 h). MYC is an attractive target in MM due to its role in disease progression. In addition, CCR1 and LGR5 were down-regulated.

CCR1 has been reported to play a central role in pathogenesis of MM as well as MM-induced osteolytic bone disease,20 whereas LGR5 has been identified as a marker of early stem cells in the intestine.21 In summary, our studies showed that MM cells are sensitive to ARV 825 and a combination of ARV 825 with synergistic small molecule inhibitors may be therapeutically effective for patients. During the final preparation of our manuscript, another manuscript was published reporting ARV 825 in MM.13 Taken together, these two studies nicely compliment each other and provide the foundation for further pre-clinical studies of both ARV 825 and MZ1 for the treatment of MM. Acknowledgments We thank Aaron Eshman and Melmed family for their invaluable support for this myeloma project. We also express gratitude to Morgan Stanley Inc. Funding This work was also supported by a grant from the Leukemia and Lymphoma Society.

Control of Intracellular Protein Levels through Modulation of the Ubiquitin Proteasome System. Angew Chemie Int Ed. 2014;53(9):2312-2330.

Saenz DT, Fiskus W, Qian Y, et al. Novel BET protein proteolysis-Targeting chimera exerts superior lethal activity than bromodomain inhibitor (BETi) against post-myeloproliferative neoplasm secondary (s) AML cells. Leukemia. 2017;31(9):1951-1961.

Sun B, Fiskus W, Qian Y, et al. BET protein proteolysis targeting chimera (PROTAC) exerts potent lethal activity against mantle cell lymphoma cells. Leukemia. 2018; 32(2):343-352.

Raina K, Lu J, Qian Y, et al. PROTACinduced BET protein degradation as a therapy for castration-resistant prostate cancer.

Proc Natl Acad Sci U S A. 2016; 113(26):7124-7129.

Winter GE, Buckley DL, Paulk J, et al.

DRUG DEVELOPMENT. Phthalimide conjugation as a strategy for in vivo target protein degradation. Science. 2015;348 (6241):1376-1381.

Zhang X, Lee HC, Shirazi F, et al. Protein targeting chimeric molecules specific for bromodomain and extra-terminal motif family proteins are active against pre-clinical models of multiple myeloma.

Leukemia. 2018;32(10):2224-2239.

Zhu YX, Braggio E, Shi C-X, et al. Cereblon expression is required for the antimyeloma activity of lenalidomide and pomalido-

mide. Blood. 2011;118(18):4771-4779.

15. Zhu YX, Shi C-X, Bruins LA, et al.

Lenalidomide, Pomalidomide and CC-220 resistant multiple myeloma: confirmation of role of CRBN/IRF4 axis and identification of multiple disruption mechanisms.

Blood. 2017;130(Suppl 1):4369.

16. Piya S, Bhattacharya S, Mu H, et al. BRD4 Proteolysis targeting chimera (PROTAC) ARV-825, causes sustained degradation of BRD4 and modulation of chemokine receptors, cell adhesion and metabolic targets in leukemia resulting in profound antileukemic effects. Blood. 2016; 128(22):748.

17. Matthews GM, Hu Y, Sheffer M, et al.

Abstract 4713: BET bromodomain degradation as a therapeutic strategy in drug-resistant multiple myeloma. Cancer Res. 2016; 76(14):4713.

18. Schuster SR, Kortuem KM, Zhu YX, et al.

The clinical significance of cereblon expression in multiple myeloma. Leuk Res. 2014; 38(1):23-28.

19. Stratikopoulos EE, Dendy M, Szabolcs M, et al. Kinase and BET Inhibitors Together Clamp Inhibition of PI3K Signaling and Overcome Resistance to Therapy. Cancer Cell. 2015;27(6):837-851.

20. Vallet S, Anderson KC. CCR1 as a target for multiple myeloma. Expert Opin Ther Targets. 2011;15(9):1037-1047.

21. Clevers H. Searching for adult stem cells in the intestine. EMBO Mol Med. 2009; 1(5):255-259.

haematologica | 2019; 104(6)

ARTICLE

Stem Cell Transplantation

Total body irradiation dose escalation decreases risk of progression and graft rejection after hematopoietic cell transplantation for myelodysplastic syndromes or myeloproliferative neoplasms Federico Monaco,1 Bart L. Scott,1,2 Thomas R. Chauncey,1,2,3 Finn B. Petersen,4 Barry E. Storer,1,2 Frederic Baron,5 Mary E. Flowers,1,2 H. Joachim Deeg,1,2 David G. Maloney,1,2 Rainer Storb1,2 and Brenda M. Sandmaier1,2

Fred Hutchinson Cancer Research Center, Seattle, WA, USA; University of Washington, Seattle, WA, USA; 3VA Puget Sound Health Care System, Seattle, WA, USA; 4LDS Hospital, Salt Lake City, UT, USA and 5University of Liege, Belgium 1

2

Ferrata Storti Foundation

Haematologica 2019 Volume 104(6):1221-1229

ABSTRACT

A

non-myeloablative regimen of fludarabine and 200 cGy total body irradiation combined with post-grafting immunosuppression with mycophenolate mofetil and a calcineurin inhibitor facilitates allogeneic hematopoietic cell transplantation from HLA-matched related or unrelated donors in older patients and/or those with comorbidities.

However, outcomes of prior studies have been disappointing in patients with myelodysplastic syndromes or myeloproliferative neoplasms due to high incidences of progression or graft failure (together termed hematopoietic cell transplantation-failure). We hypothesized that escalating the total body irradiation dose may improve the outcomes and subsequently performed a phase II total body irradiation dose-escalation trial. Patients with median age 66 years were enrolled in two arms to receive non-myeloablative conditioning followed by hematopoietic cell transplantation with total body irradiation dose escalation for excessive hematopoietic cell transplantation-failure: Arm A: myeloproliferative neoplasm/myelodysplastic syndrome low risk (n=36); and Arm B: myelodysplastic syndrome highrisk/chronic myelomonocytic leukemia (n=41). Total body irradiation dose levels were: Level-1 (300 cGy), Level-2 (400 cGy), or Level-3 (450 cGy).

Patients received intravenous fludarabine 30 mg/m2 for three days. Total body irradiation was administered on day 0 followed by infusion of peripheral blood stem cells from HLA-matched related (n=30) or unrelated (n=47) donors. Post-grafting immunosuppression with mycophenolate mofetil and cyclosporine was administered. The primary end point was day 200 hematopoietic cell transplant failure, with the objective of reducing the incidence to <20%. The primary end point was reached on Arm A at dose Level-1 (300 cGy total body irradiation) with a cumulative incidence of day 200 hematopoietic cell transplant failure of 11%, and on Arm B at dose Level-3 (450 cGy) with a cumulative incidence of day 200 hematopoietic cell transplant failure of 9%. Increasing the total body irradiation dose leads to a higher success rate with non-myeloablative conditioning by reducing relapse and rejection. Further studies are necessary to decrease non-relapse mortality, especially among patients with high-risk disease. Trial registered under clinicaltrials.gov identifier: NCT00397813.

Introduction Hematopoietic cell transplantation (HCT) is the only curative option for patients with myelodysplastic syndromes (MDS) or myeloproliferative neoplasms (MPN).1 While these disorders are usually seen in older individuals, conventional HCT after haematologica | 2019; 104(6)

Correspondence: BRENDA M. SANDMAIER bsandmai@fredhutch.org Received: June 7, 2018.

Accepted: January 2, 2019.

Pre-published: January 10, 2019. doi:10.3324/haematol.2018.199398 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/104/6/1221 Š2019 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

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

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

1221

F. Monaco et al.

intensive cytotoxic conditioning regimen has been restricted to relatively young patients, without comorbidities. In order to address these limitations, we designed a non-myeloablative (NMA) regimen specifically for older patients and in those with comorbidities. This regimen consists of fludarabine (FLU) and 200 cGy total body irradiation (TBI) combined with post-grafting immunosuppression with mycophenolate mofetil (MMF) and a calcineurin inhibitor that facilitates allogeneic HCT with HLA-matched related (MRD) or unrelated donors (URD).

Results of prior studies with this regimen have been very encouraging in most hematologic malignancies;2-5 however, in patients with chronic myelomonocytic leukemia (CMML) or previously untreated MDS/MPN, results have not been as successful. In a previous study, the cumulative incidence of HCT failure, defined as graft rejection or disease progression before day 200 after HCT, was 46% for the entire study population, 58% for patients with CMML or MDS with excess blasts (MDS-EB), and 31% for patients with MDS [comprising MDS with single or multilineage dysplasia (MDS-SLD or MDS-MLD), MDS with ring sideroblasts (MDS-RS), MDS, unclassifiable (MDS-U) MDS with <5% blasts], or MPN. The 200-day, 1-year and 2-year probabilities of progression-free survival (PFS) were 27%, 19%, and 11%, respectively, in patients with CMML or MDS-EB, and 46%, 40%, and 37%, respectively, in patients with MDS with <5% blasts or MPN. Furthermore, it was shown that early achievement of full donor T-cell chimerism was associated with a reduced risk of relapse (HR 0.5, P=0.002) in patients with hematologic malignancies receiving NMA conditioning.6 Taken together, these data suggest that increasing the intensity of the non-myeloablative conditioning regimen might delay disease progression until establishment of full donor chimerism and development of graft-versus-tumor (GvT) effects, thereby improving outcomes of patients with CMML, untreated MDS, and MPN.

Methods Study design A phase II TBI dose escalation trial (Figure 1) was conducted at Fred Hutchinson Cancer Research Center, Veterans Affairs Puget-Sound Health Care System, and Intermountain LDS Hospital. The Institutional Review Boards of the participating centers approved the protocol; all patients signed informed consent. An independent Data and Safety Monitoring Board oversaw the trial. The trial was registered at ClinicalTrials.gov, a service of the United States National Institutes of Health under number NCT00397813.

Patients Participants were eligible if they had a World Health Organization defined diagnosis of CMML, MDS, MPN or paroxysmal nocturnal hemoglobinuria (PNH) with history of life-threatening complications.

Patients were â&#x2030;Ľ50 and <75 years of age or were <50 years of age but at high risk for therapy-related toxicity from standard high-dose regimens; factors considered high risk included pre-existing conditions such as a chronic disease affecting kidneys, liver, lungs or heart, or previous failed HCT. Included patients were affected by MDS or MPN with less than 10% of blasts and not treated by 1222

myelosuppressive chemotherapy (defined as chemotherapy given with the intent of inducing a complete remission; hypomethylating agents and oral cytoreductive therapy were permitted). Patients with CMML who had progressed and received myelosuppressive chemotherapy could be enrolled if they had less than 5% of marrow blasts.

Donors were either HLA-MRD or URD; unrelated donors with only a single allele disparity for HLA-A, B, or C as defined by high-resolution typing were allowed.

Only granulocyte-colony stimulating factor (G-CSF) mobilized peripheral blood stem cells (PBSC) were permitted as a HCT source.

Conditioning regimens and immune suppression The conditioning regimens (Figure 2) were FLU (30 mg/m2/day IV) on days -4, -3, -2 and TBI of three magnitudes: 300 cGy (Level 1), 400 cGy (Level 2), or 450 cGy (Level 3) on day 0 at 6-7 cGy/min from linear accelerator, followed by HCT. For related recipients, graft-versus-host disease (GvHD) prophylaxis included cyclosporine (CSP) 5 mg/kg PO every 12 hours from day -3 to day +56 with subsequent taper to day +180 and mycophenolate mofetil (MMF) at 15 mg/kg PO every 12 hours from day 0 to day +27 and then discontinued. For unrelated recipients, GvHD prophylaxis included CSP 5 mg/kg PO every 12 hours from day -3 to day +100 with taper to day +180 and MMF at 15 mg/kg PO every eight hours from day 0 to day +40 and then tapering to day +96. Growth factors were not given during the first 21 days after HCT; infection prophylaxis and transfusion support were as per institutional guidelines. Bone marrow aspirates or biopsies and chimerism analyses were performed on days +28, +56, +84 and +180 after HCT.

Outcomes The purpose of this study was to evaluate whether a more intense but still NMA conditioning regimen could reduce the combined rates of graft rejection and disease progression (together termed HCT-failure) in this group of MDS and CMML patients, while maintaining an acceptable rate of non-relapse mortality (NRM).

The primary objective was to decrease the incidence of day-200 HCT-failure to <20%. HCT-failure in this study was defined as either progression of the underlying malignancy or graft failure. Progression was defined as recurrence of disease as detected by flow cytometry, morphology, cytogenetics, or mutation analysis. For patients with myelofibrosis, persistent fibrosis was not considered evidence of progression. Graft failure was defined as <5% donor CD3 chimerism in the peripheral blood. Secondary objectives included PFS, NRM, and the kinetics of donor engraftment.

Primary cause of death was adjudicated using previously described criteria.7 When relapse occurred, it was considered the primary cause of death regardless of other events. Additionally, acute and chronic GvHD incidences were also evaluated. Grading of acute and chronic GvHD was performed according to the established recognized standards.8 The first day of acute GvHD of a certain grade was used to calculate cumulative incidence curves for that GvHD grade.

Statistical analysis For purposes of this study, HCT-failure was defined as haematologica | 2019; 104(6)

Higher TBI dose improves NMA conditioning HCT for MDS or MPN

graft rejection (defined as <5% donor T-cell chimerism) or disease progression within 200 days of transplant.

Dose escalation was carried out independently in two groups of patients: • Arm A: patients with MPN, low-risk MDS (MDS-SLD, MDS-MLD, MDS-U, MDS-RS) or PNH; • Arm B: patients with high-risk MDS (MDS-EB-1) or CMML. In each arm, up to 24 patients were accrued to each TBI dose level, in groups of 6 patients, with an escalation rule triggered for excessive HCT failure. If 24 patients were successfully enrolled at a TBI dose level without triggering the escalation rule for HCT failure (or other stopping rules), then that dose level would be considered a success and accrual would be closed for that arm. ARM A was modified after completing accrual of 24 patients at a dose level, to allow accrual of up to 12 additional patients at the same dose level while accrual to Arm B continued.

Stopping rules for non-relapse mortality and TBI dose escalation continued to be monitored. Cohorts of 6 patients were only defined for purposes of determining dose escalation and stopping rules; the statistical analysis was based on the operating characteristics of a sequence

of cohorts of 6 patients, not on single cohorts.

All patients received FLU 30 mg/m2/day IV for three days preceding TBI. The TBI dose levels were: 1. Dose Level 1: 300 cGy TBI 2. Dose Level 2: 400 cGy TBI 3. Dose Level 3: 450 cGy TBI Dose escalation rules were imposed for HCT failure >20% at day +200. Stopping rules were imposed for NRM at day 200 of >25% in Arm A and >35% in Arm B. Overall survival (OS) and PFS were estimated by the Kaplan-Meier method. Cumulative incidences of relapse, NRM, and acute and chronic GvHD were estimated as previously described.9 All analyses were performed using SAS (SAS Institute, Cary, NC, USA).

Results Enrollment The study was opened on March 2006 and was closed to accrual on March 2017. Overall, 79 patients were screened, and 77 patients were enrolled. Patients’ and

Figure 1. Flow diagram of the progress through the phases of the study. MDS: myelodysplastic syndrome; MPN: myeloproliferative neoplasm; PNH: paroxysmal nocturnal hemoglobinuria; CMML: chronic myelomonocytic leukemia; TBI: total body irradiation; cGy: centigray; n: number.

haematologica | 2019; 104(6)

1223

F. Monaco et al. neutrophil >0.5x109/L was 13.5 days in Arm A and 13 days in Arm B. The platelet counts did not decline below 20x109/L in 9 out of 36 (25%) patients in Arm A and in 11 out of 41 (26.8%) patients in Arm B. The cumulative incidences of platelet recovery (>20x109/L) at day 84 were 97.2% in Arm A and 100% in Arm B. The median times to platelet recovery were 3.5 days in Arm A and 7 days in Arm B. There were two graft failures, both in patients enrolled in Arm A.

transplant characteristics are summarized in Table 1.

Thirty-six patients were enrolled in Arm A and 41 in Arm B.

Engraftment Neutrophils declined below 0.5x109/L in 35 out of 36 patients in Arm A and in all patients in Arm B. The cumulative incidences of neutrophil recovery at day 28 were 100% in Arm A and 97.4% in Arm B. The median time to Table 1. Patients’ characteristics.

Sex, n (%) Female Male Diagnosis, n (%) Myelofibrosis Low-risk MDS (MDS-SLD, MDS-MLD, MDS-U, MDS-RS) PNH High-risk MDS (MDS-EB-1) CMML Treatment, n (%) Dose level 1 (TBI 300 cGy) Dose level 2 (TBI 400 cGy) Dose level 3 (TBI 450 cGy) Time from diagnosis to HCT, month, median (range) Age at transplant, years, median (range) HCT-CI, n (%) 0 1–2 ≥3 Previous stem-cell transplantation, n Autologous Allogeneic Donor, n (%) MRD URD 10/10 URD 9/10 Neutrophils <0.5×109/L, days, median (range) Platelets <20×109L, days, median (range) HCT-failure before day +200, n Relapse Graft failure Relapse, n Death, n Cause of death, n Progression/relapse NRM non-GvHD related

NRM GvHD related

Follow up, months, median (range)

ARM A (n = 36)

ARM B (n = 41)

9 (25%) 27 (75%)

14 (34.1%) 27 (65.9%)

18 (50%) 17 (47.2%) 1 (2.8%) 28 (68.3%) 13 (31.7%) 39 (100%)

15.4 (4–336) 65 (37–71) Median 3 (range 0–8) 4 (11.1%) 11 (30.6%) 21 (58.3%)

12 (39.3%) 5 (12.2%) 24 (58.5%) 8.4 (1–92) 67 (52–74) Median 3 (range 0–9) 7 (17.1%) 12 (29.3%) 22 (53.6%)

0 0

2 2

16 (44.4%) 18 (50%) 2 (5.6%) 13.5 (0–28) 3.5 (0–107)

14 (34.1%) 25 (61%) 2 (4.9%) 13 (2–29) 7 (0–43)

3 2 9 24

10

9 6 Pneumonia n = 3 Sepsis n = 1 Disseminated aspergillus n = 1 Suicide n = 1 9 GvHD refractory n = 4 Pneumonia n = 2 Multiorgan failure n = 1 CNS bleeding n = 1 Sepsis n = 1 69.1 (39.1–121.8)

15 31 15 6 Pneumonia n = 2 Cardiac toxicity n = 2 Multiorgan failure n = 2 10 GvHD refractory n = 3 Pneumonia n = 2 Sepsis n = 4 Disseminated aspergillus n = 1 21 (3.4–102.6)

CMML: chronic myelomonocytic leukemia; CNS: central nervous system; GvHD: graft-versus-host disease; HCT-CI: hematopoietic cell transplantation comorbidity index; MDS: myelodysplastic syndrome; MDS-EB myelodysplastic syndrome with excess of blasts-1; MDS-RS: myelodysplastic syndrome with ring sideroblast; MDS-U: myelodysplastic syndrome unclassifiable; MLD: multilineage dysplasia; MRD: HLA-matched related donor; NRM: non-relapse mortality; PNH: paroxysmal nocturnal hemoglobinuria; SLD: single lineage dysplasia; TBI: total body irradiation; URD: HLA-matched unrelated donor.

1224

haematologica | 2019; 104(6)

Higher TBI dose improves NMA conditioning HCT for MDS or MPN

Chimerism analysis In Arm A, the day 28 median donor chimerism levels were 68% for CD3 and 100% for CD33, and the day 84 levels were 79% and 100%, respectively. In Arm B, for patients who received TBI at dose Levels-1 or 2 (300 cGy or 400 cGy), the day 28 median donor chimerism levels were 84% for CD3 and 77% for CD33, and the day 84 levels were 85% and 100%, respectively. For patients who received TBI at dose Level-3 (450 cGy) in Arm B, the day 28 median donor chimerism levels were 82% for CD3 and 100% for CD33, and the day 84 levels were 81% and 100%, respectively.

Hematopoietic cell transplantation failure and relapse In Arm A, 4 out of 36 patients given TBI at 300 cGy had HCT-failure before day 200, three due to progression and one to graft rejection, with a cumulative incidence of 11%.

The escalation rule in this arm was not reached so all patients received dose Level-1 with TBI at 300 cGy. In Arm B, 12 patients received dose Level-1 with TBI at 300 cGy and 5 had HCT-failure, all due to progression, triggering the TBI dose escalation rule. The next five patients received dose Level-2 TBI (400 cGy) and three experienced HCT-failure, all due to progression, triggering again the dose escalation rule. The cumulative incidence of HCT-failure for dose Level-1 and dose Level-2 combined was 47%. The subsequent 24 patients received dose Level-3 with TBI at 450 cGy and only two experienced HCT-failure before 200 days, with a cumulative incidence of 9%. The patientsâ&#x20AC;&#x2122; characteristics and outcomes for the high-risk patients enrolled on Arm B (separated by TBI dose) are shown in Online Supplementary Table S1 (Table 2 and Figure 3A and E).

The cumulative incidences of progression in Arm A at one year, two years, and five years was 17%, 19%, and 22%, respectively. Patients who received TBI at dose Levels-1 or 2 (300 cGy or 400 cGy) had cumulative incidences of progression at one year, two years, and five years of 53%, 53% and 53%, respectively. Patients who received TBI at dose Level-3 (450 cGy) in Arm B had cumulative incidences of progression at one year, two years, and five years of 24%, 24% and 32%, respectively.

Graft-versus-host disease The cumulative incidences of grades II-IV and grade III-IV acute GvHD at day 100 were 58% and 13%, respectively. The cumulative incidences of chronic GvHD at one year and two years were 38% and 45%, respectively: one, 21, and eight patients developed mild, moderate, and severe chronic GvHD, respectively. No statistically significant differences were seen between each arm with regards to incidence of acute or chronic GvHD (cumulative incidences in each arm are shown in Table 2) (Figure 4).

Non-relapse mortality The cumulative incidences of NRM in Arm A at one year, two years, and five years were 17%, 22%, and 43%, respectively. Patients who received TBI at dose Levels-1 or 2 (300 cGy or 400 cGy) had cumulative incidences of NRM at one year, two years, and five years of 29%, 35%, and 35%, respectively. Patients in Arm B who received TBI at dose Level-3 (450 cGy) had cumulative incidences of NRM at one year, two years, and five years of 31%, 43%, and 51%, respectively (Table 2 and Figure 3D).

Figure 2. Outline of treatment plan involving conditioning regimens and graft-versus-host disease (GvHD) prophylaxis. FLU: fludarabine; TBI: total body irradiation; HCT: hematopoietic cell transplantation; MRD: HLA-matched related donor; URD: HLA-matched unrelated donor; MMF: mycophenolate mofetil; CSP: cyclosporine; cGy: centigray; hrs: hours.

haematologica | 2019; 104(6)

1225

F. Monaco et al.

Causes of death Twenty-four of 36 patients in Arm A died; GvHD and disease progression were the primary causes of death for nine patients each. In Arm B, 31 out of 41 patients died; progression was the primary cause of death for 15 patients (Table 1).

Overall survival and progression-free survival In Arm A, OS at one year, two years, and five years were 72%, 61%, and 38%, respectively, and PFS were 68%, 58%, and 35%, respectively. Patients who received TBI at dose Levels 1 or 2 (300 cGy or 400 cGy) in Arm B had OS at one year, two years, and five years of 24%, 12%, and 12%, respectively, and PFS of 18%, 12%, and 12%, respectively. Patients in Arm B who received TBI at dose Level-3 (450 cGy) had OS at one year, two years, and five years of 60%, 37%, and 22%, respectively, and PFS of 45%, 33%, and 17%, respectively (Table 2 and Figure 3B and C).

Discussion In the early 2000s, at Fred Hutchinson Cancer Research Center, a regimen for NMA conditioning using the combination of fludarabine and 200 cGy TBI was developed.

This regimen allowed for donor cell engraftment but was characterized by a high incidence of early disease progression or graft rejection,2-5 especially for high-risk patients. In 2006, Scott et al.10 published results on 38 patients affected

by MDS or acute myeloid leukemia (AML) who received

an NMA regimen with 3-year rates of OS, PFS, and NRM of 27%, 28%, and 41%, respectively. These results were retrospectively compared with patients who underwent myeloablative conditioning, and no statistically significant differences were seen. Disease progression was the main issue with a cumulative incidence of 31% at three years.

Laport et al.11 reported data from our retrospective study on 148 patients diagnosed with MDS or MPN who underwent allogeneic HCT after NMA conditioning. PFS and OS at three years were both 27% for all patients, with a progression incidence of 41%. When stratifying data, PFS rates at three years for patients with de novo MDS (n=40), treatment-related MDS (n=25), MPN (n=27), and CMML (n=7) were 22%, 29%, 37%, and 43%, respectively, and OS rates were 20%, 27%, 43% and 43%, respectively. For all patients, NRM at three years was 32%. In both analyses, disease progression was the major cause of treatment failure and the leading cause of death; it appears that this regimen may not have conferred enough cytoreduction for adequate disease control. The data from these retrospective reviews formed the basis for the current study.

Cliff et al.12 had previously shown that an escalation in TBI dose to 15.75 Gy from 12 Gy reduced relapse/progression (13% vs. 35%). This result demonstrated that a decrease in tumor burden before transplantation led to a decreased risk of disease progression; however, most patients with MDS or MPN are not candidates for myeloablative TBI conditioning regimens either due to advanced age or comorbidities.

Taking into consideration these data, along with results from pre-clinical studies,13,14 we hypothesized that raising

Table 2. Results.

% at 100 days Overall survival Arm A TBI dose Level-1(300 cGy) Arm B TBI dose Level-1 or 2 (300–400 cGy) Arm B TBI dose Level-3 (450 cGy) Progression-free survival Arm A TBI dose Level-1(300 cGy) Arm B TBI dose Level-1 or 2 (300–400 cGy) Arm B TBI dose Level-3 (450 cGy) Relapse incidence Arm A TBI dose Level-1(300 cGy) Arm B TBI dose Level-1 or 2 (300–400 cGy) Arm B TBI dose Level-3 (450cGy) Non-relapse mortality Arm A TBI dose Level-1(300 cGy) Arm B TBI dose Level-1 or 2 (300–400 cGy) Arm B TBI dose Level-3 (450 cGy) Chronic GvHD Arm A TBI dose Level-1(300 cGy) Arm B TBI dose Level-1 or 2 (300–400 cGy) Arm B TBI dose Level-3 (450 cGy) Acute GvHD (grades 2-4) Arm A TBI dose Level-1(300 cGy) Arm B TBI dose Level-1 or 2 (300–400 cGy) Arm B TBI dose Level-3 (450 cGy) Acute GvHD (grades 3-4) Arm A TBI dose Level-1(300 cGy) Arm B TBI dose Level-1 or 2 (300–400 cGy) Arm B TBI dose Level-3 (450 cGy)

% at 1 year

% at 2 years

% at 5 years

72 24 60

61 12 37

38 12 22

68 18 45

58 12 33

35 12 17

17 53 24

19 53 24

22 53 32

17 29 31 38

22 35 43 45 53 35 40

43 35 51

58 56 65 57 13 17 12 9

TBI: total body irradiation; GvHD: graft-versus-host disease.

1226

haematologica | 2019; 104(6)

Higher TBI dose improves NMA conditioning HCT for MDS or MPN

the TBI dose from the originally used 200 cGy could improve the outcome of HCT in the NMA setting, thereby reducing the risk of disease progression in patients who are not considered candidates for myeloablative conditioning.

In agreement with this hypothesis, we showed improvement in outcome by raising the TBI dose from 200 cGy to 300 cGy in our cohort of low-risk patients (MDS-SLD, MDS-MLD, MDS-RS, MDS-U and MPN).

Similarly, the day 200 failure rate for high-risk patients (MDS-EB-1/CMML) was reduced to 9% when TBI was raised to 450 cGy while the lower TBI doses of 300 cGy

and 400 cGy failed to improve results. We achieved the primary objective of reducing transplant failure to less than 20% by day 200, which resulted in PFS at two years of 58% in the low-risk group and 33% in the high-risk group. NRM was not significantly increased after escalating TBI to 300 cGy, but 450 cGy TBI led to toxicity with a resultant NRM of 43% at two years after HCT. Limitations of the current study include the slow accrual over ten years. While donor selection and GvHD prophylaxis remained unchanged during this interval, and there were no major changes in antimicrobial treatments, radiological techniques and ancillary therapies, incremental

A

B

C

D

E

Figure 3. Patients enrolled in Arm A receiving 300 cGy total body irradiation (TBI) and patients enrolled in Arm B receiving 450 cGy TBI had reduced day 200 hematopoietic cell transplantation (HCT) failure. Transplant outcomes by Arm and TBI dose: (A) HCT-failure, (B) overall survival, (C) progression-free survival (PFS), (D) non-relapse mortality (NRM), and (E) relapse incidence.

haematologica | 2019; 104(6)

1227

F. Monaco et al.

improvements could be a bias. The slow accrual was due to competing protocols at our Centers. Another limitation regards high-risk patients who received dose level-2 TBI (400 cGy). Only five patients were enrolled but, as per protocol design, we met dose escalation rules because of transplant failure. It is possible that, with greater numbers, better results could be achieved.

The incidence of non-relapse mortality remains an issue, being high especially for high-risk patients (cumulative incidence of 43% at 2 years). There was no statistically significant difference regarding TBI-dose and GvHD incidence. Different strategies are needed to reduce the toxicity of 450 cGy TBI. Fractionation of the TBI dose was investigated in two preclinical canine experiments, which showed that fractionation of 450cGy TBI was significantly less immunosuppressive with a higher rate of graft rejection than when given as a single dose.15,16 Stelljes et al. demonstrated that a conditioning regimen of 8Gy fractionated TBI and FLU was feasible with low NRM in patients with AML (8% at 2 years for patients in CR, 37% at 2 years for patients with active disease).17 A novel approach to replace or augment TBI while reducing toxicity, consists of using targeted radioimmunotherapy.18 Monoclonal antibodies to CD45 and CD20 coupled to beta (b)-emitting radionuclides (iodine-131 or yttrium-90) have been studied19,20 showing good results but evident disadvantages (e.g. off-target effects and the need to use isolation rooms) in clinical trials. However, the recent substitution with alpha (Îą)-emitting radionuclides (bismuth213 and astatine-211) has given encouraging results,21,22 especially for astatine-211.23 Clinical trials are currently in progress using astatine-211 (clinicaltrials.gov identifier: NCT03128034).

The high incidence of GvHD irrespective of TBI dose remains a problem, especially for grades III-IV acute GvHD and chronic GvHD that reached cumulative incidences of 13% at 200 days and 45% at two years. These complications impacted OS and NRM, accounting for one-third of all deaths. Novel approaches are currently being investigated that are aimed at reducing GvHD incidence: the triple-drug combination using tacrolimus, mycophenolate mofetil, and sirolimus has been studied with encouraging results,24 and a phase III trial evaluated the addition of sirolimus to cyclosporine and MMF in order to reduce GvHD without impairing graft-versus-tumor effect with significantly less acute GvHD and less NRM with superior overall and progression-free survival (clinicaltrials.gov identifier: NCT01251575).25 Lim et al.26 recently described results from a large retrospective study by the European Group for Blood and Marrow Transplantation (EBMT) registry which included 1,333 patients with MDS, who were older than 50 years, of whom 833 (62%) underwent reduced intensity conditioning (RIC). The RIC regimens included FLU plus intermediate doses of one or two alkylating agents including busulfan, melphalan, cyclophosphamide, or thiotepa, or low-dose TBI (200-400 cGy) with or without anti-thymocyte globulin or alemtuzumab. OS, NRM and relapse incidence (RI) rates at four years were 32%, 32% and 41%, respectively. The median age of patients undergoing RIC was 59 years (range 50-74.7), which was younger than current patients, and no comorbidity data was reported.

Similarly, McClune et al.27 performed another large retrospective analysis of data of the Center for International 1228

Blood and Marrow Transplant Research (CIBMTR) involving 1,080 patients of whom 535 had MDS and received an RIC or NMA conditioning regimen before transplantation. Of this cohort, 181 patients were over 60 years of age but, again, no comorbidity scores were reported. At two years after HCT, results for patients 6064 years of age or over 65 years old included OS 45% and 38%, RFS 35% and 36%, NRM 35% and 39%, and relapse rate of 29% and 21%, respectively. Recently Scott et al.28 concluded a randomized prospective trial showing that more intensive conditioning might be better than RIC for patients with AML or MDS. A reduction in NRM for RIC (4.4% vs. 15.8% at 18 months) was offset by increased disease progression (37% vs. 3.7% at 18 months), although there were no statistically significant differences in OS and relapse-free survival between the two groups of patients.

Storb and Sandmaier29 analyzed results of eleven different retrospective studies using RIC or NMA regimens for various hematologic diseases reported by registries or individual transplant centers on nearly 9,000 patients.

With a median follow up of three years (range 1.75-5), the incidence of disease progression was 43% (range 2265%), NRM 34% (range 6-38%), and OS 38% (22-65%).

Non-myeloablative conditioning has extended the use of allogeneic HCT for patients with myeloid malignancies who were previously not considered candidates for HCT because of advanced age or comorbidities. Initial results with a regimen based on 200 cGy TBI was associated with a relatively

A

B

Figure 4. Incidences of graft-versus-host disease (GvHD) were comparable between different Arms and TBI doses. (A) Acute GvHD, (B) chronic GvHD.

haematologica | 2019; 104(6)

Higher TBI dose improves NMA conditioning HCT for MDS or MPN

high rate of disease progression or relapse. We showed that an incremental increase in TBI resulted in a lower risk of disease progression/relapse among patients with MDS, MPN, and CMML. However, the escalation of the TBI dose to 450 cGy was associated with increased NRM, offsetting the benefit achieved from decreased disease recurrence. Future studies will, therefore, employ targeted radioimmunotherapy in order to reduce relapse/progression without added toxicity and without increasing NRM. Acknowledgments We thank the patients who have participated in this trial along with the clinical staff and research staff who assisted in the con-

References 1. Scott B, Deeg HJ. Hemopoietic cell transplantation as curative therapy of myelodysplastic syndromes and myeloproliferative disorders. Best Pract Res Clin Haematol.

2006;19(3):519-522.

2. McSweeney PA, Niederwieser D, Shizuru JA, et al. Hematopoietic cell transplantation in older patients with hematologic malignancies: replacing high-dose cytotoxic therapy with graft-versus-tumor effects. Blood.

2001;97(11):3390-3400.

3. Niederwieser D, Maris M, Shizuru JA, et al.

Low-dose total body irradiation (TBI) and fludarabine followed by hematopoietic cell transplantation (HCT) from HLA-matched or mismatched unrelated donors and postgrafting immunosuppression with cyclosporine and mycophenolate mofetil (MMF) can induce durable complete chimerism and sustained remissions in patients with hematological diseases. Blood.

2003;101(4):1620-1629.

4. Maris MB, Niederwieser D, Sandmaier BM, et al. HLA-matched unrelated donor hematopoietic cell transplantation after nonmyeloablative conditioning for patients with hematologic malignancies. Blood.

2003;102(6):2021-2030.

5. Sandmaier BM, Maris M, Maloney DG, et al. Low-dose total body irradiation (TBI) conditioning for hematopoietic cell transplants (HCT) from HLA-matched related (MRD) and unrelated (URD) donors for patients with hematologic malignancies: a five-year experience.

Blood.

2003; 102(11):78-79.

6. Baron F, Maris MB, Sandmaier BM, et al.

Graft-versus-tumor effects after allogeneic hematopoietic cell transplantation with nonmyeloablative conditioning. J Clin Oncol.

2005;23(9):1993-2003.

7. Copelan E, Casper JT, Carter SL, et al. A scheme for defining cause of death and its application in the T cell depletion trial. Biol Blood Marrow Transplant. 2007; 13 (12):1469-1476.

8. Glucksberg H, Storb R, Fefer A, et al. Clinical manifestations of graft-versus-host disease in human recipients of marrow from HL-Amatched sibling donors. Transplantation.

1974;18(4):295-304.

9. Gooley TA, Leisenring W, Crowley J, Storer BE. Estimation of failure probabilities in the presence of competing risks: new representations of old estimators. Stat Med. 1999;18(6):695-706.

10. Scott BL, Sandmaier BM, Storer B, et al.

Myeloablative vs nonmyeloablative allo-

haematologica | 2019; 104(6)

11.

12.

13.

14.

15.

16.

17.

18.

19.

duct of the trial. We also thank Helen Crawford for help with manuscript preparation and submission.

Funding Support for this study was provided from grants CA078902 from National Cancer Institute and grant HL36444 from the National Heart, Lung and Blood Institute, National Institutes of Health. This work was also in part supported by the NIH/NCI Cancer Center Support Grant P30 CA015704. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Cancer Institute or the National Institutes of Health nor their subsidiary Institutes and Center.

geneic transplantation for patients with myelodysplastic syndrome or acute myelogenous leukemia with multilineage dysplasia: a retrospective analysis. Leukemia.

2006;20(1):128-135.

Laport GG, Sandmaier BM, Storer BE, et al.

Reduced-intensity conditioning followed by allogeneic hematopoietic cell transplantation for adult patients with myelodysplastic syndrome and myeloproliferative disorders.

Biol Blood Marrow Transplant.

2008;14(2):246-255.

Clift RA, Buckner CD, Appelbaum FR, et al.

Allogeneic marrow transplantation in patients with acute myeloid leukemia in first remission: A randomized trial of two irradiation regimens. Blood. 1990; 76(9):18671871.

Storb R, Yu C, Wagner JL, et al. Stable mixed hematopoietic chimerism in DLA-identical littermate dogs given sublethal total body irradiation before and pharmacological immunosuppression after marrow transplantation. Blood. 1997;89(8): 3048-3054.

Hogan WJ, Little MT, Zellmer E, et al.

Postgrafting immunosuppression with sirolimus and cyclosporine facilitates stable mixed hematopoietic chimerism in dogs given sublethal total body irradiation before marrow transplantation from DLA-identical littermates. Biol Blood Marrow Transplant.

2003;9(8):489-495.

Storb R, Raff RF, Appelbaum FR, et al.

Comparison of fractionated to single-dose total body irradiation in conditioning canine littermates for DLA-identical marrow grafts.

Blood. 1989;74(3):1139-1143.

Storb R, Raff RF, Appelbaum FR, et al.

Fractionated versus single-dose total body irradiation at low and high dose rates to condition canine littermates for DLA-identical marrow grafts. Blood. 1994;83(11): 33843389.

Stelljes M, Bornhauser M, Kroger M, et al.

Conditioning with 8-Gy total body irradiation and fludarabine for allogeneic hematopoietic stem cell transplantation in acute myeloid leukemia. Blood. 2005; 106(9):3314-3321.

Appelbaum FR, Brown P, Sandmaier B, et al.

Antibody-radionuclide conjugates as part of a myeloblative preparative regimen for marrow transplantation.

Blood.

1989; 73(8):2202-2208.

Pagel JM, Gooley TA, Rajendran J, et al.

Allogeneic hematopoietic cell transplantation after conditioning with 131 I-antiCD45 antibody plus fludarabine and lowdose total body irradiation for elderly patients with advanced acute myeloid leukemia or high-risk myelodysplastic syn-

drome. Blood. 2009;114(27):5444-5453.

20. Gopal AK, Guthrie KA, Rajendran J, et al. 90 Y-Ibritumomab tiuxetan, fludarabine, and TBI-based nonmyeloablative allogeneic transplantation conditioning for patients with persistent high-risk B-cell lymphoma.

Blood. 2011;118(4):1132-1139.

21. Sandmaier BM, Bethge WA, Wilbur DS, et al. Bismuth 213-labeled anti-CD45 radioimmunoconjugate to condition dogs for nonmyeloablative allogeneic marrow grafts.

Blood. 2002;100(1):318-326.

22. Chen Y, Kornblit B, Hamlin DK, et al.

Durable donor engraftment after radioimmunotherapy using alpha-emitter astatine211-labeled anti-CD45 antibody for conditioning in allogeneic hematopoietic cell transplantation. Blood. 2012;119(5):11301138.

23. Nakamae H, Wilbur DS, Hamlin DK, et al.

Biodistribution, myelosuppression, and toxicities in mice treated with an anti-CD45 antibody labeled with the Îą-emitting radionuclides bismuth-213 or astatine-211.

Cancer Res. 2009;69(6):2408-2415.

24. Kornblit B, Maloney DG, Storer BE, et al. A randomized phase II trial of tacrolimus, mycophenolate mofetil and sirolimus after non-myeloablative unrelated donor transplantation. Haematologica. 2014; 99(10): 1624-1631.

25. Sandmaier BM, Maloney DG, Storer BE, et al. Sirolimus combined with mycophenolate mofetil (MMF) and cyclosporine (CSP) significantly improves prevention of acute graft-versus-host-disease (GVHD) after unrelated hematopoietic cell transplantation (HCT): Results from a phase III randomized multi-center trial. Blood. 2016;128(22):506.

26. Lim Z, Brand R, Martino R, et al. Allogeneic hematopoietic stem-cell transplantation for patients 50 years or older with myelodysplastic syndromes or secondary acute myeloid leukemia. J Clin Oncol. 2010; 28(3):405-411.

27. McClune BL, Weisdorf DJ, Pedersen TL, et al. Effect of age on outcome of reducedintensity hematopoietic cell transplantation for older patients with acute myeloid leukemia in first complete remission or with myelodysplastic syndrome. J Clin Oncol.

2010;28(11):1878-1887.

28. Scott BL, Pasquini MC, Logan BR, et al.

Myeloablative versus reduced-intensity hematopoietic cell transplantation for acute myeloid leukemia and myelodysplastic syndromes. J Clin Oncol. 2017;35(11):11541161.

29. Storb R, Sandmaier BM. Nonmyeloablative allogeneic hematopoietic cell transplantation. Haematologica. 2016;101(5):521-530.

1229

ARTICLE Ferrata Storti Foundation

Haematologica 2019 Volume 104(6):1230-1236

Stem Cell Transplantation

Antilymphocyte globulin for matched sibling donor transplantation in patients with myelofibrosis

Marie Robin,1 Sylvie Chevret,2 Linda Koster,3 Christine Wolschke,4 Ibrahim Yakoub-Agha,5 Jean Henri Bourhis,6 Patrice Chevallier,7 Jan J. Cornelissen,8 Péter Reményi,9 Johan Maertens,10 Xavier Poiré,11 Charles Craddock,12 Gérard Socié,1 Maija Itälä-Remes,13 Harry C. Schouten,14 Tony Marchand,15 Jakob Passweg,16 Didier Blaise,17 Gandhi Damaj,18 Zubeyde Nur Ozkurt,19 Tsila Zuckerman,20 Thomas Cluzeau,21 Hélène Labussière-Wallet,22 Jörg Cammenga,23 Donal McLornan,24 Yves Chalandon25 and Nicolaus Kröger3

Hôpital Saint-Louis, APHP, INSERM 1131, Paris, France; 2Service de Biostatistique, Hôpital Saint-Louis, APHP, ESCTRA Team, INSERM UMR1153, Université Paris 7, France; 3 EBMT Data Office Leiden, the Netherlands; 4University Hospital Eppendorf, Hamburg, Germany; 5CHU de Lille, LIRIC, INSERM 995, Université de Lille, France; 6Institut Gustave Roussy, INSERM U1186, Université Paris Saclay, Villejuif, France; 7CHU Nantes, France; 8 Department of Hematology, Erasmus University Medical Center, Rotterdam, the Netherlands; 9St. István & St. Laszlo Hospital, Budapest, Hungary; 10University Hospital Gasthuisberg, Leuven, Belgium; 11Cliniques Universitaires St. Luc, Brussels, Belgium; 12 Centre for Clinical Haematology, Queen Elizabeth Hospital, Birmingham, UK; 13HUCH Comprehensive Cancer Center, Helsinki, Finland; 14University Hospital Maastricht, the Netherlands; 15Centre Hospitalier Universitaire de Rennes, France; 16University Hospital, Basel, Switzerland; 17Aix-Marseille University, Inserm, CNRS, Institut Paoli-Calmettes, CRCM, Marseille, France; 18CHU Caen, France; 19Gazi University Faculty of Medicine, Ankara, Turkey; 20Rambam Medical Center, Haifa, Israel; 21Université Cote d’Azur, CHU Nice, INSERM U1065, France; 22Centre Hospitalier Lyon Sud, France; 23University Hospital, Linköping, Sweden; 24Comprehensive Cancer Centre, Department of Haematology, King’s College, London, UK and 25Hôpitaux Universitaires de Genève and Faculty of Medicine, University of Geneva, Switzerland 1

Correspondence: MARIE ROBIN marie.robin@aphp.fr Received: July 11, 2018.

Accepted: January 9, 2019.

Pre-published: January 17, 2019. doi:10.3324/haematol.2018.201400 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/104/6/1230 ©2019 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

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

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

1230

ABSTRACT

T

he use of antihuman T-lymphocyte immunoglobulin in the setting of transplantation from an HLA-matched related donor is still much debated. Acute and chronic graft-versus-host disease are the main causes of morbidity and mortality after allogeneic hematopoietic stem cell transplantation in patients with myelofibrosis. The aim of this study was to evaluate the effect of antihuman T-lymphocyte immunoglobulin in a large cohort of patients with myelofibrosis (n=287). The cumulative incidences of grade II-IV acute graft-versus-host disease among patients who were or were not given antihuman T-lymphocyte immunoglobulin were 26% and 41%, respectively. The corresponding incidences of chronic graft-versushost disease were 52% and 55%, respectively. Non-adjusted overall survival, disease-free survival and non-relapse mortality rates were 55% versus 53%, 49% versus 45%, and 32% versus 31%, respectively, among the patients who were or were not given antihuman T-lymphocyte immunoglobulin. An adjusted model confirmed that the risk of acute graftversus-host disease was lower following antihuman T-lymphocyte immunoglobulin (hazard ratio, 0.54; P=0.010) while it did not decrease the risk of chronic graft-versus-host disease. The hazard ratios for overall survival and non-relapse mortality were 0.66 and 0.64, with P-values of 0.05 and 0.09, respectively. Antihuman T-lymphocyte immunoglobulin did not influence disease-free survival, graft-versus-host disease, relapse-free survival or relapse risk. In conclusion, in the setting of matched related transplantation in myelofibrosis patients, this study demonstrates that antihuman T-lymphocyte immunoglobulin decreases the risk of acute graft-versushost disease without increasing the risk of relapse. haematologica | 2019; 104(6)

ATG in myelofibrosis patients

Introduction Primary myelofibrosis and myelofibrosis secondary to polycythemia vera or essential thrombocythemia are myeloproliferative neoplasms characterized by progressive fibrosis of the bone marrow and myeloid metaplasia in the spleen and liver. Disease severity can be assessed by a number of different prognostic scoring systems, which are able to predict survival without treatment in patients with primary and secondary myelofibrosis.1–5 The risk factors usually taken into account in these scores are disease-related symptoms, the degree of cytopenia or hyperleukocytosis, peripheral or bone marrow blast excess and age. Moreover, cytogenetics and somatic mutations provide additional prognostic power to these scoring instruments.3,6–9 According to the number of risk factors, the expected median survival from diagnosis can range from more than 10 years to less than 18 months.

Allogeneic hematopoietic stem cell transplantation (HSCT) remains the only curative treatment in patients with myelofibrosis. One registry-based study demonstrated that patients with intermediate-2- or high-risk disease according to the Dynamic International Prognostic Scoring System (DIPSS) have an advantage in overall survival following transplantation and international expert consensus guidelines are in favor of transplantation in such patients.10,11 Cumulatively, overall survival after HSCT can range between 40% and 65% according to risk factors related to the disease, patient and type of donor.12– 18 Results have been considered better with transplant from an HLA-matched sibling donor than an unrelated donor. However, acute and chronic graft-versus-host disease (GvHD) remain frequent causes of death in patients with myelofibrosis undergoing HSCT, often contributing to a relatively high transplant-related mortality of around 30%.12–18 The optimal conditioning regimen and GvHD prophylaxis in these patients remain unknown. Two prospective studies of HSCT in myelofibrosis, in which the conditioning regimen and GvHD prophylaxis strategies were homogeneous, can be considered to compare GvHD rates and outcomes. In 2009, Kröger et al. reported on 103 myelofibrosis patients given conditioning with fludarabine, busulfan and the antihuman T-lymphocyte immune globulin Grafalon® at a dose of 30 mg/kg when the graft was from a matched related donor and 60 mg/kg when the donor was unrelated, combined with cyclosporine and a short course of methotrexate. With this regimen, including in vivo T-cell depletion, the rate of acute grade II-IV GvHD was relatively low (27%) and the incidence of chronic GvHD was 49%. The relapse incidence was 32% in the setting of a matched related donor and 20% with an unrelated donor; these incidences were not statistically significantly different. Rondelli et al. subsequently reported a second prospective trial regarding HSCT in myelofibrosis using a fludarabine and melphalan platform in patients transplanted from a matched related donor, with the addition of Thymoglobulin® in patients with an unrelated donor.17 Acute GvHD rates were substantial, being 38% and 41% in the sibling donor group and unrelated donor group, respectively.

Chronic GvHD rates did not differ significantly between the sibling donor (36%) and unrelated donor (38%) groups. Of particular note, mortality was dramatically higher (68%) in the group of patients who underwent unrelated donor HSCT but the effect of antilymphocyte haematologica | 2019; 104(6)

globulin (ATG) on this higher mortality risk remains undetermined. Collectively, from these two studies, it cannot be concluded unambiguously that ATG is beneficial in the setting of transplantation from a matched related donor. Recently, a randomized trial showed that ATG prevents chronic GvHD in patients with acute lymphoid leukemia or acute myeloid leukemia undergoing transplantation from an HLA-matched sibling donor following myeloablative conditioning regimens.19 Indeed, while acute GvHD was non-significantly lowered, the cumulative incidence of chronic GvHD dropped from 69% without ATG to 32% with Grafalon® without increasing the risk of relapse. In this large European Society for Blood and Marrow Transplantation (EBMT) cohort, we aimed to determine the effect of ATG in the setting of HSCT for myelofibrosis using an HLA-matched sibling donor, which is of particular importance as data remain scarce given the rarity of the disease.

Methods Consecutive patients transplanted from a matched sibling donor without ex vivo graft manipulation between 2007 and 2015 for myelofibrosis and registered in the EBMT registry were included in this study. Patients who received post-transplant treatment with cyclophosphamide or alemtuzumab as GvHD prophylaxis and those without sufficient information regarding blood cell counts prior to transplantation were excluded. A total of 287 patients were selected for the final analysis, among whom 135 received in vivo T-cell depletion while 152 did not.

The DIPSS was calculated according to the original definition.1 For some patients the data on peripheral blast count at the time of transplantation were missing; in these cases, the diagnostic blast count was used. General symptoms were either weight loss or sweating (only 2 patients had fever); data on constitutional symptoms were missing for 50 patients. Because information on the brand of drug used for T-cell depletion was not available in the registry, a stepwise hypothetical strategy was formulated to identify patients who received Thymoglobulin® and those who had received Grafalon®: ATG doses of 10 mg/kg or lower were considered as Thymoglobulin® whereas doses of 20 mg/kg or higher were considered as Grafalon® based on usual doses of each brand. This strategy was also checked by country in which the HSCT was performed, as some countries used only Grafalon®, others used only Thymoglobulin® and some used both products.

Disease-free survival was defined as survival without disease relapse or progression documented in the registry. GvHD-free, relapse-free survival (GRFS) was defined as survival without disease relapse or progression, without grade III-IV acute GvHD and without chronic extensive GvHD documented in the registry.

Failure time data were analyzed used Kaplan-Meier estimates, log-rank tests and Cox modeling unless competing risks existed, when cumulative incidence curves, the Gray test and cause-specific Cox models were used.20 When estimating the cumulative incidence of chronic GvHD, patients were censored at the time of donor lymphocyte infusion, as previously reported. Based on frailty models,21 we tested whether there was a center effect on each outcome.

The study complied with regulatory requirements, the declaration of Helsinki and Good Practice standards. Independent review boards approved the study. Patients gave written informed consent.

1231

M. Robin et al.

Results Patients and transplant characteristics The main patient, disease and transplant characteristics are described in Table 1. The median age of the participants was 56.9 years [interquartile range (IQR), 50.6-61.5 years], the minimum was 22.1 years and the maximum was 75.5 years. There was a majority of male patients (68%). Patients who were not given ATG (n=152) and those who were (n=135) had similar characteristics regarding age, gender, and type of myelofibrosis (primary or secondary) but differed for other characteristics including splenectomy before transplant (38% versus 9%), DIPSS classification (intermediate-2 or high: 59% versus 68%), conditioning regimen (Table 1) and source of stem cells (bone marrow 17% versus 2%). More patients in the ATG group received calcineurin inhibitors alone (26% versus 7%). Concerning pre-transplant therapy, five patients in the non-ATG cohort and 14 in the ATG cohort received the JAK inhibitor ruxolitinib (Novartis Pharmaceuticals, Geneva, Switzerland). Regarding the brand of ATG used in the ATG group, 37 patients received Grafalon®, 96 received Thymoglobulin® and the brand was undetermined for two patients.

Engraftment Six patients had primary graft rejection (3 in the ATG group and 3 in the non-ATG group). Four of these patients received a second HSCT and three of them were alive and in remission at the time of last reported follow-up. The cumulative incidences of neutrophil engraftment at day 60 were 96.3% [95% confidence interval (95% CI): 90.9%98.5%] and 94.1% (95% CI: 88.7%-96.9%) for the groups not given or given ATG, respectively (P=0.35). The corresponding cumulative incidences of platelet recovery at 6 months were 68.4% (95% CI: 60.3%-75.2%) and 80.3% (95% CI: 72.3%-86.1%) (P=0.09). Twenty-four patients (14 in the ATG group and 10 in the non-ATG group) had a secondary rejection at a median time of 9 months following HSCT and all but one had disease progression.

Half of them received a second HSCT, which failed to achieve a remission.

Outcomes The median time to onset of acute GvHD was 36 days.

The cumulative incidence of grade II-IV acute GvHD was significantly higher in the group of patients who did receive ATG than in the group of patients who did not: 41.4% (95% CI: 33.1%-49.5%) versus 26.2% (95% CI: 18.7%-34.3%) (P=0.0067) whereas the incidence of grade III-IV GvHD was similar in both groups (Figure 1). The median time to develop chronic GvHD was 198 days. The incidence of chronic GVHD was high (>50%) in both groups of patients (Figure 1) without there being significant differences according to whether or not ATG was administered. Rates of chronic extensive GvHD were also similar in the two groups. The cumulative incidence of relapse was 24.4% (95% CI: 16.5%-33.1%) after ATG and 18.6% (95% CI: 12.1%-26.1%) without ATG (P=0.083).

The rate of non-relapse mortality was 32.5% (95% CI: 24.4%-40.7%) with ATG versus 31% (95% CI: 20.9%41.6%) without ATG. During the follow-up period, a total of 65 patients in the non-ATG group and 44 in the ATG group died. The primary cause of death was related to myelofibrosis progression in 34% non-ATG patients and 1232

Table 1. Patient, disease and transplant characteristics.

No ATG

P value

ATG

Total number 152 135 Median age, years (IQR) 56 (50-61) 58 (51-62) Recipient gender Male (%) 100 (66) 94 (70) Female (%) 52 (34) 41 (30) Median time from diagnosis 41 (15-120) 30 (9-84) to transplant in months (IQR) Disease Primary myelofibrosis (%) 97 (64) 83 (61) Secondary myelofibrosis (%) 44 (29) 50 (37) Transformation into AML (%) 11 (7) 3 (2) Date of transplantation Before 2010 52 (34) 29 (21) 2010 and after 100 (66) 106 (79) Splenectomy before transplant (%) 42 (38) 12 (9) Lille score Low 30 (20) 31 (23) Intermediate 78 (51) 58 (43) High 44 (29) 46 (34) DIPSS score Low 21 (18) 6 (6) Intermediate-1 27 (23) 24 (25) Intermediate-2 45 (39) 32 (34) High 23 (20) 32 (34) Missing 36 41 Conditioning regimen TBI-cyclophosphamide or fludarabine 30 (20) 2 (1.5) Busulfan-cyclophosphamide 18 (12) 2 (1.5) Fludarabine-busulfan±other 37 (24) 110 (81) Fludarabine-melphalan±other 62 (41) 14 (10) FLAMSA 3 (2) 7 (5) Fludarabine-thiotepa 2 (1) 0 GvHD prophylaxis Calcineurin inhibitor alone 8 (5) 39 (29) Calcineurin inhibitor and methotrexate 63 (42) 47 (35) Calcineurin inhibitor and MMF 75 (49) 46 (34) Other 4 (3) 3 (2) Missing 1 (0.6) 0 Recipient CMV serostatus Positive 95 (63) 82 (61) Negative 57 (37) 52 (39) Missing 0 1 Conditioning regimen Reduced intensity 115 (76) 113 (84) TBI-based 39 (26) 3 (2) Source of stem cells Marrow 26 (17) 3 (2) Blood 126 (83) 132 (98) Gender Male recipient / female donor 37 (24) 44 (32) Male recipient / male donor 63 (41) 50 (37) Female recipient / female donor 21 (14) 20 (15) Female recipient / male donor 31 (20) 21 (16) Karnosfsky score, median [range] 90 [80-100] 90 (80-100] 80% or more, n (%) 142/147 (96%) 116/124 (93%)

0.07 0.53

0.13 0.07

0.02

<0.0001 0.58

0.018

< 0.0001

<0.0001

0.90

0.11 <0.0001 <0.0001

0.38

0.27

ATG: antilymphocyte globulin; IQR: interquartile range; AML: acute myeloid leukemia; DIPSS: Dynamic International Prognostic Scoring System; TBI: total body irradiation; FLAMSA: fludarabine, cytarabine and amascrine reduced intensity conditioning; GvHD: graft-versus-host disease; MMF: mycophenolate mofetil; CMV: cytomegalovirus.

haematologica | 2019; 104(6)

ATG in myelofibrosis patients

Figure 1. Acute and chronic graft-versus-host disease. The top panels show the incidences of grade II-IV and grade III-IV acute chronic graft-versus-host disease (GvHD). The bottom panels show the incidences of chronic GvHD and chronic extensive GvHD.

29% in ATG patients. The 5-year overall survival (54.7% versus 52.8%), disease-free survival (49% versus 44.7%), and GRFS (29.3% versus 23.6%) rates were not significantly different between the two groups on univariate analysis (Table 2).

Effects of antithymocyte globulin Due to disparities between the ATG and non-ATG groups, univariate analysis gave no clues on the effects of ATG. A multivariable model was generated to analyze the potential role of ATG on outcomes (Online Supplementary Table S1). Age was the strongest variable significantly associated with overall survival, disease-free survival and non-relapse mortality. Adjustments were made for age at transplantation, Lille score, Karnofsky Performance Status score, splenectomy before transplant, intensity of conditioning regimen (reduced intensity versus myeloablative) and source of stem cells (bone marrow versus peripheral blood). There was no effect of center on any outcome (Online Supplementary Table S2). Table 3 shows the effect of ATG for each outcome. The hazard ratio (HR) showed a benefit from ATG on overall survival (HR: 0.66, 95% CI: 0.43-1.00; P=0.05) and non-relpase mortality (HR: 0.64, 95% CI: 0.39-1.07; P=0.09). The incidence of grade II-IV acute GvHD was significantly lower following the use of haematologica | 2019; 104(6)

ATG (HR: 0.54, 95% CI: 0.34-0.86; P=0.01) but this was not the case for either grade III-IV acute GvHD or chronic extensive GvHD. In this model, ATG did not have a significant impact on disease-free survival, GRFS or relapse risk (see values in Table 3). Figure 2 shows the overall survival, disease-free survival and GRFS taking into account variables of the adjusted model.

Discussion While there is some evidence that in vivo ATG can protect against the occurrence of acute and chronic GvHD, which may translate into a higher probability of GRFS in patients transplanted from an HLA matched related donor,19 there are no specific data from patients with myelofibrosis undergoing HSCT, because of the small numbers of such patients. In this retrospective study on behalf of the EBMT, we analyzed the impact of ATG in the largest documented cohort of patients with myelofibrosis transplanted with an HLA-matched related donor.

Approximately half of the patients received ATG which is higher percentage than that previously reported by the Center for International Blood and Marrow Transplant Research (CIBMTR), according to which only 11% of 1233

M. Robin et al.

patients with a matched related donor received ATG.22 ATG was used less frequently before 2010 (35% versus 51%). The majority of patients received a reduced intensity conditioning regimen and the preferred source of stem cells was peripheral blood. Our study demonstrated that acute GvHD was decreased following the use of ATG but there was no impact on chronic GvHD. The lack of attenuation of the risk of chronic GVHD is in contrast to the findings of the randomized trial comparing ATG versus no ATG in the setting of matched related donor HSCT published recently by KrĂśger et al.19 However, that study included patients with acute leukemia who were given myeloablative conditioning whereas our cohort received predominantly reduced intensity conditioning regimens and the study focused only on myelofibrosis. Of note, the rate of acute GvHD, even in patients who received ATG, was relatively high in our cohort (26%) as compared to that in the prospective study cited above but not dissimilar to the rates in other studies including only patients with myelofibrosis.17,18 The rates of chronic GvHD were significantly high even after ATG; indeed, they were higher than previously reported in this disease setting. This raises the question of whether these myelofibrosis patients were more susceptible to developing chronic GvHD. We could postulate that these patients, who still had myelofibrosis slowly resolving in the first months after transplantation, had a pro-inflammatory profile able to trigger GvHD. Indeed, myelofibrosis is associated with elevated pro-inflammatory biomarkers such as those found in both autoimmune disease and immune dysregulation23â&#x20AC;&#x201C;26 and it has been demonstrated that the bone marrow remains fibrotic at 3 months following HSCT in approximately half of patients.27 Moreover, Hussain et al. reported that even in patients in whom fibrosis resolved following HSCT, the levels of pro-inflammatory cytokines and tissue remodeling factors could remain elevated.28 In contrast, other cytokines are downregulated following HSCT, such as the T-cell inhibitory receptor Tim-3 (T-cell

Table 2. Outcomes in patients with or without T-cell depletion (univariate).

Outcomes: number of events No ATG (n=152)

ATG (n=135)

P value

Neutrophil recovery 143 130 Gray: P=0.35 60-day cum incidence 94.1% (88.7-96.9) 96.3% (90.9-98.5) Platelet recovery 104 108 Gray: P=0.09 180-day cum incidence 68.4% (60.3-75.2) 80.3% (72.3-86.1) Grade II-IV acute GvHD 58 32 Gray: P=0.0067 4-month cum incidence 41.4% (33.1-49.5) 26.2% (18.7-34.3) Grade III-IV acute GVHD 18 20 Gray : P=0.47 4-month cum incidence 11.9% (7.3-17.6) 15.1% (9.6-21.7) Chronic GvHD* 75 62 Gray: P=0.47 5-year cum incidence 51.7% (43.1-59.6) 54.6% (44.5-63.7) Extensive chronic GvHD * 37 33 Gray: P=0.50 5-year cum incidence 25.8% (18.9-33.3) 28.3% (20.4-36.7) Relapse 24 29 Gray: P=0.083 5-year cum incidence 18.6% (12.1-26.1) 24.4% (16.5-33.1) Non-relapse mortality 45 31 Gray: P=0.56 5-year cum incidence 32.5% (24.4-40.7) 31.0% (20.9-41.6) Death 65 44 Logrank P=0.43 Median (95% CI) 63.4 months (39.8-NA) 64 months (44.7-NA) 5-year OS 54.7% (45.1-63.1) 52.8% (42.1-66.3) Cause of death Fisher exact: P=0.52 Relapse/progression 22 (34%) 13 (29%) Other 35 (54%) 28 (64%) Unknown 8 (12%) 3 (7%) Relapse or death 69 60 Log-rank: P=0.46 Median (95% CI) 59.5 months (29-NA) 38.1 months (23.6-NA) 5-year DFS 49.0% (40.6-59.0) 44.7% (34.7-57.4) GvHD relapse death 95 86 Log-rank: P=0.12 Median (95% CI) 9.9 months (8.2-17.7) 7.5 months (6.7-11.3) 5-year GRFS 29.3% (22.0-38.9) 23.6% (15.8-35.2) *censored at donor lymphocyte infusion. ATG: antilymphocyte globulin; cum incidence: cumulative incidence; GvHD: graft-versus-host disease; OS: overall survival; DFS: disease-free survival; GRFS: GvHD-free, relapse-free survival.

Figure 2. Adjusted survival curves. From left to right, overall survival (OS), disease-free survival (DFS), and graft-versus-host disease-free, relapse-free survival (GRFS) in patients who were given antilymphocyte globulin (red) and patients who were not given antilymphocyte globulin (black) with the 95% confidence intervals (dotted lines). Curves have been adjusted according to multiple Cox models.

1234

haematologica | 2019; 104(6)

ATG in myelofibrosis patients

immunoglobulin and mucin-domain containing-3), which may play a role in the control of GvHD.28,29 While ATG clearly decreased the risk of acute GvHD, the adjusted model showed a trend towards improved overall survival in patients who received ATG (P=0.05).

This higher risk of mortality without ATG may be explained by higher risk of acute GvHD even if the excess of mortality was not observed only in the first months post-transplant corresponding to GvHD. Treatment of GvHD and steroid-refractory GvHD may have contributed to mortality in patients who did not receive ATG. Of note, the definitions of acute and chronic GvHD in the registry were still restricted to the time of development of the disease, such that GvHD occurring in the 100 first days was considered as acute GvHD but we had no data regarding late acute GvHD which is considered as chronic GvHD in this study. Indeed, the classification of acute and chronic GvHD was not made according to the latest National Institutes of Health (NIH) consensus and chronic GvHD may have been overestimated because of the inclusion of cases of late acute GvHD.30 Our analysis was based on registry data and GvHD was not recoded a posteriori according to the NIH classification. GRFS, which captures both severe acute and severe chronic GvHD, is an important endpoint in this setting and showed no difference according to the use or not of ATG. Of note, even if the risk of chronic GvHD is mostly influenced by previous acute GvHD, other variables, such as the management of immunosuppressive therapy and cellular therapy may influence the risk of chronic GvHD. Finally, this is the first study that shows a trend to lower mortality in patients receiving ATG. Four prospective trials conducted in the setting of transplantation from unrelated donors and the aforementioned study in the matched sibling donor setting did not find a significant overall survival advantage in patients given ATG.19,31–33 In contrast, one large prospective randomized trial found that overall survival was lower in patients receiving ATG in the setting of unrelated donor transplantation (whether given reduced intensity or myeloablative conditioning).34 It must be considered however that the dose of ATG and the manufacturing process of these products may also have an impact on outcomes and differ in the various prospective trials. In the present

References 1. Passamonti F, Cervantes F, Vannucchi AM, et al. Dynamic International Prognostic Scoring System (DIPSS) predicts progression to acute myeloid leukemia in primary myelofibrosis. Blood. 2010;116(15):2857– 2858.

2. Cervantes F, Dupriez B, Pereira A, et al. New prognostic scoring system for primary myelofibrosis based on a study of the International Working Group for Myelofibrosis Research and Treatment.

Blood. 2009;113(13):2895–2901.

3. Gangat N, Caramazza D, Vaidya R, et al.

DIPSS Plus: a refined Dynamic International Prognostic Scoring System for primary myelofibrosis that incorporates prognostic information from karyotype, platelet count, and transfusion status. J Clin Oncol.

2011;29(4):392–397.

haematologica | 2019; 104(6)

Table 3. Adjusted effect of antilymphocyte globulin on outcomes; adjustment for age at transplant, Lille score, Karnofsky Performance Status, splenectomy, conditioning regimen intensity and source of stem cells.

Overall survival Relapse Non-relapse mortality Grade II-IV acute GvHD Grade III-IV acute GvHD Chronic extensive GvHD Disease-free survival GRFS

Hazard ratio (96% CI) ATG versus none

P value

0.66 (0.43-1.00) 1.31 (0.71-2.42) 0.64 (0.39-1.07) 0.54 (0.34-0.86) 1.11 (0.54-2.28) 1.17 (0.72-1.91) 0.86 (0.59-1.27) 1.05 (0.76-1.46)

0.05 0.39 0.09 0.01 0.77 0.52 0.46 0.74

ATG: antilymphocyte globulin; GvHD: graft-versus-host disease; GRFS: GvHD-free, relapse-free survival.

EBMT study, we were able to identify patients who received Thymoglobulin® or Grafalon® but due to small numbers in the subgroups, we could not draw conclusions regarding the specific impact of the individual products on outcomes. Absolute lymphocyte count may also contribute to the efficiency of ATG and this factor could not be studied here from the registry data.34 We can only postulate that myelofibrosis patients, who usually have not received intensive chemotherapy, may arrive at transplantation with subnormal lymphocyte counts, which can be targeted by ATG. With regards to relapse risk, it was not confirmed in the multivariable model that ATG increased the risk of relapse; however, relapse continued to occur late after HSCT without a real plateau occurring, highlighting the importance of long-term monitoring in myelofibrosis patients who undergo HSCT. In conclusion, this retrospective data analysis of myelofibrosis patients undergoing HSCT whose data were included in the EBMT registry confirms that in vivo ATG is able to protect against acute GvHD and possibly may decrease mortality rates. A prospective study is needed to confirm the role of ATG in myelofibrosis patients transplanted from an HLA-matched related donor.

4. Dupriez B, Morel P, Demory JL, et al.

Prognostic factors in agnogenic myeloid metaplasia: a report on 195 cases with a new scoring system. Blood. 1996;88(3):1013– 1018.

5. Cervantes F, Barosi G, Demory JL, et al.

Myelofibrosis with myeloid metaplasia in young individuals: disease characteristics, prognostic factors and identification of risk groups. Br J Haematol. 1998;102(3):684–690.

6. Guglielmelli P, Lasho TL, Rotunno G, et al.

The number of prognostically detrimental mutations and prognosis in primary myelofibrosis: an international study of 797 patients. Leukemia. 2014;28(9):1804–1810.

7. Guglielmelli P, Lasho TL, Rotunno G, et al.

MIPSS70: mutation-enhanced international prognostic score system for transplantationage patients with primary myelofibrosis. J Clin Oncol. 2018;36(4):310–318.

8. Tefferi A, Lasho TL, Finke CM, et al. CALR

9.

10.

11.

12.

vs JAK2 vs MPL-mutated or triple-negative myelofibrosis: clinical, cytogenetic and molecular comparisons.

Leukemia.

2014;28(7):1472–1477.

Vannucchi AM, Lasho TL, Guglielmelli P, et al. Mutations and prognosis in primary myelofibrosis. Leukemia. 2013;27(9):1861– 1869.

Kröger N, Giorgino T, Scott BL, et al. Impact of allogeneic stem cell transplantation on survival of patients less than 65 years of age with primary myelofibrosis. Blood.

2015;125(21):3347–3350.

Kröger NM, Deeg JH, Olavarria E, et al.

Indication and management of allogeneic stem cell transplantation in primary myelofibrosis: a consensus process by an EBMT/ELN international working group.

Leukemia. 2015;29(11):2126-2133.

Guardiola P, Anderson JE, Bandini G, et al.

Allogeneic stem cell transplantation for

1235

M. Robin et al.

13.

14.

15.

16.

17.

18.

19.

1236

agnogenic myeloid metaplasia: a European Group for Blood and Marrow Transplantation, Société Française de Greffe de Moelle, Gruppo Italiano per il Trapianto del Midollo Osseo, and Fred Hutchinson Cancer Research Center Collaborative Study. Blood. 1999;93(9):2831–2838.

Deeg HJ, Gooley TA, Flowers MED, et al.

Allogeneic hematopoietic stem cell transplantation for myelofibrosis. Blood.

2003;102(12):3912–3918.

Kerbauy DMB, Gooley TA, Sale GE, et al.

Hematopoietic cell transplantation as curative therapy for idiopathic myelofibrosis, advanced polycythemia vera, and essential thrombocythemia. Biol Blood Marrow Transplant. 2007;13(3):355–365.

Patriarca F, Bacigalupo A, Sperotto A, et al.

Allogeneic hematopoietic stem cell transplantation in myelofibrosis: the 20-year experience of the Gruppo Italiano Trapianto di Midollo Osseo (GITMO). Haematologica.

2008;93(10):1514–1522.

Robin M, Tabrizi R, Mohty M, et al.

Allogeneic haematopoietic stem cell transplantation for myelofibrosis: a report of the Société Française de Greffe de Moelle et de Thérapie Cellulaire (SFGM-TC). Br J Haematol. 2011;152(3):331–339.

Rondelli D, Goldberg JD, Isola L, et al. MPDRC 101 prospective study of reduced-intensity allogeneic hematopoietic stem cell transplantation in patients with myelofibrosis. Blood. 2014;124(7):1183–1191.

Kröger N, Holler E, Kobbe G, et al.

Allogeneic stem cell transplantation after reduced-intensity conditioning in patients with myelofibrosis: a prospective, multicenter study of the Chronic Leukemia Working Party of the European Group for Blood and Marrow Transplantation. Blood. 2009;114 (26):5264–5270.

Kröger N, Solano C, Wolschke C, et al.

Antilymphocyte globulin for prevention of chronic graft-versus-host disease. N Engl J

Med. 2016;374(1):43–53.

20. Gooley TA, Leisenring W, Crowley J, Storer BE. Estimation of failure probabilities in the presence of competing risks: new representations of old estimators. Stat Med. 1999;18(6):695–706.

21. Vaupel JW, Manton KG, Stallard E. The impact of heterogeneity in individual frailty on the dynamics of mortality. Demography.

1979;16(3):439–454.

22. Ballen KK, Shrestha S, Sobocinski KA, et al.

Outcome of transplantation for myelofibrosis. Biol Blood Marrow Transplant.

2010;16(3):358–367.

23. Kristinsson SY, Björkholm M, Hultcrantz M, Derolf ÅR, Landgren O, Goldin LR. Chronic immune stimulation might act as a trigger for the development of acute myeloid leukemia or myelodysplastic syndromes. J Clin Oncol. 2011;29(21):2897–2903.

24. Kristinsson SY, Landgren O, Samuelsson J, Björkholm M, Goldin LR. Autoimmunity and the risk of myeloproliferative neoplasms. Haematologica. 2010;95(7):1216– 1220.

25. Hasselbalch HC. Perspectives on chronic inflammation in essential thrombocythemia, polycythemia vera, and myelofibrosis: is chronic inflammation a trigger and driver of clonal evolution and development of accelerated atherosclerosis and second cancer? Blood. 2012;119(14):3219–3225.

26. Hasselbalch HC, Bjørn ME. MPNs as inflammatory diseases: the evidence, consequences, and perspectives. Mediators Inflamm. 2015;2015:102476.

27. Kröger N, Zabelina T, Alchalby H, et al.

Dynamic of bone marrow fibrosis regression predicts survival after allogeneic stem cell transplantation for myelofibrosis. Biol Blood Marrow Transplant. 2014;20(6):812– 815.

28. Hussein K, Stucki-Koch A, Alchalby H, Triviai I, Kröger N, Kreipe H. Cytokine expression pattern in bone marrow

29.

30.

31.

32.

33.

34.

microenvironment after allogeneic stem cell transplantation in primary myelofibrosis.

Biol Blood Marrow Transplant. 2016;22(4): 644–650.

Anderson AC. Tim-3: an emerging target in the cancer immunotherapy landscape.

Cancer Immunol Res. 2014;2(5):393–398.

Filipovich AH, Weisdorf D, Pavletic S, et al.

National Institutes of Health consensus development project on criteria for clinical trials in chronic graft-versus-host disease: I. Diagnosis and staging working group report.

Biol Blood Marrow Transplant.

2005;11(12):945–956.

Bacigalupo A, Lamparelli T, Bruzzi P, et al.

Antithymocyte globulin for graft-versushost disease prophylaxis in transplants from unrelated donors: 2 randomized studies from Gruppo Italiano Trapianti Midollo Osseo (GITMO). Blood. 2001;98(10):2942– 2947.

Finke J, Bethge WA, Schmoor C, et al.

Standard graft-versus-host disease prophylaxis with or without anti-T-cell globulin in haematopoietic cell transplantation from matched unrelated donors: a randomised, open-label, multicentre phase 3 trial. Lancet Oncol. 2009;10(9):855–864.

Walker I, Panzarella T, Couban S, et al.

Pretreatment with anti-thymocyte globulin versus no anti-thymocyte globulin in patients with haematological malignancies undergoing haemopoietic cell transplantation from unrelated donors: a randomised, controlled, open-label, phase 3, multicentre trial. Lancet Oncol. 2016;17(2):164–173.

Soiffer RJ, Kim HT, McGuirk J, et al.

Prospective, randomized, double-blind, phase III clinical trial of anti-T-lymphocyte globulin to assess impact on chronic graftversus-host disease-free survival in patients undergoing HLA-matched unrelated myeloablative hematopoietic cell transplantation. J Clin Oncol. 2017;35(36): 4003– 4011.

haematologica | 2019; 104(6)

ARTICLE

Platelet Biology & Its Disorders

Glycoprotein V is a relevant immune target in patients with immune thrombocytopenia

Ferrata Storti Foundation

Richard Vollenberg,1 Rabie Jouni,2 Peter A. A. Norris,3 Monika Burg-Roderfeld,4 Nina Cooper,1 Mathias J. Rummel,5 Gregor Bein,1 Irene Marini,2 Behnaz Bayat,1 Richard Burack,6 Alan H. Lazarus,3 Tamam Bakchoul2* and Ulrich J. Sachs1,7* *TB and UJS contributed equally to this work

Institute for Clinical Immunology and Transfusion Medicine, Justus Liebig University, Giessen, Germany; 2Center for Clinical Transfusion Medicine, Medical Faculty of Tübingen, Eberhard Karls University, Tübingen, Germany; 3The Canadian Blood Services & The Keenan Research Centre of St. Michael's Hospital, Toronto, ON, Canada; 4Faculty for Chemistry and Biology, Fresenius University of Applied Sciences, Idstein, Germany; 5 IVth Department of Internal Medicine (Hematology/Oncology), Justus Liebig University, Giessen, Germany; 6Department of Pathology and Laboratory Medicine, University of Rochester, NY, USA and 7Center for Transfusion Medicine and Hemotherapy and Hemostasis Center, University Hospital Giessen and Marburg, Marburg, Germany 1

Haematologica 2019 Volume 104(6):1237-1243

ABSTRACT

P

latelet autoantibody-induced platelet clearance represents a major pathomechanism in immune thrombocytopenia (ITP). There is growing evidence for clinical differences between anti-glycoprotein IIb/IIIa and anti-glycoprotein Ib/IX mediated ITP. Glycoprotein V is a well characterized target antigen in Varicella-associated and drug-induced thrombocytopenia. We conducted a systematic study assessing the prevalence and functional capacity of autoantibodies against glycoprotein V. A total of 1140 patients were included. In one-third of patients, platelet-bound autoantibodies against glycoproteins Ib/IX, IIb/IIIa, or V were detected in a monoclonal antibody immobilization of platelet antigen assay; platelet-bound autoanti-glycoprotein V was present in the majority of samples (222 out of 343, 64.7%). Investigation of patient sera revealed the presence of free autoantibodies against glycoprotein V in 13.5% of these patients by an indirect monoclonal antibody immobilization of platelet antigen assay, but in 39.6% by surface plasmon resonance technology. These antibodies showed significantly lower avidity (association/dissociation ratio 0.32±0.13 vs.

0.73±0.14; P<0.001). High- and low-avidity antibodies induced comparable amounts of platelet uptake in a phagocytosis assay using CD14+ positivelyselected human macrophages [mean phagocytic index, 6.81 (range, 4.759.86) vs. 6.01 (range, 5.00-6.98); P=0.954]. In a NOD/SCID mouse model, IgG prepared from both types of anti-glycoprotein V autoantibodies eliminated human platelets with no detectable difference between the groups from the murine circulation [mean platelet survival at 300 minutes, 40% (range, 27-55) vs. 35% (16-46); P=0.025]. Our data establish glycoprotein V as a relevant immune target in immune thrombocytopenia. We would suggest that further studies including glycoprotein V will be required before ITP treatment can be tailored according to platelet autoantibody specificity.

Introduction Immune thrombocytopenia (ITP) is an acquired hemorrhagic autoimmune disease characterized by isolated thrombocytopenia.1,2 Autoantibodies against platelet membrane glycoproteins cause platelet destruction and insufficient compensatory platelet production in the bone marrow (BM).3-5 Cytotoxic effects of T cells have also been described.6 Phagocytosis of antibody-decorated platelets via Fc-receptors or, following complement activation, via complement receptors were long-accepted concepts for the understanding of platelet destruction.6,7 Recent studies have prohaematologica | 2019; 104(6)

Correspondence: ULRICH J. SACHS ulrich.sachs@med.uni-giessen.de Received: November 5, 2018.

Accepted: March 20, 2019.

Pre-published: March 28, 2019. doi:10.3324/haematol.2018.211086 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/104/6/1237 ©2019 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

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

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

1237

R. Vollenberg et al.

vided some evidence that autoantibodies may also trigger more complex processes, such as platelet activation, platelet desialylation, or platelet apoptosis, all of which could lead to Fc-independent platelet clearance.8-11 More recently, there has also been evidence that the glycoprotein specificity of the autoantibodies could be important; for example, in a study by Li et al.,10 desialylation occurred in the presence of anti-GPIbα, but not in the presence of anti-GP IIb/IIIa antibodies.

In general, antibody identification (or the use of monoclonal antibodies in animal models) in these studies was restricted to two types of autoantibody specificities: antiGP IIb/IIIa and anti-GP Ib/IX. This was because these two glycoproteins are currently considered to be the most important autoimmune targets in ITP.12,13 However, glycoprotein V (GP V) is a major protein on the platelet membrane, with approximately 10,000 copies per platelet.14 More than 30 years ago, GP V was first thought to be the immune target of quinidine-related platelet antibodies by Stricker and Shulman,15 and Garner et al.16 described GP V as the antigen in a gold-triggered autoimmune response in patients with rheumatoid arthritis. GP V was also described as the target protein in pediatric varicella-associated thrombocytopenia.17 Some evidence for a potential role of GP V in ITP came from preliminary studies in patients with different types of thrombocytopenia.18,19 A valuable systematic study on GP V in patients with ITP was recently published,20 but whether or not anti-GP V autoantibodies contribute to thrombocytopenia in ITP remains unknown.

Here, we investigated the potential of anti-GP V autoantibodies in patients with ITP. Our work shows that autoantibodies to GP V are found in a majority of patients with ITP and can potentially cause platelet clearance mechanisms. This new information helps fill in some of the missing pathophysiological events in ITP.

Methods Adult patients with a suspected diagnosis of ITP were identified, as previously described.7 In brief, standardized questionnaires covering relevant criteria to refute or confirm a diagnosis of ITP according to the British guidelines were used.21 Patients with relevant other diagnoses that could explain thrombocytopenia, such as aplastic anemia, leukemia, lymphoma, myelodysplastic syndrome, solid tumors, liver cirrhosis, recent cardiac surgery, BM/blood stem cell transplantation, sepsis, and drug-induced thrombocytopenia, were not included. Platelet-bound and free anti-platelet autoantibodies of the IgG type were detected by the monoclonal antibody-specific immobilization of platelet antigen (MAIPA) assay, as described by Kiefel et al.22 Assay sensitivity was controlled by the use of the anti-HPA-1a World Health Organization (WHO) standard (NIBSC, Potters Bar, UK).

Leftover material was used for the additional experiments performed. All anti-GP V sera used in these experiments were negative for the presence of anti-GP IIb/IIIa, anti-GP Ib/IX, anti-GP Ia/IIa, and anti-GP IV by MAIPA. Immunoglobulin G (IgG) fractions were isolated using a commercial purification kit (MelonTM-Gel IgG Spin Purification Kit, Thermo Fisher Scientific, Waltham, MA, US).

Surface plasmon resonance (SPR) analysis allows label-free, realtime investigation of antigen-antibody interactions. This was performed on a protein interaction array system (ProteOn XPR36, Bio-Rad, Munich, Germany). Recombinant His-tagged GP V as 1238

the target protein and GP IV as an irrelevant control (R&D Systems; Life Technologies, Carlsbad, CA, USA) were immobilized onto flow cells of an HTE sensor chip. Phosphate buffered saline-tween (PBS-T) was used as running buffer for all steps. The SPR signal originates from changes in the refractive index at the chip’s surface. For antigen-antibody interactions, changes in the refractive index are linear to the number of antibodies bound.

Data were acquired with the computer software (ProteOn Manager Software, BioRad). Interaction curves were referenced by interspot, second flow cell with immobilized GP IV and monoclonal anti-GP V (MAB42, R&D Systems, 6 μg/mL) as standard.

The R700/R350 ratio was used to differentiate high-avidity (>0.5) and low-avidity (<0.5) antibody binding.23 A phagocytosis assay was performed using CD14+ positivelyselected macrophages (autoMACS Pro Separator; Miltenyi Biotec, Germany) from cryogenically stored human spleen specimens obtained from ITP patients. Healthy donor platelets were fluorescently labeled with CellTracker Green 5-chloromethylfluorescein diacetate (Thermo Fisher Scientific, MA, USA), washed, then opsonized with the ITP serum samples and added to the splenic macrophages for phagocytosis. Macrophages were observed by spinning-disc confocal microscopy under 63x objective oil immersion with differential interference contrast (DIC) and laser fluorescence (488, 647 excitation) on a Quorum multi-modal imaging system (Quorum Technologies, ON, Canada) equipped with a 50 micrometer pinhole spinning disc and an ORCA-Flash 4.0 V2 PLUS sCMOS camera. Four images were taken at the center of each well with Z-stacking every 0.33 μm with >30 stacks. Images were reconstructed in 3D for analysis using Imaris 8.0.2 (Bitplate, UK) and phagocytic index was calculated as (total engulfed platelets / splenic macrophages counted) x 100.

A NOD/SCID mouse model was used to investigate the elimination of human platelets by anti-GPV autoantibodies in ITP patients.24 In brief, NOD/SCID mice (NOD.CB17-Prkdcscid/J; Stock No. complexes, 001303) were purchased from The Jackson Laboratory (Bar Harbor, ME, USA) via Charles River, Research Models and Services (Sulzfeld, Germany). Sex- and age-matched (8-16-week old) animals were used in this study. Human platelets (200 μL, 2x109/mL) were injected into the lateral mouse tail vein.

After 30 minutes (min) a blood sample was collected by tail vein puncture to determine the baseline of circulating human platelets (100%). Subsequently, IgG fractions isolated from human sera containing anti-GPV antibodies or control sera from healthy donors were injected into the other lateral tail vein (2 mg/g body weight). The survival of human platelets in the mouse circulation was analyzed over time using flow cytometry (Cytomics FC 500; Beckman Coulter) after staining platelets with anti-human CD41PE-Cy5 (Beckman Coulter) and anti-mouse CD41-FITC (BD Biosciences, San Diego, CA, USA). Animal experiments were performed with the approval of the local authorities in Tuebingen, Germany. The study was conducted in accordance with the Declaration of Helsinki, and the use of human material was approved by the local ethics committees in Giessen, Germany and Toronto, ON, Canada.

Results Prevalence of platelet-bound autoantibodies against GP V A total of 1645 patients with no alternative reason for a low platelet count were included. The amount of autologous platelets was sufficient for a complete direct test (including all 3 glycoprotein specificities) in 1140 patients (69.3% of n=1645 patients with a clinical suspicion of haematologica | 2019; 104(6)

Glycoprotein V in ITP

Table 1. Summary of autoantibody specificities detected in 343 of 1140 immune thrombocytopenia patients, either on the surface of the patient’s platelets (platelet-bound) or free in patient serum.

Glycoprotein specificity

GP IIb/IIIa only GP Ib/IX only GP V only GP IIb/IIIa plus GP Ib/IX GP IIb/IIIa plus GP V GP Ib/IX plus GP V GP IIb/IIIa plus GP Ib/IX plus GP V Total

Platelet-bound autoantibodies N. of positive samples % 71 30 10 20 10 61 141 343

Free autoantibodies

20.7 8.8 2.9 5.8 2.9 17.8 41.1 100.0

N. of positive samples

%

6 6 6 3 3 9 12 45

13.3 13.3 13.3 6.7 6.7 20.0 26.6 99.9*

*Rounding error.

ITP). This group was further assessed in order to ensure comparability of data. Results are summarized in Table 1.

For patients with a positive test result for at least one glycoprotein, the frequency of immunization against GP V was similar to the other glycoproteins: 242 out of 343 (70.6%) patients were positive for anti-GP IIb/IIIa, 232 out of 343 (67.6%) patients were positive for anti-GP Ib/IX, and 222 out of 343 (64.7%) patients were positive for antiGP V (Kruskal-Wallis test; P=0.67) (Table 1).

Interestingly, there was also no difference in the amount of antibodies attached to GP V (antibody load), as determined by the optical density of the MAIPA assay between glycoproteins: mean values were 1.86 [95% confidence interval (CI): 1.49-2.23] for anti-GP IIb/IIIa, 1.63 (1.27-1.99) for anti-GP Ib/IX, and 1.82 (1.37-2.26) for antiGP V; Kruskal-Wallis test, P=0.77.

Prevalence and binding properties of free autoantibodies against GP V Sera from patients with any positive result in the direct MAIPA test (n=343) were further assessed by indirect MAIPA for the presence of free autoantibodies against these platelet glycoproteins. Results are summarized in Table 1. Free autoantibodies were detected in 45 out of 343 (13.1%) patient samples. The glycoprotein-specific distribution was GP IIb/IIIa (25 out of 45, 55.5%), GP Ib/IX (30 out of 45, 66.6%), and GP V (29 out of 45, 64.4%). Identified free autoantibody specificities matched the platelet-bound specificities from the same patient throughout. Addition of recombinant GP V to sera prior to testing completely blocked the detection of anti-GP V autoantibodies, but did not interfere with the detection of anti-GP IIb/IIIa or anti-GP Ib/IX autoantibodies (data not shown).

Serum IgG fractions from all 222 patients with plateletbound anti-GP V were further analyzed by SPR (Figure 1); 88 out of 222 (39.6%) patients showed specific binding to the GP V flow cell. These 88 sera included all 29 identified as containing anti-GP V by indirect MAIPA. Further analysis demonstrated that these 29 sera with autoantibodies detected both in the indirect MAIPA and in SPR were of higher avidity (R700/R350=0.73±0.14; Wilcoxon rank test, P<0.001) (Figure 1B, top panel) than the 59 sera that gave positive signals in SPR, but were negative in the indirect MAIPA assay (R700/R350=0.32±0.13) (Figure 1B, bottom haematologica | 2019; 104(6)

panel). These results indicate that SPR has better sensitivity compared to the gold standard MAIPA assay in detecting anti-GP V autoantibodies.

Autoantibody-triggered phagocytosis and in vivo platelet clearance Anti-GP V autoantibodies were grouped according to their SPR binding profiles into a “high avidity” and a “low avidity” group. IgG fractions prepared from two highavidity and two low-avidity anti-GP V antibody-containing ITP sera were tested in a phagocytosis assay using CD14 positively-selected human macrophages from ITP spleens (Figure 2). One high- and one low-avidity GP V sera induced significant platelet uptake relative to normal human serum controls (P=0.003 and P=0.026, respectively). Of those positive, high- and low-avidity antibodies induced similar amounts of platelet uptake [mean phagocytic index, 6.81 (range, 4.75-9.86) vs. 6.01 (range, 5.006.98), respectively; P=0.954]. To further assess the biological effect of anti-GP V autoantibodies on platelet destruction, the NOD/SCID mouse model was used. First, moab SW16 against human GP V was injected at two concentrations and the results verified against a murine monoclonal antibody (SZ21) specific for GPIIb/IIIa known to cause thrombocytopenia. SW16 induced similar clearance of human platelets from the murine circulation as SZ21 (mean platelet survival after 300 min, 16±5% vs. 8±8%; P=0.140) (Figure 3A). Platelet elimination was slower when SW16 was injected at a lower concentration (27±4%; P=0.018) (Figure 3A). Next, we analyzed IgG fractions isolated from ITP sera which contained anti-GP V autoantibodies only. Unexpectedly, anti-GP V reduced the survival of human platelets compared to control IgG regardless of their binding properties [median platelet survival after 300 min, “high avidity”, 35% (range, 16-46%; P=0.029) and “low avidity”: 40% (range, 27-55%; P=0.025), respectively] (Figure 3B). After 24 h, only a few injected human platelet circulated in the presence of antiGP V antibodies [median platelet survival after 1440 min, “high avidity”, 22% (range, 11-23%; P=0.0286) and “low avidity”, 20% (range, 13-24%; P=0.029) vs. 46% (range, 43-76%)] (Figure 3B). No difference in platelet elimination was observed between the two groups (P=0.229 and P=0.441, after 300 min and 1440 min, respectively). As expected, autoantibodies were generally less effective in 1239

R. Vollenberg et al.

A

removing platelets from the murine circulation than human alloantibodies (anti-HPA-1a as present in the WHO standard). These data demonstrate that anti-GP V antibodies of high or low avidity are capable of removing circulating platelets and thus represent a functionally relevant specificity of autoantibodies in ITP. To further substantiate the hypothesis that the observed effects are mediated by anti-GP V IgG, we directly compared the median human platelet survival after injection of IgG fractions prepared from one ITP serum containing anti-GP V autoantibodies only (Figure 3C), either after the absorption with recombinant glycoprotein V (rGPV; dashed line) or without (full line). The median platelet survival at t=1440 min after absorption was 48.5% (range, 44-53%) versus 18% (range, 11-20%) without absorption (P=0.028).

This experiment supports our conclusion that anti-GP V IgG is capable of removing human platelets from the murine circulation.

B

Discussion In this study, we demonstrate that GP V is a frequent immune target in ITP patients. Anti-GP V autoantibodies are detectable with the same frequency as those against GP IIb/IIIa and GP Ib/IX. Anti-GP V autoantibodies have the ability to induce a modest level of phagocytosis and to eliminate human platelets in a murine model.

C 1.0 0.8

P<0.001

0.6 0.4 0.2 0.0

Figure 1. Detection of anti-GPV autoantibodies (autoabs) by surface plasmon resonance (SPR). (A) SPR analysis was performed on a protein interaction array system. Recombinant histidine (His)-tagged GP V and GP IV (CD36) were immobilized onto HTE sensor chips. Representative curves for the interaction of monoclonal antibodies with the respective proteins are shown. (B) Representative response curves from n=6 different IgG fractions obtained from immune thrombocytopenia (ITP) patients. (Top) Reactivity of IgG fractions from n=3 sera from patients with free autoabs that were also detectable by standard serology (monoclonal antibody immobilization of platelet antigens, MAIPA). (Bottom) Reactivity of IgG fractions from n=3 sera without detectable anti-GP V by standard serology (MAIPA). Note the difference in the maximum response units (y-axis) and the different behavior of antibodies with regard to association and dissociation characteristics. (C) Comparison of the avidity of MAIPA positive (n=29) versus MAIPA negative (n=59) ITP sera detected by SPR in a box-and-whisker plot with median, interquartile range, and highest/lowest value per group. Avidity was calculated as the R700/R350 rate, where R350 indicates the maximum anti-GP V antibody binding after 350 seconds (s) of association, and R700 indicates the remaining antibody binding after additional 350 s of dissociation.

1240

Figure 2. Phagocytosis of platelets by splenic macrophages. Healthy donor platelets were opsonized with human immune thrombocytopenia (ITP) sera containing low-avidity (group A, red bars) or high-avidity (group B, blue bars) anti-GP V autoantibodies from four different patients (n=4 each). Serum from one anti-GPIIb/IIIa-positive ITP patient was used as a positive control (n=3) and healthy donor sera [normal human sera (NHS), n=4] or phosphate buffered saline (PBS) [non-opsonized (non-ops), n=4] were used as negative controls.

Human ITP splenic macrophages were isolated by CD14 positive selection from frozen adult ITP splenic single-cell suspensions and were incubated with opsonized human platelets for 40 minutes at 37°C. Phagocytosis was determined by spinning disc confocal microscopy and outside (non-phagocytosed) platelets were distinguished using an anti-GPIX antibody stain following macrophage fixation. Phagocytic index was calculated as (engulfed platelets counted / splenic macrophages counted) x 100. Error bars=Standard Deviation.

Statistical significance was calculated by one-way ANOVA against NHS unless specified. NS: not significant; *P<0.05; **P<0.01; ****P<0.0001.

haematologica | 2019; 104(6)

Glycoprotein V in ITP

Autoantibodies against platelets are considered to be the major factors for platelet clearance in ITP.25 GP IIb/IIIa and GP Ibα are generally reported as the most common antigenic targets.26,27 We were able to analyze the amount of platelet-bound, glycoprotein-specific autoantibodies in 343 patients in parallel. Whereas anti-GP V antibodies as the sole autoantibody entity were less often detected (2.9%) than the other specificities (20.7% and 8.9%, respectively), two-thirds of all ITP patients reacted with GP V in combination with other specificities. Anti-GP V was more often seen in association with anti-GP Ib/IX (61 out of 91, 67%) than with anti-GP IIb/IIIa (10 out of 8, 12%), but all entities were clearly separable.

In contrast to platelet-bound autoantibodies, free autoantibodies in patient plasma are only rarely detectable.28 Still, in our cohort, free anti-GP V was not less often detected than the other specificities, again, most frequently in association with other autoantibodies. Adding anti-GP V detection to the standard laboratory test would only mildly increase the overall test sensitivity (from 29.2% to 30.1% for platelet-bound glycoprotein specific autoantibodies and from 3.4% to 3.9% for free autoantibodies). In contrast to conventional testing by MAIPA, SPR technology significantly raised the test sensitivity, with no loss of specificity. Interestingly, we observed a clear difference in autoantibody avidity between those autoantibodies detected by standard serology plus SPR

and those detected by SPR only. To our knowledge, autoantibodies against platelets have not been investigated for their avidity before. However, we previously demonstrated that anti-HPA-1a alloantibodies against platelets may be of low avidity and escape detection by MAIPA.23 These antibodies had a comparable profile to the SPR-only autoantibodies detected in this study: a slow binding during the association and fast detachment during the dissociation phase. This suggests that these antibodies may become washed away in conventional test methods, whereas no-wash detection by SPR increases sensitivity.

Low-avidity anti-GP V autoantibodies were able to induce platelet destruction in vitro and in vivo. This finding indicates that these antibodies, which are not detectable using conventional methods, are of clinical relevance. This observation demonstrates that low sensitivity could, in fact, be an important drawback of autoantibody testing in the laboratory. Further development of methods might be useful to increase the clinical utility of platelet autoantibody testing.29 Anti-GP V autoantibodies were efficient in removing platelets, regardless of their avidity; indicating that platelets loaded with anti-GP V undergo the same fate as platelets loaded with other autoantibodies.23 The presence of anti-GP V might affect the clinical picture of ITP patients in two ways: 1) by a more efficient platelet removal because of an increased overall IgG load; or 2) by

A

B

Figure 3. Platelet elimination in a NOD/SCID mouse model. NOD/SCID mice (NOD.CB17Prkdcscid/J) were injected with freshly isolated human platelets (200 μL, 2x109/mL) into the lateral tail vein. After 30 minutes (min), blood was taken to determine the baseline of circulating human platelets (100%). Subsequently, antibodies were injected, and further blood samples were taken at 60, 120, 300 and 1440 min (24 h) after baseline, as indicated. (A) Platelet elimination in the presence of monoclonal antibodies. SZ21 against GP IIIa was used as positive control. Note that SW16 against GP V eliminated human platelets with the same kinetics as SZ21 when 10 µg were injected. Significant, but less pronounced platelet elimination was observed at a lower concentration. Data are given as Mean±Standard Deviation for three independent experiments. Murine IgG was used as negative control. (B) Platelet elimination in the presence of human anti-GP V autoantibodies. Group A: anti-GP V IgG detected by surface plasmon resonance (SPR) only (low avidity antibodies; n=3); group B: anti-GP V IgG detected by SPR and monoclonal antibody immobilization of platelet antigens (MAIPA) (high avidity antibodies, n=3). Compared to human control IgG, IgG obtained from immune thrombocytopenia (ITP) sera in both groups induced significant platelet removal. No difference in platelet elimination was observed between group A and group B. As to be expected, autoantibodies were less effective in removing platelets from the murine circulation than a human control alloantibody [anti-HPA-1a present in the World Health Organization (WHO) standard]. (C) Human ITP serum containing anti-GP V autoantibodies only was absorbed (dashed line) or not absorbed (full line) with recombinant GP V prior to IgG isolation. Platelet elimination was studied as outlined above. Data are given as Mean±Standard Deviation for three independent experiments.

C

haematologica | 2019; 104(6)

1241

R. Vollenberg et al.

functional effects of anti-GP V with subsequent changes in platelet reactivity. Since this was a laboratory-based study with one-stage clinical and laboratory data only, and no follow up, no definite conclusions can be drawn.

The low level of phagocytosis induced by these autoantibodies may hint at a unique mechanism of thrombocytopenia, or could indicate that a co-factor found in vivo but not in vitro (complement components, C-reactive protein, or serum amyloid A) is required.30 Alternatively, it is possible that the highest affinity antibodies remain bound to platelets and those in the sera have lower affinity and, therefore, trigger lower levels of phagocytosis.

Antibodies against GP V could exert different functional effects on platelets: GP V is cleaved by thrombin or, following platelet activation with collagen, by ADAM17/TACE.31,32 GP V is thought to function as a negative modulator of thrombin-induced platelet activation.33 In vivo studies in mice have demonstrated that the absence of GP V increases both platelet adhesion and aggregation; but also decreases thrombus stabilization.34 Whether any of these physiological processes are affected by anti-GP V autoantibodies is currently not known. Since we have now established GP V as an important immune target in ITP, it will be important to study whether the presence (or absence) of anti-GP V antibodies also affects treatment efficacy, as previously reported for the two other autoantibody specificities.35,36 This study has some limitations. Only ITP patients in

References 1. Cines DB, Bussel JB, Liebman HA, Luning Prak ET. The ITP syndrome: pathogenic and clinical diversity. Blood. 2009; 113(26):6511-6521.

2. Provan D, Stasi R, Newland AC, et al.

International consensus report on the investigation and management of primary immune thrombocytopenia. Blood. 2010; 115(2):168-186.

3. Harrington WJ, Minnich V, Hollingsworth JW, Moore CV. Demonstration of a thrombocytopenic factor in the blood of patients with thrombocytopenic purpura. J Lab Clin Med. 1951;38(1):1-10.

4. Nugent D, McMillan R, Nichol JL, Slichter SJ. Pathogenesis of chronic immune thrombocytopenia: increased platelet destruction and/or decreased platelet production. Br J Haematol. 2009;146(6):585-596.

5. Olsson B, Andersson PO, Jernas M, et al. Tcell-mediated cytotoxicity toward platelets in chronic idiopathic thrombocytopenic purpura. Nat Med. 2003;9(9):1123-1124.

6. Neschadim A, Kotra LP, Branch DR. Small molecule phagocytosis inhibitors for immune cytopenias. Autoimmun Rev. 2016;15(8):843-847.

7. Najaoui A, Bakchoul T, Stoy J, et al.

Autoantibody-mediated complement activation on platelets is a common finding inpatients with immune thrombocytopenic purpura (ITP). Eur J Haematol. 2012; 88(2):167-174.

8. Yan R, Chen M, Ma N, et al. Glycoprotein Ibα clustering induces macrophage-mediated platelet clearance in the liver. Thromb Haemost. 2015;113(1):107-117.

1242

whom a complete direct MAIPA test could be performed qualified. This cohort may not be representative for all ITP patients. In addition, antibodies of the IgA or IgM type, which are rarely detected in ITP,27,28,37 were not studied. We were also unable to characterize IgG subclasses in our cohort. Whereas others have shown that the majority of anti-GPIIb/IIIa autoantibodies are of the IgG1 subclass, some IgG2, 3 and 4 have been reported.38 The IgG subclass distribution of anti-GP V may differ from anti-GPIIb/IIIa. Finally, any blood sample taken from an ITP patient may not reflect the in vivo situation, since platelets sensitized with high-avidity antibodies may have been cleared (together with these antibodies) from the circulation before the sample was taken.

Despite these restrictions, we have confirmed GP V as a frequent immune target in ITP and demonstrated that anti-GP V autoantibodies are of clinical relevance since they can remove platelets from the circulation. We have also, for the first time, demonstrated that low platelet autoantibody avidity might be the main reason why current serology does not detect platelet autoantibodies more often. We would suggest that studies including GP V as an immune target are required before ITP treatment can be tailored according to platelet autoantibody specificities.

Acknowledgments The authors would like to thank Astrid Giptner, Heike Berghöfer and Renate Marschall for excellent technical support.

9. Quach ME, Dragovich MA, Chen W, et al.

Fc-independent immune thrombocytopenia via mechanomolecular signaling in platelets. Blood. 2018;131(7):787-796.

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

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

Nat Commun. 2015;6:7737.

11. Goette NP, Glembotsky AC, Lev PR, et al.

Platelet apoptosis in adult immune thrombocytopenia: insights into the mechanism of damage triggered by auto-antibodies.

PLoS One. 2016;11(8):e0160563.

12. van Leeuwen EF, van der Ven JT, Engelfriet CP, von dem Borne AE. Specificity of autoantibodies in autoimmune thrombocytopenia. Blood. 1982;59(1):23-26.

13. Kiefel V, Santoso S, Kaufmann E, MuellerEckhardt C. Autoantibodies against platelet glycoprotein Ib/IX: a frequent finding in autoimmune thrombocytopenic purpura.

Br J Haematol 1991;79(2):256-262.

14. Modderman PW, Admiraal LG, Sonnenberg A, von dem Borne AE. Glycoproteins V and Ib-IX form a noncovalent complex in the platelet membrane. J Biol Chem. 1992;267(1):364-369.

15. Stricker RB, Shuman MA. Quinidine purpura: evidence that glycoprotein V is a target platelet antigen. Blood. 1986; 67(5):1377-1381.

16. Garner SF, Campbell K, Metcalfe P, et al.

Glycoprotein V: the predominant target antigen in gold-induced autoimmune thrombocytopenia.

Blood.

2002; 100(1):344-346.

17. Mayer JLR, Beardsley DA. Varicella-associated thrombocytopenia: autoantibodies against platelet surface glycoprotein V.

Pediatr Res. 1996;40(4):615-619.

18. Joutsi L, Kekomäki R. Comparison of the direct platelet immunofluorescence test (direct PIFT) with a modified direct monoclonal antibody-specific immobilization of platelet antigens (direct MAIPA) in detection of platelet-associated IgG. Br J Haematol. 1997;96(1):204-209.

19. Joutsi-Korhonen L, Javela K, Hormila P, Kekomäki R. Glycoprotein V-specific platelet associated antibodies in thrombocytopenic patients. Clin Lab Haem. 2001; 23(5):307-312.

20. Porcelijn L, Huiskes E, Oldert G, et al.

Detection of platelet autoantibodies to identify immune thrombocytopenia: state of the art. Br J Haematol. 2018;182(3):423426.

21. British Committee for Standards in Haematology General Haematology Task Force. Guidelines for the investigation and management of idiopathic thrombocytopenic purpura in adults, children and in pregnancy. Br J Haematol. 2003;120(4):574596.

22. Kiefel V, Santoso S, Weisheit M, MuellerEckhardt C. Monoclonal antibody-specific immobilization of platelet antigens (MAIPA): a new tool for the identification of platelet-reactive antibodies. Blood. 1987; 70(6):1722-1726.

23. Bakchoul T, Kubiak S, Krautwurst A, et al.

Low-avidity anti-HPA-1a alloantibodies are capable of antigen-positive platelet destruction in the NOD/SCID mouse model of alloimmune thrombocytopenia.

Transfusion. 2011;51(11):2455-2461.

24. Bakchoul T, Walek K, Krautwurst A, et al.

Glycosylation of autoantibodies: insights into the mechanisms of immune thrombo-

haematologica | 2019; 104(6)

Glycoprotein V in ITP

25.

26.

27.

28.

29.

cytopenia. Thromb Haemost. 2013; 110(6):1259-1266.

McMillan R. The pathogenesis of chronic immune thrombocytopenic purpura.

Semin Hematol. 2007;44(4 Suppl 5):S3-S11.

Beardsley DS, Ertem M. Platelet autoantibodies in immune thrombocytopenic purpura. Transfus Sci. 1998;19(3):237-244.

He R, Reid DM, Jones CE, Shulman NR. Spectrum of Ig classes, specificities, and titers of serum antiglycoproteins in chronic idiopathic thrombocytopenic purpura.

Blood. 1994;83(4):1024-1032.

Kiefel V, Freitag E, Kroll H, Santoso S, Mueller-Eckhardt C. Platelet autoantibodies (IgG, IgM, IgA) against glycoproteins IIb/IIIa and Ib/IX in patients with thrombocytopenia. Ann Hematol. 1996;72(4):280285.

Kong X, Ye B, Yang Z, Chen B, Ling Y. Simultaneous detection of platelet-specific antibodies based on a photonic crystalencoded suspension array. Clin Chim Acta.

2016;458:72-77.

haematologica | 2019; 104(6)

30. Kapur R, Heitink-PollĂŠ KM, Porcelijn L, et al. C-reactive protein enhances IgG mediated phagocyte responses and thrombocytopenia. Blood. 2015;125(11):1793-1802.

31. Berndt MC, Phillips DR. Interaction of thrombin with platelets: purification of the thrombin substrate. Ann N Y Sci 1981; 370:87-95.

32. RabieT, Strehl A, Ludwig A, Nieswandt B. Evidence for a role of ADAM17 (TACE) in the regulation of platelet glycoprotein V. J Biol Chem. 2005;280(15):14462-14468.

33. Ramakrishnan V, Reeves PS, DeGuzman F. Increased thrombin responsiveness in platelets from mice lacking glycoprotein V. Proc Natl Acad Sci U S A. 1999; 96(23):13336-13341.

34. Ni H, Ramakrishnan V, Ruggeri ZM, Papalia JM, Phillips DR, Wagner DD. Increased thrombogenesis and embolus formation in mice lacking glycoprotein V. Blood. 2001;98(2):368-373.

35. Zeng Q, Zhu L, Tao L, et al. Relative efficacy of steroid therapy in immune thrombo-

cytopenia mediated by anti-platelet GPIIbIIIa versus GPIbÎą antibodies. Am J Hematol. 2012;87(2):206-208.

36. Peng J, Ma SH, Liu J, Hou Y, et al.

Association of autoantibody specificity and response to intravenous immunoglobulin G therapy in immune thrombocytopenia: a multicenter cohort study. J Thromb Haemost. 2014;12(4):497-504.

37. Hurlimann-Forster M, Steiner B, von Felten A. Quantitation of platelet-specific autoantibodies in platelet eluates of ITP patients measured by a novel ELISA using the purified glycoprotein complexes GPIIb/IIIa and GPIb/IX as antigens. Br J Haematol.

1997;98(2):328-335.

38. Chan H, Moore JC, Finch CN, Warkentin TE, Kelton JG. The IgG subclasses of platelet-associated autoantibodies directed against platelet glycoproteins IIb/IIIa in patients with idiopathic thrombocytopenic purpura. Br J Haematol. 2003;122(5):818824.

1243

ARTICLE Ferrata Storti Foundation

Haematologica 2019 Volume 104(6):1244-1255

Platelet Biology & Its Disorders

Downregulation of TREM-like transcript-1 and collagen receptor α2 subunit, two novel RUNX1-targets, contributes to platelet dysfunction in familial platelet disorder with predisposition to acute myelogenous leukemia Ana C. Glembotsky,1* Dominika Sliwa,2* Dominique Bluteau,2,3* Nathalie Balayn,2 Cecilia P. Marin Oyarzún,1 Anna Raimbault,2 Marie Bordas,2 Nathalie Droin,2,4 Iryna Pirozhkova,5 Valance Washington,6 Nora P. Goette,1 Rosana F. Marta,1 Rémi Favier,2,7 Hana Raslova2** and Paula G. Heller1**

Hematología Investigación, Instituto de Investigaciones Médicas “Dr. Alfredo Lanari”, Facultad de Medicina, Universidad de Buenos Aires, Consejo Nacional de Investigaciones Científicas y Tecnológicas (CONICET), Argentina; 2INSERM UMR 1170, Gustave Roussy, Université Paris-Saclay, Equipe Labellisée par la Ligue Nationale Contre le Cancer, Villejuif, France; 3Ecole Pratique des Hautes Etudes (EPHE), Paris, France; 4Gustave Roussy, Université Paris-Saclay, Genomic Platform UMS AMMICA, Villejuif, France; 5CNRS UMR 8126, Gustave Roussy, Université Paris-Saclay, Villejuif, France; 6Department of Biology, University of Puerto Rico-Rio Piedras, San Juan, Puerto Rico and 7Assistance Publique-Hôpitaux de Paris, Hôpital Trousseau, CRPP, Services d’Hématologie Biologique et Clinique, Paris, France 1

*ACG, DS and DB contributed equally to this work. **HR and PGH contributed equally to this work as senior co-authors.

ABSTRACT

Correspondence: PAULA HELLER paulaheller@hotmail.com HANA RASLOVA hana.raslova@gustaveroussy.fr Received: January 20, 2018.

Accepted: December 10, 2018.

Pre-published: December 13, 2018. doi:10.3324/haematol.2018.188904 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/104/6/1244 ©2019 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

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

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

1244

G

ermline RUNX1 mutations lead to thrombocytopenia and platelet dysfunction in familial platelet disorder with predisposition to acute myelogenous leukemia (AML). Multiple aspects of platelet function are impaired in these patients, associated with altered expression of genes regulated by RUNX1. We aimed to identify RUNX1targets involved in platelet function by combining transcriptome analysis of patient and shRUNX1-transduced megakaryocytes (MK). Down-regulated genes included TREM-like transcript (TLT)-1 (TREML1) and the integrin subunit alpha (α)-2 (ITGA2) of collagen receptor α2-beta (b)-1, which are involved in platelet aggregation and adhesion, respectively.

RUNX1 binding to regions enriched for H3K27Ac marks was demonstrated for both genes using chromatin immunoprecipitation. Cloning of these regions upstream of the respective promoters in lentivirus allowing mCherry reporter expression showed that RUNX1 positively regulates TREML1 and ITGA2, and this regulation was abrogated after deletion of RUNX1 sites. TLT-1 content was reduced in patient MK and platelets. A blocking anti-TLT-1 antibody was able to block aggregation of normal but not patient platelets, whereas recombinant soluble TLT-1 potentiated fibrinogen binding to patient platelets, pointing to a role for TLT-1 deficiency in the platelet function defect. Low levels of α2 integrin subunit were demonstrated in patient platelets and MK, coupled with reduced platelet and MK adhesion to collagen, both under static and flow conditions. In conclusion, we show that gene expression profiling of RUNX1 knock-down or mutated MK provides a suitable approach to identify novel RUNX1 targets, among which downregulation of TREML1 and ITGA2 clearly contribute to the platelet phenotype of familial platelet disorder with predisposition to AML. haematologica | 2019; 104(6)

Low TLT-1 and collagen receptor α2 in FPD/AML

Introduction The transcription factor RUNX1 is a key regulator of the megakaryocytic lineage, where it participates in a complex transcriptional network co-ordinating platelet biogenesis and function.1,2 RUNX1 co-operates with other transcriptional regulators, including GATA-1, FLI-1 and SCL at megakaryocyte (MK)-specific promoters,1 whereas co-occupancy of RUNX1 with FLI-1 and NF-E2 has been shown to prime the late MK program.3 Conditional RUNX1 inactivation in mice leads to MK maturation arrest and a substantial decline in platelet counts, highlighting the key role of RUNX1 as a master regulator of the MK lineage.4 Germline RUNX1 mutations in humans underlie familial platelet disorder with predisposition to acute myelogenous leukemia (FPD/AML), which is characterized by thrombocytopenia, platelet dysfunction, and a lifelong 30-50% predisposition to hematologic malignancies, including myeloid and lymphoid neoplasms.5 RUNX1 mutations exerting dominant-negative effects over the wild-type (WT) protein are associated with a higher leukemic rate than those acting via haploinsufficiency,6 whereas no differences in the severity of the platelet phenotype are seen between both types of mutations.7 Although once considered a rare condition, FPD/AML is now diagnosed at increasing frequency due to heightened diagnostic awareness during the workup of individuals presenting with thrombocytopenia of uncertain etiology or hereditary myeloid malignancies.

The platelet defect in FPD/AML is complex and includes abnormalities in platelet number and function, which lead to a bleeding diathesis of variable severity, ranging from mild or asymptomatic cases to a severe bleeding tendency. Thrombocytopenia is usually mild to moderate and is caused by impaired platelet production secondary to defects in multiple steps of MK development, including MK differentiation, maturation, polyploidization and proplatelet formation.2 While marked dysmegakaryopoiesis with a severe defect in proplatelet formation is observed in vitro, the presence of only mild thrombocytopenia, often at the lower limit of normal range, suggests a yet unknown compensatory mechanism in vivo. The platelet function defect is present in most, if not all, patients with FPD/AML and involves multiple abnormalities in platelet structure and activation pathways, including defective platelet aggregation and release, dense granule deficiency, associated in some pedigrees with partial alpha (α)-granule defect, and impaired αIIb-beta (b)-3 (GPIIbIIIa) activation and outside-in signaling.8,9 These abnormalities are likely due to altered expression of RUNX1-targets involved in platelet biology. The study of FPD/AML platelet samples has revealed downregulation of several RUNX1-regulated genes. Transcriptome analysis of platelets from one patient showed reduced levels of MYL9, ALOX12, PKCθ, RAB1B and PLDN,10 whereas dysregulated expression of MPL,11 MYH1012, RAB27B8 and NF-E28 has been identified by a candidate-gene approach in other pedigrees.

However, the mechanisms underlying FPD/AML platelet function defect and the effects of RUNX1 mutations on the expression of other potential genes are still not completely understood. In this study, we combined expression profiling of mature shRUNX1-transduced and FPD/AML MK to gain further insight into RUNX1-regulated genes involved in platelet function. Using this haematologica | 2019; 104(6)

approach, we identified triggering receptor expressed on myeloid cells (TREM)-like transcript (TLT)-1 and integrin subunit α2 of collagen receptor α2b1 as two novel RUNX-1 targets, whose expression was decreased in FPD/AML MK and platelets.

Methods Human samples Patients from three previously described FPD/AML pedigrees2,8,11 (Table 1 and Online Supplementary Table S1), healthy subjects, and individuals after stem cell mobilization were included. At the time of the study, patients had thrombocytopenia and/or platelet dysfunction with no evidence of myelodysplastic or leukemic transformation. Details on experiments performed on each patient are provided in Online Supplementary Table S2. The study was approved by the Ethics Committee of INSERM RBM 01-14 for the project “Network on the inherited diseases of platelet function and platelet production” in France and the Ethics Committee of the Instituto de Investigaciones Médicas “Dr. Alfredo Lanari” in Argentina. Patients and controls gave signed informed consent.

Megakaryocyte culture and transcriptome analysis CD34+ cells were isolated from cord blood, leukapheresis samples or peripheral blood of patients and healthy subjects using a magnetic cell-sorting system (AutoMACS or MiniMACS, Miltenyi Biotec SAS, Paris, France) and grown in serum free medium2 or Stem Span medium (StemCell Technologies, Vancouver, BC, Canada), supplemented with 10 ng/mL thrombopoietin (TPO) (Kirin Brewery, Tokyo, Japan or Miltenyi Biotec) and 25 ng/mL Stem Cell Factor (SCF) (Biovitrum AB, Stockholm, Sweden or Miltenyi Biotec). For culture of patient MK, 10 ng/mL IL-6 (Tebu or Miltenyi Biotec), 100 U/mL IL-3 (Novartis or R&D Systems, MN, USA) and 1 ng/mL fetal liver tyrosine kinase 3 ligand (FLT3-L) (Celldex Therapeutics or R&D Systems) were added.

For transcriptome analysis, patient and control MK were cultured as detailed above, stained on day 10 of culture with allophycocyanin (APC)-conjugated anti-CD41 and phycoerythrin (PE)-anti-CD42 antibodies (BD Biosciences, Le Pont de Claix, France), and CD41+CD42+ were sorted by flow cytometry.

CD34+ cells from leukapheresis samples were transduced on days 6 and 7 of culture with lentiviruses encoding shRUNX1_1, shRUNX1_2 and shSCR (control shRNA), and CD41+CD42+GFP+ cells were sorted on day 10 by flow cytometry, as previously described.2,13 RNA was extracted using the RNeasy Micro Kit (Qiagen, France) according to the manufacturer’s instructions. Transcriptome analysis was performed using the Agilent Whole Human Genome Microarray (see Online Supplementary Methods).

Statistical analysis For comparison between patients and controls, MannWhitney test or Wilcoxon matched pairs test were applied. For promoter activity and chromatin immunoprecipitation-polymerase chain reaction (ChIP-PCR) assays, paired t-test was used.

For assessment of the effect of a blocking anti-TLT-1 antibody on proplatelet formation, repeated measures ANOVA was used.

All statistical analyses were two-sided; P<0.05 was considered significant. The GraphPad Prism 6.01 (La Jolla, CA, USA) software was used for analysis.

Other methods are described in the Online Supplementary Appendix.

1245

A.C. Glembotsky et al.

A

C

B

D

E

G F

Figure 1. Transcriptome of RUNX1-deficient megakaryocytes (MK) reveals downregulation of TREML1 and ITGA2. (A and B) Venn diagram representation of the overlap between genes deregulated in patient MK (BII-2, BIII-1, AII-1 and AII-2), between genes deregulated in shRUNX1-transduced MK (shRUNX1_1_A, shRUNX1_1_B and shRUNX1_2), and between genes deregulated in both patient and shRUNX1-transduced MK. (A) A total of 598 genes were up-regulated in patient MK (n=4) and 289 genes were up-regulated in shRUNX1-transduced MK (n=3). A total of 43 genes were found to be up-regulated in both patient and shRUNX1transduced MK. (B) A total of 492 genes were down-regulated in patient MK (n=4) and 499 genes (527 probes) were down-regulated in shRUNX1-transduced MK (n=3). A total of 61 genes were found to be down-regulated in both patient and shRUNX1-transduced MK. shRUNX1_1 was used in two experiments (shRUNX1_1_A and shRUNX1_1_B) and shRUNX_2 in one. (C) Differentially down-regulated genes and functions associated with different biological processes. Gene ontology (GO) terms and differentially down-regulated genes are represented as nodes based on their kappa score >0.4. The node size represents the GO terms enrichment significance. (D) Microarray data for ITGA2 and TREML1 in MK from patients (AII-1 and AII-2, carrying the R174Q mutation, and BII-2 and BIII-1, with the R139X mutation) relative to healthy subjects (n=4) (horizontal dashed line). (E) Real-time polymerase chain reaction (RT-PCR) analysis of RUNX1, ITGA2 and TREML1 transcript levels normalized to PPIA during normal MK in vitro differentiation. Cells were analyzed on day (D)6, D9, and D13 of culture. Data represent Mean±Standard Deviation (SD) of three independent experiments. (F) RT-PCR analysis of RUNX1, TREML1 and ITGA2 transcript levels normalized to PPIA in MK transduced with lentiviruses encoding shRUNX1 (shRUNX1_1 and shRUNX_2) relative to control shRNA (shSCR). Data represent Mean±SD of three independent experiments. (G) RTPCR analysis of RUNX1, TREML1 and ITGA2 transcript levels normalized to HPRT and relative to control (CTRL) MK. CTRL MK: MK transduced with control lentivirus expressing Cherry and lentivirus expressing shSCR-GFP; shRUNX1_2: MK transduced with control lenvtivirus expressing Cherry and lentivirus expressing shRUNX1_2 and GFP; RUNX1mut/shRUNX1_2: MK transduced with lentivirus expressing RUNX1mut and Cherry and with lentivirus expressing shRUNX1_2 and GFP. RUNX1mut: WT RUNX1 cDNA was cloned into the lentivirus pRRL_EF1a_MCS/PGK-Cherry and mutation in four nucleotides, keeping the same aminoacids, was introduced to avoid recognition of the cDNA by shRUNX1_2. Data represent Mean±SD (n=2).

1246

haematologica | 2019; 104(6)

Low TLT-1 and collagen receptor α2 in FPD/AML Table 1. Features of familial platelet disorder with predisposition to acute myelogenous leukemia patients.

Pedigree/ patient AII-1 AII-2 BII-2 BIII-1 DII-1 DIII-1 DIII-3

Sex/Age Platelet count MPV ISTH-BAT (years) (x109/L) (fL) (score) M/38 F/39 F/37 M/2 F/51 M/30 F/27

90 116 130 106 145 115 94

9.6 10.6 9 10.2 11.3 10.5 10.4

1 2 1 2 5 6 4

ADP 55±7 37±2 24±4 40±10 31±2 30±3 34±9

Platelet aggregation (%) EPI COL 60±7 65±9 8±4 12±5 14±6 15±0 19±2

40±4 60±20 33±3 28±4 20±0 8±7 21±6

AA

RUNX1 mutation

RUNX1 RNA levels

60±5 70±10 53±10 28±4 29±40 4±6 34±41

p.R174Q p.R174Q p.R139X p.R139X p.T219Rfs*8 p.T219Rfs*8 p.T219Rfs*8

1.13* 0.92* 0.46* 2.66* 3.2** 2.9** 2.4**

MPV: mean platelet volume; ISTH-BAT: International Society on Thrombosis and Hemostasis-bleeding assessment tool; ADP: adenosine diphosphate;;EPI: epinephrine; COL: collagen; AA: arachidonic acid; nd: not done. Pedigrees A,2 B2 and D8,11 were previously described. Reference values for MPV, 8.9-12.5 fL. Platelet aggregation in response to 2 μM ADP, 1 μM epinephrine, 4 μg/mL collagen, and 1mM arachidonic acid was performed as described,8,9 and expressed as maximal light transmission percentage; Mean ± Standard Deviation values are shown; reference values, 75±9, 78±9.5, 77±11 and 79±9%, respectively. *Microarray data in patient megakaryocytes; **quantitative polymerase chain reaction data in patient platelets, relative to healthy subjects (n=4), set as 1.

Results Gene expression profiling of shRUNX1-transduced and familial platelet disorder with predisposition to acute myelogenous leukemia megakaryocytes reveals downregulation of TREML1 and ITGA2 To identify RUNX1-targets that could be involved in FPD/AML platelet dysfunction, we first performed transcriptome analysis of mature (CD41+CD42+) MK cultured from four patients, two carrying the R174Q mutation, and two with the R139X mutation. We then analyzed the transcriptome of MK cultured from normal leukapheresisderived CD34+ cells transduced with shRUNX1 at days 6 and 7 of culture in order to detect RUNX1 targets involved in late stages of MK differentiation and, more particularly, in proplatelet formation and platelet function. A significant increase in 43 genes and a decrease in 61 genes was shown in both FPD/AML and shRUNX1-transduced MK (Figure 1A and B and Online Supplementary Table S3A-F).

Analysis of up-regulated genes did not show any potential candidates for FPD/AML platelet dysfunction (Online Supplementary Table S4A and Online Supplementary Figure S1). In contrast, analysis of GO pathways revealed two down-regulated genes, TREML1 and ITGA2, that were of interest regarding their role in platelet biology (Figure 1C and D and Online Supplementary Table S4B).

TREML1 codes for TLT-1, which represents an immunoreceptor tyrosine-based inhibition motif (ITIM)containing receptor exclusively expressed in MK and platelets, where it is stored in α-granules.14 It undergoes surface translocation upon platelet activation14 and has been recently proposed to represent a more rapid and sensitive marker of platelet activation compared to P-selectin.15 TLT-1 ectodomain is released following platelet activation, leading to a naturally occurring soluble fragment (sTLT-1), which represents an abundant constituent of the platelet sheddome.16 Unlike other platelet ITIM receptors, the non-canonical TLT-1 has activating effects.17 Early work showed that, in a transiently transfected RBL-2H3 cell line, TLT-1 acts as a co-stimulatory receptor enhancing FceRI-mediated calcium signaling through recruitment of Src homology 2 domain-containing tyrosine phosphatase (SHP)-2 to its cytoplasmic ITIM domain.18 In human platelets, incubation with a blocking anti-TLT-1 antibody was shown to inhibit platelet aggregation triggered by thrombin,19 whereas, conversely, haematologica | 2019; 104(6)

sTLT-1 enhances platelet aggregation triggered by a variety of classic platelet agonists.17 Fibrinogen represents the only known ligand for TLT-1 and has been shown to bind both the full-length protein as well as the soluble form,17 and to favor fibrinogen deposition in vivo in a murine model of acute lung injury.20 Although the precise mechanism of TLT-1 action still has to be clearly defined, it has been proposed that, during platelet aggregation, fibrinogen is cross-linked by TLT-1, facilitating platelet-fibrinogen interactions and higher-order platelet aggregates, in concert with GPIIbIIIa.17 In addition, TLT-1 has been shown to interact through its cytoplasmic domain with ERM (ezrin/radixin/moesin) proteins, potentially linking fibrinogen to the platelet cytoskeleton.17 The essential role of TLT-1 in platelet function is revealed in Treml1-/- mice, which display mild thrombocytopenia, decreased platelet aggregation, and prolonged bleeding time.17 In addition to its role in hemostasis, sTLT-1 released by platelets reduces inflammation and organ damage during sepsis by counteracting leukocyte activation and platelet-neutrophil crosstalk.21 ITGA2 encodes the α2 subunit of collagen receptor α2b1, which, in concert with GPVI, mediates the plateletcollagen interaction at sites of vascular injury required for stable platelet adhesion.22 Integrin α2b1 is essential when platelets are exposed to monomeric collagen, whereas it plays a supportive role for fibrillar collagen, where GPVI is the central receptor. The complementary interplay between both collagen receptors is required for an optimal platelet response to collagen. Upregulation of ITGA2 was demonstrated in K562 cells transduced with a RUNX1 expression vector, suggesting RUNX1 may regulate ITGA2 expression.23 Thus, we next assessed TREML1 and ITGA2 expression profile during in vitro megakaryopoiesis and showed that TREML1 and ITGA2 mRNA levels increase during normal MK differentiation (Figure 1E). Using real-time PCR, we confirmed that TREML1 and ITGA2 mRNAs are decreased in mature MK after shRNA-mediated RUNX1 inhibition (Figure 1F). Moreover, transduction of a WT RUNX1 cDNA carrying a mutation that does not change the amino acid sequence but prevents its recognition by shRUNX1 (RUNX1mut) was able to rescue the inhibition in TREML1 and ITGA2 induced by shRUNX1 (Figure 1G), further demonstrating the relationship between RUNX1 and both TREML1 and ITGA2.

1247

A.C. Glembotsky et al.

TREML1 and ITGA2 are novel RUNX1 targets To investigate whether TREML1 and ITGA2 represent direct RUNX1 targets, we searched for RUNX1 and activating histone mark H3K27Ac enrichment across the entirety of these two genes by ChIP-sequencing in mature MK. Although no significant RUNX1 enrichment was shown in promoter regions, it was detected in intronic regions of both genes (Figure 2A and C). RUNX1 putative binding sites overlapping H3K27Ac were then identified

by in silico analysis. Using ChIP, we confirmed 3-4-fold enrichment for RUNX1 in two RUNX1 sites identified in TREML1 and more than 5-fold enrichment in four sites identified in ITGA2 (Figure 2B and D). To assess whether these sites are functional, we cloned these intragenic regulatory regions (TREML1_RR and ITGA2_RR) upstream of TREML1 and ITGA2 promoters into lentivirus allowing mCherry reporter expression under these promoters (Online Supplementary Figure S2A and B) and tested them in

A

B

C

D

E

F

Figure 2. TREML1 and ITGA2 are novel RUNX1 targets. (A and C) Illustration of chromatin immunoprecipitation (ChIP)-sequencing peaks across TREML1 and ITGA2 genes in primary megakaryocytes (MK) differentiated from CD34+ cells isolated from leukapheresis samples (RUNX1 and H3K27Ac). (B and D) ChIP-polymerase chain reaction (PCR) assays performed in primary MK confirming that RUNX1 directly binds TREML1 and ITGA2 intragenic regions. Genomic location of putative RUNX1 binding sites (RUNX1 BS) based on the University of California Santa Cruz (UCSC) database (CGCh37/hg19). ChIP experiments show RUNX1 binding in both (A) and (B) RUNX1 BS identified in the intragenic region of the TREML1 gene and in (A-D) RUNX1 BS identified in the intragenic region of the ITGA2 gene. RUNX1+23 enhancer was used as a positive control. Amplification of regions without RUNX1 putative BS was performed as a negative control. Data represent MeanÂąStandard Deviation (SD) of three independent experiments; *P<0.05; **P<0.01. (E and F) Functional analysis of identified RUNX1 BS. Intragenic TREML1 and ITGA2 regulatory regions were cloned upstream of TREML1 and ITGA2 promoters (RRwt_prom), respectively, in lentivirus allowing mCherry expression under these promoters and also encoding a PGK-GFP cassette, which allows selection of transduced MK. Mutagenesis was performed to delete all RUNX1 BS (RRmut_prom). MK were transduced on day 6 of culture and analyzed on day 10. (E) Results for TREML1. (F) Results for ITGA2. An identical gate was set for all conditions for each gene, including cells transduced with lentivirus carrying the promoter alone (right panel) or coupled with the wild-type (left) and mutated (middle) regulatory regions. For TREML1, Cherry expression was assessed in GFP-high CD41+ cells. Representative histograms are shown for one of four independent experiments. Median mCherry fluorescence intensity (MFI) was calculated relative to that obtained for ITGA2_prom (n=4) and TREML1_prom (n=4) constructs, respectively. Data represent MeanÂąSD of four independent experiments; *P<0.05; **P<0.01.

1248

haematologica | 2019; 104(6)

Low TLT-1 and collagen receptor Îą2 in FPD/AML

mature MK. Expression of mCherry fluorescent protein increased in both cases when TREML1_RR and ITGA2_RR were cloned upstream of the respective promoters and significantly decreased after specific deletion of RUNX1 binding sites (Figure 2E and F). These results clearly demonstrate that the identified sites are functional in mature MK and that TREML1 and ITGA2 are direct RUNX1 targets positively regulated by this transcription factor. mCherry reporter expression driven by TREML1_RR and ITGA2_RR harboring deleted RUNX1 sites did not reach the same level as for promoters alone, especially for ITGA2; this is probably because RUNX1 cooperates with other transcription factors. Indeed, ETS and

A

EVI1 binding sites and EVI1, SCL and ETS sites are identified in TREML1_RR and ITGA2_RR regions, respectively.

Moreover, at least one other site positive for H3K27Ac mark binds RUNX1 at position 52362118-52362307nt (Figure 2C). This site could also be involved in the regulation of ITGA2 by RUNX1.

TLT-1 is decreased in platelets and megakaryocytes in familial platelet disorder with predisposition to acute myelogenous leukemia Considering that the main role of TLT-1 is related to platelet function, we next assessed the levels in platelets.

First, we showed decreased TREML1 transcripts by qPCR

B

C

D

Figure 3. Levels of TLT-1 in familial platelet disorder with predisposition to acute myelogenous leukemia (FPD/AML) patient samples. (A) Real-time polymerase chain reaction (PCR) analysis of TREML1 transcript levels normalized to glyceraldehyde 3-phosphate dehydrogenase (GAPDH) in platelets from patients of pedigree D harboring the T219Rfs*8 mutation (n=3): DII-1 ( !), DIII-1 ( !), DIII-3 ( !), and healthy subjects (n=10); **P<0.01. (B) (Left) Western blot analysis of TLT-1 content in platelet lysates from patients (n=3) and controls (n=4). Membranes were probed with anti-human TLT-1 antibody and reprobed with anti-beta (β)-actin. (Right) Optical density (OD) measurement of TLT-1/ b-actin ratio by densitometry. (C) Anti-TLT-1 antibody-induced inhibition of platelet aggregation. Patient (n=3): DII-1 ( ! ), DIII-1 ( !), DIII-3 ( !), and control (n=4) washed platelets were incubated with an anti-TLT-1 blocking antibody (Ab) or vehicle, then challenged with thrombin and platelet aggregation was recorded in a Lumi-Aggregometer. Percentage inhibition in platelet aggregation induced by the anti-TLT-1 Ab was calculated relative to the vehicle (set as 100%). Median values and interquartile range are shown; *P<0.05. (Right) Representative examples of aggregation traces of control and patient (DII-1) platelets incubated with anti-TLT-1 Ab or vehicle and challenged with thrombin. As patient platelets had reduced response to thrombin, thrombin concentration was titrated until a response was achieved. (D) Potentiation of thrombin-induced fibrinogen binding by recombinant soluble (rs) TLT-1. Patient (n=2) and control (n=2) washed platelets were incubated with rsTLT-1 or vehicle and exposed to increasing concentrations of thrombin. Mean fluorescence intensity (MFI) is expressed as arbitrary units (MFU). Median values and range are shown. (Right) Representative histograms and corresponding MFI values.

haematologica | 2019; 104(6)

1249

A.C. Glembotsky et al.

in platelets from a third pedigree harboring the T219Rfs*8 mutation (pedigree D) (Figure 3A). Western blot analysis of platelet samples confirmed a marked decrease in TLT-1 content (Figure 3B), revealing for the first time downregulation of a member of the family of ITIM-bearing receptors in patients with RUNX1 mutations. We have previously shown that patients from this pedigree display partial deficiency of α-granules,8 where TLT-1 is located, and this finding was associated with a mild decrease in α-granule protein TSP-1.8 Heterogeneous platelet content of other α-granule proteins was demonstrated by other authors, including reduced levels of PF4, which is regulated by RUNX1,24 and preserved levels of b-TG and PDGF.25 More recently, RUNX1 deficiency has been linked to abnormal ER-to-Golgi trafficking and protein sorting leading to low von Willebrand factor (vWF) α-granule content.26 Although down-regulated TREML1 gene expression secondary to RUNX1 loss-of-function seems to be the major mechanism leading to the low TLT-1 found in this study, we cannot exclude the possibility that αgranule defects could contribute in part to the profound reduction in TLT-1 protein.

Whereas incubation of normal platelets with a blocking anti-TLT-1 antibody inhibited thrombin-induced platelet aggregation, as previously reported,19 it had no effect on thrombin-induced aggregation of FPD/AML platelets (Figure 3C), further demonstrating the absence of relevant amounts of TLT-1 in patient platelets. Conversely, consistent with the functional role of sTLT-1 in platelet activation,17 and as previously shown for platelet adhesion and actin polymerization on fibrinogen matrices,27 a recombinant soluble fragment (rsTLT-1) was able to potentiate thrombin-induced fibrinogen binding to normal platelets (Figure 3D). In the absence of rsTLT-1, thrombin-induced fibrinogen binding was lower in patients from pedigree D compared to controls, as previously reported.8 This defect was partially corrected by incubation of patient platelets with rsTLT-1 (Figure 3D), pointing to a role for TLT-1 deficiency in the platelet function defect. However, several other abnormalities are also involved, possibly contributing to the variability in platelet aggregation tests among patients (Table 1).

As shown in this work, TREML1 gene expression levels are low in normal early megakaryopoiesis and increase markedly along MK maturation reaching higher expression levels in mature MK. Double immunofluorescence labeling of TLT-1 and vWF in mature (CD41+CD42+) MK from one

control and 2 patients (DII-1 and DIII-3) confirmed that TLT-1 is mainly localized in α-granules (Figure 4), as previously reported.14 However, part of TLT-1 does not co-localize with vWF and is present in a different subpopulation of α-granules. Interestingly, a distinct staining pattern of TLT1 and P-selectin was recently shown in mouse and human platelets and mouse MK, suggesting differential compartmentalization of these proteins within α-granules,15 as previously shown for other proteins packaged in platelets.28 TLT-1-positive granules were less abundant or absent in patient compared to control MK (Figure 4).

The role of TLT-1 in normal MK is not known. TLT-1deficient mice display a 20% decrease in platelet counts,17 although the underlying mechanism has not been explored. In order to gain insight into this, we incubated human cord blood or leukapheresis-derived MK with a blocking anti-TLT-1 antibody or a control IgG. There was no significant difference in megakaryocyte output and maturation in the presence of anti-TLT-1 compared to control, whereas proplatelet formation was reduced (Online Supplementary Figure S3), suggesting TLT-1 may have a role in normal proplatelet formation. However, further study will be required to definitively establish this issue and to determine whether TLT-1 deficiency contributes to defective platelet production in FPD/AML.

Levels of integrin subunit α2 and collagen adhesion are decreased in platelets in familial platelet disorder with predisposition to acute myelogenous leukemia

Transcript levels of ITGA2, coding for the α2 integrin subunit of collagen receptor α2b1, were also shown to be decreased in platelets from pedigree D by qPCR (Figure 5A). Accordingly, surface expression of α2 (GPIa) was substantially reduced, as revealed by analysis of platelet-rich plasma (Figure 5B) and whole blood flow cytometry (Online Supplementary Table S5), and further confirmed by western blot (Figure 5C). The reduction in surface α2 was associated with a decrease in platelet surface expression of the heterodimeric b1 subunit (GPIIa), whereas GPVI, GPIIbIIIa, GPIb-IX were preserved (Table 2), indicating a selective abnormality in the α2b1 complex. The reduction in b1 is probably due to the concomitant decrease in its α2 partner, as b1 subunit (ITGB1) mRNA levels were preserved (Online Supplementary Figure S4). In addition, platelet surface levels of α5 and α6 integrin subunits, which also heterodimerize with b1, were normal or in the lower normal limit (Table 2).

Table 2. Surface platelet glycoprotein expression.

Patient AII-1 AII-2 BII-2 BIII-1 DII-1 DIII-1 DIII-3 Ref. range

GPIIa (b1)

GPVI

0.48 0.65 nd nd 0.43 0.77 0.65 0.81-1.19

nd nd nd nd 0.79 0.89 0.80 0.73-1.27

Relative fluorescence intensity (patient/control ratio) GPIIb GPIIIa GPIb GPIX 0.98 0.96 0.90 0.94 0.79 0.79 0.98 0.78-1.22

1.09 1.07 1.01 0.97 1.02 0.91 1.03 0.78-1.22

1.44 1.87 1.99 2.37 1.51 1.42 1.58 0.65-1.35

1.22 1.47 1.85 2.07 1.84 1.74 1.28 0.81-1.19

α5 integrin

α6 integrin

nd nd nd nd 0.95 1.29 0.98 0.85-1.12

nd nd nd nd 0.78 1.32 0.78 0.70-1.55

GP: glycoprotein. Ref. ; reference; nd; not done. GPIIa represents the b1 integrin subunit of the collagen receptor α2b1. Relative fluorescence intensity (RFI) was calculated as the ratio between the corresponding antibody and the isotypic control. Results are expressed as the ratio between RFI in each patient and a simultaneously assayed control sample.

The reference range was established by the Mean ± two Standard Deviations of ten healthy subjects who were simultaneously studied.

1250

haematologica | 2019; 104(6)

Low TLT-1 and collagen receptor Îą2 in FPD/AML

Figure 4. Levels of TLT-1 in megakaryocytes (MK) in familial platelet disorder with predisposition to acute myelogenous leukemia. Double immunofluorescence labeling of TLT-1 and von Willebrand factor (vWF) in mature MK grown from peripheral blood CD34+ cells from one control and 2 patients (DII-1 and DIII-3).

Representative confocal microscopy images of sorted CD41+D42+ cells stained with rabbit anti-TLT-1, mouse anti-vWF, Phallodin-Alexa 633 and 4â&#x20AC;˛,6-diamidino-2phenylindole. Anti-rabbit Alexa-488 and anti-mouse Alexa-546 were used as secondary antibodies. White arrows indicate staining for TLT-1 in granules that do not contain vWF.

haematologica | 2019; 104(6)

1251

A.C. Glembotsky et al.

Low α2b1 has been described in ANKRD26-related thrombocytopenia (RT),29 which shares several features with FPD/AML. However, whereas we confirmed that reduced α2 was restricted to some but not all ANKRD26RT patients (Online Supplementary Figure S5), all FPD/AML patients studied showed this defect. Interestingly, α2 defi-

B

C

(RFI)

A

ciency was recently reported in two other FPD/AML pedigrees.30,31 Study of a larger cohort is required to determine whether low α2 is a constant feature of FPD/AML and could be useful as a screening tool in this setting.

Consistent with the essential role of α2b1 in platelet interaction with monomeric collagen I,22,32 patient platelet

D

F

E

G

Figure 5. Levels of integrin subunit alpha (α)-2 and collagen adhesion in platelets in familial platelet disorder with predisposition to acute myelogenous leukemia (FPD/AML). (A) Real-time polymerase chain reaction (PCR) analysis of ITGA2 transcript levels normalized to GAPDH in patients from pedigree D harboring the T219Rfs*8 mutation (n=3): DII-1 ( !), DIII-1 ( !), DIII-3 ( !), and healthy subjects (n=10); **P<0.01. (B) (Left) Platelet surface levels of integrin subunit alpha (α)-2 by flow cytometry. Results are expressed as the ratio between fluorescence intensity obtained with anti-CD49b antibody and the corresponding isotype control (relative fluorescence intensity) for patients (n=4): DII-1 ( !), DIII-1 ( !), DIII-3 ( !), AII-1 ( )! and controls (n=15). Median values and interquartile range are shown; ***P<0.001. (Right) Representative histograms of integrin α2 in patient (dashed line) and control (thick continuous black line) platelets. Isotype controls in the patient and control are shown by superimposed solid gray and empty gray histograms, respectively. (C) Western blot analysis of platelet integrin subunit α2. Platelet lysates from patients (n=3) and controls (n=4) were subjected to sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE); membranes were probed with anti-human integrin α2 antibody and reprobed with anti-beta (b)-actin. (Bottom) Densitometric analysis of integrin α2/b-actin ratio. (D) Platelet adhesion to monomeric and fibrillar type I collagen in patients (n=3: DII-1; DIII-1; DIII-3) and controls (n=5). Platelets were stained with Phalloidin-fluorescein isothiocyanate (FITC) labeled peptide, and the number of adherent platelets per field at 1000x magnification was counted. Bars represent median values and interquartile range; *P<0.05. (E) Representative images of platelet adhesion to monomeric and fibrillar collagen are shown. (F and G) Platelet accumulation under flow. Platelets from controls (n=3) and patients (n=3) were perfused over fibrillar collagen I-coated coverslips through a flow chamber at 1 dyn/cm2, labeled with phalloidin-tetramethylrhodamine and platelet surface coverage was determined using Image J software. (F) Representative images of control and patient platelets obtained at 400x magnification are shown. AU: arbitrary units.

1252

haematologica | 2019; 104(6)

Low TLT-1 and collagen receptor α2 in FPD/AML

A

B

vs.

adhesion to this substrate was severely impaired (Figure 5D and E). Moreover, platelet adhesion was also, albeit less markedly, reduced over fibrillar collagen I (Figure 5D and E). Considering that GPVI, which has a more prominent role over fibrillar collagen, was preserved, other abnormalities in FPD/AML platelets may contribute to defective adhesion to fibrillar collagen in addition to low α2b1. In contrast, platelet adhesion to convulxin, which relies on GPVI, and to fibrinogen, which depends mainly on GPIIbIIIa, were largely preserved (Online Supplementary Table S6). As an approach to mimic the in vivo conditions, we studied platelet aggregate formation under flow, which revealed a substantial decrease in surface area coverage over a collagen substrate for patient compared to control platelets (Figure 5F and G). Overall, these abnormalities suggest that primary hemostasis might be impaired in FPD/AML, although the clinical relevance of this finding still has to be determined. On the other hand, defective collagen-induced aggregation (Table 1) cannot be explained only by decreased α2b1, as collagen preparations used in aggregometry depend mainly on GPVI. Given this, the aggregation defect in response to collagen and other platelet agonists may also rely on the role played by TLT-1 in fibrinogen binding, platelet aggregate formation and stabilization.

Megakaryocytes in familial platelet disorder with predisposition to acute myelogenous leukemia display low levels of α2 integrin subunit and decreased adhesion to collagen I Platelet production is a tightly regulated process governed by the close interaction between MK and bone marrow (BM) extracellular matrix (ECM) proteins. In addition to its abundance at the vascular bed, collagen I is a crucial component of the BM ECM. MK-collagen I interaction is of critical importance in restraining proplatelet formation at the osteoblastic niche, thus preventing premature platelet release into the interstitial space and allowing normal platelet production into the lumen of BM sinusoids.33 Ligation of the α2b1 receptor in MK is required for stress fiber formation and adhesion over collagen I, as shown in both human and mouse MK,33,34 whereas although α2b1 integrin is involved in collagen I-induced inhibition of proplatelet formation in human MK,33 it does not seem to be essential in mouse MK, where GPVI mediates the inhibitory signal.34 Considering the key function of α2b1 in MK behavior over collagen, we next studied MK α2 surface expression and found decreased levels on patient mature (CD41+CD42+) MK (Figure 6A). There was a trend towards reduced MK (CD41+) adhesion to fibrillar type I collagen, which was confirmed after sorting the mature (CD41+CD42+) MK population, whereas adhesion to fibrinogen was preserved (Figure 6B and C). In addition to low α2 levels, other defects in FPD/AML MK could contribute to reduced collagen adhesion. Virtual absence of proplatelet formation from patient cells2 prevented us from assessing whether physiological collagen I-inhibition in thrombopoiesis was affected. The role of α2b1 in platelet production in vivo remains controversial as, unexpectedly, Itga2-/- mice show normal platelet counts.32 Increased MK numbers found in these mice may represent a compensatory mechanism and provide an explanation for the absence of thrombocytopenia in this model. We have previously shown that FPD/AML patients show decreased MK output from hematopoietic progenitors, haematologica | 2019; 104(6)

C

Figure 6. Levels of integrin subunit alpha (α)-2 and collagen adhesion in megakaryocytes (MK) in familial platelet disorder with predisposition to acute myelogenous leukemia (FPD/AML). (A) (Left) Flow cytometry analysis of integrin α-2 expression in mature (CD41+CD42+) MK grown from peripheral blood CD34+cells from patients (n=3): DII-1 ( !), DIII-1 ( ! ), DIII-3 ( !), and healthy subjects (n=5). Median relative fluorescence intensity values and interquartile range are depicted; *P<0.05. (Right) Representative histograms of CD49b expression in patient (dashed line) and control (thick continuous black line) MK. Isotype controls in the patient and control are shown by superimposed solid gray and empty gray histograms, respectively. (B) MK adhesion to fibrillar type I collagen and fibrinogen in patients. The number of adherent cells was expressed as percentage of simultaneously assayed controls, set as 100%. (Left) Results for MK (CD41+) adhesion to collagen in patients (n=3: DII-1, DIII-1 and DIII-3). In separate experiments, CD41+CD42+ cells were sorted by flow cytometry (n=2: DIII-3 and BIII-1), allowed to adhere to collagen or fibrinogen-coated surfaces; (middle and right) results are shown. Bars represent median values and interquartile range; P=not significant. (C) Representative images of patient and control adherent mature (CD41+CD42+) MK stained with phalloidin-fluorescein isothiocyanate (FITC) and 4′,6-diamidino-2-phenylindole.

1253

A.C. Glembotsky et al. delayed MK maturation, and low ploidy levels.2 On this basis, it seems reasonable to consider the possibility that, in vivo, FPD/AML MK may not be able to fully compensate for the decrease in α2 and, in this scenario, α2 downregulation could possibly contribute to impaired platelet production in patients in addition to the above-mentioned MK abnormalities.

Considering that FPD/AML is a heterogeneous condition, study of additional pedigrees would be useful to determine whether patients harboring RUNX1 mutations different from those included in this study display similar defects.

In conclusion, gene expression analysis of RUNX1 knock-down or mutated MK proved to be a suitable approach to identify novel RUNX1 targets involved in platelet biology. TREML1 and ITGA2 may now be added to the growing list of genes regulated by RUNX1 in the megakaryocytic lineage. Down-regulated expression of these genes in patients contributes to the platelet defects induced by RUNX1 mutations. These findings highlight the key role of RUNX1 as a master regulator of the MK lineage, where it modulates the expression of a diverse array of genes crucial to platelet production and function.

They also help further unravel the molecular mechanisms underlying the FPD/AML phenotype.

References 1. Tijssen MR, Cvejic A, Joshi A, et al.

Genome-wide analysis of simultaneous GATA1/2, RUNX1, FLI1, and SCL binding in megakaryocytes identifies hematopoietic regulators. Dev Cell. 2011;20(5):597-609.

2. Bluteau D, Glembotsky AC, Raimbault A, et al. Dysmegakaryopoiesis of FPD/AML pedigrees with constitutional RUNX1 mutations is linked to myosin II deregulated expression. Blood. 2012;120(13):27082718.

3. Zang C, Luyten A, Chen J, Liu XS, Shivdasani RA. NF-E2, FLI1 and RUNX1 collaborate at areas of dynamic chromatin to activate transcription in mature mouse megakaryocytes. Sci Rep. 2016;6:30255.

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

5. Schlegelberger B, Heller PG. RUNX1 deficiency (familial platelet disorder with predisposition to myeloid leukemia, FPDMM). Semin Hematol. 2017;54(2):7580.

6. 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.

7. Antony-Debré I, Manchev VT, Balayn N, et al. Level of RUNX1 activity is critical for leukemic predisposition but not for thrombocytopenia. Blood. 2015;125(6):930-940.

8. Glembotsky AC, Bluteau D, Espasandin YR, et al. Mechanisms underlying platelet function defect in a pedigree with familial

1254

9.

10.

11.

12.

13.

14.

15.

16.

Funding This study was supported by a French grant from the Ligue National Contre le Cancer (LNCC, équipe labellisée to HR, 2013 and 2016), the European grant ERA-NET (to C. Balduini, 2013), the Argentinian grant from the National Agency for Scientific and Technological Research (to PH, 2012), a grant from the Fondation Nelia et Amadeo Barletta (to PH, 2017) and the cooperation program between France and Argentina, Ecos-Sud-Ministerio de Ciencia, Tecnología e Innovación Productiva (MINCyT) (to HR and PH, 2016). DS was supported by a postdoctoral fellowship from LNCC. The microarray was funded by the apprenticeship tax from Gustave Roussy, Villejuif, France.

Acknowledgments We thank the patients and their families for participating in this study. We thank P. Rameau for flow cell sorting and flow cytometry analysis (PFIC, Gustave Roussy, Villejuif, France), G. Meurice for transcriptome analysis (UMS AMMICA, INSERM US23/CNRS UMS S3665, Gustave Roussy, Villejuif, France), Gabriel Correa (Instituto Lanari, University of Buenos Aires, Argentina) for coagulation studies, Mirta Schattner (IMEXCONICET, Buenos Aires, Argentina) for help with the flow chamber assay and Daniela Ayala (UE IDIM-CONICET) for help with Western blot experiments.

platelet disorder with a predisposition to acute myelogenous leukemia: potential role for candidate RUNX1 targets. J Thromb Haemost. 2014;12(5):761-772.

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

Sun L, Gorospe JR, Hoffman EP, Rao AK. Decreased platelet expression of myosin regulatory light chain polypeptide (MYL9) and other genes with platelet dysfunction and CBFA2/RUNX1 mutation: insights from platelet expression profiling. J Thromb Haemost. 2007;5(1):146-154.

Heller PG, Glembotsky AC, Gandhi MJ, et al. Low Mpl receptor expression in a pedigree with familial platelet disorder with predisposition to acute myelogenous leukemia and a novel AML1 mutation.

Blood. 2005;105(12):4664-4670.

Antony-Debré I, Bluteau D, Itzykson R, et al. MYH10 protein expression in platelets as a biomarker of RUNX1 and FLI1 alterations. Blood. 2012;120(13):2719-2722.

Gilles L, Guièze R, Bluteau D, et al.

P19INK4D links endomitotic arrest and megakaryocyte maturation and is regulated by AML-1. Blood. 2008;111(8):40814091.

Washington AV, Schubert RL, Quigley L, et al. A TREM family member, TLT-1, is found exclusively in the alpha-granules of megakaryocytes and platelets. Blood. 2004; 104(4):1042-1047.

Smith CW, Raslan Z, Parfitt L, et al. TREMlike transcript 1: a more sensitive marker of platelet activation than P-selectin in humans and mice.

Blood Adv. 2018;2(16):2072-2078.

Gattis JL, Washington AV, Chisholm MM,

17.

18.

19.

20.

21.

22.

23.

24.

et al. The structure of the extracellular domain of triggering receptor expressed on myeloid cells like transcript-1 and evidence for a naturally occurring soluble fragment. J Biol Chem. 2006;281(19):13396-13403.

Washington AV, Gibot S, Acevedo I, et al.

TREM-like transcript-1 protects against inflammation-associated hemorrhage by facilitating platelet aggregation in mice and humans. J Clin Invest. 2009;119(6):14891501.

Barrow AD, Astoul E, Floto A, et al.

Cutting edge: TREM-like transcript-1, a platelet immunoreceptor tyrosine-based inhibition motif encoding costimulatory immunoreceptor that enhances, rather than inhibits, calcium signaling via SHP-2. J Immunol. 2004;172(10):5838-5842.

Giomarelli B, Washington VA, Chisholm MM, et al. Inhibition of thrombin-induced platelet aggregation using human singlechain Fv antibodies specific for TREM-like transcript-1. Thromb Haemost. 2007; 97(6):955-963.

Morales-Ortíz J, Deal V, Reyes F, et al. TLT1 is a prognostic indicator in ALI/ARDS and prevents tissue damage in the lungs in a mouse model. Blood. 2018;132(23):24952505.

Derive M, Bouazza Y, Sennoun N, et al.

Soluble TREM-like transcript-1 regulates leukocyte activation and controls microbial sepsis. J Immunol. 2012;188(11):5585-5592.

Nieswandt B, Watson SP. Platelet-collagen interaction: is GPVI the central receptor?

Blood. 2003;102(2):449-461.

Elagib KE, Racke FK, Mogass M, Khetawat R, Delehanty LL, Goldfarb AN. RUNX1 and GATA-1 coexpression and cooperation in megakaryocytic differentiation. Blood.

2003;101(11):4333-4341.

Aneja K, Jalagadugula G, Mao G, Singh A, Rao AK. Mechanism of platelet factor 4

haematologica | 2019; 104(6)

Low TLT-1 and collagen receptor Îą2 in FPD/AML

(PF4) deficiency with RUNX1 haplodeficiency: RUNX1 is a transcriptional regulator of PF4. J Thromb Haemost. 2011; 9(2):383-391.

25. Weiss HJ, Witte LD, Kaplan KL, et al.

Heterogeneity in storage pool deficiency: studies on granule-bound substances in 18 patients including variants deficient in alpha-granules, platelet factor 4, beta-thromboglobulin, and platelet-derived growth factor. Blood. 1979;54(6):1296-1319.

26. Jalagadugula G, Goldfinger LE, Mao G, Lambert MP, Rao AK. Defective RAB1Brelated megakaryocytic ER-to-Golgi transport in RUNX1 haplodeficiency: impact on von Willebrand factor. Blood Adv. 2018;2(7):797-806.

27. Morales J, Villa K, Gattis J, et al. Soluble TLT-1 modulates platelet-endothelial cell

haematologica | 2019; 104(6)

28.

29.

30.

31.

interactions and actin polymerization.

Blood Coagul Fibrinolysis. 2010;21(3):229236.

Chatterjee M, Huang Z, Zhang W, et al.

Distinct platelet packaging, release, and surface expression of proangiogenic and antiangiogenic factors on different platelet stimuli. Blood. 2011; 117(14):3907-3911.

Noris P, Perrotta S, Seri M, et al. Mutations in ANKRD26 are responsible for a frequent form of inherited thrombocytopenia: analysis of 78 patients from 21 families.

Blood. 2011;117(24):6673-6680.

Perez Botero J, Chen D, Cousin MA, et al.

Clinical characteristics and platelet phenotype in a family with RUNX1 mutated thrombocytopenia. Leuk Lymphoma.

2017;58(8):1963-1967.

De Rocco D, Melazzini F, Marconi C, et al.

Mutations of RUNX1 in families with inherited thrombocytopenia. Am J Hematol. 2017;92(6):E86-E88.

32. HoltkĂśtter O, Nieswandt B, Smyth N, et al.

Integrin alpha 2-deficient mice develop normally, are fertile, but display partially defective platelet interaction with collagen.

J Biol Chem. 2002;277(13):10789-10794.

33. Sabri S, Jandrot-Perrus M, Bertoglio J, et al.

Differential regulation of actin stress fiber assembly and proplatelet formation by alpha2beta1 integrin and GPVI in human megakaryocytes. Blood. 2004; 104(10): 3117-3125.

34. Semeniak D, Kulawig R, Stegner D, et al.

Proplatelet formation is selectively inhibited by collagen type I through Syk-independent GPVI signaling. J Cell Sci. 2016; 129(18):3473-3484.

1255

ARTICLE Ferrata Storti Foundation

Haematologica 2019 Volume 104(6):1256-1267

Platelet Biology & Its Disorders

High-throughput elucidation of thrombus formation reveals sources of platelet function variability Johanna P. van Geffen,1 Sanne L.N. Brouns,1 Joana Batista,2,3 Harriet McKinney,2,3 Carly Kempster,2,3 Magdolna Nagy,1 Suthesh Sivapalaratnam,2,4 Constance C.F.M.J. Baaten,1 Nikki Bourry,1 Mattia Frontini,2,3,5 Kerstin Jurk,6 Manuela Krause,7 Daniele Pillitteri,7 Frauke Swieringa,1 Remco Verdoold,1 Rachel Cavill,8 Marijke J. E. Kuijpers,1 Willem H. Ouwehand,2,3,5,9,10 Kate Downes2,3,9* and Johan W.M. Heemskerk1*

Department of Biochemistry, Cardiovascular Research Institute Maastricht (CARIM), Maastricht University, the Netherlands; 2Department of Haematology, University of Cambridge, Cambridge Biomedical Campus, UK; 3National Health Service Blood and Transplant (NHSBT), Cambridge Biomedical Campus, UK; 4The Royal London Haemophilia Centre, London, UK; 5BHF Centre of Excellence, Division of Cardiovascular Medicine, Cambridge University Hospitals, Cambridge Biomedical Campus, UK; 6Center for Thrombosis and Hemostasis (CTH), University Medical Center of the Johannes Gutenberg University Mainz, Germany; 7DKD Helios Klinik Wiesbaden, Germany; 8Department of Data Science & Knowledge Engineering, Faculty of Humanities and Sciences, Maastricht University, the Netherlands; 9NIHR BioResource, University of Cambridge, Cambridge Biomedical Campus, UK and 10Department of Human Genetics, The Wellcome Sanger Institute, Hinxton, Cambridge, UK 1

*KD and JWMH contributed equally to this work.

ABSTRACT

Correspondence: JOHAN W. M. HEEMSKERK jwm.heemskerk@maastrichtuniversity.nl KATE DOWNES kd286@cam.ac.uk Received: May 31, 2018.

Accepted: December 5, 2018.

Pre-published: December 13, 2018. doi:10.3324/haematol.2018.198853 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/104/6/1256 ©2019 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

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

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

1256

I

n combination with microspotting, whole-blood microfluidics can provide high-throughput information on multiple platelet functions in thrombus formation. Based on assessment of the inter- and intra-subject variability in parameters of microspot-based thrombus formation, we aimed to determine the platelet factors contributing to this variation. Blood samples from 94 genotyped healthy subjects were analyzed for conventional platelet phenotyping: i.e. hematologic parameters, platelet glycoprotein (GP) expression levels and activation markers (24 parameters). Furthermore, platelets were activated by ADP, CRP-XL or TRAP. Parallel samples were investigated for whole-blood thrombus formation (6 microspots, providing 48 parameters of adhesion, aggregation and activation). Microspots triggered platelet activation through GP Ib-V-IX, GPVI, CLEC-2 and integrins.

For most thrombus parameters, inter-subject variation was 2-4 times higher than the intra-subject variation. Principal component analyses indicated coherence between the majority of parameters for the GPVI-dependent microspots, partly linked to hematologic parameters, and glycoprotein expression levels. Prediction models identified parameters per microspot that were linked to variation in agonist-induced αIIbb3 activation and secretion. Common sequence variation of GP6 and FCER1G, associated with GPVI-induced αIIbb3 activation and secretion, affected parameters of GPVIand CLEC-2-dependent thrombus formation. Subsequent analysis of blood samples from patients with Glanzmann thrombasthenia or storage pool disease revealed thrombus signatures of aggregation-dependent parameters that were subject-dependent, but not linked to GPVI activity. Taken together, this high-throughput elucidation of thrombus formation revealed patterns of inter-subject differences in platelet function, which were partly related to GPVI-induced activation and common genetic variance linked to GPVI, but also included a distinct platelet aggregation component.

haematologica | 2019; 104(6)

High-throughput elucidation of thrombus formation

Introduction Whole-blood-based microfluidics methods, measuring thrombus formation under flow, are increasingly used as proxy measurements for in vivo models of thrombosis.

Such flow assays provide valuable mechanistic information on the consequences of loss-of-function or gain-offunction mutations of key platelet signaling proteins for arterial thrombosis and hemostasis.1-4 Until recently, collagen-coated surfaces were used for such measurements, so as to approximate common collagen-dependent models of arterial thrombosis in vivo.5,6 Recently, we showed that by using multiple, microspotted surfaces this method can be extended to multiparameter measurements of thrombus formation,7 thus elucidating platelet dysfunction in patients with a range of bleeding diatheses.7-9 Commonly, detection of an heritable or acquired platelet function impairment is made by conventional approaches, such as light transmission aggregometry,10, 11 flow cytometry,12 and the PFA-100.13 However, such tests are limited by a low throughput and requirement of relatively large volumes of blood, if multiple agonists at various concentrations need to be tested. These limitations can be overcome by higher-throughput, well-plate-based tests of platelet aggregation or flow cytometric analysis.14,15 For the multiparameter measurement of thrombus formation, however, still little is known about the detection capability to identify (small differences in) platelet phenotypes in healthy subjects and groups of patients.

Two recent whole genome association studies have identified over 640 independent single nucleotide variants

Table 1. Overview of microspot surfaces (M) and parameters (P) in flow assays; as well as platelet activation (A) markers in flow cytometry.

Microspot surface M1 M2 M3 M4 M5 M6

GPIb, GPVI, α2b1 GPIb, GPVI, α2b1 GPIb + α6b1 GPIb + GPVI, α2b1 GPIb + CLEC-2 GPIb + αIIbb3

collagen type I (VWF)* collagen type III (VWF)* VWF + laminin VWF-BP + GFOGER-(GPO)n VWF-BP + rhodocytin VWF-BP + fibrinogen

Brightfield / Fluorescence parameters P1 P2 P3 P4 P5 P6 P7 P8

Platelet receptors involved

platelet surface area coverage (%SAC) platelet aggregate (%SAC) thrombus morphological score thrombus multilayer score thrombus contraction score PS exposure (%SAC) secretion (P-selectin positive, %SAC) integrin αIIbb3 activation (%SAC)

Range 0 – 66.47 0 – 47.84 0–5 0–3 0–3 0 – 22.71 0 – 63.44 0 – 48.05

Normalized 0 – 10 0 – 10 0 – 10 0 – 10 0 – 10 0 – 10 0 – 10 0 – 10

Flow cytometry activation (A), secretion (Sec), integrin αIIbb3 activation (Int) A1 A2 A3 A4

unstimulated ADP-stimulated CRP-XL-stimulated TRAP-stimulated

*(VWF), von Willebrand factor from plasma.

haematologica | 2019; 104(6)

A1-Sec, A1-Int A2-Sec, A2-Int A3-Sec, A3-Int A4-Sec, A4-Int

that are associated with quantitative platelet traits (count, mean volume, distribution width of mean volume, and mass or crit).16,17 Several of the variants appeared to be linked to alterations in platelet activation tendency, in particular by using flow cytometric assessment of agonistinduced integrin αIIbb3 activation and P-selectin expression (measuring platelet secretion) in the genotyped individuals. Detailed studies revealed single nucleotide variants which associated with altered platelet expression levels of the collagen receptor, glycoprotein (GP)VI, and with altered GPVI-induced activation responses.18-20 In the present study we aimed to evaluate, in a cohort of genetically defined healthy subjects, the underlying reasons of inter-individual variability in a microspot-based multiparameter assay of thrombus formation. The results were therefore related to a set of 24 other variables, including hematologic parameters, platelet glycoprotein expression levels and platelet activation markers.

Prediction models were built to link specific variables.

Blood samples from patients with bleeding disorders were used to interpret the observed relations.

Methods A detailed description of the methods is available in the Online Supplement.

Blood donors and blood collection Studies were approved by the Maastricht University Medical Center Ethics Committee and the Cambridge East Research Ethics Committee (Genetic analysis of platelets in healthy individuals, REC ref 10/H0304/65). Healthy subjects (laboratory population) in cohort 1 (n=10) donated three blood samples at 2- to 4-week intervals. Genotyped healthy subjects (cohort 2, n=94) were analyzed in a period of 2 weeks. These subjects were registered in the National Institute for Health Research (NIHR) BioResource (Unicorn-2 study). Genotyping of subjects in cohort 2 was performed as described elsewhere.20 The subjects’ demographics are indicated in Online Supplementary Table S1.

Included patients had confirmed Glanzmann thrombasthenia: GT1: ITGA2B c.[2943G>A], [2943G>A], p. (=) (homozygous splice mutation); GT2: ITGA2B c.213C>G; p.P71A, c.2051T>G, p.L684A (compound heterozygous point mutations); GT3: ITGA2B, c.621C>T; p.T176I (homozygous point mutation). Two other patients had a confirmed quantitative delta-storage pool disease (reduced mepacrine capture/release): SPD1, SPD2. All patients and the three day-control subjects had normal blood cell counts (Online Supplementary Data File 1D).

High-throughput microfluidics

Glass coverslips were coated with three microspots of 0.5 μL (3 mm center-to-center distance) using a high-precision mold. Using two sets of coverslips, a total of six different microspots (M1-6) were applied for whole-blood perfusion (see Table 1).

Microfluidics assays were performed as described previously,7 with minor modifications. Details of the standard operating procedures are provided in the Online Supplement. Post-staining of thrombi was performed with FITC-labeled anti-fibrinogen monoclonal antibody (1:100, Dako, F0111, Santa Clara, CA, USA), Alexa Fluor (AF)568 annexin A5 (1:200, Molecular Probes), and AF647 anti-CD62P monoclonal antibody (1:80, Biolegend, London, UK). Representative brightfield and tri-color fluorescence images were taken with an EVOS-FL microscope (Life Technologies), equipped with GFP, RFP and Cy5 dichroic cubes 1257

J.P. Van Geffen et al.

and an Olympus UPLSAPO 60x oil-immersion objective. Images were analyzed for the parameters described in Table 1, using semiautomated scripts written in Fiji. Test variability was 5-8%, depending on the type of surface and parameter.

Platelet immunophenotyping and platelet activation by flow cytometry Platelet immunophenotyping and platelet activation were performed basically as described previously.21 The platelet activation parameters that were analyzed are presented in Table 1.

Results Multiparameter assessment of whole-blood thrombus formation on microspots under flow High-throughput microfluidics has been used for indepth characterization of platelet dysfunction in patients with bleeding disorders. The technique uses microspotcoated flow chambers for multiparameter measurement of thrombus formation during whole-blood perfusion at defined wall-shear rates.7,9 In this study, we further standardized this method for inter-subject analysis of platelet function, using blood samples from healthy subjects, with results expected to represent the normal range observed in a population. Procedures included (see also the Online Supplement and de Witt et al.22): (i) microspot coating using a high-precision mold; (ii) strictly controlled conditions of blood drawing, anticoagulation and storage; (iii) pulse-free blood perfusion through a Maastricht flow chamber at a defined shear rate; (iv) defined time protocols for rinsing, staining and capturing of brightfield and fluorescence microscopic images; (v) pre-defined scripts for consistent analysis of all image sets; (vi) a gallery of exemplary images for the scoring of thrombus parameters; and (vii) comparative analysis by trained personnel performing flow runs and blind image analysis. Out of a list of 52 different surfaces,7 we selected six microspots (M1-6) to provide the most discriminative information on small changes in thrombus formation. For each surface, eight outcome parameters (P1-8) were defined, as indicated in Table 1.

Employing this standardized procedure, after wholeblood perfusion over each of the microspots, representative images were taken to visualize and assess: platelet adhesion (P1), platelet aggregation and thrombus morphology (P2-5); platelet phosphatidylserine exposure, as a marker of procoagulant activity (P6); P-selectin expression, to measure secretion (P7); and fibrinogen binding, to report on integrin αIIbb3 activation (P8). Image sets for a typical healthy control subject are given in Figure 1A. Blood perfusion over microspot M1 (collagen type I) resulted in the formation of large thrombi with contracted aggregates of platelets, high in activation markers (phosphatidylserine exposure, P-selectin expression, αIIbb3 activation), linked to a relatively high GPVI signal.7 Microspot M2 (collagen type III) produced smaller platelet aggregates with less pronounced activation markers, corresponding to more limited GPVI signaling.

Microspot M3 [von Willebrand factor (VWF) + laminin] gave a monolayer of platelets with P-selectin expression and αIIbb3 activation, but essentially no phosphatidylserine exposure, indicative of the primary adhesive role of the laminin receptor, integrin α6b1. Microspot M4 contained a combination of collagen-derived peptides [VWF1258

BP + GFOGER-(GPO)n] giving similar thrombi as on collagen type I. Microspot M5 (VWF-BP + rhodocytin) also produced full aggregates with contraction of platelets expressing activation markers. Microspot M6 (VWF-BP + fibrinogen) triggered mainly platelet adhesion with the scattered presence of small platelet aggregates, showing limited P-selectin expression and αIIbb3 activation. The observations for M1, M4 and M5 are in agreement with the capability of GPVI and CLEC-2 adhesive surfaces to support full thrombus formation in flow assays.7 To evaluate the performance of the standardized method, we collected three different blood samples from ten healthy donors (cohort 1), and determined the coefficients of variation for each of the parameters per microspot (coated M1, M2 and M6). For the majority of the parameters, intra-assay variability of duplicate measurements was 5-8%. For microspots M1 and M2, the median intra-individual coefficients of variation over the three bleeds were 15% and 18%, respectively, which is considered acceptable for whole-blood assays (Figure 1B).

For microspot M6, a higher median intra-individual coefficient of variation of 37% was obtained, likely due to the fact that parameter values on this ‘weak’ surface are low.

This initial analysis indicated that inter-individual coefficients of variation for the microspots were about twice as high as the intra-individual coefficients of variation (Online Supplementary Table S3).

High-throughput platelet phenotyping by multiparameter assessment of thrombus formation in combination with platelet count and platelet activation markers Subsequently, thrombus formation was assessed on microspots M1-6 with blood samples from 94 genotyped healthy subjects from the NIHR BioResource (cohort 2, all blood type O). Donors of either sex had a median age of 64 years (Online Supplementary Table S1). Using the same microfluidics device, brightfield and fluorescence images were recorded and analyzed for all eight parameters (P18) per microspot. Comparison of the inter-individual coefficients of variation for the 94 samples to the intraindividual ones for a subset of ten of these samples indicated, for the parameters of microspots M1-3, approximately 2-4 times higher inter-individual coefficients of variation. This ratio was 2.5-1.5 times higher for the M46 microspots (Figure 2A). Heat mapping of the normalized values per parameter and microspot showed major differences between the 94 blood samples (Figure 2B).

Taken together, these data indicated that a substantial part of the measured values contained a subject-dependent component.

For all 94 subjects, parallel measurements were performed to assess: (i) hematologic parameters using a Sysmex XN-1000 analyzer; (ii) expression levels of platelet membrane proteins; and (iii) platelet activation tendencies by flow cytometry. Hematologic parameters included white and red blood cell counts, hematocrit, platelet count, platelet crit, and mean platelet volume.

Surface membrane proteins assessed in unstimulated platelets included major adhesive glycoprotein complexes: integrin αIIbb3 (CD41a, CD41b, CD61), GPIb-V-IX (CD42a, CD42b), integrin α2b1 (CD29, CD49b), GPIV (CD36), GPVI and the surface-expressed protein tyrosine phosphatase (CD148). Platelet activation tendency (integrin activation and α-granule secretion) was measured haematologica | 2019; 104(6)

High-throughput elucidation of thrombus formation

following stimulation via P2Y1/12 receptors (with ADP), GPVI (CRP-XL), or the PAR1 thrombin receptor (TRAP) in the presence of aspirin and apyrase (where appropriate), to suppress autocrine-dependent secondary activation.

For these platelet traits, heat mapping of the scaled values again showed major inter-individual differences (Figure 2C), similarly as previously described.23

Comparative analysis of high-throughput parameters For systematic evaluation of the eight parameters of thrombi formed on six microspots for the 94 subjects, we performed a correlational analysis across all microspot and parameter combinations. As shown in Figure 3A,B, the majority of parameters tended to correlate positively per microspot, suggesting a 'thrombus profile' represented by multiple parameters (median P values for all 6 microspots

A

B

Figure 1. Microscopic imaging of platelet thrombus formation on six different microspots and variability analysis. (A) Representative images after flow of whole blood from a representative healthy subject over series of microspots M1-6 (composition as indicated). For the brightfield images, scored values are indicated for parameters P3 (thrombus morphological score), P4 (thrombus multilayer score), and P5 (thrombus contraction score). Bars, 20 Îźm. The definition of all parameters is given in Table 1. (B) Three separate blood samples from ten healthy subjects (cohort 1), taken at intervals of 2-4 weeks, were used to assess thrombus formation on microspots M1, M2, and M6. Intra- and inter-individual coefficients of variance (CV) are plotted per microspot and parameter. PS: phosphatidylserine.

haematologica | 2019; 104(6)

1259

J.P. Van Geffen et al.

5.22 x 10-5). Furthermore, positive correlations were also present across multiple microspots. For instance, parameters P2-5 (indicative of thrombus morphology and platelet aggregation) correlated between M1 (collagen I) or M2 (collagen III) with M4 [GFOGER-(GPO)n + VWF-BP] and M5 (rhodocytin + VWF-BP), i.e. all GPVI- and CLEC-2activating surfaces (P2-5; median P values 4.39 x 10-22).

Furthermore, several parameters (P2-8) correlated between M2 (collagen III) and M3 (laminin + VWF, P=0.44-1.04 x 10-11). Typically, parameters for microspot

A

M6 (fibrinogen + VWF-BP) showed no more than poor correlation with those of other surfaces.

Multiparameter correlation and regression analysis were also performed to compare subjects’ age and sex, hematologic data, platelet glycoprotein expression levels and agonist-induced platelet activation markers by flow cytometry. The interaction matrices displayed in Figure 3C,D show the expected correlation of sex (much stronger than age) with red blood cell and platelet counts (P<0.014), but not with other traits. Subjects’ age correlated weakly with

B

C

Figure 2. Inter-subject differences in platelet thrombus formation and other platelet traits. (A) Using blood samples from 94 genotyped healthy subjects (cohort 2), parameters of thrombus formation were assessed on microspots M16. Duplicate samples were analyzed for ten of these subjects. The ratios of interversus intra-individual coefficients of variance per parameter and microspot are shown. (B) Heatmap of normalized parameters per microspot and per subject (rows). Scaling of 0-10 was performed per parameter across all surfaces. (C) Heatmap of additional platelet traits of the cohort of 94 subjects (scaling 0-10).

Hematologic variables: white blood cell count (WBC), red blood cell count (RBC), hematocrit (HCT), platelet count (PLT), platelet crit (PCT = count x size), mean platelet volume (MPV); platelet glycoprotein expression (integrin αIIbb3: CD41a, CD41b, CD61; GPIb-V-IX: CD42a, CD42b; integrin α2b1: CD29, CD49b; CD36; GPVI; CD148). In addition, activation markers of unstimulated, ADP-, CRP-XL- or TRAP-stimulated platelets (A1-4), regarding integrin αIIbb3 activation (Int) and secretion (P-selectin expression, Sec). For details of coding of microspots (M), parameters (P) and activation markers (A), see Table 1. Raw data are provided in Online Supplementary Data File 1A-C.

1260

haematologica | 2019; 104(6)

High-throughput elucidation of thrombus formation

platelet integrin activation. As required, red blood cell count versus hematocrit, and platelet count versus platelet crit (i.e. count x platelet size), were highly dependent variables (P<5.67 x 10-14). Platelet size (mean platelet volume) correlated positively with the expression levels of surface glycoproteins (P=0.03-1.92 x 10-14). In addition, the majority of surface protein expression levels correlated signifi-

cantly with the components of glycoprotein complexes, integrin αIIbb3 (CD41a, CD41b, CD61), GPIb-V-IX (CD42a, CD42b) and integrin α2b1 (CD29, CD49b) (P=0.012-6.84 x 10-11). The exception was GPVI, which showed a lower level of correlation with other glycoproteins.

Regression analysis of the markers of agonist-induced platelet activation identified associations between integrin

A

B

C

D

Figure 3. Interactions between parameters of thrombus formation, hematology, platelet surface proteins and activation markers. For 94 genotyped healthy subjects (cohort 2), multiple quantitative traits of thrombus formation, blood cell and platelet parameters, and platelet activation tendency were compared by regression analysis. For coding of microspots (M), parameters (P) and activation markers (A), see Table 1. Other abbreviations are explained in Figure 2. (A, B) Correlation matrices for parameters of thrombus formation. (A) Heat mapped –log P values, in which dark colors indicate highly significant correlations (white offset at P=0.05). (B) Heat mapped Pearson correlation coefficients R, in which dark colors indicate high positive or negative correlations. (C, D) Correlation matrices for age, sex and platelet quantitative traits (hematology variables, platelet glycoprotein expression levels and platelet activation markers). (C) Heat mapped –log P values, with colors as in panel A. (D) Heat mapped Pearson correlation coefficients R, with colors as in panel C. Full statistical data are provided in Online Supplementary Data File 2A-D.

haematologica | 2019; 104(6)

1261

J.P. Van Geffen et al.

activation and secretion in response to either ADP (A2) or CRP-XL (A3) and, to a lesser extent, in response to TRAP (A4) (Figure 3C,D). A clear relationship was present between GPVI expression level and CRP-XL-induced activation markers (P=0.010 for integrin activation; P<0.0001 for secretion). Furthermore, except for TRAP-induced secretion, all platelet activation markers were associated with glycoprotein expression levels (median P=0.010) and, to a certain extent, with platelet size. Taken together, these results pointed to a subject-dependent component of platelet size and activation, independent of the type of agonist.

Comparative analysis of parameters of thrombus formation with other platelet traits Principle component analysis (PCA), using the dataset of 94 subjects, was performed to further analyze the relation-

A

C

ships between the parameters of thrombus formation and measured platelet quantitative traits. As a way to visualize the results, the relative contributions of components 1 and 2 (C1, C2) were heat mapped for each of the microspots and parameters. The resulting heatmaps (Figure 4) can be read as indicating the parameters of thrombus formation that tend to cluster together per component (in dark color), when compared to other sets of variables.

PCA was first applied to compare parameters P1-8 for each microspot with subjects’ age and sex (Figure 4A).

Component 1 (91.1%) showed a high correlation for all microspots regarding the parameters P1 (platelet adhesion) and P2-5 (collectively reflecting platelet aggregation and thrombus morphology). Neither age nor sex contributed to this component. Component 2 (1.5%) showed a cluster of the platelet activation parameters P6-8 (procoagulant activity, secretion and integrin activation). Only

B

D

E

Figure 4. Principal component analysis to reveal correlations between variables of thrombus formation and quantitative platelet traits. Principle component analyses (PCA) of mean centered data from 94 healthy subjects (cohort 2), after univariate scaling as represented in Figure 2B,C. In order to reveal patterns of jointly contributing factors to thrombus formation, scaled data from six microspots (M1-6) and eight parameters (P1-8) were combined in a PCA with sets of other variables from the 94 subjects. Heatmaps show relative contributions of each of the parameters to the first two components, C1 and C2. For coding of microspots (M), parameters (P), and platelet activation markers (A), see Table 1. (A) PCA of the M × P matrix in combination with subjects’ age and sex. (B) PCA combined with hematologic variables. (C) PCA combined with glycoprotein surface expression levels. (D) PCA combined with integrin αIIbb3 activation and secretion markers of agonist-stimulated platelets. (E) PCA of only activation markers of agonist-stimulated platelets with glycoprotein surface expression levels. Color bars of relative contributions to C1 and C2, ranging from 0 to 1. White boxes indicate no relation, while dark red boxes indicate a large contribution for the indicated parameters. Raw data are presented in Online Supplementary Data File 2E.

1262

haematologica | 2019; 104(6)

High-throughput elucidation of thrombus formation

for microspot M3, did the subjects’ age, but not sex contribute to component 2. Together, these results underscore the earlier conclusion that parameters of platelet aggregation across microspots covariate.

Combined PCA of microspot parameters and hematologic parameters (Figure 4B) resulted in a similar configuration of component 1 (97.5%) as above, with in particular high contributions of microspots M1-6 for P2-5. Platelet count and platelet crit appeared in component 1 for microspots M1-2, while these variables contributed to component 2 (1.1%) for M3-6.

Combined PCA of microspot parameters and glycoprotein expression levels (Figure 4C) revealed a marked contribution of most CD variables (although not of CD36) to the first component (95.5%) regarding M1, M2 and M6, suggesting that these expression levels were relevant to thrombus formation on these microspots. Component 2 (0.8%) contained most of the glycoprotein expression levels, particularly on M3 and M5 spots.

The combined PCA of microspot parameters with agonist-induced platelet secretion and integrin activation (Figure 4D) revealed a component 1 (89.0%), again showing large contributions for M1-5 on parameters P1-5 (platelet adhesion, aggregation and thrombus morphology). For microspot M6, component 1 also included the platelet activation markers (A2-4, i.e., ADP, CRP-XL and TRAP). Interestingly, component 2 (1.9%) revealed large contributions of the platelet activation parameters on all other microspots M1-5. These results suggested that microspot M6 (VWF-BP + fibrinogen) is the most sensitive to variation in common agonist-induced platelet activation traits, regardless of the agonist. A final PCA, combining glycoprotein expression markers with platelet activation markers, revealed strong correlations for the variables linked to integrin αIIbb3 (Figure 4E).

Together, these results suggested that a considerable part of the variability in thrombus outcomes between the 94 subjects was linked, in a surface-dependent way, to hematologic traits (platelet count and crit), expression patterns of glycoproteins and to agonist-induced secretion and integrin αIIbb3 activation. Another part seemed to be linked, across surfaces, to platelet aggregation and thrombus morphology.

Prediction models of platelet traits contributing to variation in thrombus formation Based on the PCA, predictive models were built to identify the quantitative platelet traits that correlate with thrombus formation parameters. In a first set of partial least squares (PLS) models, we aimed to find the covariance between the individual thrombus parameters and the other platelet parameters. By predicting one variable of the thrombus microspot - parameters matrix at a time, 14 (of 48) parameters had a reduction in the mean square error of prediction, meaning that they could be predicted more accurately than just using the mean (in cross validation). Eleven of the 14 models appeared to be robust (single component, or improvement with an additional component). Most of these models captured a limited amount of the variation, although, for parameter P1, three models (M4P1, M5P1, M2P7) predicted >5% (6-11%) of the total variation (Figure 5A). This analysis explained a small part of the variance, although with a focus on platelet adhesion (P1) across the surfaces. Given the limited biological insight of this effort, additional modeling was performed. haematologica | 2019; 104(6)

Since thrombus formation on three out of the six microspots (M1, M2, M4) was GPVI-dependent, additional prediction models were made to relate the values of GPVI-induced integrin αIIbb3 activation (A3-Int) and secretion (A3-Sec) to the thrombus parameters. All models were checked by leaving-one-out cross-validations. To predict GPVI-induced integrin activation, an orthogonal PLS model was built with two components. This resulted in a b matrix for each of the microspots and parameters (Figure 5Bi). For the collagen microspots (M1-2), parameters of platelet adhesion and activation (P1, P6-7) gave a positive weight to the prediction, as well as the majority of parameters of M6 (VWF-BP + fibrinogen). Interestingly, for the strongest GPVI-dependent surfaces (M1, M4), parameters determining platelet aggregation under flow (P2-5, P8) weighed negatively on the prediction.

Concerning GPVI-induced secretion, the most suitable PLS model had a single principal component. The resulting b matrix was similar to that predicting GPVI-induced integrin activation, albeit with more negative predictive weights for some M5-6 parameters (Figure 5Bii). Taken together, this analysis of the collagen surfaces indicates a positive relationship between GPVI-induced platelet integrin activation/secretion (assessed by flow cytometry) and platelet adhesion and activation in a thrombus.

Prediction of genetic variants of platelet proteins associated with variation in thrombus formation Genome-wide association studies have identified several hundreds of genetic variants associated with quantitative platelets traits.16 Epigenetic mapping has revealed that many of these variants are regulatory and lie within superenhancer regions that are implicated in megakaryocyte differentiation and platelet production.20 For three variants associated with platelet traits, we analyzed whether these were also associated with thrombus formation parameters. Univariate linear regression analysis for the whole MP matrix was used to identify associations with the following single nucleotide variants: rs1613662 (GP6), rs3557 (FCER1G) and rs2363877 (VWF-CD9) (Figure 5C).

The single nucleotide variant rs1613662 is a non-synonymous variant in the GP6 gene, while rs3557 is located in the 3’ untranslated region of FCER1G (Fcγ receptor chain, a co-receptor of GPVI) in a megakaryocytic superenhancer region. Subjects carrying the major allele of either variant (rs1613662, AA; and rs3557, TT) have higher levels of platelet GPVI, and higher CRP-XL-induced platelet activation.20 For the present data set, regression analysis indicated inter-allelic differences, in the same direction, in thrombus formation parameters at the collagen surfaces M1-2 for P1 (platelet adhesion) with rs1613662 (Figure 5Ci, ii). In addition, for rs3557 we identified associations at M1-2 and M4, i.e., the other GPVIdependent surface, regarding P8 (integrin activation).

Unexpectedly, alleles of both variants were also associated with parameters at the CLEC-2-dependent microspot M5 (VWF-BP + rhodocytin).

The variant rs2363877 is located in a megakaryocytespecific super-enhancer site that interacts with the gene promoters of VWF and CD9. This single nucleotide variant is linked to opposite changes in expression levels of platelet-stored VWF and the surface levels of the tetraspanin CD9.20 Here an allelic association was identified at microspot M5 for secretion (P7) (Figure 5Ciii).

Trends were also seen for platelet adhesion (P1) at the 1263

J.P. Van Geffen et al.

microspots M1 and M5. Interestingly, the rs2363877 allele linked to increased thrombus parameters at M1 was also associated with high CD9 levels, but low platelet VWF levels.24 Altogether, these analyses identified novel variation in platelet thrombus parameters associated with GP6 signaling for GPVI/collagen-dependent platelet adhesion (P1) and activation (P6-8).

Thrombus signature based on parameters of platelet aggregation To further evaluate the inter-subject differences in thrombus formation, blood was obtained from three patients with Glanzmann thrombasthenia (confirmed mutations in ITG2B, defective integrin αIIbb3 expression), two patients with delta storage pool disease (reduced dense granule secretion), and three healthy day-control subjects. Application of the multiparameter test with the patients’ blood samples revealed identifiable patterns of altered thrombus formation on all microspots (raw data in

A

Data File 1D). Interestingly, the altered patterns seen with the three Glanzmann samples (Online Supplementary Figure S1A-C) reinforced the earlier PCA results that, across microspots, the values of P2-5 (and to a lesser extent P8) tended to cluster. All of these parameters relate to platelet aggregation and contraction, i.e. platelet functions known to be impaired in these patients. For the patients with storage pool disease, similar but less striking patterns of changes were observed (Online Supplementary Figure S1D,E).

Subsequently, we determined whether this cluster of parameters also provides information on the thrombus signature for healthy subjects. The summative scaled value Σ(P2-5, P8) was calculated per microspot and for all microspots. Supervised clustering, ranking the 94 subjects according to this summation, did indeed point to patterns of high and low platelet aggregation, which extended over multiple microspots (Figure 6A). For the three Glanzmann patients, these summative values were very low, as

Figure 5. Prediction models explaining variation in thrombus formation. (A) Partial least squares (PLS) models determining the covariance for each of the individual thrombus parameters (M1-M6, P1-P8) and all other platelet traits. Fourteen (from 48) parameters showed a relevant prediction, capturing 0.2-10.9% (blue color intensity) of the variation. (B) Beta matrix per individual parameter for the PLS models of GPVI-induced platelet activation (unit variance scaled, mean centered data): (i) GPVIinduced αIIbb3 activation (A3-Int, 2 components orthogonal PLS); (ii) GPVI-induced secretion (A3-Sec, 1 component PLS). Positive and negative weights are indicated by different colors. (C) Matrix of significance per parameter (quantile normalized linear regression, and likelihood ratio test per allelic score), expressed as P values, for the following genetic variants: (i) GP6, rs1613662 (AA, GA, GG; n = 63, 29, 1); (ii) FCER1G, rs3557 (TT, TG, GG; n = 67, 24, 2); (iii) VWF-CD9, rs2363877 (GG, GA, AA; n = 20, 47, 26). Significance is indicated by green color intensity and *P<0.05.

B

C

1264

haematologica | 2019; 104(6)

High-throughput elucidation of thrombus formation

expected. For the two patients with storage pool disease, these values appeared to be in the lower ranges of normal.

To further assess the subject-dependent component of this thrombus signature, we evaluated the intra- and interindividual variations in ÎŁ(P2-5, P8). This analysis indicated 2 to 5 times higher inter-individual coefficients of variation (Figure 6B). Interestingly, this measure of the thrombus signature was not related to GPVI expression of GPVIinduced platelet responses, thus pointing to other factors determining the platelet aggregation profile under flow conditions.

Discussion Multiparameter assessment of thrombus formation using microfluidic assays has proven to be relevant for the assessment of platelet dysfunction in patients with different bleeding diatheses.7-9 In the present study, we further developed this technique to assess the sources of variability in thrombus formation between individuals with normal hemostasis. We evaluated platelet quantitative traits as well as thrombus formation in a cohort of 94 genetically defined healthy subjects. Regression analysis identified an overall strong correlation between most parameters per microspot surface. When comparing different microspots, especially parameters of platelet aggregation and throm-

A

bus morphology (P2-5) correlated with each other for GPVI- and CLEC-2-activating surfaces (M1-2, M4-5).

These parameters appear to describe subject-dependent differences in thrombus formation or thrombus signature, such as was also deduced from strong alterations in these parameters in blood from patients with Glanzmann thrombasthenia. Comparative matrices of other traits did not reveal age or sex as determinants of thrombus formation, in accordance with previous analyses.20,25,26 A clear relation was however observed between GPVI expression levels and CRP-XL-induced platelet activation markers, which is in agreement with an early report.27 Several PCA were applied to compare the matrix of thrombus formation parameters with other platelet traits.

An effect of platelet count and crit was seen for thrombus formation on collagen-containing microspots (M1-2), in agreement with an earlier conclusion that platelet count is a regulatory determinant of collagen-dependent thrombus formation.9 Regression analysis confirmed the predicted associations between mean platelet volume and expression levels of most platelet membrane glycoproteins. In addition, associations were seen between platelet activation markers after ADP, CRP-XL or TRAP activation and glycoprotein expression levels.

A PCA of thrombus parameters plus glycoprotein expression levels revealed marked contributions of most CD markers to the first component (95.5%) for M1, M2

Figure 6. Identification of thrombus signatures across microspots. (A) Supervised clustering of thrombus parameters P2-5 and P8 for microspots M1-6, aligned as indicated. Ranking of data from 94 subjects (cohort 2), patients with storage pool disease (SPD1-2) or Glanzmann thrombasthenia (GT1-3) and day-controls (C1-3) was according to the sum of normalized thrombus parameters of all surfaces ÎŁ(P2-5, P8). Order of subjects (compare Figure 1B): 92, 49, 33, 4, 57, 35, 9, 19, 6, 8, 15, 24, 78, 50, 16, 42, 20, 65, C3, 72, 41, 44, 76, 71, 18, 25, 26, 52, 40, 10, C2, 51, 5, 12, 28, 29, 36, 67, 23, 32, 38, 2, 88, 17, 54, 59, 89, 81, 60, 80, 11, 82, 73, 79, 69, 55, 87, 48, 47, 66, 53, 7, 70, 93, 64, 1, 45, 46, 30, 91, 3, 94, 77, 68, 43, 34, 86, 62, 58, 37, 75, 63, 27, 39, 83, C1, 22, 85, 90, 21, 13, 31, 61, 74, 84, 56, SPD1, SPD2, GT1, GT2, GT3. Note the overall consistency of the platelet aggregation-linked parameters (P2-5, P8) per subject. (B) Intra- and inter-individual coefficients of variance (CV) of summative value ÎŁ(P25, P8) per microspot (cohort 2), together with ratios indicating high subject-dependency of this thrombus signature.

B

haematologica | 2019; 104(6)

1265

J.P. Van Geffen et al.

and M6, suggesting that these expression levels determine, at least partially, thrombus formation on these microspots. The PCA with platelet activation markers did not reveal clear links with thrombus parameters, except for the weakest surface M6. Together with the relatively high intra-subject variance of this surface, this finding suggests that M6 parameters detect small changes in the activation tendency of platelets, possibly related to the quality of the blood sample.

A PLS prediction model was developed to assess the extent to which the platelet traits can explain the measured inter-subject variation in thrombus formation.

Results revealed a limited predictability for platelet adhesion (P1) at various microspots (M4-5). Overall, 1-2% of the variance could be predicted in this way, likely because of the multivariate and multidirectional nature of these platelet traits. As a more targeted and powerful approach, regression models were built to predict the M-P matrices from subject-dependent values of GPVI-induced integrin activation and secretion. As expected, the CRP-XLinduced integrin activation and secretion measures were positively associated with GPVI-dependent microspot (M1-2) parameters of platelet adhesion and activation (P1, P7). A negatively weighted prediction was seen for parameters of M4-5 (VWF-BP + GFOGER-(GPO)n; VWF-BP + rhodocytin), possibly because of the relatively large roles of α2b1 and CLEC-2, respectively, on these surfaces.

Regarding GPVI-induced integrin activation, a positively weighted prediction was seen for thrombi on the αIIbb3dependent microspot M6, suggesting that the variation on this surface had different causes.

Regression analysis also indicated associations for single nucleotide variants that are linked to alterations in platelet size or GPVI-induced platelet activation.20,27 For two single nucleotide variants linked to GPVI expression, rs1613662 (GP6) and rs3557 (super-enhancer for FCER1G), allelic associations were identified with the GPVI-dependent microspots (M1-2, M4) for platelet adhesion (P1) and

References 1. Stegner D, Nieswandt B. Platelet receptor signaling in thrombus formation. J Mol Med (Berl). 2011;89(2):109-121.

2. Versteeg HH, Heemskerk JW, Levi M, Reitsma PH. New fundamentals in hemostasis. Physiol Rev. 2013;93(1):327-358.

3. Swieringa F, Kuijpers MJ, Heemskerk JW, van der Meijden PE. Targeting platelet receptor function in thrombus formation: the risk of bleeding. Blood Rev. 2014;28(1):9-21.

4. Von Hundelshausen P, Agten SM, Eckardt V, et al. Chemokine interactome mapping enables tailored intervention in acute and chronic inflammation. Sci Transl Med. 2017;9(384).

5. Ruggeri ZM. Platelet adhesion under flow.

Microcirculation. 2009;16:58-83.

6. Heemskerk JWM, Sakariassen KS, Zwaginga JJ, et al. Collagen surfaces to measure thrombus formation under flow: possibilities for standardization. J Thromb Haemost. 2011;9(4):856-858.

7. De Witt SM, Swieringa F, Cavill R, et al.

1266

8.

9.

10.

11.

12.

platelet activation (P8, integrin activation) parameters. In addition, single nucleotide variant associations were observed for thrombus formation on the CLEC-2-dependent microspot M5, for unclear reasons. Other authors have used microfluidics assays to show that subjectdependent differences in platelet calcium fluxes contribute to a variation in collagen-dependent thrombus formation.28 This work supports our findings of the presence of a subject-dependent factor in GPVI-induced platelet activation and thrombus formation on collagen surfaces.

In contrast, no associations were seen for the single nucleotide variant of the VWF-CD9 locus. Plasma levels of VWF are known to determine the thrombus outcome in flow assays.29 However, the variant rs2363877 only modifies platelet-accumulated VWF, rather than plasma VWF.20 A limitation of our study is that, despite the measurement of over 70 different blood and platelet traits, the number of healthy subjects was restricted to 94, thus limiting the statistical power and the precise assignment of the meaning of all individual parameters. Further work in larger cohorts of healthy subjects, and in patients with known platelet and plasma disorders, will add to the characterization of many of these parameters.

Acknowledgments Support was obtained from the Cardiovascular Center (HVC), Maastricht University Medical Center, the Center for Translational Molecular Medicine (Incoag/Mikrobat), Interreg Euregio Meuse-Rhin (Polyvalve), the National Institute for Health Research (NIHR) Cambridge Biomedical Research Centre and the National Health Service Blood and Transplant (NHSBT). We gratefully acknowledge the participation of all NIHR BioResource volunteers, and thank the NIHR BioResource center and staff for their contribution. We also thank the National Institute for Health Research and NHS Blood and Transplant for funding support.

KD is supported as a HSST trainee by Health Education England.

Funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Identification of platelet function defects by multi-parameter assessment of thrombus formation. Nat Commun. 2014;5:4257.

Mattheij NJ, Braun A, van Kruchten R, et al.

Survival protein anoctamin-6 controls multiple platelet responses including phospholipid scrambling, swelling, and protein cleavage. FASEB J. 2016;30(2):727-737.

Nagy M, Mastenbroek TG, Mattheij NJ, et al. Variable impairment of platelet functions in patients with severe, genetically linked immune deficiencies. Haematologica.

2018;103(3):540-549.

Harrison P, Mackie I, Mumford A, et al.

Guidelines for the laboratory investigation of heritable disorders of platelet function. Br J Haematol. 2011;155(1):30-44.

Dawood BB, Lowe GC, Lordkipanidze M, et al. Evaluation of participants with suspected heritable platelet function disorders including recommendation and validation of a streamlined agonist panel. Blood.

2012;120(25):5041-5049.

Goodall AH, Appleby J. Flow-cytometric analysis of platelet-membrane glycoprotein expression and platelet activation. Methods

Mol Biol. 2004;272:225-253.

13. Hayward CP, Harrison P, Cattaneo M, Ortel TL, Rao AK. Platelet function analyzer (PFA)-100 closure time in the evaluation of platelet disorders and platelet function. J Thromb Haemost. 2006;4(2):312-319.

14. Lordkipanidze M, Lowe GC, Kirkby NS, et al. Characterization of multiple platelet activation pathways in patients with bleeding as a high-throughput screening option: use of 96-well Optimul assay. Blood.

2014;123(8):e11-22.

15. Dovlatova N, Lordkipanidzé M, Lowe GC, et al. Evaluation of a whole blood remote platelet function test for the diagnosis of mild bleeding disorders. J Thromb Haemost.

2014;12(5):660-665.

16. Gieger C, Radhakrishnan A, Cvejic A, et al.

New gene functions in megakaryopoiesis and platelet formation. Nature. 2011;480 (7376):201-207.

17. Astle WJ, Elding H, Jiang T, et al. The allelic landscape of human blood cell trait variation and links to common complex disease. Cell.

2016;167(5):1415-1429.

18. Joutsi-Korhonen L, Smethurst PA, Rankin A,

haematologica | 2019; 104(6)

High-throughput elucidation of thrombus formation

et al. The low-frequency allele of the platelet collagen signaling receptor glycoprotein VI is associated with reduced functional responses and expression. Blood. 2003;101(11): 4372-4379.

19. Watkins NA, O'Connor MN, Rankin A, et al. Definition of novel GPVI polymorphisms and major difference in haplotype frequencies between populations by a combination of in-depth exon resequencing and genotyping with tag single nucleotide polymorphisms. J Thromb Haemost. 2006;4(6):11971205.

20. Petersen R, Lambourne JJ, Javierre BM, et al.

Platelet function is modified by common sequence variation in megakaryocyte super enhancers. Nat Commun. 2017;8:16058.

21. Chen L, Kostadima M, Martens JH, et al.

Transcriptional diversity during lineage commitment of human blood progenitors.

haematologica | 2019; 104(6)

Science. 2014;345(6204):1251033.

22. De Witt S, Swieringa F, Cosemans JM, Heemskerk JW. Thrombus formation on microspotted arrays of thrombogenic surfaces. Nat Protocol Exchange. 2014;2014: 3309#.

23. Garner SF, Furnell A, Kahan BC, et al. Platelet responses to agonists in a cohort of highly characterised platelet donors are consistent over time. Vox Sang. 2017;112(1):18-24.

24. Marx I, Lenting PJ, Adler T, Pendu R, Christophe OD, Denis CV. Correction of bleeding symptoms in von Willebrand factor-deficient mice by liver-expressed von Willebrand factor mutants. Arterioscler Thromb Vasc Biol. 2008;28(3):419-424.

25. Eshel-Green T, Berny MA, Conley RB, McCarty OJ. Effect of sex difference on platelet adhesion, spreading and aggregate formation under flow. Thromb Haemost.

2009;102(5):958-965.

26. Favaloro EJ, Franchini M, Lippi G. Aging hemostasis: changes to laboratory markers of hemostasis as we age. Semin Thromb Hemost. 2014;40(6):621-633.

27. Jones CI, Bray S, Garner SF, Stephens J, et al.

A functional genomics approach reveals novel quantitative trait loci associated with platelet signaling pathways. Blood.

2009;114(7):1405-1416.

28. Flamm MH, Colace TV, Chatterjee MS, et al.

Multiscale prediction of patient-specific platelet function under flow. Blood.

2012;120(1):190-198.

29. Neeves KB, Onasoga AA, Hansen RR, et al.

Sources of variability in platelet accumulation on type 1 fibrillar collagen in microfluidic flow assays. PloS one. 2013;8(1): e54680.

1267

ARTICLE Ferrata Storti Foundation

Haematologica 2019 Volume 104(6):1268-1276

Hemostasis

Generation of anti-idiotypic antibodies to detect anti-spacer antibody idiotopes in acute thrombotic thrombocytopenic purpura patients An-Sofie Schelpe,1 Elien Roose,1 Bérangère S. Joly,2,3 Inge Pareyn,1 Ilaria Mancini,4,5 Marina Biganzoli,4,5 Hans Deckmyn,1 Jan Voorberg,6 Rob Fijnheer,7 Flora Peyvandi,4,5 Simon F. De Meyer,1 Paul Coppo,8 Agnès Veyradier2,3 and Karen Vanhoorelbeke1

Laboratory for Thrombosis Research, IRF Life Sciences, KU Leuven Campus Kulak Kortrijk, Belgium; 2Service d’Hématologie Biologique, Hôpital Lariboisière, Assistance Publique-Hôpitaux de Paris, France; 3EA3518, Institut Universitaire d’Hématologie SaintLouis, Université Paris Diderot, France; 4Fondazione IRCCS Ca’ Granda Ospedale Maggiore Policlinico, Angelo Bianchi Bonomi Hemophilia and Thrombosis Center, Università degli Studi di Milano, Italy; 5Department of Pathophysiology and Transplantation, Fondazione Luigi Villa, Milan, Italy; 6Department of Molecular and Cellular Hemostasis, Sanquin-AMC Landsteiner Laboratory, Amsterdam, the Netherlands; 7Department of Internal Medicine, Meander Medical Center, Amersfoort, the Netherlands and 8Sorbonne Universités, Service d'Hématologie et Centre de Référence des Microangiopathies Thrombotiques (CNR-MAT), Hôpital Saint Antoine, Assistance Publique-Hôpitaux de Paris, France 1

ABSTRACT

I

Correspondence: KAREN VANHOORELBEKE Karen.Vanhoorelbeke@kuleuven.be Received: September 6, 2018.

Accepted: November 30, 2018.

Pre-published: December 6, 2018. doi:10.3324/haematol.2018.205666 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/104/6/1268 ©2019 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

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

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

1268

n autoantibody-mediated autoimmune diseases, autoantibody profiling allows patients to be stratified and links autoantibodies with disease severity and outcome. However, in immune-mediated thrombotic thrombocytopenic purpura (iTTP) patients, stratification according to antibody profiles and their clinical relevance has not been fully explored. We aimed to develop a new type of autoantibody profiling assay for iTTP based on the use of anti-idiotypic antibodies. Anti-idiotypic antibodies against 3 anti-spacer autoantibodies were generated in mice and were used to capture the respective anti-spacer idiotopes from 151 acute iTTP plasma samples. We next deciphered these anti-spacer idiotope profiles in iTTP patients and investigated whether these limited idiotope profiles could be linked with disease severity. We developed 3 anti-idiotypic antibodies that recognized particular idiotopes in the anti-spacer autoantibodies II-1, TTP73 or I-9, that are involved in ADAMTS13 binding; 35%, 24% and 42% of patients were positive for antibodies with the II-1, TTP73 and I-9 idiotopes, respectively. Stratifying patients according to the corresponding 8 anti-spacer idiotope profiles provided a new insight into the anti-spacer II-1, TTP73 and I-9 idiotope profiles in these patients. Finally, these limited idiotope profiles showed no association with disease severity. We successfully developed 3 anti-idiotypic antibodies that allowed us to determine the profiles of the anti-spacer II-1, TTP73 and I-9 idiotopes in iTTP patients.

Increasing the number of patients and/or future development of additional anti-idiotypic antibodies against other anti-ADAMTS13 autoantibodies might allow idiotope profiles of clinical, prognostic value to be identified.

Introduction In autoantibody-mediated autoimmune diseases, patients develop autoantibodies against self-antigens.1 The autoantibody response can be directed to multiple self-antigens like in systemic sclerosis,2 Sjögren syndrome,3 and type 1 diabetes4 or to a single self-antigen like myasthenia gravis5 and Graves disease.6 Patients suffering from the autoimmune disorder immune-mediated thrombotic thrombocytopenic purpura (iTTP) present with an autoantibody response against one antigen: the von Willebrand factor (VWF) cleaving protease ADAMTS13 (A Disintegrin And haematologica | 2019; 104(6)

Anti-spacer idiotope profiles in iTTP patients

Metalloprotease with ThromboSpondin type 1 repeats, member 13).7,8 Deficiency in ADAMTS13 leads to accumulation of hyper-active ultra-large VWF multimers that spontaneously interact with platelets. The resulting microthrombi block arterioles and capillaries, which leads to severe thrombocytopenia, hemolytic anemia and organ failure. The VWF cleaving protease ADAMTS13 consists of 14 domains: the metalloprotease (M), disintegrin-like (D), cysteine-rich (C) and spacer (S) domains, 8 thrombospondin type 1 repeats (T1-8) and 2 CUB domains.9 It is known that the anti-ADAMTS13 autoimmune response in iTTP patients is polyclonal but 80-100% of patients possess autoantibodies targeting the cysteine-rich and spacer domain.7,10-12 The standard treatment for iTTP is plasma exchange (PEX), often in combination with immunosuppressive agents (mainly steroids and rituximab).8 Recently, the anti-VWF nanobody caplacizumab, used as front-line therapy together with PEX, hastened TTP recovery, opening promising perspectives to improve the prognosis of the disease.13,14 Splenectomy is only performed in the most severe patients, when other measures have failed.8,15 Since autoimmune diseases manifest differently among patients and have a chronic course with recurring acute bouts, biomarkers are identified that allow patient stratification to predict disease outcome and prognosis and to adapt specific treatment.16 Obviously, autoantibodies are useful biomarkers in autoimmune diseases and autoantibody profiling has been shown to be valuable in stratifying patients with autoimmune disorders.17,18 On the one hand, autoantibody profiling approaches are based on the

binding of the patient autoantibodies to the disease causing antigen (recombinant proteins, fragments of these, or peptides).19,20 Whereas, on the other hand, autoantibody profiling can be performed independently of the antigen using anti-idiotypic antibodies that recognize autoantibodies that bind to the antigen (Figure 1).21 Anti-idiotypic antibodies can be generated by immunizing mice with purified or cloned antigen-binding antibodies.22-24 Antibodies that bind to particular idiotopes involved in antigen binding can then be used to detect specific autoantibodies in patient plasma or serum.21 Finally, even if the disease-causing antigen is not known, antibody profiling can lead to the identification of disease-linked peptides using next-generation sequencing25 and mass spectrometry26,27 of the total antibody response in autoimmune disease patients.

Furthermore, iTTP is a chronic disease with a variable disease outcome and risk of relapse.28 Levels of ADAMTS13 activity, anti-ADAMTS13 autoantibody subtypes, ADAMTS13 antigen levels or a combination of these have been used to identify patient groups with a worse disease outcome or a higher risk of relapse.28-35 Although the outcome of the different studies varies, it has been shown, for example, that an ADAMTS13 activity <10% during acute disease is linked with an increased risk of relapse,35 and that presenting anti-ADAMTS13 autoantibody and ADAMTS13 antigen levels predict prognosis.31 In addition, prognostic scoring systems based on clinical and/or laboratory parameters have been set up to predict severe cases and patients at risk; from 1987 with the Rose index,36,37 to more recently with the PLASMIC

Figure 1. Anti-idiotypic antibodies directed against different idiotopes in autoantibodies. A representative autoantibody is illustrated with the variable regions of heavy (VH) and light (VL) chains and the constant regions of heavy (CH) and light (CL) chains. Variable regions consist of framework regions and complementarity determining regions (CDRs).

The CDRs are unique among antibodies and consist of idiotopes that are involved in binding to the (self)-antigen (dark blue dots) and idiotopes that are not involved in binding to the (self)-antigen (light blue dots). All other regions are conserved regions (gray) between different antibodies, and make up the framework regions of the VH and VL and the constant regions of CH and CL chains. Antiidiotypic antibodies (Abs) bind to idiotopes involved in (self)-antigen binding, shown in dark blue, whereas anti-idiotypic antibodies that bind to idiotopes not involved in binding to the (self)-antigen are shown in light blue. Anti-conserved region antibodies are in gray.

haematologica | 2019; 104(6)

1269

A.S. Schelpe et al. score,38 and the score by Benhamou et al.39 The predictive model set up by Benhamou et al. takes into account age, lactate dehydrogenase (LDH) levels, and cerebral involvement, and detects early death in acquired severe ADAMTS13 deficiency-associated idiopathic TTP.39 However, in iTTP, autoantibody profiling to stratify patients has not yet been fully explored.

In this study, we developed an autoantibody profiling assay for iTTP using anti-idiotypic antibodies that recognize particular idiotopes on anti-ADAMTS13 autoantibodies, idiotopes that are involved in ADAMTS13 binding (Figure 1). Since the ADAMTS13 spacer domain seems to be the main immunogenic region targeted in these patients,29 we generated an anti-idiotypic antibody against 3 available cloned human anti-spacer autoantibodies. The selected anti-idiotypic antibodies were then used to screen 151 iTTP plasmas for the presence of autoantibodies with the same idiotopes across patients, which resulted in stratification of iTTP patients according to these anti-spacer idiotope profiles. We next investigated in a subgroup of 95 patients whether certain anti-spacer idiotope profiles could be linked with disease severity.

Patients’ samples Detailed information about the 151 iTTP plasma samples can be found in the Online Supplementary Methods. The study protocol was approved by the Medical Ethical Committee of the University Medical Center Utrecht (Utrecht, the Netherlands), the Ethics Committee of Hospital Pitié-Salpêtrière and Hospital SaintAntoine (Paris, France), and the Ethics Committee of Fondazione IRCCS Ca’ Granda Ospedale Maggiore Policlinico (Milan, Italy).

The study was carried out in accordance with the Declaration of Helsinki.

ELISA to identify the presence of anti-spacer idiotope profiles in plasmas of acute iTTP patients using the newly developed anti-idiotypic antibodies Murine anti-idiotypic antibody 17H9 (anti-II-1 antibody), 9G12 (anti-TTP73 antibody), or 7D10 (anti-I-9 antibody) were coated at 5 μg/mL. After blocking, patient plasma (starting dilution 10%, v/v) was added and diluted 1:2. Bound patient antibodies were detected with HRP-labeled anti-human IgG1-4 antibodies (Sanquin).

Statistical analysis

Methods

Graphpad Prism v.5.03 software (GraphPad Software Inc., San Diego, CA, USA) was used for statistical analysis. Further details of the methods used are available in the Online Supplementary Methods.

Immunization strategy and characterization of anti-II-1, anti-TTP73 and anti-I-9 antibodies

Results

Anti-II-1, anti-TTP73 and anti-I-9 antibodies were developed by immunizing BALB/c mice (Janvier Labs, Le Genest-Saint-Isle, France) with the cloned human anti-spacer autoantibodies II-1,40 TTP73, or I-9,41 respectively (see the immunization strategy in the Online Supplementary Appendix). The binding of purified anti-II-1, anti-TTP73 or anti-I-9 antibodies to II-1, TTP73, and I-9, respectively, and to the conserved regions (Figure 1, gray) in human immunoglobulin G (IgG) antibodies were identified using ELISA.

ELISA to identify anti-II-1, anti-TTP73 and anti-I-9 antibodies that inhibit the binding of anti-spacer autoantibodies II-1, TTP73, or I-9 to ADAMTS13, respectively Human anti-spacer autoantibodies II-1, TTP73 or I-9 (constant final EC50: 0.04, 0.85 and 0.04 μg/mL, respectively) (see Online Supplementary Methods), were pre-incubated with a 1:2 dilution series of murine anti-II-1, anti-TTP73, or anti-I-9 antibodies (final start concentration 10 μg/mL) respectively, in a pre-blocked plate.

After 30 minutes, samples were transferred to a recombinant human (rh)ADAMTS13 [2.7 μg/mL in phosphate buffered saline (PBS)] coated plate. Bound human anti-spacer autoantibodies II-1, TTP73, or I-9 were detected using a mixture of HRP-labeled antihuman IgG1-4 (IgG1: 1/20,000 and IgG2-4: 1/2,000; Sanquin, Amsterdam, the Netherlands).

ELISA to study the binding of the anti-idiotypic antibodies to the anti-spacer idiotopes of II-1, TTP73, and I-9 Murine anti-idiotypic antibodies 17H9 (anti-II-1 antibody), 9G12 (anti-TTP73 antibody) and 7D10 (anti-I-9 antibody) were coated at 5 μg/mL in carbonate/bicarbonate coating buffer (50 mM Na2CO3/NaHCO3, pH 9.6). After blocking, human anti-spacer autoantibodies II-1, TTP73, and I-9 were added at a starting concentration of 1 μg/mL and further diluted 1:2. Bound anti-spacer autoantibodies were detected by adding a mixture of HRP-labeled anti-human IgG1-4 antibodies (Sanquin).

1270

Development of anti-idiotypic antibodies against idiotopes in anti-spacer autoantibodies II-1, TTP73 or I-9 involved in ADAMTS13 binding To generate anti-idiotypic antibodies recognizing particular idiotopes in anti-spacer autoantibodies involved in ADAMTS13 binding, 3 cloned human anti-spacer autoantibodies with different epitopes and inhibitory characteristics were available: II-1,40 TTP7342 and I-941 (see Online Supplementary Methods) and were used to immunize BALB/c mice. As the injected anti-spacer autoantibodies are full IgG antibodies in which the variable regions are grafted onto a human IgG1 constant region,40,41 the mice developed antibodies that either recognized conserved regions (e.g. constant regions: CH and CL and framework regions in VH and VL) (Figure 1, gray parts) or idiotopes in the complementarity determining regions (CDRs) of the VH and VL of II-1, TTP73 and I-9 (Figure 1, dark and light blue dots). We obtained 1 mouse monoclonal antibody that recognized anti-spacer autoantibody II-1, 2 that recognized anti-spacer autoantibody TTP73 and 10 that recognized anti-spacer autoantibody I-9 (Figure 2A) as the generated antibodies bound to the coated anti-spacer autoantibodies II-1, TTP73 or I-9, respectively. To identify which of the generated monoclonal antibodies recognized the conserved part of the human autoantibodies (CH, CL and framework regions in VH and VL) (Figure 1, gray parts), their binding to a pool of purified human IgG antibodies was studied. Monoclonal antibody 17H9 recognizing antispacer autoantibody II-1 did not recognize the conserved part of the coated human IgG antibodies (Figure 2B), while 1 of the monoclonal antibodies (20H3) recognizing antispacer autoantibody TTP73 and 9 of the monoclonal antibodies (1E6, 5C8, 6C9, 7E8, 9F9, 9G9, 9H4, 11F7, and 14G6) recognizing anti-spacer autoantibody I-9 did bind to the conserved part of the coated human IgG antibodies haematologica | 2019; 104(6)

Anti-spacer idiotope profiles in iTTP patients

(Figure 2B). Therefore, monoclonal antibodies 17H9, 9G12 and 7D10 are anti-idiotypic antibodies that target idiotopes in the CDRs of VH and VL of anti-spacer autoantibody II-1, TTP73, or I-9, respectively (Figure 2B).

We next aimed to identify if the anti-idiotypic antibodies recognizing particular idiotopes in the anti-spacer autoantibodies II-1, TTP73, and I-9 are anti-idiotypic antibodies that are involved in ADAMTS13 binding (Figure 1, dark blue antibody). To do so, we used a competition ELISA where we studied if the binding of anti-spacer autoantibodies II-1, TTP73 and I-9 could be inhibited by their respective anti-idiotypic antibody. The 3 developed anti-idiotypic antibodies (17H9, 9G12 and 7D10) inhibited the binding of their respective anti-spacer autoantibodies

(II-1, TTP73 and I-9) to rhADAMTS13 (Figure 2C).

In conclusion, we developed 3 anti-idiotypic antibodies that recognize particular idiotopes in the anti-spacer autoantibodies II-1, TTP73 and I-9 that are involved in ADAMTS13 binding, as they strongly inhibit the binding of anti-spacer autoantibodies II-1, TTP73 or I-9, respectively, to rhADAMTS13.

Anti-idiotypic antibodies and their binding to idiotopes in II-1, TTP73 and I-9 Since anti-spacer autoantibodies II-1 and I-9 have overlapping but different epitopes (see Online Supplementary Methods),43 they will have both shared and unique idiotopes. We, therefore, investigated whether anti-idio-

A

B

C

Figure 2. Development and characterization of anti-idiotypic antibodies that inhibit the binding of respectively anti-spacer autoantibody II-1, TTP73 or I-9 to rhADAMTS13. (A) Binding of purified murine anti-II-1 (red), anti-TTP73 (green) and anti-I-9 (orange) antibodies to coated human anti-spacer autoantibodies II-1, TTP73 or I-9. Bound murine anti-II-1, anti-TTP73 and anti-I-9 antibodies were detected using GAM-HRP. Murine anti-II-1, anti-TTP73 or anti-I-9 antibody binding was expressed as relative absorbance values [mean ± Standard Deviation (SD), n=3] with absorbance of the respective positive controls (sera of mice immunized with either II-1, TTP73 or I-9) set as 1. (B) Binding of purified murine anti-II-1 (red), anti-TTP73 (green) and anti-I-9 (orange) antibodies to a coated human IgG pool. Bound murine anti-II-1, anti-TTP73 and anti-I-9 antibodies were detected using GAM-HRP. Murine anti-II-1, anti-TTP73 or anti-I-9 antibody binding was expressed as relative absorbance values (mean±SD, n=3) with absorbance of the respective positive controls (sera of mice immunized with either II-1, TTP73 or I-9) set as 1. Binding to coated human IgG pool indicates that the murine antibodies bind to the conserved regions of antibodies. (C) Inhibition of anti-spacer autoantibody binding to rhADAMTS13 by anti-idiotypic antibodies. A 1:2 dilution of murine anti-II-1 (red), anti-TTP73 (green) or anti-I-9 antibodies (orange) were pre-incubated with constant amounts of anti-spacer autoantibody II-1, TTP73 and I-9, respectively, before addition to a rhADAMTS13 coated 96-well microtiter plate. Bound II-1 (red), TTP73 (green) and I-9 (orange) antibodies were detected using a mixture of anti-human IgG1-4-HRP. Data are expressed as % binding (mean±SD, n=3) of anti-spacer autoantibodies II-1 (red), TTP73 (green) or I-9 (orange) to rhADAMTS13 in the presence of the competing murine anti-II-1 (17H9), anti-TTP73 (9G12) or anti-I-9 (7D10) antibody relative to the binding of II-1, TTP73 or I-9 in the absence of anti-idiotypic antibodies (dotted lines). OD: optical density.

haematologica | 2019; 104(6)

1271

A.S. Schelpe et al.

typic antibodies developed against anti-spacer autoantibody II-1 (17H9) and I-9 (7D10) recognized shared or unique idiotopes in II-1 and I-9. As a control, we included the anti-idiotypic antibody 9G12 developed against the anti-spacer autoantibody TTP73, which does not have an overlapping epitope with II-1 and I-9.

The anti-idiotypic antibody against anti-spacer autoantibody II-1 (17H9) recognized a unique idiotope in II-1 as it only captured II-1 and not anti-spacer autoantibodies I9 and TTP73 (Figure 3A). As expected, the anti-idiotypic antibody against TTP73 (9G12) also recognized a unique idiotope in anti-spacer autoantibody TTP73 as it only captured TTP73 and not anti-spacer autoantibodies II-1 and I9 (Figure 3B). In contrast, the anti-idiotypic antibody (7D10) against the anti-spacer I-9 idiotope captured both anti-spacer autoantibody I-9 and II-1 (Figure 3C) showing

that anti-idiotypic antibody 7D10 recognizes a common idiotope in II-1 and I-9.

In conclusion, these data show that the anti-idiotypic antibodies against anti-spacer autoantibody II-1 (17H9) and TTP73 (9G12) recognize a unique idiotope in II-1 and TTP73, respectively, whereas the anti-idiotypic antibody developed against anti-spacer autoantibody I-9 (7D10) recognizes an idiotope present in both anti-spacer autoantibodies II-1 and I-9 (Figure 3).

Identification of anti-spacer idiotope profiles in plasmas of acute iTTP patients using the newly developed anti-idiotypic antibodies As a first step, we screened the plasma of 151 iTTP patients for the presence or absence of the anti-spacer II-

A

B

C

1272

Figure 3. Anti-idiotypic antibodies and their binding to the anti-spacer idiotopes of II-1, TTP73 and I-9.

Binding of human anti-spacer autoantibodies (autoAbs) II-1 (red), TTP73 (green), I-9 (orange) and of a pool of human IgG antibodies (negative control, black) to coated murine anti-idiotypic antibody (Ab) 17H9 developed against II-1 (A), 9G12 developed against TTP73 (B) and 7D10 developed against I-9 (C). Bound human antispacer autoantibodies II-1, TTP73 and I-9 were detected using a antihuman IgG1-4. Data are expressed as relative absorbance values (mean+Standard Deviation, n=3) [da sottolineare il simbolo +]with absorbance of binding of II-1, TTP73 and I-9 at 1 Îźg/mL to their respective anti-idiotypic antibodies set as 1. OD: optical density.

haematologica | 2019; 104(6)

Anti-spacer idiotope profiles in iTTP patients

1, TTP-73 and I-9 idiotopes using the 3 newly developed anti-idiotypic antibodies. In a second step, we stratified the patients according to their anti-spacer idiotope profile.

The 151 iTTP plasma samples were all collected during an acute iTTP episode (see the Online Supplementary Methods for details). All patients presented with severe ADAMTS13 deficiency (< 10% activity) and detectable anti-ADAMTS13 IgG titers. Anti-ADAMTS13 IgG titers ranged from 16 to ≥100 IU/mL (median: 87 IU/mL) (Figure 4). Of the 151 iTTP patients, 34% (52 out of 151) were positive for antibodies with the anti-spacer II-1 idiotope (recognized by anti-idiotypic antibody 17H9) (Figure 4A, red dots) with median anti-spacer II-1 idiotope levels of 47 ng/mL (Figure 4A, red squares). Twenty-five percent (37 out of 151) of the patients were positive for antibodies with anti-spacer TTP73 idiotope (recognized by anti-idiotypic antibody 9G12) (Figure 4B, green dots) with median anti-spacer TTP73 idiotope levels of 174 ng/mL (Figure 4B, green squares). Forty-two percent (63 out of 151) of the patients were positive for antibodies with anti-spacer I-9 idiotope (recognized by anti-idiotypic antibody 7D10) (Figure 4C, orange dots) with median anti-spacer I-9 idiotope levels of 57 ng/mL (Figure 4C, orange squares).

We next stratified the acute iTTP patients according to their anti-spacer idiotope profile (Figure 5). The 8 possible profiles correspond to the presence of either 1, 2, 3 or none of the 3 anti-spacer idiotopes. All 8 anti-spacer idiotope profiles were identified in the iTTP patient cohort (n=151) (Figure 5). In 28% (42 out of 151) of the patients, only one particular idiotope could be detected in the plasma, with 8% (12 out of 151) having the II-1 idiotope (profile 1), 4% (6 out of 151) having the TTP73 idiotope (profile 2), and 16% (24 out of 151) having the I-9 idiotope (profile 3). In 19% (28 out of 151) of the patients, 2 idiotopes were identified in their antibody repertoire, with 5% (7 out of 151) having II-1 and TTP73 idiotopes (profile 4), 10% (15 out of 151) having II-1 and I-9 idiotopes (profile 5), and 4% (6 out of 151) having I-9 and TTP73 idiotopes (profile 6). In 12% (18 out of 151) of the patients, all 3 idiotopes were present in their antibody repertoire (profile 7). In 42% (63 out of 151) of the patients, none of the 3 idiotopes were detected (profile 8).

In conclusion, using the 3 developed anti-idiotypic antibodies, we here for the first time unraveled the specific II1, TTP73, and I-9 idiotope profiles in iTTP patients and showed that 58% of the patients had antibodies with II-1, TTP73, and I-9 idiotopes in their plasma, and this in different combinations, while 42% of the patients were negative for these idiotopes.

come was previously identified in the patients at time of diagnosis by determining a score defined by Benhamou et al.39 This score (either 1, 2, 3 or 4) is a risk score for early death in TTP based on three factors related to clinical and biological presentation (age, high LDH levels, and cerebral involvement). A score of ≥3 has a positive predictive value for mortality (patients at risk of 30-day mortality after

A

B

C

Anti-spacer idiotope profiles and their possible link with disease severity We next investigated whether the identified anti-spacer idiotope profiles (Figure 5) could be linked with disease severity, although the number of patients per profile group was rather low and we only screened for the presence or absence of 3 anti-spacer idiotopes. As a measure of disease severity, we studied disease outcome and applied treatment strategy. This part of the study was performed on the 95 patients of the French Reference Center for Thrombotic Microangiopathy, as detailed information on laboratory, clinical and outcome parameters were available for these patients (Online Supplementary Table S2).

We first analyzed whether the anti-spacer idiotope profiles could be linked with disease outcome. Disease outhaematologica | 2019; 104(6)

Figure 4. Total anti-ADAMTS13 IgG titers and anti-spacer II-1, TTP73 and I-9 idiotope concentrations in acute immune-mediated thrombotic thrombocytopenic purpura (iTTP) patients. (A-C) Total anti-ADAMTS13 IgG titers (IU/mL) were determined via TECHNOZYM® except for three samples (ID 124, 128 and 130) which were determined via an in house developed anti-ADAMTS13 IgG ELISA (§ in Online Supplementary Table S1). Anti-spacer II-1 (red square) (A), TTP73 (green square) (B), and I-9 (orange square) (C) idiotope concentrations were determined by adding patient plasma to coated murine anti-idiotypic antibody 17H9 (A), 9G12 (B) or 7D10 (C). Bound human autoantibodies were detected using a mixture of anti-human IgG1-4-HRP. A dilution series of human anti-spacer autoantibodies II-1 (A), TTP73 (B) or I-9 (C) was used as a calibration curve to determine idiotope concentrations (ng/mL). Median is represented for total anti-ADAMTS13 IgG titers and II-1, TTP73 and I-9 idiotope concentrations.

1273

A.S. Schelpe et al.

treatment initiation) and a score <3 has a negative predictive value.39 To check whether the disease outcome parameter could be linked with specific anti-spacer idiotope profiles, we performed χ2-based analysis. However, none of the anti-spacer idiotope profiles could be linked with a score of ≥3 (χ2, not significant) (Figure 6A). In line with this, there was no link between the anti-spacer idiotope profiles and the individual factors related to the score by Benhamou et al.39 (age: ANOVA, not significant; cerebral involvement and high LDH levels: χ2, not significant) (Online Supplementary Figure S1).

We next used the same approach to investigate whether anti-spacer idiotope profiles could be linked with the applied treatment strategy. We, therefore, compared the anti-spacer idiotope profiles in patients treated with PEX with/without rituximab and patients treated with PEX with/without rituximab and additional treatment(s) (either steroids or other immunosuppressive drugs, e.g. cyclophosphamide, bortezomib; or/and caplacizumab or/and splenectomy) (Online Supplementary Table S2).

However, also treatment could not be linked with antispacer idiotope profiles (χ2, not significant) (Figure 6B).

Discussion In this study, we successfully generated three anti-idiotypic antibodies that specifically recognized the idiotopes of anti-spacer autoantibodies II-1, TTP73, and I-9. With this anti-idiotypic assay, we could for the first time identify the presence or absence of anti-spacer II-1, TTP73, and I-9 idiotopes in iTTP patients. In addition, grouping the patients according to the absence or presence of one, two or three of the anti-spacer idiotopes revealed an until now unknown insight into the anti-spacer II-1, TTP73 and I-9 idiotopes in these patients. Although the resulting idiotope profiles could not be linked with disease severity, our data show that anti-idiotypic antibodies are interesting tools to determine an antibody profile in patients with any autoimmune disease.

Many studies have used groups of ADAMTS13 domains to identify which ADAMTS13 domains (e.g.

MDTCS, MDT, CS, T2-C2, T2-T8, C1-C2) are targeted by anti-ADAMTS13 autoantibodies in individual iTTP patients. All these studies concluded that the immune response in iTTP patients is polyclonal with an immunodominant epitope in the cysteine-rich/spacer domain.8,1012,29,43-45 Antibody profiling based on these data stratifies patients according to either the presence or absence of anti-ADAMTS13 antibodies against certain domain(s).

Only two studies investigated the link between domain specificity of anti-ADAMTS13 antibodies and disease severity or platelet counts. Thomas et al.29 stratified iTTP patients according to having either anti-MDTCS or antiT2-C2 autoantibodies but could not identify a link with disease severity. On the other hand, Zheng et al.10 reported an inverse correlation between the presence of IgG antibodies against the T2-T8 and/or C1-C2 domains and platelet counts on admission. In our study, we used antiidiotypic antibodies to stratify iTTP patients according to the presence or absence of anti-ADAMTS13 antibodies with specific idiotopes. By using an anti-idiotypic antibody, we can, therefore, investigate whether a specific anti-ADAMTS13 idiotope is present or absent in an individual iTTP patient. Indeed, with our three anti-idiotypic 1274

antibodies, we determined the previously unknown antispacer II-1, TTP73, and I-9 idiotope profiles in 151 iTTP patients in acute phase. Eighteen of the 151 iTTP patients had all three anti-spacer idiotopes in their plasma, 63 patients had none of the anti-spacer idiotopes, and 70 patients had either one or a combination of two of the anti-spacer idiotopes in their plasma, showing that the presence of these three anti-spacer idiotopes is not a common feature in iTTP patients. In addition, the anti-spacer autoantibody II-140 used in this study is a well characterized iTTP patient autoantibody that targets the R568F592-R660-Y661-Y665 epitope in the ADAMTS13 spacer domain43 and is a strong inhibitor of ADAMTS13 activity.40 Although approximately 50% of the iTTP patients have inhibitory anti-ADAMTS13 autoantibodies,29,46 it is still not known if all these patients have a II-1 idiotope in their plasma. Using our anti-idiotypic antibody against the anti-spacer II-1 idiotope, we now provide a novel insight into the incidence of this anti-spacer II-1 idiotope in iTTP patients. Indeed, our study showed that only 34% of the patients had this anti-spacer idiotope in their plasma. A wider understanding of the diversity of inhibitory antiADAMTS13 autoantibodies that target the R568-F592R660-Y661-Y665 epitope is important in view of the development of a targeted antibody therapy. In addition, anti-idiotypic antibodies allow the study of epitope spreading observed in iTTP patients by following the presence of specific idiotopes over time. An additional advantage of using anti-idiotypic antibodies for antibody profiling is that the antigen itself is not needed for the profiling assay.21,47 Production of recombinant ADAMTS13 and its fragments in the case of iTTP is more expensive and complex than producing and purifying murine anti-idiotypic antibodies.

Finally, we investigated whether we could establish a link between these anti-spacer idiotope profiles and disease severity (disease outcome and applied treatment strategy). However, the current idiotope profiles did not allow specific profiles that are linked with disease severity to be identified. On the one hand, this could be due to the

Figure 5. Anti-spacer idiotope profiles in acute immune-mediated thrombotic thrombocytopenic purpura (iTTP) patients. Acute iTTP patients (n=151) were stratified according to the presence (+) or absence (-) of II-1, TTP73 and I-9 idiotopes, as determined in Figure 4. n: number.

haematologica | 2019; 104(6)

Anti-spacer idiotope profiles in iTTP patients

A

B

Figure 6. Disease outcome and treatment according to the anti-spacer idiotope profiles. Stratification of the 95 acute immune-mediated thrombotic thrombocytopenic purpura (iTTP) patients of the French Reference Center for Thrombotic Microangiopathy according to the eight idiotope profiles for (A) the scoring system of Benhamou et al.39 (score < 3, white bars; score ≥ 3, black bars) and for (B) treatment with plasma exchange (PEX) with/without rituximab (white bars) or PEX with/without rituximab and additional treatment(s) (black bars). †Three patients died before treatment initiation and were excluded from the study (Online Supplementary Table S2).

relatively low number of patients per idiotope profile.

Therefore, increasing the number of patients in each idiotope profile could show a link between certain profiles and disease severity. On the other hand, although the majority of iTTP patients do have autoantibodies against the cysteine-rich/spacer domain, autoantibodies targeting other regions within or outside the cysteine-rich/spacer domain could be important, as the immune response is polyclonal. Therefore, multiple anti-idiotypic antibodies recognizing a large number of anti-ADAMTS13 autoantibodies might be needed to identify autoantibody profiles in iTTP that predict disease outcome or that are linked with treatment. We are, therefore, currently expanding our panel of anti-idiotypic antibodies with anti-idiotypic antibodies recognizing anti-ADAMTS13 autoantibodies outside the spacer domain to identify autoantibody profiles of clinical, prognostic value.

The strength of autoantibody profiling to predict disease severity and outcome in an autoimmune disorder where autoantibodies are developed against a single selfantigen has been clearly demonstrated, for example, in myasthenia gravis. Indeed, it has been shown that the

References 1. Ludwig RJ, Vanhoorelbeke K, Leypoldt F, et al. Mechanisms of Autoantibody-Induced Pathology. Front Immunol. 2017;8:603.

2. Günther J, Rademacher J, van Laar JM, Siegert E, Riemekasten G. Functional autoantibodies in systemic sclerosis. Semin Immunopathol. 2015;37(5):529-542.

3. Mavragani CP, Moutsopoulos HM. Sjögren syndrome. CMAJ. 2014;186(15):E579-586.

4. Taplin CE, Barker JM. Autoantibodies in type 1 diabetes. Autoimmunity. 2008;41 (1):11-18.

5. Luo J, Taylor P, Losen M, de Baets MH, Shelton GD, Lindstrom J. Main immuno-

haematologica | 2019; 104(6)

presence of autoantibodies against a specific epitope in AChR is linked with disease severity in these patients.48,49 Therefore, the development of anti-idiotypic antibodies against anti-ADAMTS13 autoantibodies that are linked with disease severity, outcome, and relapse remains a promising approach to personalize treatment of iTTP patients.

In conclusion, we have shown that anti-idiotypic antibodies are useful to unravel anti-spacer autoantibody specificity in iTTP patients. Moreover, this approach is broadly applicable and can, therefore, be used to perform autoantibody profiling in any antibody-mediated autoimmune disease.

Funding This paper was supported by the following grant(s): the KU Leuven grants OT/14/071 and PF/10/014, and the European Framework Program for Research and Innovation (Horizon2020 Marie Sklodowska Curie Innovative training network PROFILE grant 675746). ASS is supported by a PhD grant from the Agency Innovation and Entrepreneurship (VLAIO, www.iwt.be), Flanders, Belgium (141136).

genic region structure promotes binding of conformation-dependent myasthenia gravis autoantibodies, nicotinic acetylcholine receptor conformation maturation, and agonist sensitivity. J Neurosci. 2009;29(44): 13898-13908.

6. Chazenbalk GD, Pichurin P, Chen C-R, et al.

Thyroid-stimulating autoantibodies in Graves disease preferentially recognize the free A subunit, not the thyrotropin holoreceptor. J Clin Invest. 2002;110(2):209-217.

7. Tersteeg C, Verhenne S, Roose E, et al.

ADAMTS13 and anti-ADAMTS13 autoantibodies in thrombotic thrombocytopenic purpura - current perspectives and new treatment strategies. Expert Rev Hematol.

2016;9(2):209-221.

8. Kremer Hovinga JA, Coppo P, Lämmle B, Moake JL, Miyata T, Vanhoorelbeke K. Thrombotic thrombocytopenic purpura.

Nat Rev Dis Prim. 2017;3:17020.

9. Zheng X, Chung D, Takayama TK, Majerus EM, Sadler JE, Fujikawa K. Structure of von Willebrand Factor-cleaving Protease (ADAMTS13), a Metalloprotease Involved in Thrombotic Thrombocytopenic Purpura.

J Biol Chem. 2001;276(44):41059-41063.

10. Zheng XL, Wu HM, Shang D, et al.

Multiple domains of ADAMTS13 are targeted by autoantibodies against ADAMTS13 in patients with acquired idiopathic thrombotic thrombocytopenic

1275

A.S. Schelpe et al.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

1276

purpura. Haematologica. 2010;95(9):15551562.

Klaus C, Plaimauer B, Studt J-D, et al.

Epitope mapping of ADAMTS13 autoantibodies in acquired thrombotic thrombocytopenic purpura. Blood. 2004;103(12):45144519.

Luken BM, Turenhout EAM, Hulstein JJJ, Van Mourik JA, Fijnheer R, Voorberg J. The spacer domain of ADAMTS13 contains a major binding site for antibodies in patients with thrombotic thrombocytopenic purpura. Thromb Haemost. 2005;93(2):267-274.

Peyvandi F, Scully M, Kremer Hovinga JA, et al. Caplacizumab for Acquired Thrombotic Thrombocytopenic Purpura. N Engl J Med. 2016;374(6):511-522.

Peyvandi F, Scully M, Kremer Hovinga JA, et al. Caplacizumab reduces the frequency of major thromboembolic events, exacerbations and death in patients with acquired thrombotic thrombocytopenic purpura. J Thromb Haemost. 2017;15(7):1448-1452.

Beloncle F, Buffet M, Coindre J-P, et al.

Splenectomy and/or cyclophosphamide as salvage therapies in thrombotic thrombocytopenic purpura: the French TMA Reference Center experience. Transfusion.

2012; 52(11):2436-2444.

Purnamawati K, Ong JA, Deshpande S, et al. The Importance of Sex Stratification in Autoimmune Disease Biomarker Research: A Systematic Review. Front Immunol. 2018; 9:1208.

Damoiseaux J, Andrade LE, Fritzler MJ, Shoenfeld Y. Autoantibodies 2015: From diagnostic biomarkers toward prediction, prognosis and prevention. Autoimmun Rev. 2015;14(6):555-563.

Hueber W, Utz PJ, Steinman L, Robinson WH. Autoantibody profiling for the study and treatment of autoimmune disease.

Arthritis Res. 2002;4(5):290.

Ayoglu B, Schwenk JM, Nilsson P. Antigen arrays for profiling autoantibody repertoires. Bioanalysis. 2016;8(10):1105-1126.

Fathman CG, Soares L, Chan SM, Utz PJ. An array of possibilities for the study of autoimmunity. Nature. 2005;435(7042):605-611.

Sullivan MA, Wentworth T, Kobie JJ, Sanz I. Anti-idiotypic monobodies for immune response profiling. Methods. 2012;58(1):6268.

Sanches J de S, De Aguiar RB, Parise CB, Suzuki JM, Chammas R, de Moraes JZ. Anti-bevacizumab idiotype antibody vaccination is effective in inducing vascular endothelial growth factor-binding response, impairing tumor outgrowth. Cancer Sci. 2016;107(4):551-555.

Hu L, Liu A, Chen W, Yang H, Wang X, Chen F. A non-toxic enzyme-linked immunosorbent assay for aflatoxin B 1 using anti-idiotypic antibodies as substitutes. J Sci Food Agric. 2017;97(5):16401645.

Lubahn BC, Reisner HM. Characterization of a monoclonal anti-idiotype antibody to human anti-factor VIII antibodies. Proc Natl Acad Sci U S A. 1990;87(21):8232-8236.

Robinson WH. Sequencing the functional antibody repertoire--diagnostic and thera-

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

peutic discovery. Nat Rev Rheumatol. 2015; 11(3):171-182.

Maat P, Van Duijn M, Brouwer E, et al. Mass spectrometric detection of antigen-specific immunoglobulin peptides in paraneoplastic patient sera. J Autoimmun. 2012;38(4):354360. de Costa D, Broodman I, Calame W, et al.

Peptides from the Variable Region of Specific Antibodies Are Shared among Lung Cancer Patients. PLoS One. 2014;9(5): e96029.

Coppo P, Wolf M, Veyradier A, et al.

Prognostic value of inhibitory antiADAMTS13 antibodies in adult-acquired thrombotic thrombocytopenic purpura. Br J Haematol. 2006;132(1):66-74.

Thomas MR, de Groot R, Scully MA, Crawley JTB. Pathogenicity of AntiADAMTS13 Autoantibodies in Acquired Thrombotic Thrombocytopenic Purpura.

EBioMedicine. 2015;2(8):942-952.

Ferrari S, Scheiflinger F, Rieger M, et al.

Prognostic value of anti-ADAMTS13 antibody features (Ig isotype, titer, and inhibitory effect) in a cohort of 35 adult French patients undergoing a first episode of thrombotic microangiopathy with undetectable ADAMTS13 activity. Blood.

2007;109(7): 2815-2822.

Alwan F, Vendramin C, Vanhoorelbeke K, et al. Presenting ADAMTS13 antibody and antigen levels predict prognosis in immunemediated thrombotic thrombocytopenic purpura. Blood. 2017;130(4):466-471.

Yang S, Jin M, Lin S, Cataland S, Wu H. ADAMTS13 activity and antigen during therapy and follow-up of patients with idiopathic thrombotic thrombocytopenic purpura: correlation with clinical outcome.

Haematologica. 2011;96(10):1521-1527.

Vesely SK, George JN, Lämmle B, et al.

ADAMTS13 activity in thrombotic thrombocytopenic purpura--hemolytic uremic syndrome: relation to presenting features and clinical outcomes in a prospective cohort of 142 patients. Blood. 2003;102 (1):60-68.

Peyvandi F, Lavoretano S, Palla R, et al.

ADAMTS13 and anti-ADAMTS13 antibodies as markers for recurrence of acquired thrombotic thrombocytopenic purpura during remission. Haematologica. 2008; 93(2):232-239.

Kremer Hovinga JA, Vesely SK, Terrell DR, Lämmle B, George JN. Survival and relapse in patients with thrombotic thrombocytopenic purpura. Blood. 2010;115(8):150011; quiz 1662.

Rose M, Eldor A. High incidence of relapses in thrombotic thrombocytopenic purpura: Clinical Study of 38 Patients. Am J Med. 1987;83(3):437-444.

Wyllie BF, Garg AX, Macnab J. Thrombotic thrombocytopenic purpura/haemolytic uraemic syndrome: a new index predicting response to plasma exchange. Br J Haematol. 2006;132(2):204-209.

Bendapudi PK, Hurwitz S, Fry A, et al.

Derivation and external validation of the PLASMIC score for rapid assessment of adults with thrombotic microangiopathies:

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

a cohort study. Lancet Haematol. 2017;4(4): e157-e164.

Benhamou Y, Assie C, Boelle P-Y, et al.

Development and validation of a predictive model for death in acquired severe ADAMTS13 deficiency-associated idiopathic thrombotic thrombocytopenic purpura: the French TMA Reference Center experience.

Haematologica.

2012;97(8):1181-1186.

Pos W, Luken BM, Kremer Hovinga JA, et al. VH1-69 germline encoded antibodies directed towards ADAMTS13 in patients with acquired thrombotic thrombocytopenic purpura. J Thromb Haemost.

2009;7(3):421-428.

Luken BM, Kaijen PHP, Turenhout EAM, et al. Multiple B-cell clones producing antibodies directed to the spacer and disintegrin/thrombospondin type-1 repeat 1 (TSP1) of ADAMTS13 in a patient with acquired thrombotic thrombocytopenic purpura.

J Thromb Haemost.

2006;4(11):2355-2364.

Roose E, Vidarsson G, Kangro K, et al. AntiADAMTS13 Autoantibodies against Cryptic Epitopes in Immune-Mediated Thrombotic Thrombocytopenic Purpura.

Thromb Haemost. 2018;118(10):1729-1742.

Pos W, Sorvillo N, Fijnheer R, et al. Residues arg568 and phe592 contribute to an. antigenic surface for anti-adamts13 antibodies in the spacer domain. Haematologica. 2011; 96(11):1670-1677.

Yamaguchi Y, Moriki T, Igari A, et al.

Epitope analysis of autoantibodies to ADAMTS13 in patients with acquired thrombotic thrombocytopenic purpura.

Thromb Res. 2011;128(2):169-173.

Grillberger R, Casina VC, Turecek PL, Zheng XL, Rottensteiner H, Scheiflinger F. Anti-ADAMTS13 IgG autoantibodies present in healthy individuals share linear epitopes with those in patients with thrombotic thrombocytopenic purpura.

Haematologica. 2014;99(4):e58-60.

Mariotte E, Azoulay E, Galicier L, et al.

Epidemiology and pathophysiology of adulthood-onset thrombotic microangiopathy with severe ADAMTS13 deficiency (thrombotic thrombocytopenic purpura): a cross-sectional analysis of the French national registry for thrombotic microangiopathy. Lancet Haematol. 2016;3(5):e237245.

Bulashev AK, Borovikov SN, Serikova SS, Suranshiev ZA, Kiyan VS, Eskendirova SZ. Development of an ELISA using anti-idiotypic antibody for diagnosis of opisthorchiasis. Folia Parasitol (Praha). 2016;63.

Masuda T, Motomura M, Utsugisawa K, et al. Antibodies against the main immunogenic region of the acetylcholine receptor correlate with disease severity in myasthenia gravis. J Neurol Neurosurg Psychiatry.

2012;83(9):935-940.

Huijbers MG, Lipka AF, Plomp JJ, Niks EH, van der Maarel SM, Verschuuren JJ. Pathogenic immune mechanisms at the neuromuscular synapse: the role of specific antibody-binding epitopes in myasthenia gravis. J Intern Med. 2014;275(1):12-26.

haematologica | 2019; 104(6)

ARTICLE

Coagulation & Its Disorders

The Khorana score for prediction of venous thromboembolism in cancer patients: a systematic review and meta-analysis

Ferrata Storti Foundation

Frits I. Mulder,1,2 Matteo Candeloro,3 Pieter W. Kamphuisen,1 Marcello Di Nisio,3 Patrick M. Bossuyt,2 Noori Guman,2 Kirsten Smit,2 Harry R. Büller,2 and Nick van Es2 on behalf of the CAT-prediction collaborators

Tergooi Hospitals, Department of Internal Medicine, Hilversum, the Netherlands; Amsterdam UMC, University of Amsterdam, Department of Vascular Medicine, Amsterdam Cardiovascular Science, Amsterdam, the Netherlands; 3University G. D’Annunzio, Department of Medicine and Ageing Sciences, Chieti, Italy

1 2

Haematologica 2019 Volume 104(6):1277-1287

ABSTRACT

W

e aimed to evaluate the performance of the Khorana score in predicting venous thromboembolic events in ambulatory cancer patients. Embase and MEDLINE were searched from January 2008 to June 2018 for studies which evaluated the Khorana score. Two authors independently screened studies for eligibility, extracted data, and assessed risk of bias. Additional data on the 6-month incidence of venous thromboembolism were sought by contacting corresponding authors. The incidence in each Khorana score risk group was estimated with random effects meta-analysis. A total of 45 articles and eight abstracts were included, comprising 55 cohorts enrolling 34,555 ambulatory cancer patients. For 27,849 patients (81%), 6-month follow-up data were obtained. Overall, 19% of patients had a Khorana score of 0 points, 64% a score of 1 or 2 points, and 17% a score of 3 or more points. The incidence of venous thromboembolism in the first six months was 5.0% (95%CI: 3.9-6.5) in patients with a low-risk Khorana score (0 points), 6.6% (95%CI: 5.6-7.7) in those with an intermediate-risk Khorana score (1 or 2 points), and 11.0% (95%CI: 8.8-13.8) in those with a high-risk Khorana score (3 points or higher). Of the patients with venous thromboembolism in the first six months, 23.4% (95%CI: 18.4-29.4) had been classified as high risk according to the Khorana score. In conclusion, the Khorana score can be used to select ambulatory cancer patients at high risk of venous thromboembolism for thromboprophylaxis; however, most events occur outside this high-risk group.

Introduction Venous thromboembolism (VTE) is a burdensome and frequent complication in patients with active cancer. The estimated overall 12-month incidence is approximately 6-8% but varies widely across tumor types.1,2 VTE is associated with substantial morbidity and mortality,3 decreases quality of life,4 and can lead to interruption or discontinuation of cancer treatment. Although thromboprophylaxis effectively reduces the risk of VTE,5 current guidelines recommend against its routine use in ambulatory cancer patients, probably due to the high number that require treatment, the fear of bleeding, and the considerable burden associated with daily injections of low-molecular-weight heparins.6 Risk stratification tools may help to reduce the number requiring treatment by guiding selection of cancer patients at high risk of VTE. An ideal risk score would help clinicians identify both patients with a negligible risk as well as those at very high risk needing intervention. The best-known risk stratification tool is the Khorana score, which was introduced in 2008. This score assigns points to five clinical and pre-chemotherapy laboratory parameters: primary tumor site (+1 or 2 haematologica | 2019; 104(6)

Correspondence: FRITS I. MULDER f.i.mulder@amc.nl Received: October 12, 2018.

Accepted: January 2, 2019.

Pre-published: January 3, 2019. doi:10.3324/haematol.2018.209114 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/104/6/1277 ©2019 Ferrata Storti Foundation Material published in Haematologica is covered by copyright.

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

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

1277

F.I. Mulder et al. points), platelet count of 350x109/L or more (+1 point), hemoglobin concentration of 100 g/L or lower or use of erythropoiesis-stimulating agents (+1 point), leukocyte count of 11x109/L or higher (+1 point), and a Body Mass Index of 35 kg/m2 or higher (+1 point) (Table 1).7 A sum score of 0 points classifies patients as being at low risk of VTE, 1 or 2 points at intermediate risk, and those with 3 or more points at high risk. The Khorana score is endorsed by the latest guideline updates of the American Society of Clinical Oncology and the National Comprehensive Cancer Network to select ambulatory cancer patients for thromboprophylaxis.6,8 Over 50 studies have evaluated the score since its publication, but reported results were often conflicting. A clear interpretation of these findings is further hampered by the substantial variation in study design, cancer types included, and duration of follow up, ranging from a median of 2 to 79 months.9,10 To obtain valid and interpretable summary estimates of the performance of the Khorana score, based on the evidence available, we performed a systematic review and meta-analysis, specifically focusing on 6-month follow-up outcomes of all published relevant studies by obtaining additional data, thereby minimizing between-study heterogeneity. Our findings provide physicians with clinically useful data on the absolute risks of VTE associated with a low-, intermediate-, and high-risk Khorana score in ambulatory patients with cancer.

Methods This report adheres to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidance (See checklist in Online Supplementary Table S1).11

Search strategy and data collection A comprehensive search was performed in Embase and MEDLINE from January 2008 to June 2018 to identify studies that had evaluated the Khorana score in ambulatory cancer patients. In addition, studies presented as abstracts at conferences of the American Society of Hematology (ASH) or the International Society on Thrombosis and Haemostasis (ISTH) were identified by a manual search. Two reviewers (FIM and MC) independently screened studies and assessed bias with the Quality in Prognosis Studies (QUIPS) tool.12 The search strategy is shown in Online Supplementary Table S2, and a full explanation of study selection, data extraction, and bias assessment is provided in Online Supplementary list 1.

Table 1. Khorana risk score.

Patients’ characteristics

Risk score

Site of cancer Very high risk (stomach, pancreas) High risk (lung, lymphoma, gynecological, bladder, or testicular) Prechemotherapy platelet count ≥350 x 109/L Prechemotherapy hemoglobin level <100 g/L or use of red cell growth factors Prechemotherapy leukocyte count >11 x 109/L Body Mass Index ≥35 kg/m2

2 1 1 1 1 1

or high (3 or more points) Khorana score. VTE was defined as the composite of radiologically confirmed symptomatic or incidental distal or proximal lower-extremity deep-vein thrombosis, upperextremity deep-vein thrombosis, or pulmonary embolism. Studies with a fixed follow-up time less than six months in their study design were not included in the analysis of the 6-month outcomes.

The derivation cohort of the Khorana score was excluded from analysis.7 As currently ongoing clinical trials (clinicaltrials.gov identifier: 02048865 and 02555878) select patients with a score of 2 or more for thromboprophylaxis; the primary outcome was also assessed for this alternative positivity threshold. Secondary outcome measures included the proportion of patients with VTE during overall follow up, the proportion of VTE occurring in the highrisk group, and the relative risk of VTE for patients with a highrisk score (≥3 points) versus those with a low-to-intermediate risk score (0-2 points) in the first six months and during complete follow up. A sensitivity analysis was performed restricted to studies not judged to be at high risk of bias in any of the domains.

A random effects model with logit transformation and inverse variance weighting was used to calculate summary estimates.

Forest plots are presented with back-transformed study-specific estimates and corresponding 95% confidence and prediction intervals. Between-study heterogeneity was assessed by calculating tau-squared (τ2) using restricted maximum likelihood estimation.

Differences between subgroups were tested for significance with a χ2 test. P<0.05 was considered statistically significant.

Publication bias was explored with a funnel plot using the relative risk between high- and low-to-intermediate risk patients on the xaxis.13 Analyses were performed with R computing software, version 3.4.3 (R Foundation for Statistical Computing, Vienna, Austria; www.r-project.org), in particular using the meta package version 4.9-0.

Results

Additional data

Search results

Because the number of events are expected to increase with the duration of follow up, we evaluated the incidence of VTE during a pre-specified follow-up duration to minimize between-study heterogeneity in observation time. Since the majority of venous thromboembolic events occur in the first six months after start of chemotherapy,1 this 6-month follow-up period was considered most relevant. Corresponding authors of included studies not reporting the 6-month period were contacted and invited to provide additional data for this period.

The database and manual search yielded 1,826 unique articles and 53 abstracts, of which 1,641 were excluded on the basis of title and abstract (Figure 1). Another 50 studies were excluded after full-text assessment because the Khorana score was not reported (n=31), VTE incidence was not reported (n=6), the study population only comprised patients with VTE (n=6), the cohort was a duplicate report (n=5), or the study had a case-control design (n=2).

A total of 45 articles and eight abstracts were included in the analysis, comprising 55 cohorts and 34,555 ambulatory cancer patients, of whom 2,386 (6.9%) were diagnosed with VTE during follow up. Most studies included patients with various tumors (n=22; 42%), while others had confined recruitment to patients with gastrointestinal

Statistical analysis The primary outcome measure was the proportion of cancer patients who developed VTE during the first six months of study follow up in those with a low (0 points), intermediate (1-2 points), 1278

haematologica | 2019; 104(6)

Predicting cancer-associated venous thromboembolism

Table 2. Studies with relevant characteristics.

Author (year)

Type

Abdel-Razeq (2017)25 Article 26 Ades (2015) Article Austin (2017)27 Abstract Ayyappan (2016)28 Abstract Bezan-Graz (2017)10 Article Bezan-Zurich (2017)10 Article 29 Borchmann (2016) Abstract Cella (2017)30 Article Ferroni (2015)31 Article Ferroni (2012)32 Article 33 Fuentes (2017) Article George (2011)34,35 Article Guadagni (2017)36 Article Kearney (2009)37 Abstract Khorana (2017)38 Article 39 Khorana (2014) Article Khorana-cohort 2 (2008)7 Article Kim (2012)40 Article Kruger (2017)41 Article Kuderer (2017)42 Article 43 Kuk (2017) Article Kunapareddy (2017)44 Abstract Lim (2015)45 Article Lubberts (2016)46 Article Lustig (2015)47 Article Mandala (2012)9 Article Mansfield (2016)48 Article Misch (2013)49 Article Moore (2011)50 Article Munoz-Martin (2018)51 Article 52 Munoz-Martin (2014) Article Noble (2015)53,54 Article Panizo (2015)55 Article Papaxoinis56 Article Park (2017)57 Article Patel (2015)58 Article Pelzer (2013)24,59 Article Petitto (2017)60 Abstract Posch (2016)61 Article 62 Ramos (2016) Article Ruch (2012)63 Abstract Rupa-Matysek (2018)64 Article Rupa-Matysek (2018)65 Article Santi (2017)66 Article Sohal (2016)67 Abstract Srikanthan cohort 1 (2015)68 Article Srikanthan cohort 2 (2015)68 Article

Study Newly VTE Cancer Median designâ&#x20AC;¡ diagnosed screening type follow up cancer only before duration study start (months) Retrospective Retrospective Retrospective Retrospective Retrospective Retrospective Retrospective Prospective Prospective Retrospective Retrospective Prospective Retrospective Retrospective Prospective Prospective Prospective Retrospective Retrospective Prospective Retrospective Prospective Retrospective Prospective Prospective Prospective Retrospective Retrospective Retrospective Prospective Retrospective Prospective Prospective Retrospective Prospective Prospective Prospective Prospective Prospective Retrospective Retrospective Retrospective Retrospective Prospective Prospective Retrospective Retrospective

No No NR Yes NR NR No NR Yes Yes Yes NR Yes NR No Yes No Yes No NR Yes NR Yes Yes No No NR No NR NR No Yes No Yes Yes Yes Yes Yes No Yes NR Yes Yes Yes NR Yes Yes

No Various 40 No Colorectal 27.5 No Various 12 No DLBCL 46 No Testicular 79.2 No Testicular NR No HL 12 Yes Various 8.3 No Various 9.2 No Lung 6.9 No Gastric 21.3 NR Various 6 No Gastrointestinal 11.0 No Various NR Yes Various 2.8 Yes Various 3.7 No Various 2.4 No Various 18.9 No Pancreatic 9.2 No Lung 6.0 No Ovarian NR No Various 7.9 No DLBCL 41.9 No Testicular 33.0 No Various 3.0 No Various 2.0 No Lung 15.2 No Glioma NR No Various NR No Various 6.0 No Pancreatic 9.5 NR Lung 6,0 No Various 3.0 No Gastrointestinal 43.0 No Gastric 10.8 No Prostate 24.0 No Pancreatic 12.0 No Various 6.0 No Various 24.0 No Urothelial 8.6 No Pancreatic 8.8 No DLBCL 37.0 No Lung 14.0 No NHL 6.0 No Colorectal 6.0 No Testicular NR No Testicular NR

Study population*

Total follow up: patients with VTE, n (%)

First 6 months: patients with VTE, n (%)

1,677 151 740 241 586 303 5,409 827 810 108 108 1,553 342 112 48 35 1,365 90 111 1,780 57 191 322 72 580 1,412 658 38 932 389 73 1,068 841 526 241 948 144 553 1,594 943 85 428 118 1,189 1,593 207 105

96 (5.7%) 35 (23.2%) 72 (9.7%) 45 (18.7%) 30 (5.1%) 21 (6.9%) 169 (3.1%) 52 (6.3%) 54 (6.7%) 16 (14.8%) 9 (8.3%) 53 (3.4%) 32 (9.4%) 23 (20.5%) 10 (20.8%) 8 (22.9%) 28 (2.1%) 15 (16.7%) 16 (14.4%) 111 (6.2%) 5 (8.8%) 25 (13.1%) 29 (9.0%) 4 (5.6%) 35 (6.0%) 56 (4.0%) 79 (12.0%) 4 (10.5%) 168 (18.0%) 71 (18.3%) 22 (30.1%) 69 (6.5%) 43 (5.1%) 49 (9.3%) 23 (9.5%) 58 (6.1%) 21 (14.6%) 28 (5.1%) 127 (8.0%) 89 (9.4%) 19 (22.4%) 64 (15.0%) 20 ((16.9%) 15 (1.3%) 86 (5.4%) 20 (9.7%) 10 (9.5%)

83 (4.9%) 15 (9.9%) 64 (8.6%) 29 (12.0%) 27 (4.6%) 21 (6.9%) 158 (2.9%) 38 (4.6%) 43 (5.3%) 14 (13.0%) 4 (3.7%) 53 (3.4%) 24 (7.0%) NR 10 (20.8%) 8 (22.9%) NR NR 11 (9.9%) 111 (6.2%) NR 25 (13.1%) 25 (7.8%) 3 (4.2%) 35 (6.0%) NR 44 (6.7%) NR NR 71 (18.3%) 14 (19.2%) 69 (6.5%) 43 (5.1%) 49 (9.3%) 14 (5.8%) 41 (4.3%) 21 (14.6%) NR 91 (5.7%) 55 (5.8%) NR 35 (8.2%) NR 15 (1.3%) 86 (5.4%) 20 (9.7%) 10 (9.5%) continued on next page

haematologica | 2019; 104(6)

1279

F.I. Mulder et al. continued from previous page

Author (year)

Tafur (2015)69 van Es (2017)70 van Es (2017)71 Vathiotis72 Verso (2012)23 Wang (2017)73 Yust-Katz (2015)74 Zahir (2017)75

Type

Article Article Article Article Article Article Article Article

Study Newly VTE Cancer Median designâ&#x20AC;Ą diagnosed screening type follow up cancer only before duration study start in months Prospective Prospective Retrospective Retrospective Prospective Retrospective Retrospective Retrospective

Yes No Yes Yes No NR Yes No

Study population*

Total follow up: patients with VTE, n (%)

First 6 months: Patients with VTE, n (%)

241 843 147 130 381 270 440 400

29 (12.0%) 53 (6.3%) 20 (13.6%) 13 (10.0%) 15 (3.9%) 16 (5.9%) 64 (14.5%) 42 (10.5%)

24 (10.0%) 53 (6.3%) 13 (8.8%) 7 (5.4%) 15 (3.9%) 11 (4.1%) NR 42 (10.5%)

No Various 10.4 No Various 6.0 No Pancreatic 7.7 No Lung 3.7 No Various 3.7 No Hepatocellular 11.9 No Glioblastoma NR No Various NR

VTE: venous thromboembolism; n: number; DLBCL: diffuse large B-cell lymphoma; NR:not reported; NHL: non-Hodgkin lymphoma.

(n=12; 23%), lung (n=6; 11%), urogenital (n=6; 11%), hematologic (n=5; 9%), or central nervous system cancer (n=2; 4%). Almost half of the studies had a prospective design (n=25; 47%); the majority also included incidentally detected VTE as outcome event (n=32; 60%). Study group size ranged from 35 to 5,409 patients. Median follow-up duration ranged from 2 to 79 months. Key study characteristics of included studies are shown in Table 2.

The 6-month follow-up data were reported in eight of the included studies. For 11 studies, no additional data were obtained after contacting the corresponding author because the authors did not reply despite reminders (n=8), were not able to retrieve the data (n=1), or where not willing to share the data (n=2). For 34 studies, additional data were obtained, yielding available 6-month data for 27,849 of the available 34,555 patients (81%).

Risk of bias Using the pre-specified Quality in Prognosis Studies (QUIPS) criteria, 25 studies were judged to be at high risk of bias for one or more of the bias domains. All eight included abstracts and four articles were judged to be at high risk of bias because of insufficient reporting on methods. Other reasons were a high risk of bias in the applicability of the Khorana score (n=1), patient selection (n=4), outcome (n=3), study attrition (n=2), participation (n=4), prognostic factor measurement (n=3), outcome measurement (n=5), and confounding factors (n=4). Online Supplementary Table S4 summarizes the risk of bias assessment for all studies. A funnel plot did not indicate evidence of publication bias (Online Supplementary Figure S1).

Risk classification by the Khorana score Overall, 6,319 patients (19%) had a Khorana score of 0 points (low risk), 21,172 patients (64%) a score of 1 or 2 points (intermediate risk), and 5,614 patients (17%) a score of 3 or more points (high risk). The group with a Khorana score of 0 or 1 point comprised 15,107 patients (53%), and the group with a score of 2 points or higher 13,148 (47%).

Incidence of venous thromboembolism in the Khorana score risk groups The incidence of VTE in the first 6-month period was 5.0% (95%CI: 3.9-6.5) in patients with a low-risk Khorana score (0 points), 6.6% (95%CI: 5.6-7.7) in those with an 1280

intermediate-risk Khorana score (1 or 2 points), and 11.0% (95%CI: 8.8-13.8) in those with a high-risk Khorana score (3 points or higher) (Table 3 and Figure 2A-C). The relative risk of VTE in the first six months was 1.8 (95%CI: 1.52.1) for patients with a score of 3 or higher compared to those with a score of 2 or lower (Online Supplementary Figure S2).

In the high-risk Khorana score group, the reported 6month risk of VTE was lower in studies including patients with lung cancer (6.4%; 95%CI: 4.9-8.4) or hematologic malignancies (7.1%; 95%CI: 2.6-18.4) compared to studies with gastrointestinal (13.0%; 95%CI: 8.5-19.6), urogenital cancer (18.2%; 95%CI: 8.6-34.6), or various cancers (11.5%; 95%CI: 8.6-15.3, lung vs. various, P=0.0008; hematologic vs. various, P=0.000). The 6-month incidence in the group with a Khorana score of 1 point or lower was 5.5% (95%CI: 4.5-6.9) compared to 8.9% (95%CI: 7.310.8) in the group with a score of 2 or more points, corresponding to a relative risk of 1.5 (95%CI: 1.3-1.8).

During the overall study follow-up period, that ranged from a median of two to 79 months, the summary incidence of VTE was 5.7% (95%CI: 4.2-7.9) in patients with a low-risk Khorana score (0 points), 8.6% (95%CI: 7.310.2) in those with an intermediate-risk Khorana score (1 or 2 points), and 14.0% (95%CI:11.7-16.7) in those with a high-risk Khorana score (3 points or higher) (Table 3 and Online Supplementary Figure S3A-C).

Distribution of venous thromboembolic events over the Khorana score risk groups Of all patients who developed VTE in the first six months, 23.4% (95%CI: 18.4-29.4) had been classified as high risk with the Khorana score (3 points or higher). All other thromboembolic events occurred in the intermediate- or low-risk groups (76.6%; 95%CI:70.6 -81.6). For the total follow-up duration, the proportion of events occurring in the high-risk group was 23.7% (95%CI: 18.7-29.5).

Sensitivity analyses Results were consistent in the sensitivity analysis in which studies judged to be at high risk of bias in one or more of the bias domains were excluded (Table 3). When excluding these studies, the 6-month risks of VTE in patients with a Khorana score of 0, 1 to 2, and 3 points or higher were 4.6% (95%CI: 3.2-6.5), 6.1% ((95%CI: 5.07.4), and 11.1% (95%CI: 8.3-14.7), respectively. The incihaematologica | 2019; 104(6)

Predicting cancer-associated venous thromboembolism

Figure 1. PRISMA flow chart. ASH: American Society of Hematology; ISTH: International Society on Thrombosis and Haemostasis.

dence in the group with a score of 2 points or higher was 8.3% (95%CI: 6.4-10.7). The relative risk of patients with a score of 3 or higher compared to those with a lower score was 1.9 (95%CI: 1.5-2.3).

Discussion This systematic review and meta-analysis examined the performance of the Khorana score in predicting VTE in over 34,000 patient ambulatory patients with various types of cancer. To minimize between-study heterogeneity and obtain clinically relevant estimates, the main analysis was restricted to the first six months of follow up.

During this period, the summary estimate of the risk of VTE in patients with a high-risk Khorana score was 11.0%, which was significantly higher than in those with a low-risk (5.0%) or intermediate-risk (6.6%) score. These findings indicate that the Khorana score may help clinicians in selecting patients at high risk of VTE for thromboprophylaxis, which is in support of the suggestions presented in current guidelines.

The analyses also highlight several limitations of the score. Within the high-risk group, the estimated risk of haematologica | 2019; 104(6)

VTE was considerably lower for patients with lung cancer and hematologic malignancies than for those with other cancer types (Figure 2C). Hence, the Khorana score appears to be less informative for these two large groups of patients. Furthermore, the VTE incidence in patients with a low-to-intermediate risk score was 5-7%, which indicates that the residual risk in this group is still substantial.

Therefore, the Khorana score is of limited use in ruling out a future venous thromboembolic event. Lastly, the Khorana score is designed to select patients in the high-risk group for thromboprophylaxis. However, about one in four (23.4%, 95%CI: 18.4-29.4) of the venous thromboembolic events occur in patients with a high-risk Khorana score. This means that a substantial amount of cancer patients with subsequent venous thromboembolic events will not be identified with this form of risk stratification, and will, therefore, not benefit from thromboprophylaxis.

A major strength of this study is the additional data obtained from 34 studies on the 6-month incidence of VTE after starting chemotherapy, representing 81% of cancer patients in the available relevant literature. This approach minimized between-study heterogeneity related to the broad range of reported median follow-up durations. We considered this 6-month period to be clinically 1281

F.I. Mulder et al.

A

B

1282

haematologica | 2019; 104(6)

Predicting cancer-associated venous thromboembolism

C

Figure 2. Venous thromboembolism incidence in the low-, intermediate-, and high-risk group over six months. Venous thromboembolism incidence in the low-risk (A), intermediate-risk (B), and high-risk (C) groups according to the Khorana score, over six months follow up.

most relevant. Prediction of VTE only for the first few months of chemotherapy may be too short, since the risk remains elevated throughout the first six months. On the other hand, the Khorana score calculated with prechemotherapy laboratory data likely predicts less well for longer term (>6 months) than for shorter term intervals.

The inclusion of more than 50 studies enabled the metaanalysis for various subgroups of cancer patients, showing that the performance of the Khorana score varies across tumor types. A potential limitation is the substantial proportion of studies judged to be at high risk of bias (Online Supplementary Table S4). However, the sensitivity analyses restricted to studies at low risk of bias did not materially alter the results (Table 3). When the analysis was restricted to studies with a prospective design or to studies without systematic VTE screening preceding study, results were comparable (data not shown). Additional data for the first haematologica | 2019; 104(6)

six months could not be obtained for eleven studies, possibly introducing sampling bias. We believe, however, that the magnitude of this risk of bias is at best modest since 6month data were available in the final analyses for 81% of all patients. Some studies included more types of venous thromboembolic events than specified in our primary outcome. However, these types of venous thromboembolic events occur infrequently. A large proportion of the studies (n=32, 60%) included incidentally detected VTE, unlike the outcome in the derivation study of the Khorana score.7 However, we believe these events should also be considered since clinical outcomes in patients with incidental VTE are similar to those with symptomatic events.14-16 Consequently, international guidelines regard incidental VTE events as clinically relevant and recommend anticoagulant treatment, as for patients with symptomatic VTE.6,17 Despite minimizing bias due to differ1283

F.I. Mulder et al.

ences in follow up by using 6-month outcome data, considerable residual heterogeneity was observed in the analyses. This is expected in meta-analyses of predictive model performance, especially when evaluating risk assessment tools across various cancers.18 Nonetheless, we believe the presented estimates overall and for subgroups by cancer type are the most reliable ones based on the current literature, and can help clinicians to decide whether to use the score in their practice.

Two currently ongoing randomized trials use the Khorana score to select cancer patients at high risk of VTE for thromboprophylaxis (clinicaltrials.gov identifier: 02048865 and 02555878). Interestingly, these studies apply a positivity threshold of 2 points rather than the conventional 3 points. Our analyses demonstrate that this approach increases the proportion of patients classified as high risk (17-47%) while in parallel decreasing the absolute risk of VTE in this group (11-9%). As a conse-

Table 3. Summary estimates for 6-month and total follow-up duration.‡

Khorana score 0 % (95% CI) 6 months follow-up duration 5.0 (3.9-6.5) Total study follow-up duration* 5.7 (4.2-7.9) Low and moderate bias studies only 6 months follow-up duration 4.6 (3.2-6.5) Total study follow-up duration* 4.5 (3.0-6.7)

Khorana score 1-2 % (95% CI)

Incidence of VTE Khorana score ≥3 % (95% CI)

Khorana score ≤1 % (95% CI)

Relative risk versus lower Proportion risk groups of all VTE Khorana Khorana Khorana Khorana score≥2 score ≥3 score ≥2 score ≥3 % (95% CI) (95% CI) (95% CI) % (95% CI)

6.6 (5.6-7.7) 11.0 (8.8-13.8) 5.5 (4.5-6.9) 8.9 (7.3-10.8) 1.8 (1.5-2.1) 1.5 (1.3-1.8) 23.4 (18.4-29.4) 8.6 (7.3-10.2) 14.0 (11.7-16.7) 6.8 (5.2-8.9) 11.3 (9.4-13.4) 1.7 (1.5-2.0) 1.5 (1.3-1.8) 23.7 (18.7-29.5) 6.1 (5.0-7.4) 11.1 (8.3-14.7) 5.0 (4.0-6.3) 8.3 (6.4-10.7) 1.9 (1.5-2.3) 1.6 (1.3-2.0) 24.4 (17.8-32.5 7.6 (6.0-9.5) 13.5 (10.7-16.8) 6.3 (4.9-8.1) 10.6 (8.4-13.2) 1.8 (1.4-2.2) 1.5 (1.2-1.9) 22.9 (17.2-29.9)

‡ Estimates were derived from random effects meta-analysis. *Total follow-up duration varied substantially complicating interpretation of the results at total follow-up duration.

CI: confidence interval; VTE: venous thromboembolism.

A

B

5.0%

C

1284

D

Figure 3. Estimated incidence of venous thrombosis and proportion in the high-risk group over six months. Estimated incidence of venous thrombosis (A and C) and proportion of venous thromboembolic events allocated to the high-risk group (B and D). When considering two points or more as high-risk (C and D) instead of three points or more (traditional threshold, A and B), the proportion of venous thromboembolic events allocated to the high risk groups increases, but also results in a lower incidence. VTE: venous thromboembolism.

haematologica | 2019; 104(6)

Predicting cancer-associated venous thromboembolism

quence, the proportion of thromboembolic events that occur in the high-risk group increases from 23% to 55% (Figure 3). It is a matter of debate whether the 9% risk of VTE during the first six months is considered high enough to justify thromboprophylaxis.

The primary aim of risk stratification with the Khorana score is to select cancer patients with a high risk of VTE suitable for long-term thromboprophylaxis. A meta-analysis of randomized trials that compared low-molecularweight heparins in prophylactic doses in cancer patients with placebo showed an absolute risk reduction of approximately 50% during a median follow-up of ten months (RR 0.54; 95%CI: 0.38-0.75), with an increase in major bleeding events (RR 1.44; 95%CI: 0.98-2.11).19 As the estimated 6-month incidence of VTE in cancer patients with a high Khorana score is 11.0%, thromboprophylaxis with low-molecular-weight heparins for cancer patients in this group could result in a number requiring treatment of approximately 19 when extrapolating the relative risk reduction of 0.54. When considering patients with 2 points or more as high-risk, thromboprophylaxis with low-molecular-weight heparins could result in a number requiring treatment of 24. Recent trials showed an acceptable safety profile of therapeutic doses of direct oral anticoagulants in cancer patients compared to low-molecularweight heparins.20,21 Since their oral administration makes these drugs more convenient, long-term thromboprophylaxis would be less burdensome and, therefore, more likely to be accepted by clinicians and patients. Whether the safety and efficacy of prophylactic doses of direct oral anticoagulants are comparable to that of low-molecularweight heparin in cancer patients needs to be established.

The present meta-analysis shows that the Khorana score can select high-risk patients for thromboprophylaxis overall. These findings indicate that the Khorana score may help clinicians in selecting patients at high risk of VTE for thromboprophylaxis, which is in support of the suggestions presented in some guidelines and could accelerate their implementation in clinical practice. However, several limitations of the Khorana need to be taken into account, including the different in predicted performance across cancer types and the modest proportion of patients with VTE assigned to the high-risk group. Several other VTE prediction tools for cancer patients have been introduced, which may have a better performance than the Khorana score;22-24 these scores, however, require prospective validation. Development of risk prediction models for bleeding events in patients with prophylactic anticoagu-

References 1. Timp JF, Braekkan SK, Versteeg HH, Cannegieter SC. Epidemiology of venous thrombosis. Blood. 2013;122(10):17121723.

2. Cohen AT, Katholing A, Rietbrock S, Bamber L, Martinez C. Epidemiology of first and recurrent venous thromboembolism in patients with active cancer.

Thromb Haemost. 2016;117(1):57-65.

3. Khorana AA, Francis CW, Culakova E, Kuderer NM, Lyman GH. Thromboembolism is a leading cause of death in cancer patients receiving outpa-

haematologica | 2019; 104(6)

lants could help to carefully weigh the benefit risk tradeoff for thromboprophylaxis in cancer patients. In addition, future prediction tools should aim to address the limitations of the Khorana score, as outlined by this analysis.

Novel biomarkers or genetic information from tumor biopsies could improve prediction of VTE and, therefore, merit investigation.

CAT-prediction collaborators Abdel-Razeq H, King Hussein Cancer Center, Jordan; Ades S, University of Vermont, Burlington, VT, USA; Ayappan SR, The Ohio State University Comprehensive Cancer Center (OSUCCC-James), Columbus, OH, USA; Borchmann S, University Hospital Cologne, Germany; Cella CA, Federico II University, Naples, Italy; Fankhauser CD, University Hospital Zürich, Switzerland; Ferroni P, San Raffaele Roma Open University, Italy; Fuentes HE, John Stronger Jr. Hospital, Chicago, IL, USA; Kruger S, Ludwig-Maximilians-University of Munich, Germany; Lim SH, Samsung Medical Center, Seoul, Republic of Korea; Lubberts S, University Medical Center Groningen, the Netherlands; Lustig DB, University of Ottawa, ON, Canada; Mansfield AS, Mayo Clinic, Rochester, MN, USA; Munõz Martín AJ, Medical Oncology Service, Hospital General Universitario Gregorio Marañón, Madrid, Spain; Noble S, Cardiff University, UK; Panizo E, University Clinic of Navarra, Pamplona, Spain; Papaxoinis G, Department of Medical Oncology, The Christie NHS Foundation Trust, Manchester, UK; Park K, Pusan National University Yangsan Hospital, Republic of Korea; Patel JN, Levine Cancer Institute, NC, USA; Posch F, Medical University of Vienna, Austria; Ramos JD, University of Washington, Seattle, WA, USA; Roselli M, University of Rome Tor Vergata, USA; Santi R, A.O.SS. Antonio e Biagio e Cesare Arrigo of Alessandria, Italy; Sohal D, Cleveland Clinic, OH, USA; Srikanthan A, Princess Margaret Cancer Centre, University of Toronto, ON, USA; Tafur AJ, University of Oklahoma Health Sciences Center, USA; Terbuch A, Medical University, Graz, Austria; Thomas M, University College London Hospitals NHS Foundation Trust, UK; Vathiotis O, Oncology Unit, Sotiria General Hospital, University of Athens, Greece; Wang R, Graduate School of Tianjin Medical University, PRC; Zahir MN, Aga Khan University Hospital, Pakistan.

Funding This study was supported by an unrestricted grant from LeoPharma. The sponsor had no influence on study design, data collection, analysis, writing of the manuscript, or in the decision to submit the manuscript for publication.

tient chemotherapy. J Thromb Haemost.

2007;5(3):632-634.

4. Lloyd AJ, Dewilde S, Noble S, Reimer E, Lee AYY. What Impact Does Venous Thromboembolism and Bleeding Have on Cancer Patients’ Quality of Life? Value Heal. 2018;21(4):449-455.

5. Di Nisio M, Porreca E, Ferrante N, Otten HM, Cuccurullo F, Rutjes AW. Primary prophylaxis for venous thromboembolism in ambulatory cancer patients receiving chemotherapy. Cochrane Database Syst Rev. 2012;2(12):CD008500.

6. Lyman GH, Bohlke K, Khorana AA, et al.

Venous thromboembolism prophylaxis and

treatment in patients with cancer: American Society of Clinical Oncology clinical practice guideline update 2014. J Clin Oncol. 2015;33(6):654-656.

7. Khorana AA, Kuderer NM, Culakova E, Lyman GH, Francis CW. Development and validation of a predictive model for chemotherapy-associated thrombosis.

Blood. 2008;111(10):4902-4907.

8. Streiff MB, Holmstrom B, Ashrani A, et al.

Cancer-associated venous thromboembolic disease, version 1.2015: Featured updates to the NCCN Guidelines. JNCCN J Natl Compr Cancer Netw. 2015;13(9):1079-1095.

9. Mandalá M, Clerici M, Corradino I, et al.

1285

F.I. Mulder et al.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

1286

Incidence, risk factors and clinical implications of venous thromboembolism in cancer patients treated within the context of phase I studies: The “sendo experience.”

Ann Oncol. 2012;23(6):1416-1421.

Bezan A, Posch F, Ploner F, et al. Risk stratification for venous thromboembolism in patients with testicular germ cell tumors.

PLoS One. 2017;12(4):1-15.

Moher D, Liberati A, Tetzlaff J, et al.

Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. PLoS Med. 2009;6(7):e1000097.

Hayden J a, Windt D a Van Der, Cartwright JL, Co P. Research and reporting methods annals of internal medicine assessing bias in studies of prognostic factors. Ann Intern Med. 2013;158(4):280-286.

Sterne JAC, Harbord RM. Funnel plots in meta-analysis. Stata J. 2004;4(2):127-141. den Exter PL, Hooijer J, Dekkers OM, Huisman M V. Risk of recurrent venous thromboembolism and mortality in patients with cancer incidentally diagnosed with pulmonary embolism: a comparison with symptomatic patients. J Clin Oncol.

2011;29(17):2405-2409.

Font C, Carmona-Bayonas A, Beato C, et al. Clinical features and short-term outcomes of cancer patients with suspected and unsuspected pulmonary embolism: The EPIPHANY study. Eur Respir J. 2017;49(1):1600282.

Sahut D’Izarn M, Caumont Prim A, Planquette B, et al. Risk factors and clinical outcome of unsuspected pulmonary embolism in cancer patients: a case-control study. J Thromb Haemost. 2012; 10(10):2032-2038.

Kearon C, Akl EA, Comerota AJ, et al.

Antithrombotic therapy for VTE disease: Antithrombotic therapy and prevention of thrombosis, 9th ed: American College of Chest Physicians evidence-based clinical practice guidelines. Chest. 2012;141(2 SUPPL.):419-494.

Debray TPA, Damen JAAG, Snell KIE, et al.

A guide to systematic review and metaanalysis of prediction model performance.

BMJ. 2017;356:i6460.

Di Nisio M, Porreca E, Ferrante N, Otten HM, Cuccurullo F, Rutjes AW. Primary prophylaxis for venous thromboembolism in ambulatory cancer patients receiving chemotherapy. Cochrane Database Syst Rev. 2012;(2):CD008500.

Young AM, Marshall A, Thirlwall J, et al.

Comparison of an oral factor Xa inhibitor with low molecular weight heparin in patients with cancer with venous thromboembolism: results of a randomized trial (SELECT-D). J Clin Oncol. 2018; 36(20):2017-2023.

Raskob GE, van Es N, Verhamme P, et al.

Edoxaban for the treatment of cancer-associated venous thromboembolism. N Engl J Med. 2018;378(7):615-624.

Pabinger I, van Es N, Heinze G, et al. A clinical prediction model for cancer-associated venous thromboembolism: a development and validation study in two independent prospective cohorts. Lancet Haematol.

2018;5(7):e289-e298.

Verso M, Agnelli G, Barni S, Gasparini G, LaBianca R. A modified Khorana risk assessment score for venous thromboembolism in cancer patients receiving chemotherapy: the Protecht score. Intern Emerg Med. 2012;7(3):291-292.

Pelzer U, Sinn M, Stieler J, Riess H. Primäre medikamentöse thromboembolieprophy-

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

laxe bei ambulanten patienten mit fortgeschrittenem pankreaskarzinom unter chemotherapie? Dtsch Med Wochenschr.

2013;138(41):2084-2088.

Abdel-Razeq H, Mansour A, Abdulelah H, et al. Thromboembolic events in cancer patients on active treatment with cisplatinbased chemotherapy: another look!

Thromb J. 2018;16:2.

Ades S, Kumar S, Alam M, et al. Tumor oncogene (KRAS) status and risk of venous thrombosis in patients with metastatic colorectal cancer. J Thromb Haemost .2015;13(6):998-1003.

Austin K, Borrowman J, Blake L, Scully M, Thomas M. Retrospective cohort study of venous thromboembolism (VTE) rates in ambulatory cancer patients, and association with Khorana Score. Res Pract Thromb Haemost. 2017;11-1451.

Ayyappan SR, Gupta V, Diamond A, et al.

Venous thromboembolic events in diffuse large B cell lymphoma patients: risk factors and outcomes. Blood. 2016; 128(22):3611.

Borchmann S, Hude I, Müller H, et al. thrombosis in Hodgkin lymphoma patients: incidence, time points, risk factors and impact of stage and treatment. Blood.

2016;128(22):4151.

Cella CA, Di Minno G, Carlomagno C, et al. Preventing Venous Thromboembolism in Ambulatory Cancer Patients: The ONKOTEV Study. Oncologist. 2017; 22(5):601-608.

Ferroni P, Riondino S, Formica V, et al.

Venous thromboembolism risk prediction in ambulatory cancer patients: clinical significance of neutrophil/lymphocyte ratio and platelet/lymphocyte ratio. Int J Cancer.

2015;136(5):1234-1240.

Ferroni P, Martini F, Portarena I, et al. Novel high-sensitive D-Dimer determination predicts chemotherapy-associated venous thromboembolism in intermediate risk lung cancer patients. Clin Lung Cancer.

2012;13(6):482-487.

Fuentes HE, Oramas DM, Paz LH, Wang Y, Andrade XA, Tafur AJ. Venous thromboembolism is an independent predictor of mortality among patients with gastric cancer. J Gastrointest Cancer. 2017;1-7.

George D, Agnelli G, Fisher W, et al.

Venous thromboembolism (VTE) prevention with semuloparin in cancer patients initiating chemotherapy: benefit-risk assessment by VTE risk in SAVE-ONCO. Blood. 2011;118(21):206 LP-206.

Agnelli G, George DJ, Kakkar AK, et al.

Semuloparin for thromboprophylaxis in patients receiving chemotherapy for cancer.

N Engl J Med. 2012;366(7):601-609.

Guadagni F, Riondino S, Formica V, et al.

Clinical significance of glycemic parameters on venous thromboembolism risk prediction in gastrointestinal cancer. World J Gastroenterol. 2017;23(28):5187-5195.

Kearney JC, Rossi S, Glinert K, Henry DH. Venous thromboembolism (VTE) and survival in a cancer chemotherapy outpatient clinic: a retrospective chart review validation of a VTE predictive Model. Blood.

2009;114(22):2503.

Khorana AA, Francis CW, Kuderer NM, et al. Dalteparin thromboprophylaxis in cancer patients at high risk for venous thromboembolism: a randomized trial. Thromb Res. 2017;151:89-95.

Khorana AA, Rubens D, Francis CW. Screening high-risk cancer patients for VTE: A prospective observational study.

Thromb Res. 2014;134(6):1205-1207.

40. Kim SY, Burns ZT, Henry DH. The assessment of thrombotic risk using a predictive model in metastatic cancer patients undergoing first-line therapy. Thromb Res. 2012; 130(6):967-970.

41. Kruger S, Haas M, Burkl C, et al. Incidence, outcome and risk stratification tools for venous thromboembolism in advanced pancreatic cancer – A retrospective cohort study. Thromb Res. 2017;157:9-15.

42. Kuderer NM, Poniewierski MS, Culakova E, et al. Predictors of venous thromboembolism and early mortality in lung cancer: results from a global prospective study (CANTARISK). Oncologist. 2018;23(2):247-255.

43. Kuk A, Magnowska M, Suchy W, et al.

Retrospective evaluation of Thromboembolism risk in ovarian cancer patients treated with Bevacizumab. Target Oncol. 2017;12(4):495-503.

44. Kunapareddy G, Switzer B, Conces M, et al. Implementation of an Electronic medical record tool for early detection of deep vein thrombosis in the Ambulatory oncology Setting: the Cleveland Clinic Experience.

Blood. 2017;130(Suppl 1):3361.

45. Lim SH, Woo SY, Kim S, Ko YH, Kim WS, Kim SJ. Cross-sectional study of patients with diffuse large B-cell lymphoma: assessing the effect of host status, tumor burden and inflammatory activity on venous thromboembolism. Cancer Res Treat.

2015;48(1):1-10.

46. Lubberts S, Boer H, Altena R, et al. Vascular fingerprint and vascular damage markers associated with vascular events in testicular cancer patients during and after chemotherapy. Eur J Cancer. 2016;63:180-188.

47. Lustig D Ben, Rodriguez R, Wells PS. Implementation and validation of a risk stratification method at the Ottawa Hospital to guide thromboprophylaxis in ambulatory cancer patients at intermediate-high risk for venous thrombosis.

Thromb Res. 2015;136(6):109-1102.

48. Mansfield AS, Tafur AJ, Wang CE, Kourelis TV, Wysokinska EM, Yang P. Predictors of active cancer thromboembolic outcomes: validation of the Khorana score among patients with lung cancer. J Thromb Haemost. 2016;14(9):1773-1778.

49. Misch M, Czabanka M, Dengler J, et al. Ddimer elevation and paresis predict thromboembolic events during bevacizumab therapy for recurrent malignant glioma.

Anticancer Res. 2013;33(5):2093-2098.

50. Moore RA, Adel N, Riedel E, et al. High incidence of thromboembolic events in patients treated with cisplatin-based chemotherapy: A large retrospective analysis. J Clin Oncol. 2011;29(25):3466-3473.

51. Muñoz Martín AJ, Ortega I, Font C, et al.

Multivariable clinical-genetic risk model for predicting venous thromboembolic events in patients with cancer. Br J Cancer. 2018; 118(8):1056-1061.

52. Muñoz Martín AJ, García Alfonso P, Rupérez Blanco AB, et al. Incidence of venous thromboembolism (VTE) in ambulatory pancreatic cancer patients receiving chemotherapy and analysis of Khorana’s predictive model. Clin Transl Oncol. 2014; 16(10):927-930.

53. Noble S, Robbins A, Alikhan R, Hood K, Macbeth F. Prediction of venous thromboembolism in lung cancer patients receiving chemotherapy. Int Soc Thromb Haemost. 2015;131-997.

54. Macbeth F, Noble S, Evans J, et al.

Randomized phase III trial of standard therapy plus low molecular weight heparin in

haematologica | 2019; 104(6)

Running Title

55.

56.

57.

58.

59.

60.

61.

patients with lung cancer: FRAGMATIC trial. J Clin Oncol. 2016;34(5):488-494.

Panizo E, Alfonso A, García-Mouriz A, et al. Factors influencing the use of thromboprophylaxis in cancer outpatients in clinical practice: A prospective study. Thromb Res. 2015;136(6):1145-1148.

Papaxoinis G, Kamposioras K, Germetaki T, et al. Predictive factors of thromboembolic complications in patients with esophagogatric adenocarcinoma undergoing preoperative chemotherapy. Acta Oncol. 2018; 57(6):1-9.

Park K, Ryoo B-Y, Ryu M-H, et al. Incidence of venous thromboembolism and the role of D-dimer as predictive marker in patients with advanced gastric cancer receiving chemotherapy: A prospective study. World J Gastrointest Oncol. 2017;9(4):176-183.

Patel JN, Jiang C, Hertz DL, et al.

Bevacizumab and the risk of arterial and venous thromboembolism in patients with metastatic, castration-resistant prostate cancer treated on Cancer and Leukemia Group B (CALGB) 90401 (Alliance). Cancer.

2015;121(7):1025-1031.

Pelzer U, Opitz B, Deutschinoff G, et al.

Efficacy of prophylactic low-molecular weight heparin for ambulatory patients with advanced pancreatic cancer: Outcomes from the CONKO-004 trial. J Clin Oncol. 2015;33(18):2028-2034.

Sanchez Petitto G, Escalante CP, Richardson MN, Rojas Hernandez C. modified Khorana models for prediction of cancer-associated venous thromboembolism: an exploratory study. Blood. 2017; 130(Suppl 1):4635.

Posch F, Riedl J, Reitter E-M, et al.

haematologica | 2019; 104(6)

62.

63.

64.

65.

66.

67.

68.

Hypercoagulabilty, venous thromboembolism, and death in patients with cancer.

Thromb Haemost. 2016;115(04):817-826.

Ramos JD, Casey MF, Bamias A, et al. The Khorana Score in predicting venous thromboembolism for patients with metastatic urothelial carcinoma and variant histology treated with chemotherapy. Clin Appl Thromb. 2017; 23(7):755-760.

Ruch JM, Bellile E, Hawley AE, Anderson MA, Wakefield TW, Sood SL. Prediction of venous thromboembolism (VTE) in patients with Pancreatic cancer using clinical data, biomarkers, and VTE risk models.

Blood. 2012;120(21):3398.

Rupa-Matysek J, Gil L, Kaźmierczak M, Barańska M, Komarnicki M. Prediction of venous thromboembolism in newly diagnosed patients treated for lymphoid malignancies: validation of the Khorana Risk Score. Med Oncol. 2018;35(1):1–8.

Rupa-Matysek J, Lembicz M, Rogowska EK, Gil L, Komarnicki M, Batura-Gabryel H. Evaluation of risk factors and assessment models for predicting venous thromboembolism in lung cancer patients. Med Oncol. 2018;35(5):1-10.

Santi RM, Ceccarelli M, Bernoccol E, et al.

Khorana score and histotype predicts incidence of early venous thromboembolism in non-Hodgkin lymphomas.

Thromb Haemost. 2017;117(8):1615-1621.

Sohal DPS, Kuderer NM, Shepherd FA, et al. Predictors of venous thmromboembolism in colorectal cancer: results from a global prospective study. Blood. 2016; 128(22):1422.

Srikanthan A, Tran B, Beausoleil M, et al.

Large retroperitoneal lymphadenopathy as

69.

70.

71.

72.

73.

74.

75.

a predictor of venous thromboembolism in patients with disseminated germ cell tumors treated with chemotherapy. J Clin Oncol. 2015;33(6):582-587.

Tafur AJ, Dale G, Cherry M, et al.

Prospective evaluation of protein C and factor VIII in prediction of cancer-associated thrombosis. Thromb Res. 2015; 136(6):1120-1125. van Es N, Di Nisio M, Cesarman G, et al.

Comparison of risk prediction scores for venous thromboembolism in cancer patients: A prospective cohort study.

Haematologica. 2017;102(9):1494-1501. van Es N, Franke VF, Middeldorp S, Wilmink JW, Büller HR. The Khorana score for the prediction of venous thromboembolism in patients with pancreatic cancer.

Thromb Res. 2017;150:30-32.

Vathiotis I, Dimakakos EP, Boura P, et al.

Khorana Score: new predictor of early mortality in patients With lung adenocarcinoma. Clin Appl Thromb. 2018;24(8):13471351.

Wang Y, Attar BM, Fuentes HE, Yu J, Zhang H, Tafur AJ. Performance of Khorana Risk Score for prediction of Venous thromboembolism in patients with hepatocellular carcinoma. Clin Appl Thromb Hemost. 2018; 24(3):471-476.

Yust-Katz S, Mandel JJ, Wu J, et al. Venous thromboembolism (VTE) and glioblastoma.

J Neurooncol. 2015;124(1):87-94.

Zahir MN, Shaikh Q, Shabbir-Moosajee M, Jabbar AA. Incidence of venous thromboembolism in cancer patients treated with cisplatin based chemotherapy - a cohort study. BMC Cancer. 2017;17(1):1-8.

1287