Page 1


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

Information for readers, authors and subscribers Haematologica (print edition, pISSN 0390-6078, eISSN 1592-8721) publishes peer-reviewed papers on all areas of experimental and clinical hematology. The journal is owned by a non-profit organization, the Ferrata Storti Foundation, and serves the scientific community following the recommendations of the World Association of Medical Editors (www.wame.org) and the International Committee of Medical Journal Editors (www.icmje.org). Haematologica publishes editorials, research articles, review articles, guideline articles and letters. Manuscripts should be prepared according to our guidelines (www.haematologica.org/information-for-authors), and the Uniform Requirements for Manuscripts Submitted to Biomedical Journals, prepared by the International Committee of Medical Journal Editors (www.icmje.org). Manuscripts should be submitted online at http://www.haematologica.org/. Conflict of interests. According to the International Committee of Medical Journal Editors (http://www.icmje.org/#conflicts), “Public trust in the peer review process and the credibility of published articles depend in part on how well conflict of interest is handled during writing, peer review, and editorial decision making”. The ad hoc journal’s policy is reported in detail online (www.haematologica.org/content/policies). Transfer of Copyright and Permission to Reproduce Parts of Published Papers. Authors will grant copyright of their articles to the Ferrata Storti Foundation. No formal permission will be required to reproduce parts (tables or illustrations) of published papers, provided the source is quoted appropriately and reproduction has no commercial intent. Reproductions with commercial intent will require written permission and payment of royalties. Detailed information about subscriptions is available online at www.haematologica.org. Haematologica is an open access journal. Access to the online journal is free. Use of the Haematologica App (available on the App Store and on Google Play) is free. For subscriptions to the printed issue of the journal, please contact: Haematologica Office, via Giuseppe Belli 4, 27100 Pavia, Italy (phone +39.0382.27129, fax +39.0382.394705, E-mail: info@haematologica.org). Rates of the International edition for the year 2018 are as following: Print edition

Institutional Euro 600

Personal Euro 150

Advertisements. Contact the Advertising Manager, Haematologica Office, via Giuseppe Belli 4, 27100 Pavia, Italy (phone +39.0382.27129, fax +39.0382.394705, e-mail: marketing@haematologica.org). Disclaimer. Whilst every effort is made by the publishers and the editorial board to see that no inaccurate or misleading data, opinion or statement appears in this journal, they wish to make it clear that the data and opinions appearing in the articles or advertisements herein are the responsibility of the contributor or advisor concerned. Accordingly, the publisher, the editorial board and their respective employees, officers and agents accept no liability whatsoever for the consequences of any inaccurate or misleading data, opinion or statement. Whilst all due care is taken to ensure that drug doses and other quantities are presented accurately, readers are advised that new methods and techniques involving drug usage, and described within this journal, should only be followed in conjunction with the drug manufacturer’s own published literature. Direttore responsabile: Prof. Edoardo Ascari; Autorizzazione del Tribunale di Pavia n. 63 del 5 marzo 1955. Printing: Press Up, zona Via Cassia Km 36, 300 Zona Ind.le Settevene - 01036 Nepi (VT)


haematologica calendar of events

Journal of the European Hematology Association Published by the Ferrata Storti Foundation EHA-SAH Hematology Tutorial on lymphoid Malignancies and Plasma Cell Dyscrasias September 14-15, 2018 Buenos Aires, Argentina 14th Educational Course of the Lymphoma Working Party European Society for Blood and Marrow Transplantations (EBMT) Lymphoma Working Party Chairs: S Montoto, A Sureda, L Bento September 27-28, 2018 Palma, Spain EHA-SWG Scientific Meeting on Aging and Hematology Chair: D Bron October 12-14, 2018 Warsaw, Poland

Calendar of Events updated on June 4, 2018


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

Table of Contents Volume 103, Issue 7: July 2018 Cover Figure Peripheral blood smear from a patient with thrombotic microangiopathic anemia showing schistocytes, microspherocytes and polychromatic macrocytes. Courtesy of Prof. Rosangela Invernizzi.

Editorials 1093

Linking histone methylation, transcription rates, and stem cell robustness Justin C. Wheat and Ulrich Steidl

1094

Complication free survival long-term after hemopoietic cell transplantation in thalassemia Emanuele Angelucci

1096

Still a role for second-line chemoimmunotherapy in chronic lymphocytic leukemia? Jennifer R Brown

Review Article 1099

Dissecting the pathophysiology of immune thrombotic thrombocytopenic purpura: interplay between genes and environmental triggers Johana Hrdinovรก et al.

Articles Hematopoiesis

1110

Setd2 regulates quiescence and differentiation of adult hematopoietic stem cells by restricting RNA polymerase II elongation Yile Zhou et al.

Red Cell Biology & its Disorders

1124

Endothelin type A receptors mediate pain in a mouse model of sickle cell disease Brianna Marie Lutz et al.

1136

Proteomic analysis of plasma from children with sickle cell anemia and silent cerebral infarction Sanjay Tewari et al.

Complications in Hematology

1143

Late effects after hematopoietic stem cell transplantation for b-thalassemia major: the French national experience Ilhem Rahal et al.

Bone Marrow Failure

1150

Circulating exosomal microRNAs in acquired aplastic anemia and myelodysplastic syndromes Valentina Giudice et al.

Myeloproliferative Disorders

1160

JAK2V617F-bearing vascular niche enhances malignant hematopoietic regeneration following radiation injury Chi Hua Sarah Lin et al.

Acute Myeloid Leukemia

1169

The chromatin-remodeling factor CHD4 is required for maintenance of childhood acute myeloid leukemia Yaser Heshmati et al.

Haematologica 2018; vol. 103 no. 7 - July 2018 http://www.haematologica.org/


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

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

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

Scientific Latin

haematologicus (adjective) = related to blood

Scientific Latin

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

Modern English

The oldest hematology journal, publishing the newest research results. 2016 JCR impact factor = 7.702

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


haematologica Journal of the European Hematology Association Published by the Ferrata Storti Foundation Non-Hodgkin Lymphoma

1182

Gene expression profiling reveals a close relationship between follicular lymphoma grade 3A and 3B, but distinct profiles of follicular lymphoma grade 1 and 2 Heike Horn et al.

1191

The outcome of peripheral T-cell lymphoma patients failing first-line therapy: a report from the prospective, International T-Cell Project Monica Bellei et al.

Chronic Lymphocytic Leukemia

1198

Low-count monoclonal B-cell lymphocytosis persists after seven years of follow up and is associated with a poorer outcome Ignacio Criado et al.

1209

Efficacy of bendamustine and rituximab as first salvage treatment in chronic lymphocytic leukemia and indirect comparison with ibrutinib: a GIMEMA, ERIC and UK CLL FORUM study Antonio Cuneo et al.

Plasma Cell Disorders

1218

Repurposing tofacitinib as an anti-myeloma therapeutic to reverse growth-promoting effects of the bone marrow microenvironment Christine Lam et al.

1229

Plasma cell proliferative index predicts outcome in immunoglobulin light chain amyloidosis treated with stem cell transplantation M. Hasib Sidiqi et al.

Platelet Biology & Its Disorders

1235

Platelet Munc13-4 regulates hemostasis, thrombosis and airway inflammation Eduardo I. Cardenas et al.

Coagulation & its Disorders

1245

C-reactive protein and risk of venous thromboembolism: results from a population-based case-crossover study Gro Grimnes et al.

Letters to the Editor Letters are available online only at www.haematologica.org/content/103/7.toc

e279

Improved erythroid differentiation of multiple human pluripotent stem cell lines in microcarrier culture by modulation of Wnt/b-Catenin signaling Jaichandran Sivalingam et al. http://www.haematologica.org/content/103/7/e279

e284

Expanding the phenotypic and genetic spectrum of radioulnar synostosis associated hematological disease Amanda Walne et al. http://www.haematologica.org/content/103/7/e284

e288

Prognostic impact of the absence of biallelic deletion at the TRG locus for pediatric patients with T-cell acute lymphoblastic leukemia treated on the Medical Research Council UK Acute Lymphoblastic Leukemia 2003 trial Nadine Farah et al. http://www.haematologica.org/content/103/7/e288

e293

Combination of common and novel rare NUDT15 variants improves predictive sensitivity of thiopurine-induced leukopenia in children with acute lymphoblastic leukemia Yiping Zhu et al. http://www.haematologica.org/content/103/7/e293

Haematologica 2018; vol. 103 no. 7 - July 2018 http://www.haematologica.org/


haematologica Journal of the European Hematology Association Published by the Ferrata Storti Foundation e296

Efficacy and safety of high-dose etoposide cytarabine as consolidation following rituximab methotrexate temozolomide induction in newly diagnosed primary central nervous system lymphoma in immunocompetent patients Rudy Birsen et al. http://www.haematologica.org/content/103/7/e296

e300

Incidence and predictors of infection among patients prior to treatment of chronic lymphocytic leukemia: a Danish nationwide cohort study Michael Asger Andersen et al. http://www.haematologica.org/content/103/7/e300

e304

Long-term follow up of the CLL2007FMP trial evaluating fludarabine and cyclophosphamide in combination with either rituximab or alemtuzumab in previously untreated patients with chronic lymphocytic leukemia Pierre Feugier et al. http://www.haematologica.org/content/103/7/e304

e307

Ibrutinib withdrawal symptoms in patients with Waldenström macroglobulinemia Jorge J. Castillo et al. http://www.haematologica.org/content/103/7/e307

e311

The impact of the introduction of bortezomib on dialysis independence in multiple myeloma patients with renal impairment: a nationwide Dutch population-based study Berdien E. Oortgiesen et al. http://www.haematologica.org/content/103/7/e311

e315

Recurrent thrombosis in patients with antiphospholipid antibodies treated with vitamin K antagonists or rivaroxaban Ida Martinelli et al. http://www.haematologica.org/content/103/7/e315

Case Reports Case Reports are available online only at www.haematologica.org/content/103/7.toc

e318

Anti-PD1 associated fulminant myocarditis after a single pembrolizumab dose: the role of occult pre-existing autoimmunity Nicolas Martinez-Calle et al. http://www.haematologica.org/content/103/7/e318

e322

[18F]Florbetaben PET-CT confirms AL amyloidosis in a patient with Waldenström’s Macroglobulinemia Thomas A. Fox et al. http://www.haematologica.org/content/103/7/e322

e325

Successful treatment of disseminated Rosai-Dorfman disease with siltuximab Hannah Lee et al. http://www.haematologica.org/content/103/7/e325

Comments Comments are available online only at www.haematologica.org/content/103/7.toc

e329

Immune failure, infection and survival in chronic lymphocytic leukemia Kyle R. Crassini et al. http://www.haematologica.org/content/103/7/e329

e330

Immune failure, infection and survival in chronic lymphocytic leukemia in Denmark Michael Asger Andersen et al. http://www.haematologica.org/content/103/7/e330

Haematologica 2018; vol. 103 no. 7 - July 2018 http://www.haematologica.org/


EDITORIALS Linking histone methylation, transcription rates, and stem cell robustness Justin C. Wheat1 and Ulrich Steidl1,2,3,4 1

Department of Cell Biology; 2Department of Medicine; 3Albert Einstein Cancer Center and 4Ruth L. and David S. Gottesman Institute for Stem Cell Research and Regenerative Medicine, Albert Einstein College of Medicine - Montefiore Medical Center, Bronx, NY, USA E-mail: ulrich.steidl@einstein.yu.edu doi:10.3324/haematol.2018.196089

A

s blocked differentiation is a hallmark of most tumors, significant efforts have been made to understand the regulation of gene networks during normal differentiation and how perturbation of these regulatory processes contribute to tumor initiation. An emerging question of interest within this arena is how the core transcriptional machinery and the epigenome interact during transcription. While strong correlations between certain histone marks and transcriptional activity have been found, critical questions remain, including the information encoded in these marks, whether these marks are cause or consequence of polymerase activity, and what functional cost is incurred by stem cells when these marks are perturbed. In this issue of the Journal, Zhou et al.1 study the hematopoietic stem cell defects associated with genetic deletion of the mammalian H3K36 tri-methyltransferase, Setd2. While embryonic deletion of Setd2 within the endothelial/hematopoietic lineages by a Tie2-Cre is lethal, deletion with Vav and the inducible Mx1-Cre systems permitted the study of this critical enzyme during hematopoiesis. The Authors found significant multilineage leukopenia, multiple erythroid dysplasias, and ultimately bone marrow failure in knockout mice. Interestingly, this study found an expansion of early erythroblasts and mature megakaryocytes in the bone marrow, and a macrocytic anemia with thrombocytosis in the peripheral blood, indicating that the erythroid/megakaryocytic lineage specification is at least partially maintained despite Setd2 loss. Indeed, the putative number of preCFU-E (as defined by FACS) and BFU-E (defined by colony assay) were increased in Setd2 knockout mice, indicating that the peripheral anemias do not arise from a reduction in progenitor cells committing to the erythroid lineage. Characterization of the HSC compartment in these mice also demonstrated substantial defects in repopulating capacity in transplantation experiments and suggest that the erythroid-megakaryocytic bias of Setd2 KO cells is intrinsic to HSC. Finally, Setd2 HSC were found to cycle more than wild-type HSC and were more sensitive to challenge with 5-FU. In total, Setd2 appears to play an essential role in normal HSC biology. Next, Zhou et al.1 established the molecular mechanism underlying these dramatic phenotypic findings. First, they found significant changes in both the mRNA and protein levels of other H3K36 methyltransferases, with Ash11 decreasing and Nsd1/2/3 increasing. As these enzymes catalyze the mono- and di-methylation reactions on H3K36, these authors speculated that Nsd enzymes may compete with Setd2 at H3K36. In cell lines, they found that overexpression of Nsd proteins phenocopied the molecular aberrations seen with Setd2 loss of function, consistent with a mutual antagonism between these methyltransferases. Moreover, perturbation of the mono/di- to tri-methylation status of H3K36 also lead to changes in other histone marks correlated with active transcription, such as increased H3K79me2 and H3K4me3, and decreased H3K27me3. As both H3K36 and H3K79 have been strongly linked to polymerase processivity, Zhou et al.1 looked at the distribution and abundance of initiated and actively elongating polymerase, indicated by the presence of haematologica | 2018; 103(7)

phosphorylation marks on the CTD at serine 5 and serine 2, respectively. Intriguingly, these Authors show robust increases in the Ser2P after deletion of Setd2, as well as increased ChIP signal within the exonic region of elongation complex sensitive genes such as Mycc. These changes in polymerase processivity led to specific changes in gene expression, with increased expression of terminal erythroid genes such as Gata1. Finally, pharmacological inhibition of other super elongation complex enzymes, including DOT1L, BRD4, and CDK9, fully reversed the transcriptional and phenotypic changes observed in Setd2 knockout HSC. Prior work integrating Global Run-On sequencing (Gro-Seq) and ChIP-seq of H3K36me3 has demonstrated substantial anticorrelation between this histone mark and polymerase elongation rates.2 The study by Zhou et al.1 is, to our knowledge, the first to demonstrate in vivo evidence for a directional and putatively causal link of this mark with polymerase kinetics. Moreover, this study provides significant evidence for the importance of Setd2, and consequently H3K36me3, to hematopoietic stem cell function. A growing body of literature has demonstrated that the core transcriptional machinery is not a simple digital switch responding to upstream interactions of transcription factors, but rather a highly regulated, multi-step process that behaves differently within different cells and in different contexts (reviewed by Coulon et al.,3 Goodrich and Tjian,4 and Jonkers and Lis5). The study by Zhou et al.1 confirms those results in the context of hematopoiesis and stem cell biology, as well as prompts a number of important questions that must now be resolved. Why are erythroid cells more robust to changes in Setd2 activity? What facilitates crosstalk between the CTD and H3K36 methylation? What biophysical parameters change to increase transcription elongation, and how does polymerase “read� differences in K36 methylation status? And how do different genes differentially respond to perturbations in this single mark? While many questions remain, this study by Zhou et al.1 provides substantial impetus for the continued exploration of the fundamental questions surrounding the core transcriptional machinery, its tissue-specific regulation, and its role in modulating stem cell function.

References 1 Zhou Y, Yan X, Feng X et al. Setd2 regulates quiescence and differentiation of adult hematopoietic stem cells by restricting RNA polymerase II elongation. Haematologica. 2018;103(7):1110-1123. 2 Jonkers I, Kwak H, Lis JT. Genome-wide dynamics of Pol II elongation and its interplay with promoter proximal pausing, chromatin, and exons. Elife. 2014;3:e02407. 3. Coulon A, Chow CC, Singer RH, Larson DR. Eukaryotic transcriptional dynamics: from single molecules to cell populations. Nat Rev Genet. 2013;14(8):572-584. 4. Goodrich JA, Tjian R. Unexpected roles for core promoter recognition factors in cell-type-specific transcription and gene regulation. Nat Rev Genet. 2010;11(8):549-558. 5. Jonkers I, Lis JT. Getting up to speed with transcription elongation by RNA polymerase II. Nat Rev Mol Cell Biol. 2015;16(3):167-177.

1093


Editorials

Complication free survival long-term after hemopoietic cell transplantation in thalassemia Emanuele Angelucci Hematology and Transplant Center, Ospedale Policlinico San Martino, IRCCS, Genoa, Italy E-mail: emnang@tin.it doi:10.3324/haematol.2018.196071

T

halassemia patients have witnessed and have been the protagonists of two extraordinary events in the progress of their treatment during recent decades. Medical therapy (transfusion and regular chelation) has dramatically changed their prognosis from a fatal disease in childhood to a chronic disease into adulthood with an open, undefined prognosis. Hematopoietic cell transplant has changed the paradigm of thalassemia, introducing for the first time in medical history the notion of a cure for this congenital disease. Allogeneic hematopoietic cell transplantation (HCT) has become a standard of care for the cure of transfusiondependent thalassemia patients, with thousands undergoing this curative approach worldwide. Transplant has expanded from the industrialized countries and is now performed in several parts of the world, including those where the disease is most prevalent.1 As transplantation is a curative approach usually performed in childhood, with a limited but not absent risk in a non-neoplastic disease in which prolonged survival (decades) can be achieved with conventional medical therapy, data on long-term real-life complications are essential. In this issue of the Journal, Rahal et al.2 and the French co-operative group2 report on these much needed data. Of the over 134 patients transplanted in the period 1984-2012 (median age at transplant 5.9 years, interquartile range 3-11 years) in 21 French centers, 107 were alive and well two years after HCT; 2 subsequently died from chronic graft-versus-host disease (GvHD) and 6 were later lost to follow up. The remaining 99 patients were part of these detailed analyses on long-term complications. Almost all patients had been transplanted in childhood after a myeloablative conditioning and almost all had received hematopoietic cells from an HLA identical sibling. After a median follow up of 12 years (interquartile range 7-19 years) 11% of these patients presented thyroid dysfunction, 5% diabetes, and 2% cardiac failure. Hypogonadism was present in 56% of females and 14% of males. As expected, females who experienced normal puberty were younger at transplant compared to those who experienced delayed puberty. Almost half of the females aged 20 years or over had spontaneous and successful pregnancy after transplantation, confirming another single center report.3 Interestingly, no secondary cancer, delayed graft failure with thalassemia recurrence or transplant-related mortality were registered. As correctly reported by the Authors, and as also expressed in other recent reports,4 the issue of secondary cancers probably requires a longer follow up to be confirmed. However, the limited incidence of chronic GvHD allows for some slight optimism in this setting. The data regarding absence of long-term thalassemia 1094

recurrence do not confirm recent isolated case reports (EP Alessandrino and C Giardini, 2018, personal communications), but even in this case, longer follow up would probably be necessary. Most patients have been successfully treated for iron overload even if the suggested target of normal transferrin saturation has not been completely reached in all of them.5 The great strength of this report is that it includes 73% of the transplanted patients and 93% of the patients who have survived for at least two years in the multi-center (21 transplant centers) experience of an entire large nation and, therefore, provides reliable real-life data on a population judged suitable for transplantation in the years indicated without other forms of selection. Moreover, conditioning, age at transplant, and follow-up methods were substantially uniform for almost all the patients. This report provides the most uniform world-wide long-term analyses from the entire transplant activity of one country. Another great strength of this work is that the Authors successfully managed to separate iron overload thalassemia-related long-term complication from transplantrelated complication. Because of the homogeneous donor selection, and the very limited incidence of chronic GvHD, the reported transplant-related complication could likely be related to the intensity of the conditioning regimen, and mainly to the use of high doses of the alkylating agent busulfan. This makes this analysis likely predictive for long-term complication in gene therapy programs in hemoglobinopathies6 requiring myeloablative preparative regimens. There are three unavoidable limitations to a general use of these data. One is that patients were transplanted at a very young age (following the indications of the Pesaro Group), and it is likely that patients transplanted at an older age would present many more long-term thalassemia-related complications.7,8 Moreover, patients reported here were transplanted more than a decade ago with a standard myeloablative regimen. Today, less toxic preparative regimens have been developed,9-11 and their long-term complications could be different and, hopefully, of a lower grade. Finally, 91% of patients transplanted in France were transplanted from an HLA identical sibling, and therefore long-term complications of transplants from different donor types could not be completely represented in this research. Nevertheless, taking into account the above reported limitations, these data can be used to set up rational screening programs for patients transplanted in childhood. There are several important considerations arising from this article. haematologica | 2018; 103(7)


Editorials

Figure 1. Factors to be considered in deciding hematopoietic cell transplant versus gene therapy. GvHD: graft-versus-host disease.

1) Once again it has been confirmed that transplantation in thalassemia should be performed as soon as possible, not only to maximize transplant outcome, but also to minimize long-term complications. The data on spontaneous puberty make this issue very clear.2 2) Long-term complications are less than those expected in medically-treated thalassemia12 and, with the exception of hypogonadism, are of limited incidence with a quality of life similar to that of a matched normal population.13 3) Several reports of spontaneous maternities have been published, but this is the first report with epidemiology data including probability of maternity and paternity after an analysis of almost the entire transplanted population. 4) For the first time, the issue of adolescents being overweight is reported post transplant for a congenital disease. Clearly this observation requires more in-depth analysis and prospective dedicated studies with comparison to normal population data. 5) The problem of iron overload in the post-transplant setting has been resolved, and most patients can easily achieve normal iron burden5 thus avoiding long-term iron toxicity complications.14 6) Like iron overload, other complications related to thalassemia can be treated after transplantation, such as hepatitis C and B virus infection, with the therapies available today. 7) Specific follow-up guidelines and screening recommendations can be proposed and can be used specifically for this category of patients in order to prevent / cure the complications that are known today.15,16 8) Recent data on the emerging gene therapy approach clearly indicate that a myeloablative preparative regime is necessary to allow gene-modified autologous stem cell engraftment. Therefore, if the long follow up confirms current safety data on the cellular product,17 complications relating to the conditioning regimen, such as those reported here, can be foreseen and similar screening programs can be set up. haematologica | 2018; 103(7)

Lastly, this article provides a further contribution to the debate on the therapeutic decision-making process for thalassemia patients. This debate, which has so far been limited to medical therapy and transplantation, will soon be carried forward by another great innovation in the treatment of this disease: gene therapy.17 Even if this approach has so far only demonstrated complete clinical effectiveness in non β0/β0 thalassemia, the door has been left ajar and will certainly be thrown wide open soon. Regardless of the problem of gene therapy costs, which will probably significantly compromise its wide applicability, these data reporting long-term limited complication and no cancer after HCT can be of enormous help in establishing the correct therapeutic approach. As discussed by many authors, transplantation, medical therapy, and now gene therapy, is an individual and highly personal decision.18 Because of this, there have been no randomized prospective trials and it is likely that none will ever be performed. Several different medical, personal, and socio-economic aspects must be considered, taking into account the impressive epidemiology and the social conditions of countries were this disease is prevalent.19 Figure 1 reports the factors to be considered today in approaching transplantation or gene therapy for the cure of transfusion-dependent thalassemia. In my personal opinion, this report by Rahal et al.2 reinforces the position of transplantation in this difficult decision-making process, particularly in the case of young children when an HLA identical sibling donor is available.

References 1 Baronciani D, Angelucci E, Potschger U, et al. Hemopoietic stem cell transplantation in thalassemia: a report from the European Society for Blood and Bone Marrow Transplantation Hemoglobinopathy Registry, 2000-2010. Bone Marrow Transplant. 2016;51(4):536-541. 2. Rahal I, Galambrun C, Bertrand Y, et al. Late effects after hematopoietic stem cell transplantation for beta-thalassemia major: the French national experience. Haematologica. 2018;103(7):1143-1149.

1095


Editorials 3. Santarone S, Natale A, Olioso P, et al. Pregnancy outcome following hematopoietic cell transplantation for thalassemia major. Bone Marrow Transplan. 2017;52(3):388-393. 4. Santarone S, Pepe A, Meloni A, et al. Secondary solid cancer following hematopoietic cell transplantation in patients with thalassemia major. Bone Marrow Transplant. 2018;53(1):39-43. 5. Angelucci E, Pilo F. Management of iron overload before, during, and after hematopoietic stem cell transplantation for thalassemia major. Ann NY Acad Sci. 2016;1368(1):115-121. 6. Ferrari G, Cavazzana M, Mavilio F. Gene Therapy Approaches to Hemoglobinopathies. Hematol Oncol Clin North Am. 2017;31(5): 835-852. 7. Lucarelli G, Clift RA, Galimberti M, et al. Marrow transplantation for patients with thalassemia: results in class 3 patients. Blood. 1996;87 (5):2082-2088. 8. Angelucci E, Pilo F, Coates TD. Transplantation in thalassemia: Revisiting the Pesaro risk factors 25 years later. Am J Hematol. 2017;92 (5):411-413. 9. Bernardo ME, Piras E, Vacca A, et al. Allogeneic hematopoietic stem cell transplantation in thalassemia major: results of a reduced-toxicity conditioning regimen based on the use of treosulfan. Blood. 2012;120(2):473-476. 10. Poomthavorn P, Chawalitdamrong P, Hongeng S, et al. Gonadal function of beta-thalassemics following stem cell transplantation conditioned with myeloablative and reduced intensity regimens. J Pediatr Endocrinol Metab. 2013;26(9-10):925-932. 11. Anurathapan U, Pakakasama S, Mekjaruskul P, et al. Outcomes of thalassemia patients undergoing hematopoietic stem cell transplantation by using a standard myeloablative versus a novel reduced-toxicity conditioning regimen according to a new risk stratification. Biol Blood Marrow Transplant. 2014;20(12):2066-2071.

12. Borgna-Pignatti C, Rugolotto S, De Stefano P, et al. Survival and complications in patients with thalassemia major treated with transfusion and deferoxamine. Haematologica. 2004;89(10):1187-1193. 13. La Nasa G, Caocci G, Efficace F, et al. Long-term health-related quality of life evaluated more than 20 years after hematopoietic stem cell transplantation for thalassemia. Blood. 2013;122(13):2262-2270. 14. Coates TD, Carson S, Wood JC, Berdoukas V. Management of iron overload in hemoglobinopathies: What is the appropriate target iron level? Ann NY Acad Sci. 2016;1368(1):95-106. 15. Shenoy S, Angelucci E, Arnold SD, et al. Current Results and Future Research Priorities in Late Effects after Hematopoietic Stem Cell Transplantation for Children with Sickle Cell Disease and Thalassemia: A Consensus Statement from the Second Pediatric Blood and Marrow Transplant Consortium International Conference on Late Effects after Pediatric Hematopoietic Stem Cell Transplantation. Biol Blood Marrow Transplant. 2017;23(4):552-561. 16. Shenoy S, Gaziev J, Angelucci E, et al. Late Effects Screening Guidelines after Hematopoietic Cell Transplantation (HCT) for Hemoglobinopathy: Consensus Statement From the Second Pediatric Blood and Marrow Transplant Consortium International Conference on Late Effects after Pediatric HCT. Biol Blood Marrow Transplant. 2018 Apr 10. [Epub ahead of print]. 17. Thompson AA, Walters MC, Kwiatkowski J, et al. Gene Therapy in Patients with Transfusion-Dependent beta-Thalassemia. N Engl J Med. 2018;378(16):1479-1493. 18. Angelucci E. Hematopoietic stem cell transplantation in thalassemia. Hematology Am Soc Hematol Educ Program. 2010;2010:456-462. 19. Modell B, Darlison M. Global epidemiology of haemoglobin disorders and derived service indicators. Bulletin of the World Health Organization. 2008, p. 480-7.

Still a role for second-line chemoimmunotherapy in chronic lymphocytic leukemia? Jennifer R Brown CLL Center, Division of Medical Oncology, Dana-Farber Cancer Institute and Harvard Medical School, Boston, MA, USA E-mail: jennifer_brown@dfci.harvard.edu doi:10.3324/haematol.2018.196014

I

n this issue of the Journal, Cuneo et al. report a retrospective observational study of the efficacy of bendamustine rituximab (BR) given as first salvage therapy for chronic lymphocytic leukemia (CLL) patients within the GIMEMA and ERIC networks.1 Among 237 patients, the median progression-free survival (PFS) was an excellent 25 months and the median time to next treatment 31.3 months. Predictors of shorter PFS in multivariable analysis included del(17p), unmutated IGHV and advanced stage. Cuneo et al. further performed a matched adjusted comparison of overall survival (OS) between the subset of BR-treated patients without del(17p) who had received front-line chemoimmunotherapy (CIT), and similar patients who had received ibrutinib second-line in named patient programs in the UK and Italy. Interestingly, there was no difference in OS, with 63% alive in the ibrutinib group at 36 months, as compared to 74.4% in the BR group.1 At first glance these data may seem surprising, as ibrutinib has had OS benefit in both the RESONATE trial,2 comparing ibrutinib with ofatumumab in relapsed refractory CLL, and in the RESONATE-2 trial,3 comparing ibrutinib with chlorambucil in previously untreated CLL. It is important to note that the control arms of both of these trials did not unfortunately represent particularly effective 1096

therapy, especially in comparison to the BR presented here; they were also both relatively small studies. A US Intergroup trial comparing ibrutinib to ibrutinib rituximab to BR for front-line therapy of CLL in older patients has completed accrual and results are awaited. While these data from randomized trials are invaluable, they often do not capture the full picture of a new therapy, hence the value of observational studies like that of Cuneo et al.1 Eligibility, particularly for phase III trials, is typically strict, resulting in a selected healthy patient population. This may be particularly true of the ibrutinib randomized studies. For example, although the median age of patients receiving ibrutinib in RESONATE-2 was 73 years, only 31% had a cumulative illness rating score (CIRS) over 6, indicating a very low level of comorbidity for their age.3 Why does this matter? Although ibrutinib is often said to be well-tolerated among older patients with comorbidities, the data supporting this claim are actually quite limited, and a recent multicenter retrospective study has found that a CIRS score of over 6 was in fact associated with inferior event-free survival and OS, as well as increased risk of dose reduction or discontinuation, among ibrutinib-treated patients.4 Other real-world analyses with ibrutinib, as well as longer follow up of the prospective trials, have also made haematologica | 2018; 103(7)


Editorials

it clear that many patients do not tolerate extended therapy. A multicenter retrospective analysis of 616 ibrutinibtreated patients reported a 41% discontinuation rate with a median time to ibrutinib discontinuation of seven months.5 The predominant cause of discontinuation was toxicity, including atrial fibrillation, arthralgia, rash, infection and pneumonitis. The median PFS for the entire cohort of predominantly relapsed patients was 36 months, as compared to the recently reported 51-month median PFS in the phase Ib/II ibrutinib clinical trial.6 Additionally, with more mature follow up of the ibrutinib clinical trials, discontinuation rates are rising and are beginning to look more similar to the earlier real-world data. In the admittedly small previously untreated cohort of the phase Ib/II study,6 the discontinuation rate has reached 45% at five years, although PFS is 92%, indicating that these discontinuations are not predominantly due to progressive disease. More frequent discontinuation, dose holds and dose reductions seem likely to explain some of the differences in outcome between ibrutinib clinical trials and real-world reports. The real-world analysis presented by Cuneo et al.1 focuses on an interesting CLL patient niche that has perhaps been neglected, namely those receiving first salvage therapy. These patients are often pooled with more heavily pre-treated patients, making it difficult to assess their outcomes. The BR population in this study has a median age of 70 years and is reasonably healthy based on Eastern Cooperative Oncology Group (ECOG) performance status, comorbidities and creatinine clearance.1 Their disease, however, is advanced, with 78.6% in advanced stage, 91% with bulky lymphadenopathy, and 73% with unmutated IGHV. High-risk FISH is relatively limited at 20.8% with del(11q) and 12.6% with del(17p), although approximately half had progressed within 36 months of their prior therapy. In comparison to the ibrutinib arm of RESONATE, for example, these patients are much less heavily pre-treated and at much lower risk according to FISH, yet have more advanced disease according to stage and bulk. Despite the latter, and likely related to the former, they did well with BR, with a median PFS of 25 months that compares favorably to the previously reported 15-18month PFS for BR in the first or second salvage settings.7,8 As expected, PFS was worse for genetically higher risk patients, but reached 40.4 months in those without del(17p), with mutated IGHV, and Rai stage 0-2 disease. The overall analysis, and particularly the comparison to ibrutinib, is certainly limited by its retrospective nature and potential differences in the patient populations under comparison. For the comparative analysis, Cuneo et al.1 wisely chose to focus on patients without del(17p) who had had front-line CIT, and in doing so the cohorts were statistically comparable for age, ECOG, response to firstline therapy, and IGHV, although the ibrutinib cohort still had more patients with less than 36 months from first-line therapy (76.1% vs. 59.1%). Furthermore, the ibrutinib cohort also showed a trend to more ECOG-2 patients: 17.4%, compared to 8.1% in the BR cohort. The former had an estimated 2-year OS of 35%, much less than the 73% OS of ibrutinib-treated ECOG 0-1 patients. In this context, a detailed listing of the causes of death among the ibrutinib-treated patients would be helpful. In addition, haematologica | 2018; 103(7)

the ibrutinib cohort in particular is quite small, and several early deaths due to infection or Richter’s syndrome, of unclear relationship to ibrutinib, may also have affected the OS curve. While these differences certainly confound the results, nonetheless the findings are provocative in demonstrating comparable OS between second-line ibrutinib and BR in two approximately matched real-world patient populations. Despite its limitations, the Cuneo et al.1 study demonstrates that six months of CIT can be very effective second-line therapy in appropriate patients, and challenges the increasingly widespread belief that if ibrutinib is not used first line, it should certainly be used second line. Ibrutinib’s initial overwhelming efficacy was evident particularly among very heavily pre-treated patients with 17p and 11q deletions9 whose response to traditional CIT is dismal. The Authors of this paper note that the real-world data with ibrutinib show similar duration of therapy and benefit in first versus later relapses, suggesting that relative ibrutinib benefit in the real world is greater in more heavily pre-treated patients. I would agree that the still limited data currently available generally support this claim, although the relative proportion of discontinuations for disease progression is higher in later line patients, and it is hard to ignore the clinical trial results demonstrating longer PFS in less heavily pre-treated patients, at least in the relapsed setting.10 In a cross-trial comparison performed with 24-month follow up and excluding del(17p) patients, the PFS was similar for first-line patients in RESONATE-2 to that of second-line patients in RESONATE,11 but better than that of later line RESONATE patients. A further issue raised by the clinical trial data is whether ibrutinib or idelalisib should be added if later-line BR were given, since the addition of either drug improved PFS and possibly OS among a more heavily pre-treated higher risk patient population.12,13 Further complicating this landscape are the recently reported MURANO data in which venetoclax-rituximab greatly improved PFS compared to BR in relapsed CLL patients, most of whom had one prior therapy, albeit with a higher risk profile including del(17p) and TP53 mutation.14 How then to reconcile the clinical trial and real-world data, while also taking into account patient preference for time-limited therapy and cost considerations? Notably absent from this discussion has been a deeper focus on risk stratification as well as individualized patient management, including comorbidities and performance status, in selecting therapy. The relative benefit of ibrutinib and other novel agents is certainly greatest in higher risk disease, and, at least with ibrutinib, among patients able to remain on drug for extended times. Yet to date, only del(17p) has been widely accepted as altering therapy choice and trial design, with the result that most of our CLL trials enroll a broad patient population which may not be well stratified for disease risk or directly comparable to the population enrolled on other studies or seen in our clinics. The most important example here is IGHV mutation status, which clearly predicts long-term benefit from FCR CIT,15-17 yet has not been widely incorporated into our thinking about relapsed or older patients, despite evidence, as in this paper,1 that the mutated subgroup can often respond well to a diversity of therapies. 1097


Editorials

Additionally, the negative impact of genomic complexity as measured by complex karyotype18 or multiple driver mutations19,20 has become increasingly clear, and these features would likely add further nuance to a patient-specific risk stratification. Increasingly, I take into account all of these disease features, as well as the patient’s age, general health, preferences and reason for needing therapy, when considering their therapeutic options. The time has come for clinical trials and long-term population-based studies informed by this deeper risk stratification in order to understand the natural history of sequential therapy choices in these increasingly differentiated unique patient subgroups. Only in this way can we best serve and advise our patients among the myriad of choices.

References 1 Cuneo A, Follows G, Rigolin GM, et al. Efficacy of bendamustine and rituximab as first salvage treatment in chronic lymphocytic leukemia and indirect comparison with ibrutinib: a GIMEMA, ERIC and UK CLL FORUM study. Haematologica. 2018;103(7):1209-1217. 2. Byrd JC, Brown JR, O'Brien S, et al. Ibrutinib versus ofatumumab in previously treated chronic lymphoid leukemia. N Engl J Med. 2014;371(3):213-23. 3. Burger JA, Tedeschi A, Barr PM, et al. Ibrutinib as Initial Therapy for Patients with Chronic Lymphocytic Leukemia. N Engl J Med. 2015;373(25):2425-2437. 4. Gordon M, Churnetski M, Alqahtani H, et al. Comorbidities Predict Inferior Outcomes in Chronic Lymphocytic Leukemia Treated With Ibrutinib. Cancer. 2018 [In press] 5. Mato AR, Nabhan C, Thompson MC, et al. Toxicities and outcomes of 616 ibrutinib-treated patients in the United States: a real-world analysis. Haematologica. 2018;103(5):874-879. 6. O'Brien S, Furman RR, Coutre S, et al. Single-Agent Ibrutinib in Treatment-Naive and Relapsed/Refractory Chronic Lymphocytic Leukemia: A 5-Year Experience. Blood. 2018;131(17):1910-1919. 7. Fischer K, Cramer P, Busch R, et al. Bendamustine combined with rituximab in patients with relapsed and/or refractory chronic lymphocytic leukemia: a multicenter phase II trial of the German Chronic Lymphocytic Leukemia Study Group. J Clin Oncol. 2011;29(26):35593566.

1098

8. Fornecker LM, Aurran-Schleinitz T, Michallet AS, et al. Salvage outcomes in patients with first relapse after fludarabine, cyclophosphamide, and rituximab for chronic lymphocytic leukemia: the French intergroup experience. Am J Hematol. 2015;90(6):511-514. 9. Byrd JC, Furman RR, Coutre SE, et al. Targeting BTK with ibrutinib in relapsed chronic lymphocytic leukemia. N Engl J Med. 2013;369(1):3242. 10. Brown JR, Hillmen P, O'Brien S, et al. Extended follow-up and impact of high-risk prognostic factors from the phase 3 RESONATE study in patients with previously treated CLL/SLL. Leukemia. 2018;32(1):83-91. 11. O'Brien S, Byrd J, Hillmen P, et al. Outcomes with ibrutinib by line of therapy in patients with CLL: Analyses from phase III data. J Clin Oncol. 2016;34(suppl):Abstr 7520. 12. Chanan-Khan A, Cramer P, Demirkan F, et al. Ibrutinib combined with bendamustine and rituximab compared with placebo, bendamustine, and rituximab for previously treated chronic lymphocytic leukaemia or small lymphocytic lymphoma (HELIOS): a randomised, double-blind, phase 3 study. Lancet Oncol. 2016;17(2):200-211. 13. Zelenetz AD, Barrientos JC, Brown JR, et al. Idelalisib or placebo in combination with bendamustine and rituximab in patients with relapsed or refractory chronic lymphocytic leukaemia: interim results from a phase 3, randomised, double-blind, placebo-controlled trial. Lancet Oncol. 2017;18(3):297-311. 14. Seymour JF, Kipps TJ, Eichhorst B, et al. Venetoclax-Rituximab in Relapsed or Refractory Chronic Lymphocytic Leukemia. N Engl J Med. 2018;378(12):1107-1120. 15. Fischer K, Bahlo J, Fink AM, et al. Long-term remissions after FCR chemoimmunotherapy in previously untreated patients with CLL: updated results of the CLL8 trial. Blood. 2016;127(2):208-215. 16. Rossi D, Terzi-di-Bergamo L, De Paoli L, et al. Molecular prediction of durable remission after first-line fludarabine-cyclophosphamide-rituximab in chronic lymphocytic leukemia. Blood. 2015;126(16):19211924. 17. Thompson PA, Tam CS, O'Brien SM, et al. Fludarabine, cyclophosphamide, and rituximab treatment achieves long-term disease-free survival in IGHV-mutated chronic lymphocytic leukemia. Blood. 2016;127(3):303-309. 18. Thompson PA, O'Brien SM, Wierda WG, et al. Complex karyotype is a stronger predictor than del(17p) for an inferior outcome in relapsed or refractory chronic lymphocytic leukemia patients treated with ibrutinib-based regimens. Cancer. 2015;121(20):3612-3621. 19. Landau DA, Tausch E, Taylor-Weiner AN, et al. Mutations driving CLL and their evolution in progression and relapse. Nature. 2015;526(7574): 525-530. 20. Puente XS, Bea S, Valdes-Mas R, et al. Non-coding recurrent mutations in chronic lymphocytic leukaemia. Nature. 2015;526(7574):519-524.

haematologica | 2018; 103(7)


REVIEW ARTICLE

Dissecting the pathophysiology of immune thrombotic thrombocytopenic purpura: interplay between genes and environmental triggers

Ferrata Storti Foundation

Johana Hrdinová,1,2,3 Silvia D’Angelo,4,5 Nuno A. G. Graça,1,6 Bogac Ercig,1,2,3 Karen Vanhoorelbeke,4 Agnès Veyradier,7,8 Jan Voorberg 1and Paul Coppo8,9,10

Department of Plasma Proteins, Sanquin-Academic Medical Center Landsteiner Laboratory, Amsterdam, the Netherlands; 2PharmaTarget B.V., Maastricht, the Netherlands; 3Department of Biochemistry, Cardiovascular Research Institute Maastricht (CARIM), Maastricht University, the Netherlands; 4Laboratory for Thrombosis Research, IRF Life Sciences, KU Leuven Campus Kulak Kortrijk, Belgium; 5Protobios LLC, Tallinn, Estonia; 6Icosagen Cell Factory OÜ, Ülenurme Vald, Tartumaa, Estonia; 7 Service d'Hématologie Biologique and EA3518, Groupe Hospitalier Saint LouisLariboisière, Assistance Publique - Hôpitaux de Paris, Université Paris Diderot, France; 8 Centre de Référence des Microangiopathies Thrombotiques, Hôpital Saint-Antoine, APHP, Paris, France; 9Service d’Hématologie, Assistance Publique – Hôpitaux de Paris, France and 10Sorbonne Université, UPMC Univ Paris 06, France 1

Haematologica 2018 Volume 103(2):1099-1109

ABSTRACT

A

lthough outstanding progress has been made in understanding the pathophysiology of thrombotic thrombocytopenic purpura (TTP), knowledge of the immunopathogenesis of the disease is only at an early stage. Anti-ADAMTS13 auto-antibodies were shown to block proteolysis of von Willebrand factor and/or induce ADAMTS13 clearance from the circulation. However, it still remains to identify which immune cells are involved in the production of antiADAMTS13 autoantibodies, and therefore account for the remarkable efficacy of the B-cell depleting agents in this disease. The mechanisms leading to the loss of tolerance of the immune system towards ADAMTS13 involve the predisposing genetic factors of the human leukocyte antigen class II locus DRB1*11 and DQB1*03 alleles as well as the protective allele DRB1*04, and modifying factors such as ethnicity, sex and obesity. Future studies have to identify why these identified genetic risk factors are also frequently to be found in the healthy population although the incidence of immune-mediated thrombotic thrombocytopenic purpura (iTTP) is extremely low. Moreover, the development of recombinant ADAMTS13 opens a new therapeutic era in the field. Interactions of recombinant ADAMTS13 with the immune system of iTTP patients will require intensive investigation, especially for its potential immunogenicity. Better understanding of iTTP immunopathogenesis should, therefore, provide a basis for the development of novel therapeutic approaches to restore immune tolerance towards ADAMTS13 and thereby better prevent refractoriness and relapses in patients with iTTP. In this review, we address these issues and the related challenges in this field.

Introduction Thrombotic thrombocytopenic purpura (TTP) is a devastating disease resulting from a severe deficiency in the von Willebrand factor (VWF)-cleaving protease ADAMTS13. This deficiency causes the accumulation of ultra-large VWF multimers in the circulation and the formation of thrombi in the microvasculature under high shear stress conditions. When left untreated, these microthrombi cause multi-organ failure and lead to death. In the acquired immune-mediated form of haematologica | 2018; 103(7)

Correspondence: paul.coppo@aphp.fr

Received: January 11, 2018. Accepted: April 13, 2018. Pre-published: April 19, 2018. doi:10.3324/haematol.2016.151407 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/103/7/1099 ©2018 Ferrata Storti Foundation Material published in Haematologica is covered by copyright. All rights are reserved to the Ferrata Storti Foundation. Use of published material is allowed under the following terms and conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode. Copies of published material are allowed for personal or internal use. Sharing published material for non-commercial purposes is subject to the following conditions: https://creativecommons.org/licenses/by-nc/4.0/legalcode, sect. 3. Reproducing and sharing published material for commercial purposes is not allowed without permission in writing from the publisher.

1099


J. HrdinovĂĄ et al.

TTP (iTTP), patients develop antibodies (Abs) against ADAMTS13 that enhance its clearance or inhibit its VWF processing activity.1 Therapeutic plasma exchange (TPE) greatly improved the fatal outcome of iTTP leading to survival rates of more than 80%.1 As iTTP is an autoimmune disease, steroids were used together with TPE.1 Over the last few years, the use of the B-cell depleting agent rituximab (MabtheraÂŽ, Roche) as a more targeted immunomodulator led to a reduction in TPE duration and to efficient prevention of 1-year relapses. The pre-emptive administration of rituximab is also increasingly used in patients with a persistently severe acquired ADAMTS13 deficiency, and otherwise in remission, to prevent long-term relapses.2-4 However, the efficacy of rituximab is only transient, and in up to 50% of cases, additional courses of rituximab are required to maintain a detectable ADAMTS13 activity. Moreover, 10-15% of patients are primarily unresponsive to rituximab, or experience a subsequent refractoriness after an initial response.3 The pathophysiological mechanisms underlying these different scenarios in iTTP, as well as the specific B- and T-cell and plasmacytic subpopulations involved in the reoccurrence of anti-ADAMTS13 Abs after rituximab, still remain unknown. Besides the increasing use of immunomodulators, a recombinant form of ADAMTS13 has passed through a phase I clinical trial5 and should soon be available for the treatment of the congenital form of TTP. The use of recombinant ADAMTS13 in iTTP to over-ride inhibitory Abs, in combination with the antiVWF nanobody caplacizumab, probably represents the next breakthrough in the management of this disease.6 However, the use of a recombinant ADAMTS13 in iTTP may involve the potential risk of boosting inhibitor titers by the activation of ADAMTS13 specific memory B and T cells, as suggested previously.7 Taking into consideration these challenges, it is crucial to understand in more detail the mechanisms leading to the loss and re-establishment of self-tolerance of the immune system towards ADAMTS13. In this review, we address the current knowledge on the immunopathogenesis of iTTP and the potential forthcoming challenges in the field. We also provide evidence that iTTP represents an illustrative model of multistep disease resulting from the combination of genetic risk factors for autoimmunity and environmental precipitating factors.

Anti-ADAMTS13 Abs: physical and functional features, mechanisms of pathogenicity The presence of anti-ADAMTS13 Abs in plasma of iTTP patients with an inhibitory activity towards the VWF cleavage activity of normal plasma was first demonstrated by isolating IgGs via protein A-Sepharose, as well as on protein G-Sepharose column chromatography.8,9 Isolated IgGs were later shown to bind to ADAMTS13 in ELISA.10 It is now known that anti-ADAMTS13 Abs result in a profound deficiency in ADAMTS13 activity by two main mechanisms: inhibitory (neutralizing) Abs block the proteolytic activity of ADAMTS13 towards VWF, whereas non-inhibitory Abs increase ADAMTS13 clearance from the circulation by forming immune-complexes.11-13 ADAMTS13 antigen (Ag) levels are decreased in most patients, suggesting that Ab-mediated ADAMTS13 depletion is an important pathogenic mechanism underlying severe loss of enzyme activity.11-13 It is 1100

likely that both inhibitory and non-inhibitory Abs promote ADAMTS13 clearance. Inhibitory IgGs that account for the majority of auto-Abs found in patients with iTTP are mainly directed against the spacer domain of ADAMTS13, while those solely targeting the carboxyterminal domains are in vitro non-inhibitory.14 Whether immune complexes may in addition activate the complement system and bind to cellular Fc receptors, thereby promoting inflammation, endothelial activation with thrombosis, and relapse, represent attractive mechanisms requiring further exploration.15-17 Retrospective studies based on large series of iTTP patients have shown that, during the first acute event, anti-ADAMTS13 Abs of IgG type (detected either by functional or immunological methods) were identified in approximately 75-90% of patients.18-27 Of all subclasses of anti-ADAMTS13 IgG detected in patients with iTTP, IgG4 was the most prevalent (90%) followed by IgG1 (53%), IgG2 (50%), and IgG3 (33%). There was an inverse correlation between the frequency and abundance of IgG4 and IgG1 Abs.17,28 Anti-ADAMTS13 Abs of IgM and IgA type are found in only approximately 10% of iTTP patients, mainly in association with IgG.21,29,30 Interestingly, anti-ADAMTS13 IgG have been reported to be present in 5% of healthy individuals, possibly sharing some linear epitopes with iTTP patients.21,22,31 Importantly, these Abs in healthy individuals are noninhibitory against ADAMTS13, which is likely due to a lower affinity towards the protein.31 However, whether individuals with non-pathogenic anti-ADAMTS13 Abs are prone to develop iTTP still has to be established. It is nevertheless tempting to speculate that non-pathogenic anti-ADAMTS13 IgG Abs could precede clinical disease onset in iTTP, as reported for other autoimmune diseases such as systemic lupus erythematosus (SLE).32 In this scenario, the mechanisms turning non-pathogenic antiADAMTS13 Abs into pathogenic Abs could involve an enhanced frequency of somatic hypermutation of IgG memory B cells, as well as epitope spreading, underlining the need for future studies in this direction. It would also be of interest to assess whether non-pathogenic antiADAMTS13 Abs occur more frequently in healthy individuals carrying the HLA susceptibility alleles DRB1*11 and DQB1*03, and if these individuals are more prone to develop inhibitory autoantibodies from non-pathogenic antibodies. In a study including 160 patients with miscellaneous diseases but no severe ADAMTS13 deficiency,21 anti-ADAMTS13 Abs of IgG type were also found in 20% of patients with a thrombotic microangiopathy (TMA) other than iTTP, in 8% of patients with various causes of thrombocytopenia, in 13% of patients with SLE, and in 5% of patients with the anti-phospholipid syndrome (APS). In addition, anti-ADAMTS13 Abs of IgM type were also detected in 18% of SLE and APS patients.21 Indeed, conventional inhibitor and ELISA assays allow for assessment of anti-ADAMTS13 Abs in a significant proportion of patients with various TMAs. Therefore, attribution of the mechanistic role and clinical significance of antiADAMTS13 Abs alone remains challenging.21 The Abs directed against ADAMTS13 spacer domain use the heavy chain (VH) gene segment VH1-69 in 75% of cases.14,33,34 The clinical significance of the restricted VH169 germline gene segment has been observed in neutralizing Abs directed toward a highly conserved region in the haematologica | 2018; 103(7)


Immune TTP: interplay between haplotypes & environment

hemagglutinin ectodomain of the influenza virus35 and in patients with B-cell lymphoma after chronic hepatitis C infection.36 Sequence analysis of anti-ADAMTS13 IgGs revealed unique heavy-chain complementary determining region 3 motifs, of which some were shared by unrelated patients, suggesting that the autoimmune response in iTTP is antigen-driven.37 Several animal models of iTTP have been developed by directly injecting ADAMTS13 inhibiting Abs.38-41 Importantly, some of these models highlight the role of auto-Abs in the pathology of iTTP, as administration of the antibody alone (passive transfer) was enough to trigger the hallmarks of iTTP, as described for the baboon model.38

Clinical relevance of anti-ADAMTS13 Abs in iTTP treatment The identification of anti-ADAMTS13 antibodies as the main mechanism of ADAMTS13 deficiency in iTTP prompted an evaluation of B-cell depleting therapies in the management of the disease. Several studies showed that rituximab at the acute phase of the disease, in association with TPE and steroids, shortens the time to response.2,33 Upon treatment with rituximab, TPE duration usually does not exceed 30 days, whereas before the era of rituximab, 25% of patients required TPE for more than one month.2,42 Moreover, the administration of rituximab during the acute phase results in fewer relapses in the 12-18 months following remission. Beyond this time, however, relapses can reoccur as a consequence of the reappearance of anti-ADAMTS13 Abs along with peripheral B-cell recovery.2,42,43 The persistence of a severe ADAMTS13 deficiency following the acute phase has been associated with a high risk of relapse,3,29,44 and this has led investigators to propose pre-emptive B-cell depleting strategies. Although based on comparative studies, this strategy proved its efficiency in protecting patients from fullblown relapses by maintaining ADAMTS13 activity within normal values.3 This strategy implies, therefore, a regular assessment of ADAMTS13 activity during long-term follow up.

Prognostic value of anti-ADAMTS13 Abs The role of anti-ADAMTS13 Abs in iTTP prognosis has still not been fully defined. Anti-ADAMTS13 Abs with a strong inhibitory activity were typically associated with more plasma volume to achieve remission and with an increased risk of relapse, as well as with a lower survival rate in some studies (but not all).19,27,29,45 Also, the combination of several anti-ADAMTS13 Abs isotypes (IgG, IgM, and IgA), including very high IgA titers, and the presence of high levels of IgG1 combined with undetectable levels of IgG4, has been associated with poor prognosis and an increased risk of mortality.17,28 Recently, higher mortality rates were found in individuals with higher antiADAMTS13 Abs and lower ADAMTS13 antigen levels; patients with anti-ADAMTS13 IgG Abs in the upper quartile and ADAMTS13 antigen in the lowest quartile had the highest mortality (27.3%).46 Epitope mapping studies have shown that patients typically display auto-Abs against the cysteine-rich/spacer domains of ADAMTS13, while more than 50% of patients have auto-Abs against the CUB domains or the TSP2-8 domains.47 A major antibody epitope in the spacer domain has been resolved at the amino acid level; residues haematologica | 2018; 103(7)

R568, F592, R659, R660, Y661 and Y665 in the spacer exosite-3 constitute a conformational epitope that is targeted by the majority of anti-spacer Abs.30,48-52 Auto-Abs directed toward the terminal TSP2-8 repeats (+/- CUB domains) have been associated with lower platelet counts;53 however, others have found no correlation between these parameters, suggesting that the “specificity profile” of patients’ auto-Abs is not a major determinant of their pathophysiology.13 Furthermore, the “specificity profile” of these auto-Abs may change with time (“epitope spreading”), even in patients subjected to immunomodulatory treatments. Importantly, anti-ADAMTS13 Abs have been shown to persist during iTTP remission, either in free form or within immune complexes, in spite of a detectable ADAMTS13 activity.16

Human leukocyte antigen molecules as predisposing factors to iTTP; from serotype to immunochip Immune-mediated thrombotic thrombocytopenic purpura occurs after the development of a specific adaptive immune response targeting ADAMTS13. The human leukocyte antigen (HLA) system has an important role as a genetic risk factor in autoimmune diseases, and a similar association could thus be assumed for iTTP, especially considering cases of iTTP in siblings.54,55 The presence of class-switched high-affinity anti-ADAMTS13 Abs implies co-operation between T and B lymphocytes. Activation of ADAMTS13-specific CD4+ T cells requires recognition of ADAMTS13 peptides bound to HLA molecules by T-cell receptors, thereby potentially implicating HLA molecules as predisposing factors for iTTP. In the 90’s, the first possible association between iTTP/the hemolytic-uremic syndrome (HUS) and HLA was reported.56 More specifically, study of the frequency of supertypic antigens DR52 and DR53, which are linked to the HLA-DRB3 and HLA-DRB4 genes respectively, indicated under-representation of DR53 in patients with iTTP compared to healthy individuals. Interestingly, molecular testing did not reveal any association with specific HLA-DRB3 alleles, suggesting the association was not with DR52 but with absence of the DR53 antigen, which could, therefore, be protective against iTTP.56 In 2010, two studies reported for the first time an association of HLA class II with iTTP.57,58 In a cohort of European patients, HLA-DRB1*04 and its associated HLADRB4-encoded serotype HLA-DR53 were protective against the development of iTTP. The decreased frequency of HLA-DRB4 in iTTP patients as compared to healthy controls appeared to be attributable to the specific reduction in the frequency of the linked HLA-DRB1*04 allele. In addition, the frequencies of HLA-DRB1*11 and HLADQB1*0301 were found to be higher in patients with iTTP when compared to healthy controls, suggesting these alleles to be risk factors for the onset of iTTP, although a primary role for HLA-DQB1*0301 could not be excluded.57 Reconstruction of HLA-haplotypes resulted in statistically higher frequency of DQB1*03-DRB1*11 haplotype in iTTP when compared to healthy controls. Highresolution typing of HLA-DRB1*11 revealed that both DRB1*1101 and DRB1*1104 were over-represented in iTTP, suggesting that the generic DRB1*11 is the predisposing factor.58 Other groups reported comparable findings,59,60 and provided additional protective haplotypes (DRB1*07-DQB1*02 and DRB1*13-DQB1*06) or haplotypes associated with susceptibility (DRB1*15-DQB1*06) 1101


J. Hrdinovรก et al.

for iTTP.60 Recently, additional evidence for the involvement of the HLA-DRB1*11 and HLA-DQB1*03 haplotype in the development of iTTP was provided from a case of familial iTTP on 2 first-degree relatives (mother and daughter) who both carried HLA-DRB1*1101/DRB1*1104 and the linked HLA-DQB1*03 allele. Interestingly, previous reports had shown associations between DRB1*11 and certain clinical conditions such as systemic sclerosis, early-onset juvenile chronic arthritis and sarcoidosis. In accordance with these findings, such associated conditions could be observed in our iTTP registry.61,62 The wide spectrum of autoimmune diseases sharing DRB1*11 clearly argues for the existence of additional risk factors, which may determine the specific clinical features of those diseases. Mancini et al. employed immunochip analysis to identify susceptibility loci in the HLA region for iTTP.63 A common variant rs6903608 was shown to confer 2.6-fold increased risk of development of iTTP.63 The rs6903608 is an intron of pseudogene HLA-DRB9 that maps to an intergenic region between HLA-DRA and HLA-DRB5 (Figure 1). Unexpectedly, HLA-DRB1*11 and HLA-DQB1*03 were not identified as risk factors in this study. It was proposed that rs6903608 is in linkage disequilibrium with HLA-DRB1*11, thereby masking a potential contribution of this allele.63 Interestingly, imputation analysis suggested that HLA-DQB1*0503 was also associated with iTTP.63 This particular allele had not been linked to iTTP in previous studies.

Potential mechanisms of HLA association with onset of iTTP The observed association between HLA alleles and the development of autoimmunity has been considered to arise from recognition of self-peptides by low affinity CD4+ T cells that have escaped negative selection in the thymus.64 It has recently been shown that differences in the intrinsic stability of HLA-DQ proteins may be linked to the onset of autoimmunity.65 Several HLA-DQ proteins have been shown to be poorly expressed on the cell surface; immuno-dominant epitopes that may potentially bind to these HLA-DQ proteins may, therefore, not be sufficiently presented, allowing CD4+ T cells to escape the negative selection in the thymus. This would result in the appearance of potentially auto-reactive CD4+ T cells in the periphery that could contribute to the onset of autoimmunity.65 It is possible that the increased frequency of HLADQB1*03 in patients with iTTP is caused by the inability of HLA-DQB1*03-containing DQ proteins to efficiently present immune-dominant self-peptides. This would result in a defective elimination of self-reactive CD4+ T cells during negative selection in the thymus. In this respect, the single nucleotide polymorphism (rs6903608) that has been linked to the onset of iTTP (see above), is located in the HLA locus that affects expression of a number of MHC class II subunits.63 More precisely, this single nucleotide polymorphism is located in close proximity to the DRB9 pseudogene, upstream of DRA gene. Although this has not been established, the rs6903608 single

Figure 1. The genes of the human MHC locus involved in immune-mediated thrombotic thrombocytopenic purpura (iTTP) susceptibility. Localization of genes of the two HLA alleles, HLADQB1*03 and HLA-DRB1*11, previously described as risk factors for iTTP, as well as the approximate localization of single polymorphism rs6903608, which appears to be in linkage disequilibrium with HLA-DRB1*11 and which has also been linked to the onset of iTTP.97

1102

haematologica | 2018; 103(7)


Immune TTP: interplay between haplotypes & environment

nucleotide polymorphism may affect the expression of the DRA gene, thereby modulating the amount of MHC class II on antigen-presenting cells. As outlined by Miyadera et al., increased stability of MHCII-peptide complexes may also confer protection for autoimmunity by a more rigorous elimination of self-reactive CD4+ T cells recognizing immune-dominant self-epitopes in the thymus (Figure 2).65 This mechanism may potentially provide an explanation for the protective effect of HLA-DRB1*04 on iTTP development. Future studies are needed to address whether HLA-DR or HLA-DQ stability modulates CD4+ T-cell responses in iTTP.

Presentation of ADAMTS13-derived peptides to CD4+ T cells Activation of ADAMTS13-specific CD4+ T cells requires uptake of ADAMTS13 by antigen-presenting cells (APCs) and presentation of ADAMTS13-derived peptides on HLA molecules. As the HLA-DRB1*11 was identified as a risk factor for the development of iTTP, investigation of ADAMTS13-derived peptides that are preferentially presented on MHC class II of healthy individuals positive for this allele was performed.66 The results of this study revealed peptides with a core sequence FINVAPHAR to be presented on MHC class II domains of HLA-DRB1*11 positive individuals. A number of individuals positive for HLA-DQB1*03 were also included in the study and peptides with core sequence LIRDTHSLR were found to be

presented by APCs. Interestingly, both peptide sequences originate from the CUB2 domain of ADAMTS13. However, when the APCs of these patients were pulsed with higher concentrations of ADAMTS13 (500 nM), the presented peptides within HLA-DRB1*11 or DQB1*03 donors remained essentially the same, but in patients with other alleles, some presented peptides belonging to other ADAMTS13 domains (i.e. TSP2-8 repeats; the metalloprotease or the spacer domains), in addition to the abovementioned core peptides FINVAPHAR and LIRDTHSLR. The results of this study thus suggest that FINVAPHAR and LIRDTHSLR-containing peptides contribute to the onset of iTTP.66 The next step was to investigate whether CD4+ T cells from HLA-typed patients with iTTP would recognize these peptides.67 In vitro stimulated peripheral blood mononuclear cells (PBMCs) of healthy donors and iTTP patients with FINVAPHAR (NAGGCRLFINVAPHARIAIH) or ASYILIRD (EGANASYILIRDTHSLRTTA) peptides were analyzed by flow cytometry. The results confirmed the initial assumption of FINVAPHAR peptide being able to activate CD4+ T cells of HLADRB1*11 expressing patients, whereas ASYILIRD resulted in activation of CD4+ T cells of HLA-DQB1*03 positive patients (Figure 3).67 In a separate study, Gilardin et al. identified another CUB2 domain-derived peptide with sequence GDMLLLWGRLTWRKM (1239-1253) as a potential DRB1*01 and DRB1*11 restricted T-cell epitope.68 This peptide was previously found to be presented

Figure 2. The possible role of MHC class II stability in the onset of immune-mediated thrombotic thrombocytopenic purpura (iTTP). (A) Intrinsically unstable HLADRB1*11 molecules are rapidly endocytosed, thus limiting the exposure of HLA-DRB1*11-loaded peptides on medullary thymic epithelial cells (mTEC) to CD4+ T cells. The proposed instability of HLA-DRB1*11/peptide complexes results in inefficient removal of self-reactive CD4+ T cells (through induction of apoptosis) in the thymus, and the appearance of potentially self-reactive CD4+ T cells in the peripheral lymphoid system.98,99 (B) Intrinsically stable HLA-DRB1*04 molecules are retained for prolonged periods of time on the surface of mTEC, thereby efficiently eliminating self-reactive CD4+ T cells. Consequently, the number of self-reactive CD4+ T cells escaping the negative selection in thymus will be very limited, which could account for the protective role of HLA-DRB1*04 in the development of iTTP.

haematologica | 2018; 103(7)

1103


J. HrdinovĂĄ et al.

by APCs from a DRB1*0401/DRB1*1301 positive donor which were pulsed with 500 nM ADAMTS13.66 In a recent study, we explored the repertoire of ADAMTS13derived peptides that were presented on HLA-DQ.69 In 4 of 9 donors analyzed, ADAMTS13-derived peptides presented on HLA-DQ were identified. Three HLADQB1*0301 positive donors were included in this study, and ADAMTS13-derived peptides were only found in one donor.69 The identified HLA-DQ-presented ADAMTS13 peptides originated from different domains; a CUB1 domain-derived peptide was presented in 3 of 8 donors. In the same study, the re-evaluation of HLA-DR-presented peptide repertoires confirmed the presentation of FINVAPHAR-derived peptides on HLA-DRB1*11.69 A large collection of ADAMTS13 derived peptides appeared to be presented on HLA-DR when compared to a previous study.66 Overall, the presented peptides were derived from multiple domains of ADAMTS13, with, however, a dominant contribution of CUB1/2-derived peptides to the total repertoire.69 To summarize, the specific peptides of ADAMTS13 recognized preferentially by HLA-DRB1*11 and DQB1*03 have now been identified. However, we need to understand how the recognition of these peptides by specific HLA molecules leads to the production of pathogenic antiADAMTS13 antibodies. Comparison of autoantibody profiles in animal models possessing/lacking risk MHC II alleles for iTTP and immunized with ADAMTS13 might help address this question.

Patients with iTTP are prone to develop autoimmunity Immune-mediated thrombotic thrombocytopenic purpura is associated with another autoimmune disease in up to 20% of cases. Other autoimmune diseases can occur years before iTTP, during long-term follow up, or con-

comitantly with iTTP. SLE can be observed in up to 10% of iTTP patients and represents the most common associated autoimmune disease, followed by Sjogren’s syndrome, presented by 3% of iTTP patients. Antinuclear Abs can be identified in 50% of cases. The presence of additional autoimmune diseases has no impact on the outcome of an acute iTTP episode. The presence of anti-double stranded DNA Abs or anti-SSA Abs at iTTP diagnosis is significantly associated with the development of an additional autoimmune disease during follow up.25,61,62,70

Modifying factors: role of ethnicity, sex, obesity, and others The over-representation of women and blacks in most iTTP registries in Western countries highlighted sex and ethnic disparities in this disease, further suggesting the involvement of specific genetic risk factors.23,71,72 In black patients, iTTP is typically associated with an increased risk of exacerbations at the acute phase but with less fatal outcomes (death rate 2.7% vs. 11.6% white patients in our experience), although initial presentation and prognosis is comparable to that of white patients,72,73 raising the hypothesis of a differential tolerance and/or adaptation to tissue ischemia between both ethnic groups.74 The increased prevalence of iTTP in blacks could at least in part result from the naturally low prevalence of the protective allele DRB1*04 in this ethnic group.72 The striking predominance of iTTP in women of childbearing age and during pregnancy raises the hypothesis that estrogen may favor the occurrence of iTTP. As reported in SLE, it is likely that estrogen increases the risk of iTTP in genetically predisposed women by elevating type1 interferon production and favoring the survival of autoreactive B cells.75 Similarly, obesity could represent a risk factor for iTTP as a result of increased peripheral

Figure 3. The onset of immune-mediated thrombotic thrombocytopenic purpura (iTTP). ADAMTS13 is endocytosed by antigen-presenting cell [in this figure, dendritic cell (DC)] and processed to peptides that are subsequently loaded on MHC-II molecules. As described previously, ADAMTS13-derived peptides FINVAPHAR and ASYILIRD were found to be presented on HLA-DRB1*1101 and HLA-DQB1*03, respectively. In the case of presence of specific autoreactive CD4+ T cells, the complex MHC-II/peptide will be recognized by TCR, which will cause the activation of the CD4+ T cell. Such activated CD4+ T cells will then provide help to autoreactive B cells that will result in a production of ADAMTS13-specific auto-Abs.

1104

haematologica | 2018; 103(7)


Immune TTP: interplay between haplotypes & environment

aromatization of androgens to estrogens, but also higher leptin levels, in this population of patients.71,76 In addition, obese healthy individuals have an increased prevalence of non-inhibitory ADAMTS13 auto-Abs, although no differences in ADAMTS13 activity and antigen levels were found with lean people.77 Taken together, these findings stress the role of modifying factors including hormones, cytokines and other mediators involved in the loss of tolerance against ADAMTS13; these require further study to be better understood mechanistically. There is also a need to specify whether iTTP patients with associated conditions, including pregnancy, connective tissue diseases or obesity, also have specific genetic risk factors as compared to patients with idiopathic iTTP. Von Willebrand Factor (VWF) is another important modifier of TTP susceptibility, either in the congenital or in the autoimmune form. Raised VWF levels are required to induce TTP in Adamts13-/- mice, but varying the concentration between 20 and 120 U/mL does not appear to affect the occurrence or severity of the disease, suggesting that a threshold level of VWF is sufficient, and that higher levels confer little additional risk. However, humans appear to be more sensitive to changes in VWF levels than mice. Women with inherited ADAMTS13 deficiency frequently develop TTP during pregnancy, which probably results from the progressive increase in VWF levels throughout the gestation period. Moreover, obese individuals have higher levels of VWF, providing further evidence for an association between obesity and TTP. Thus,

changes in VWF secretion, multimer distribution, and plasma level may trigger TTP.78 Among additional modifiers of TTP, blood group O was found over-represented in iTTP cohorts, suggesting that blood group O could be a risk factor for TTP.79

ADAMTS13 conformation in iTTP The precise mechanism triggering the formation of antiADAMTS13 Abs in previously healthy individuals still has to be identified. As mentioned earlier, the majority of antispacer auto-Abs that develop in patients with iTTP is directed toward an exosite composed of residues R568, F592, R660, Y661 and Y665.30 Under quiescent conditions, the exosite within the spacer domain is covered by the CUB1/2 domains maintaining ADAMTS13 in a so-called closed conformation.80-82 Binding of VWF to the carboxyterminal CUB domains induces a conformational activation of ADAMTS13 which results in the exposure of the spacer domain exosite composed of residues R568, F592, R660, Y661 and Y665 which now becomes available for binding to the unfolded VWF A2 domain.83 Apart from VWF, murine monoclonal Abs can induce a conversion of ADAMTS13 from a closed to an open conformation.80,82 Conservative mutations of the spacer domain exosite created a variant with increased proteolytic activity toward its substrate VWF.84 Recently, it was shown that this socalled gain-of-function (GoF) variant is present in an open conformation, which promotes binding of the spacer domain exosite to the unfolded A2 domain of VWF.81

Figure 4. Conformational changes of ADAMTS13. [Mp: metalloprotease (red); Dis: disintegrin-like domain (green); Cys-rich: cysteine-rich domain (orange); TSRs: thrombospondin-type 1 repeats (light and dark gray); CUB1 domain (cyan); CUB2 domain (magenta)]. (A) In the blood circulation of healthy individuals and immune-mediated thrombotic thrombocytopenic purpura (iTTP) patients in remission ADAMTS13 is present in closed conformation with CUB1/2 domains covering the spacer domain. Binding of plasma factors (e.g. activating anti-ADAMTS13 auto-Abs against the TSR5-8-CUB1/2 domains80,82) (orange part of the anti-ADAMTS13 antibodies) results in a conformational change: the opening of the ADAMTS13.83,100 (B) Open conformation of ADAMTS13 exposes the B-cell epitope localized in the spacer domain, making it accessible for additional anti-ADAMTS13 Abs.

haematologica | 2018; 103(7)

1105


J. Hrdinová et al.

Figure 5. Presentation of ADAMTS13 peptides on MHC class II. (A) Three-dimensional model structure of HLA-DRB1*11 (graphic representation in gray) with FINVAPHAR peptide (stick representation in magenta with cyan surface) from ADAMTS13 CUB2 domain. (B) Three-dimensional model structure of HLA-DQB1*03 (graphic representation in gray) with LIRDTHSLR peptide (stick representation in cyan with magenta surface) from ADAMTS13 CUB2 domain.

These results imply that the major B-cell epitope in the spacer domain is not accessible for binding of auto-Abs when ADAMTS13 circulates in a so-called closed conformation. Because the knowledge on conformational state of ADAMTS13 in patient plasma could be of help in the diagnosis and prognosis of iTTP, Roose et al. designed an assay to monitor this.85 Their results showed that ADAMTS13 adopts an altered (more open) conformation in iTTP patients in the acute phase of the disease while ADAMTS13 was closed in the majority of iTTP patients in remission, in healthy donors, and patients with sepsis or HUS. Hence, acute iTTP is not only characterized by severe thrombocytopenia, hemolytic anemia, ADAMTS13 activity less than 10%, and presence of anti-ADAMTS13 Abs, but also by the presence of an open ADAMTS13 conformation. It now remains to be determined which factors (antiADAMTS13 Abs and/or other plasma factors) open ADAMTS13. However, once the conformation of ADAMTS13 is changed during acute iTTP, antiADAMTS13 Abs recognizing not only non-cryptic epitopes but also cryptic epitopes in the spacer domain can now bind, leading to the inhibition and/or clearance of ADAMTS13 (Figure 4). Whether the conformational change in ADAMTS13 also leads to an (additional) immune response, remains to be determined.85

The autoimmune response against ADAMTS13: an infection-driven event? Many studies have suggested a link between infections and the onset of iTTP. First, infectious agents can be precipitating factors in patients with a severe ADAMTS13 deficiency, through endothelial activation. No specific agents could be identified,86 as opposed to HUS, which was typically associated to specific strains of the bacteria Escherichia coli. Endothelial activation in this context may involve nucleosomes that derive at least in part from neutrophil extracellular traps (NETS), networks made of nuclear DNA, histones, granular and cytoplasmic proteins that are released by neutrophils in response to infections.87 In response to these agonists, endothelial cells release high molecular weight VWF multimers leading to thrombi in the microvasculature of most organs. Human neutrophil peptides released from activated and degranulated neu1106

trophils can also alter ADAMTS13 activity.88 This scenario, termed the “two-hit model”, was evidenced from animal models and illustrates the interaction between a genetic background (a severe constitutive ADAMTS13 deficiency or a propensity to develop an immune-mediated ADAMTS13 deficiency) and environmental factors, especially infections and inflammation.89 Second, infections may potentially result in the presentation of pathogen-derived peptides that are homologous to ADAMTS13-derived peptides such as the CUB2 domainderived peptide FINVAPHAR.90,91 CD4+ T cells developing in response to a challenge by pathogens may, therefore, cross-react with ADAMTS13 peptides presented on MHC class II, resulting in their activation and proliferation. This phenomenon has been designated molecular or epitope mimicry.92 As discussed earlier, our group identified CD4+ T cells reactive to ADAMTS13 peptides derived from the CUB2 domain in patients with iTTP.67 These peptides with amino acid sequence FINVAPHAR and ASYILIRD (Figure 5) may share epitope mimicry with microbiomederived peptides. CD4+ T cells targeting these microbiome-derived peptides may cross-react with ADAMTS13 peptides presented on risk alleles for iTTP, thereby contributing to the onset of iTTP. Another hypothesis developed from the observed association between infections and iTTP is a gene-environment interaction process involving microbial effectors that activate endothelial and polymorphonuclear cells. Therefore, functionally relevant molecules belonging to the innate immune response pathways could be important modulators of iTTP initiation in the context of infection. Variants of the anti-infectious innate immunity sensor Toll-like receptor (TLR)-9 were reported to be more represented in iTTP patients negative for the HLADRB1*11 Class II susceptibility allele.86 TLR-9 harboring specific variants could, therefore, be more prone to activate endothelial and polymorphonuclear cells and produce Th1 cytokines in a context of infection, precipitating an iTTP episode.86 Similarly, anecdotal responses in patients with severe iTTP following the administration of the complement blocker eculizumab suggest a role for complement deregulation in iTTP pathophysiology.93 Unusually large VWF multimers may activate the alternahaematologica | 2018; 103(7)


Immune TTP: interplay between haplotypes & environment

tive complement pathway to promote the generation of terminal complement complexes (C5b-9).94 These observations therefore deserve systematic explorations to unravel the role of the alternative complement pathway and its regulators (CFH, CFI, MCP and thrombomodulin) in iTTP pathophysiology. Interestingly, HLA-DRB1*11 was reported to be protective against tuberculosis, whereas HLA-DRB1*04 was associated with an increased risk of tuberculosis and severe malaria.95,96 These observations raise the intriguing possibility that autoimmunity against ADAMTS13 leading to iTTP could represent the cost of an efficient ancestral immune response selected to fight against historically harmful infectious agents.

Perspectives: future directions Despite the considerable progress made in unravelling the role of ADAMTS13 in primary hemostasis, our under-

References 1. Kremer Hovinga JA, Coppo P, Lämmle B, Moake JL, Miyata T, Vanhoorelbeke K. Thrombotic thrombocytopenic purpura. Nat Rev Dis Prim 2017;317020. 2. Scully M, McDonald V, Cavenagh J, Hunt BJ. A phase 2 study of the safety and efficacy of rituximab with plasma exchange in acute acquired thrombotic thrombocytopenic purpura. Blood. 2011;118(7):1746-1754. 3. Hie M, Gay J, Galicier L, et al. Preemptive rituximab infusions after remission efficiently prevent relapses in acquired thrombotic thrombocytopenic purpura. Blood. 2014;124(2):204-210. 4. Westwood J-P, Thomas M, Alwan F, et al. Rituximab prophylaxis to prevent thrombotic thrombocytopenic purpura relapse: outcome and evaluation of dosing regimens. Blood Adv. 2017;1(15):1159 LP-1166. 5. Scully M, Knobl P, Kentouche K, et al. Recombinant ADAMTS-13: first-in-human pharmacokinetics and safety in congenital thrombotic thrombocytopenic purpura. Blood. 2017;130(19):2055-2063. 6. 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. 7. Kremer Hovinga JA, Voorberg J. Improving on nature: redesigning ADAMTS13. Blood. 2012;119(16):3654-3655. 8. Furlan M, Robles R, Solenthaler M, Lä B. Acquired Deficiency of von Willebrand Factor-Cleaving Protease in a Patient With Thrombotic Thrombocytopenic Purpura. Blood. 1998;91(8):2839-2846. 9. Tsai HM, Lian EC. Antibodies to von Willebrand factor-cleaving protease in acute thrombotic thrombocytopenic purpura. N Engl J Med. 1998;339(22):1585-1594. 10. Scheiflinger F, Kno P, Trattner B, Plaimauer B, Mohr G, Dockal M. Nonneutralizing IgM and IgG antibodies to von Willebrand factor – cleaving protease ( ADAMTS-13 ) in a patient with thrombotic thrombocytopenic purpura. Blood. 2003;102(9):3241-3243.

haematologica | 2018; 103(7)

standing of the immunopathogenesis of iTTP is still not complete. Future studies will have to reveal a reason for the paradoxically extremely low incidence of the disease considering the frequent occurrence of identified genetic risk factors within the HLA-class II locus in the healthy population. Another field of investigation involves the mechanisms determining the reoccurrence of autoreactive lymphocytes in patients following immunomodulation with B-cell depleting therapies in order to better anticipate relapses. Moreover, a new therapeutic area is being opened in the field with the development of the recombinant ADAMTS13; its interactions with the immune system of iTTP patients will require further investigation. Acknowledgements The authors would like to thank the Horizon 2020 Framework program for Research and Innovation of the European Union for funding this work under 675746 (PROFILE).

11. Crawley JTB, Scully MA. Thrombotic thrombocytopenic purpura: basic pathophysiology and therapeutic strategies. Hematology Am Soc Hematol Educ Program. 2013;2013:292-299. 12. Sadler JE. What’s new in the diagnosis and pathophysiology of thrombotic thrombocytopenic purpura. Hematology Am Soc Hematol Educ Program. 2015;2015:631-636. 13. Thomas MR, de Groot R, Scully MA, Crawley JTB. Pathogenicity of AntiADAMTS13 Autoantibodies in Acquired Thrombotic Thrombocytopenic Purpura. EBioMedicine. 2015;2(8):942-952. 14. Ostertag EM, Kacir S, Thiboutot M, et al. ADAMTS13 autoantibodies cloned from patients with acquired thrombotic thrombocytopenic purpura: 1. Structural and functional characterization in vitro. Transfusion. 2016;56(7):1763-1774. 15. Lotta LA, Valsecchi C, Pontiggia S, et al. Measurement and prevalence of circulating ADAMTS13-specific immune complexes in autoimmune thrombotic thrombocytopenic purpura. J Thromb Haemost. 2014;12(3): 329-336. 16. Ferrari S, Palavra K, Gruber B, et al. Persistence of circulating ADAMTS13-specific immune complexes in patients with acquired thrombotic thrombocytopenic purpura. Haematologica. 2014;99(4):779-787. 17. Ferrari S, Knöbl P, Kolovratova V, et al. Inverse correlation of free and immune complex-sequestered anti-ADAMTS13 antibodies in a patient with acquired thrombotic thrombocytopenic purpura. J Thromb Haemost. 2012;10(1):156-158. 18. Bianchi V, Robles R, Alberio L, Furlan M, Lammle B. Von Willebrand factor-cleaving protease (ADAMTS13) in thrombocytopenic disorders: a severely deficient activity is specific for thrombotic thrombocytopenic purpura. Blood. 2002;100(2):710713. 19. Zheng XL, Kaufman RM, Goodnough LT, Sadler JE. Effect of plasma exchange on plasma ADAMTS13 metalloprotease activity, inhibitor level, and clinical outcome in patients with idiopathic and nonidiopathic thrombotic thrombocytopenic purpura. Ann Intern Med. 2004;103(11):4043-4049.

20. Peyvandi F, Ferrari S, Lavoretano S, Canciani MT, Mannucci PM. Von Willebrand factor cleaving (ADAMTS-13) and ADAMTS-13 neutralizing autoantibodies in 100 patients with thrombotic thrombocytopenic purpura. Br J Haematol. 2004;127(4):433-439. 21. Rieger M, Mannucci PM, Kremer Hovinga JA, et al. ADAMTS13 autoantibodies in patients with thrombotic microangiopathies and other immunomediated diseases. Blood. 2005;106(4):1262-1267. 22. Tsai HM, Raoufi M, Zhou W, et al. ADAMTS13-binding IgG are present in patients with thrombotic thrombocytopenic purpura. Thromb Haemost 2006;95(5):886892. 23. Scully M, Yarranton H, Liesner R, et al. Regional UK TTP Registry: Correlation with laboratory ADAMTS 13 analysis and clinical features. Br J Haematol. 2008;142(5):819826. 24. Fujimura Y, Matsumoto M. Registry of 919 patients with thrombotic microangiopathies across Japan: database of Nara Medical University during 1998-2008. Intern Med. 2010;49(1):7-15. 25. 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):e237-245. 26. Joly BS, Stepanian A, Leblanc T, et al. Childonset and adolescent-onset acquired thrombotic thrombocytopenic purpura with severe ADAMTS13 deficiency: a cohort study of the French national registry for thrombotic microangiopathy. Lancet Haematol. 2016;3(11):e537-e546. 27. Kremer Hovinga JA, Vesely SK, Terrell DR, Lammle B, George JN. Survival and relapse in patients with thrombotic thrombocytopenic purpura. Blood. 2010;115(8):150011; quiz 1662. 28. Ferrari S, Mudde GC, Rieger M, Veyradier A, Kremer Hovinga JA, Scheiflinger F. IgG subclass distribution of anti-ADAMTS13 antibodies in patients with acquired thrombotic thrombocytopenic purpura. J Thromb

1107


J. Hrdinová et al. Haemost 2009;7(10):1703-1710. 29. 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. 30. 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. 31. Grillberger R, Casina VC, Turecek PL, Long Zheng X, 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):2-4. 32. Arbuckle MR, McClain MT, Rubertone M V, et al. Development of autoantibodies before the clinical onset of systemic lupus erythematosus. N Engl J Med. 2003;349(16):15261533. 33. Luken BM, Kaijen PHP, Turenhout EA, 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. 34. 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. 35. Sui J, Hwang WC, Perez S, et al. Structural and functional bases for broad-spectrum neutralization of avian and human influenza A viruses. Nat Struct Mol Biol. 2009;16(3):265-273. 36. Marasca R, Vaccari P, Luppi M, et al. Immunoglobulin gene mutations and frequent use of VH1-69 and VH4-34 segments in hepatitis C virus-positive and hepatitis C virus-negative nodal marginal zone B-cell lymphoma. Am J Pathol. 2001;159(1):253261. 37. Schaller M, Vogel M, Kentouche K, Lämmle B, Kremer Hovinga JA. The splenic autoimmune response to ADAMTS13 in thrombotic thrombocytopenic purpura contains recurrent antigen-binding CDR3 motifs. Blood 2014;124(23):3469-3479. 38. Vanhoorelbeke K, De Meyer SF. Animal models for thrombotic thrombocytopenic purpura. J Thromb Haemost. 2013;11 (Suppl 1):2-10. 39. Ostertag EM, Bdeir K, Kacir S, et al. ADAMTS13 autoantibodies cloned from patients with acquired thrombotic thrombocytopenic purpura: 2. Pathogenicity in an animal model. Transfusion. 2016;56(7): 1775-1785. 40. Tersteeg C, Schiviz A, De Meyer SF, et al. Potential for Recombinant ADAMTS13 as an Effective Therapy for Acquired Thrombotic Thrombocytopenic Purpura. Arterioscler Thromb Vasc Biol. 2015;35(11):2336-2342. 41. Deforche L, Tersteeg C, Roose E, et al. Generation of Anti-Murine ADAMTS13 Antibodies and Their Application in a Mouse Model for Acquired Thrombotic Thrombocytopenic Purpura. PLoS One. 2016;11(8):e0160388. 42. Froissart A, Buffet M, Veyradier A, et al. Efficacy and safety of first-line rituximab in severe, acquired thrombotic thrombocy-

1108

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

57.

topenic purpura with a suboptimal response to plasma exchange. Experience of the French Thrombotic Microangiopathies Reference Center. Crit Care Med. 2012;40(1):104-111. Page EE, Kremer Hovinga JA, Terrell DR, Vesely SK, George JN. Rituximab reduces risk for relapse in patients with thrombotic thrombocytopenic purpura. Blood. 2016;127 (24):3092-3094. 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. 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. 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. Klaus C, Plaimauer B, Studt JD, et al. Epitope mapping of ADAMTS13 autoantibodies in acquired thrombotic thrombocytopenic purpura. Blood. 2004;103(12):4514-4519. Luken BM, Turenhout EAM, Kaijen PHP, et al. Amino acid regions 572-579 and 657-666 of the spacer domain of ADAMTS13 provide a common antigenic core required for binding of antibodies in patients with acquired TTP. Thromb Haemost. 2006;96 (3):295-301. Pos W, Crawley JTB, Fijnheer R, Voorberg J, Lane D a., Luken BM. An autoantibody epitope comprising residues R660, Y661, and Y665 in the ADAMTS13 spacer domain identifies a binding site for the A2 domain of VWF. Blood. 2010;115(8):1640-1649. 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. Jin S-Y, Skipwith CG, Zheng XL. Amino acid residues Arg(659), Arg(660), and Tyr(661) in the spacer domain of ADAMTS13 are critical for cleavage of von Willebrand factor. Blood. 2010;115(11): 2300-2310. Casina VC, Hu W, Mao J-H, et al. High-resolution epitope mapping by HX MS reveals the pathogenic mechanism and a possible therapy for autoimmune TTP syndrome. Proc Natl Acad Sci. 2015;112(31):9620-9625. Long Zheng X, Wu HM, Shang D, et al. Multiple domains of ADAMTS13 are targeted by autoantibodies against ADAMTS13 in patients with acquired idiopathic thrombotic thrombocytopenic purpura. Haematologica. 2010;95(9):15551562. Studt JD, Hovinga JAK, Radonic R, et al. Familial acquired thrombotic thrombocytopenic purpura: ADAMTS13 inhibitory autoantibodies in identical twins. Blood. 2004;103(11):4195-4197. Wallace DC, Lovric A, Clubb JS, Carseldine DB. Thrombotic thrombocytopenic purpura in four siblings. Am J Med. 1975;58(5):724734. Joseph G, Smith KJ, Hadley TJ, et al. HLADR53 protects against thrombotic thrombocytopenic purpura/adult hemolytic uremic syndrome. Am J Hematol. 1994;47(3):189193. Scully M, Brown J, Patel R, Mcdonald V,

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

68.

69.

70.

71.

Brown CJ, Machin S. Human leukocyte antigen association in idiopathic thrombotic thrombocytopenic purpura: Evidence for an immunogenetic link. J Thromb Haemost. 2010;8(2):257-262. Coppo P, Busson M, Veyradier A, et al. HLA-DRB1*11: A strong risk factor for acquired severe ADAMTS13 deficiencyrelated idiopathic thrombotic thrombocytopenic purpura in Caucasians. J Thromb Haemost. 2010;8(4):856-859. John M-L, Hitzler W, Scharrer I. The role of human leukocyte antigens as predisposing and/or protective factors in patients with idiopathic thrombotic thrombocytopenic purpura. Ann Hematol. 2012;91(4):507-510. Lombardi AM, Pulini S, Passeri C, et al. Familial acquired thrombotic thrombocytopenic purpura: immunogenetic link with HLA-DRB1*11 and DQB1*03 antigens. Br J Haematol. 2017 Oct 26. [Epub ahead of print] Coppo P, Bengoufa D, Veyradier A, et al. Severe ADAMTS13 deficiency in adult idiopathic thrombotic microangiopathies defines a subset of patients characterized by various autoimmune manifestations, lower platelet count, and mild renal involvement. Medicine (Baltimore). 2004;83(4):233-244. Coppo P, Schwarzinger M, Buffet M, et al. Predictive features of severe acquired ADAMTS13 deficiency in idiopathic thrombotic microangiopathies: the French TMA reference center experience. PLoS One. 2010;5(4):e10208. Mancini I, Ricaño-Ponce I, Pappalardo E, et al. Immunochip analysis identifies novel susceptibility loci in the human leukocyte antigen region for acquired thrombotic thrombocytopenic purpura. J Thromb Haemost. 2016;14(12):2356-2367. Bergseng E, Dørum S, Arntzen MØ, et al. Different binding motifs of the celiac disease-associated HLA molecules DQ2.5, DQ2.2, and DQ7.5 revealed by relative quantitative proteomics of endogenous peptide repertoires. Immunogenetics. 2015;67 (2):73-84. Miyadera H, Ohashi J, Lernmark Å, Kitamura T, Tokunaga K. Cell-surface MHC density profiling reveals instability of autoimmunity-associated HLA. J Clin Invest 2015;125(1):275-291. Sorvillo N, van Haren SD, Kaijen PH, et al. Preferential HLA-DRB1*11-dependent presentation of CUB2-derived peptides by ADAMTS13-pulsed dendritic cells. Blood. 2013;121(17):3502-3510. Verbij FC, Turksma AW, de Heij F, et al. CD4+ T cells from patients with acquired thrombotic thrombocytopenic purpura recognize CUB2 domain-derived peptides. Blood. 2016;127(12):1606-1609. Gilardin L, Delignat S, Peyron I, et al. The ADAMTS13(1239-1253) peptide is a dominant HLA-DR1-restricted CD4+ T-cell epitope. Haematologica. 2017;102(11):18331841. Hrdinova J, Verbij FC, Kaijen PHP, et al. Mass spectrometry-assisted identification of ADAMTS13-derived peptides presented on HLA-DR and HLA-DQ. Haematologica. 2018 Mar 22. [Epub ahead of print]. Roriz M, Landais M, Desprez J, et al. Risk Factors for Autoimmune Diseases Development After Thrombotic Thrombocytopenic Purpura. Medicine (Baltimore). 2015;94(42):e1598. George JN, Vesely SK, Terrell DR. The Oklahoma Thrombotic Thrombocytopenic Purpura-Hemolytic Uremic Syndrome

haematologica | 2018; 103(7)


Immune TTP: interplay between haplotypes & environment

72.

73.

74. 75.

76.

77.

78. 79.

80.

81.

(TTP-HUS) Registry: A Community Perspective of Patients with Clinically Diagnosed TTP-HUS. Semin Hematol. 2004;41(1):60-67. Martino S, Jamme M, Deligny C, et al. Thrombotic thrombocytopenic purpura in Black people: Impact of ethnicity on survival and genetic risk factors. PLoS One. 2016;11(7):3-12. Cataland SR, Yang S Bin, Witkoff L, et al. Demographic and ADAMTS13 biomarker data as predictors of early recurrences of idiopathic thrombotic thrombocytopenic purpura. Eur J Haematol. 2009;83(6):559564. Sheikh N, Sharma S. Impact of ethnicity on cardiac adaptation to exercise. Nat Rev Cardiol. 2014;11(4):198-217. Hughes GC, Choubey D. Modulation of autoimmune rheumatic diseases by oestrogen and progesterone. Nat Rev Rheumatol. 2014;10(12):740-751. Diamanti-Kandarakis E, Bergiele A. The influence of obesity on hyperandrogenism and infertility in the female. Obes Rev. 2001;2(4):231-238. Lombardi AM, Fabris R, Scarda A, et al. Presence of anti-ADAMTS13 antibodies in obesity. Eur J Clin Invest. 2012;42(11):11971204. Sadler JE. Pathophysiology of thrombotic thrombocytopenic purpura. Blood. 2017;130 (10):1181-1188. Terrell DR, Motto DG, Kremer Hovinga JA, Lammle B, George JN, Vesely SK. Blood group O and black race are independent risk factors for thrombotic thrombocytopenic purpura associated with severe ADAMTS13 deficiency. Transfusion. 2011;51(10):22372243. Muia J, Zhu J, Gupta G, et al. Allosteric activation of ADAMTS13 by von Willebrand factor. Proc Natl Acad Sci USA. 2014;111(52):18584-18589. South K, Luken BM, Crawley JTB, et al. Conformational activation of ADAMTS13. Proc Natl Acad Sci USA. 2014;111(52): 18578-18583.

haematologica | 2018; 103(7)

82. Deforche L, Roose E, Vandenbulcke A, et al. Linker regions and flexibility around the metalloprotease domain account for conformational activation of ADAMTS-13. J Thromb Haemost. 2015;13(11):2063-2075. 83. Gao W, Anderson PJ, Sadler JE. Extensive contacts between ADAMTS13 exosites and von Willebrand factor domain A2 contribute to substrate specificity. Blood. 2008;112(5):1713-1719. 84. Jian C, Xiao J, Gong L, et al. Gain-of-function ADAMTS13 variants that are resistant to autoantibodies against ADAMTS13 in patients with acquired thrombotic thrombocytopenic purpura. Blood. 2012;119 (16):3836-3843. 85. Roose E, Schelpe AS, Joly BS, et al. An open conformation of ADAMTS13 is a hallmark of acute acquired thrombotic thrombocytopenic purpura. J Thromb Haemost. 2018;16(2):378-388. 86. Morgand M, Buffet M, Busson M, et al. High prevalence of infectious events in thrombotic thrombocytopenic purpura and genetic relationship with toll-like receptor 9 polymorphisms: experience of the French Thrombotic Microangiopathies Reference Center. Transfusion. 2014;54(2):389-397. 87. Fuchs TA, Kremer Hovinga JA, Schatzberg D, Wagner DD, Lammle B. Circulating DNA and myeloperoxidase indicate disease activity in patients with thrombotic microangiopathies. Blood. 2012;120(6):1157-1164. 88. Pillai VG, Bao J, Zander CB, et al. Human neutrophil peptides inhibit cleavage of von Willebrand factor by ADAMTS13: a potential link of inflammation to TTP. Blood. 2016;128(1):110-119. 89. Motto DG, Chauhan AK, Zhu G, et al. Shigatoxin triggers thrombotic thrombocytopenic purpura in genetically susceptible ADAMTS13-deficient mice. J Clin Invest. 2005;115(10):2752-2761. 90. Verbij FC, Fijnheer R, Voorberg J, Sorvillo N. Acquired TTP: ADAMTS13 meets the immune system. Blood Rev. 2014;28(6):227234. 91. Sorvillo N, Van Haren SD, Kaijen PH, et al.

Preferential HLA-DRB1*11–dependent presentation of CUB2-derived peptides by ADAMTS13-pulsed dendritic cells. Blood. 2013;121(17):3502-3510. 92. 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. 93. Chapin J, Weksler B, Magro C, Laurence J. Eculizumab in the treatment of refractory idiopathic thrombotic thrombocytopenic purpura. Br J Haematol. 2012;157(6):772774. 94. Feng S, Kroll MH, Nolasco L, Moake J, Afshar-Kharghan V. Complement activation in thrombotic microangiopathies. Br J Haematol. 2013;160(3):404-406. 95. Osafo-Addo AD, Koram KA, Oduro AR, Wilson M, Hodgson A, Rogers WO. HLADRB1*04 allele is associated with severe malaria in Northern Ghana. Am J Trop Med Hyg. 2008;78(2):251-255. 96. Tong X, Chen L, Liu S, et al. Polymorphisms in HLA-DRB1 gene and the risk of tuberculosis: a meta-analysis of 31 studies. Lung. 2015;193(2):309-318. 97. Mancini I, Ricaño-Ponce I, Pappalardo E, et al. Immunochip analysis identifies novel susceptibility loci in the human leukocyte antigen region for acquired thrombotic thrombocytopenic purpura. J Thromb Haemost. 2016;14(12):2356-2367. 98. Liu GY, Fairchild PJ, Smith RM, Prowle JR, Kioussis D, Wraith DC. Low avidity recognition of self-antigen by T cells permits escape from central tolerance. Immunity. 1995;3(4):407-415. 99. James EA, Kwok WW. Low-affinity major histocompatibility complex-binding peptides in type 1 diabetes. Diabetes. 2008;57(7):1788-1789. 100. South K, Freitas MO, Lane DA. A model for the conformational activation of the structurally quiescent metalloprotease ADAMTS13 by von willebrand factor. J Biol Chem. 2017;292(14):5760-5769.

1109


ARTICLE

Hematopoiesis

Ferrata Storti Foundation

Setd2 regulates quiescence and differentiation of adult hematopoietic stem cells by restricting RNA polymerase II elongation Yile Zhou,1,2 Xiaomei Yan,2 Xiaomin Feng,2 Jiachen Bu,2,3 Yunzhu Dong,2 Peipei Lin,2 Yoshihiro Hayashi,2 Rui Huang,2 Andre Olsson,4 Paul R. Andreassen,2 H. Leighton Grimes,4 Qian-fei Wang,3 Tao Cheng,5 Zhijian Xiao,5 Jie Jin1,* and Gang Huang2,*

Haematologica 2018 Volume 103(7):1110-1123

Department of Hematology, The First Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, China; 2Division of Pathology and Experimental Hematology and Cancer Biology, Cincinnati Children’s Hospital Medical Center, OH, USA; 3Laboratory of Genome Variations and Precision Bio-Medicine, Beijing Institute of Genomics, Chinese Academy of Sciences, China; 4Division of Immunobiology and Center for Systems Immunology,Cincinnati Children’s Hospital Medical Center, OH, USA; and 5State Key Laboratory of Experimental Hematology, Institute of Hematology and Blood Diseases Hospital and Center for Stem Cell Medicine, Chinese Academy of Medical Sciences and Peking Union Medical College, Tianjin, China 1

*JJ and GH contributed equally to this study as joint senior authors

ABSTRACT

S

Correspondence: gang.huang@cchmc.org or jiej0503@zju.edu.cn Received: Juanary 3, 2018. Accepted: April 6, 2018. Pre-published: April 12, 2018.

doi:10.3324/haematol.2018.187708 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/103/7/1110

ET domain containing 2 (Setd2), encoding a histone methyltransferase, is associated with many hematopoietic diseases when mutated. By generating a novel exon 6 conditional knockout mouse model, we describe an essential role of Setd2 in maintaining the adult hematopoietic stem cells. Loss of Setd2 results in leukopenia, anemia, and increased platelets accompanied by hypocellularity, erythroid dysplasia, and mild fibrosis in bone marrow. Setd2 knockout mice show significantly decreased hematopoietic stem and progenitor cells except for erythroid progenitors. Setd2 knockout hematopoietic stem cells fail to establish long-term bone marrow reconstitution after transplantation because of the loss of quiescence, increased apoptosis, and reduced multiple-lineage terminal differentiation potential. Bioinformatic analysis revealed that the hematopoietic stem cells exit from quiescence and commit to differentiation, which lead to hematopoietic stem cell exhaustion. Mechanistically, we attribute an important Setd2 function in murine adult hematopoietic stem cells to the inhibition of the Nsd1/2/3 transcriptional complex, which recruits super elongation complex and controls RNA polymerase II elongation on a subset of target genes, including Myc. Our results reveal a critical role of Setd2 in regulating quiescence and differentiation of hematopoietic stem cells through restricting the NSDs/SEC mediated RNA polymerase II elongation.

©2018 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.

1110

Introduction Hematopoietic stem cells (HSCs) are characterized by their capability for selfrenewal and multi-potency.1,2 Hematopoiesis is dynamically controlled by the interplay of transcriptional and epigenetic networks, while dysregulation of these networks can lead to unfitness of hematopoiesis, cellular transformation, and hematologic diseases. Multiple drugs targeting epigenetic modulators have shown promising effects on certain hematopoietic diseases.3,4 Thus, a better understanding of how the epigenome is regulated in hematopoiesis may provide insights that can improve the treatment of hematologic disorders. haematologica | 2018; 103(7)


Setd2 regulates hematopoietic stem cells

Histone H3K36 methylation is one of the most prominent epigenetic modifications that are associated with gene activation. In yeast, Set2 is the sole H3K36 methyltransferase, which is responsible for all three methylation events and can interact with RNA polymerase II (RNA Pol II).5 Set2 contains several conserved domains. One of them is the SET domain, which is the catalytic domain for H3K36 methylations. Another important domain is the SRI domain, which binds to serine 2 (Ser2) and serine 5 (Ser5) doubly phosphorylated carboxyl terminal domain (CTD) repeats of RNA Pol II.6 The human ortholog of Set2, SETD2, was first isolated from human CD34+ hematopoietic stem/progenitor cells (HSPCs).7 SETD2 mainly works as H3K36 tri-methyltransferase, while H3K36me1 and H3K36me2 are catalyzed by other methyltransferases. To date, 7 other HMT enzymes have been reported to methylate H3K36, including NSD1, NSD2, NSD3, and ASH1L.8 NSD1/2/3 and ASH1L can methylate H3K36 to generate H3K36me1 and H3K36me2. The NSDs have been reported as oncogenic drivers in many cancers including leukemia. Furthermore, NSDs could regulate WNT, MYC, and NFκB to affect various physiological or pathological processes.9 It has been reported that Setd2 is required for murine embryonic stem cells (mESCs) differentiation toward endoderms and endoderm development during murine embryonic development,10 while Setd2-/- resulted in embryonic lethality at E10.5-11.5.11 SETD2 was identified as a tumor suppressor, as loss-of-function (LOF) mutations of SETD2 have been found in many human cancers, including leukemia and lymphoma.12-15 Previously, we have reported that there are SETD2 mutations in 6% of acute leukemia with 22% enriched in MLL-rearranged leukemia.16 However, the roles of SETD2 in adult HSPCs and normal hematopoiesis have not been fully studied. To understand the mechanisms of how Setd2 regulates the normal hematopoiesis, by using a novel conditional knockout model, we revealed a unique and critical role of Setd2 in regulating quiescence and differentiation of adult HSCs through restricting NSDs/SEC mediated RNA polymerase II elongation.

Methods Animals Setd2f/f (B6, CD45.2) mice were generated by Cincinnati Children’s Hospital Transgenic Core. Vav1-Cre, Mx1-Cre, Tie2Cre mice were purchased from Jackson Laboratory. All mice were housed in the rodent barrier facility at Cincinnati Children’s Hospital Medical Center (CCHMC).

Small molecular inhibitors treatment The CD117 positive selection of bone marrow (BM) cells was performed using magnetic CD117 microbeads (Miltenyi 130091-224) following the manufacturer's instructions. The CD117 positive fractions were cultured in medium (Stemspan+100 ng/mL SCF+100 ng/mL TPO) and treated with JQ1 500 nM, EPZ-5676 1uM, BAY 1143572 400 nM for 24 and 48 hours (h). The inhibitors were from the following companies: JQ1 (SigmaAldrich, SML0974), EPZ-5676 (Selleckchem, S7062), BAY 1143572 (MedChem Express, HY-12871). Details of the methods used are available in the Online Supplementary Appendix. haematologica | 2018; 103(7)

Results Generating a novel Setd2 conditional knockout allele Setd2 is involved in the ESCs differentiation and vascular formation during embryonic development. Setd2-/- mice are embryonic lethal.10 Thus, we generated a Setd2 conditional knockout allele by inserting two LoxP sites flanking Setd2 exon6, which encodes part of the SET domain. Deletion of exon 6 could result in frame-shift and trigger nonsense-mediated decay (NMD) of the mutant mRNA transcript (Online Supplementary Figure S1A). Three Cre transgenic lines were used: Tie2-Cre, Mx1-Cre, and Vav1Cre. Tie2-Cre mice display Cre activities in both endothelial cells and hematopoietic cells.17 However, we were unable to develop any Setd2f/f/Tie2-Cre mice by intercrossing Setd2f/w/Tie2-Cre mice with Setd2f/f mice in multiple litters (Online Supplementary Table S1), while polyinosinicpolycytidylic acid (pIpC) induced Setd2f/f/Mx1-Cre mice and Setd2f/f/Vav1-Cre mice are viable and fertile. Thus, we focused on Mx1-Cre and Vav1-Cre alleles to achieve Setd2f/f deletion in the hematopoietic system. First, to confirm Setd2 deletion, the mice were genotyped using tail tissue and peripheral blood by genomic PCR (Online Supplementary Figure S1B and C). The LoxP insertion and Setd2 deletion were confirmed in Setd2f/f and Setd2f/f/Vav1-Cre mice. Subsequently, the Setd2 expressions were confirmed to be dramatically decreased at both mRNA and protein levels in Setd2f/f/Vav1-Cre mice and pIpC induced Setd2f/f/Mx1-Cre mice BM cells (Figure 1A, Online Supplementary Figure S1D, and data not shown). Consistent with the role of Setd2 in regulating H3K36 methylation, global H3K36me3 was significantly reduced in BM cells of Setd2 knockout mice (Figure 1A).

Setd2Δ/Δ mice showed leukopenia, anemia, erythroid dysplasia, increased thrombopoiesis, and mild BM fibrosis Setd2f/f/Vav1-Cre mice are born small and pale. When the circulating blood count (CBC) was checked at eight weeks, they showed leukopenia, macrocytic anemia, and increased platelet count compared to the control littermates (Figure 1B-D). To exclude the possibility that the phenotype is contingent on deletion early in fetal hematopoiesis, we induced excision in 6- to 10-week old Setd2f/f/Mx1-Cre mice with pIpC injection. We found similar phenotypes two weeks after pIpC injection (Online Supplementary Figure S2A-C). Consistent with peripheral blood (PB) phenotype, the Setd2Δ/Δ mice, both Setd2f/f/Vav1-Cre and Setd2f/f/Mx1-Cre models, had 30% fewer nucleated BM cells, enlarged spleens, and obviously shrunken thymuses (Figure 1E-G and Online Supplementary Figure S2D), which were also confirmed by pathology. There are significantly increased erythroblasts and mature megakaryocytes in Setd2Δ/Δ mice BM compared to the controls (Online Supplementary Table S2). Notably, the percentage of erythroblasts gradually increases with aging and could even reach up to 80% in some mice (data not shown). Erythroid dysplasia could also be observed. Compared with the round nuclei of erythroblasts in control mice, erythroblasts in Setd2Δ/Δ mice showed frequent multi-nucleation, nuclear budding, nuclear fragments, and more cells in mitosis (Figure 1H). However, no obvious dysplasia was observed in other myeloid lineages or megakaryocytes. In addition to erythroid dysplasia, the histology showed increased reticulin 1111


Y. Zhou et al.

A

C

B

D

E

G

F

H

I

Figure 1. Setd2Δ/Δ mice showed leukopenia, anemia, erythroid dysplasia, increased thrombopoesis and mild bone marrow (BM) fibrosis. (A) Setd2 and H3K36me3 protein levels were determined by immunoblotting using c-Kit+ BM cells. Representative data were from 3 independent experiments. [N=3; mean±Standard Deviation (SD)]. (B) Complete blood count of Setd2f/f/Vav1-Cre mice, showing reduced white blood cells, lymphocytes, neutrophils, and platelets. [N=8 mice per genotype; mean±Standard Error of Mean (SEM)]. (C) Representative photos of Wright’s stained peripheral blood smear of Setd2f/f and Setd2f/f/Vav1-Cre mice. (D) Complete blood count of Setd2f/f/Vav1-Cre mice, showing reduced red blood cells, hemoglobin content, red blood cell specific volume (HCT), mean corpuscular hemoglobin concentration (MCHC), but increased mean corpuscular volume of red cells (MCV) and mean corpuscular hemoglobin (MCH). (N=8 mice per genotype; mean±SEM). (E) Representative photos of bones (tibia and fibula), spleens, and thymuses in Setd2f/f and Setd2f/f/Vav1-Cre mice. (F) BM cellularity, spleen weight, and thymus weight of Setd2f/f and Setd2f/f/Vav1-Cre mice. (N≥4 mice per genotype; mean±SEM). (G) Representative photos of hematoxylin & eosin-stained sections from the sternum, spleens, thymuses of Setd2f/f and Setd2f/f/Vav1-Cre mice. (H) Dysplastic erythroid cells can be found in BM cytospin: megaloblastic erythroid precursors, dysplatic erythroid precursors with multi-nucleation, nuclear fragments, or nuclear budding. In addition, erythroid cells can be caught in mitosis. (I). Representative photos of reticulin-stained sections from sternum of Setd2f/f and Setd2f/f/Vav1-Cre mice.

1112

haematologica | 2018; 103(7)


Setd2 regulates hematopoietic stem cells staining in Setd2Δ/Δ mice BM (Figure 1I). Compared with some scattered punctual and linear reticulin in the control, scattered linear reticulin with loose network and some focal density increases in reticulin could be found in Setd2Δ/Δ BM within four months, which could be classified into mild BM fibrosis (grade 1 or 2).18

Setd2Δ/Δ mice showed profound reduction of myeloid, lymphoid, and megakaryocytic progenitors, but significantly increased erythroid progenitors

To understand leukopenia and anemia in Setd2Δ/Δ mice, we first examined the BM progenitor populations by flow cytometry. Significant reductions in the absolute number of CLP, Pre-GM, GMP, Pre-MegE, and MkP were found, while the absolute number of Pre CFU-E was dramatically increased (Figure 2A and B). It is noteworthy that Setd2Δ/Δ mice showed anemia in PB but significantly increased Pre CFU-Es in BM. To understand the remarkable differences between these two phenotypes, a detailed analysis of erythroid differentiation was performed. There were increased proportions of nucleated erythroblasts accompanied by a decreased proportion of enucleated erythrocytes in Setd2Δ/Δ mice (Online Supplementary Figure S3A and B), indicating the defective terminal erythroid differentiation. Meanwhile, a reduction of MkPs in BM, accompanied by an increase in platelet counts in PB, was observed in Setd2Δ/Δ mice. In the analysis of polyploidy using BM CD41+ cells, the Setd2Δ/Δ mice displayed significantly increased distributions of hyper-polyploidy (16N and 32N) cells and reduced distributions of low- to intermediate-ploidy (2N-8N) cells (Online Supplementary Figure S3C and D). In addition, more megakaryocytes were observed in both BM cytospins and spleen histology slides in Setd2Δ/Δ mice compared with control (Online Supplementary Table S2 and Online Supplementary Figure S3E). To determine HPC functional activity besides phenotypic changes, we performed colony-forming unit (CFU) assays, which showed that almost all types of colony numbers were decreased except the burst-forming uniterythroid (BFU-E). Interestingly, the BFU-E could even be detected in the 4th replating, while all other colonies stopped growing after three incidences of replating in the controls (Figure 2C). To further confirm the erythroidrelated results, the CFU-E assays were performed in M3334 medium, which contains erythropoietin (EPO) only. After 48 h, Setd2Δ/Δ BM cells showed significantly increased BFU-E/CFU-E colony frequencies and the colonies were larger in size compared with colonies from the controls (Figure 2D). These results indicate that Setd2 is critical in maintaining normal HPC numbers and lineage specification.

Depletion of phenotypic and functional HSCs in Setd2Δ/Δ mice Next, we examined the bone marrow HSC populations. Significant reductions, in absolute number and frequency, of LSKs, SLAM-HSCs, and MPPs were found in Setd2Δ/Δ mice compared with the controls (Figure 3A and B). To determine the HSC activity, a series of bone marrow transplantation (BMT) assays were performed. We first evaluated the HSC function in a competitive bone marrow transplantation assay (CBMT). Lethally irradiated CD45.1+ recipient mice were transplanted using an equal number of BM cells from both CD45.1+ competitors and haematologica | 2018; 103(7)

CD45.2+ Setd2f/f or Setd2f/f/Mx1-Cre. Sixteen weeks after CBMT, Setd2Δ/Δ cells were outcompeted to less than 1% in PB (Figure 3C). In the BM, Setd2Δ/Δ failed to support longterm reconstitution of the Gr1+CD11b+ population (myeloid), B220+ population (B cells), and CD3+ population (T cells). Analysis of BM LSKs showed a complete absence of Setd2Δ/Δ LSKs (Figure 3D). Similar results were observed by using Setd2/Vav1-Cre BM cells in CBMT assay (Figure 3E). To exclude the possibility that BM microenvironment defects (such as endothelial cells and stromal BM cells) may contribute to HSC dysfunction, we transplanted BM cells from Setd2f/f or Setd2f/f/Mx1-Cre mice into lethally irradiated CD45.1+ recipient mice. We found similar engraftment four weeks after transplantation, around 90% engraftment in both groups. Then we deleted Setd2 in donor-derived grafts with pIpC injection. We found decreased donor-derived cell chimerism in Setd2Δ/Δ mice (Figure 3F). The PB phenotypes of BMT mice are similar to the primary knockout mice, which also manifested leukopenia, macrocytic anemia, and increased platelet counts (Figure 3G). However, when non-competitive transplantation with Setd2f/f/Vav1-Cre BM cells were performed, all the recipients died within 75 days (Figure 3H). When complete PB count was checked at 28 days post BMT, recipient mice showed severe pancytopenia, indicating the failure of BM reconstitution (Online Supplementary Figure S4A). These results indicate that Setd2Δ/Δ HSCs have intrinsic defects in BM reconstitution.

Setd2Δ/Δ HSCs show reduced self-renewal and quiescence, but increased proliferation, apoptosis, and differentiation.

To identify the Setd2Δ/Δ HSC functions under chemotherapeutic stress, we next challenged Setd2Δ/Δ and control mice with 5-fluorouracil (5-FU). In the 8-day recovery group, a single 5-FU treatment resulted in a 10fold reduction of BM cellularity and a 2-fold reduction of SLAM-HSCs in the control, but a 20-fold reduction of BM cellularity and a 5-fold reduction in Setd2Δ/Δ mice (Figure 4A). In the 5-FU weekly treated group, Setd2Δ/Δ mice could tolerate 2 cycles of 5-FU injections, while all the control mice could tolerate 3 cycles (Figure 4B). These results indicated that Setd2Δ/Δ mice were more sensitive to 5-FU. 5-FU kills dividing cells but spares quiescent cells, such as stem cells, and subsequently forces HSCs to proliferate to reconstitute the BM; thus Setd2Δ/Δ HSCs might have intrinsic defects in maintaining normal quiescence. Next, the cell cycle status was assessed. Setd2Δ/Δ HSCs had a markedly reduced G0 fraction and increased entries into G1 and S/G2/M phases of the cell cycle (Figure 4C). Setd2Δ/Δ SLAM-HSCs also exhibited increased incorporation of BrdU into the DNA, indicative of more cycling cells (Online Supplementary Figure S4B). These results suggested that Setd2Δ/Δ HSCs could not maintain a normal quiescent state and subsequently enter the cell cycle. Also, apoptotic status was assessed in Setd2Δ/Δ and control mice. There were significantly increased proportions of Annexin V+ SLAM-HSC, LSK, and LK cells, which demonstrated that Setd2Δ/Δ HSPCs underwent more cell death (Figure 4D and Online Supplementary Figure S4C). To investigate the differentiation potential of Setd2Δ/Δ HSCs, single SLAM-HSCs were sorted and cultured in cytokine-containing medium. With this culture condi1113


Y. Zhou et al.

tion, 4 lineages could be observed, erythroid cells, megakaryocytes, neutrophils, and macrophages, in the single SLAM-HSC generated clones after 10-14 days. First, Setd2Δ/Δ SLAM-HSCs showed decreased clonogenicities, as there were significantly less HSCs that could gen-

erate clones in each plate in the Setd2Δ/Δ group compared with the control, while the sorting efficiency was comparable between these two groups (Figure 4E). In addition, Setd2Δ/Δ SLAM-HSCs produced fewer frequencies of 4-lineage clones compared with the control, accompanied

A

B

C

D

Figure 2. Setd2Δ/Δ mice showed profound reduction of myeloid, lymphoid and megakaryocyte progenitors but significantly increased erythroid progenitors. (A) Flow cytometry analysis of Setd2f/f and Setd2f/f/Vav1-Cre mice bone marrow (BM) cells. (B) Absolute number of hematopoietic progenitor cell (HPC) populations. [Setd2f/f =4, Setd2f/f/Vav1-Cre=5; mean±Standard Error of Mean (SEM)]. (C) Colony-forming cell (CFU) using BM cells from Setd2f/f or Setd2f/f/Vav1-Cre. 2x104 cells were plated in M3434 in triplicate and colonies were scored every seven days. GEMM: granulocyte, erythroid, macrophage, megakaryocyte colony; GM: granulocyte/macrophage; G/M: granulocyte or macrophage; BFU-E: burst formation unit-erythroid. Representative data were from 3 independent experiments. (N=3; mean±Standard Deviation (SD)]. (D) CFU-erythroid (CFU-E) assay using BM cells from Setd2f/f or Setd2f/f/Vav1-Cre. 5x105 cells were plated in M3334 in triplicate and colonies were scored 48 hours later. Representative data were from 3 independent experiments. (N=3; mean±SD).

1114

haematologica | 2018; 103(7)


Setd2 regulates hematopoietic stem cells

Setd2Δ/Δ HSPCs show a loss of stem cell identity and an increase in differentiation toward progenitors

with increased frequencies of 3-, 2-, and 1-lineage clones (Figure 4E). These results indicated that Setd2Δ/Δ SLAMHSCs lost normal clonogenicity and multi-lineage differentiation potential, which indicated that Setd2Δ/Δ SLAMHSCs were in a more differentiated state.

A

C

There was a 4-fold reduction of SLAM-HSC numbers in Setd2Δ/Δ mice but a complete loss of HSC functions in the BMT assay. Thus, we consider it unlikely that the reduc-

B

D

E

F

G

H

Figure 3. Setd2Δ/Δ mice had depletion of phenotypic and functional hematopoietic stem cells (HSCs). (A) Flow cytometry analysis of Setd2f/f and Setd2f/f/Vav1-Cre mice bone marrow (BM) cells. (B) Absolute number of HSC populations. (Setd2f/f =4, Setd2f/f/Vav1-Cre=5; mean±Standard Error of Mean (SEM)] (C). Experimental strategy: Setd2f/f or Setd2f/f/Mx1-Cre (pIpC injected) CD45.2 BM cells (1.5×106 cells each) was injected into irradiated (7.5+4.25Gy) B6-CD45.1 recipients, with B6-CD45.1 competitor BM (1.5×106 cells each). Peripheral blood (PB) was analyzed 2-16 weeks after competitive transplantation. Representative data were from 2 independent experiments. (N=8 each genotype; mean±SEM). (D) CD45.1/CD45.2 chimerism in BM LSKs, Gr1+CD11b+ myeloid cells, B220+ B cells, CD3+ T cells 16 weeks after transplantation. Representative data were from 2 independent experiments. (N=8 each genotype; mean±SEM). (E) Experimental strategy: Setd2f/f or Setd2f/f/Vav1-Cre CD45.2 BM cells (1.5×106 cells each) was injected into irradiated (7.5+4.25Gy) B6-CD45.1 recipients, with B6-CD45.1 competitor BM (1.5×106 cells each). PB was analyzed 4-16 weeks after competitive transplantation. Representative data were from 2 independent experiments. (N=8 each genotype; mean±SEM). (F and G) Experimental strategy: Setd2f/f or Setd2f/f/Mx1-Cre CD45.2 BM was injected into irradiated (7.5+4.25Gy) B6-CD45.1 recipients (2×106 cells per genotype), with B6-CD45.1 helper BM (1×105 cells). pIpC was injected two weeks after BMT. Peripheral blood were analyzed 0-10 weeks after pIpC injection. Representative data were from 2 independent experiments. (N=8 each genotype; mean±SEM). (H) Experimental strategy: Setd2f/f or Setd2f/f/Vav1-Cre CD45.2 BM was injected into irradiated (7.5+4.25Gy) B6-CD45.1 recipients (2×106 cells per genotype), with B6-CD45.1 helper BM (1×105 cells), survival conditions were monitored. (N=6 each genotype; mean±SEM).

haematologica | 2018; 103(7)

1115


Y. Zhou et al.

A

B

C

D

E

Figure 4. Setd2Δ/Δ hematopoietic stem cells (HSCs) show reduced quiescence, but increased apoptosis and differentiation. (A) Experimental strategy: Setd2f/f and Setd2f/f/Vav1-Cre mice were injected with 150 mg/kg 5-fluorouracil (FU) and sacrificed eight days later (left). Statistical analyses of the bone marrow (BM) cellularity and absolute number of long-term (LT)-HSCs analyzed by flow cytometry (right). Representative data were from 3 independent experiments. (N=6 each genotype; mean±Standard Error of Mean (SEM)]. (B) Experimental strategy: Setd2f/f and Setd2f/f/Vav1-Cre mice were injected with 150 mg/kg 5-FU weekly and monitored for survival (left). Survical curve (right) (n=6 each genotype). (C) Flow cytometry analysis of Setd2f/f and Setd2f/f/Vav1-Cre BM cells with Ki-67 and 7-AAD. Gating strategy is shown in one representative FACS blots per genotype (left). Summary of statistical analyses showed decreased G0 distribution in Setd2f/f/Vav1-Cre LT-HSCs (right). Representative data were from 3 independent experiments. (N=6 each genotype; mean±SEM). (D) Flow cytometry analysis of Setd2f/f and Setd2f/f/Vav1-Cre BM cells with Annexin V and 7-AAD. Gating strategy is shown in one representative FACS blots per genotype (left). Summary of statistical analyses showed increased distribution into AnnexinV positive fraction in Setd2f/f/Vav1-Cre SLAM-HSCs. Representative data were from 3 independent experiments. (N=6 each genotype; mean ± SEM). (E) The bar figure shows the number of clones generated by Setd2f/f and Setd2f/f/Vav1-Cre single LT-HSCs per 60-well plate (top left). Pie chart shows the relative frequencies of 4-lineage, 3-lineage, 2-lineage, and 1-lineage clones generated from Setd2f/f and Setd2f/f/Vav1-Cre single LT-HSCs (bottom left). Wright’s stained cytospin showed representative pictures of 4-lineage, 3-lineage, 2-lineage, and 1-lineage clones generated from Setd2f/f and Setd2f/f/Vav1-Cre single LT-HSCs (right). Representative data were from 2 independent experiments. (N=4 each genotype; mean±SEM).

1116

haematologica | 2018; 103(7)


Setd2 regulates hematopoietic stem cells

tion of SLAM-HSC is the sole reason to explain the complete loss of HSC functions. We used CD150 and CD48 to define SLAM-HSC, which remains heterogeneous, as reported in recent studies.20-22 We further analyzed quiescent long-term HSCs with additional defined surface markers besides CD48 and CD150: CD135, CD34, CD201, and CD49b. The results showed that there was a dramatic reduction of CD201+CD49b- fraction and a significantly increased entry of CD201-CD49b- fraction in the CD48-CD150+CD135-CD34-LSK populations in Setd2Δ/Δ mice compared with the control (Figure 5A). It has been reported that CD201+ is an accurate marker in SLAMHSCs to define the quiescent stem cell pool which is capable of repopulation in bone marrow transplantation,23,24 and CD201+CD49b-CD135-CD34- SLAM-HSC population is reputed, as the “true HSC”, to have the longest-term self-renewal capacity in deep quiescence.25 Our data indicate that the real quiescent fractions are further reduced in the decreased SLAM-HSC population in Setd2Δ/Δ mice. To explore the molecular mechanisms underlying HSC regulation by Setd2, we performed RNA-seq using LSKs from Setd2f/f and Setd2Δ/Δ mice. Unbiased Gene Set Enrichment Analysis (GSEA) revealed that the HSC, longterm HSC, and short-term HSC signature genes26 were all significantly down-regulated, while intermediate and late progenitor signature genes26 were significantly up-regulated in Setd2Δ/Δ LSKs (Figure 5B), implying that immune-phenotypically defined Setd2Δ/Δ HSCs lost the stem cell identity and differentiated toward multipotent progenitors (MPP). Gene ontology analysis of top differentially expressed genes indicated that lineage development/differentiation-related genes were significantly up-regulated in Setd2Δ/Δ LSKs (Figure 5C and D, and Online Supplementary Table S3), including Gata1, Gata3, and Klf1, which are important in HSC differentiation toward myeloid and lymphoid lineages. Collectively, loss of Setd2 could induce LT-HSCs to exit from quiescence and commit to differentiation, leading to the exhaustion of LT-HSCs, IT-HSCs, MPP1, and the differentiation to the MPP2, MPP3, and MPP4 populations (Figure 5E).

Setd2Δ/Δ HSPCs show increased Nsds and RNA Pol II elongation associated phosphorylation changes To understand whether loss of Setd2 and H3K36me3 would affect other H3K36-methyltransferases and subsequent methylation states of H3K36, the expression levels of 4 other most closely related enzymes were assessed using LSK cells from BM. Our data showed that Ash1l was decreased, but Nsd1/2/3 were all significantly increased at both mRNA and protein levels (Figure 6A and Online Supplementary Figure S5A). Interestingly, when we overexpressed WT NSD2, or gain-of-function (GOF) mutant of NSD2 (E1099K) in a murine Mll-Af9 leukemia cell line, both WT NSD2 and the GOF of NSD2-E1099K showed similar H3K36me3/2 changes to LOF of Setd2 (Online Supplementary Figure S5B), which implies that Setd2 and Nsds actually antagonize each other’s function. LOF SETD2 and GOF NSD2 in human leukemia may result in similar transcriptional dysregulation. Indeed, we found that H3K36me1 and H3K36me2 were dramatically increased, correlated with the up-regulated Nsd1/2/3 (Figure 6B). In addition, H3K4me3 and H3K79me2 were also significantly increased, while H3K27me3 was slightly decreased (Figure 6B), which indicated the promoting haematologica | 2018; 103(7)

transcriptional elongation of RNA polymerase II. The significant increase in elongation-associated phosphorylation changes [RNA pol II (Ser5P) and pol II (Ser2P)] were further confirmed by immunoblotting (Figure 6C). Thus, we hypothesized that Setd2 knockout up-regulates the RNA pol II transcriptional elongation to activate a subset of genes, which could affect the identity and functions of HSCs. Our bulk RNA-seq was performed with LSK cells due to limited cell numbers. Next we aimed to define a subset of genes that were up-regulated in Setd2Δ/Δ SLAM-HSCs. To identify some candidate genes, the expression profiles of Setd2 and transcriptional elongation related genes and complexes were checked using a published database.27 The results showed that BET family genes and some wellknown super elongation regulating genes (Myc, Mycn, Myb etc.) were significantly up-regulated during HSC differentiation (Online Supplementary Figure S6). Thus, we performed RT-PCR on these up-regulated genes with sorted Setd2Δ/Δ SLAM-HSCs. The results showed the dramatic upregulation of Gata1, Gata3, and Klf1, which was consistent with our RNA-seq data (Figure 6D). At the same time, we noticed that Myc was also significantly up-regulated in Setd2Δ/Δ SLAM-HSCs (Figure 6D). Myc is well known to be very sensitive to RNA pol II promoter-proximal pausing and releasing from pausing by elongation changes. Importantly, the phenotypes of Setd2Δ/Δ HSCs recapitulated the Myc overexpression situation. It has been reported that enforced expression of Myc in SLAM-HSCs promotes differentiation at the expense of self-renewal, inducing exit from quiescence, increased apoptosis, and failure to reconstitute BM in BMT assay,28 which phenocopies Setd2Δ/Δ HSCs. Also, we confirmed the dramatic upregulation of Gata1, Gata3, Klf1, and Myc at the protein level (Figure 6C). As Myc is a well-studied gene, which regulates the entry and exit from stem cell quiescence during development, we first confirmed that the significant increase of Myc is due to enhanced RNA pol II elongation. ChIP-qPCR assays of Setd2, H3K36-related histone modifications, and Pol II were performed at the Myc locus using c-kit+ cells from Setd2f/f and Setd2Δ/Δ mice. The results showed that there was a significantly higher enrichment of pol II (Ser5P) and Pol II (Ser2P) occupancy along the whole gene body, promoter and enhancer regions of Myc in Setd2Δ/Δ HSPCs compared with control.29,30 Meanwhile, H3K36me2 occupancy also showed a significantly higher enrichment along the gene body, both promoter and enhancers, while the H3K36me1 mainly increased along the enhancer region. As expected, enrichment of Setd2 and H3K36me3 was dramatically reduced, especially at the gene body region (Figure 6E).

Setd2Δ/Δ deficiencies could be partially rescued by super elongation complex-related inhibitors Next, we tested whether the increased gene expressions, such as Myc, could be reversed by epigenetic drugs. The c-kit+ Setd2Δ/Δ cells were treated in an in vitro culture assay with super elongation complex (SEC) related inhibitors for 24 and 48 h: JQ1 (Brd4 inhibitor), EPZ-5676 (Dot1l inhibitor), and BAY 1143572 (p-TEFb/CDK9 inhibitor). The elevated H3K36me1/2 marks were not affected 48 h after treatment (Figure 7A), while the expression levels of pol II (Ser2P), pol II (Ser5P), Gata1, Gata3, and Myc were significantly decreased in all 3 drug-treated 1117


Y. Zhou et al.

groups (Figure 7B). This indicated that Nsds-mediated H3K36me1/2 modifications were upstream of SEC complex recruitment and releasing promoter-proximal pausing of pol II. Meanwhile, SEC complex inhibitors could not affect other genes such as Gapdh and β-actin (Figure 7B).

Besides the changes of protein expression levels, the functional changes were also assessed. The c-kit+ Setd2Δ/Δ cells were treated in vitro for 24 h with 3 inhibitors respectively, and LSK populations were further gated to analyze the apoptotic and cell cycle status. There were significant-

A

B

C

E

D

Figure 5. Setd2Δ/Δ hematopoietic stem cells (HSCs) show loss of stem cell identity and increased differentiation toward progenitors. (A) Flow cytometry analysis of Setd2f/f and Setd2f/f/Vav1-Cre mice bone marrow (BM) cells. [N=3 each genotype; mean±Standard Error of Mean (SEM)]. (B) Gene Set Enrichment Analysis (GSEA) for genes affected in the LSKs of Setd2f/f and Setd2f/f/Vav1-Cre mice, after RNA-seq analyses. (A-E) Enrichment of HSC, long-term (LT)-HSC, short-term (ST)-HSC, early progenitors, intermediate progenitors, and late progenitors gene sets in LSKs of Setd2f/f and Setd2f/f/Vav1-Cre mice, respectively. All gene sets are from GSEA molecular signature database.26 (C and D) Up-regulated gene ontology analysis of differentially expressed genes. (E) Diagram of Setd2Δ/Δ HSPC differentiation.

1118

haematologica | 2018; 103(7)


Setd2 regulates hematopoietic stem cells

ly decreased Annexin V+ fractions and increased G0 fractions in all 3 drug-treated groups compared with PBS treated group (Figure 7C-F), which indicates that SEC complex inhibitors could partially rescue the HSC functional deficiencies in Setd2Δ/Δ cells. However, the in vitro single-cell

A

B

differentiation assays failed due to the toxicity of those inhibitors in a long-term treatment. Collectively, based on our results and published data, we propose a model for Setd2 function in HSCs (Figure 8). Setd2 loss leads to the significant upregulation of Nsds,

C

D

E

Figure 6. Setd2Δ/Δ hematopoietic stem cells (HSCs) show increased Nsds and RNA Pol II elongation associated phosphorylation changes. (A) Nsd1/2/3 and β-actin levels were determined by immunoblotting using bone marrow (BM) LSK cells. (B and C) H3K36me1/2, H3K4me3, H3K79me2, H3K27me3, H3, RNA pol II (Ser2P), pol II (Ser5P), Gata1, Gata3, Klf1, Myc, and β-actin levels were determined by immunoblotting using BM LSK cells from Setd2f/f and Setd2f/f/Vav1-Cre mice. (D) Relative gene expression levels were determined by qrt-PCR using flow sorted SLAM-HSCs from Setd2f/f and Setd2f/f/Vav1-Cre mice. Representative data were from 3 independent experiments. [N=6 each genotype; mean±Standard Error of Mean (SEM)]. (E) ChIP-qPCR assays of Setd2, Setd2 related histone modifications, and pol II phosphorylated forms [Pol II (Ser5P) and pol II (Ser2P)] on Myc locus was determined with c-kit+ BM cells from Setd2f/f and Setd2f/f/Vav1-Cre mice. (Setd2f/f=5 and Setd2f/f/Vav1-Cre=8; mean±SEM from 2 independent experiments).

haematologica | 2018; 103(7)

1119


Y. Zhou et al.

which could recruit Brd4,31-34 whereas Brd4 could recruit DOT1L complex and p-TEFb complex to increase p-TEFbdependent phosphorylation of pol II CTD and stimulate transcription from promoters that have promoter-proximal pausing.35-37 Such enhanced pol II elongation could result in the upregulation of a subset of genes including Myc, which controls the balance of quiescence and differentiation in HSCs.

A

Discussion In this study, we generated a novel conditional knockout allele (Setd2Δ/Δ). These mice manifested leukopenia, macrocytic anemia, increased platelet, and erythroid dysplasia in BM, which are comparable to the phenotypes of myelodysplastic syndromes (MDS) associated with isolated del(5q). It has been reported that rare SETD2 mutations

B

C

D

E

F

Figure 7. Setd2Δ/Δ deficiencies could be partially rescued by super elongation complex-related inhibitors. (A and B) Treatment of c-Kit+ bone marrow (BM) cells from Setd2f/f/Vav1-Cre mice with JQ1 500nM, EPZ-5676 1uM, BAY 1143572 400 nM for 24 h (h). Then cells were collected to determine the proteins levels of H3K36me1/2, H3K4me3, H3K79me2, H3K27me3, H3, RNA pol II (Ser2P), pol II (Ser5P), Gata1, Gata3, Myc, and β-actin by immunoblotting. (C and D) Flow cytometry analysis of drugtreated c-kit+ cells from Setd2f/f and Setd2f/f/Vav1-Cre mice with Annexin V and 7-AAD. Gating strategy is shown in one representative FACS blot per genotype. Summary of statistical analyses showed decreased distribution into AnnexinV positive fraction in drug-treated groups. Representative data were from 3 independent experiments. (Setd2f/f=6 and Setd2f/f/Vav1-Cre=12). (E and F) Flow cytometry analysis of drug treated c-kit+ cells from Setd2f/f and Setd2f/f/Vav1-Cre mice with Ki67 and Hoechst blue. Gating strategy is shown in one representative FACS blot per genotype. Summary of statistical analyses showed increased distribution into G0 fraction in drug-treated groups. Representative data are from 3 independent experiments; (Setd2f/f=6 and Setd2f/f/Vav1-Cre=12).

1120

haematologica | 2018; 103(7)


Setd2 regulates hematopoietic stem cells

were identified in MDS patients,38 but whether SETD2 is involved in MDS development is still unclear. To examine whether SETD2 is down-regulated in MDS del(5q), we checked SETD2 expression levels in an MDS del(5q) cohort with 3 different probes. However, no significant differences between del(5q) and the control groups were found (Online Supplementary Figure S5A). Besides erythroid dysplasia and BM fibrosis,39 there are increased proportions of erythroblasts in the BM. Moreover, as the Setd2Δ/Δ mice became older, the percentage of erythroblasts accumulated and could even reach up to 80% in some mice, which is somewhat similar to the BM of acute erythroid leukemia (pure erythroid type).40 It is clear that Setd2 plays a critical role in maintaining the identity and functions of HSCs. Setd2Δ/Δ mice showed reduced HSCs and capability of BM reconstitution after transplantation. Setd2Δ/Δ HSCs showed loss of quiescence, increased apoptosis, and reduced multi-potent differentiation potential. Unbiased GSEA and GO analysis also indicated the upregulation of lineage development/differentiation pathways and related genes. Thus, there could be two explanations for the dramatic reduction in HSC numbers. First, some HSCs exited from quiescence and committed to differentiation. The balance between selfrenewal and differentiation is important for the maintenance of the stem cell pool. Pushing HSCs to differentiate

would come at the expense of self-renewal and lead to the exhaustion of HSCs. The other reason could be that some HSCs directly underwent cell death. However, Setd2Δ/Δ mice did not progress to pancytopenia and BM failure without a stress challenge; this is in line with the finding that, at steady stage, a limited number of HSPCs would be sufficient to maintain normal hematopoiesis.41 Setd2, a histone methyltransferse, regulates H3K36me3. The direct effect after Setd2 knockout is the impact on histone modifications and other closely related H3K36 methyltransferases. Thus, the expression levels of Ash1l, Nsd1/2/3, and related histone markers were assessed first. The results showed significantly up-regulated Nsd1/2/3, accompanied with increased H3K36me1 and H3K36me2. Recently, we found that downregulation of SETD2 leads to a global elevation of DOT1L-mediated H3K79me2 in MLL-AF9 leukemia.42 Consistent with this finding, we observed a dramatic increase in H3K79me2 and H3K4me3, implying the promoting of transcriptional elongation. The enhanced RNA pol II elongation was further confirmed by the up-regulated pol II (Ser2P) and pol II (Ser5P) phosphorylations. We confirmed the significant upregulation of Gata1, Gata3, Klf1, and Myc in Setd2Δ/Δ SLAM-HSCs. These subsets of genes are sensitive to enhanced pol II elongation. The enhanced elongation after Setd2 knockout was

Figure 8. The diagram of our working model. In normal adult hematopoietic stem cells (HSCs), Setd2, responsible for H3K36me3, could repress Nsds, which are responsible for H3K36me1/2. Nsds interact with Brd4, p-TEFb, and Dot1l to stimulate transcriptional elongation. On the other hand, Setd2 binds to pol II (Ser2P) and pol II (Ser5P) doubly modified CTD repeats. Thus, a subset of genes, such as Myc, is maintained at a proper level to keep the balance between quiescence and differentiation of adult stem cells (top). In Setd2Δ/Δ HSCs, Setd2 loss leads to the upregulation of NSDs, which would further enhance the Pol II phosphorylation and elongation, resulting in the upregulation of Myc. When treated with Brd4/Dot1l/p-TEFb inhibitors, the pol II (Ser2), pol II (Ser5), and the expressions of the Myc could be down-regulated (bottom).

haematologica | 2018; 103(7)

1121


Y. Zhou et al.

clearly indicated by globally elevated marks. The higher enrichment of pol II (Ser5P) and pol II (Ser2P) occupancy were also confirmed on Myc locus in Setd2Δ/Δ HSPCs. Setd2 and H3K36me3 generally mark the active genes; however, our surprising findings indicated that Setd2-H3K36me3 restrict the pol II elongation. To connect the Setd2 loss to enhanced elongation, we observed upregulations of Nsd1/2/3 after Setd2 knockout. There is much literature showing that NSDs could interact with BRD4, which could bridge to the SEC and DOT1l complex.31-34 Thus, we proposed a regulatory model in which there is a crosstalk between Setd2 and Nsds. Loss of Setd2 leads to the upregulation of Nsds. Meanwhile, Nsds could interact with Brd4, SEC, and Dot1l complex to enhance the elongation, and results in the upregulation of a subset of target genes that regulate quiescence and differentiation of HSCs. In summary, using our novel Setd2 conditional knockout allele, we revealed unique roles of Setd2 in regulating quiescence and differentiation of HSCs. Our study not only provides us with a deeper understanding of Setd2 functions in HSCs, but also a better understanding of SETD2 functions during leukemic transformation and solid tumors. Along with using the KDM4 inhibitor to

References 1. Kiel MJ, Yilmaz OH, Iwashita T, et al. SLAM family receptors distinguish hematopoietic stem and progenitor cells and reveal endothelial niches for stem cells. Cell. 2005;121(7):1109-1121. 2. Orkin SH, Zon LI. Hematopoiesis: an evolving paradigm for stem cell biology. Cell. 2008;132(4):631-644. 3. Dawson MA, Kouzarides T. Cancer epigenetics: from mechanism to therapy. Cell. 2012;150(1):12-27. 4. Oike T, Ogiwara H, Amornwichet N, Nakano T, Kohno T. Chromatin-regulating proteins as targets for cancer therapy. J Radiat Res. 2014;55(4):613-628. 5. Strahl BD, Grant PA, Briggs SD, et al. Set2 is a nucleosomal histone H3-selective methyltransferase that mediates transcriptional repression. Mol Cell Biol. 2002;22(5):12981306. 6. McDaniel SL, Strahl BD. Shaping the cellular landscape with Set2/SETD2 methylation. Cell Mol Life Sci. 2017;74(18):3317-3334. 7. Mao M, Fu G, Wu JS, et al. Identification of genes expressed in human CD34(+) hematopoietic stem/progenitor cells by expressed sequence tags and efficient fulllength cDNA cloning. Proc Natl Acad Sci USA. 1998;95(14):8175-8180. 8. Wagner EJ, Carpenter PB. Understanding the language of Lys36 methylation at histone H3. Nat Rev Mol Cell Biol. 2012;13(2):115126. 9. Vougiouklakis T, Hamamoto R, Nakamura Y, Saloura V. The NSD family of protein methyltransferases in human cancer. Epigenomics. 2015;7(5):863-874. 10. Zhang Y, Xie S, Zhou Y, et al. H3K36 histone methyltransferase Setd2 is required for murine embryonic stem cell differentiation toward endoderm. Cell Rep. 2014;8(6):19892002. 11. Hu M, Sun XJ, Zhang YL, et al. Histone H3

1122

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

restore the H3K36me3,43 inhibiting the elongation complex or downstream targets, could be effective for LOF mutation SETD2 in cancer and could also benefit bi-allele SETD2 mutant patients. Our data also indicate that GOF mutation of NSDs, such as NSD2 mutations or NSD1/3 translocations, might follow the same regulatory dysregulation as LOF of SETD2 on pol II elongation in leukemia and other cancers. Acknowledgments We would like to thank Cincinnati Children’s Research Flow Cytometry Core (RFCC), Cincinnati Children’s Veterinary Services, and CCHMC mouse core. We would also like to thank Damien Reynaud and George Mike Freudiger for their help on this project. This work was supported by grants from the Ministry of Science and Technology of China (2016YFA0100600) (to TC), the National Natural Science Foundation of China (81421002) (to TC), CAMS Initiative for Innovative Medicine (2016-I2M-1-017) (to TC), Leukemia Research Innovative Team of Zhejiang Province (2011R50015) (to JJ), National Natural Science Foundation of China (81370643-H0812) (to JJ), the CFK (to GH), National Institutes of Health (NIH) (R21CA187276) (to GH).

lysine 36 methyltransferase Hypb/Setd2 is required for embryonic vascular remodeling. Proc Natl Sci USA. 2010;107(7):2956-2961. Fahey CC, Davis IJ. SETting the Stage for Cancer Development: SETD2 and the Consequences of Lost Methylation. Cold Spring Harb Perspect Med. 2017;7(5). Zhang J, Ding L, Holmfeldt L, et al. The genetic basis of early T-cell precursor acute lymphoblastic leukaemia. Nature. 2012;481 (7380):157-163. Parker H, Rose-Zerilli MJ, Larrayoz M, et al. Genomic disruption of the histone methyltransferase SETD2 in chronic lymphocytic leukaemia. Leukemia. 2016;30(11):21792186. Moffitt AB, Ondrejka SL, McKinney M, et al. Enteropathy-associated T cell lymphoma subtypes are characterized by loss of function of SETD2. J Exp Med. 2017;214(5):1371-1386. Zhu X, He F, Zeng H, et al. Identification of functional cooperative mutations of SETD2 in human acute leukemia. Nat Genet. 2014;46(3):287-293. Speck NA, Iruela-Arispe ML. Conditional Cre/LoxP strategies for the study of hematopoietic stem cell formation. Blood Cells Mol Dis. 2009;43(1):6-11. Kvasnicka HM, Beham-Schmid C, Bob R, et al. Problems and pitfalls in grading of bone marrow fibrosis, collagen deposition and osteosclerosis - a consensus-based study. Histopathology. 2016;68(6):905-915. Cheng T, Rodrigues N, Shen H, et al. Hematopoietic stem cell quiescence maintained by p21cip1/waf1. Science. 2000;287(5459):1804-1808. Oguro H, Ding L, Morrison SJ. SLAM family markers resolve functionally distinct subpopulations of hematopoietic stem cells and multipotent progenitors. Cell Stem Cell. 2013;13(1):102-116. Raaijmakers MH, Scadden DT. Divided within: heterogeneity within adult stem cell pools. Cell. 2008;135(6):1006-1008.

22. Wilson A, Laurenti E, Oser G, et al. Hematopoietic stem cells reversibly switch from dormancy to self-renewal during homeostasis and repair. Cell. 2008;135(6):1118-1129. 23. Balazs AB, Fabian AJ, Esmon CT, Mulligan RC. Endothelial protein C receptor (CD201) explicitly identifies hematopoietic stem cells in murine bone marrow. Blood. 2006;107(6):2317-2321. 24. Kent DG, Copley MR, Benz C, et al. Prospective isolation and molecular characterization of hematopoietic stem cells with durable self-renewal potential. Blood. 2009;113(25):6342-6350. 25. Schroeder T. Hematopoietic stem cell heterogeneity: subtypes, not unpredictable behavior. Cell Stem Cell. 2010;6(3):203-207. 26. Ivanova NB, Dimos JT, Schaniel C, et al. A stem cell molecular signature. Science. 2002;298(5593):601-604. 27. Cabezas-Wallscheid N, Buettner F, Sommerkamp P, et al. Vitamin A-Retinoic Acid Signaling Regulates Hematopoietic Stem Cell Dormancy. Cell. 2017;169(5):807823 e819. 28. Wilson A, Murphy MJ, Oskarsson T, et al. cMyc controls the balance between hematopoietic stem cell self-renewal and differentiation. Genes Dev. 2004;18(22): 2747-2763. 29. Herranz D, Ambesi-Impiombato A, Palomero T, et al. A NOTCH1-driven MYC enhancer promotes T cell development, transformation and acute lymphoblastic leukemia. Nat Med. 2014;20(10):1130-1137. 30. Bahr C, von Paleske L, Uslu VV, et al. A Myc enhancer cluster regulates normal and leukaemic haematopoietic stem cell hierarchies. Nature. 2018;553(7689):515-520. 31. Zhang Q, Zeng L, Shen C, et al. Structural Mechanism of Transcriptional Regulator NSD3 Recognition by the ET Domain of BRD4. Structure. 2016;24(7):1201-1208. 32. Sarai N, Nimura K, Tamura T, et al. WHSC1 links transcription elongation to HIRA-

haematologica | 2018; 103(7)


Setd2 regulates hematopoietic stem cells

mediated histone H3.3 deposition. EMBO J. 2013;32(17):2392-2406. 33. Shen C, Ipsaro JJ, Shi J, et al. NSD3-Short Is an Adaptor Protein that Couples BRD4 to the CHD8 Chromatin Remodeler. Mol Cell. 2015;60(6):847-859. 34. Rahman S, Sowa ME, Ottinger M, et al. The Brd4 extraterminal domain confers transcription activation independent of pTEFb by recruiting multiple proteins, including NSD3. Mol Cell Biol. 2011;31(13):26412652. 35. Jang MK, Mochizuki K, Zhou M, et al. The bromodomain protein Brd4 is a positive regulatory component of P-TEFb and stimulates RNA polymerase II-dependent transcription. Mol Cell. 2005;19(4):523-534.

haematologica | 2018; 103(7)

36. Yang Z, Yik JH, Chen R, et al. Recruitment of P-TEFb for stimulation of transcriptional elongation by the bromodomain protein Brd4. Mol Cell. 2005;19(4):535-545. 37. 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. 38. Haferlach T, Nagata Y, Grossmann V, et al. Landscape of genetic lesions in 944 patients with myelodysplastic syndromes. Leukemia. 2014;28(2):241-247. 39. Della Porta MG, Malcovati L. Myelodysplastic syndromes with bone marrow fibrosis. Haematologica. 2011;96(2): 180-183.

40. 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. 41. Busch K, Klapproth K, Barile M, et al. Fundamental properties of unperturbed haematopoiesis from stem cells in vivo. Nature. 2015;518(7540):542-546. 42. Bu J, Chen A, Yan X, et al. SETD2-mediated crosstalk between H3K36me3 and H3K79me2 in MLL-rearranged leukemia. Leukemia. 2018;32(4):890-899. 43. Mar BG, Chu SH, Kahn JD, et al. SETD2 alterations impair DNA damage recognition and lead to resistance to chemotherapy in leukemia. Blood. 2017;130(24):2631-2641.

1123


ARTICLE

Red Cell Biology & its Disorders

Ferrata Storti Foundation

Endothelin type A receptors mediate pain in a mouse model of sickle cell disease

Brianna Marie Lutz,1,2 Shaogen Wu,1 Xiyao Gu,1 Fidelis E. Atianjoh,1,3 Zhen Li,1 Brandon M. Fox,4 David M. Pollock4 and Yuan-Xiang Tao1,2,5

Department of Anesthesiology, New Jersey Medical School, Rutgers, The State University of New Jersey, Newark, NJ, USA; 2Rutgers Graduate School of Biomedical Sciences, New Jersey Medical School, The State University of New Jersey, Newark, NJ, USA; 3Intensive Care Unit, MedStar Southern Maryland Hospital Center, Clinton, MD, USA; 4Cardio-Renal Physiology and Medicine, Department of Medicine, University of Alabama at Birmingham, AL, USA and 5Neuroscience Research Institute, Zhengzhou University Academy of Medical Sciences, Henan, China 1

Haematologica 2018 Volume 103(7):1124-1135

ABSTRACT

S

Correspondence:

ickle cell disease is associated with acute painful episodes and chronic intractable pain. Endothelin-1, a known pain inducer, is elevated in the blood plasma of both sickle cell patients and mouse models of sickle cell disease. We show here that the levels of endothelin-1 and its endothelin type A receptor are increased in the dorsal root ganglia of a mouse model of sickle cell disease. Pharmacologic inhibition or neuron-specific knockdown of endothelin type A receptors in primary sensory neurons of dorsal root ganglia alleviated basal and post-hypoxia evoked pain hypersensitivities in sickle cell mice. Mechanistically, endothelin type A receptors contribute to sickle cell disease-associated pain likely through the activation of NF-κB-induced Nav1.8 channel upregulation in primary sensory neurons of sickle cell mice. Our findings suggest that endothelin type A receptor is a potential target for the management of sickle cell disease-associated pain, although this expectation needs to be further verified in clinical settings.

yt211@njms.rutgers.edu

Received: December 21, 2017. Accepted: March 13, 2018. Pre-published: March 15, 2018. doi:10.3324/haematol.2017.187013 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/103/7/1124 ©2018 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.

1124

Introduction Sickle cell disease (SCD) results from an amino acid substitution in the β globin chain of the oxygen carrying molecule, hemoglobin.1 Approximately 100,000 Americans currently suffer from SCD, and over 80,000 SCD related hospitalizations occur annually resulting in over 450 million dollars in healthcare costs.2 Much of these healthcare costs and hospitalizations stem from acute vaso-occlusive episodes in which patients often report extreme pain.3 Pain is a hallmark of SCD and correlates with morbidity and disease severity.4 In addition to acute painful episodes, some SCD patients report persistent/chronic pain.5 Treatment for SCDassociated pain includes pain medications such as opioids, but pain relief may remain inadequate for many SCD patients. Additionally, repeated and/or prolonged administration of these drugs has potential side effects.6 Understanding the causes of pain in SCD may bring forward new and more efficient therapeutic strategies for the management of SCD pain. Endothelin-1 (ET-1) is a 21 amino acid peptide released from endothelial cells, immune cells, and neurons.7 ET-1 binds to endothelin type A (ETA) and endothelin type B (ETB) G-protein coupled receptors.7 Classically, ET-1 acts as a vaso-constrictor, but recent studies have shown that ET-1 can also induce pain in humans and rodents.8,9 In the dorsal root ganglia (DRG), ETA receptors are expressed predominantly in the primary sensory neurons, while ETB receptors are detected in glial cells.10 Blocking ETA receptors attenuates ET-1-induced nerve fiber activation and pain behavior in rodents.7,11,12 ET-1-induced activation of the ETA receptor has been found to increase intracellular calcium release, potentiate transient receptor potential vanilloid 1 current, and alter the functioning of tetrodotoxin-resistant (TTX-R) sodium channels in DRG neurons,8 but detailed mechanisms of how ET-1 induces pain are not fully understood. haematologica | 2018; 103(7)


Mouse model of sickle cell disease

In SCD patients, blood plasma ET-1 levels are elevated at a basal level and increase further during an acute vasoocclusive crisis.13,14 Elevated ET-1 and its subsequent activation of ETA receptors have been demonstrated to contribute to SCD pathology, including renal injury15 and pulmonary hypertension.16 The level of ET-1 mRNA is also increased in the DRG of SCD mice.17 However, whether and how this elevated ET-1 in SCD DRG contributes to SCD-associated pain remains elusive. Given that ETA receptor antagonists have been used in phase II/III clinical trials for cancer treatment18,19 and SCD-related pulmonary hypertension,16 identifying the role of ET1 and its ETA receptors in SCD-associated pain may provide a new avenue for treatment of SCD-associated pain in patients. Humanized mouse models of SCD display persistent pain hypersensitivity that can worsen with exposure to hypoxia.17,20,21 In the present study, using the humanized Townes (HbSS) and Berkeley (BerkSS) mouse models of SCD, we first examined whether local pharmacologic inhibition or conditional knockout of ETA receptors in DRG neurons affected pain hypersensitivity in SCD mice. We then examined whether SCD mice expressed elevated levels of ET-1 and ETA receptors in DRG neurons. Finally, we unveiled a possible molecular mechanism by which DRG ETA receptors mediate SCD pain through the activation of NF-ÎşB and subsequent increase of Nav1.8 expression in the DRG.

Methods Animals All procedures used in this study have been approved by the Rutgers New Jersey Medical School Animal Care and Use Committee and adhere to the ethical guidelines of the National Institutes of Health and the International Association for the Study of Pain. All experiments were designed to minimize animal suffering and the number of animals used, and were conducted in a blind manner. Detailed animal information can be found in Online Supplementary Methods.

Behavioral analysis Mechanical testing, thermal testing, and cold testing were conducted on the same day with at least 1 hour rest periods between tests as described.22 The conditioned place-preference (CPP) test was carried out as described22 and in the Online Supplementary Methods.

Hypoxia/Reoxygenation Hypoxia/reoxygenation exposure for the induction of vasoocclusion was conducted in the same manner as previously described.20 The detailed procedures can be found in the Online Supplementary Methods.

Genetic Knockdown and Bone Marrow Transplantation Chimeric mouse generation was completed using a previously described protocol, which showed the development of the sickle cell phenotype in BerkSS bone marrow-recipient mice beginning at 2 months after transplantation.23 The detailed procedures can be found in the Online Supplementary Methods. Isoelectric focusing, cell culture, immunofluorescence, Western blotting, qRT-PCR, ChIP assay, luciferase assay, electrophysiological recording, and statistical analysis methods can be found in the Online Supplementary Methods. haematologica | 2018; 103(7)

Results Pain hypersensitivity in SCD mouse models We first characterized basal pain behaviors in our colony of Townes humanized SCD mice. HbSS and HbAA mice (as a control) aged 4-6 months were used. Like male BerkSS mice (Online Supplementary Figure S2A, B), male HbSS mice displayed significant bilateral increases in basal paw withdrawal frequencies (PWF) in response to 0.16 g and 0.4 g von Frey filaments (Figure 1 A, B), and reductions in paw withdrawal latencies (PWL) to both thermal and cold stimuli when compared to their HbAA littermates (Figure 1 C, D). Similar pain hypersensitivities were seen in female HbSS mice (Online Supplementary Figure S3 A-D). We then examined whether pain hypersensitivities stemmed from an increase in DRG neuronal excitability in HbSS mice. Compared to HbAA mice, the medium and small DRG neurons of HbSS mice showed increases of 9.85 mV and 8.24 mV, respectively, in the resting membrane potentials (Figure 2A) and decreases of 36% and 52%, respectively, in the current thresholds for action potential generation (Figure 2B). The number of action potentials evoked by a stimulation of ≼ 200 pA in medium and small neurons was significantly higher in HbSS mice (Figure 2C, D), although both mice displayed similar membrane input resistances and other action potential parameters, (Online Supplementary Table S1). In addition, a larger percentage of small and medium HbSS DRG neurons exhibited spontaneous activity compared to HbAA DRG neurons (Figure 2 E, F). The frequency of spontaneous action potential generation was also higher in HbSS small and medium DRG neurons compared to HbAA neurons (Figure 2 G). These indicators of heightened excitability were not seen in large DRG neurons (Figure 2 A-G).

Effect of DRG ETA receptor inhibition on SCD pain

To address the role of DRG ETA receptors in SCD-associated pain, we subcutaneously (s.c.) injected ABT-627, a selective and specific ETA receptor antagonist,24 into the plantar side of the left hindpaw. Single administration of 5 nmol ABT-627, 2 hours before behavioral testing under normoxic conditions, led to the attenuation of ipsilateral mechanical (Figure 3A, B; Online Supplementary Figure S3A, B), thermal (Figure 3C; Online Supplementary Figure S3C) and cold (Figure 3D; Online Supplementary Figure S3D) pain hypersensitivities in the male and female HbSS mice. This effect is dose-dependent (Online Supplementary Figure S4). ABT-627 did not alter the basal mechanical, thermal and cold responses of the male and female HbAA mice (Figure 3 A-D; Online Supplementary Figure S3 A-D). Similar ABT627 efficacy was evident in male BerkSS mice (Online Supplementary Figure S2 A-D). Exposure to hypoxia/reoxygenation in SCD mice has been implicated as a method of inducing vaso-occlusion and mimicking acute painful episodes in SCD patients.20 We further performed behavior analyses following hypoxia/reoxygenation exposure in SCD mice. Male HbSS mice exhibited similar responses to the 0.4 g von Frey filament (Figure 3B), thermal (Figure 3C), and cold (Figure 3D) stimuli before and after hypoxia/reoxygenation. However, after hypoxia/reoxygenation, bilateral PWF to the 0.16 g von Frey filament increased by 20% in male HbSS mice compared to basal values (Figure 3A). ABT-627 administered to the unilateral hind paw immediately after hypox1125


B.M. Lutz et al.

ia exposure, led to the attenuation of ipsilateral mechanical, thermal, and cold pain hypersensitivities in male HbSS mice (Figure 3 A-D), although it did not affect basal or post-hypoxia ipsilateral responses in HbAA mice (Figure 3A-D) or contralateral responses in HbAA and HbSS mice (Online Supplementary Figure S5). Similar ABT-627 efficacy was seen in male BerkSS mice after hypoxia/reoxygenation exposure (Online Supplementary Figure S6 A-D). There were no sex-based differences in pain hypersensitivities and ABT-627 efficacy after hypoxia/reoxygenation exposure (Figure 3 A-D, Online Supplementary Figure S7). Spontaneous ongoing pain in HbSS mice using a CPP paradigm was also examined. HbSS mice, but not HbAA mice, given vehicle once a day for 4 days prior to and during CPP testing (Online Supplementary Figure S1), spent significantly more time in the lidocaine-paired chamber after conditioning compared to their pre-test time (Figure 3 E, F). When given ABT-627 systemically (i.p) once a day for four days prior to and during testing (Online Supplementary Figure S1), neither HbSS nor HbAA mice displayed a significant lidocaine-paired chamber preference (Figure 3 E, F).

Effect of knockdown of DRG ETA receptors on SCD pain

The effects of ABT-627 hindpaw administration described above suggest the role of peripheral ETA receptors in SCD-associated pain; however, ABT-627 lacks celltype specificity and may also have off-target effects. As a complementary approach, we generated a DRG neuronspecific ETA receptor knockout mouse by crossing an ETAflox/flox mouse with an Advillincre/+ mouse in which Crerecombinase is expressed only in Advillin-positive neurons exclusive to the DRG and trigeminal ganglia. DRG-

A

C

specific knockdown of ETA receptors was validated using Western blotting and immunostaining (Online Supplementary Figure S8 A, B). Two months after bone marrow transplantation, donor hemoglobin expression in the recipient mice was confirmed using isoelectric focusing (Online Supplementary Figure S8 C). Mice receiving BerkSS bone marrow only expressed human βs globin (βSS). BerkAA bone marrow recipients expressed the normal human β globin (βAA; Online Supplementary Figure S8 C) and were used as a control. Mouse hemoglobin (Hb) was not found in the blood of recipient mice (Online Supplementary Figure S8 C), confirming the knockdown of mouse globin. Bilateral evoked pain hypersensitivities as evidenced by the increase in PWF to mechanical stimuli and the decrease in PWL to thermal and cold stimuli developed in the ETAflox/flox mice expressing βSS (Figure 4 A-D) 2 months after bone marrow transplantation. These pain hypersensitivities had a tendency of aggravation after hypoxia/reoxygenation (Figure 4 A-D). ETAcre/flox mice expressing βSS failed to develop these pain hypersensitivities after bone marrow transplantation in both normoxic and post-hypoxia/reoxygenation conditions (Figure 4 AD). ETAflox/flox mice and ETAcre/flox mice transplanted with normal human βAA maintained normal basal responses after bone marrow transplantation under normoxia and hypoxia/reoxygenation conditions (Figure 4 A-D).

Increased ET-1 and ETA receptor expression in SCD DRG Our behavioral observations indicate the significance of DRG ETA receptors in pain hypersensitivity in HbSS mice. We next examined the expression of the ETA receptor and

B

D

Figure 1. Male Townes HbSS mice display evoked pain hypersensitivity. Paw withdrawal frequencies (PWF) to a 0.16 g low force (A) and a 0.4 g medium force (B) von Frey filaments and paw withdrawal latencies (PWL) to thermal (C) and cold (D) stimuli on both left and right hind paws of 4 month old male HbAA and HbSS mice. n = 10 mice/genotype, **P< 0.01 vs. the corresponding HbAA group.

1126

haematologica | 2018; 103(7)


Mouse model of sickle cell disease

its ligand, ET-1, in the DRG and peripheral tissue of HbSS mice. Edn1 mRNA (encoding ET1) and ET-1 protein were elevated in HbSS mouse DRG (Figure 5 A, B). Ednra mRNA (encoding ETA receptor) expression in the DRG did not differ between HbAA and HbSS mice, but the level of ETA receptor protein significantly increased in HbSS DRG (Figure 5 A, B). In accordance with previous studies,10,25 ETA receptors were expressed exclusively in the neurons of DRG. Approximately 43% of ETA receptorlabeled neurons are positive for CGRP, 36% for IB4, and only 12% for NF200 (Figure 5 C). Interestingly, the immunostaining for pre-pro ET-1, a precursor of ET-1 peptide, was also restricted to DRG neurons (Online Supplementary Figure S9). About 36% of pre-pro ET-1labeled neurons are positive for CGRP, 56% for IB4, and 7% for NF200 (Figure 5C). The percentage of neurons expressing ETA receptors or pre-pro ET-1 within the DRG of HbSS mice was 45% and 32%, respectively, higher than those in the HbAA DRG (Figure 5 D, E). Unexpectedly, there was no significant difference in hindpaw expression of ET-1 or ETA receptors between HbAA and HbSS mice (Figure 5F).

Effect of ETA receptor inhibition on Nav1.8 expression and channel activity in SCD DRG Previous studies have shown that ET-1 can alter the activity of TTX-R sodium channels in rat DRG neurons.26 Nav1.8, a TTX-R sodium channel in DRG, participates in

A

chronic inflammatory pain genesis.27 We first determined if Nav1.8 expression differed between HbAA and HbSS DRG. The levels of Scn10a mRNA (encoding Nav1.8) and Scn11a mRNA (encoding Nav1.9, another TTX-R sodium channel) were elevated in HbSS DRG (Figure 6A). The amount of Nav1.8 protein and the percentage of Nav1.8positive neurons were also increased in HbSS DRG (Fig. 6 B, C). Four days of hindpaw ABT-627 administration abolished the increase in Nav1.8 protein in HbSS DRG (Figure 6 B), indicating that increased Nav1.8 requires ETA receptor activation in SCD (HbSS) DRG. Furthermore, we analyzed Nav1.8 activity in HbSS DRG neurons. Whole cell voltage-clamp recordings were performed in freshly disassociated small DRG neurons since Nav1.8 is expressed predominantly in small DRG neurons.28 Neurons were held at -60 mV, given a depolarizing step at 50 ms from -55 mV to +40 mV with a 5 mV increment (Figure 7A). Under this protocol with the presence of 500 nM TTX, Nav1.9 current was inhibited to keep the Nav1.8 current integrated.29 Nav1.8 current density from the HbSS DRG neurons significantly increased compared to HbAA DRG neurons (Figure 7A, B). Bilateral subcutaneous ABT-627 administered daily over 4 days prior to DRG collection produced a larger reduction in Nav1.8 current density in HbSS DRG compared to HbAA DRG (Figure 7A, B). When tested at -15 mV, ABT-627 reduced Nav1.8 current by approximately 300 pA in HbSS DRG neurons, with no change in the HbAA DRG neurons

C

B

D

E

F

G

Figure 2. Townes HbSS DRG neurons display increased excitability and spontaneous activity. (A and B) Resting membrane potential (RMP, A) and current threshold for pulse (Ithreshold, B). n = 24 large, 22 medium, and 21 small neurons from the HbAA group (7 mice). n = 25 large, 23 medium, and 25 small neurons from the HbSS group (6 mice). **P<0.01 vs. the corresponding HbAA group. (C) Representative trace of evoked action potentials in small DRG neurons. (D) Numbers of evoked action potentials (APs) in large, medium, and small DRG neurons from the HbAA and HbSS groups after application of different currents as indicated. Numbers of the recorded neurons are the same as in A. *P<0.05, **P<0.01 vs. the same stimulation intensity in the HbAA group. (E) Representative trace of spontaneous activity in small DRG neurons from the HbAA and HbSS groups. (F and G) Percentage of DRG neurons that had spontaneous activity (SA, F) and frequency of spontaneous activity (G). n = 23 large, 22 medium, and 25 small DRG neurons for the HbAA group (7 mice). n = 26 large, 25 medium, and 22 small DRG neurons from the HbSS group (6 mice). *P<0.05, **P<0.01 vs. the corresponding HbAA group.

haematologica | 2018; 103(7)

1127


B.M. Lutz et al.

(Figure 7C). These data further indicate that the increase in Nav1.8 current from HbSS DRG neurons is ETA receptordependent. We also examined the gating properties of Nav1.8 in HbSS DRG neurons. HbSS DRG neurons showed a hyperpolarizing potential shift compared to HbAA DRG neurons under the conditions of either activation or inactivation of Nav1.8 (Figure 7 D, E). The halfmaximal activation voltage (V1/2) was -20.35 ± 1.16 mV in the HbAA mice and -24.79 ± 0.93 mV in the HbSS mice (Figure 7D), whereas the half-maximal inactivation voltage (V1/2) was -37.56 ± 1.32 mV in the HbAA mice and 45.80 ± 1.80 mV in the HbSS mice (Figure 7E). In vivo

A

administration of ABT-627 as described above did not lead to changes in the gating properties of Nav1.8 in HbAA or HbSS mice. HbSS DRG neurons likely display ETA receptor-independent changes in the gating properties of Nav1.8 channels.

NF-κ B-triggered Nav1.8 expression in HbSS DRG neurons How does ETA receptor activation mediate an increase in Nav1.8 expression and activity in SCD (HbSS) DRG neurons? The transcription factor, NF-κB, may be upstream of Nav1.8 upregulation in DRG neurons and is

B

D

C

D

E

F

Figure 3. Hindpaw administration of ABT-627 can alleviate pain hypersensitivity and spontaneous ongoing pain under normoxic and hypoxic conditions in male HbSS mice. (A-D) Effect of hindpaw injection of ABT-627 on paw withdrawal frequencies (PWF) to a 0.16 g low force (A) and a 0.4g medium force (B) and paw withdrawal latencies (PWL) to heat (C) and cold (D) stimuli on the ipsilateral side of HbAA mice and HbSS mice under normoxic and hypoxic conditions. Veh: vehicle. ABT: ABT-627. n=6 mice/genotype/treatment. *P<0.05, **P<0.01 vs. the corresponding HbAA group. ##P<0.01 vs. the HbSS baseline value. $P<0.05 vs. the corresponding vehicle-treated HbSS group. (E and F) Effect of hindpaw injection of ABT-627 (ABT) on ongoing pain as assessed by the conditioned place preference paradigm. n=6-8 mice/group. **P<0.01 vs. the corresponding pre-conditioning (E) or the HbAA plus vehicle (Vel) group (F). ##P<0.01 vs. the corresponding post-conditioning (post) from the HbSS plus vehicle group or the HbSS plus vehicle group (F).

1128

haematologica | 2018; 103(7)


Mouse model of sickle cell disease

activated by ET-1.30,31 To confirm this relationship among ET-1, ETA receptors, and Nav1.8 channels in DRG neurons, we cultured HbAA DRG neurons in the presence of ET-1 peptide. ET-1 stimulation of HbAA DRG neurons for 24 hours resulted in increases in the levels of Nav1.8 protein (Figure 8A) and phosphorylated p65 protein (an indicator of NF-κB activation32), a NF-κB subunit (Figure 8B). Co-incubation with ABT-627 or pyrrolidine dithiocarbamate (PDTC), an NF-κB-specific inhibitor, prevented these increases (Figure 8A and 8B). Furthermore, we examined the expression of phosphorylated p65 in in vivo mice injected intraperitoneally (i.p) with vehicle or ABT-627 once daily for 4 days. Phosphorylated p65 levels in HbSS DRG significantly increased after vehicle injection, but did not significantly change after ABT-627 administration compared to that in the HbAA mice after vehicle injection (Figure 8C). Total p65 protein levels did not significantly differ among treatment groups (Figure 8C). These findings indicate that NF-κB activation and Nav1.8 upregulation in HbSS DRG are ETA receptor-dependent and may be involved in SCD-associated pain. This conclusion is fur-

ther supported by our behavioral observations that single i.p administration of PDTC, a specific inhibitor of NF-κB activation, or A-803467, a specific antagonist of Nav1.8,33 alleviated mechanical allodynia and/or thermal hyperalgesia in HbSS mice, without affecting the basal responses of HbAA mice (Figure 8D; Online Supplementary Figure S10). We then examined whether activated NF-κB is directly linked to increased Nav1.8 expression in HbSS DRG. Using ChIP analyses, a fragment within the Scn10a promoter was amplified from a complex immunoprecipitated with an anti-p65 antibody (Figure 8E), indicating the binding of p65 to the Scn10a gene promoter in the DRG. This binding activity increased in the HbSS DRG compared to HbAA DRG, as shown by a 7-fold increase in band density (Figure 8E). To examine whether NF-κB activation directly regulates Scn10a transcriptional activity, we performed a dual luciferase assay using a mouse neuronal cell line (CAD cells)34 transfected with a luciferase reporter vector containing the Scn10a gene promoter. Since phorbol 12-myristate 13-acetate (PMA) can activate NF-κB through protein kinase C (PKC) activation,35 we applied

A

B

C D

Figure 4. Male HbSS mice with conditional genetic knockdown of DRG ETA receptors fail to develop pain hypersensitivity before and after hypoxia/reoxygenation exposure. Paw withdrawal frequency (PWF) in response to a 0.16 g low mechanical force (A) and a 0.4 g medium mechanical force (B) and paw withdrawal latency (PWL) in response to heat (C) and cold (D) stimuli on both left and right sides of ETAfl/fl mice and ETAcre/fl mice 2 months after BerkSS (SS) or BerkAA (AA) bone marrow transplantation (BMT). n = 6-8 mice/group. **P<0.01 vs. the corresponding baseline value. ##P<0.01 vs. the corresponding BMT + Normoxia value.

haematologica | 2018; 103(7)

1129


B.M. Lutz et al.

PMA into the medium. PMA stimulation increased the activity of the Scn10a gene promoter by 3.5-fold as compared to the vehicle-treated group (Figure 8F). Co-administration of PDTC or bisindolylmaleimide I (BIM), a PKC inhibitor, prevented this increase (Figure 8F). BIM or PDTC alone did not affect basal luciferase activity (Figure 8F). These results indicate that the increased Scn10a gene activity is a specific response to PKC/NF-ÎşB activation. Single-cell RT-PCR analysis revealed that 40% of individual small DRG neurons co-expressed Scn10a mRNA, Ednra mRNA, and Rela mRNA (encoding p65) (Figure 8G). Taken together, our findings suggest the participation of NF-ÎşB in ETA receptor-dependent Nav1.8 upregulation in HbSS DRG neurons (Online Supplementary Figure S11).

Pain is a major symptom reported by SCD patients. Currently, painkillers such as opioids are prescribed for SCD pain management but repeated and/or prolonged administration of these drugs has limited analgesic efficacy and the potential to produce side effects.6 Identifying novel and specific targets for pain management is essential for improving SCD patient care. We showed here that the expression of ET-1 and its ETA receptor increases in the DRG of SCD mice. Sensory neuron-specific knockdown or local inhibition of DRG ETA receptors alleviated basal and post-hypoxia mechanical allodynia and thermal/cold hyperalgesia in SCD mice. Although the exact mechanism

B

A

D

C

E

Discussion

F

Figure 5. ET-1 and ETA receptor expression increases in the DRG of Townes HbSS mice. (A) Expression of Edn1 mRNA (encoding ET-1) and Ednra mRNA (encoding ETA receptor) in the L3-L4 DRG. n=3 mice/genotype. **P<0.01 vs. the corresponding HbAA mice. (B) Expression of ET-1 and ETA receptor proteins in the L3-L4 DRG. Left: representative Western blots. Right: a summary of densitometric analysis. n=3 mice/genotype. *P<0.05 and **P<0.01 vs. the corresponding HbAA mice. (C) Double immunostaining shows the co-localization of ETA receptors (ETAR) or pre-pro ET-1 (ET-1) with CGRP, IB4, and NF200 in the L3-L4 DRG of HbAA mice (for ETAR) or HbSS mice (for ET-1). Scale bar: 50 mm. (D and E) Number of ETA receptors (ETAR)-positive neurons or pre-pro ET-1 (ET-1)-positive neurons in the L4 DRG from HbAA mice and HbSS mice. D: representative immunostaining images. E: a summary of statistical analysis of the number of positive neurons. n=3 mice/genotype. **P<0.01 vs. the HbAA mice. Scale bar: 50 mm. (F) Expression of ET-1 and ETA receptor proteins in hind paw from HbAA mice and HbSS mice. Left: representative Western blots. Right: a summary of densitometric analysis. n=3 mice/genotype.

1130

haematologica | 2018; 103(7)


Mouse model of sickle cell disease

underlying these effects requires further validation in SCD mouse models, our study suggests that ETA receptors are key players in SCD-associated pain. The present study used two humanized mouse models of SCD, the Berkeley and Townes, to analyze SCD-associated pain. Although pain-like behaviors in SCD mouse models may not exactly mimic pain in SCD patients due to several confounding environmental and emotional variables, analyzing pain hypersensitivity in SCD murine models is a productive tool for identifying the underlying mechanisms of SCD-associated pain. These mice displayed significant mechanical allodynia, thermal/cold hyperalgesia, and spontaneous ongoing pain consistent with previous reports.36 Using age-matched male and female HbSS and HbAA littermates, we found no significant sex-based differences in basal or post-hypoxia evoked pain hypersensitivity or ABT-627 efficacy. This aligns with clinical reports that found no significant differences in self-reported pain experiences between maleand female SCD patients.37,38 Interestingly, Townes HbSS mice did not display a similar degree of pain exacerbation following hypoxia/reoxygenation compared to the BerkSS mice. This may be related to the presence of a human β-globin locus control region (LCR) that leads to hemoglobin switching similar to humans in the Townes HbSS mice, although both models express human globins. In the Townes model, relative g-globin expression (g/g+βS) at birth is between 30% to 50%, then β-globin expression dominates at 1 month of age.39 In contrast, the Berkeley model exhibits a complete g-globin to β-globin switch in utero leading to greater prenatal death and low birth rates. Given that higher g-globin expression is associated with diminished SCD symptoms,40 the early expression of βsglobin in BerkSS mice may cause more severe symptoms of SCD, which could explain their significant pain exacerbation following hypoxia/reoxygenation. An increase in ETA receptor expression may be attributed to the elevated ET-1 in the DRG neurons of HbSS

A

C

mice. The levels of ET-1 mRNA, ET-1 protein, and ETA receptor protein increased in HbSS DRG. Consistent with previous studies,10,25 ET-1 and ETA receptors are predominantly expressed in DRG neurons, although ET-1 expression in immune cells found in HbSS DRG cannot be ruled out.41 Elevated ET-1 could trigger increased ETA receptor expression in DRG possibly through increased receptor cycling to the membrane and/or enhanced transcription or translation.42,43 Since we found no significant changes in ETA receptor mRNA in HbSS DRG, it would be expected that, under constantly elevated ET-1 conditions, ET-1/ETA receptor cycling coupled with ET-1-triggered activation of intracellular signals enhances ETA receptor translation through unknown mechanisms in HbSS DRG. The mechanism of ETA receptor upregulation in HbSS DRG remains to be further investigated. ETA receptors are required for increased Nav1.8 expression in HbSS DRG, since ETA receptor inhibition completely blocked increases in Nav1.8 protein and current in HbSS DRG. However, whether the participation of DRG ETA receptors in the pain hypersensitivity of HbSS mice is mediated by DRG Nav1.8 upregulation in HbSS mice requires further investigation as the present data showed only an association between activation of the ETA receptors and upregulation of Nav1.8 in the DRG of HbSS mice. Interestingly, ETA receptor inhibition did not alleviate the hyperpolarizing shift in the gating properties of Nav1.8 channels in HbSS DRG. Previous studies reported that phosphorylation of Nav1.8 channels by protein kinase A (PKA), but not PKC, could affect the channel’s gating properties.44,45 The altered gating of Nav1.8 channels in the HbSS DRG is likely the result of activation of PKA by an ETA receptor-independent signaling mechanism. It is worth noting that the leftward shift of 4.44 mV in the HbSS DRG would only result in a modest increase in sodium current.26 Given that ABT-627 fully attenuated the increase in Nav1.8 current in the HbSS DRG, an ETA -receptor independent mechanism of Nav1.8 channel

B

D

Figure 6. Nav1.8 expression is elevated in Townes HbSS DRG and is ETA receptor-dependent. (A) Expression of Scn10a mRNA (encoding Nav1.8) and Scn11a mRNA (encoding Nav1.9) in the L3-L4 DRG from HbAA mice and HbSS mice. n=3 mice/genotype. *P<0.05 vs. the corresponding HbAA mice. (B) Effect of hindpaw injection of ABT-627 or vehicle on expression of Nav1.8 protein in the L3-L4 DRG from HbAA mice and HbSS mice. Left: representative Western blots. Right: a summary of densitometric analysis. n=3 mice/group. *P<0.05 vs. the vehicle-treated HbAA mice. (C) Number of Nav1.8-labeled neurons in the L4 DRG from HbAA mice and HbSS mice. Left: representative immunostaining images. Right: a summary of statistical analysis of the number of Nav1.8-labeled neurons. Scale bar: 50 mm. n = 3 mice/genotype. **P<0.01 vs. the HbAA mice.

haematologica | 2018; 103(7)

1131


B.M. Lutz et al.

phosphorylation may not be a significant contributor to Nav1.8-mediated neuronal excitability and behavioral pain hypersensitivity. Additionally, we observed that the resting membrane potential of HbSS DRG neurons was elevated. This suggests a possible decrease in potassium channel expression or function. ET-1 has been shown to depress delayed rectifier potassium currents in rat DRG neurons.46 This depression of potassium currents may be involved in increased excitability of SCD neurons and the pain hypersensitivity of SCD mice found in this study. Thus, it is of interest to further examine whether the expression of other ion channels including delayed rectifi-

er potassium channels differs between HbAA and HbSS DRG neurons and whether these differences are ETA receptor-dependent. NF-κB may be a key modulator of ETA receptor-dependent Nav1.8 upregulation in HbSS DRG. NF-κB is expressed in small and medium neurons of the DRG.47 Under inflammatory conditions, PKCε activation has been linked to Nav1.8 upregulation in DRG neurons48 and NF-κB activation.49 A recent study also reported that chemokine ligand 2 can induce NF-κB activation via PKC and lead to Nav1.8 upregulation in DRG neurons.30 Since ETA receptor activation leads to the downstream activa-

B

A

C

D

E

Figure 7. Nav1.8 channel activity increases in Townes HbSS DRG neurons. (A) Representative traces of Nav1.8 in small DRG neurons from the HbAA mice, HbSS mice, and these mice injected with ABT-627. (B) I-V curves in small DRG neurons from HbAA mice (24 neurons), HbSS mice (21 neurons), HbAA mice given ABT-627 (21 neurons), and HbSS mice given ABT-627 (23 neurons). *P<0.05, **P<0.01 vs. the corresponding HbAA group. (C) When tested at -15mV, HbSS mice exhibit an increase in Nav1.8 current in small DRG neurons compared to HbAA mice. This increase is abolished in HbSS mice given ABT-627. n=25 neurons for the HbAA group, 22 for the HbSS group, 21 for the HbAA with ABT-627 group, and 23 for the HbSS with ABT-627 group. *P<0.05 vs. the HbAA group. #P<0.05 vs. the HbSS group. (D) Activation curve for Nav1.8. Curves are fitted by Boltzmann function. E0.5 = -20.35 ± 1.16 mV for the HbAA group (n = 21 neurons), -24.79 ± 0.93 mV for the HbSS group (n=24 neurons), -21.34 ± 0.65 mV for the HbAA with ABT-627 group (n=21 neurons), and -24.37 ± 0.82 mV for the HbSS with ABT-627 group (n=23 neurons). (E) Inactivation curve for Nav1.8. Curves are fitted by Boltzmann function. E0.5 = -37.56 ± 1.32mV for the HbAA group (n=27 neurons), -45.80 ± 1.80 mV for the HbSS group (n=24 neurons), -39.79 ± 1.11 mV for the HbAA with ABT-627 group (n=19 neurons), and -41.05 ± 1.86 mV for the HbSS with ABT-627 group (n=24 neurons).

1132

haematologica | 2018; 103(7)


Mouse model of sickle cell disease

tion of PKCε and PKCd,50 it is likely that ETA receptors contribute to increased Nav1.8 expression through NF-κB activation in HbSS DRG. NF-κB activation was indeed enhanced in ET-1-stimulated cultured DRG neurons and in vivo HbSS DRG, and this enhanced activation was ETA receptor-dependent. NF-κB binding to the Scn10a

A

(Nav1.8) promoter region increased in HbSS DRG. Additionally, PKC-induced NF-κB activation led to increased Scn10a promoter activity in mouse neuronal cells. These findings suggest that the Nav1.8 upregulation found in HbSS DRG is the result of ET-1/ETA receptorinduced NF-κB activation. However, the involvement of

B

D

C

E

F

G

Figure 8. ET-1 induced activation of ETA receptors leads to the upregulation of Nav1.8 and activation of NF-κB in HbSS DRG neurons. (A) Cultured HbAA DRG neurons exposed to 400 nmol of ET-1 peptide for 24 hr displayed an increase in Nav1.8 protein. Co-administration of ABT-627 (1 mmol) prevented this increase. Left: Representative western blots. Right: Summary of densitometric analysis. n=3 biological repeats. **P<0.01 vs. the vehicle (Veh) group. (B) Cultured HbAA DRG neurons exposed to 400 nmol of ET-1 peptide for 90 minutes displayed an increase in phosphorylated p65 (p-p65). Co-administration of ABT-627 (1 mmol) 10 minutes prior to ET-1 stimulation prevented the increase in phosphorylated p65 protein. The amount of total p65 protein did not change among the treated groups. Left: Representative western blots. Right: Summary of densitometric analysis n=3 biological repeats. **P<0.01 vs. the vehicle (Veh) group. (C) The levels of phosphorylated p65 (p-p65), increase in the L3-L4 DRG of the vehicle-treated HbSS mice. This increase is not seen in the ABT-627-treated HbSS mice. Left: representative Western blots. Right: a summary of densitometric analysis. n=3 mice/group. *P<0.01 vs. the vehicle-treated HbAA group, ##P<0.01 vs. the vehicle-treated HbSS group. (D) 45 mins after intraperitoneal injection of the NF-κB -specific inhibitor, PDTC, increased response to a low force (0.16 g) von Frey filament is alleviated in the HbSS mice. n = 6 mice/group. *P<0.05 vs. the baseline in HbAA mice. #P<0.05 vs. the baseline in HbSS mice. (E) Scn10a promoter fragments immunoprecipitated by rabbit anti-p65 in the L3-L5 DRG from HbSS (SS) and HbAA (AA) mice. Top: representative gel image. Bottom: a summary of densitometric analysis. Input: total purified fragments. M: ladder marker. n=3 mice/genotype. **P<0.01 vs. the HbAA mice. (F) Scn10a promoter activity in mouse CAD cells transfected with a luciferase reporter vector containing a Scn10a promoter and treated with drugs as indicated. PMA: the PKC activator. BIM: the PKC-specific inhibitor. PDTC: the NFκB-specific inhibitor. n=3 repeats/treatment. One-way ANOVA on Ranks, **P<0.01 vs. the vehicle group. #P<0.05 vs. the PMA-treated group. (G) Single cell RT-PCR analysis shows the co-expression of Scn10a mRNA, Ednra mRNA (encoding ET-1), and Rela mRNA (encoding p65) in individual small DRG neurons from HbAA mice. Gapdh mRNA is used as a positive control. M: ladder marker. n=3 repeats.

haematologica | 2018; 103(7)

1133


B.M. Lutz et al.

other transcription factors or molecules in ET-1/ETA receptor-mediated Nav1.8 upregulation in HbSS DRG could not be excluded. In conclusion, this study provides evidence for one possible mechanism by which ET-1/ETA receptors contribute to SCD-associated pain likely through the NF-ÎşB-triggered upregulation of Nav1.8 in primary sensory neurons. We identified the ability of ABT-627, an ETA receptorspecific antagonist,24 to alleviate basal and post-hypoxia evoked pain hypersensitivity in SCD mice. At the dosages used, ABT-627 produced pain relief that persisted for the entire testing period (at least 7 hours) with no noticeable side effects. Given that ABT-627 and similar

References 1. Bunn HF. Pathogenesis and Treatment of Sickle Cell Disease. N Engl J Med. 1997;337(11):762-769. 2. Brozovic M, Davies S. Management of sickle cell disease. Postgrad Med J. 1987;63(742):605-609. 3. Manwani D, Frenette PS. Vaso-occlusion in sickle cell disease: pathophysiology and novel targeted therapies. Hematology Am Soc Hematol Educ Program. 2013;2013: 362-369. 4. Smith WR, Penberthy LT, Bovbjerg VE, et al. Daily assessment of pain in adults with sickle cell disease. Ann Intern Med. 2008;148(2):94-101. 5. Brandow AM, Farley RA, Panepinto JA. Neuropathic pain in patients with sickle cell disease. Pediatr Blood Cancer. 2014;61(3):512-517. 6. Gupta M, Msambichaka L, Ballas SK, Gupta K. Morphine for the treatment of pain in sickle cell disease. ScientificWorldJournal. 2015;2015:540154. 7. Chichorro JG, Fiuza CR, Bressan E, Claudino RF, Leite DF, Rae GA. Endothelins as pronociceptive mediators of the rat trigeminal system: Role of ETA and ETB receptors. Brain Res. 2010;1345:73-83. 8. Barr TP, Kam S, Khodorova A, Montmayeur JP, Strichartz GR. New perspectives on the endothelin axis in pain. Pharmacol Res. 2011;63(6):532-540. 9. Ergul S, Brunson CY, Hutchinson J, et al. Vasoactive factors in sickle cell disease: In vitro evidence for endothelin-1-mediated vasoconstriction. Am J Hematol. 2004; 76(3):245-251. 10. Pomonis JD, Rogers SD, Peters CM, Ghilardi JR, Mantyh PW. Expression and localization of endothelin receptors: implications for the involvement of peripheral glia in nociception. J Neurosci. 2001;21(3): 999-1006. 11. Motta EM, Chichorro JG, Rae GA. Role of ETA and ETB endothelin receptors on endothelin-1-induced potentiation of nociceptive and thermal hyperalgesic responses evoked by capsaicin in rats. Neurosci Lett. 2009;457(3):146-150. 12. Gokin AP, Fareed MU, Pan HL, Hans G, Strichartz GR, Davar G. Local injection of endothelin-1 produces pain-like behavior and excitation of nociceptors in rats. J Neurosci. 2001;21(14):5358-5366. 13. Graido-Gonzalez E, Doherty JC, Bergreen EW, Organ G, Telfer M, McMillen MA. Plasma endothelin-1, cytokine, and prostaglandin E2levels in sickle cell disease

1134

14. 15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

ETA receptor antagonists have been used in phase II/III clinical trials for cancer treatment18,19 and are beneficial for SCD-related pulmonary hypertension,16 targeting DRG ETA receptors may have potential therapeutic value for SCD-associated pain management. Funding This work was supported by NIH grants (R01NS094664, R01NS094224, R01DA033390, and U01HL117684) to YXT and by a NIH research Fellowship (F31NS092310) to BML at Rutgers New Jersey Medical School. We thank Dr. Roger Howell at Rutgers New Jersey Medical School for his assistance with animal radiation.

and acute vaso-occlusive sickle crisis. Blood. 1998;92(7):2551-2555. Rybicki AC, Benjamin LJ. Increased levels of endothelin-1 in plasma of sickle cell anemia patients. Blood. 1998;92(7):2594-2596. Heimlich JB, Speed JS, O'Connor PM, et al. Endothelin-1 contributes to the progression of renal injury in sickle cell disease via reactive oxygen species. Br J Pharmacol. 2016; 173(2):386-395. Minniti CP, Machado RF, Coles WA, Sachdev V, Gladwin MT, Kato GJ. Endothelin receptor antagonists for pulmonary hypertension in adult patients with sickle cell disease. Br J Haematol. 2009;147(5):737-743. Zappia KJ, Garrison SR, Hillery CA, Stucky CL. Cold hypersensitivity increases with age in mice with sickle cell disease. Pain. 2014;155(12):2476-85. Carducci MA, Manola J, Nair SG, et al. Atrasentan in patients with advanced renal cell carcinoma: a phase II trial of the ECOG-ACRIN Cancer Research Group (E6800). Clin Genitourin Cancer. 2015; 13(6):531-539. Quinn DI, Tangen CM, Hussain M, et al. Docetaxel and atrasentan versus docetaxel and placebo for men with advanced castration-resistant prostate cancer (SWOG S0421): a randomised phase 3 trial. Lancet Oncol. 2013;14(9):893-900. Cain DM, Vang D, Simone DA, Hebbel RP, Gupta K. Mouse models for studying pain in sickle disease: effects of strain, age, and acuteness. Br J Haematol. 2012;156(4):535544. Cataldo G, Rajput S, Gupta K, Simone DA. Sensitization of nociceptive spinal neurons contributes to pain in a transgenic model of sickle cell disease. Pain. 2015;156(4):722730. Zhao JY, Liang L, Gu X, et al. DNA methyltransferase DNMT3a contributes to neuropathic pain by repressing Kcna2 in primary afferent neurons. Nat Commun. 2017; 8, 14712. Turhan A, Weiss LA, Mohandas N, Coller BS, Frenette PS. Primary role for adherent leukocytes in sickle cell vascular occlusion: A new paradigm. Proc Natl Acad Sci USA. 2002;99(5):3047-3051. Jarvis MF, Wessale JL, Zhu CZ, et al. ABT627, an endothelin ET(A) receptor-selective antagonist, attenuates tactile allodynia in a diabetic rat model of neuropathic pain. Eur J Pharmacol. 2000;388(1):29-35. Stosser S, Agarwal N, Tappe-Theodor A, Yanagisawa M, Kuner R. Dissecting the functional significance of endothelin A receptors in peripheral nociceptors in vivo

via conditional gene deletion. Pain. 2010; 148(2):206-214. 26. Zhou Z, Davar G, Strichartz G. Endothelin1 (ET-1) selectively enhances the activation gating of slowly inactivating tetrodotoxinresistant sodium currents in rat sensory neurons: a mechanism for the pain-inducing actions of ET-1. J Neurosci. 2002; 22(15):6325-6330. 27. Joshi SK, Mikusa JP, Hernandez G, et al. Involvement of the TTX-resistant sodium channel Nav 1.8 in inflammatory and neuropathic, but not post-operative, pain states. Pain. 2006;123(1-2):75-82. 28. Wang W, Gu J, Li YQ, Tao YX. Are voltagegated sodium channels on the dorsal root ganglion involved in the development of neuropathic pain? Mol Pain. 2011;7:16. 29. Gu XY, Liu BL, Zang KK, et al. Dexmedetomidine inhibits tetrodotoxinresistant Nav1.8 sodium channel activity through Gi/o-dependent pathway in rat dorsal root ganglion neurons. Mol Brain. 2015;8:15. 30. Zhao R, Pei GX, Cong R, Zhang H, Zang CW, Tian T. PKC-NF-kB are involved in CCL2-induced Nav1.8 expression and channel function in dorsal root ganglion neurons. Biosci Rep. 2014;34(3). 31. Cianfrocca R, Tocci P, Semprucci E, et al. Arrestin 1 is required for endothelin-1induced NF-kB activation in ovarian cancer cells. Life Sci. 2014;118(2):179-184. 32. Wang P, Qiu W, Dudgeon C, et al. PUMA is directly activated by NF-kappaB and contributes to TNF-alpha-induced apoptosis. Cell Death Differ. 2009;16(9):1192-1202. 33. Jarvis MF, Honore P, Shieh CC, et al. A803467, a potent and selective Nav1.8 sodium channel blocker, attenuates neuropathic and inflammatory pain in the rat. Proc Natl Acad Sci USA. 2007;104(20):85208525. 34. Qi Y, Wang JKT, McMillian M, Chikaraishi DM. Characterization of a CNS cell line, CAD, in which morphological differentiation is initiated by serum deprivation. J Neurosci. 1997;17(4):1217-1225 35. Moscat J, Diaz-Meco MaT, Rennert P. NF-ÎşB activation by protein kinase C isoforms and B-cell function. EMBO Rep. 2003;4(1):31-36. 36. He Y, Wilkie DJ, Nazari J, et al. PKCd-targeted intervention relieves chronic pain in a murine sickle cell disease modeld. J Clin Invest. 2016;126(8):3053-3057. 37. Fosdal MB. Perception of pain among pediatric patients with sickle cell pain crisis. J Pediatr Oncol Nurs. 2014;32(1):5-20. 38. McClish D, Levenson JL, Penberthy LT, et al. Gender differences in pain and health-

haematologica | 2018; 103(7)


Mouse model of sickle cell disease

39.

40.

41.

42.

43.

care utilization for adult sickle cell patients: The PiSCES Project. J Womens Health (Larchmt). 2006;15(2):146-154. Ryan TM, Ciavatta DJ, Townes TM. Knockout-transgenic mouse model of sickle cell disease. Science. 1997;278(5339):873876. Nevitt SJ, Jones AP, Howard J. Hydroxyurea (hydroxycarbamide) for sickle cell disease. Cochrane Database Syst Rev. 2017;4:CD002202. Ehrenreich H, Andereson RW, Fox CH, et al. Endothelins, peptides with potent vasoactive properties, are produced by human macrophages. J Exp Med. 1990; 172(6):1741-1748. Bremnes T, Paasche JD, Mehlum A, Sandberg C, Bremnes B, Attramadal H. Regulation and intracellular trafficking pathways of the endothelin receptors. J Biol Chem. 2000;275(23):17596-17604. Hansen-Schwartz J, Svensson CL, Xu CB,

haematologica | 2018; 103(7)

44.

45.

46.

47.

Edvinsson L. Protein kinase mediated upregulation of endothelin A, endothelin B and 5-hydroxytryptamine 1B/1D receptors during organ culture in rat basilar artery. Br J Pharmacol. 2002;137(1):118-126. Gold M, Levine J, Correa A. Modulation of TTX-R INa by PKC and PKA and their role in PGE2-induced sensitization of rat sensory neurons in vitro. J Neurosci. 1998;18(24):10345-10355. England S, Bevan S, Docherty RJ. PGE2 modulates the tetrodotoxin-resistant sodium current in neonatal rat dorsal root ganglion neurones via the cyclic AMP-protein kinase A cascade. J Physiol. 1996;495(Pt 2):429-440. Feng B, Strichartz G. Endothelin-1 raises excitability and reduces potassium currents in sensory neurons. Brain Res Bull. 2009; 79(6):345-350. Berti-Mattera LN, Larkin B, Hourmouzis Z, Kern TS, Siegel RE. NF-κB subunits are dif-

ferentially distributed in cells of lumbar dorsal root ganglia in naive and diabetic rats. Neurosci Lett. 2011;490(1):41-45. 48. Cang CL, Zhang H, Zhang YQ, Zhao ZQ. PKCepsilon-dependent potentiation of TTX-resistant Nav1.8 current by neurokinin-1 receptor activation in rat dorsal root ganglion neurons. Mol Pain. 2009;5:33. 49. Satoh A, Gukovskaya AS, Nieto JM, et al. PKC-d and -ε regulate NF-κB activation induced by cholecystokinin and TNF-alpha in pancreatic acinar cells. Am J Physiol Gastrointest Liver Physiol. 2004; 287(3):G582-591. 50. Clerk A, Bogoyevitch MA, Anderson MB, Sugden PH. Differential activation of protein kinase C isoforms by endothelin-1 and phenylephrine and subsequent stimulation of p42 and p44 mitogen-activated protein kinases in ventricular myocytes cultured from neonatal rat hearts. J Biol Chem. 1994; 269(52):32848-32857.

1135


ARTICLE

Red Cell Biology & its Disorders

Ferrata Storti Foundation

Proteomic analysis of plasma from children with sickle cell anemia and silent cerebral infarction

Sanjay Tewari,1 George Renney,2 John Brewin,1 Kate Gardner,1 Fenella Kirkham,3 Baba Inusa,4 James E Barrett,5 Stephan Menzel,1 Swee Lay Thein,6 Malcolm Ward2 and David C. Rees1

Haematologica 2018 Volume 103(7):1136-1142

1 Red Cell Biology Unit, King’s College Hospital, King’s College London, UK; 2Proteomics Laboratory, Institute of Psychiatry, King’s College London, UK; 3Department of Neurosciences, Institute of Child Health, University College Hospital, London, UK; 4Evelina Children’s Hospital, Guy’s and St Thomas’ Hospital, London, UK; 5Division of Health & Social Care Research, King's College London, UK; 6Sickle Cell Branch, National Heart, Lung and Blood Institute, National Institutes of Health, Bethesda, MD, USA

ABSTRACT

S

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/103/7/1136

ilent cerebral infarction is the most common neurological abnormality in children with sickle cell anemia, affecting 30-40% of 14 year olds. There are no known biomarkers to identify children with silent cerebral infarcts, and the pathological basis is also unknown. We used an unbiased proteomic discovery approach to identify plasma proteins differing in concentration between children with and without silent cerebral infarcts. Clinical parameters and plasma samples were analysed from 51 children (mean age 11.8 years, range 6-18) with sickle cell anemia (HbSS). A total of 19 children had silent cerebral infarcts and 32 normal MRI; the children with silent infarcts had lower HbF levels (8.6 vs. 16.1%, P=0.049) and higher systolic blood pressures (115 vs. 108.6, P=0.027). Plasma proteomic analysis showed 13 proteins increased more than 1.3 fold in the SCI patients, including proteins involved in hypercoagulability (α2-antiplasmin, fibrinogen−g chain, thrombospondin-4), inflammation (α2-macroglobulin, complement C1s and C3), and atherosclerosis (apolipoprotein B-100). Higher levels of gelsolin and retinol-binding protein 4 were also found in the population with silent infarcts, both of which have been linked to stroke. We investigated the genetic basis of these differences by studying 359 adults with sickle cell disease (199 with silent cerebral infarcts, 160 normal MRIs), who had previously undergone a genome-wide genotyping array. None of the genes coding for the differentially expressed proteins were significantly associated with silent infarction. Our study suggests that silent cerebral infarcts in sickle cell anemia may be associated with higher systolic blood pressure, lower HbF levels, hypercoagulability, inflammation and atherosclerotic lipoproteins.

©2018 Ferrata Storti Foundation

Introduction

Correspondence: david.rees2@nhs.net.

Received: January 10, 2018. Accepted: March 14, 2018. Pre-published: March 15, 2018.

doi:10.3324/haematol.2018.187815

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

Sickle cell anemia (SCA) is the most common cause of stroke in childhood.1 The abnormal sickle hemoglobin (HbS) polymerizes when deoxygenated, damaging red cells, causing vaso-occlusion and vascular endothelial dysfunction.2 A cascade of pathological processes follows, including inflammation, hemolysis, anemia, oxidative stress, reperfusion injury, hypercoagulability, and nitric oxide deficiency.3 Survival is reduced by 20-30 years4,5 and vasculopathic complications include pulmonary hypertension, priapism, and cerebrovascular disease.6 The latter is the major cause of morbidity in children and takes two forms: overt stroke often associated with large vessel disease,7 and silent cerebral infarcts (SCI) of less certain pathology.8 Without intervention, overt stroke has a peak incidence of 1.02/100 patient years between 2 and 5 years,9 although primary prevention using transcrahaematologica | 2018; 103(7)


Proteomics of silent cerebral infarction in SCA

nial Doppler (TCD) scanning and blood transfusion reduces this by 90%.10 The implementation of TCD-based stroke prevention has led to a significant fall in overt strokes in many countries.11,12 SCI is more common than overt stroke with MRI showing relevant lesions in 13% of two year olds, 25% of six year olds, and 30-40% of 14 year olds.13 Although by definition SCIs are not associated with overt symptoms, they cause significant morbidity, including a reduction in IQ, defects in executive function, epilepsy, and increased risk of further SCIs and overt strokes.13 Early detection of SCI is useful to allow cognitive assessment, educational support and therapeutic intervention. The Silent Cerebral Infarct Multi-Center Clinical Trial (SIT Trial), a randomised controlled study, showed that regular blood transfusion reduced the incidence of recurrent infarction in children with SCIs compared to the observational arm.14 The pathophysiology of SCI is unclear. It is often stated that SCIs are caused by small vessel disease, although there is no direct evidence for this. There are no postmortem studies looking at the histological changes corresponding to the MRI appearances of SCI, although older studies identified small necrotic lesions in the subcortex of the brain, possibly representing SCIs.15 It is also possible that SCIs are areas of demyelination, or linked to venous sinus thrombosis.13 SCIs may also be caused by difficulties maintaining constant blood flow to the brain, leading to watershed infarction precipitated by acute anemia or hypoxia.16 The lack of pathophysiological understanding makes it difficult to develop new therapeutic strategies. Here, we used an unbiased proteomic approach to identify plasma proteins associated with SCI and potentially linked to the underlying pathophysiology. It is not easy to identify children with or at increased risk of SCI. Baseline data from the SIT Trial showed that steady state hemoglobin <7.6g/dl, steady state systolic blood pressure >104mmHg and male sex were all associated with increased risk of SCI, although a model combining all three significant factors had weak predictive powers, with a C statistic of about 0.6.8 Abnormal transcranial Doppler velocities are not reliably associated with SCI, although there is a possible association with stenosis of the cervical internal carotid artery.17 Neuropsychometry can detect some children with SCIs, but is time consuming and not easily available. MRI is currently the only way of diagnosing SCI, although before 7 years of age this usually requires sedation or general anesthetic, which carries significant risks in children with SCD and is unsuitable for screening. MRI is also not widely available in many African countries where SCA is most prevalent. It would therefore be beneficial to develop a simple, screening test for SCI, potentially based on blood testing. Proteomics can detect, identify and quantify very low concentrations of proteins. For example, a recent study suggested that urine proteomics might help in the diagnosis of acute stroke by detecting brain peptides released by ischemia.18 The presence of SCI may suggest small blood vessel disease and a tendency to infarct brain tissue, which is likely to be a chronic process, releasing detectable brain proteins into plasma. Plasma protein profiles could also differ between those with and without SCI because of causative pathological factors, including coagulopathy, inflammation, hypoxia and vasculopathy; these differences may be acquired or inherited, and reflect genetic haematologica | 2018; 103(7)

predisposition to silent cerebral infarction. We investigated the possibility that there might be differences in the plasma proteome between children with and without SCIs, with a view to understanding more about the pathophysiology of SCIs and identifying potentially useful biomarkers. We further investigated the hypothesis that genetic factors account for the differences in concentrations of the various proteins, using data from a genome wide association study (GWAS) in a separate cohort of patients with SCA and SCIs.

Methods Patients and setting The National Research Ethics Committee approved the study (reference 13/LO/0709). Children were recruited from clinics at Kingâ&#x20AC;&#x2122;s College Hospital and the Evelina Childrenâ&#x20AC;&#x2122;s Hospital. Approximately 1000 different children are seen annually in these clinics with a further 500 seen in outreach clinics.19 The aim was to recruit a total of 50 patients, with equal numbers of those with SCIs and controls with normal brain MRIs. Children known to have SCIs from previous MRI scans were selectively recruited. Eligible patients without previous MRIs were recruited sequentially in clinic, on the basis that 20-30% of these children would have SCIs, and combining the two groups would result in approximately 25 children in each group. SCI and control groups were not specifically matched for age and gender. Inclusion criteria were: sickle cell anemia (HbSS), age 8-18 years old, normal or conditional TCD velocities (<200cm/s) (10), able to tolerate MRI scan without sedation. Exclusion criteria were: history of overt stroke, blood transfusion in last 4 months, serious co-existing disorder (renal failure, HIV, hepatitis, malignancy, autoimmune disease, chronic infection, cardiac disease, long-term medication), acute complications requiring hospital attendance in last 3 months. Patients on hydroxyurea (HU) were recruited to both arms. Routine clinical and laboratory data were collected.

Brain MRI and neurological assessment Brain MRIs were performed to identify the presence of SCIs using standardised definitions20 (see Online Supplementary Materials). On the day of blood sampling, all children underwent standardised neurological examination by experienced clinicians to document the absence of neurological abnormalities suggestive of an overt stroke. All children had routine TCD scans performed at least annually, and these were used in this study.

Proteomic analysis Blood samples were collected on all children at the time of consent. A data-driven proteomics approach was used to discover protein changes in the plasma associated with SCI. We used an established workflow combining isobaric Tandem Mass Tags (TMT) reagents with off-gel fractionation, followed by high sensitivity rapid throughput mass spectrometry (MS) and LC-MS/MS. (details in Online Supplementary Materials). Relative amounts of each peptide were compared between the SCI and control groups, and P values were calculated for those peptides showing greater than 1.3 fold differences.

Targeted plasma analysis for biomarkers of neurological disease Plasma samples were also analysed for 92 protein biomarkers known to be implicated in neurological disease, using a multiplexed proximity extension assay21 (see Online Supplementary Materials for full list of biomarkers). 1137


S. Tewari et al.

Data analysis Statistical significance was assessed using appropriate tests to compare means with adjustments for multiple comparisons when appropriate. Statistical analyses were performed in the proteomics laboratory and medical statistics department of Public Health Sciences at Kingâ&#x20AC;&#x2122;s College London.

Genetic association of protein biomarkers with SCIs The genes coding for plasma proteins which differed significantly between children with and without SCIs were identified using standard databases. Genetic association analysis was then undertaken in a separate adult cohort at KCH, evaluating the variants within each gene and their association with SCI 22 (details in Online Supplementary Materials).

Results A total of 55 patients consented to take part in the study, although no blood samples were taken from one, and three further patients did not have an MRI scan performed. Plasma from 54 patients underwent proteomic analysis, although only data from the 51 patients (22 female) with MRI results were analysed. In total, 19 (37%) had SCIs and 32 (63%) normal brain MRIs, deviating slightly from the aim of equal numbers in each group because of recruitment of children without previous MRIs, the majority of whom had normal brain imaging (Figure 1). A total of 15 children were identified as having SCIs from scans performed as part of this study, and four children were recruited as a result of previous MRI scans showing SCIs; in these four children, there was a median delay of 20 months (range 9 â&#x20AC;&#x201C; 40 months) between MRI scan and blood sampling for proteomic analysis. One child

had conditional velocities and 50 normal velocities on TCD scanning.

Basic clinical and laboratory data Relatively more males (13/29, 44%) than females (6/22, 27%) were in the SCI group, although this was not significantly different (Chi square, P=0.199). There was no significant difference in HU use between the SCI and control groups: controls 11/32 (36%) on HU, SCI 7/19 (37%) on HU (Chi square, P=0.86). G6PD assays were available for 45 patients, and there was no association with SCI: G6PD deficiency was present in 3/29 (10%) controls and 2/16 (12%) with SCIs (Chi square, P=0.83). Basic laboratory and clinical characteristics are summarized in Table 1. Mean HbF levels were significantly lower (6.1% versus 8.6%, P=0.049), and mean systolic blood pressures significantly higher (115 versus 108.6, P=0.027) in those with SCIs compared to controls.

Plasma proteomic analysis A total of 4662 different peptides were identified, although only 1312 were present in all 51 patient samples, corresponding to 346 different proteins. Forty-four peptides from 13 different proteins were present at more than 1.3 times the concentration in the SCI patients compared to controls (Table 2), whereas 41 peptides from 4 proteins were more than 1.3 higher in the control population (Table 3). Proteins differentially expressed in the SCI group included those implicated in coagulation, lipid and inflammatory pathways.

Genetic association of identified biomarkers with SCI We explored the possibility that differences in concentrations of proteins (Tables 2 and 3) between SCI and con-

Figure 1: Diagram showing the flow of patients through the study.

1138

haematologica | 2018; 103(7)


Proteomics of silent cerebral infarction in SCA

trol populations were inherited, using a different sample of adult patients who had undergone brain MRI and a genome-wide genotyping array. Of 359 patients with brain MRIs and MEGAchip data, 199 had SCI and 160 did not. Regions of interest were analysed for the candidate genes identified in the proteomic study; the most significant genetic variant from each region is shown in Table 4. No genetic variant in the candidate genes was significantly associated with SCI after correcting for multiple testing. A variant in gelsolin was closest to achieving significance (P=0.00029, threshold for significance 0.00009).

Table 1. Comparison of laboratory and clinical measurements of those with and without SCI. P values from t-tests, with significant results shown in bold (P<0.05), not corrected for multiple comparisons.

Age (years) Hb (g/L) HbF (%) MCH (pg) Neutrophils (x109/L) Reticulocytes (%) Bilirubin AST Creatinine LDH Diastolic bp (mmHg) Systolic bp (mmHg)

N.

Controls mean

s.d.

Silent cerebral infarction N. mean s.d.

32 32 32 32 32

11.5 83.4 8.6 28.0 4.7

3.17 8.45 5.1 4.7 1.6

19 19 18 19 19

12.4 86.1 6.1 28.6 4.4

3.82 11.1 3.6 4.0 1.8

P value 0.40 0.33 0.049 0.68 0.55

Concentrations of 92 known neurological biomarkers were measured in the plasma of 54 patients who consented to the study and gave blood samples. Fifty-one patients had MRI scans available and were analysed. In order to control for the potential effect of hydroxyurea, a logistic regression model was fitted for each biomarker; the dependent variable was SCI status, and the biomarker and hydroxyurea were included as covariates in each model. Seven biomarkers were significantly associated with SCI, but after correction for multiple hypotheses none of these were significant (Table 5). Several biomarkers were strongly correlated with each other and a heatmap of the correlation matrix indicated a moderate level of redundancy between them. A principal component analysis found that approximately 90% of the variance could be attributed to the first 30 principal components. None of the principal components were found to be differentially

Table 3. List of all peptides and proteins present at significantly decreased concentrations (>0.77 fold) in children with SCIs compared to controls. Where more than one peptide was identified, the most significant difference is given. Concentrations are normalized against standard controls.

Protein 27

12.3

3.7

15

12.4

4.7

0.99

32 32 32 32 26

40.5 57.5 31.9 567 66.5

15.5 26.0 9.9 135 6.9

19 19 19 18 14

42.9 56.0 36.4 565 66.5

20.5 19.0 11.8 148 4.9

0.64 0.84 0.14 0.97 0.99

26

108.6

7.6

14

115

9.6

0.027

Number Normalized Normalized Fold P value different concentration concentration change peptides in controls in SCI in SCI

Ig alpha-1 chain 1 Ig gamma-1 chain 35 Ig gamma- 3 chain 1 Platelet basic protein 4

0.757 1.74 1.08 1.865

0.541 1.09 0.821 1.27

0.71 0.63 0.76 0.68

0.021 0.018 0.035 0.015

Table 4. Association of candidate genes for SCI identified in the proteomic part of the study. The most significant genetic variants are shown. None of the variants were significant after adjustment for multiple comparisons.

Table 2. List of all peptides and proteins present at significantly increased concentrations (>1.3 fold) in children with SCIs compared to controls. Where more than one peptide was identified, the most significant difference is given. Concentrations are normalized against standard controls.

Protein

Targeted plasma analysis

Number Normalized Normalized Fold P value different concentration concentration change peptides in controls in SCI in SCI

α-2-antiplasmin 3 Albumin 1 α-2-macroglobulin 7 Apolipoprotein A-IV 7 Apolipoprotein B-100 1 Complement C1s Complement C3 2 Fibrinogen g chain 1 Gelsolin 10 Inter-α−trypsin 2 inhibitor heavy chain-1 Pigment epithelium1 derived factor Retinol-binding protein 4 3 Thrombospondin-4 2

0.635 0.505 0.601 0.28 0.303 0.42 0.414 0.936 0.411 0.833

0.914 0.703 0.832 0.421 0.428 0.565 0.561 1.25 0.577 1.14

1.44 1.39 1.38 1.50 1.41 1.34 1.35 1.33 1.40 1.36

0.004 0.005 0.01 0.0002 0.014 0.014 0.004 0.01 0.00006 0.002

1.14

1.5

1.32

0.006

0.465 0.577

0.625 0.763

1.34 1.32

0.002 0.001

haematologica | 2018; 103(7)

Protein

Gene Chromosome Number Symbol of SNPs analysed

Alpha-2-macroglobulin A2M Albumin ALB Apoliprotein B-100 APOB ComplementC1s C1S Complement C3 C3 Fibrinogen gamma chain FGG Gelsolin GSN Ig alpha-1 chain IGHA1 Ig gamma-1 chain IGHG1 Ig gamma-3 chain IGHG3 Inter-alpha-trypsin ITIH1 inhibitor heavy chain Pigment epithelium-derived PEDF/S factor ERPINF1 Platelet basic protein PPBP Retinol-binding protein 4 RBP4 Alpha-2-antiplasmin SERPINF2 Thrombospondin 4 THBS4

P value of most significant genetic variant

9 4 2 12 19 4 9 14 14 14 3

431 89 180 360 449 23 607 25 43 26 75

0.0140 0.0016 0.0587 0.0016 0.0257 0.0258 0.0003 0.2699 0.0609 0.0488 0.0180

17

117

0.0495

4 10 17 5

4 99 70 387

0.4393 0.0040 0.0233 0.0207 1139


S. Tewari et al.

expressed with respect to SCI status, either with or without controlling for HU status.

Table 5. Known biomarkers of neurological disease showing significant differences between SCI and control populations, although none of these maintain significance when correcting for multiple tests.

Unadjusted P value Corrected P value

Discussion Our study found that the presence of SCI in SCA is associated with higher systolic blood pressures, an association that was also found in the SIT trial.8 Blood pressure is relatively low in SCA,23 possibly related to chronic anemia and subsequent vasodilatation, and the levels found in our study, with a mean of 115mmHg in the SCI group, would not be considered hypertensive. The exact way in which relative systolic hypertension is associated with SCI in children with SCA is unknown, although systolic hypertension is linked with other vascular complications in SCA (overt stroke,23 renal insufficiency,24 and high tricuspid jet velocity24). Systolic hypertension is also a risk factor for SCI in the non-sickle population.25 The association between relative systolic hypertension and SCI in SCA seems robust, and is biologically plausible, although there is currently no evidence on whether it is beneficial to lower blood pressure in children with SCA, and at what level such treatment should be considered. We also found that higher HbF levels were protective against SCI. Increased HbF levels lessen the severity of SCA, with a clear association with decreased pain and longer life expectancy,2 but a more complex association with cerebrovascular complications. In the SIT Trial, HbF was not associated with SCI in a logistic regression model8 and analysis of data from the Cooperative Study of Sickle Cell Disease also showed no protective effect of higher HbF levels.26 Other studies have suggested that there may be neurological benefit to having higher HbF levels. In France, HbF levels were higher in children without SCIs compared to those with (16.5% vs. 13.1%, P=0.02);27 a Dutch study found that although HbF did not protect against the occurrence of SCIs, higher levels were associated with a smaller volume of white matter hyperintensities;28 a study on Brazilian patients showed protection against stroke was associated with genetic determinants of high HbF levels, particularly BCL11A alleles.29 It seems likely that the inherited ability to make more HbF offers some protection against neurological injury in SCA, although this may not always be reflected in the measured HbF percentage, particularly in small studies. The unbiased proteomic discovery part of this study identified 44 peptides from 17 different proteins which were significantly up- or down-regulated compared to levels in the control population, without correction for multiplicity (Tables 2 and 3). Many of these proteins can be plausibly implicated in pathogenesis of SCI and fall in to three broad groups.

Prothrombotic proteins Four proteins which were found at higher levels in the SCI patients are known to be prothrombotic. α2-antiplasmin is a serine protease inhibitor which inactivates plasmin and decreases the rate of fibrinolysis; high levels are associated with increased ischemic stroke in the non-sickle population.30 Similarly, α2-macroglobulin has prothrombotic properties via the inactivation of plasmin, and elevated levels have been linked to stroke and white matter lesions in non-sickle patients.31 The fibrinogen g chain was increased 1.33 fold in the SCI population, and is at the cen1140

Von Willebrand factor C domain 2 containing Protein Smooth muscle-associated protein 2 ADAM metalloproteinase domain 22 ADAM metalloproteinase domain 23 Scavenger receptor class A member 5 TNF receptor superfamily member 12A Bone morphogenetic protein 4

0.03

0.62

0.02 0.04 0.04 0.03 0.02 0.03

0.62 0.62 0.62 0.62 0.62 0.62

ter of fibrin polymerization, interaction with platelets and regulation of factor XIII activity.32 Increased levels have been associated with vasculopathy, higher levels being increased in inflammation and also possibly inherited.33 Fourthly, thrombospondin-4 was increased 1.32 fold with SCI. Thrombospondins mediate cellular adhesion to other cells and matrix, and are implicated in many thrombotic and vasculopathic processes. Thrombospondin-1 levels have been found to be increased in children with SCA in association with both SCI34 and overt stroke,35 although thrombospondin-4 has not previously been studied in this context. These findings support the potential use of anticoagulant and antiplatelet agents to prevent and treat SCI in SCA.

Pro-inflammatory proteins

Elevated levels of α2-macroglobulin, complement C1s and complement C3 in the SCI patients suggest that there is increased inflammation, which is known to predispose towards vasculopathy and stroke. As mentioned previously, α2-macroglobulin is a prothrombotic protease inhibitor, but is also involved in inflammatory states, as a carrier for interleukin-6.31 Both complements C1s and C3 are elevated in inflammation, and high levels of C3 are a known risk factor for atherosclerosis and stroke in the general population.36 Inflammation is known to be an important component of pathophysiology in SCA, and the role of antiinflammatory agents in preventing SCIs should be explored further.

Lipoproteins Apolipoprotein B-100 is the main component of low density lipoprotein cholesterol (LDL-C) and high levels are a well-established risk factor for atherosclerosis. Various studies have shown that drugs which reduce LDL-C also reduce the rate of cardiovascular events, including ischemic stroke,37 suggesting a possible therapeutic option for children with SCA and ischemic stroke for drugs such as statins. We also found that apolipoprotein A-IV levels were higher in the SCI population, which is perhaps paradoxical in that it is thought to be protective against coronary syndromes,38 although it has not been directly implicated in cerebrovascular disease or studied in children. A recent study suggested that higher levels were associated with reduced glomerular filtration rates,39 and this could be one explanation for the differences in our study. haematologica | 2018; 103(7)


Proteomics of silent cerebral infarction in SCA

Other interesting candidates Several other interesting candidates were identified, both shedding light on the pathophysiology of SCI and with the potential to act as useful biomarkers. Gelsolin binds to and remodels actin by both severing and capping the protein; it is present in both cytoplasm and plasma. Increased plasma levels have been identified as a poor prognostic marker in ischemic stroke.40 Further significant actions in the context of SCI include a role in promoting platelet formation and activation.41 In our genetic analysis, a genetic variant in the gelsolin gene showed a trend towards being associated with SCI. Retinol binding protein-4 is a lipocalin transporting vitamin A from the liver to other tissues, including the brain; plasma levels have been found to be elevated in atherosclerosis and also ischemic stroke.42,43 Pigment epithelium-derived factor is multifunctional with anti-angiogenic properties, secreted by different tissues including adipocytes. It is potentially significant in the development of SCI because of its ability to induce insulin resistance, and inflammation and proliferation of muscle cells.44 It is harder to know the significance of some other proteins, including the lower levels of immunoglobulin chains and platelet basic protein in SCI, although immunodeficiency has been associated with primary and recurrent stroke in the general pediatric population.45 Several plausible biomarkers for SCI were identified in the targeted plasma analysis, including von Willebrand factor C domain, although none of these reached statistical significance when corrected for multiple assays. Similarly, none of the candidate proteins identified by proteomics showed a convincing genetic association with SCI in a separate cohort of 359 patients with SCA, although the pathophysiology of SCIs seen in the adults used in this part of the study may differ from that in the children used in the proteomic part. This may also reflect the relatively small size of this study, or that the changes seen in proteins of those with SCIs are secondary to other pathological events rather than inherited quantitative trait loci. The strength of our study is that it used an unbiased

References 1. Earley CJ, Kittner SJ, Feeser BR, et al. Stroke in children and sickle-cell disease: BaltimoreWashington Cooperative Young Stroke Study. Neurology. 1998;51(1):169-176. 2. Brousse V, Makani J, Rees DC. Management of sickle cell disease in the community. BMJ. 2014;348:g1765. 3. Rees DC, Gibson JS. Biomarkers in sickle cell disease. Br J Haematol. 2012;156(4):433445. 4. Platt OS, Brambilla DJ, Rosse WF, et al. Mortality in sickle cell disease. Life expectancy and risk factors for early death. N Engl J Med. 1994;330(23):1639-1644. 5. Gardner K, Douiri A, Drasar E, et al. Survival in adults with sickle cell disease in a highincome setting. Blood. 2016;128(10):14361438. 6. Kato GJ, Gladwin MT, Steinberg MH. Deconstructing sickle cell disease: reappraisal of the role of hemolysis in the development of clinical subphenotypes. Blood Rev. 2007;21(1):37-47.

haematologica | 2018; 103(7)

proteomics approach to identify proteins involved in the pathogenesis of SCI, together with a combination of targeted plasma analysis and genetic association studies. Our findings suggest that proinflammatory and prothrombotic states contribute to the development of SCI, with possible abnormalities in lipoproteins. Previous studies to identify biomarkers for SCI in SCD have been summarized by Lance et al.35 They identified nine studies measuring a range of different candidates, although no studies used an unbiased proteomic approach. In total, 17 potential candidates were identified in this review, again with a predominance of factors suggesting coagulation activation and increased inflammation.35 None of the identified candidates were the same as our candidates, although one study found increased levels of thrombospondin-1 in children with SCIs and SCA.34 Although our study is relatively small, in keeping with proteomic discovery approaches, some significant factors have emerged. We confirmed the importance of hypertension as a risk factor for SCI, together with support for higher HbF levels being protective, suggesting that antihypertensives and hydroxyurea may be of benefit. The utility of individual biomarkers, such as gelsolin or Îą-2antiplasmin, to act as clinically useful biomarkers in screening for SCI needs to be confirmed in larger, prospective studies. Previous studies have shown increased inflammation and coagulation in children with SCD,46,47 possibly linked to small blood vessel disease,48 and our study suggests that this is more marked in children who also have SCIs. This relatively prothrombotic and proinflammatory state associated with SCIs suggests that it may be useful to study anti-inflammatory and antiplatelet agents in clinical trials of SCI prevention. Ongoing clinical trials of antiplatelet agents,49 statins50 and canakinumab are particularly relevant in this context. Acknowledgements The authors would like to thank the patients involved in this study and the Stroke Association for funding this research (Grant TSA 2012/06).

7. Helton KJ, Adams RJ, Kesler KL, et al. Magnetic resonance imaging/angiography and transcranial Doppler velocities in sickle cell anemia: results from the SWiTCH trial. Blood. 2014;124(6):891-898. 8. DeBaun MR, Sarnaik SA, Rodeghier MJ, et al. Associated risk factors for silent cerebral infarcts in sickle cell anemia: low baseline hemoglobin, sex, and relative high systolic blood pressure. Blood. 2012;119(16):36843690. 9. Ohene-Frempong K, Weiner SJ, Sleeper LA, et al. Cerebrovascular accidents in sickle cell disease: rates and risk factors. Blood. 1998;91(1):288-294. 10. Adams RJ, McKie VC, Hsu L, et al. Prevention of a first stroke by transfusions in children with sickle cell anemia and abnormal results on transcranial Doppler ultrasonography. N Engl J Med. 1998;339(1):5-11. 11. Bernaudin F, Verlhac S, Arnaud C, et al. Impact of early transcranial Doppler screening and intensive therapy on cerebral vasculopathy outcome in a newborn sickle cell anemia cohort. Blood. 2011;117(4):11301140; quiz 1436.

12. Fullerton HJ, Adams RJ, Zhao S, Johnston SC. Declining stroke rates in Californian children with sickle cell disease. Blood. 2004;104(2):336-339. 13. DeBaun MR, Armstrong FD, McKinstry RC, Ware RE, Vichinsky E, Kirkham FJ. Silent cerebral infarcts: a review on a prevalent and progressive cause of neurologic injury in sickle cell anemia. Blood. 2012;119(20):45874596. 14. DeBaun MR, Gordon M, McKinstry RC, et al. Controlled trial of transfusions for silent cerebral infarcts in sickle cell anemia. N Engl J Med. 2014;371(8):699-710. 15. Rothman SM, Fulling KH, Nelson JS. Sickle cell anemia and central nervous system infarction: a neuropathological study. Ann Neurol. 1986;20(6):684-690. 16. Dowling MM, Quinn CT, Plumb P, et al. Acute silent cerebral ischemia and infarction during acute anemia in children with and without sickle cell disease. Blood. 2012;120(19):3891-3897. 17. Deane CR, Goss D, Bartram J, et al. Extracranial internal carotid arterial disease in children with sickle cell anemia.

1141


S. Tewari et al. Haematologica. 2010;95(8):1287-1292. 18. Dawson J, Walters M, Delles C, Mischak H, Mullen W. Urinary proteomics to support diagnosis of stroke. PLoS One. 2012;7 (5):e35879. 19. Day TG, Thein SL, Drasar E, et al. Changing pattern of hospital admissions of children with sickle cell disease over the last 50 years. J Pediatr Haematol Oncol. 2011;33(7):491495. 20. Casella JF, King AA, Barton B, et al. Design of the silent cerebral infarct transfusion (SIT) trial. Pediatr Hematol Oncol. 2010;27(2):6989. 21. Assarsson E, Lundberg M, Holmquist G, et al. Homogenous 96-plex PEA immunoassay exhibiting high sensitivity, specificity, and excellent scalability. PLoS One. 2014;9(4): e95192. 22. Cheverud JM, Vaughn TT, Pletscher LS, et al. Genetic architecture of adiposity in the cross of LG/J and SM/J inbred mice. Mamm Genome. 2001;12(1):3-12. 23. Rodgers GP, Walker EC, Podgor MJ. Is "relative" hypertension a risk factor for vasoocclusive complications in sickle cell disease? Am J Med Sci. 1993;305(3):150-156. 24. Gordeuk VR, Sachdev V, Taylor JG, Gladwin MT, Kato G, Castro OL. Relative systemic hypertension in patients with sickle cell disease is associated with risk of pulmonary hypertension and renal insufficiency. Am J Hematol. 2008;83(1):15-18. 25. Vermeer SE, Koudstaal PJ, Oudkerk M, Hofman A, Breteler MM. Prevalence and risk factors of silent brain infarcts in the population-based Rotterdam Scan Study. Stroke. 2002;33(1):21-25. 26. Kinney TR, Sleeper LA, Wang WC, et al. Silent cerebral infarcts in sickle cell anemia: a risk factor analysis. The Cooperative Study of Sickle Cell Disease. Pediatrics. 1999;103(3):640-645. 27. Bernaudin F, Verlhac S, Arnaud C, et al. Chronic and acute anemia and extracranial internal carotid stenosis are risk factors for silent cerebral infarcts in sickle cell anemia. Blood. 2015;125(10):1653-1661. 28. van der Land V, Mutsaerts HJ, Engelen M, et al. Risk factor analysis of cerebral white matter hyperintensities in children with sickle cell disease. Br J Haematol. 2016;172 (2):274-284.

1142

29. Leonardo FC, Brugnerotto AF, Domingos IF, et al. Reduced rate of sickle-related complications in Brazilian patients carrying HbFpromoting alleles at the BCL11A and HMIP2 loci. Br J Haematol. 2016;173(3):456-460. 30. Reed GL, Houng AK, Singh S, Wang D. alpha2-Antiplasmin: new insights and opportunities for ischemic stroke. Semin Thromb Hemost. 2017;43(2):191-199. 31. Nezu T, Hosomi N, Aoki S, et al. Alpha2macroglobulin as a promising biomarker for cerebral small vessel disease in acute ischemic stroke patients. J Neurol. 2013;260(10):2642-2649. 32. Mosesson MW. Fibrinogen gamma chain functions. J Thromb Haemost. 2003;1(2): 231-238. 33. Williams SR, Hsu FC, Keene KL, et al. Shared genetic susceptibility of vascularrelated biomarkers with ischemic and recurrent stroke. Neurology. 2016;86(4):351-359. 34. Faulcon LM, Fu Z, Dulloor P, et al. Thrombospondin-1 and L-selectin are associated with silent cerebral infarct in children with sickle cell anaemia. Br J Haematol. 2013;162(3):421-424. 35. Lance EI, Casella JF, Everett AD, BarronCasella E. Proteomic and biomarker studies and neurological complications of pediatric sickle cell disease. Proteomics Clin Appl. 2014;8(11-12):813-827. 36. Niculescu F, Rus H. The role of complement activation in atherosclerosis. Immunol Res. 2004;30(1):73-80. 37. Ray KK, Ginsberg HN, Davidson MH, et al. Reductions in atherogenic lipids and major cardiovascular events: a pooled analysis of 10 ODYSSEY trials comparing alirocumab with control. Circulation. 2016;134(24): 1931-1943. 38. Kronenberg F, Stuhlinger M, Trenkwalder E, et al. Low apolipoprotein A-IV plasma concentrations in men with coronary artery disease. J Am Coll Cardiol. 2000;36(3):751-757. 39. Mack S, Coassin S, Vaucher J, Kronenberg F, Lamina C, Apo AIVGC. Evaluating the causal relation of ApoA-IV with diseaserelated traits - A bidirectional two-sample Mendelian randomization study. Sci Rep. 2017;7(1):8734. 40. Garcia-Berrocoso T, Penalba A, Boada C, et al. From brain to blood: new biomarkers for ischemic stroke prognosis. J Proteomics.

2013;94:138-148. 41. Silacci P, Mazzolai L, Gauci C, Stergiopulos N, Yin HL, Hayoz D. Gelsolin superfamily proteins: key regulators of cellular functions. Cell Mol Life Sci. 2004;61(19-20): 2614-2623. 42. Llombart V, Garcia-Berrocoso T, Bustamante A, et al. Plasmatic retinol-binding protein 4 and glial fibrillary acidic protein as biomarkers to differentiate ischemic stroke and intracerebral hemorrhage. J Neurochem. 2016;136(2):416-424. 43. Sasaki M, Otani T, Kawakami M, Ishikawa SE. Elevation of plasma retinol-binding protein 4 and reduction of plasma adiponectin in subjects with cerebral infarction. Metabolism. 2010;59(4):527-532. 44. Famulla S, Lamers D, Hartwig S, et al. Pigment epithelium-derived factor (PEDF) is one of the most abundant proteins secreted by human adipocytes and induces insulin resistance and inflammatory signaling in muscle and fat cells. Int J Obes (Lond). 2011;35(6):762-772. 45. Ganesan V, Prengler M, Wade A, Kirkham FJ. Clinical and radiological recurrence after childhood arterial ischemic stroke. Circulation. 2006;114(20):2170-2177. 46. van der Land V, Peters M, Biemond BJ, Heijboer H, Harteveld CL, Fijnvandraat K. Markers of endothelial dysfunction differ between subphenotypes in children with sickle cell disease. Thromb Res. 2013;132(6):712-717. 47. Chen J, Hobbs WE, Le J, Lenting PJ, de Groot PG, Lopez JA. The rate of hemolysis in sickle cell disease correlates with the quantity of active von Willebrand factor in the plasma. Blood. 2011;117(13):3680-3683. 48. Colombatti R, De Bon E, Bertomoro A, et al. Coagulation activation in children with sickle cell disease is associated with cerebral small vessel vasculopathy. PLoS One. 2013;8(10):e78801. 49. Heeney MM, Hoppe CC, Abboud MR, et al. A multinational trial of prasugrel for sickle cell vaso-occlusive events. N Engl J Med. 2016;374(7):625-635. 50. Hoppe C, Jacob E, Styles L, Kuypers F, Larkin S, Vichinsky E. Simvastatin reduces vaso-occlusive pain in sickle cell anaemia: a pilot efficacy trial. Br J Haematol. 2017;177(4):620-629.

haematologica | 2018; 103(7)


ARTICLE

Complications in Hematology

Late effects after hematopoietic stem cell transplantation for β-thalassemia major: the French national experience

Ilhem Rahal,1 Claire Galambrun,1 Yves Bertrand,2 Nathalie Garnier,2 Catherine Paillard,3 Pierre Frange,4 Corinne Pondarré,5 Jean Hugues Dalle,6 Regis Peffault de Latour,7 Mauricette Michallet,8 Dominique Steschenko,9 Despina Moshous,4 Patrick Lutz,3 Jean Louis Stephan,10 Pierre Simon Rohrlich,11 Ibrahim Yakoub-Agha,12 Françoise Bernaudin,5 Christophe Piguet,13 Nathalie Aladjidi,14 Catherine Badens,15 Claire Berger,10 Gérard Socié,7 Cécile Dumesnil,16 Marie Pierre Castex,17 Marilyne Poirée,11 Anne Lambilliotte,12 Caroline Thomas,18 Pauline Simon,19 Pascal Auquier,20 Gérard Michel,1 Anderson Loundou,20 Imane Agouti15 and Isabelle Thuret,1,15

Service d’Hémato-Oncologie Pédiatrique, Hôpital d’Enfant de la Timone, Assistance Publique des Hôpitaux de Marseille; 2Service d'Hématologie et Immunologie Pédiatrique, Institut d'Hématologie et d'Oncologie Pédiatrique, Lyon; 3Service d’Hémato-Oncologie Pédiatrique, CHU de Strasbourg - Hôpital de Hautepierre; 4Service d’Immunologie Hématologie Pédiatrique, CHU Necker-Enfants Malades, Assistance Publique Hôpitaux de Paris; 5Service de Pédiatrie, Centre de Référence de la Drépanocytose, Centre Hospitalier Intercommunal de Créteil (CHIC); 6Service d’Immunologie Hématologie, Hôpital Robert Debré, Assistance Publique Hôpitaux de Paris; 7Service d’Hémato-Oncologie - Greffe, Hôpital Saint Louis, Assistance Publique Hôpitaux de Paris; 8Service d’Hématologie, Centre Hospitalier Lyon Sud, Pierre-Bénite; 9Service d’Hémato-Oncologie Pédiatrique, CHRU Nancy, Hôpitaux de Brabois, Vandœuvre-lès-Nancy; 10Service d’Immuno-Hématologie et Oncologie Pédiatrique, CHU de Saint-Étienne, Saint-Priest-en-Jarez; 11Service d’Hémato-Oncologie Pédiatrique, Hôpital l'Archet 2, CHU de Nice; 12Service de Maladies du Sang, CHRU LilleHôpital Claude Huriez; 13Service d’Hémato-Oncologie Pédiatrique, Hôpital de la Mère et de l’Enfant, CHU de Limoges; 14Service de Pédiatrie Médicale, Groupe Hospitalier Pellegrin Enfants, Bordeaux; 15Centre de Référence Thalassémie, Hôpital d’Enfant de la Timone, Assistance Publique des Hôpitaux Marseille; 16Service d’Immuno-Hématologie et Oncologie Pédiatrique, CHU-Hôpitaux de Rouen; 17Service d’Hémato-Oncologie Pédiatrique, Hôpital Des Enfants, CHU de Toulouse; 18Service d’Hématologie Pédiatrique, Hôpital EnfantAdolescent, CHU Nantes; 19Service d’Hémato-Oncologie Pédiatrie, CHRU Jean Minjoz, Besançon and 20Service de Santé Publique, Assistance Publique des Hôpitaux Marseille et Université Aix-Marseille, France

Ferrata Storti Foundation

Haematologica 2018 Volume 103(7):1143-1149

1

Correspondence: isabelle.thuret@ap-hm.fr

ABSTRACT

I

n this retrospective study, we evaluate long-term complications in nearly all β-thalassemia-major patients who successfully received allogeneic hematopoietic stem cell transplantation in France. Ninety-nine patients were analyzed with a median age of 5.9 years at transplantation. The median duration of clinical follow up was 12 years. All conditioning regimens were myeloablative, most were based on busulfan combined with cyclophosphamide, and more than 90% of patients underwent a transplant from a matched sibling donor. After transplantation, 11% of patients developed thyroid dysfunction, 5% diabetes, and 2% heart failure. Hypogonadism was present in 56% of females and 14% of males. Female patients who went on to normal puberty after transplant were significantly younger at transplantation than those who experienced delayed puberty (median age 2.5 vs. 8.7 years). Fertility was preserved in 9 of 27 females aged 20 years or older and 2 other patients became pregnant following oocyte donation. In addition to patient’s age and higher serum ferritin levels at transplantation, time elapsed since transplant was significantly associated with decreased height growth in multivariate analysis. Weight growth increased after transplantation particularly in females, 36% of adults being overweight at last evaluation. A comprehensive long-term monitoring, especially of endocrine late effects, is required after hematopoietic stem cell transplantation for thalassemia.

haematologica | 2018; 103(7)

Received: November 9, 2017. Accepted: March 23, 2018. Pre-published: March 29, 2018.

doi:10.3324/haematol.2017.183467 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/103/7/1143 ©2018 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


I. Rahal et al.

Introduction In β-thalassemia, absent or reduced synthesis of the βglobin chain results in ineffective erythropoiesis and peripheral hemolysis. Anemia of the most severe form of the disease, known as β-thalassemia major (β-TM) or transfusiondependent thalassemia, is treated with lifelong red blood cell transfusions associated with chelation therapy in order to limit chronic complications and premature deaths related to iron overload. If this conventional therapy has dramatically improved survival and quality of life of patients,1,2 allogeneic hematopoietic stem cell transplantation (HSCT) is, in clinical practice, the only curative treatment. Only recently, patients were treated with β-globin gene therapy using autologous hematopoietic stem cells (HSCs) modified by lentiviral vectors.3 Hematopoietic stem cell transplantation has been successfully performed over the last 30 years4-7 with current thalassemia-free survival rates of 80-90% in children transplanted with HLA-matched sibling donor (MSD) before the onset of complications related to their disease or to the supportive treatment.8,9 Thalassemia is a rare disease in France. HSCT results from 1985 to 2007 were reported for 108 β-TM patients with 87% survival rate and, for patients treated after 2004, 85% thalassemia-free survival.10 Hematopoietic stem cell transplantation potentially results in a better long-term quality of life than that observed in patients treated with regular transfusion and chelation therapy.11,12 However, β-TM patients are exposed to late transplant-related complications particularly when transplant is performed in older children, adolescents or adults and in patients who have received inadequate chelation therapy before HSCT.12-14 Occurrence of late hepatic, endocrine and cardiovascular complications have been described, related to past and residual iron overload (IO) as well as conditioning toxicity, viral infections and chronic graft-versus-host disease (GvHD). Few studies have analyzed the long-term health status after HSCT including fertility in thalassemia patients. The present report includes almost all patients with βTM who successfully received an allogeneic HSCT in France between 1985 and 2012 and were alive at least two years after HSCT. The aim of this national retrospective study was to evaluate over time the long-term outcomes in β-TM patients after allogeneic HSCT using post-transplant medical examination data, long-term treatment records, and laboratory test results.

effects. Neither graft failure nor death occurred after two years post transplant.

Patients’ characteristics at HSCT The clinical characteristics of the 99 patients analyzed are reported in Table 1. Age at HSCT ranged from 8 months to 26 years old (median age 5.9 years). No patient had diabetes or thyroid dysfunction. Only one patient who was transplanted at 19 years of age had pre-existing IO-related cardiomyopathy. One male and 2 females were treated for hypogonadism. The median duration of clinical follow up after transplantation was 11.9 years (range 2-30 years).

Transplantation procedure All conditioning regimens were myeloablative, for the most part based on busulfan combined with cyclophosphamide (BuCy). In the first years of the program, 3 patients received irradiation (Table 1). Six patients underwent a second allogeneic HSCT after a median time of 2.8 years after the first transplantation due to graft failure. Ninety-one percent of transplants were from HLA-MSD. Sixty-seven patients received anti-thymocyte globulin as part of conditioning. All patients received cyclosporine A as GvHD prophylaxis, combined with methotrexate in 58 patients. Grade II-IV acute GvHD occurred in 22 patients. Chronic GvHD occurred within two years post transplant in 14 patients (limited in 9, extensive in 5). Within two years after successful HSCT, immunosuppressive treatment was stopped in 94 patients (median time dura-

Table 1. Patients’ and HSCT characteristics.

Number of patients

99

Male/female 45/ 54 Age at transplantation, years, median (IQR) 5.9 (3.1-11.2) Age at last assessment, years, median (IQR) 20 (14.2-28.3) Follow-up duration, years, median (IQR) 11.9 (7-19.3) Serum ferritin level before HSCT µg/L, median (IQR) 1400 (835-2250) Splenectomy 30 Height SDS at transplantation, median (IQR) -0.2 (-1.66, 0.29) Weight SDS at transplantation, median (IQR) - 0.18 (-1.62, 0.88) Pesaro classification (<18 years) - Class 1/Class 1 or 2/Class 2 30/15/39 - Class 2 or 3/ Class 3 4/3

Puberty at transplantation in female patients, n - Ongoing puberty or post-pubertal - Delayed puberty and aged ≥ 13 years - Pre-pubertal

54 6 2+2* 44

Type of donors, n

Methods This retrospective non-interventional study was approved by the national regulatory authorities (CCTIRS ref. 13.425 / CNIL n. 2009-674) and was partly based on data collected in the national registry of β-TM patients.15 Between December 1985 and December 2012, 134 patients had received an allogeneic HSCT for β-TM in 21 French transplantation centers. Fifteen patients died within two years post transplant. Twelve patients resumed regular transfusions after graft failure and 107 of 134 patients were alive at least two years after successful HSCT. Six were not analyzed in the long-term study (e.g. because they had returned to their home country or were lost to follow up) and 2 died of chronic GvHD early in the third year post transplant. Finally, 99 patients were studied for transplant-related long-term 1144

- MSD/matched other related - URD

91/5 3

Source of stem cells, n - Bone marrow - Cord blood/cord blood + bone marrow - Peripheral blood stem cells

84 11/3 1

Conditioning regimen (MAC = 100%), n - Busulfan + cyclophosphamide 86 (oral=52) - Busulfan + fludarabine ± thiotepa 10 - Others including irradiation (5 or 6 Gy thoracoabdominal 3 or total body irradiation). HSCT: hematopoietic stem cell transplantation; IQR: interquartile range; SDS: standard deviation scores; MSD: matched sibling donor; URD: unrelated donor; MAC: myeloablative conditioning; n: number. *Sex hormone replacement for hypogonadism.

haematologica | 2018; 103(7)


Late effects after hematopoietic stem cell transplantation for thalassemia major

tion of 9 months). Forty-four patients underwent phlebotomy and/or received chelation therapy after transplant.

Definition of methods and end points Late effects data documented by physicians were collected through visits to reference or transplant centers. Collected data included medical examination results, long-term treatment, and laboratory tests (serum ferritin, creatinine and hormone levels). Measurements of height and weight were converted to standard deviation scores (SDS) using French references.16 Delay of puberty, hypogonadism, being overweight, hypothyroidism, heart failure, and diabetes were defined using standard criteria (see Online Supplementary Methods).

were treated with corticosteroids for GvHD (Table 2). One patient, with arrhythmia and cardiomyopathy before transplant, regained normal heart rhythm and function after HSCT. Two patients who received a single conditioning with BuCy (200 mg/Kg) at the age of 13 and 4 years developed cardiac insufficiency 84 and 116 months, respectively, after HSCT. The first patient, now aged 39 years, has a moderate cardiac insufficiency whereas the other, who experienced a more severe disease course, is still undergoing treatment at the age of 20 years (Table 2). Their serum ferritin levels at HSCT were 370 and 1510 mg/L, respectively. No cardiac MRI was available at onset of cardiac symptoms to allow investigation of a possible cardiac iron overload.

Statistical analysis Continuous variables were reported as mean±Standard Deviation (SD) or as median and interquartile range (IQR) for nonnormal distribution. Wilcoxon signed rank test was used to compare sample median. As repeated measurements were made on the same statistical units (several measurements for each patient), univariate and multivariate linear mixed-effects models were used.17 Those variables significantly associated with outcome and those that were marginally significant (P<0.10) in univariate analysis were included into multivariate analysis. For all analysis, a twotailed test was used; P<0.05 was considered significant. All statistical analysis was performed using IBM SPSS Statistics v.20 (IBM SPSS Inc., Chicago, IL, USA).

Results Thyroid, diabetes and heart Eleven patients (11%) developed thyroid complications after HSCT (Table 2). The spectrum of thyroid complications was broad. Seven of 11 patients with a median serum ferritin level at transplant of 1560 mg/L had subclinical or overt hypothyroidism; this was transient in 2 cases. Two patients developed nodules or cysts without biological abnormalities and 2 other patients an autoimmune thyroid disease. Only 3 of 90 patients who received a single transplant with no irradiation developed permanent hypothyroidism. No patient experienced thyroid carcinoma. Five patients (5%) had diabetes mellitus after transplantation; their median age at HSCT was 13.7 years (range 1.826) and median serum ferritin level 1085 mg/L. Two patients

Growth In multivariate analysis, older age at the time of transplantation and, to a lesser extent, higher serum ferritin levels inversely correlated to height SD scores after transplant (Online Supplementary Table S1). Patient’s sex was not found to affect height SDS evolution after transplantation. Height SDS also decreased with time (P<0.001). Forty-nine patients (30 females and 19 males) had reached their full-grown height at last follow up. The median SDS for final height was of -1.4 (range -3 to 1.3) in males and -1.1 (range -3 to 3) in females. The multivariate analysis revealed that, unlike height, weight SDS increased with time (P<0.001). This increase was more prominent in females compared to males (P=0.003) (Online Supplementary Table S1). At last follow up, 36% of the 49 adult patients (11 females and 7 males) were overweight [Body Mass Inedx (BMI) >25 kg/m2]. Four adult females were obese with a BMI of over 30 kg/ m2.

Pubertal development in females At last evaluation, 43 of 54 females were assessable for puberty. For 6 of 43 patients, puberty was reached or ongoing at HSCT: all had secondary amenorrhea after transplant and 5 had hypogonadism (hypergonadotropic in 4 patients). Four of 43 females had delayed puberty at HSCT: all of them subsequently developed hypogonadism. Thirty-three of 43 females were pre-pubertal at transplant. One third (12 of 33) experienced spontaneous and normal puberty after one HSCT performed at a median age of 2.5 years. Only one patient had hypergonadotropic

Table 2. Thyroid complications, diabetes and impaired cardiac function.

Number of patients

Median time HSCT-disease (months, range)

Conditioning

11 7** 1* 1

68 [11-164] 102 30

Bu-Cy, TLI (n=1) Bu-Cy Bu-Cy

2

n=1, 226; n=1, 276

TBI/Bu-Cy

Diabetes

5*

78 [3-249]

Impaired LVEF (≤ 50%)

2

n=1, 84; n=1, 116

Bu-Cy (n=4)° TBI (n=1) Bu-Cy

Thyroid complications Hypothyroidism Hashimoto disease Hyperthyroidism (Grave disease) Hemorrhagic pseudocysts /nodules

a/c GvHD

Treatment

0/0 L-Thyroxin (n=7) L-Thyroxin + Surgery (n=1) Carbimazole and L-Thyroxin Surgery (n=2) 2/1 0

Insulin (n=4) OAD (n=1) ACEI (n=1) Diuretics (n=2)

HSCT: hematopoietic stem cell transplantation; a/cGvHD: acute/chronic graft-versus-host disease; Bu-Cy: busulfan and cyclophosphamide; TBI: total body irradiation; TLI: total lymphocyte irradiation; LVEF: left ventricular ejection fraction; OAD: oral anti-diabetic; irr: irradiation; ACEI: angiotensin-converting enzyme inhibitor. *Second transplant for one patient. °One patient had received irradiation for extra medullary hematopoiesis before HSCT. **Second transplant for 2 patients.

haematologica | 2018; 103(7)

1145


I. Rahal et al.

hypogonadism. Delayed puberty was observed in 21 of 33 patients; most of these cases presented hypergonadotropic hypogonadism (Table 3). The patients who spontaneously started their puberty were significantly younger at transplant compared to those who had delayed puberty [median age 2.5 years (range 1.6-5.9) vs. 8.7 years (range 1.7-12), respectively; P<10-3].

Pubertal development in males Twenty-nine of 45 males were assessed for puberty. Five of 29 males were post-pubertal at HSCT: one patient who had required hormonal replacement therapy before HSCT remained on treatment after HSCT for hypergonadotropic hypogonadism. The 2 of 29 males with ongoing puberty at transplant completed normal puberty. Among the 22 males who were pre-pubertal at transplant, only 4 (18%) had delayed puberty and 3 developed hypogonadism after transplant (hypogonadotropic in 2 cases). Eighteen of 22 patients (82%) transplanted at a median age of 5.9 years started puberty spontaneously.

Fertility and pregnancy Among the 27 females aged 20 years and over at last evaluation, 11 (40%) had had at least one successful pregnancy after transplant. Sixteen successful pregnancies were recorded with a median age of 26 years (22-33 years) at delivery. Two patients had benefited from oocyte donation; both had had delayed puberty and post-HSCT hypogonadism. Among the 9 remaining patients, 3 experienced normal puberty after HSCT and 6 had delayed puberty. It is worthy of note that 5 of 9 patients were diagnosed with hypergonadotropic hypogonadism. Among the 21 males aged over 20 years at last visit and evaluable for fatherhood, 4 (19%) fathered at least one child; 3 had experienced normal puberty and one patient had delayed puberty, hypergonadotropic hypogonadism and oligoasthenozoospermia. He and his partner had benefited from in vitro fertilization, which had resulted in a fullterm pregnancy and delivery.

recovered spontaneously. Five patients developed liver complications: 3 had liver fibrosis, one nodular regenerative hyperplasia, and one focal nodular hyperplasia; none of them developed hepatocellular carcinoma. At last visit, only 3 patients still had limited chronic GvHD that did not require any treatment, but another patient developed severe bronchiolitis obliterans. Two patients presented psychiatric disorders (one schizophrenia, one paranoia). No secondary malignancy was recorded. Creatinine levels (n=99) at a median time of 11 years after transplant were within the normal range for sex and age groups in all patients except for one 14-year old male patient with a chronic kidney disease stage 2 (96 mmoles/L). Another patient with diabetes developed a chronic proteinuria (2 gr/L) without renal insufficiency. Proteinuria was not routinely investigated after transplant in the study population.

Ongoing medication Half of the patients were on long-term treatment at last evaluation. Hormonal therapy (sex hormone replacement, thyroid hormone or insulin therapy) was prescribed for 34 patients, antibiotic therapy for 17, and cardiac treatment for 2. One patient with mixed chimerism was receiving longterm treatment with erythropoietin. The only patient receiving systemic immunosuppressive therapy at last evaluation was treated for auto-inflammatory arthritis.

Serum ferritin and hemoglobin levels Mean serum ferritin level at last evaluation was 405 mg/L±295. Thirty-seven patients were treated with phlebotomy, 7 with chelation therapy, and 11 with both. In multivariate analysis, serum ferritin levels after transplant significantly decreased with time and with the use of phlebotomy/iron chelation therapy. Serum ferritin levels after transplant were higher in older patients and/or in patients with high serum ferritin levels at HSCT (Online Supplementary Table S2). Median hemoglobin value at last evaluation was 125 g/L (range 86-170 g/L). All patients were free of transfusion, and only one patient received erythropoietin therapy.

Other complications Other relevant long-term late effects were encountered. Eleven patients had acquired hepatitis C virus (HCV) infection before transplant and had a positive HCV-RNA after HSCT. At last evaluation, 3 of 11 patients remained positive (2 of 3 did not require antiviral treatment), 7 of 11 became HCV-RNA negative after an antiviral treatment, and one

Discussion Nearly all β-TM patients successfully treated in France with allogeneic HSCT were assessed for late effects with a long follow up after transplantation (median duration of follow up 12 years). The vast majority of patients were trans-

Table 3. Gonadal dysfunction after hematopoietic stem cell transplantation (HSCT) in female patients.

Number Median age of patients at HSCT (years) Ongoing puberty or post-pubertal at HSCT Delayed puberty and aged ≥ 13 years at HSCT Pre-pubertal and aged < 13 years at HSCT Normal puberty Delayed puberty

Median age at assessment (years)

Spontaneous menarche

Secondary amenorrhea

Hypogonadism

Pregnancy (≥20 years)

6

19

31

-

6

4+1*

1 /6

4

13.6

26.5

0

-

2+2*

1°/2

2.5 8.7

18.5 21

12 5

3** 4

1 15

3 /5 5 + 1° /14

33 12 21

*Hypogonadotropic hypogonadism. °Oocyte donation. **Transient in 2 cases.

1146

haematologica | 2018; 103(7)


Late effects after hematopoietic stem cell transplantation for thalassemia major

planted early in childhood from MSD and all received myeloablative conditioning regimen (MAC), most often BuCy. At last evaluation, hypogonadism, defined as low estradiol levels or need for long-term sex hormone replacement therapy, was observed in 58% of female patients. Hypogonadism was hypergonadotropic in 84% of cases, the few cases of hypogonadotropic hypogonadism being observed in female patients who were post-pubertal or over 13 years at transplant. After transplant for thalassemia, ovarian failure has been reported with a frequency ranging from 50% to 100% (Table 4).18-25 Here, we report that gonadal dysfunction generally resulted from the busulfanrelated ovarian toxicity rather than IO which would lead to

hypogonadotropic hypogonadism. In several studies of βTM patients, older age at HSCT (>7 years) has been associated with more frequent post-transplant hypogonadism.14,20,22,24,25 This observation can be explained by the fact that the older the patient at HSCT, the higher the pretransplant exposure to IO, but also by a possible reduced gonadal toxicity to busulfan in very young children. The pool of oocytes is limited and decreases from birth,26 and pre-pubertal gonadal quiescence is gonadal-protective in children receiving chemotherapy.27 High-dose busulfanbased conditioning regimens are known to induce amenorrhea and elevated gonadotropin levels in almost all postmenarcheal women and at least 50% of pre-pubertal

Table 4. Review of the literature on long-term complications in β-thalassemia major (β-TM) transplanted patients.

Study reference

Uni/multicentric (number of patients)

Follow up, years [range]

End points

Results

Santarone39 2018 Santarone31 2017

Unicentric (122) Unicentric (75)

24 [4-34] 24 [10-33]

Cancer Pregnancy

Caocci38 2017

Multicentric (258)

11 [1-30]

Unicentric (40)

Cumulative incidence at 10 years

Multicentric (176)

7 [1-20]

Cancer Pregnancy Diabetes Growth Gonadal dysfunction Thyroid Pregnancy Cancer Growth Cardiac, renal

8 cases 40% women (n=15 including 2 after oocyte donation) 21% partners of male patients (n=8) 3 cases 6 women and 6 partners of male patients 2 cases Mean decrease in height SDS of -0.84* ° 55% gonadal dysfunction° 7.5% hypothyroidism One female patient No case Similar mean height and weight z-score > 4 years post HSCT/BL No cardiac complications, 20% renal complications (proteinuria/ elevated serum creatinine) 37% abnormal FSH or LH or testosterone level No case No case Similar mean height SDS after HSCT and at BL° +0.6 mean increase in weight SDS * 14.6% low FSH, LH, testosterone or estradiol ° 10% hypothyroidism 4 cases 5 cases of chronic renal failure 2 (liver and kidney) 14% women (n=6), 17% partners of male patients (n=11) 77% females and 48% males: raised basal FSH (+/- raised LH) or low estradiol/ testosterone response to hCG 9% diabetes 52% final height SDS < -2 11% cardiac events 80% females and 36% males (absence of pubertal development and elevated FSH/LH level) ° 11% hypothyroidism 2 cases 4 women and 2 partners of male patients Mean decrease in height SDS of 0.59* 100% females and 50% males (abnormal puberty) 1 case Growth rate deceleration after transplant. Median final height around -0.5 in females and -1 SDS in males 66% females and 37% males (abnormal puberty) One female patient

See25 2017

Chaudhury24 2017

Aldemir-Kocabas22 2014

La Nasa12 2013

Unicentric (41)

5.4 [2-10]

Multicentric (109)

22.8 [11-30]

Poomthavorn23 2013

Unicentric (47)

Khalil21 2012

Unicentric (47)

6 [1-10.6] 7 [2-11.6]

Di Bartolomeo11 2008

Unicentric (90)

15 [1-24]

Li19 2004

Unicentric (32)

5 [2-8.8]

De Sanctis18 2002

Unicentric (68)

3

Gonadal dysfunction Pregnancy Thyroid Growth Gonadal dysfunction Thyroid Cancer Kidney Organ transplant Pregnancy Gonadal dysfunction Diabetes Growth Heart Gonadal dysfunction Thyroid Cancer Pregnancy Growth Gonadal dysfunction Diabetes Growth Gonadal dysfunction Pregnancy

SDS: standard deviation scores; FSH: follicle stimulating hormone; LH: luteinizing hormone; hCG: human chorionic gonadotropin. *After hematopoietic stem cell transplantation (HSCT) compared to baseline at HSCT (BL). °Young age at HSCT is protective.

haematologica | 2018; 103(7)

1147


I. Rahal et al.

females.28-30 In our study, females who spontaneously started their puberty after transplant were significantly younger at transplant than those who experienced delayed puberty. In this report, the frequency of hypogonadism in male patients (low testosterone levels or long-term hormone replacement therapy) was 14%, less than that usually reported in transplanted thalassemia patients (Table 4). This could be underestimated since more sensitive criteria such as inhibin levels, gonadotrophin-releasing hormone tests or semen analysis were not available. A few pregnancies after HSCT for thalassemia have been reported but recently Santarone et al.31 described in a monocentric study 15 women who became pregnant after HSCT. We also observed that fertility was preserved in at least one-third of female patients aged 20 years and over. The proportions of females requiring ovarian stimulation or who tried to conceive without success are not known. Surprisingly, in our study, several females with delayed puberty and hypogonadism became pregnant. All women who became pregnant received oral busulfan (median dose 14-16 mg/kg), which is known to have a wide intra- and inter-patient pharmacokinetic variability. Consequently, fertility should be re-assessed according to the more recent procedure of use, i.e. intravenous busulfan. Fertility in male patients (n=4) was also partially preserved. We found that the rate of growth of β-TM patients was impaired after HSCT; height was influenced both by age and by serum ferritin levels at the time of transplant. These 2 variables reflect the impact of IO. This result is in agreement with most studies (Table 4),14,32 although others reported patients catching up in the first years following HSCT.5,19 Time elapsed after transplant also negatively influences the rate of growth, suggesting that a long-term follow up is necessary to assess the impact on height growth. We report a median loss from transplant of approximately one SD for patients reaching their full-grown height. Conditioning may contribute to growth delay in β-TM transplanted patients since a moderate decrease in height growth has also been observed in patients with hematologic malignancies receiving busulfan-based conditioning.33 Few data are available about weight development after HSCT in β-TM patients. In our study, weight SDS increased with time after transplant, especially in women. Being overweight appeared to be more frequent in thalassemia adult patients treated with HSCT (36% of patients at last evaluation) compared to those receiving conventional therapy (14.6% of adult patients; data from the French β-thalassemia registry, personal communication, 2017) or those receiving Bu-based conditioning for childhood leukemia.34 This result leads us to propose accurate investigation of the metabolic syndrome after HSCT in thalassemia patients. The frequency of hypothyroidism after HSCT ranged from 0% to 11% in thalassemia patients (Table 4). In this study, thyroid complications affected 11% of patients. It appeared to be mainly related to the cytotoxic effect of conditioning as half of these patients had received irradiation or a second transplant. Their age and serum ferritin levels at transplant were similar to those in the whole study population. It should be noted that hypothyroidism has been reported after treatment with BuCy.35,36

1148

We found a 5% rate of diabetes mellitus after transplant compared to 0-9% in previous studies on transplanted βTM patients.19,21,25 Potential risk factors for diabetes mellitus are: IO (because patients developing diabetes tended to be older at HSCT), conditioning therapy administration, use of corticosteroids for GvHD management. Cardiac complications have rarely been reported after a standard BuCy conditioning and are usually related to residual IO in β-TM patients.21,37 The presence of cardiac IO could not be ruled out in the 2 patients who developed cardiac dysfunction as cardiac MRI T2* was not then routinely used in France. Nonetheless, one patient transplanted at just four years of age with low serum ferritin values at and after HSCT developed cardiac failure. Three studies have reported several cases of cancer 10-25 years after HSCT for thalassemia, mostly cancer of the oral cavity and thyroid carcinoma.12,38,39 Recently, 8 cases of secondary solid cancer (SSC) have been reported among 112 patients, mostly children transplanted from an MSD.39 SSC occurred at a median time of 18 years after HSCT, stressing the need for a very long-term monitoring of TM survivors after HSCT. The length of follow up after HSCT may be too short in our study to record such cases. In addition, chronic GvHD, reported as an independent risk factor for secondary solid tumor,39 in our study affected only 3 patients with untreated limited disease at last evaluation. Monitoring of thyroid through regular ultrasound, treatment of HCV infection, and of residual IO may also contribute to the lack of secondary malignancy. Hepatic, psychiatric, or pulmonary complications were also observed in few patients. It is worth noting that HCV was successfully treated after transplant in nearly all patients with active infection. In summary, long-term complications were mainly related to the conditioning regimen in our study population where most patients were transplanted in the early phase of their disease. Pre-transplant ferritin levels were not elevated, and only few patients had IO-related clinical complications. Moreover residual IO was treated in 44% of cases by phlebotomy and/or iron chelation, these 2 modalities of treatment being efficient after HSCT.40-42 Reduced intensity or reduced toxicity conditioning based on treosulfan,43 fludarabine, and/or use of low doses of busulfan were investigated in β-TM patients.23,44 Long-term toxicity results of these studies are not yet available. In our report, Bu-based conditioning was myeloablative in all cases. Indeed, in our national experience, MAC was required in order to limit graft failure.8 In current gene therapy trials, conditioning with high doses of Bu also appeared necessary to allow corrected autologous cells to graft. Although not usually severe or life threatening, long-term effects after HSCT are frequent and diverse in TM patients, half of them undergoing long-term treatment, especially hormonal replacement. National and international guidelines describing comprehensive long-term monitoring should be established for thalassemia patients treated with HSCT. Acknowledgments This research was supported by AORC APHM 2011 (Appel d’Offre de Recherche Clinique).

haematologica | 2018; 103(7)


Late effects after hematopoietic stem cell transplantation for thalassemia major

References 1. Angelucci E, Barosi G, Camaschella C, et al. Italian Society of Hematology Practice Guidelines for the Management of Iron Overload in Thalassemia Major and Related Disorders. Haematologica. 2008;93(5):741752. 2. Rachmilewitz EA, Giardina PJ. How I treat thalassemia. Blood. 2011;118(13):3479–3488 3. Cavazzana M, Antoniani C, Miccio A. Gene therapy for -hemoglobinopathies. Mol Ther. 2017;25(3):1142-1154. 4. Thomas ED, Buckner CD, Sanders JE, Papayannopoulou T, et al. Marrow Transplantation for Thalassaemia. Lancet. 1982;2(8292):227-229. 5. Lucarelli G, Polchi P, Galimberti M, et al. Marrow Transplantation for Thalassaemia Following Busulphan and Cyclophosphamide. Lancet. 1985;1(8442): 1355-1357. 6. Lucarelli G, Galimberti M, Polchi P, et al. Marrow Transplantation in Patients with Thalassemia Responsive to Iron Chelation Therapy. N Engl J Med. 1993;329(12):840844. 7. Angelucci E. Hematopoietic stem cell transplantation in thalassemia. Hematology Am Soc Hematol Educ Program. 2010;2010:456462 8. Sabloff M, Chandy M, Wang Z, et al. HLAmatched sibling bone marrow transplantation for beta-thalassemia major. Blood. 2011;117(5):1745-1750. 9. Baronciani D, Angelucci E, Potschger U, et al. Hematopoietic stem cell transplantation in thalassemia: a report from the European Society for Blood and Bone Marrow Transplantation Hemoglobinopathy Registry, 2000-2010. Bone Marrow Transplant. 2016;51(4):536-541. 10. Galambrun C, Pondarré C, Bertrand Y, et al. French Multicenter 22-Year Experience in Stem Cell Transplantation for BetaThalassemia Major: Lessons and Future Directions. Biol Blood Marrow Transplant. 2013;19(1):62-68. 11. Di Bartolomeo P, Santarone S, Di Bartolomeo E, et al. Long-Term Results of Survival in Patients with Thalassemia Major Treated with Bone Marrow Transplantation. Am J Hematol. 2008;83(7):528-530. 12. La Nasa G, Caocci G, Efficace F, et al. LongTerm Health-Related Quality of Life Evaluated More than 20 Years after Hematopoietic Stem Cell Transplantation for Thalassemia. Blood. 2013;122(13):2262-2270. 13. De Simone M, Olioso P, Di Bartolomeo P, et al. Growth and Endocrine Function Following Bone Marrow Transplantation for Thalassemia. Bone Marrow Transplant. 1995;15(2):227-233. 14. Shenoy S, Angelucci E, Arnold SD, et al. Current Results and Future Research Priorities in Late Effects after Hematopoietic Stem Cell Transplantation (HCT) for Children with Sickle Cell Disease and Thalassemia: a Consensus Statement From the Second Pediatric Blood and Marrow Transplant Consortium International Conference on Late Effects after Pediatric HCT. Biol Blood Marrow Transplant. 2017;23(4):552-561. 15. Thuret I, Pondarré C, Loundou A, et al. Complications and treatment of patients with -thalassemia in France: results of the National Registry. Haematologica. 2010; 95(5): 724-729. 16. Sempé M, Capron JP. [“Chronos”: analysis of bone maturation by an automated numerical

haematologica | 2018; 103(7)

method]. Pediatrie. 1979;34(8):833-845. 17. Brown H, Prescott R. Applied Mixed Models in Medicine, 2nd Ed. John Wiley& Sons Ltd: Chichester, UK; 2006. 18. De Sanctis, V. Growth and Puberty and Its Management in Thalassaemia. Horm Res. 2002;58 (Suppl 1):72-79. 19. Li CK, Chik KW, Wong GW, Cheng PS, Lee V, Shing MM. Growth and Endocrine Function Following Bone Marrow Transplantation for Thalassemia Major. Pediatr Hematol Oncol. 2004;21(5):411-419. 20. Vlachopapadopoulou E, Kitra V, Peristeri J, et al. Gonadal Function of Young Patients with Beta-Thalassemia Following Bone Marrow Transplantation. J Pediatr Endocrinol Metab. 2005;18(5):477-483. 21. Khalil A, Zaidman, Elhasid R, Peretz-Nahum M, Futerman B, Ben-Arush M. Factors influencing outcome and incidence of late complications in children who underwent allogeneic Hematopoietic Stem Cell Transplantation for Hemoglobinopathy. Pediatr Hematol Oncol. 2012;29(8):694-703. 22. Aldemir-Kocaba B, Tezcan-Karasu G, Bircan I, Bircan O, Akta -Samur A, Ye ilipek MA. Evaluating the Patients with Thalassemia Major for Long-Term Endocrinological Complications after Bone Marrow Transplantation. Pediatr Hematol Oncol. 2014;31(7):616-623. 23. Poomthavorn P, Chawalitdamrong P, Hongeng S, et al. Gonadal Function of BetaThalassemics Following Stem Cell Transplantation Conditioned with Myeloablative and Reduced Intensity Regimens. J Pediatr Endorinol Metab. 2013;26(9-10):925-932. 24. Chaudhury S, Ayas M, Rosen C, et al. A Multicenter Retrospective Analysis Stressing the Importance of Long-Term Follow-Up after Hematopoietic Cell Transplantation for -Thalassemia. Biol Blood Marrow Transplant. 2017;10(10):1695-1700. 25. See WQ, Tung JY, Cheuk DK, et al. Endocrine complications in patients with transfusiondependent thalassaemia after haematopoietic stem cell transplantation. Bone Marrow Transplantat. 2018;53(3):356-360. 26. Baker TG. A quantitative and cytological study of germ cells in human ovaries. Proc R Soc Lond B Biol Sci. 1963;158:417-433. 27. Rivkees SA, Crawford JD. The Relationship of Gonadal Activity and ChemotherapyInduced Gonadal Damage. JAMA. 1988;259(14):2123-2125. 28- Afify Z, Shaw PJ, Clavano-Harding A, Cowell CT. Growth and Endocrine Function in Children with Acute Myeloid Leukaemia after Bone Marrow Transplantation Using Busulfan/Cyclophosphamide. Bone Marrow Transplant. 2000;25(10):1087-1092. 29. Allewelt H, El-Khorazaty J, Mendizabal A, et al. Late Effects after Umbilical Cord Blood Transplantation in Very Young Children after Busulfan-Based, Myeloablative Conditioning. Biol Blood Marrow Transplant. 2016;22(9):1627-1635. 30. Cho WK, Lee JW, Chung NG, Jung MH, et al. Primary Ovarian Dysfunction after Hematopoietic Stem Cell Transplantation during Childhood: Busulfan-Based Conditioning Is a Major Concern. J Pediatr Endocrinol Metab. 2011;24(11-12):10311035. 31. Santarone S, Natale A, Olioso P, et al. Pregnancy Outcome Following Hematopoietic Cell Transplantation for Thalassemia Major. Bone Marrow Transplantation 2017;52(3):388-393. 32. De Simone M, Verrotti A, Iughetti L, et al.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

Final Height of Thalassemic Patients Who Underwent Bone Marrow Transplantation during Childhood. Bone Marrow Transplant. 2001;28(2):201-205. Bernard F, Bordigoni P, Simeoni MC, et al. Height Growth during Adolescence and Final Height after Haematopoietic SCT for Childhood Acute Leukaemia: The Impact of a Conditioning Regimen with BU or TBI. Bone Marrow Transplant. 2009;43(8):637642. Bernard F, Auquier P, Herrmann I, et al. Health Status of Childhood Leukemia Survivors Who Received Hematopoietic Cell Transplantation after BU or TBI: An LEA Study. Bone Marrow Transplant. 2014;49(5):709-716. Michel G, Socié G, Gebhard F, et al. Late Effects of Allogeneic Bone Marrow Transplantation for Children with Acute Myeloblastic Leukemia in First Complete Remission: The Impact of Conditioning Regimen without Total-Body Irradiation--a Report from the Société Française de Greffe de Moelle. J Clin Oncol. 1997;15(6):22382246. Sanders JE, Hoffmeister PA, Woolfrey AE, et al. Thyroid function following hematopoietic cell transplantation in children: 30 years' experience. Blood. 2009;113(2):306-308. Mariotti E, Angelucci E, Agostini A, Baronciani D, Sqarbi E, Lucarelli G. Evaluation of cardiac status in iron-loaded thalassaemia patients following bone marrow transplantation: improvement in cardiac function during reduction in body iron burden. Br J Haematol. 1998;103(4):916-921. Caocci G, Orofino MG, Vacca A, et al. Longterm survival of beta thalassemia major patients treated with hematopoietic stem cell transplantation compared with survival with conventional treatment. Am J Hematol. 2017;92(12):1303-1310. Santarone S, Pepe A, Meloni A, et al. Secondary solid cancer following hematopoietic cell transplantation in patients with thalassemia major. Bone Marrow Transplant. 2018;53(1):39-43. Angelucci E, Muretto P, Lucarelli G, et al. Phlebotomy to reduce iron overload in patients cured of thalassemia by bone marrow transplantation. Italian Cooperative Group for Phlebotomy Treatment of Transplanted Thalassemia Patients. Blood. 1997;90(3):994-998. Angelucci E, Pilo F. Management of iron overload before, during, and after hematopoietic stem cell transplantation for thalassemia major. Ann NY Acad Sci. 2016;1368(1):115121. Inati A, Kahale M, Sbeiti N, et al. One-year results from a prospective randomized trial comparing phlebotomy with deferasirox for the treatment of iron overload in pediatric patients with thalassemia major following curative stem cell transplantation. Pediatr Blood Cancer. 2017;64(1):188-196. Bernardo ME, Piras E, Vacca A, et al. Allogeneic Hematopoietic Stem Cell Transplantationin Thalassemia Major: Results of a Reduced-Toxicity Conditioning Regimen Based on the Use of Treosulfan. Blood. 2012;120(2):473-476. Anurathapan U, Pakakasama S, Mekjaruskul P, et al. Outcomes of Thalassemia Patients Undergoing Hematopoietic Stem Cell Transplantation by Using a Standard Myeloablative versus a Novel ReducedToxicity Conditioning Regimen according to a New Risk Stratification. Biol Blood Marrow Transplant. 2014;20(12):2066-2071.

1149


ARTICLE

Bone Marrow Failure

Ferrata Storti Foundation

Circulating exosomal microRNAs in acquired aplastic anemia and myelodysplastic syndromes

Valentina Giudice,1 Lauren G. Banaszak,1 Fernanda Gutierrez-Rodrigues,1 Sachiko Kajigaya,1 Reema Panjwani,1 Maria del Pilar Fernandez Ibanez,1 Olga Rios,1 Christopher K. Bleck,2 Erin S. Stempinski,2 Diego Quinones Raffo,1 Danielle M. Townsley1 and Neal S. Young1

Haematologica 2018 Volume 103(7):1150-1159

Hematology Branch and 2Electron Microscopy Core Facility, National Heart, Lung, and Blood Institute (NHLBI), National Institutes of Health (NIH), Bethesda, MD, USA

1

ABSTRACT

E

Correspondence: valentina.giudice@nih.gov

Received: October 20, 2017. Accepted: April 18, 2018. Pre-published: April 19, 2018.

xosomal microRNAs modulate cancer cell metabolism and the immune response. Specific exosomal microRNAs have been reported to be reliable biomarkers of several solid and hematologic malignancies. We examined the possible diagnostic and prognostic values of exosomal microRNAs in two human bone marrow failure diseases: aplastic anemia and myelodysplastic syndromes. After screening 372 microRNAs in a discovery set (n=42) of plasma exosome samples, we constructed a customized PCR plate, including 42 microRNAs, for validation in a larger cohort (n=99). We identified 25 differentially expressed exosomal microRNAs uniquely or frequently present in aplastic anemia and/or myelodysplastic syndromes. These microRNAs could be related to intracellular functions, such as metabolism, cell survival, and proliferation. Clinical parameters and progression-free survival were correlated to microRNA expression levels in aplastic anemia and myelodysplastic syndrome patients before and after six months of immunosuppressive therapy. One microRNA, mir-126-5p, was negatively correlated with a response to therapy in aplastic anemia: patients with higher relative expression of miR-126-5p at diagnosis had the shortest progression-free survival compared to those with lower or normal levels. Our findings suggest utility of exosomal microRNAs in the differential diagnosis of bone marrow failure syndromes. (Registered at clinicaltrials.gov identifiers: 00260689, 00604201, 00378534, 01623167, 00001620, 00001397, 00217594).

doi:10.3324/haematol.2017.182824 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/103/7/1150 Š2018 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

Introduction The bone marrow (BM) failure syndromes are a heterogeneous group of hematologic diseases characterized by peripheral blood (PB) cytopenia due to hematopoietic stem and progenitor cell (HSPC) destruction and dysfunction, and/or constitutional syndromes due to genetic lesions.1-3 Blood count improvement after immunosuppressive therapies (IST) implicates autologous immunemediated HSPC destruction in aplastic anemia (AA).4 Myelodysplastic syndromes (MDS) are a heterogeneous group of clonal pre-malignant diseases characterized by ineffective hematopoiesis, progressive cytopenias, increased risk of developing acute myeloid leukemia (AML), and poor overall survival.2,5 However, approximately 10-15% of MDS cases share clinical and pathological features with AA, in a subtype known as hypocellular MDS; conversely, 10-15% of patients with BM failure ultimately develop secondary MDS after IST, suggesting that AA and MDS have common mechanisms in pathophysiology and disease progression.6,7 Despite the discovery of new molecular biomarkers, the differential diagnosis between BM failure syndromes and their prognosis is still challenging.8 In an effort to improve the definition of these hematologic disorders, circulating microRNAs (miRNAs, small non-coding RNA molecules of about 22 nucleotides), haematologica | 2018; 103(7)


Exosomal miRNAs in AA and MDS

have been described as potential biomarkers of AA and MDS.9,10 There is evidence that miRNAs are crucial in controlling and modulating immunity, promoting survival and growth of malignant cells, and in cancer metastasis.11,12 Some studies have demonstrated the presence of cancerspecific miRNAs in plasma or serum, suggesting utility of these molecules as potential biomarkers of diseases.13,14 Exosomes are small extracellular vesicles directly released into the extracellular space by all cell types;15 RNAs (protected from degradation16) and proteins are contained within these vesicles, allowing the transfer of genetic material and signaling proteins to recipient cells.17,18 Indeed, exosomal miRNAs influence biological functions, and they have been proposed as specific and reliable biomarkers of many malignant disorders, including colon cancer, melanoma, and AML, and to track minimal residual diseases in hematologic malignancies.19-27 In the current work, we investigated plasma exosomal miRNAs as potential minimally-invasive biomarkers of AA and MDS. Detection of specific exosomal miRNAs in these diseases might aid in understanding the pathophysiology of BM failure as well as diagnosis and prognosis.

until further use. To confirm that RNA was confined within exosomes, exosomes extracted from several samples were treated with RNase A, as described in the Online Supplementary Appendix.

Methods

Results

Human plasma samples

Exosomal miRNA content profiling shows a distinct signature in SAA and MDS

Whole PB was collected in ethylenediaminetetraacetic acid (EDTA) tubes from patients and healthy subjects after informed consent was obtained in accordance with the Declaration of Helsinki28 and protocols approved by the National Heart, Lung, and Blood Institute Institutional Review Board [National Institutes of Health (NIH), Bethesda, MD, USA]. All patients had a diagnosis of severe AA (SAA) or MDS according to standard criteria.29,30 Healthy controls were recruited from donors of the NIH Clinical Center Department of Transfusion Medicine. A discovery set (n=42), which was used for an initial screening of an exosomal miRNA signature, included 16 healthy controls, 16 SAA patients, and 10 MDS patients. For a validation set (n=99), 36 healthy controls, 54 SAA patients, and 20 MDS patients were recruited. Additionally, 10 patients from the discovery set and 30 SAA patients from the validation set were screened for exosomal miRNA expression before and after six months of IST. Clinical characteristics are summarized in Online Supplementary Table S1. Specimens were collected at the time of diagnosis and after six months of IST. After centrifugation at 2000 RPM for 10 min, plasma was collected and stored at -80°C until use.

Exosome extraction

Isolation of exosomes from 800 mL of plasma was performed using the PureExo Exosome Isolation kit (101Bio, Palo Alto, CA, USA) according to the manufacturer’s instructions. Flow-through was collected and stored at -80°C until further use. To confirm the presence of exosomes in the flow-through, protein content, particle size, CD63 expression, and transmission electron microscopy were assessed (Online Supplementary Appendix and Online Supplementary Figure S1).31

RNA extraction and cDNA synthesis To obtain high-quality RNA, a bead-based RNA purification protocol was performed using the Direct-zol™ -96 MagBead RNA kit (Zymo Research, Irvine, CA, USA) according to the manufacturer’s instructions with minor modifications. Purified RNA was subjected to cDNA synthesis with the miScript® II RT kit (Qiagen, Hilden, Germany), and undiluted cDNA stored at -20°C haematologica | 2018; 103(7)

miRNA profiling Initial screening of 372 miRNAs and 12 miRNA controls was performed in the discovery set using the miScript® miRNA PCR Array Human Serum & Plasma 384HC array (MIHS-106ZE, Qiagen) and the miScript® SYBR® Green PCR kit (Qiagen). Data were analyzed using the real-time thermal cycler 7900HT Fast Real-Time PCRs (Applied Biosystems, Thermo Fisher Scientific). A custom miScript miRNA PCR array (CMIHS02531E, Qiagen) including 42 candidate miRNAs and 6 controls was designed for validation.

Statistical analysis Data were analyzed using Prism (v.7.02; GraphPad software, La Jolla, CA, USA). miScript miRNA PCR Array Data Analysis software was utilized to analyze data from miRNA PCR arrays. Ingenuity Pathway Analysis (v.33559992, Qiagen Bioinformatics)32 and miRWalk 2.033 software were used for pathway and/or prediction analyses.

Previous studies investigated circulating miRNA expression levels in the plasma of AA and MDS patients,9,10 but no characteristic exosomal miRNAs have been described to date in these diseases. We first screened a large number of miRNAs (372 miRNAs associated with serum and plasma and 12 control miRNAs; miScript® miRNA PCR Array MIHS-3106Z; http://www.sabiosciences. com/genetable.php?pcatn=MIHS-3106Z) in plasma exosome samples in the discovery set by principal component analysis (PCA). We found different miRNAs in SAA (7 upand 4 down-regulated) and MDS (15 up- and 7 down-regulated) (Figure 1A and B). Exosomal miRNA expression profiles were compared between SAA and MDS, resulting in 42 miRNAs (21 and 21 up-regulated in MDS and SAA, respectively) (Figure 1C).

Validation of miRNA content profiles For validation of results obtained in the discovery set, 42 candidate miRNAs were selected and examined in a larger cohort of SAA, MDS, and healthy controls (n=99) (see Online Supplementary Table S1 for clinical characteristics and Online Supplementary Table S2 for selected miRNAs). To assess reaction quality and for data normalization, 6 control miRNAs (miRTC, PPC, SNORD61, miR-339-3p, miR-211-5p, and miR-30c-5p) were included in the custom PCR array plate, and analysis was performed as described in the Online Supplementary Appendix. PCA of the validation set revealed 9 miRNAs present in SAA (6 up- and 3 down-regulated) (Figure 2A). When MDS patients were compared to healthy controls, 15 and 5 miRNAs were significantly up-regulated and down-regulated, respectively (Figure 2B). In addition, MDS patients were compared to the SAA group, and 20 miRNAs were found to be differentially expressed (Figure 2C). Hierarchical clustering displayed differences in exosomal miRNA signatures between SAA and MDS (Figure 2D). Relative expression (Log2FC) values of 42 selected 1151


V. Giudice et al.

miRNAs were calculated for each sample; data were compared among SAA, MDS, and control groups by one-way ANOVA using the Kruskal-Wallis test and multiple comparisons by the two-stage linear step-up procedure of Benjamini, Krieger, and Yekutieli (Figure 3 and Online Supplementary Figure S2). miR-196a-5p and miR-196b-5p levels were higher in SAA (P=0.009 and P=0.002, respectively) and MDS (P<0.0001 and P=0.0003, respectively) compared to controls, but there were no differences between SAA and MDS groups (details of the analysis in

Online Supplementary Table S3). miR-4267, miR-19b-3p, and miR-1180-3p levels were lower in both SAA and MDS patients compared to controls. miR-378i was increased in both SAA and MDS, with higher levels in MDS. miR-5325p was increased in SAA (P=0.026) but not in MDS (P=0.098) compared to controls. Relative levels of 13 miRNAs were higher in MDS compared to controls and SAA patients, while miR-126-5p and miR-382-5p were lowest in MDS. miR-3200-3p was decreased in SAA and increased in MDS patients compared to controls. miR-

A

B

C

1152

Figure 1. miRNAs in severe aplastic anemia (SAA) and myelodysplastic syndromes (MDS) patients compared to healthy controls (HC) in the discovery set. Principal component analysis was performed to compare 384 exosomal miRNA expression levels: SAA versus HC (A), MDS versus HC (B), and MDS versus AA (C). Results are shown using volcano plots and tables. In volcano plots, x- and yaxes show estimated expression difference measured in Log2(FC) and the significance of the expression difference measured in -Log10(P-value), respectively. In the plots, horizontal and vertical lines indicate cut-off of significance (P<0.05) and expression levels greater than Âą1.5fold regulation (FR), respectively. For each comparison, miRNAs with Âą1.5 FR and P<0.1 are displayed in correspondent tables in which P<0.05 is highlighted in bold.

haematologica | 2018; 103(7)


Exosomal miRNAs in AA and MDS

423-3p, miR-1193, and miR-143-3p were higher in MDS patients relative to SAA and controls. Other miRNAs did not show variations in either SAA or MDS.

Association of exosomal miRNAs with diseases To assess specificity and sensitivity of exosomal miRNAs for the diagnosis of SAA and MDS, a receiver operating characteristic (ROC) curve analysis was employed using the validation set samples (Figure 3 and Online Supplementary Table S3). miR-196a-5b [area under the curve (AUC), 0.74], miR-196b-5p (AUC, 0.74), miR4267 (AUC, 0.71), miR-378i (AUC, 0.75), miR-19b-3p (AUC, 0.68), miR-1180-3p (AUC, 0.64), miR-423-3p (AUC, 0.63), miR-532-5p (AUC, 0.65), miR-574-3p (AUC, 0.64), and miR-3200-3p (AUC, 0.77) showed strong associations with SAA. Other miRNAs were not statistically significantly associated with the disease. In MDS patients, 21 exosomal miRNAs displayed strong association with the disease (Figure 3B). Strong association was observed in all 7 miRNAs present in both SAA and MDS patients with the highest AUC value (0.99) of miR-378i, followed by miR-574-3p (AUC, 0.87), miR196a-5p (AUC, 0.85), miR-3200-3p (AUC, 0.83), miR196b-5p (AUC, 0.79), miR-1180-3p (AUC, 0.79), miR-4267 (AUC, 0.74), and miR-19b-3p (AUC, 0.67).

Correlation with clinical parameters Pearson correlation analysis was used to investigate correlation of exosomal miRNAs with clinical parameters of hemoglobin (Hb), white blood cell (WBC) count, platelet

A

B

(Plt) count, absolute neutrophil count (ANC), absolute lymphocyte (ALC) and absolute reticulocyte counts (ARC), and lactate dehydrogenase (LDH) level in MDS patients and in SAA patients (before and after IST) from the validation set (Online Supplementary Table S4). Correlation with LDH was included in the analysis because higher LDH has been related to hemolysis and to disease progression in many malignant hematologic disorders.34 No association was found between any exosomal miRNA and Hb at diagnosis and after treatment. miR-5743p and miR-4274 were positively associated with WBC (r=-0.332, P=0.021) and Plt (r=0.330, P=0.021) counts at diagnosis, respectively, but not after IST. miR-15a-3p (r=0.312, P=0.026), miR-532-5p (r=0.368, P=0.010), and miR-26b-3p (r=0.364, P=0.017) positively correlated with an LDH level before therapy. miR-103a-3p and miR-29c3p positively correlated with ANC before IST, while miR126-5p negatively correlated after treatment (r=-0.326, P=0.040). miR-4651 was positively related to ARC before therapy, but no miRNAs were found after IST. However, SAA patients after treatment displayed negative correlations between ALC and several miRNAs: miR-3200-3p, miR-196a-5p, miR-28-3p, miR-133b, miR-26b-3p, let-7b5p, miR-133a-3p, and miR-106b-5p (Online Supplementary Table S4). In MDS, miR-1180-3p was positively correlated with a Hb level (r=0.483, P=0.036) and WBC count (r=0.561, P=0.013), while miR-3200-3p (r=0.963, P=0.002), miR-196b-5p (r=0.485, P=0.035), miR-378i (r=0.498, P=0.030), and miR-1260a (r=0.495, P=0.037) only to WBC count. No miRNAs showed a correlation with Plt count or

C

D

Figure 2. Validation of miRNA signatures in severe aplastic anemia (SAA) and myelodysplastic syndromes (MDS) patients. Principal component analysis was employed to compare the 48 miRNA expression levels in the validation set: SAA versus healthy controls (HC) (A), MDS versus HC (B), and MDS versus SAA (C). These results are shown with volcano plots in a similar manner as described in Figure 1. (D) Hierarchical clustering visualizes the 48 exosomal miRNAs in SAA, MDS, SAAresponders, SAA-non-responders, and HC. A red-green color scale indicates normalized miRNA expression levels (red: maximum; green: minimum).

haematologica | 2018; 103(7)

1153


V. Giudice et al.

ARC. miR-223-3p and miR-19b-3p were positively associated with an LDH level. miR-3200-3p was also positively associated with ANC (r=0.911, P=0.012), and miR-196a5p, miR-15a-3p, miR-133b, miR-106b-5p, and miR-4267 with ALC (Online Supplementary Table S4).

miR-126-5p as a candidate biomarker of response to IST in SAA We next sought to investigate effects of IST on exosomal miRNA profiles after six months of treatment in 40 SAA patients. PCA was performed to detect miRNAs in SAA patients before and after IST (Online Supplementary Figure S3). miR-143-3p (P=0.033), miR-324-3p (P=0.001),

miR-1180-3p (P=0.009), miR-126-5p (P=0.008), and miR382-5p (P=0.009) were significantly decreased after treatment (Online Supplementary Figure S4). Subsequently, SAA patients were classified into either SAA-responders (complete or partial response) or SAA-non-responders after six months of therapy based on standard clinical parameters, and miRNA expression profiles were then compared before and after IST. No variations were observed after IST for any miRNAs in SAA-non-responders. In SAAresponders, miR-4651 (P=0.015) and miR-126-5p (P=0.025) were significantly reduced after treatment (Figure 4A). However, only miR-126-5p (AUC, 0.79) displayed a strong association with the response to IST in

A

B

Figure 3. Exosomal miRNAs in severe aplastic anemia (SAA) and myelodysplastic syndromes (MDS) patients. Relative expression levels of the 48 miRNAs were calculated as Log2FC and shown for each group [SAA, MDS, and healthy controls (HC)]. (A) 7 exosomal miRNAs differentially expressed in both SAA and MDS compared to HC. (B) Receiver operating characteristic (ROC) curves for 4 miRNAs in AA and MDS. The ROC curve of the miRNA panel was generated based on the predicted probability for each patient and using the healthy group as a control. P<0.05 was considered statistically significant. AUC: area under the curve; CI: Confidence Interval.

1154

haematologica | 2018; 103(7)


Exosomal miRNAs in AA and MDS

SAA patients by ROC analysis (Figure 4B). When exosomal miRNA levels of SAA patients were compared at post treatment to those of healthy controls, miR-126-5p was significantly decreased in responders compared to healthy controls (Figure 4C). However, no differences were observed before therapy. SAA patients were classified based on fold regulation (FR) values before treatment, and progression-free survival (PFS) of each group was determined as shown in Figure 4D. When miRNA levels were undetermined, patients were removed from the analysis. There were no

A

significant variations in PFS based on relative expression of miR-196a-5p (P=0.239), miR-196b-5p (P=0.317), miR4267 (P=0.362), miR-378i (P=0.751), miR-19b-3p (P=0.093), miR-1180-3p (P=0.069), and miR-3200-3p (P=0.744). However, SAA patients with higher miR-1265p at diagnosis experienced decreased PFS (8.5 months; n=6; median follow up 6.9 months; range 3.2-25.3 months) compared to those with normal or lower levels (n=36 and n=11; median follow up 22.7 and 37.4 months; range 0.3-109.3 and 3-109.5 months, respectively; P=0.013) (Figure 4D).

B

C

D

Figure 4. Exosomal miRNAs and their correlation with prognosis in severe aplastic anemia (SAA) responders. Relative expression levels were calculated as Log2FC for all 48 miRNAs in SAA patients and compared before and after immunosuppressive therapies (IST). (A) Only miR-4651 and miR-126-5p were significantly decreased after treatment in SAA-responders. (B) For these 2 exosomal miRNAs, receiver operating characteristic (ROC) curves were generated as described in Figure 3 using a healthy group as a control. (C) Relative expression of differentially expressed exosomal miRNAs at diagnosis and after treatment in responders (R) and nonresponders (NR) of SAA patients were compared to healthy controls (HC) using one-way ANOVA with Kruskal-Wallis and the two-stage linear step-up procedure of Benjamini, Krieger and Yekutieli tests. Data are shown as mean+Standard Deviation (SD). (D) After calculation of progression-free survival (PFS), patients were divided into three groups according to miRNA relative expression shown as Log2FC at diagnosis of selected miRNAs. P<0.05 was considered statistically significant. AUC: area under the curve; FR: fold regulation.

haematologica | 2018; 103(7)

1155


V. Giudice et al.

Pathway analysis of exosomal miRNAs Using available software,35 increased or decreased miRNAs in each group (SAA, MDS, and SAA-responder patients) were interpolated (Figure 5). Venn diagrams displayed that miR-532-5p was unique in SAA, miR-4651

unique in SAA-responders, 14 miRNAs were unique in MDS, 7 miRNAs common in SAA and MDS, and miR126-5p was common in MDS and SAA-responders to IST (Figure 5A). Next, predicted target genes for each miRNA were identified using miRWalk 2.0 database, followed by

A

B

C

D

Figure 5. Pathway analysis using differentially expressed exosomal miRNAs. (A) VENNY (an interactive tool for comparing lists with Venn Diagrams) was used to find common or unique miRNAs among severe aplastic anemia (SAA), myelodysplastic syndromes (MDS), and SAA-responder patients. miRNAs classified into individual groups are listed accordingly. Red: increased exosomal miRNAs; blue: decreased exosomal miRNAs; miR-3200-3p is shown in black because of different expression profiles between SAA (down-regulated) and MDS (up-regulated). Predicted targeted genes of miRNAs exclusively expressed in SAA or MDS were used for pathway analysis by IPA software. Top 10 pathways in SAA (B) and the top 20 in MDS (C) are shown. (D) Venn diagram shows the number of unique or common pathways in SAA and MDS and a list of the 15 common signaling pathways.

1156

haematologica | 2018; 103(7)


Exosomal miRNAs in AA and MDS

filtering to extract genes that were exclusively present in SAA, MDS, or SAA-responders, or present in both MDS and SAA or SAA-responders (Figure 5B and C, Online Supplementary Table S5 and Online Supplementary Figure S5). Common genes targeted by the 7 shared miRNAs but not present in MDS were selected for analysis, and the top 20 pathways were reported (Figure 5D). miR-4651 and miR-126-5p were used for pathway analysis in SAAresponders, after removing the common targeted genes from MDS and SAA at diagnosis (Online Supplementary Figure S5C). The analysis revealed involvement of several intracellular functions, related to cell cycle, DNA damage response, intracellular signaling, and metabolic pathways.

Discussion We investigated exosomal miRNA profiles in SAA and MDS in order to find potential biomarkers for diagnosis and disease progression in BM failure syndromes. Based on screening of 372 miRNAs in the discovery set of plasma exosome samples, a custom miRNA PCR plate was designed, including 42 miRNAs for validation in a larger cohort. The analysis revealed 25 exosomal miRNAs that were uniquely or commonly present in SAA and/or MDS patients; they were involved in several biological functions, such as HSC differentiation. Recently published work from our laboratory describes circulating miRNAs using whole plasma samples of AA and MDS patients;9 however, no distinctive signatures have been reported for exosomal miRNAs to date. In our current study, we identified exosomal miRNAs exclusively present in SAA (miR-532-5p), MDS (14 miRNAs), and SAA responders to IST (miR-4651) patients, or common to SAA and MDS (7 miRNAs), and to SAA responders to IST and MDS (miR-126-5p). The miRNAs we identified have not been reported to be different in AA and/or MDS plasma samples. However, circulating miRNAs are composed of passively released nucleic acids from different cell types,15,36 and actively secreted in exosomes. Exosomal miRNAs are not randomly loaded into vesicles; rather, mature miRNAs are specifically sorted by diverse mechanisms and based on sequence, reflecting the tissue of origin.17,20,24 Because they are tissue-specific and stable under different conditions, exosomal miRNAs have been proposed as better biomarkers. Their significance in AA and MDS has not yet been examined. Plasma circulating miRNAs have been related to extracellular pro-inflammatory signaling pathways, such as Toll-like receptor or tumor necrosis factor Îą.9 Dysregulated exosomal miRNAs in our cohort were associated with numerous intracellular functions. Circulating miR-532-5p (unique in SAA) is related to response to chemotherapy in AML, and associated with decreased expression of Runt-related (RUNX) 3 protein in melanoma.37,38 miR-196a and miR-196b (present in our SAA and MDS cases) are differentially expressed in the long-term HSC compartment compared to more mature populations, and miR-196 induces the promotion of HSC differentiation through the inhibition of Homeobox (HOX) family members.39 miR-19b (decreased in our SAA and MDS cases) together with other miRNAs promotes the differentiation of progenitor cells into lymphoid precursors through inhibition of PTEN.39 Additionally, cytokines modulate miRNA expression, such as miR-196 haematologica | 2018; 103(7)

by interferons during hepatitis C infection.40 Based on this, increased exosomal miR-532-5p and miR-196, and decreased miR-19b in our cohort of SAA and MDS patients may indicate attempted expansion of the stem and progenitor cell compartments. miR-1180-3p has an anti-apoptotic function through inhibition of the BCL protein family;41 this miRNA was decreased in our SAA and MDS, patients more apoptotic in marrow with active disease. Other exosomal miRNAs present in our cohort do not have known functions in hematopoiesis or the immune response. There is increasing evidence of essential roles of circulating and exosomal miRNAs in modulating cancer cell metabolism, the immune response, and altering the local and distant tumor environments.12,17,42 miR-126-5p and its 3â&#x20AC;&#x2122; UTR counterpart are regulators of angiogenesis, cell metabolism, and glucose homeostasis, and they are down-regulated in many tumors. By targeting the PI3K/AKT/mTOR pathway, overexpression of miR-126 results in reduction of proliferation, driving malignant and normal HSCs into quiescence.43,44 Decreased exosomal miR-126 is not only related to increased proliferation of leukemic cells and differentiation44 but also decreases autophagy in cancer cells, leading to accumulation of damaged mitochondria and reactive oxygen species.44,45 In SAA, increased amino acid biosynthesis has been reported.46 It is plausible that miR-126-5p modulates metabolic programming during BM failure and recovery, especially as miR-126-5p was decreased in MDS and SAA responders to IST. Pathway analysis of predicted target genes using miRWalk 2.0 and IPA software revealed that downstream signaling could be regulated by exosomal miRNAs, as for MAPK, p38, JAK/STAT, and ERK. These pathways can be activated by many different extracellular stimuli, such as cytokines and growth factors, and exosomal miRNAs by interfering with this signaling could impair intracellular responses.12,17 However, miRNAs from circulating exosomes may be derived from different tissues, such as BM or T cells. Therefore, additional functional studies and exosomal miRNA profiling in BM samples will resolve the origin and biological functions of these miRNAs during active disease. Utilization of circulating and exosomal miRNAs in clinical practice is uncertain, mainly due to technical issues such as data normalization, exosome preparation, and RNA extraction.9,47-49 For circulating miRNAs, normalization is performed using small nucleolar RNAs (snoRNAs) and other methods such as GeNormPlus, NormFinder, and global mean of miRNA expression. For exosomal miRNAs, there are no clear guidelines for data normalization. RNU6 used for fresh samples for miRNA normalization is not a reliable endogenous control for frozen samples.47 The mean expression value is often applied for normalization47 but employing only one method could result in missing significant differences.48 In our study, we normalized our data using the geometric mean of 2 control miRNAs within each group: one snoRNA (SNORD61) and one miRNA (miRNA-211-5p) homogeneously expressed in the discovery set. This approach achieved high concordance across plates and for replicates. Ultracentrifugation remains the standard assay for extracellular vesicles isolation and enrichment, as no commercial kit achieves a high-yield exosome with more than 99% purity and therefore allows exclusion of lipoproteins. Treating extracted exosomes with RNase or proteinase K can 1157


V. Giudice et al.

increase the purity of samples, but there are no generally accepted guidelines for such manipulations.49 In consideration of these limitations, we utilized the PureExo Exosome Isolation kit for vesicle extraction because the ultra-centrifugation step can be eliminated and exosomes can be obtained from small volume biological samples. Biomarkers allow better evaluation of disease in general and improve prognosis, thus helping clinicians in the decision-making process. In SAA, age, sex, and pre-treatment PB counts are established as prognostically valuable markers for response to IST.50 In the current study, we observed positive correlation of some exosomal miRNAs with WBC and Plt counts, and LDH levels at diagnosis in SAA, but not with pre-treatment PB counts. These results suggest that their expression was not affected by transfusion history or disease severity, and that miRNAs could be used as independent diagnostic or prognostic markers. Similarly, positive correlation of miRNAs with clinical parameters was also seen in MDS. Exosomal miRNAs are available from PB, and they are considered reliable biomarkers because they are protected from degradation and specifically loaded into vesicles from proliferating or apoptotic cells. We propose further study of measurement of exosomal miRNAs in marrow

References 1. Young NS, Calado RT, Scheinberg P. Current concepts in the pathophysiology and treatment of aplastic anemia. Blood. 2006;108(8):2509-2519. 2. Fenaux P, Adès L. How we treat lower-risk myelodysplastic syndromes. Blood. 2013;121(21):4280-4286 3. 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. 4. Young NS. Hematopoietic cell destruction by immune mechanisms in aquired aplastic anemia. Semin Hematol. 2000;37(1):314. 5. Zeidan AM, Al Ali N, Barnard J, et al. Comparison of clinical outcomes and prognostic utility of risk stratification tools in patients with therapy-related vs. de novo myelodysplastic syndromes: A report on behalf of the MDS clinical research consortium. Leukemia. 2017;31(6):1391-1397. 6. Maciejewski JP, Selleri C. Evolution of clonal cytogenetic abnormalities in aplastic anemia. Leuk Lymphoma. 2004;45(3):433-440. 7. Serio B, Risitano A, Giudice V, Montuori N, Selleri C. Immunological derangement in hypocellular myelodysplastic syndromes. Transl Med UniSa. 2014;8:31-42. 8. DeZern AE, Sekeres MA. The challenging world of cytopenias: distinguishing myelodysplastic syndromes from other disorders of marrow failure. Oncologist. 2014;19(7):735-745. 9. Hosokawa K, Kajigaya S, Feng X, et al. A plasma microRNA signature as a biomarker for acquired aplastic anemia. Haematologica. 2017;102(1):69-78. 10. Butler JT, Abdelhamed S, Kurre P. Extracellular vesicles in the hematopoietic microenvironment. Haematologica.

1158

failure syndromes in general. We have identified different miRNA signatures in SAA and MDS, and a candidate biomarker of responsiveness to IST. However, a larger cohort of patients combined with an ultra-pure exosome isolation technique are needed to validate the clinical utility of exosomal miRNAs. Acknowledgments The authors would like to acknowledge the support of the Electron Microscopy Core of the National Heart, Lung, and Blood Institute (NHLBI) at National Institutes of Health (NIH, Bethesda, MD, USA). The authors would like to thank Xingmin Feng and Sawa Ito (Hematology Branch, NHLBI, NIH), Nastaran Rezaie and Anisha Kharkia (Qiagen, Germantown, MD, USA), and Leonid Margolis (National Institute of Child Health and Human Development, NIH, Bethesda, MD, USA) for technical assistance; and Kinneret Broder and Therese Intrater (Hematology Branch, NHLBI, NIH) for assistance in obtaining samples from healthy volunteers. Funding This research was supported by the Intramural Research Program of the NIH, National Heart, Lung, and Blood Institute.

2018;103(3):382-394. 11. Wang WT, Chen YQ. Circulating miRNAs in cancer: from detection to therapy. J Hematol Oncol. 2014;7:86. 12. Tomasetti M, Santarelli L, Neuzil J, Dong L. MicroRNA regulation of cancer metabolism: role in tumour suppression. Mitochondrion. 2014;19 Pt A:29-38. 13. Grasedieck S, Sorrentino A, Langer C, et al. Circulating microRNAs in hematological diseases: principles, challenges, and perspectives. Blood. 2013;121(25):4977-4984. 14. Wang J, Chen J, Sen S. MicroRNA as Biomarkers and Diagnostics. J Cell Physiol. 2016;231(1):25-30. 15. Théry C, Ostrowski M, Segura E. Membrane vesicles as conveyors of immune responses. Nat Rev Immunol. 2009;9(8):581-593. 16. Ge Q, Zhou Y, Lu J, Bai Y, Xie X, Lu Z. miRNA in plasma exosome is stable under different storage conditions. Molecules. 2014;19(2):1568-1575. 17. Tomasetti M, Lee W, Santarelli L, Neuzil J. Exosome-derived microRNAs in cancer metabolism: possible implications in cancer diagnostics and therapy. Exp Mol Med. 2017;49(1):e285. 18. Franquesa M, Hoogduijn MJ, Ripoll E, et al. Update on controls for isolation and quantification methodology of extracellular vesicles derived from adipose tissue mesenchymal stem cells. Front Immunol. 2014;5:525. 19. Saleem SN, Abdel-Mageed AB. Tumorderived exosomes in oncogenic reprogramming and cancer progression. Cell Mol Life Sci. 2015;72(1):1-10. 20. Boyiadzis M, Whiteside TL. Information transfer by exosomes: A new frontier in hematologic malignancies. Blood Rev. 2015;29(5):281-290. 21. Zhou J, Wang S, Sun K, Chng WJ. The emerging roles of exosomes in leukemo-

22.

23.

24.

25.

26.

27.

28.

29.

30.

geneis. Oncotarget. 2016;7(31):5069850707. Turpin D, Truchetet ME, Faustin B, et al. Role of extracellular vesicles in autoimmune diseases. Autoimmun Rev. 2016;15(2):174-183. Buzas EI, György B, Nagy G, Falus A, Gay S. Emerging role of extracellular vesicles in inflammatory diseases. Nat Rev Rheumatol. 2014;10(6):356-364. Wang J, De Veirman K, Faict S, et al. Multiple myeloma exosomes establish a favourable bone marrow microenvironment with enhanced angiogenesis and immunosuppression. J Pathol. 2016; 239(2):162-173. Matsumura T, Sugimachi K, Iinuma H, et al. Exosomal microRNA in serum is a novel biomarker of recurrence in human colorectal cancer. Br J Cancer. 2015;113(2):275-281. Pfeffer SR, Grossmann KF, Cassidy PB, et al. Detection of Exosomal miRNAs in the Plasma of Melanoma Patients. J Clin Med. 2015;4(12):2012-2027. Hornick NI, Huan J, Doron B, et al. Serum Exosome MicroRNA as a MinimallyInvasive Early Biomarker of AML. Sci Rep. 2015;5:11295. World Medical Association. World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects. JAMA. 2013;310(20):21912194. International Agranulocytosis and Aplastic Anaemia Study. Incidence of aplastic anemia: the relevance of diagnostic criteria. By the International Agranulocytosis and Aplastic Anemia Study. Blood. 1987;70(6):1718-1721. Greenberg PL, Stone RM, Al-Kali A, et al. Myelodysplastic Syndromes, Version 2.2017, NCCN Clinical Practice Guidelines in Oncology. J Natl Compr Canc Netw. 2017;15(1):60-87.

haematologica | 2018; 103(7)


Exosomal miRNAs in AA and MDS

31. Théry C, Amigorena S, Raposo G, Clayton A. Isolation and characterization of exosomes from cell culture supernatants and biological fluids. Curr Protoc Cell Biol. 2006;Chapter 3:Unit 3.22. 32. Krämer A, Green J, Pollard J Jr, Tugendreich S. Causal analysis approaches in Ingenuity Pathway Analysis. Bioinformatics. 2014; 30(4):523-530. 33. Dweep H, Gretz N. miRWalk2.0: a comprehensive atlas of microRNA-target interactions. Nat Methods. 2015;12(8):697. 34. Dumontet, C, Drai J, Bienvenu J, et al. Profiles and prognostic values of lactate dehydrogenase iso-enzymes in patients with non-hodgkin's lymphoma. Leukemia. 1999;13(5):811-817. 35. Oliveros JC. Venny. An interactive tool for comparing lists with Venn's diagrams. (2007-2015). Available from: http:// bioinfogp.cnb. csic.es/tools/venny/ index. html 36. Cortez MA, Bueso-Ramos C, Ferdin J, Lopez-Berestein G, Sood AK, Calin GA. MicroRNAs in body fluids--the mix of hormones and biomarkers. Nat Rev Clin Oncol. 2011;8(8):467-477. 37. Mosakhani N, Räty R, Tyybäkinoja A, Karjalainen-Lindsberg ML, Elonen E, Knuutila S. MicroRNA profiling in

haematologica | 2018; 103(7)

38.

39. 40.

41.

42.

43.

44.

chemoresistant and chemosensitive acute myeloid leukemia. Cytogenet Genome Res. 2013;141(4):272-276. Kitago M, Martinez SR, Nakamura T, Sim MS, Hoon DS. Regulation of RUNX3 tumor suppressor gene expression in cutaneous melanoma. Clin Cancer Res. 2009;15(9):2988-2994. Montagner S, Dehó L, Monticelli S. MicroRNAs in hematopoietic development. BMC Immunol. 2014;15:14. Pedersen IM, Cheng G, Wieland S, et al. Interferon modulation of cellular microRNAs as an antiviral mechanism. Nature. 2007;449(7164):919-922. Tan G, Wu L, Tan J, et al. MiR-1180 promotes apoptotic resistance to human hepatocellular carcinoma via activation of NF- B signaling pathway. Sci Rep. 2016;6:22328. Chan B, Manley J, Lee J, Singh SR. The emerging roles of microRNAs in cancer metabolism. Cancer Lett. 2015;356(2 Pt A):301-308. Ebrahimi F, Gopalan V, Smith RA, Lam AK. miR-126 in human cancers: clinical roles and current perspectives. Exp Mol Pathol. 2014;96(1):98-107. Lechman ER, Gentner B, Ng SW, et al. miR126 Regulates Distinct Self-Renewal

45.

46.

47.

48.

49.

50.

Outcomes in Normal and Malignant Hematopoietic Stem Cells. Cancer Cell. 2016;29(2):214-228. Tomasetti M, Monaco F, Manzella N, et al. MicroRNA-126 induces autophagy by altering cell metabolism in malignant mesothelioma. Oncotarget. 2016; 7(24):36338-36352. Zhong P, Zhang J, Cui X. Abnormal metabolites related to bone marrow failure in aplastic anemia patients. Genet Mol Res. 2015;14(4):13709-13718. Manier S, Liu CJ, Avet-Loiseau H, et al. Prognostic role of circulating exosomal miRNAs in multiple myeloma. Blood. 2017;129(17):2429-2436. Bockmeyer CL, Säuberlich K, Wittig J, et al. Comparison of different normalization strategies for the analysis of glomerular microRNAs in IgA nephropathy. Sci Rep. 2016;6:31992. Witwer KW, Buzás EI, Bemis LT, et al. Standardization of sample collection, isolation and analysis methods in extracellular vesicle research. J Extracell Vesicles. 2013;2. Narita A, Kojima S. Biomarkers for predicting clinical response to immunosuppressive therapy in aplastic anemia. Int J Hematol. 2016;104(2):153-158.

1159


ARTICLE

Myeloproliferative Disorders

Ferrata Storti Foundation

JAK2V617F-bearing vascular niche enhances malignant hematopoietic regeneration following radiation injury Chi Hua Sarah Lin,1* Yu Zhang,2* Kenneth Kaushansky3 and Huichun Zhan1,4

Department of Medicine, Stony Brook School of Medicine, NY, USA; 2Biopharmaceutical R&D Center, Chinese Academy of Medical Sciences & Peking Union Medical College, Suzhou, China; 3Office of the Sr. Vice President, Health Sciences, Stony Brook School of Medicine, NY, USA and 4Northport VA Medical Center, Northport, NY, USA 1

Haematologica 2018 Volume 103(7):1160-1168

*CHSL and YZ contributed equally to this work.

ABSTRACT

M

Correspondence: huichun.zhan@stonybrookmedicine.edu or Huichun.Zhan@va.gov Received: December 1, 2017. Accepted: March 14, 2018. Pre-published: March 22, 2018.

yeloproliferative neoplasms are clonal stem cell disorders characterized by hematopoietic stem/progenitor cell expansion. The acquired kinase mutation JAK2V617F plays a central role in these disorders. Abnormalities of the marrow microenvironment are beginning to be recognized as an important factor in the development of myeloproliferative neoplasms. Endothelial cells are an essential component of the hematopoietic vascular niche. Endothelial cells carrying the JAK2V617F mutation can be detected in patients with myeloproliferative neoplasms, suggesting that the mutant vascular niche is involved in the pathogenesis of these disorders. Here, using a transgenic mouse expressing JAK2V617F specifically in all hematopoietic cells (including hematopoietic stem/progenitor cells) and endothelial cells, we show that the JAK2V617F-mutant hematopoietic stem/progenitor cells are relatively protected by the JAK2V617F-bearing vascular niche from an otherwise lethal dose of irradiation during conditioning for stem cell transplantation. Gene expression analysis revealed that chemokine (C-X-C motif) ligand 12, epidermal growth factor, and pleiotrophin are up-regulated in irradiated JAK2V617F-bearing endothelial cells compared to wild-type cells. Our findings suggest that the mutant vascular niche may contribute to the high incidence of disease relapse in patients with myeloproliferative neoplasms following allogeneic stem cell transplantation, the only curative treatment for these disorders.

doi:10.3324/haematol.2017.185736

Introduction

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/103/7/1160

The chronic Philadelphia chromosome (Ph1) negative myeloproliferative neoplasms (MPNs) are clonal stem cell disorders characterized by hematopoietic stem/progenitor cell (HSPC) expansion and overproduction of mature blood cells. The acquired kinase mutation JAK2V617F plays a central role in MPNs. However, the mechanisms responsible for the malignant HSPC expansion in MPNs are not fully understood, limiting the effectiveness of current treatment. Although the etiology of dysregulated hematopoiesis has been mainly attributed to the molecular alterations within the HSPC compartment, abnormalities of the marrow microenvironment are beginning to be recognized as an important factor in the development of MPNs.1-5 The diseased niche could impair normal hematopoiesis and favor the competing malignant stem cells, which could contribute to the poor donor engraftment and high incidence of disease relapse following allogeneic stem cell transplantation (SCT), the only curative treatment for patients with MPNs.2,6-8 Endothelial cells (ECs) are an essential component of the hematopoietic niche and most HSPCs reside close to a marrow sinusoid (the â&#x20AC;&#x153;perivascular nicheâ&#x20AC;?).9 MPNs are characterized by increased marrow angiogenesis compared to normal marrow.10-12 Although the existence and cell of origin of endothelial progenitors is still a subject of debate, JAK2V617F mutation can be detected in endothelial progenitors derived from the hematopoietic lineage (the so-called endothelial cell

Š2018 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.

1160

haematologica | 2018; 103(7)


JAK2V617F-bearing vascular niche in MPNs

colony-forming units; CFU-ECs or Hill) and, in some reports, in the true endothelial colony-forming cells (ECFC) based on in vitro assays.13-17 JAK2V617F mutation is also present in isolated liver or spleen ECs from patients with MPNs.15,18 Previously, we and others have shown that JAK2V617F-bearing ECs are critical in the development of the bleeding abnormalities in a murine model of JAK2V617F-positive MPNs in which JAK2V617F is expressed in all hematopoietic cells and endothelial cells.19 In addition, we have reported that the JAK2V617F -bearing vascular niche promotes the expansion of the JAK2V617F HSPCs in preference to JAK2WT HSPCs.20,21 All of these observations suggest that ECs are involved in the pathogenesis of MPNs. In the present study, using the hematopoietic and endothelial specific Tie2-Cre system and different marrow transplantation models, we demonstrate that JAK2V617F-mutant HSPCs are relatively protected by the JAK2V617F-bearing vascular niche from the otherwise lethal irradiation administered during conditioning for marrow transplantation. Taken together, our studies indicate that the mutant vascular niche could contribute to the poor donor cell engraftment and the high incidence of disease relapse well known to occur in patients with MPNs after allogeneic SCT. Therefore, targeting the altered hematopoietic vascular niche could provide more effective therapies for patients with MPNs.

Methods Experimental mice JAK2V617F Flip-Flop (FF1) mice22 were provided by Radek Skoda (University Hospital, Basal, Switzerland) and Tie2-Cre mice23 by Mark Ginsberg (University of California, San Diego, USA). The FF1 mice were crossed with Tie2-Cre mice to express JAK2V617F specifically in hematopoietic cells and ECs (Tie2/FF1 mice). All mice used were crossed onto a C57BL/6 background and were bred in a pathogen-free mouse facility at Stony Brook University. CD45.1+ congenic mice (SJL) were purchased from Taconic Inc. (Albany, NY, USA). Animal experiments were performed in accordance with the guidelines provided by the Institutional Animal Care and Use Committee.

Stem cell transplantation assays The effects of the JAK2V617F-bearing vascular niche on MPN hematopoiesis were studied in vivo using marrow transplantation assays. First, we transplanted wild-type (WT) CD45.1 marrow cells into lethally irradiated (950cGy)24,25 8-14-week old Tie2/FF1 mice or WT controls (CD45.2). Peripheral blood was obtained every four weeks after transplantation, and CD45.1 donor chimerism and complete blood counts were measured. To study the effects of HSPC JAK2V617F mutation on HSPC radioprotection, we generated a chimeric murine model with JAK2V617F-mutant HSPCs and WT vascular niche by transplanting JAK2V617F marrow cells (CD45.2) into lethally irradiated (950cGy) WT recipients (CD45.1). The transplantation of CD45.2 WT marrow cells into CD45.1 WT recipients served as a control. Following hematopoietic recovery and full donor cell engraftment, each set of mice were irradiated with 300cGy to create a radiation injury. Two hours later, marrow Lineageneg (Lin-) HSPCs were isolated using Lineage Cell Depletion Kit (Miltenyi Biotec, San Diego, CA, USA) for evaluation of cellular apoptosis and cell cycle status. For competitive marrow transplantation experiments, 5x105 post-irradiated marrow cells (CD45.2) were injected intrahaematologica | 2018; 103(7)

venously together with 1x105 competitor CD45.1 WT marrow cells into lethally irradiated (950 cGy) CD45.1 recipients. Peripheral blood was obtained every four weeks after transplantation, and CD45.2 chimerism was measured. To study the effects of EC JAK2V617F mutation on HSPC radioprotection, we generated a chimeric murine model with WT HSPCs and JAK2V617F-bearing vascular niche by transplanting WT marrow cells (CD45.1) into lethally irradiated (950cGy) Tie2/FF1 recipients (CD45.2). The transplantation of CD45.1 WT marrow cells into CD45.2 WT recipients served as a control. Following hematopoietic recovery and full donor cell engraftment, each set of mice were irradiated with 300cGy to create a radiation injury. Two hours later, marrow Lineageneg (Lin-) HSPCs were isolated for evaluation of cellular apoptosis. Additional details of the methods used can be found in the Online Supplementary Methods.

Results Expression of JAK2V617F in Tie2+ cells protects marrow HSPCs from lethal irradiation Mice expressing Cre under the control of the Tie2 promoter (Tie2-Cre) were crossed with JAK2V617F Flip-Flop (FF1) mice to generate mice bearing human JAK2V617F expression specifically in endothelial and hematopoietic cells (Tie2/FF1). The Tie2/FF1 mice develop an MPN-like phenotype with neutrophilia, thrombocytosis, significant splenomegaly, and greatly increased marrow vascular density, megakaryopoiesis, and numbers of HSPCs.19,20 To investigate the effects of the JAK2V617F-bearing vascular niche on MPN hematopoiesis in vivo, WT CD45.1 marrow cells were transplanted directly into lethally irradiated (950cGy) Tie2/FF1 mice or age-matched littermate control mice (CD45.2) (n=12 in each group) (Figure 1A). During a 3-month follow up, while all WT control recipients displayed full donor engraftment, 7 of 12 (approx. 60%) Tie2/FF1 recipient mice displayed recovery of JAK2V617F-mutant hematopoiesis (mixed donor/recipient chimerism) ten weeks after transplantation (Figure 1B). We followed some of the Tie2/FF1 and WT control recipients (n=7 in each group) for more than eight months. In contrast to the Tie2/FF1 recipients with full donor engraftment, the mixed chimeric mice developed neutrophilia, thrombocytosis, and splenomegaly (Figure 1C and D), similar to what has been observed in the primary Tie2/FF1 mice.19,26 Flow cytometry analysis revealed that JAK2V617F-mutant CD45.2+EPCR+CD48â&#x20AC;&#x201C;CD150+ (ESLAM) cells, which is a highly purified long-term repopulating HSPC population in normal and in MPN marrow,27,28 are significantly expanded in the mixed chimeric mice compared to Tie2/FF1 recipients with full donor engraftment or WT recipients (Figure 1E). In virtually all our transplantation experiments performed over the past four years, we have used 950cGy radiation24 and have seen virtually 100% donor engraftment in every recipient. In contrast, 7 of 12 mice in our Tie2/FF1 recipients of normal marrow demonstrated mixed chimerism with an average of 23% recipient cells in peripheral blood at fourteen weeks following 950cGy irradiation and marrow transplantation, and developed an MPN phenotype resembling the primary Tie2/FF1 mice during more than eight months of follow up. These findings suggest that the JAK2V617F-mutant HSPCs in Tie2/FF1 mice are relatively protected from the otherwise 1161


C.H.S. Lin et al. A

B

C

D

F

E

G

Figure 1. JAK2V617F hematopoietic stem/progenitor cells (HSPCs) in the Tie2/FF1 mice are protected from lethal irradiation. (A) Scheme of direct marrow transplantation. (B) Peripheral blood CD45.1 chimerism following transplantation of wild-type (WT) CD45.1 marrow cells into lethally irradiated Tie2/FF1 mice or WT control mice (CD45.2) (n=12 in each group). (C) During more than eight months of follow up, Tie2/FF1 recipients with mixed chimerism developed both neutrophilia and thrombocytosis. (D) Spleens collected eight months following transplantation did not display significant weight differences between Tie2/FF1 recipients with full donor chimerism and WT recipients. In contrast, Tie2/FF1 recipients with mixed chimerism developed moderate splenomegaly compared to WT recipients (spleen weight 196 mg vs. 80 mg; P=0.010). (E) The frequency of normal marrow donor-derived E-SLAM (CD45.1) cells in the Tie2/FF1 recipients was unchanged from WT recipient mice. In contrast, JAK2V617F-mutant recipient-derived E-SLAM (CD45.2) cells were significantly expanded in Tie2/FF1 recipients with mixed chimerism. (F) Schematic diagram of irradiation and analysis of Tie2/FF1 and WT control mice. (G) Linâ&#x20AC;&#x201C; HSPC cell apoptosis in Tie2-cre control or Tie2/FF1 mice after 300cGy irradiation (n=2 in each group). *P<0.05.

1162

haematologica | 2018; 103(7)


JAK2V617F-bearing vascular niche in MPNs

A

B

C

D

Figure 2. The JAK2V617F-mutant hematopoietic stem/progenitor cell (HSPC) is more sensitive to radiation-induced apoptosis than JAK2WT HSPC. (A) Outline of experiment design to generate a chimeric murine model with JAK2V617F-mutant HSPCs and wild-type (WT) vascular niche. (B) After 300cGy irradiation, cell apoptosis (Annexin V+7AAD–) was significantly increased in JAK2V617F Lin– HSPCs (black bar) compared to JAK2WT Lin– HSPCs (gray bar). (C) Unirradiated JAK2V617F (black) Lin– HSPCs proliferated to a greater extent than JAK2WT (gray) Lin– HSPCs in vitro. (D) Peripheral blood donor chimerism following a competitive repopulation assay in which 5x105 post-irradiated JAK2WT or JAK2V617F marrow cells (CD45.2) were injected together with 1x105 competitor CD45.1 WT marrow cells into lethally irradiated CD45.1 WT recipients (n=4 in each group). *P<0.05.

lethal irradiation administered during conditioning for marrow transplantation. To confirm this hypothesis, we irradiated primary Tie2/FF1 mice or Tie2-cre control mice with 300cGy, and two hours later their marrow LinHSPCs were isolated for evaluation of cellular apoptosis (Figure 1F). We found that the JAK2V617F HSPCs in the mutant vascular niche (i.e. Tie2/FF1 mice) had significantly less cellular apoptosis compared to JAK2V617F HSPCs in WT vascular niche (i.e. control mice) (12.1% vs. 25.8%; P=0.043) (Figure 1G). Therefore, the JAK2V617F-mutant HSPCs in Tie2/FF1 mice are relatively protected from lethal irradiation, which could be responsible for the reported high incidence of disease relapse in patients undergoing allogeneic SCT for MPNs.2,6,7

The JAK2V617F-mutant HSPC is more sensitive to radiation-induced apoptosis than JAK2WT HSPC Tie2-Cre mice express Cre recombinase in both ECs and hematopoietic cells. To investigate whether the radioprotection phenotype noted in the prior experiments is due to the JAK2V617F mutation in Tie2/FF1 HSPCs, we generated a chimeric murine model with JAK2V617F-mutant HSPCs and a WT vascular niche by transplanting Tie2/FF1 marrow cells into WT recipients. The transplantation of WT marrow cells into WT recipients served as a control. Following hematopoietic recovery and full donor cell engraftment, each set of mice were again irradiated with 300cGy to create a radiation injury, and two hours later, marrow Lineageneg (Lin-) HSPCs were isolated for evaluation of cellular apoptosis and cell cycle status. In our previous study, recipient mice of Tie2/FF1 marrow developed a MPN phenotype by eight weeks post transplantation with significant thrombocytosis and neutrophilia.19 Therefore, we irradiated the mice at six weeks post transplant in this study before the development of any clinical phenotype (data not shown) (Figure 2A). We found that cell haematologica | 2018; 103(7)

apoptosis was significantly increased in the JAK2V617Fmutant HSPCs compared to JAK2WT HSPCs (47.3% vs. 20.3%; P=0.005) (Figure 2B). This result is consistent with the observation that unirradiated JAK2V617F Lin– HSPCs proliferate to a greater extent than JAK2WT Lin– HSPCs in serum-free medium in vitro (5.8-fold; P=0.000006) (Figure 2C), and therefore are predictably more sensitive to radiation-induced apoptosis. There was no significant difference in cell cycle status between the JAK2V617F HSPCs and JAK2WT HSPCs after irradiation. These data suggest that, in the WT vascular niche, the JAK2V617F-mutant HSPC is more (not less) sensitive to radiation-induced apoptosis than are JAK2WT HSPCs. To further test the effect of irradiation on HSPC function in the WT vascular niche, we performed a competitive repopulation assay in which 5x105 post-irradiated marrow cells (CD45.2 JAK2WT or CD45.2 JAK2V617F) were injected intravenously together with 1x105 competitor CD45.1 WT marrow cells into lethally irradiated (950 cGy) CD45.1 recipients (Figure 2A). Since the presence of JAK2V617F mutation in HSPCs may affect the cell’s longterm proliferation, we focused on donor cell chimerism in the early phase of engraftment. During an 8-week posttransplant follow up, there was no difference in CD45.2 donor chimerism between the recipients of post-irradiated JAK2V617F marrow cells and recipients of post-irradiated JAK2WT marrow cells, suggesting that the engraftment potential of post-irradiated JAK2V617F HSPCs (or at least the short-term HSPCs) do not differ from JAK2WT HSPCs (Figure 2D).

JAK2V617F-bearing ECs protect HSPCs from lethal irradiation We next studied the effects of an EC JAK2V617F mutation on hematopoietic radioprotection. Lin- marrow HSPCs were isolated from WT or Tie2/FF1 mice and cul1163


C.H.S. Lin et al. A

C

B

D

Figure 3. The JAK2V617F-bearing endothelial cells (ECs) protect JAK2V617F hematopoietic stem/progenitor cells (HSPCs) from lethal irradiation. (A and B) After 300cGy irradiation, there were higher total cell numbers (1.6-fold; P=0.026) (A) and hematopoietic progenitors (1.3-fold; P=0.010) (B) from JAK2V617F Lin– HSPCs cultured on JAK2V617F-bearing ECs compared to their being cultured on JAK2WT ECs. Cell numbers were shown as the relative ratio compared to JAK2WT Lin– HSPCs cultured on JAK2WT ECs under the same conditions. Data are from 3 independent experiments. (C) Outline of experiment design to generate a chimeric murine model with JAK2WT HSPCs and mutant vascular niche which is then subjected to sublethal irradiation. (D) CD45.1 WT Lin– HSPC cell apoptosis in WT or JAK2V617F-mutant vascular niche after 300cGy irradiation (n=4 in each group).

tured on primary EC feeder layers derived from WT or Tie2/FF1 (JAK2V617F) murine lungs. The Lin- HSPC-EC co-cultures were irradiated with 300cGy ex vivo and cell number was counted within 24 hours of irradiation. We observed higher total cell numbers (1.6-fold; P=0.026) and hematopoietic progenitors (1.3-fold; P=0.010) from JAK2V617F HSPCs cultured on JAK2V617F-bearing ECs compared to their being cultured on JAK2WT ECs, suggesting that the JAK2V617F-bearing vascular niche contributes directly to JAK2V617F-mutant HSPC radioprotection (Figure 3A and B). No significant difference was observed in cell numbers or hematopoietic progenitors between JAK2WT HSPCs cultured on JAK2V617F-bearing ECs and their being cultured on JAK2WT ECs in vitro after irradiation. To further investigate the effects of JAK2V617F-bearing vascular niche on the response of HSPCs to radiation injury in vivo, we generated a chimeric murine model with WT HSPCs and JAK2V617F-bearing vascular niche by transplanting WT marrow cells (CD45.1) into lethally irradiated (950cGy) Tie2/FF1 recipients (CD45.2). The transplantation of CD45.1 WT marrow cells into CD45.2 WT recipients served as a control (Figure 3C). Based on our observation that Tie2/FF1 recipients did not develop any significant recovery of the JAK2V617F-mutant hematopoiesis until ten weeks after transplantation (Figure 1B), each set of chimeric mice were irradiated with 300cGy at 6-10 weeks following transplantation to create 1164

a radiation injury. Two hours later, marrow Lin– HSPCs were isolated for evaluation of cellular apoptosis. We found that the WT Lin- HSPC cell apoptosis was decreased in the JAK2V617F-mutant vascular niche compared to WT vascular niche (12.7% vs. 19.7%; P=0.034) (Figure 3D). Taken together, these data suggest that the JAK2V617F-bearing vascular niche contributes directly to HSPC radioprotection. In contrast to our observations in vivo, where JAK2V617F-mutant HSPCs had increased apoptosis compared to JAK2WT HSPCs in the WT vascular niche (Figure 2A and B), cell numbers or progenitor numbers from the JAK2V617F HSPCs and JAK2WT HSPCs were similar when cultured on JAK2WT ECs (Figure 3A and B). Similarly, while the JAK2V617F-bearing vascular niche is protective for WT HSPCs in vivo (Figure 3C and D), JAK2V617F EC did not significantly protect WT HSPC in vitro (Figure 3A and B). These results are likely due to the different cell-cell interactions and niche factors between the in vitro culture condition and in vivo microenvironment.

The JAK2V617F mutation alters vascular niche function to contribute to HSPC radioprotection Next, we investigated how the JAK2V617F mutation alters EC function in the vascular niche to protect HSPCs from radiation injury. In our previous studies, we found that JAK2V617F-bearing ECs proliferate to a greater extent than JAK2WT ECs and display significantly increased haematologica | 2018; 103(7)


JAK2V617F-bearing vascular niche in MPNs

A

Figure 4. The JAK2V617F mutation alters vascular niche function to contribute to hematopoietic stem/progenitor cell (HSPC) radioprotection. (A) After 300cGy irradiation, cell apoptosis was higher in the JAK2WT endothelial cells (ECs) than JAK2V617F ECs. Data are from 1 of 3 independent experiments that gave similar results. (B and C) The expression levels of CXCL12, EGF and PTN in unirradiated (B) and irradiated (C) JAK2V617F-bearing ECs compared to irradiated JAK2WT ECs. Gene expression is shown as the relative fold-change compared with the JAK2WT EC expression which was set as 1. (D) Representative histogram plots and flow cytometry quantitative analysis of phosphorylated EGFR expression in irradiated JAK2WT HSPCs [from wild-type (WT) control mice] and JAK2V617F HSPCs (from Tie2/FF1 mice) (n=2 in each group).

B

C

D

angiogenesis in vitro compared to JAK2WT ECs. In addition, the tubular structures formed by the JAK2V617Fbearing ECs in vitro were more stable than those from JAK2WT ECs.21 In this study, we found that JAK2V617F lung ECs displayed less cell apoptosis in vitro after 300cGy irradiation compared to JAK2WT ECs (7.7% vs. 19.5%; P=0.026) (Figure 4A). It has long been known that hematopoietic recovery following lethal irradiation requires an intact vasculature.29-34 Therefore, the increased cell proliferation and/or decreased apoptosis could expand the vascular niche in JAK2V617F-bearing mice, which in turn contributes to the hematopoietic radioprotection we have observed in the Tie2/FF1 recipient mice. CXCL12 is an essential niche factor important for both HSPC maintenance and HSPC survival after radiation injury.35-38 Epidermal growth factor (EGF) and pleiotrophin (PTN), two other factors secreted by the vascular niche, haematologica | 2018; 103(7)

have been shown to play important roles in the regulation of HSPC regeneration following radiation injury. 39-41 Recently, we demonstrated that the expression level of CXCL12 was up-regulated in JAK2V617F-bearing marrow ECs compared to wild-type ECs, which could mediate the clonal expansion of JAK2V617F HSPCs, via the up-regulated CXCL12 receptor CXCR4, over JAK2WT HSPCs.20 To understand the EC signals responsible for HSPC radioprotection in the Tie2/FF1 recipient mice, we measured the expression levels of CXCL12, EGF, and PTN in both nonirradiated and irradiated JAK2WT and JAK2V617F lung ECs. qPCR analysis confirmed that there was upregulation of CXCL12 (2.5-fold; P=0.0001), EGF (4.0-fold; P=0.011) and PTN (11.4-fold; P=0.00001) in irradiated JAK2V617Fbearing ECs compared to irradiated JAK2WT ECs (Figure 4B and C). Furthermore, quantitative flow cytometry analysis showed that the proportion of marrow 1165


C.H.S. Lin et al.

CD150+CD48– HSPCs expressing phosphorylated EGFR was increased in irradiated JAK2V617F HSPCs (from Tie2/FF1 mice) as compared to irradiated JAK2WT HSPCs (from control mice) (2.7-fold; P=0.042), suggesting that EGFR signaling activity was up-regulated in irradiated JAK2V617F HSPCs (Figure 4D). These results suggest that the JAK2V617F-bearing vascular niche contributes directly to HSPC radioprotection, possibly by its elaboration of soluble niche factors.

Discussion Vascular ECs are a major component of the HSPC niche (the “vascular niche”) and provide many key factors that are required for HSPC maintenance.9 Patients with MPNs are characterized by increased marrow angiogenesis compared to normal marrow.10-12 ECs carrying the JAK2V617F mutation can be detected in patients with MPNs, suggesting that ECs are involved in the pathogenesis of MPNs.15,18 Here, by using the hematopoietic and endothelial specific Tie2-Cre system and different marrow transplantation models, we have been able to highlight the importance of JAK2V617F-bearing ECs in MPN disease relapse, which is seen in up to 40% of patients (especially after reduced intensity conditioning) following allogeneic SCT, the only curative treatment for MPNs.6,8,42-44 It has long been known that hematopoietic recovery following lethal irradiation requires an intact vasculature.29-34 Following radiation injury, co-culture of irradiated HSPCs with ECs can rescue HSPCs with multilineage reconstituting capacity.45,46 Our previous study has demonstrated that JAK2V617F-bearing ECs proliferate to a greater extent than JAK2WT ECs in vitro.21 In this study, we show that the JAK2V617F-bearing ECs display less cell apoptosis in vitro after irradiation compared to JAK2WT ECs. In addition, the JAK2V617F-mutant Lin– HSPCs produce more cells and hematopoietic colonies after irradiation when cultured on JAK2V617F-bearing ECs compared to their being cultured on JAK2WT ECs. Moreover, the expression levels of CXCL12, EGF, and PTN, which are important niche factors involved in HSPC maintenance and/or HSPC regeneration following radiation injury, were up-regulated in irradiated JAK2V617F-bearing ECs compared to JAK2WT ECs (Figures 3 and 4). These results suggest that the JAK2V617F-bearing vascular niche contribute directly to HSPC radioprotection. Consistent with these findings, 7 of 12 mice in our Tie2/FF1 recipients of normal marrow demonstrated mixed chimerism of an average 77% donor in peripheral blood cells at fourteen weeks following transplantation (Figure 1B). Previously we reported that the JAK2V617F-bearing vascular niche promotes the expansion of the JAK2V617F HSPCs in preference to JAK2WT HSPCs and the development of marrow fibrosis.20 Since graft failure or poor graft function in MPN patients after SCT is most likely due to marrow fibrosis,44 our work has demonstrated that the mutant vascular niche can contribute to the poor donor cell engraftment and the high incidence of disease relapse, the two major causes of treatment-related morbidity and mortality associated with allogeneic SCT in patients with MPNs.2,6,7,44 The Tie2-Cre mice express Cre recombinase in both ECs and hematopoietic cells. Although an EC-specific Cre (e.g. VEcadherin-Cre) would allow us to distinguish the specific role of ECs in HSPC radioprotection, we chose to 1166

use Tie2-Cre as it mimics the human MPNs in which both the HSPCs and ECs harbor the JAK2V617F mutation. In order to determine whether the radioprotection phenotype we have observed in the Tie2/FF1 mice is also due to the JAK2V617F mutation in HSPCs, we generated a chimeric murine model with JAK2V617F-mutant HSPCs and a WT vascular niche using marrow transplantation. We found that, in the WT vascular niche, the JAK2V617Fmutant Lin- HSPC is more (not less) sensitive to radiationinduced apoptosis than JAK2WT HSPCs. Although there have been reports that EC infusion could augment hematopoietic recovery following myeloablative injury, transplanted ECs exert their pro-regenerative effect transiently, and there is no evidence that donor marrow ECs could engraft and achieve long-term reconstitution in the recipient marrow vascular niche.31,32,47 Therefore, analysis six weeks post transplantation (Figure 2A and B) is unlikely to be affected by “carry-over” ECs from the Tie2/FF1 donor at the time of transplantation. In addition, the engraftment potential of irradiated JAK2V617F-mutant HSPCs does not differ from irradiated JAK2WT HSPCs (Figure 2). The results of these studies suggest that the radioprotection phenotype we have observed in the Tie2/FF1 recipients is unlikely to be due solely to the presence of the JAK2V617F mutation in HSPCs. In contrast, in another chimeric murine model with WT HSPCs and JAK2V617F-bearing vascular niche, WT Lin– HSPC is less sensitive to radiation-induced apoptosis in the JAK2V617F-mutant vascular niche compared to WT vascular niche (Figure 3). These results suggest that JAK2V617F-bearing vascular niche contributes directly to HSPC radioprotection. We could not exclude the possibility that altered interactions between the JAK2V617F HSPCs and JAK2V617F ECs contribute to HSPC radioprotection in the Tie2/FF1 mice. Indeed, no significant difference was observed in cell numbers or hematopoietic progenitors between JAK2WT HSPCs cultured on JAK2V617F-bearing ECs compared to their being cultured on JAK2WT ECs after irradiation (Figure 3A and B). This observation suggests that the JAK2V617F-bearing vascular niche by itself may not be sufficient to account for the radioprotection phenotype. Rather, it is most likely that specific cell-cell interactions involving the stem cells and niche ECs are required to provide the radioprotection of JAK2V617F HSPCs when present in a JAK2V617F vascular niche, as exemplified by the up-regulated EGF-EGFR signaling reported in this study (Figure 4). Systemic analysis of HSPC and EC proteins using either quantitative proteomics or antibodybased arrays, along with specific knock-out mouse models would be required to further investigate the interactions between HSPCs and ECs in JAK2V617F-bearing MPNs in vitro and in vivo. Although the JAK2V617F mutation has only been reported in liver and spleen ECs from patients with MPNs,15,18 it is very probably also present in their marrow ECs, considering that liver, spleen, and marrow are all hematopoietic organs during embryonic development and/or throughout adulthood. In most MPN patients, the stem cell compartment in MPN is heterogeneous with the presence of both JAK2 wild-type clones and the JAK2V617F mutant clones. We hypothesize that the vascular niche in MPN patients is also heterogeneous with the co-existence of both normal and mutant ECs. Since the JAK2V617F mutation is present in all HSPCs and ECs haematologica | 2018; 103(7)


JAK2V617F-bearing vascular niche in MPNs

from birth in the Tie2/FF1 mice, our murine model may not present the same acquired clonality and heterogeneous vascular niche features that characterize patients with MPNs. Nonetheless, our study has demonstrated that the JAK2V617F-bearing vascular niche can protect the JAK2V617F HSPCs from the otherwise lethal irradiation administered during conditioning for marrow transplantation, which provides a mechanism for the high incidence of disease relapse in MPN patients after allogeneic SCT. The optimal conditioning regimen for MPN patients undergoing SCT has still not been determined.44,48 As most current conditioning regimens for SCT are not restricted to only radiation, further investigation using murine models with different quantities of mutant ECs versus WT ECs will be required to fully understand the effects of the JAK2V617F-bearing vascular niche on mutant HSPC

References 1. Walkley CR, Olsen GH, Dworkin S, et al. A microenvironment-induced myeloproliferative syndrome caused by retinoic acid receptor gamma deficiency. Cell. 2007; 129(6):1097-1110. 2. Schepers K, Pietras EM, Reynaud D, et al. Myeloproliferative neoplasia remodels the endosteal bone marrow niche into a selfreinforcing leukemic niche. Cell Stem Cell. 2013;13(3):285-299. 3. Arranz L, Sanchez-Aguilera A, MartinPerez D, et al. Neuropathy of haematopoietic stem cell niche is essential for myeloproliferative neoplasms. Nature. 2014; 512(7512):78-81. 4. Mager LF, Riether C, Schurch CM, et al. IL33 signaling contributes to the pathogenesis of myeloproliferative neoplasms. J Clin Invest. 2015;125(7):2579-2591. 5. Walkley CR, Shea JM, Sims NA, Purton LE, Orkin SH. Rb regulates interactions between hematopoietic stem cells and their bone marrow microenvironment. Cell. 2007;129(6):1081-1095. 6. Kroger N, Holler E, Kobbe G, et al. Allogeneic stem cell transplantation after reduced-intensity conditioning in patients with myelofibrosis: a prospective, multicenter study of the Chronic Leukemia Working Party of the European Group for Blood and Marrow Transplantation. Blood. 2009;114(26):5264-5270. 7. Rondelli D, Goldberg JD, Isola L, et al. MPD-RC 101 prospective study of reduced-intensity allogeneic hematopoietic stem cell transplantation in patients with myelofibrosis. Blood. 2014;124(7):11831191. 8. 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. 9. Morrison SJ, Scadden DT. The bone marrow niche for haematopoietic stem cells. Nature. 2014;505(7483):327-334. 10. Medinger M, Skoda R, Gratwohl A, et al. Angiogenesis and vascular endothelial growth factor-/receptor expression in

haematologica | 2018; 103(7)

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

expansion and HSPC resistance to lethal irradiation and the cytotoxic chemotherapies commonly used in SCT for patients with MPNs. Acknowledgments The authors thank Todd Rueb and Rebecca Connor (Flow Cytometry Core Facility, Stony Brook University, NY, USA) for their assistance with the flow cytometry experiments. We would also like to thank Dr. Yupo Ma (Stony Brook University, NY, USA) for his continuous support throughout this work. YZ is supported by the State Scholarship Fund from Chinese Scholarship Council. Funding This research was supported by the Veterans Affairs award IK2BX001559 (HZ) and National Heart, Lung, and Blood Institute grant R01 HL134970 (HZ).

myeloproliferative neoplasms: correlation with clinical parameters and JAK2-V617F mutational status. Br J Haematol. 2009;146(2):150-157. Boveri E, Passamonti F, Rumi E, et al. Bone marrow microvessel density in chronic myeloproliferative disorders: a study of 115 patients with clinicopathological and molecular correlations. Br J Haematol. 2008;140(2):162-168. Gianelli U, Vener C, Raviele PR, et al. VEGF expression correlates with microvessel density in Philadelphia chromosome-negative chronic myeloproliferative disorders. Am J Clin Pathol. 2007;128(6):966-973. Yoder MC, Mead LE, Prater D, et al. Redefining endothelial progenitor cells via clonal analysis and hematopoietic stem/progenitor cell principals. Blood. 2007;109(5):1801-1809. Teofili L, Martini M, Iachininoto MG, et al. Endothelial progenitor cells are clonal and exhibit the JAK2(V617F) mutation in a subset of thrombotic patients with Ph-negative myeloproliferative neoplasms. Blood. 2011;117(9):2700-2707. Rosti V, Villani L, Riboni R, et al. Spleen endothelial cells from patients with myelofibrosis harbor the JAK2V617F mutation. Blood. 2013;121(2):360-368. Sozer S, Ishii T, Fiel MI, et al. Human CD34+ cells are capable of generating normal and JAK2V617F positive endothelial like cells in vivo. Blood Cells Mol Dis. 2009;43(3):304-312. Piaggio G, Rosti V, Corselli M, et al. Endothelial colony-forming cells from patients with chronic myeloproliferative disorders lack the disease-specific molecular clonality marker. Blood. 2009;114(14):3127-3130. Sozer S, Fiel MI, Schiano T, Xu M, Mascarenhas J, Hoffman R. The presence of JAK2V617F mutation in the liver endothelial cells of patients with BuddChiari syndrome. Blood. 2009; 113(21):5246-5249. Etheridge SL, Roh ME, Cosgrove ME, et al. JAK2V617F-positive endothelial cells contribute to clotting abnormalities in myeloproliferative neoplasms. Proc Natl Acad Sci USA. 2014;111(6):2295-2300. Zhan H, Lin CHS, Segal Y, Kaushansky K.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

The JAK2V617F-bearing vascular niche promotes clonal expansion in myeloproliferative neoplasms. Leukemia. 2018; 32(2):462-469. Lin CH, Kaushansky K, Zhan H. JAK2V617F-mutant vascular niche contributes to JAK2V617F clonal expansion in myeloproliferative neoplasms. Blood Cells Mol Dis. 2016;62:42-48. Tiedt R, Hao-Shen H, Sobas MA, et al. Ratio of mutant JAK2-V617F to wild-type Jak2 determines the MPD phenotypes in transgenic mice. Blood. 2008;111(8):39313940. Constien R, Forde A, Liliensiek B, et al. Characterization of a novel EGFP reporter mouse to monitor Cre recombination as demonstrated by a Tie2 Cre mouse line. Genesis. 2001;30(1):36-44. Duran-Struuck R, Dysko RC. Principles of bone marrow transplantation (BMT): providing optimal veterinary and husbandry care to irradiated mice in BMT studies. J Am Assoc Lab Anim Sci. 2009;48(1):11-22. Grahn D, Hamilton KF. Genetic Variation in the Acute Lethal Response of Four Inbred Mouse Strains to Whole Body XIrradiation. Genetics. 1957;42(3):189-198. Sangkhae V, Etheridge SL, Kaushansky K, Hitchcock IS. The thrombopoietin receptor, MPL, is critical for development of a JAK2V617F-induced myeloproliferative neoplasm. Blood. 2014;124(26):3956-3963. Kiel MJ, Yilmaz OH, Iwashita T, Yilmaz OH, Terhorst C, Morrison SJ. SLAM family receptors distinguish hematopoietic stem and progenitor cells and reveal endothelial niches for stem cells. Cell. 2005;121 (7):1109-1121. Kent DG, Copley MR, Benz C, et al. Prospective isolation and molecular characterization of hematopoietic stem cells with durable self-renewal potential. Blood. 2009;113(25):6342-6350. Hooper AT, Butler JM, Nolan DJ, et al. Engraftment and reconstitution of hematopoiesis is dependent on VEGFR2mediated regeneration of sinusoidal endothelial cells. Cell Stem Cell. 2009;4(3):263-274. Doan PL, Russell JL, Himburg HA, et al. Tie2(+) bone marrow endothelial cells regulate hematopoietic stem cell regeneration

1167


C.H.S. Lin et al.

31.

32.

33.

34.

35.

36.

1168

following radiation injury. Stem Cells. 2013;31(2):327-337. Chute JP, Muramoto GG, Salter AB, et al. Transplantation of vascular endothelial cells mediates the hematopoietic recovery and survival of lethally irradiated mice. Blood. 2007;109(6):2365-2372. Salter AB, Meadows SK, Muramoto GG, et al. Endothelial progenitor cell infusion induces hematopoietic stem cell reconstitution in vivo. Blood. 2009;113(9):2104-2107. Li B, Bailey AS, Jiang S, Liu B, Goldman DC, Fleming WH. Endothelial cells mediate the regeneration of hematopoietic stem cells. Stem Cell Res. 2010;4(1):17-24. Himburg HA, Sasine J, Yan X, Kan J, Dressman H, Chute JP. A Molecular Profile of the Endothelial Cell Response to Ionizing Radiation. Radiat Res. 2016; 186:141-52. Ding L, Saunders TL, Enikolopov G, Morrison SJ. Endothelial and perivascular cells maintain haematopoietic stem cells. Nature. 2012;481(7382):457-462. Ding L, Morrison SJ. Haematopoietic stem cells and early lymphoid progenitors occupy distinct bone marrow niches. Nature. 2013;495(7440):231-235.

37. Greenbaum A, Hsu YM, Day RB, et al. CXCL12 in early mesenchymal progenitors is required for haematopoietic stem-cell maintenance. Nature. 2013;495(7440):227230. 38. Foudi A, Jarrier P, Zhang Y, et al. Reduced retention of radioprotective hematopoietic cells within the bone marrow microenvironment in CXCR4-/- chimeric mice. Blood. 2006;107(6):2243-2251. 39. Doan PL, Himburg HA, Helms K, et al. Epidermal growth factor regulates hematopoietic regeneration after radiation injury. Nat Med. 2013;19(3):295-304. 40. Himburg HA, Muramoto GG, Daher P, et al. Pleiotrophin regulates the expansion and regeneration of hematopoietic stem cells. Nat Med. 2010;16(4):475-482. 41. Himburg HA, Doan PL, Quarmyne M, et al. Dickkopf-1 promotes hematopoietic regeneration via direct and niche-mediated mechanisms. Nat Med. 2017;23(1):91-99. 42. Guardiola P, Anderson JE, Bandini G, et al. Allogeneic stem cell transplantation for agnogenic myeloid metaplasia: a European Group for Blood and Marrow Transplantation, Societe Francaise de Greffe de Moelle, Gruppo Italiano per il

43.

44. 45.

46.

47.

48.

Trapianto del Midollo Osseo, and Fred Hutchinson Cancer Research Center Collaborative Study. Blood. 1999; 93(9):2831-2838. Ballen KK, Shrestha S, Sobocinski KA, et al. Outcome of transplantation for myelofibrosis. Biol Blood Marrow Transplant. 2010;16(3):358-367. Kroger N. Current Challenges in Stem Cell Transplantation in Myelofibrosis. Curr Hematol Malig Rep. 2015;10(4):344-350. Chute JP, Clark W, Saini A, Wells M, Harlan D. Rescue of hematopoietic stem cells following high-dose radiation injury using ex vivo culture on endothelial monolayers. Mil Med. 2002;167(2 Suppl):74-77. Chute JP, Fung J, Muramoto G, Erwin R. Ex vivo culture rescues hematopoietic stem cells with long-term repopulating capacity following harvest from lethally irradiated mice. Exp Hematol. 2004;32(3):308-317. Poulos MG, Crowley MJ, Gutkin MC, et al. Vascular Platform to Define Hematopoietic Stem Cell Factors and Enhance Regenerative Hematopoiesis. Stem Cell Reports. 2015;5(5):881-894. Spivak JL. Myeloproliferative Neoplasms. N Engl J Med. 2017;376(22):2168-2181.

haematologica | 2018; 103(7)


ARTICLE

Acute Myeloid Leukemia

The chromatin-remodeling factor CHD4 is required for maintenance of childhood acute myeloid leukemia

Ferrata Storti Foundation

Yaser Heshmati,1 Gözde Türköz,1 Aditya Harisankar,1 Shabnam Kharazi,2 Johan Boström,3 Esmat Kamali Dolatabadi,1 Aleksandra Krstic,2 David Chang,1 Robert Månsson,2,4 Mikael Altun,3 Hong Qian1 and Julian Walfridsson1

Center for Hematology and Regenerative Medicine, Department of Medicine; 2Center for Hematology and Regenerative Medicine, Department of Laboratory Medicine; 3 Research Division of Translational Medicine and Chemical Biology, Department of Medical Biochemistry and Biophysics, and 4Hematology Center, Karolinska Institutet, Karolinska University Hospital, Stockholm, Sweden 1

Haematologica 2018 Volume 103(7):1169-1181

ABSTRACT

E

pigenetic alterations contribute to leukemogenesis in childhood acute myeloid leukemia and therefore are of interest for potential therapeutic strategies. Herein, we performed large-scale ribonucleic acid interference screens using small hairpin ribonucleic acids in acute myeloid leukemia cells and non-transformed bone marrow cells to identify leukemia-specific dependencies. One of the target genes displaying the strongest effects on acute myeloid leukemia cell growth and less pronounced effects on nontransformed bone marrow cells, was the chromatin remodeling factor CHD4. Using ribonucleic acid interference and CRISPR-Cas9 approaches, we showed that CHD4 was essential for cell growth of leukemic cells in vitro and in vivo. Loss of function of CHD4 in acute myeloid leukemia cells caused an arrest in the G0 phase of the cell cycle as well as downregulation of MYC and its target genes involved in cell cycle progression. Importantly, we found that inhibition of CHD4 conferred anti-leukemic effects on primary childhood acute myeloid leukemia cells and prevented disease progression in a patient-derived xenograft model. Conversely, CHD4 was not required for growth of normal hematopoietic cells. Taken together, our results identified CHD4 as a potential therapeutic target in childhood acute myeloid leukemia.

Correspondence: julian.walfridsson@ki.se

Received: November 7, 2017. Accepted: March 23, 2018. Pre-published: March 29, 2018. doi:10.3324/haematol.2017.183970

Introduction Acute myeloid leukemia (AML) is a stem cell disease, characterized by rare leukemia-initiating cells (LICs) with increased self-renewal capacity that can propagate rapidly, growing immature myeloid blast cells with limited differentiation capacity.1,2 The LICs are largely resistant to chemotherapy and therefore many patients will ultimately relapse, which accounts for the leading cause of death in AML.3 The genetic, epigenetic and transcriptomic landscape in AML differs significantly between adults and children. Many of the causative lesions identified in adult AML (e.g., IDH1 and DNMT3A mutations) are rare events in childhood AML, whereas other gene mutations are more frequent in childhood AML (e.g., MYC, IKFZ1 and EZH2).4,5 The molecular differences between adult and childhood AML also include alterations in chromosomal copy number, translocations, different micro ribonucleic acid (miRNA) and messenger (m)RNA expression levels as well as epigenetic patterns.4 For example, translocations involving the MLL gene comprise 15% to 20% of all childhood acute myeloid leukemia (AML) cases. In contrast, only around 5% of adult AML patients carry MLL-rearrangements.6 These differences manifest as dissimilar biological characteristics, clinical behavior and different response to treatment.7 haematologica | 2018; 103(7)

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/103/7/1169 ©2018 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.

1169


Y. Heshmati et al.

The chromodomain helicase DNA-binding protein 4 (CHD4) is an adenosine triphosphate (ATP)-ase dependent chromatin remodeling factor and a component of the nucleosome remodeling and histone deacetylation (NuRD) complex and plays an important role in epigenetic transcriptional gene regulation.8 CHD4 has been linked to oncogenic processes, including control of cell cycle progression,9-12 cancer metastasis, epithelial-to-mesenchymal transition,13 and epigenetic suppression of tumor suppressor genes.13 Although the role of CHD4 in AML is largely unknown, inhibition of the chromatin remodeler has been reported to reduce AML tumor formation and sensitize AML cells to genotoxic drugs via the increased accessibility of DNA and impaired double strand break repair.14 A large number of functional screens have identified essential genes in various cancer cells,15 including AML.1627 However, AML-specific vulnerabilities have not been studied in detail. In the study herein, we performed loss of function screens on a large scale in AML cells and nontransformed bone marrow cells (BMs) in order to identify potential AML-specific vulnerabilities. CHD4 was identified as being required for cell growth and disease progression for primary childhood AML patient samples, but not for primary blood cells. Inhibition of CHD4 resulted in a downregulation of MYC and its target genes as well as a growth arrest in the G0 phase of the cell cycle.

Methods Cell growth assays of primary childhood AML samples The investigation was conducted in accordance with the ethical standards and according to the Declaration of Helsinki and to national and international guidelines, and has been approved by the authors' institutional review board. Culturing of the childhood samples was carried out as previously reported.28 MS-5 cells (DSMZ) were radiated at 80 Gy and plated at a density of 10,000 cells/well in MyeloCult media H5100 (STEMCELL Technologies Inc.) in a collagen I Cellware 96-well plate (Corning), two to three days before plating the cells. 10,000-20,000 cells suspended in MyeloCult media supplemented with recombinant human interleukin-6 (rhIL-6), recombinant human interleukin-3 (rhIL-3), recombinant human Fms-like tyrosine kinase 3/fetal liver kinase2 (rhFl3/Flk-2) ligand, recombinant human thrombopoietin (rhTPO), recombinant human stem-cell factor (rhSCF) and recombinant human granulocyte colony-stimulating factor (rhGCSF; STEMCELL Technologies Inc.) at a concentration of 20 ng/mL, were added to each well. The cells were maintained at normoxic conditions and effects in cell growth (LICs and leucocytes) were determined by flow cytometric analysis (see Online Supplementary Table S1 for antibodies).

Flow cytometric analysis and sorting Flow cytometric analysis was performed with a 4-laser BD LSRFortessa. Primary childhood AML cells were harvested and incubated in anti-CD16/32 (Fc-block) antibodies against mouse (Biolegend) and human (ChromPure Mouse IgG, Jackson ImmmunoResearch) for 20 minutes on ice. Then, the cells were stained with: human CD45, CD34, CD38 and lineage antibodies (CD20, CD4, CD8, CD2, CD56, CD235b, CD3 and CD19) and incubated on ice for 20 minutes (see Online Supplementary Table S1 for antibodies). Dead cells were excluded using the Near-IR Live/Dead marker (Invitrogen). Human CD45 positive cells were analyzed by a high-throughput automated plate reader (BD LSRFortessa). 1170

For the cell growth competition assays, cells were harvested and washed with cold phosphate-buffered saline (PBS) and thereafter stained with Near-IR Live/Dead marker in a 96-well plate in 80 ml of PBS and 2% fetal bovine serum (FBS). A highthroughput automated plate reader was used to detect the absolute number of live cells. To determine the level of engraftment of human AML cells in transplanted NOD scid g mice expressing human SCF, GM-CSF, and IL-3 (NSG-SGM3), BMs were isolated from the tibia and femur. The isolated BMs were incubated in mouse and human FC blocking antibodies for 20 minutes on ice. After washing, the cells were stained with human anti-CD45 on ice for 20 minutes. Then cells were incubated with Near-IR Live/Dead marker to detect live cells. Analysis was performed by FlowJo Version 9.3.3 software (Tree Star Inc.).

Results Identification of target genes selectively required for growth of MLL-AF9 rearranged AML cells by large-scale short hairpin RNA (shRNA) screening To identify novel target genes that are required for growth of AML cells expressing the MLL-AF9 fusion oncogene, we performed shRNA-based screens of two human AML cell lines (NOMO-1 cells derived from an adult, and THP-1 cells derived from a 1-year-old child) and a mouse AML cell line, all of which carried the MLLAF9 translocation. Non-transformed mouse FactorDependent Continuous Paterson Laboratories (Factor dependent cell-Paterson [FDCP]-mix) BMs were used as control cells in the screens.29 The RNA interference (RNAi) screening systems consisted of approximately 27000 lentiviral shRNAs targeting around 5400 putative disease-associated and drug targets (Cellecta Inc.). As outlined in Figure 1A, the barcoded lentiviral libraries were transduced as a pool into respective cell lines. The cells were harvested at an initial time point and after ten cell divisions. Next-generation sequencing of the polymerase chain reaction (NGS PCR) amplified barcodes from genomic DNA was followed by deconvolution and normalization of the data (Online Supplementary Table S2). To identify target genes that were selectively required for AML cell growth, we first determined the ratio of each individual barcode in the mouse AML cells compared to the mouse FDCP-mix cells after ten cell divisions. We identified 1082 target genes with at least a fivefold higher effect in cell growth in the mouse AML cells compared to the FDCP-mix cells (Figure 1B). By comparing these 1082 target genes with the corresponding human target genes of the human THP-1 and NOMO-1 cells (Figure 1C,D), we identified 34 shRNA-target genes that overlapped with all three screens (Figure 1E). Notably, 201 of the 289 total target genes from the screens of the two human AML cell lines overlapped (P=2.4E-215, hypergeometric test), indicating a conserved functional importance of these genes in AML maintenance (Figure 1E). Although most of the identified target genes had not been reported to have a role in AML, some have previously been linked to the disease, including MED12,30 USP7,31 FIP1L1,32 and SMC1A.33 Additionally, a significant proportion of the genes (15 of 34) were targeted by multiple shRNAs (Figure 1F). Gene ontology analysis revealed that the 34 target genes were associated with a broad haematologica | 2018; 103(7)


CHD4 is required for maintenance of childhood acute myeloid leukemia

range of functions and protein classes in different compartments of the cells (Online Supplementary Figure S1). Genetic alterations of the 34 target genes occurred frequently in various cancers (Online Supplementary Figure S2). In addition, TMPRSS3 (amplifications), ZFHX3 (deletions), P4HA2 (deletions) and SMC1A (missense mutations and amplifications) were found in more than 2% of AML patient samples.34 Remarkably, mRNA expression

was significantly correlated with the decreased overall survival of AML for eight of the 34 target genes.35 Moreover, 17 of the target genes were found to be markedly transcriptionally upregulated in MLLrearranged AML patient samples compared to healthy hematopoietic stem and progenitor cells (HSPCs).36 Thus, the data suggest that the screens were valid for identification of novel target genes in MLL-AF9 rearranged AML.

A

B

C

E

D

F

Figure 1. Large-scale shRNA-based RNAi screens identify novel target genes required for cell growth of AML cells. A. Schematic illustration of the shRNA-based screening procedure. B. Bar chart representing ratios of barcodes from 27104 individual shRNAs after screens in murine MLL-AF9 transformed AML cells compared to murine FDCP-mix BMs, after 10 cell divisions (T10). The 1086 barcodes with at least a five-fold lower representation in the murine MLL-AF9 AML cells compared to the murine FDCP-mix cells are highlighted in red. C and D. Bar charts representing differences in ratios of barcodes from 27491 individual shRNAs (THP-1) and 27104 (NOMO-1) after pooled loss of function screens after T10. Individual shRNAs targeting CHD4 are highlighted in red. E. Venn diagram illustrating the overlap between target genes that are five-fold more depleted in mouse AML cells compared to mouse FDCP-mix control cells, five-fold depleted target genes in THP-1 cells and NOMO-1 cells after T10. F. List of the 34 target genes overlapping in all three screens in the Venn diagram from Figure 1E together with information about the number of individual shRNAs resulting in at least a five-fold reduction in abundance of barcodes in the screens of the human AML and the mean fold reduction in abundance of barcodes in the screens after T10. FDCP: factor dependent cell-Paterson; AML: acute myeloid leukemia; PCR: polymerase chain reaction; shRNA: short hairpin RNA.

haematologica | 2018; 103(7)

1171


Y. Heshmati et al.

A

B

D

E

F

I

G

C

H

J

K

Figure 2. CHD4 inhibition impairs cell growth of leukemic cell lines and disease progression in vivo. A. Schematic drawing of the in vitro cell growth competition assay. B and F. Bar charts represent real time PCR analysis of mRNA levels after shRNA-based knockdown of CHD4, relative to control cells transduced with vectors expressing scrambled shRNA in MLL-rearranged THP-1, NOMO-1, and murine AML cells (Figure 2B), or in non-MLL rearranged HL-60, K-562 and NB-4 cells (Figure 2F), at 72 hours post transduction. Transduced cells were used for flow cytometry and growth competition assays (Figure 2D, E, H, I). mRNA levels were normalized to UBC. The data is represented as the mean ÂąS.E.M., ***P<0.005, ****P<0.001 (unpaired t-test), n=3. C and G. Western blot analyses of the endogenous CHD4 and Actin levels with and without knockdown of CHD4 in the indicated cells. D and H. Representative flow cytometry plots showing percentage of live RFP+ CHD4 knockdown cells in red (CHD4 shRNA-RFP) relative to GFP+ control cells (Sc shRNA-GFP), at the indicated days after transduction. E and I. Line charts depict the percentage of live RFP+ cells normalized to 100% at the initial time point (Day one) of the indicated cells determined by flow cytometry analysis at the indicated time points. J. Bar charts represent real time PCR analysis of mRNA levels after shRNA-based knockdown of Chd4 using two independent lentiviral vectors, relative to control cells transduced with vectors expressing scrambled shRNA in mouse AML cells that were subsequently transplanted into recipient mice for survival studies (Figure 2K) at 72 hours post transduction. mRNA levels were normalized to Hprt. The data is represented as the mean ÂąS.E.M., ****P<0.001 (unpaired t-test), n=3. K. Kaplan-Meier survival curves of wild-type (wt) C57BL/6 mice transplanted with 200,000 of murine MLL-AF9 transformed AML cells transduced with the two individual shRNA against CHD4, or the negative control vector. Percent survival and the number of animals for each cohort (time to euthanasia of moribund animals), were plotted against time, in days. Mean time survival of animals transplanted with control vector was 20 days, whereas it was 32 and 43 days for animals transplanted with cells transduced with the two vectors against CHD4 (***P<0.005). Seven animals were used for each experiment. AML: acute myeloid leukemia; shRNA: short hairpin RNA; Sc: scramble control; RFP: red fluorescent protein; GFP: green fluorescent protein.

1172

haematologica | 2018; 103(7)


CHD4 is required for maintenance of childhood acute myeloid leukemia

A

D

B

E

C

Figure 3. CHD4 inhibition causes minor effect in cell growth of normal blood cells in vitro. A. Bar charts represent real time PCR analysis of mRNA levels after shRNAbased knockdown of Chd4 relative to control cells transduced with vectors expressing scrambled shRNA, in sorted primary murine cKit+ BMs used in the competitive proliferation assays (Figure 3B,C), at 72 hours post transduction. mRNA levels were normalized to Hprt. The data is represented as the mean ±S.E.M., ****P<0.001 (unpaired t-test), n=3. B. Flow cytometry charts are representative plots of the percentage of live RFP+ Chd4 knockdown cKit+ cells (Chd4 shRNA-RFP) relative to GFP+ control cells (Sc shRNA-GFP) at the indicated time points. C. Line chart of a representative experiment of the percentage of Chd4 shRNA-RFP+ vs. Sc shRNA-GFP+ BMs propagated in suspension, normalized to 50% at the initial time point (Day one), determined by flow cytometry analysis at the indicated time points. D. Bar charts represent real time PCR analysis of mRNA levels after shRNA-based knockdown of CHD4, relative to control cells transduced with vectors expressing scrambled shRNA, in primary human CD34 enriched UCBs used in the cell growth assays (Figure 3E), 72 hours post transduction. mRNA levels were normalized to UBC. The data is represented as the mean ±S.E.M., ***P<0.005 (unpaired t-test), n=3. E. Bar chart of the total number of viable primary UCBs, transduced with shRNA against CHD4 (CHD4-shRNA), or a negative control vector (Sc-shRNA). The UCBs were propagated in suspension with supplemented cytokines and growth factors for 14 days. The number of viable UCBs was determined by flow cytometric analysis. **P<0.01 (unpaired t-test), n=3. mRNA: messenger ribonucleic acid; RFP: red fluorescent protein; GFP: green fluorescent protein; shRNA: short hairpin RNA.

CHD4 is required for cell growth of leukemia cells and disease progression in vivo In the screens, multiple shRNA vectors against CHD4 strongly inhibited expansion of the two human AML cell lines (Figure 1C,D). Moreover, three out of six shRNAs targeting CHD4 in the screens caused more pronounced reduction in cell growth in the murine AML cells compared to the FDCP-mix control cells (fold change: 0.19, 0.32, 0.38, respectively) (Online Supplementary Table S2). Both CHD4 and the NuRD complex have been reported to be required for cell growth of various types of cancer cells.9,13,14,37-39 Taken together with the fact that CHD4 is a potentially “druggable” enzyme, our studies proceeded to focus on this epigenetic factor. To validate the importance of CHD4 in AML, we performed a pairwise cell growth competition assay, allowing monitoring of control and CHD4 targeted cells under the same conditions (Figure 2A). shRNA-based inhibition using two different vectors resulted in efficient reduction in mRNA (Figure 2B; Online Supplementary Figure S3A) and protein levels (Figure 2C; Online Supplementary Figure S3B) of CHD4 as compared to control cells transduced with a negative control vector. Longitudinal flow cytometric analysis revealed that THP-1, NOMO-1 and mouse AML cells subjected to CHD4 knockdown by two independent shRNAs were strongly inhibited in cell growth compared to control cells (Figure 2D,E; Online Supplementary Figure S3C,D). Given the demonstrated importance in MLL-rearranged AML cells, we then investigated whether CHD4 was also required for cell growth of leukemic cells carrying differhaematologica | 2018; 103(7)

ent types of genetic lesions. Knockdown of CHD4 with two distinct shRNAs caused a significant reduction in mRNA (Figure 2F; Online Supplementary Figure S3E) as well as protein levels (Figure 2G; Online Supplementary Figure S3F), and prevented cell growth of the HL-60 (promyelocytic leukemia cells), K562 cells (chronic myelogenous leukemia) and NB-4 (acute promyelocytic leukemia), to a similar degree as that of the MLL-AF9 rearranged AML cells (Figure 2H,I; Online Supplementary Figure S3G,H). To investigate if CHD4 also has a role in AML disease progression, we used a murine immune competent transplantation model.40 As previously reported, wild-type (wt) mice that received transplants of MLL-AF9 transformed AML mouse cells, that were transduced with a negative control vector, resulted in pathologic and clinical manifestations of human AML within 19 to 21 days post transplantation.40 In contrast, wt recipient mice that received transplants of MLL-AF9 transformed AML cells that were transduced with two individual shRNA vectors, which mediated efficient suppression in CHD4 mRNA levels (Figure 2J), survived significantly longer than mice transplanted with control cells (Figure 2K).

CDH4 is not required for cell growth of normal hematopoietic cells To assess the importance of Chd4 in normal hematopoietic cells, we performed shRNA-based knockdown experiments in primary mouse BMs, which resulted in efficient reduction in Chd4 mRNA levels compared to the control cells (Figure 3A). Using the cell growth 1173


Y. Heshmati et al.

assay described in Figure 2A, dynamic flow cytometric analysis displayed minor differences in cell growth between the control cells and the Chd4 depleted cells (Figure 3B,C). Similarly, CHD4 suppressed CD34 positive umbilical cord blood cells (UCBs) (Figure 3D), showed a modest reduction in the number of viable cells compared to control cells, after 14 days of culturing

(Figure 3E), and as compared to the reduction in cell growth of CHD4 targeted AML cells (Figure 2; Online Supplementary Figure S3). Together, these results show that CHD4 is critical for cell growth of various types of leukemic cells and AML progression in vivo, but not for the proliferation and survival of normal primary hematopoietic cells.

A

B

E

H

C

D

F

G

I

Figure 4. CRISPR-Cas9 mediated disruption of CHD4 prevents growth of human AML cells. A. Schematic image depicting the homology of the two gRNAs used for targeting of the coding region of the human CHD4 gene. The protospacer adjacent motifs (PAM) are highlighted in blue letters and the gRNA target sequences are highlighted in bold black letters. B and F. Bar charts of real time PCR quantification of isolated genomic DNA from the targeted CHD4 gene and control locus after CRISPR-Cas9 disruption of CHD4 using two individual gRNAs, relative to the cells transduced with control vectors, in THP-1 Cas9 (Figure 4B) or MV4-11 Cas9 expressing cells (Figure 4F), at 72 hours post induction of gRNA expression. Real time PCR quantification of Cas9 targeted CHD4 loci was normalized to a control locus positioned upstream of the CHD4 gene. The transduced cells were used in the growth competition assays (Figure 4D,E,H,I). The data is represented as the mean ÂąS.E.M., ***P<0.005, ****P<0.001 (unpaired t-test), n=3. C and G. Western blot analyses and quantification of the endogenous CHD4 and Actin levels in the indicated cells, with and without knockdown of CHD4, at 72 hours post transduction. D and H. Representative flow cytometry charts of the percentage of live mCherry+ THP1 (Figure 4D) or MV4-11 cells (Figure 4H) expressing Cas9, transduced with CHD4 gRNAs expressing iRF670 (CHD4-gRNA), relative to control cells transduced with empty vectors expressing BFP (control gRNA), at the indicated days after induction of gRNA expression. E and I. Line charts of the relative ratio, in percentage, of THP-1 (Figure 4E) or MV4-11 cells (Figure 4I) transduced with CHD4 gRNA-iRFP670+ knockout constructs or BFP+ control vectors, at the indicated time points using flow cytometry analysis. gRNA: guide RNA; RFP: red fluorescent protein.

1174

haematologica | 2018; 103(7)


CHD4 is required for maintenance of childhood acute myeloid leukemia

A

B

C

D

E

F

Figure 5. CHD4 inhibition prevents growth of patient-derived childhood AML cells ex vivo and in vivo. A. Bar charts show real time PCR analysis of mRNA levels after shRNA-based knockdown of CHD4, relative to control cells transduced with vectors expressing scrambled shRNA, of the indicated primary childhood AML samples used in the proliferation assays (Figure 5B-D), at 72 hours post transduction. mRNA levels were normalized to UBC. The data is represented as the mean ±S.E.M., **P<0.01, ***P<0.005, ****P<0.001 (unpaired t-test), n=3. B. Flow cytometry charts of expression profiles of CD34, CD38, Lin, and CD45 of cells transduced with CHD4 shRNA expressing RFP (CHD4-shRNA) or a negative control vector expressing scrambled shRNA and GFP (Control shRNA). The patient-samples were cocultured with murine MS-5 stromal cells until the control cells reached confluence, or up to five weeks of maintenance. The number of viable CD45+ or LinCD34+CD38–, were determined by flow cytometric analysis. Live cell populations negative for Live/dead Near-IR dye was used for flow cytometric analysis. The numbers in each square represent the percentage of cells within each cell population. C and D. Bar chart showing the total numbers of live CD45+ (Figure C) and LinCD34+CD38– (Figure D) primary childhood AML cells transduced with CHD4 shRNA, or a control vector, as indicated. *P<0.05, **P<0.01, ***P<0.005, ****P<0.001 (unpaired t-test). E. Bar charts depict real time PCR analysis of mRNA levels after shRNA-based knockdown of CHD4, relative to control cells transduced with vectors expressing scrambled shRNA, of the indicated primary childhood AML samples used for the xenograft experiments (Figure F) at 72 hours post transduction. mRNA levels were normalized to UBC. The data is represented as the mean ±S.E.M., **P<0.01 (unpaired t-test), n=3. F. Scatter plot of percent of engrafted primary childhood AML cells of two childhood patient samples, transduced with shRNA against CHD4, or a negative control vector, that were transplanted into humanized NSG-SGM3 recipient mice. Each animal was transplanted by intrafemoral injection with a single cell dose of up to 300,000 unfractionated primary cells. The childhood AML BMs were harvested after eight weeks post transplantation and the level of engraftment was determined by flow cytometric analysis. Each symbol represents an individual transplanted mouse, and each sample was transplanted in triplicates. *P<0.05, ****P<0.001 (unpaired t-test). Sc: scramble control; shRNA: short hairpin RNA; ns: non-significant; AML: acute myeloid leukemia.

haematologica | 2018; 103(7)

1175


Y. Heshmati et al.

A

B

C

D

E

F

G

H

I

J

K

L

Figure 6. CHD4 inhibition in AML cells causes an arrest in the G0 phase of the cell cycle. A. Bar charts represent real time PCR analysis of mRNA levels after shRNAbased knockdown of CHD4 in MV4-11, Kasumi-1 and AML-193 cells (knockdown of mRNA levels of CHD4 in THP-1 cells used in this assay is depicted in Figure 2A). The cells were used for apoptosis and cell cycle analysis (Figure 6C-L), relative to control samples transduced with negative control vectors expressing scrambled shRNA, at 72 hours post transduction. mRNA levels were normalized to UBC. The data is represented as the mean ±S.E.M., ****P<0.001 (unpaired t-test), n=3. B. Western blot analysis of CHD4 and Actin levels in MV4-11, AML-193 and Kasumi-1 cells, with and without knockdown of CHD4, (knockdown levels of CHD4 protein levels in the THP-1 cells used in this assay is depicted in Figure 2B). C. Representative flow cytometry charts of the THP-1 cells transduced with CHD4 shRNAs, or a negative control vector, stained with Ki-67 and DAPI, at 72 hours after transduction. Each cell cycle phase is highlighted in red. D-G. Bar graphs of flow cytometric quantification as the percentage of cells in each population of the indicated cell lines transduced with CHD4 shRNA or a control vector, at 72 hours post CHD4. The data is presented as mean ±S.E.M., ***P<0.005, ****P<0.001 (unpaired t-test), n=3. H. Representative flow cytometry charts of THP-1 cells transduced with CHD4 shRNAs, or a negative control vector, stained with near-IR and Annexin V, at 72 hours after transduction. I-L. Bar graphs illustrating flow cytometric quantification as the percentage of cells in each population of the indicated cell lines transduced with CHD4 shRNA or a control vector, at 72 hours post transduction. The data is presented as mean ±S.E.M. **P<0.01, ***P<0.005, ****P<0.001, (unpaired t-test), n=3. DAPI; 4',6-diamidino-2-phenylindole; FITC: fluorescein isothiocyanate; ns: non-significant; Sc: scramble control; shRNA: short hairpin RNA; AML: acute myeloid leukemia; mRNA: messenger RNA.

1176

haematologica | 2018; 103(7)


CHD4 is required for maintenance of childhood acute myeloid leukemia

Clustered regularly interspaced short palindromic repeats (CRISPR)-Cas9 mediated loss of function of CHD4 prevents cell growth of AML cells

growth of bulk cancer cells/leukocytes and LICs of primary childhood AML BMs with different genetic lesions.

To exclude the possibility that the observed effects of the CHD4 inhibition was caused by off target effects, we took advantage of the highly specific CRISPR-Cas9 genome editing technology.41 Accordingly, we generated MV4-11 and THP-1 AML cells constitutively expressing Cas9, and transduced the cells with two individual inducible vectors expressing guide RNA (gRNA) homologous to the CHD4 locus (Figure 4A). Induction of expression of both gRNAs resulted in a significant disruption of the CHD4 coding region (Figure 4B) and reduction of protein levels (Figure 4C), in THP-1 AML cells. Although the reduction in protein levels in some experiments were relatively moderate, the specific CRISPR-Cas9 targeting of CHD4 caused a marked reduction in live THP-1 cells, confirming the essential role of CHD4 in AML cell growth (Figure 4D,E). Comparable results were obtained using the same approach with the human MV4-11 AML cells (Figure 4F-I) and the mouse MLL-AF9 AML cells (Online Supplementary Figure S4), thereby further validating that CHD4 is required for AML cell growth.

CHD4 suppression induces anti-leukemic effects in primary childhood AML cells in vivo

CHD4 suppression impairs maintenance of primary childhood AML cells ex vivo We demonstrated that CHD4 inhibition prevented growth of childhood AML cell lines (Figure 2 and Figure 4). Thus, we next investigated the importance of the inhibition of CHD4 in a cohort of childhood AML patient samples (Online Supplementary Table S3). To enable maintenance of the primary human childhood AML cells with high viability, including LICs, we utilized a stromal cell co-culture system for the cell growth assays.28 shRNAmediated knockdown of CHD4 using two independent constructs proved to be efficient for four out of six samples from which we could extract good quality mRNA (Figure 5A; Online Supplementary Figure S5A), and indeed from one sample we had enough material for western blot analysis (Online Supplementary Figure S5B). Flow cytometric analysis showed that leukocytes (CD45+ cells) and LICs (Lin-CD34+CD38– cells) of primary childhood AML samples transduced with a negative control vector could be maintained on stromal feeder layers with high viability (80-90%) up to five weeks after transduction (Figure 5B-D; Online Supplementary Figure S5C,D). In contrast, shRNA-mediated targeting of CHD4 in a primary childhood sample carrying the MLL translocation (Online Supplementary Table S3, sample “AML 6”) prevented maintenance of leukocytes and LICs compared to the control sample (Figure 5C,D). We then sought to investigate if CHD4 was required for maintenance of childhood AML samples carrying alternative genetic lesions (Online Supplementary Table S3). With the exception of LICs in one patient sample (sample “AML 2”, Figure 5D), the suppression of CHD4 with two independent shRNAs prevented the maintenance of leukocytes and LICs in all of the tested non-MLL rearranged childhood AML samples compared to the control cells (Figure 5B-D; Online Supplementary Figure S5C,D). Although all non-MLL rearranged samples did not robustly expand, the suppression of maintenance was similar to that of the MLLrearranged primary childhood AML sample (“AML 6”) (Figure 5B-D; Online Supplementary Figure S5C,D). These results show that the inhibition of CHD4 prevents haematologica | 2018; 103(7)

To investigate whether CHD4 also was required for growth of primary childhood AML cells in vivo, we used a humanized NSG-SGM3.42 Patient sample “AML 2” and “AML 7” (Online Supplementary Table S3), transduced with CHD4 targeting shRNA vectors or control vectors (Figure 5E), were intrafemurally transplanted into recipient mice. NSG-SGM3 mice transplanted with childhood AML cells transduced with control vectors resulted in a significant level of engraftment for both transplanted patient samples (median value of 5.7% +/- 2, and 9.6% +/- 2), determined by flow cytometric analysis of CD45+ cells, eight weeks post transplantation (Figure 5F). In sharp contrast, primary childhood AML cells transduced with shRNA vectors efficiently knocking down CHD4 mRNA (Figure 5E) displayed significantly lower levels of leukemic cell engraftment compared to the control cells (median value of 1.6, 0.03, P=0.0152, P<0.001, respectively) (Figure 5F). Thus, the targeting of CHD4 inhibits cell proliferation of primary childhood AML cells and AML progression in vivo.

CHD4 controls cell cycle progression of AML cells To investigate the cellular mechanisms by which CHD4 was required for AML cell growth in childhood AML, we used two MLL rearranged (THP-1 and MV4-11) and two non-MLL rearranged (AML-193 and Kasumi-1) childhood AML cell lines, and analyzed the effects in apoptosis and the cell cycle upon suppression of CHD4 expression. shRNA-mediated inhibition of CHD4 resulted in a significant reduction of mRNA (Figure 2A and Figure 6A) and protein levels (Figure 2B and Figure 6B), and a dramatic accumulation of all four cell lines in the G0 phase of the cell cycle, compared to the control cells. In addition, all cell lines displayed a decrease of cells in G1, whereas less pronounced effects was observed in the S and G2/M phase, compared to the control cells (Figure 6C-G). The robust arrest in G0 when CHD4 was inhibited (Online Supplementary Figure S3A,B; Online Supplementary Figure S6A, B), was confirmed in all four cell lines using an independent shRNA against CHD4 (Online Supplementary Figure S6C-F). In contrast to the strong effects in cell cycle progression, CHD4 depleted cells (Figure 2A,B and Figure 6A,B) displayed modest levels of early apoptotic cells (Annexin V+NIR–), and late apoptotic cells (Annexin V+NIR+) (Figure 6G-J). Likewise, knockdown of CHD4 using an independent shRNA (Online Supplementary Figure S3A,B and Online Supplementary Figure S6A,B) resulted in the same relatively low levels of apoptosis in all cell lines (Online Supplementary Figure S6G-J). Taken together, these data indicate that the primary cellular mechanism, whereby inhibition of CHD4 prevented cell growth of AML cells, is mediated via a growth arrest of the cells in the G0 phase of cycle progression, rather than induction of apoptosis.

CHD4 suppression induces an expression profile correlating to MYC targets To delineate the underlying molecular mechanisms whereby CHD4 is required for the growth of AML cells, 1177


Y. Heshmati et al. A

B

C

D

E

F

G

H

I

Figure 7. CHD4 inhibition in AML cells induces an expression profile linked to MYC and cell cycle progression. A. Heatmap of 100 genes with most significant changes in gene expression of THP-1 cells transduced with shRNA targeting CHD4, or a negative control vector, 72 hours post transduction. The data represent clustering of the individual experiments. B. List of the ten most significant KEGG gene set pathways correlating to gene expression changes resulting from shRNA-based inhibition of CHD4, relative to a negative control vector, in THP-1 cells. C-H. Enrichment plots of gene set enrichment analysis (GSEA) of mRNA gene expression profiling in response to CHD4 knockdown in THP-1 cells, demonstrating significant normalized enrichment score (NES) between MYC target V2 (Figure 7C), MYC targets V1 (Figure 7D), and E2F targets (Figure 7E), or to S phase (Figure 7F), synthesis of DNA (Figure 7G), assembly of the pre-replicative complex (Figure 7H). The bar charts represent the top ranked correlations in the predefined MSigDB: H hallmark collections (Figure 7C-E), or significant correlations to CP Reactome collection (Figure 7F-H). The heatmap on the right in C-E shows the relative level of gene expression (red = high, blue = low) of the most significant genes in the leading edge subset. I. Bar charts show real time PCR analysis of mRNA levels of genes showing changes in RNA-Seq of THP-1 cells, after shRNA-based knockdown of CHD4, relative to control cells transduced with vectors expressing scrambled shRNA, at 72 hours post transduction. mRNA levels were normalized to UBC. The data is represented as the mean ÂąS.E.M., *P<0.05, **P<0.01, ***P<0.005, ****P<0.001 (unpaired t-test), n=3. KEGG: Kyoto encyclopedia of genes and genomes; Sc: scramble control; FDR: false discovery rate.

1178

haematologica | 2018; 103(7)


CHD4 is required for maintenance of childhood acute myeloid leukemia

we performed RNA-Seq analysis of human THP-1 AML cells transduced with shRNAs targeting CHD4. Density and boxplot analysis of the samples illustrated an even distribution of counts per million (CPM) after normalization and clustering analysis (Online Supplementary Figure S7A,B). Heatmap clustering of the top 100 upregulated or downregulated genes (Figure 7A) of the RNA-Seq data showed a high degree of reproducibility between the triplicates, and a significant differentiation between the AML CHD4 knockdown cells and the cells transduced with a scramble control. Differentially expressed genes were those in which the mRNA levels were changed >1 or log2 fold change <-1; P<0.05; false discovery rate (FDR) <0.05 (Online Supplementary Table S4). Consistent with a role in gene repression,8 a majority of the genes were found to be upregulated upon knockdown of CHD4 (1011 genes were upregulated whereas 413 genes were downregulated). CHD4 and some of its previously reported target genes (e.g., IGF2, TGFB1 and PDGFA)38 were among those with the most significant changes in mRNA levels (Online Supplementary Table S4). Gene set enrichment analysis (GSEA) was used to investigate whether the transcriptome profile generated by the inhibition of CHD4 was associated to the collection of annotated gene sets (i.e., The Molecular Signatures Database [MSigDB]).43 The Kyoto encyclopedia of genes and genomes (KEGG) pathway analysis revealed that genes deregulated upon CHD4 suppression were enriched for gene sets such as spliceosome, proteasome and base excision repair (Figure 7B). Intriguingly, the expression changes in response to the inhibition of CHD4 compared to the Molecular Signatures Database (MSigDB) hallmark gene sets were most significantly correlated with the gene signature of two individual subgroups of MYC targets data sets (FDR Q-value=0.0 and normalized enrichment score [NES]=-3.3; FDR Qvalue=0.0, NES=-2.9, respectively), and to the MYC target E2F transcription factor (E2F) and its target genes with cell cycle related functions (FDR Q-value=0.0, NES=-2.6) (Figure 7C-E). Consistent with the link to MYC and its important role in cell cycle regulation, GSEA analysis using the reactome gene sets and comparisons to changes in gene expression upon CHD4 knockdown revealed significant associations with several gene sets involving cell cycle progression, including S phase (NES=-2.7; FDR Q=0), synthesis of DNA (NES=-3.3; FDR Q=6.64E-04) and assembly of the pre-replicative complex (NES=-2.4; FDR Q=0.00162) (Figure 7F-H). Further analysis of the RNA-Seq data revealed that inhibition of CHD4 caused decreased mRNA levels of MYC and several of its target genes involved in G1/S cell cycle transition, including cyclin D1, D2, E1, E2F1, and E2F2. Conversely, the negative cell cycle regulator p27 was shown to be upregulated. In contrast, other MYC targets with alternative roles in other stages of the cell cycle, such as cyclin A2, B1, E2, displayed less pronounced changes in mRNA levels (Online Supplementary Table S4). To validate the RNA-Seq findings, we performed an additional shRNA-based knockdown of CHD4 in THP-1 cells. qPCR analysis confirmed the RNA-Seq data, but with even more significant changes in the gene expression of MYC and an asset of its targets genes known to be involved in cell cycle progression (Figure 7I). Consistent with this, an additional RNA-Seq analysis of THP-1 cells haematologica | 2018; 103(7)

transduced with an independent shRNA against CHD4 again showed significant correlations to MYC and E2F targets. In addition, qPCR-based validation of the RNASeq data confirmed the previous results in Figure 7I and the observed deregulation of MYC and its downstream targets (Online Supplementary Table S4; Online Supplementary Figure S8). Thus, inhibition of CHD4 was significantly associated with MYC targets and gene sets involved in S phase cell cycle progression.

Discussion In this study, we performed loss of function screens on a large scale in AML cells and non-transformed BMs. We identified the epigenetic factor CHD4 as being essential for maintenance of LICs and disease progression of childhood AML, but not for normal hematopoietic cells. CHD4 inhibition in AML cells caused downregulation of MYC and its target genes and an arrest in the G0 phase of cell cycle progression. It is of utmost importance that future treatments for AML selectively target the cancer cells without harming normal cells. Accordingly, we showed that CHD4 is required for cell growth of leukemic cells carrying various genetic lesions and for disease progression using an immune competent mouse model. However, CHD4 is not essential for primary normal murine BMs or for normal human UCBs. Our findings are supported by previous reports showing that inhibition of CHD4 is not crucial for normal hematopoietic cell growth,14 but has nonessential functions in self-renewal and lineage choice in normal hematopoiesis.44,45 Interestingly, this selectivity seems to be conserved in breast cancer.9 Indeed, CHD4 has previously been demonstrated to be required for growth of a broad range of cancer cells,9,13,37,38,46,47 including colony formation capacity of AML cells,14 implying that CHD4 may represent a cancer-specific dependency in a wider repertoire of tumors. Most importantly, our results highlight a novel and essential role for CHD4 in maintenance of childhood AML in vitro and in vivo. The use of appropriate co-culture systems and a patient-derived xenograft mouse model for childhood AML allowed us to demonstrate that the essential role of CHD4 was consistently manifested in patient samples carrying diverse types of genetic lesions as well as the LICs. Intriguingly, the importance of CHD4 in cancer-initiating cells has also been reported in hepatocellular carcinoma47 and glioblastoma,38 indicating that the role of CHD4 in these cells that drive tumor growth may be more general than previously anticipated. CHD4/NuRD has been shown to control cell cycle progression in a p53 dependent manner,10-12 or in a p53 independent manner,9,48 and inhibition of CHD4 was shown to cause a cell cycle arrest in G1/S.11 In the present study, inhibition of CHD4 resulted in repression of MYC and its target genes involved in cell cycle progression and consequently caused a G0 cell cycle arrest. In support of this, inhibition of MYC has been reported to cause a G0/G1 block in the cell cycle in human lymphoid and myeloid cells.49 Moreover, CHD4 has been found to directly bind to the MYC promoter in glioblastoma cells and inhibition of CHD4 resulted in a downregulation of MYC.38 In addition, MYC was also part of a set of genes suggested to have a role in colony formation in AML cells.14 1179


Y. Heshmati et al.

Herein, we demonstrated that CHD4 is essential for maintenance of childhood AML and the LICs that are responsible for the emergence and development of the disease. Our data indicate that the importance of CHD4 in childhood AML may be mediated in part by promoting the expression of the MYC oncogene and its target genes. The cancer-specific dependency found in our studies show that CHD4 may represent a promising therapeutic target to battle childhood AML. Acknowledgments We thank Ying Qu for help with RNA-Seq analysis. We also thank Tim Somervaille for kindly providing the MLL-AF9 vector and Marco Herold for kindly providing the Cas9-mCherry vector. We would also like to thank NOPHO and Josefine Palle for providing the primary childhood AML cells as well as

References 1. Bonnet D, Dick JE. Human acute myeloid leukemia is organized as a hierarchy that originates from a primitive hematopoietic cell. Nat Med. 1997;3(7):730-737. 2. Lapidot T, Sirard C, Vormoor J, et al. A cell initiating human acute myeloid leukaemia after transplantation into SCID mice. Nature. 1994;367(6464):645-648. 3. Ishikawa F, Yoshida S, Saito Y, et al. Chemotherapy-resistant human AML stem cells home to and engraft within the bonemarrow endosteal region. Nat Biotechnol. 2007;25(11):1315-1321. 4. Farrar JE, Schuback HL, Ries RE, et al. Genomic profiling of pediatric acute myeloid leukemia reveals a changing mutational landscape from disease diagnosis to relapse. Cancer Res. 2016; 76(8):2197-2205. 5. Ho PA, Kutny MA, Alonzo TA, et al. Leukemic mutations in the methylationassociated genes DNMT3A and IDH2 are rare events in pediatric AML: a report from the Children's Oncology Group. Pediatr Blood Cancer. 2011;57(2):204-209. 6. Balgobind BV, Zwaan CM, Pieters R, Van den Heuvel-Eibrink MM. The heterogeneity of pediatric MLL-rearranged acute myeloid leukemia. Leukemia. 2011; 25(8):1239-1248. 7. Creutzig U, van den Heuvel-Eibrink MM, Gibson B, et al. Diagnosis and management of acute myeloid leukemia in children and adolescents: recommendations from an international expert panel. Blood. 2012; 120(16):3187-3205. 8. Denslow SA, Wade PA. The human Mi2/NuRD complex and gene regulation. Oncogene. 2007;26(37):5433-5438. 9. D'Alesio C, Punzi S, Cicalese A, et al. RNAi screens identify CHD4 as an essential gene in breast cancer growth. Oncotarget. 2016;7(49):80901-80915. 10. Luo J, Su F, Chen D, Shiloh A, Gu W. Deacetylation of p53 modulates its effect on cell growth and apoptosis. Nature. 2000;408(6810):377-381. 11. O'Shaughnessy A, Hendrich B. CHD4 in the DNA-damage response and cell cycle progression: not so NuRDy now. Biochem Soc Trans. 2013;41(3):777-782. 12. Polo SE, Kaidi A, Baskcomb L, Galanty Y, Jackson SP. Regulation of DNA-damage responses and cell-cycle progression by the chromatin remodelling factor CHD4. EMBO J. 2010;29(18):3130-3139.

1180

Cecilia Götherström and the National Cord Blood Bank at Karolinska University Hospital for providing the UCBs. We would like to acknowledge the MedH Core Flow Cytometry facility (Karolinska Institutet), supported by KI/SLL, for providing cell sorting services, cell analysis services, technical expertise and scientific input. We would also like to thank the Affymetrix core facility at Neo, BEA, Bioinformatics and Expression Analysis, which is supported by the board of research at the Karolinska Institute and the research committee at the Karolinska hospital. Funding This work was supported by the Wallenberg Foundation, The Swedish Cancer Society, the Magnus Bergvalls Foundation, the Karolinska Institutet, the Åke Wibergs Foundation, and Dr Åke Olsson Foundation for Hematological Research.

13. Xia L, Huang W, Bellani M, et al. CHD4 has oncogenic functions in initiating and maintaining epigenetic suppression of multiple tumor suppressor genes. Cancer Cell. 2017;31(5):653-668.e7. 14. Sperlazza J, Rahmani M, Beckta J, et al. Depletion of the chromatin remodeler CHD4 sensitizes AML blasts to genotoxic agents and reduces tumor formation. Blood. 2015;126(12):1462-1472. 15. Schmidt EE, Pelz O, Buhlmann S, Kerr G, Horn T, Boutros M. GenomeRNAi: a database for cell-based and in vivo RNAi phenotypes, 2013 update. Nucleic Acids Res. 2013;41(Database issue):D1021-1026. 16. Zuber J, Shi J, Wang E, et al. RNAi screen identifies Brd4 as a therapeutic target in acute myeloid leukaemia. Nature. 2011;478(7370):524-528. 17. Wermke M, Camgoz A, PaszkowskiRogacz M, et al. RNAi profiling of primary human AML cells identifies ROCK1 as a therapeutic target and nominates fasudil as an antileukemic drug. Blood. 2015; 125(24):3760-3768. 18. Tzelepis K, Koike-Yusa H, De Braekeleer E, et al. A CRISPR dropout screen identifies genetic vulnerabilities and therapeutic targets in acute myeloid leukemia. Cell Rep. 2016;17(4):1193-1205. 19. Sroczynska P, Cruickshank VA, Bukowski JP, et al. shRNA screening identifies JMJD1C as being required for leukemia maintenance. Blood. 2014;123(12):18701882. 20. Puram RV, Kowalczyk MS, de Boer CG, et al. Core circadian clock genes regulate leukemia stem cells in AML. Cell. 2016;165(2):303-316. 21. Porter CC, Kim J, Fosmire S, et al. Integrated genomic analyses identify WEE1 as a critical mediator of cell fate and a novel therapeutic target in acute myeloid leukemia. Leukemia. 2012;26(6):1266-1276. 22. Li H, Mar BG, Zhang H, et al. The EMT regulator ZEB2 is a novel dependency of human and murine acute myeloid leukemia. Blood. 2017;129(4):497-508. 23. Jude JG, Spencer GJ, Huang X, et al. A targeted knockdown screen of genes coding for phosphoinositide modulators identifies PIP4K2A as required for acute myeloid leukemia cell proliferation and survival. Oncogene. 2015;34(10):1253-1262. 24. Chang JY, Ngai PK, Priestle JP, Joss U, Vosbeck KD, van Oostrum J. Identification of a reactive lysyl residue (Lys103) of

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

recombinant human interleukin-1 beta. Mechanism of its reactivity and implication of its functional role in receptor binding. Biochemistry. 1992;31(11):2874-2878. Chan SM, Thomas D, Corces-Zimmerman MR, et al. Isocitrate dehydrogenase 1 and 2 mutations induce BCL-2 dependence in acute myeloid leukemia. Nat Med. 2015; 21(2):178-184. Carey A, Edwards DKt, Eide CA, et al. Identification of interleukin-1 by functional screening as a key mediator of cellular expansion and disease progression in acute myeloid leukemia. Cell Rep. 2017; 18(13):3204-3218. Banerji V, Frumm SM, Ross KN, et al. The intersection of genetic and chemical genomic screens identifies GSK-3alpha as a target in human acute myeloid leukemia. J Clin Invest. 2012;122(3):935-947. Griessinger E, Anjos-Afonso F, Pizzitola I, et al. A niche-like culture system allowing the maintenance of primary human acute myeloid leukemia-initiating cells: a new tool to decipher their chemoresistance and self-renewal mechanisms. Stem Cells Transl Med. 2014;3(4):520-529. Spooncer E, Heyworth CM, Dunn A, Dexter TM. Self-renewal and differentiation of interleukin-3-dependent multipotent stem cells are modulated by stromal cells and serum factors. Differentiation. 1986;31(2):111-118. Bhagwat AS, Roe JS, Mok BYL, Hohmann AF, Shi J, Vakoc CR. BET Bromodomain inhibition releases the mediator complex from select cis-regulatory elements. Cell Rep. 2016;15(3):519-530. Noguera NI, Song MS, Divona M, et al. Nucleophosmin/B26 regulates PTEN through interaction with HAUSP in acute myeloid leukemia. Leukemia. 2013; 27(5):1037-1043. Metzgeroth G, Walz C, Score J, et al. Recurrent finding of the FIP1L1-PDGFRA fusion gene in eosinophilia-associated acute myeloid leukemia and lymphoblastic T-cell lymphoma. Leukemia. 2007; 21(6):1183-1188. Thol F, Bollin R, Gehlhaar M, et al. Mutations in the cohesin complex in acute myeloid leukemia: clinical and prognostic implications. Blood. 2014;123(6):914-920. Cancer Genome Atlas Research N. Genomic and epigenomic landscapes of adult de novo acute myeloid leukemia. N Engl J Med. 2013;368(22):2059-2074.

haematologica | 2018; 103(7)


CHD4 is required for maintenance of childhood acute myeloid leukemia

35. Mizuno H, Kitada K, Nakai K, Sarai A. PrognoScan: a new database for metaanalysis of the prognostic value of genes. BMC Med Genomics. 2009;2:18. 36. Bagger FO, Sasivarevic D, Sohi SH, et al. BloodSpot: a database of gene expression profiles and transcriptional programs for healthy and malignant haematopoiesis. Nucleic Acids Res. 2016;44(D1):D917-924. 37. Bohm M, Wachtel M, Marques JG, et al. Helicase CHD4 is an epigenetic coregulator of PAX3-FOXO1 in alveolar rhabdomyosarcoma. J Clin Invest. 2016; 126(11):4237-4249. 38. Chudnovsky Y, Kim D, Zheng S, et al. ZFHX4 interacts with the NuRD core member CHD4 and regulates the glioblastoma tumor-initiating cell state. Cell Rep. 2014;6(2):313-324. 39. Guillemette S, Serra RW, Peng M, et al. Resistance to therapy in BRCA2 mutant cells due to loss of the nucleosome remodeling factor CHD4. Genes Dev. 2015;29(5): 489-494. 40. Somervaille TC, Cleary ML. PU.1 and Junb: suppressing the formation of acute myeloid

haematologica | 2018; 103(7)

41. 42.

43.

44.

45.

leukemia stem cells. Cancer Cell. 2006; 10(6):456-457. Cong L, Ran FA, Cox D, et al. Multiplex genome engineering using CRISPR/Cas systems. Science. 2013;339(6121):819-823. Billerbeck E, Barry WT, Mu K, Dorner M, Rice CM, Ploss A. Development of human CD4+FoxP3+ regulatory T cells in human stem cell factor-, granulocyte-macrophage colony-stimulating factor-, and interleukin3-expressing NOD-SCID IL2Rgamma(null) humanized mice. Blood. 2011;117(11): 3076-3086. Subramanian A, Tamayo P, Mootha VK, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A. 2005;102(43):15545-15550. Prasad P, Lennartsson A, Ekwall K. The roles of SNF2/SWI2 nucleosome remodeling enzymes in blood cell differentiation and leukemia. Biomed Res Int. 2015; 2015:347571. Yoshida T, Hazan I, Zhang J, et al. The role of the chromatin remodeler Mi-2beta in hematopoietic stem cell self-renewal and

46.

47.

48.

49.

multilineage differentiation. Genes Dev. 2008;22(9):1174-1189. Mohd-Sarip A, Teeuwssen M, Bot AG, et al. DOC1-dependent recruitment of NURD reveals antagonism with SWI/SNF during epithelial-mesenchymal transition in oral cancer cells. Cell Rep. 2017;20(1):61-75. Nio K, Yamashita T, Okada H, et al. Defeating EpCAM(+) liver cancer stem cells by targeting chromatin remodeling enzyme CHD4 in human hepatocellular carcinoma. J Hepatol. 2015;63(5):1164-1172. Buddaseth S, Gottmann W, Blasczyk R, Huyton T. Dysregulation of cell cycle control caused by overexpression of the oncogene pp32r1 (ANP32C) and the Tyr>His mutant pp32r1Y140H. Biochim Biophys Acta. 2013;1833(5):1212-1221. Wickstrom EL, Bacon TA, Gonzalez A, Freeman DL, Lyman GH, Wickstrom E. Human promyelocytic leukemia HL-60 cell proliferation and c-myc protein expression are inhibited by an antisense pentadecadeoxynucleotide targeted against cmyc mRNA. Proc Natl Acad Sci USA. 1988; 85(4):1028-1032.

1181


ARTICLE

Non-Hodgkin Lymphoma

Ferrata Storti Foundation

Haematologica 2018 Volume 103(7):1182-1190

Gene expression profiling reveals a close relationship between follicular lymphoma grade 3A and 3B, but distinct profiles of follicular lymphoma grade 1 and 2 Heike Horn,1 Christian Kohler,2 Raphael Witzig,3 Markus Kreuz,4 Ellen Leich,5 Wolfram Klapper,6 Michael Hummel,7 Markus Loeffler,4 Lorenz Trümper,8 Rainer Spang,2 Andreas Rosenwald5 and German Ott;3 for the Molecular Mechanisms in Malignant Lymphomas (MMML) Network Project

Dr. Margarete Fischer-Bosch-Institute of Clinical Pharmacology, Stuttgart and University of Tübingen; 2Statistical Bioinformatics Department, Institute of Functional Genomics, University of Regensburg; 3Department of Clinical Pathology, Robert Bosch Krankenhaus, Stuttgart; 4Institute for Medical Informatics, Statistics and Epidemiology, University of Leipzig; 5Institute of Pathology, University of Würzburg, and Comprehensive Cancer Center Mainfranken, Würzburg; 6Department of Pathology, Hematopathology Section, University Hospital Schleswig-Holstein Campus Kiel/Christian-Albrechts University Kiel; 7Institute of Pathology, Campus Benjamin Franklin, Charité– Universitätsmedizin Berlin and 8Department of Hematology and Oncology, Georg-August University of Göttingen, Germany 1

*HH and CK contributed equally to this work. A full list of MMML members is provided in the Online Supplementary Appendix.

ABSTRACT

A

Correspondence: german.ott@rbk.de

Received: October 24, 2017. Accepted: March 15, 2018. Pre-published: March 22, 2018. doi:10.3324/haematol.2017.181024 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/103/7/1182 ©2018 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.

1182

linear progression model of follicular lymphomas (FL) FL1, FL2 and FL3A has been favored, since FL3A often co-exist with an FL1/2 component. FL3B, in contrast, is thought to be more closely related to diffuse large B-cell lymphoma (DLBCL), and both are often simultaneously present in one tumor (DLBCL/FL3B). To obtain more detailed insights into follicular lymphoma progression, a comprehensive analysis of a well-defined set of FL1/2 (n=22), FL3A (n=16), FL3B (n=6), DLBCL/FL3B (n=9), and germinal center B-cell-type diffuse large B-cell lymphoma (n=45) was undertaken using gene expression profiling, immunohistochemical stainings and genetic analyses by fluorescence in situ hybridization. While immunohistochemical (CD10, IRF4/MUM1, Ki67, BCL2, BCL6) and genetic profiles (translocations of BCL2, BCL6 and MYC) delineate FL1-3A from FL3B and DLBCL/FL3B, significant differences were observed between FL1/2 and FL3A upon gene expression profiling. Interestingly, FL3B turned out to be closely related to FL3A, not categorizing within a separate gene expression cluster, and both FL3A and FL3B showed overlapping profiles in between FL1/2 and diffuse large B-cell lymphoma. Finally, based upon their gene expression pattern, DLBCL/FL3B represent a composite form of FL3B and DLBCL, with the majority of samples more closely resembling the latter. The fact that gene expression profiling clearly separated FL1/2 from both FL3A and FL3B suggests a closer biological relationship between the latter. This notion, however, is in contrast to immunohistochemical and genetic profiles of the different histological FL subtypes that point to a closer relationship between FL1/2 and FL3A, and separates them from FL3B.

Introduction Follicular lymphoma (FL) comprises approximately 30% of B-cell non-Hodgkin lymphomas (B-NHL) and represents the most common type of indolent B-NHL. FL originates from germinal center B cells (GCB) characterized by expression of CD10 and BCL6. FL consists of a mixture of centroblasts and centrocytes, the relative ratio of which determines the histological grade. While FL1/2 and FL3A consist of centrocytes and centroblasts (the difference between them being the numhaematologica | 2018; 103(7)


Gene expression profiling of FL subtypes

ber of centroblasts), FL3B only harbors centroblasts and centrocytes are not present.1 Although criteria for the histological grading of FL are well-defined,1,2 its precise assessment is challenging even for expert hematopathologists, in some cases resulting in interobserver variability in daily routine diagnostics.3,4 On the genetic level, approximately 85% of FL are characterized by the hallmark translocation t(14;18)(q32;q21), resulting in the juxtaposition of BCL2 to the IGH gene locus and, subsequently, to constitutive overexpression of BCL2 and inhibition of apoptosis. The evolution of novel cell clones with modified growth potential, morphologically often characterized by a higher number of centroblasts and/or by an increased proliferation index, is characteristic of progression of FL.2 Approximately 30% of FL transform to a more aggressive lymphoma, usually diffuse large B-cell lymphoma (DLBCL), which is typically associated with inferior clinical outcome.5 However, the genetic mechanisms underlying the progression and transformation of FL are poorly understood. Since FL grade 3A often co-exist with an FL1/2 component, and harbor the t(14;18) in approximately 60% of cases, a linear progression model of FL1, FL2 and FL3A has been developed, although FL3A does not necessarily evolve from FL1/2.3 FL3B, on the other hand, is presumed to be more closely related to DLBCL, and both FL3B and DLBCL are often simultaneously present in a lymph node.2,6,7 Although a molecular characterization of FL3A and FL3B versus FL1/2 has been attempted in the past, many reports have only addressed either immunohistochemical and/or genetic differences.6,8,9 The main goal of the present study, therefore, was the comprehensive genetic analysis of a well-defined set of FL3A and FL3B and their comparison with related entities such as GCBtype DLBCL and DLBCL with an FL3B component by gene expression profiling, immunohistochemistry, and genetic analysis by fluorescence in situ hybridization (FISH).

Methods Sample selection and histological grading All samples were collected by the Molecular Mechanisms in Malignant Lymphomas (MMML) network project, for which central and local ethical approval had been obtained. Due to the retrospective nature of the study, patients had been treated with various chemotherapy regimens, including (although only in a few cases) rituximab. Altogether, 98 tumor samples were included: 12 FL1, 10 FL2, 16 FL3A, 6 FL3B with a purely follicular growth pattern, 9 DLBCL with an additional FL3B component (DLBCL/FL3B), and 45 DLBCL of GCB-type, as determined by gene expression profiling.10 All tumor samples were classified and graded on the basis of routine hematoxylin and eosin (H&E), Giemsa and Perjodic acid Schiff (PAS) stainings according to the criteria of the World Health Organization (WHO) classification of tumors of hematopoetic and lymphoid tissues within a panel review process conducted by expert reference hematopathologists of the MMML.1,10

Immunohistochemical staining, fluorescence in situ hybridization and gene expression profiling Paraffin sections were immunostained with antibodies against CD20, CD10, BCL2, BCL6, IRF4/MUM1 and Ki67 as previously described.10,11 For the detection of BCL2-, BCL6- and haematologica | 2018; 103(7)

MYC-translocations, FISH was performed as described in the Online Supplementary Appendix.12 Gene expression profiles were generated as described in the Online Supplementary Appendix.10

Statistical analysis Differential gene expression analysis, ANOVA, determination of gene expression indices, classification analyses, and evaluation of the clinical outcome were assessed as described in the Online Supplementary Appendix.

Results Immunohistochemical and FISH analyses delineate FL1-3A from FL3B and DLBCL/FL3B According to FISH analyses, BCL2 breaks were the predominant genetic feature of FL1/2 (18 of 22, 82%). The number of cases with BCL2 alterations was lower in FL3A (12 of 16, 75%), FL3B (3 of 6, 50%) and in GCB-DLBCL (13 of 43, 30%), and BCL2 breaks were only infrequently detected in DLBCL/FL3B (1 of 9, 11%) (Table 1 and Online Supplementary Figure S1A). The substantial difference in the incidence of BCL2 breaks between FL3B and DLBCL/FL3B in comparison with FL1/2 or FL3A did not achieve statistical significance due to the small number of cases. Rearrangements of the BCL6 gene locus were most frequently encountered in DLBCL/FL3B (4 of 9, 44%) and in GCB-DLBCL (15 of 44, 34%), but were also detected in FL1/2 (2 of 22, 9%), FL3A (5 of 16, 31%), and FL3B (1 of 6, 17%) (Table 1 and Online Supplementary Figure S1A). Signal constellations indicative of an MYC break were most frequently observed in DLBCL/FL3B (2 of 9, 22%) and, to a lesser extent, also in FL3B (1 of 6, 17%) and GCBDLBCL (4 of 44, 9%). Occasional MYC alterations were also detected in FL1/2 (1 of 22, 5%) and FL3A (1 of 16, 6%) (Table 1 and Online Supplementary Figure S1A). According to immunohistochemistry, CD10 positive samples (>25% positive cells) were equally distributed within FL1/2 (13 of 21, 62%), FL3A (8 of 12, 67%), and FL3B/DLBCL (4 of 7, 57%), and, to a lesser extent, in GCB-DLBCL (14 of 42, 33%). CD10 was also expressed in 2 of 2 FL3B tested (Table 1 and Online Supplementary Figure S1B). While none of the FL1/2 were IRF4/MUM1 positive (â&#x2030;Ľ26%), reactivity for this protein was significantly increased in FL3A (4 of 8, 50%; P<0.001), FL3B (2 of 3, 67%; P<0.001), and DLBCL/FL3B (2 of 6, 33%; P<0.05) (Table 1). With increasing grade, the number of cases with high Ki67 indices (â&#x2030;Ľ70%) rose. While 3 of 20 FL1/2 (15%) showed reactivity for Ki67 of 70% or over, such a staining pattern was observed in 4 of 15 FL3A (27%) and in 3 of 5 FL3B (60%; P<0.05). A higher proportion of DLBCL/FL3B (7 of 8, 88%) were Ki67-high, although the difference did not reach significance when compared with FL3B or DLBCL (20 of 39, 51%) (Table 1 and Online Supplementary Figure S1B). Tumor samples with high BCL2 (>50%) expression were equally distributed among the different subtypes. FL3B and GCB-DLBCL showed the lowest numbers of BCL2 expressing cases (2 of 4, 50% and 24 of 43, 56%, respectively, vs. a mean frequency of 84% in the other lymphoma subtypes) (Table 1). All samples showed high numbers of BCL6-expressing cells ranging from 71% to 100% (Table 1). To summarize, FISH and immunohistochemical profiles pointed to a profound biological difference between FL1/2 1183


H. Horn et al. Table 1. Clinical data, immunohistochemistry and FISH analysis of all folicular lymphoma (FL) subsets.

FL1/2 FL3A

N. of cases 22 16 Median age, 58 57 years (range) (38-78) (36-71) Male/female 7/8 5/7 Median OS, months 69 72 BCL2-Break (%) 18/22 12/16 (82) (75) BCL6-Break (%) 2/22 5/16 (9) (31) MYC-Break (%) 1/22 1/16 (5) (6) CD10 (%) 13/21 8/12 (62) (67) IRF4/MUM1 (%) 0/16 4/8 (0) (50) Ki67 (%) 3/20 4/15 (15) (27) BCL2 (%) 15/20 12/15 (75) (80) BCL6 (%) 15/17 12/13 (88) (92)

FL3B

FL3B/ DLBCL

GCBDLBCL

6 9 53 57 (39-66) (34-80) 4/0 3/6 55 98 3/6 1/9 (50) (11) 1/6 4/9 (17) (44) 1/6 2/9 (17) (22) 2/2 4/7 (100) (57) 2/3 2/6 (67) (33) 3/5 7/8 (60) (88) 2/4 5/7 (50) (71) 3/3 5/7 (100) (71)

44 60 (8-85) 24/19 45 13/43 (30) 15/44 (34) 4/44 (9) 14/42 (33) 20/39 (51) 29/42 (69) 24/43 (56) 34/39 (87)

P FL1/2 FL1/2 FL1/2 FL1/2 vs. vs. vs. vs. FL3A FL3B DLBCL/FL3B DLBCL

FL3A vs. FL3B

FL3B FL3B vs. vs. DLBCL/FL3B DLBCL

ns ns ns ns

ns ns ns ns

ns ns ns 0.0007

ns ns ns 0.00004

ns ns ns ns

ns ns ns ns

ns ns ns ns

ns

ns

0.023

0.029

ns

ns

ns

ns

ns

ns

ns

ns

ns

ns

ns

ns

ns

0.03

ns

ns

ns

0.014

0.0001

ns

ns

ns

0.0009 0.00005 ns

0.0359

0.00006

0.00003

ns

ns

ns

ns

ns

ns

ns

ns

ns

ns

ns

ns

ns

ns

ns

ns

ns

DLBCL: diffuse large B-cell lymphoma; OS: overall survival; N: number; ns: not significant.

and the other FL-subgroups, with FL3A showing features more similar to FL1/2 than the other subtypes.

Results of gene expression profiling within different FL subtypes by unsupervised analysis In order to clarify whether FL subtypes defined by cytomorphology do also harbor unique gene expression profiles, hierarchical clustering was performed in 44 FL: FL1 (n=12), FL2 (n=10), FL3A (n=16) and FL3B (n=6). With regard to the 100 most variably expressed genes, however, the different samples did not cluster together according to their pre-defined histological grade (Online Supplementary Figure S2A and B). In order to clarify the relationship of tumors with a follicular background, particularly FL3B versus GCB-DLBCL, hierarchical clustering of the 44 FL samples and 45 GCB-DLBCL was performed using two different strategies. 1) The first round of analysis was based on the 100 most variably expressed genes, and 2) the second round focussed on the 500 most variably expressed genes. However, the GCB-DLBCL samples did not cluster together in either of the two approaches, and the 6 FL3B samples were scattered in between FL1-3A and DLBCL (Online Supplementary Figure S2C and D).

Gene expression profiling reveals a close relationship between FL3A and FL3B with supervised analysis In the next step, differential gene expression analysis (DEA) was performed to compare the gene expression profiles (GEP) of tumor subtypes and to align the results to different molecular features. To this end, the differential 1184

gene expression between: i) FL1/2 and FL3A; ii) FL3A and FL3B; iii) FL3B and DLBCL/FL3B; iv) FL3B and DLBCL; v) FL3A and DLBCL; and vi) FL1/2 and DLBCL was assessed. The GEPs of FL1/2, FL3A and FL3B differed significantly from DLBCL (with 7059 probeIDs mapping to 5027 annotated unique genes), 5093 (3798 genes) and 840 (691 genes) differentially expressed, respectively. Intriguingly, comparison of FL1/2 and FL3A revealed significant differences in GEPs (Figure 1A), while both FL3A and FL3B, as well as FL3B and DLBCL/FL3B, showed similar expression patterns (data not shown). DEA between FL1 versus FL2 failed to disclose significantly up- or down-regulated genes. Since differential GEPs had been described for t(14;18)-positive and t(14;18)-negative FL, respectively,13 the t(14;18) status for each sample was included in the final model to account for potentially hidden confounding effects. This ensures that the final list of genes differentially expressed between FL1/2 and FL3A does indeed reflect differences in regulation between these two FL entities and is therefore not due to the sample's individual t(14;18) status. Comparing FL1/2 and FL3A, 643 differentially expressed genes were identified; of those, 519 genes were up-regulated in FL3A and 125 genes up-regulated in FL1/2. A robust estimate of the median expression of all 643 regulated genes was calculated, resulting in a single value per sample, termed expression index. The FL1/2 subgroup tended to have mainly negative values since the majority of regulated features were indeed down-regulated. In contrast, the samples in the FL3A subtype were mostly positive. Obviously, the haematologica | 2018; 103(7)


Gene expression profiling of FL subtypes

A

C

B

D

Figure 1. Differential gene expression of FL1/2 and FL3A is independent of t(14;18)-status. Heatmap visualizing all 747 probe IDs differentially expressed between the histological follicular lymphoma (FL) subtypes FL1/2 and FL3A. Gene expression is shown as a pseudocolored representation of log expression ratio (=fold change), with yellow being above and blue being below the median level of gene expression in each row, as shown by the color scale (A). A robust estimate of the median expression of all 747 differentially expressed probe IDs was calculated resulting in a single value per sample (=expression index). While the indices of the FL1/2 subgroup were mainly negative (=down-regulated gene expression), the values for the FL3A group were mostly positive (=up-regulated gene expression). In each boxplot, the diamond symbol represents the mean index value (B). Separating the FL1/2 group into its individual components revealed that the mean indices between FL1 and FL2 are almost identical to FL3A (C). The frequency of the t(14;18) is almost balanced in FL1/2 (75%) and FL3A (82%), and also the mean gene expression indices of t(14;18)-positive and t(14;18)-negative FL are almost identical. This supports the notion that the different expression profiles of FL1/2 and FL3A were not the result of differences in the t(14;18)-status (D).

two means of the two groups were quite distinct, and this had also been supported by the results from DEA (Figure 1B). Splitting up the FL1/2 group into its two components (FL1 and FL2) showed the mean indices of these two groups to be almost identical (Figure 1C). The list of differentially expressed genes in FL1/2 and FL3A should be independent of the t(14;18)-translocation status, suggesting that the mean gene expression indices of both t(14;18)-positive and t(14;18)-negative samples, should be balanced. Of note, almost identical indices of differentially expressed genes were observed between FL1/2 and FL3A with regard to their t(14;18)-status. Thus, a confounding effect mediated by t(14;18) can be excluded (Figure 1D). As can be concluded from 519 genes up-regulated in FL3A and 125 genes up-regulated in FL1/2, different signaling pathways are active in the respective subtypes. While FL1/2 is dominated by the expression of genes involved in microenvironmental interactions (as for example cell-cell-adhesion and T-cell proliferation), FL3A is characterized by the expression of genes related to RNA transport and regulation, cell cycle, and DNA repair (Online Supplementary Tables S1 and S2, and Online Supplementary Figures S3 and S4). Since it might be argued that the differential gene expression observed between FL1/2 and FL3A might only haematologica | 2018; 103(7)

reflect enhanced proliferation in the latter, previously published proliferation signatures14,15 were investigated within the present study cohort. A comparison of genes contained in the known proliferation signature (n=592)14 with the 643 differentially expressed genes in FL1/2 and FL3A/B showed only a small overlap in 105 genes (approx. 16%). Applying a ‘proliferation index’ based upon the genes of the published signature14 to our FL cases revealed a highly heterogeneous spectrum of the indices in FL1/2, indicating 10 of 22 (45%) with low (≤0) and 12 of 22 (55%) samples with high proliferation index (>0) (Online Supplementary Figure S5). Furthermore, differential GEP between FL1/2 and FL3A might reflect the non-malignant stroma, as indicated also by enhanced expression of microenvironmental genes in FL1/2. By analyzing a previously published stromal signature16 within the present data set, the highest median stroma index was indeed observed in FL1/2 (Online Supplementary Figure S6). No significant differences were observed when comparing GEPs between FL3A and FL3B. Moreover, upon applying a supervised ANOVA approach including FL1/2, FL3A and FL3B, as previously described,8 12 genes could be identified as significantly differentially expressed between FL1/2 and FL3A (Figure 2A); the majority of these genes are involved in cell proliferation, DNA repair, cellular metabolism and intracellular protein trafficking (Online 1185


H. Horn et al. A

B

C

Figure 2. The gene expression profiles of FL3B and FL3A seem to be closely related. A supervised ANOVA analysis identified 13 probe IDs (=12 unique genes) that were most variable between FL1/2, FL3A and FL3B. The majority of FL3B samples more closely resembled the expression pattern of FL3A (A), which was also evident when performing hierarchical clustering based upon the expression of the 13 probe IDs. The legend indicates how many samples belong to an individual entity [eg. FL1(12) represents 12 FL1 samples] (B). Application of a linear classifier that separates FL1/2 from FL3A and subsequently predicts class membership of six FL3B cases revealed that the majority of these FL3B (4 of 6, 67%) show a “FL3A”-like Gene Expression Profile. This result further indicates that the close relationship between FL3A and FL3B is not due to a methodological artefact introduced by ANOVA analysis because the classification approach uses a distinct strategy. The dashed vertical line at Y intercept at 50% separates the ’FL1/2’-like and ’FL3A’-like Gene Expression Profiles. Each follicular lymphoma (FL) entity is represented by an individual symbol (circle: FL1; triangle FL3A; square FL3B). The vertical dotted lines separate the three FL entities (C).

Supplementary Table S3). Interestingly, however, and in contrast to previous findings,8 in this analysis, 5 of 6 FL3B samples showed a GEP similar to the FL3A cases (Figure 2A) which was also shown when applying hierarchical clustering to FL subtypes based upon the 12 most differentially expressed genes (Figure 2B). To further substantiate these findings, supervised classification analysis was performed establishing a linear classifier to separate FL1/2 from FL3A in a test setting and to predict class membership of FL3B in the validation set. The relative frequency (as %) of being classified as FL1/2 (>50%) is indicated in Figure 2C. While 17 of 22 FL1/2 (77%) were clearly classified as such, the GEP of 5 FL1/2 (23%, MPI-629, MPI-772, MPI-776, MPI-777 and MPI889) was intermediate between FL1/2 and FL3A. With regard to FISH and/or immunohistochemical parameters, there was no difference in the FL1/2 classified as such and the 5 spiking FL1/2, most notably also not in respect of the proliferation index as measured by Ki67 (data not shown). Moreover, 13 of 16 FL3A (81%) were explicitly classified as such, while GEP of 2 FL3A (13%, MPI-632 and MPI643) tended towards the profile of FL1/2. One FL3A (6%, MPI-867) was distinctly classified as FL1/2. All 3 ‘misclassified’ FL3A harbored only low numbers of Ki67-positive cells (mean: 18%, range 5-30%) as the only striking feature when compared with their ‘classical’ counterparts 1186

(mean: 63%, range 30-90%). Interestingly, categorization of FL3B samples assigned 4 cases (4 of 6, 67%; MPI-660, MPI-661, MPI-721 and MPI-817) as FL3A, while 1 FL3B (17%, MPI-649) was classified as FL1/2, and the expression pattern of the remaining sample (MPI-667) was intermediate between FL1/2 and FL3B. When considering genetic and immunohistochemical features, the 2 ‘misclassified’ FL3B were IRF4/MUM1 expression negative, although a high proportion of FL3B showed positivity for this marker (Table 1). A decreased number of Ki67-positive cells, however, was not observed in these ‘misclassified’ FL3B samples.

Gene expression profiles of FL3B/DLBCL cluster in between FL3B and GCB-DLBCL To asses whether FL3B and DLBCL/FL3B show a GEP more similar to FL or DLBCL, classification analysis was undertaken by devising a linear classifier allowing for the distinction of FL1/2 and DLBCL, since GEP clearly separated these entities, with 7059 probe IDs (5027 genes) being differentially expressed. Subsequently, the linear classifier was used to predict class-membership of FL3A, FL3B and FL3B/DLBCL (Figure 3A). GEPs of FL1/2 and DLBCL appeared quite distinct. A homogenous GEP of FL1/2 and DLBCL was observed with only a few exceptions. Of note, 1 DLBCL (MPI-226) was classified as FL1/2 in haematologica | 2018; 103(7)


Gene expression profiling of FL subtypes

A

B

Figure 3. Classification analysis to follicular lymphoma (FL) subtypes FL3A, FL3B and DLBCL/FL3B into either the category of FL1/2 or germinal center B-cell diffuse large B-cell lymphoma (GCB-DLBCL). While the Gene Expression Profiles (GEPs) were clearly separated for FL1/2 and GCB-DLBCL, no clear distinction was seen for FL3A and FL3B. DLBCL/FL3B tended to cluster in the group of GCB-DLBCL (A). Schematic overview of FL subtype categorization based on GEP. While FL1/2 and GCB-DLBCL indicate a homogeneous expression profiling clearly separated from each other, the expression pattern of FL3A and FL3B is in between FL and DLBCL. DLBCL/FL3B are more closely related to DLBCL than to FL (B).

approximately 75% of all iterations. Eleven of 16 (69%) FL3A clustered within the group of FL1/2 (Figure 3A). Considering genetic features of the resulting 5 ‘misclassified’ FL3A, the frequency of BCL6 breaks in this group (3 of 5, 60%) was more similar to the DLBCL group than to FL1/2 (GCB-DLBCL: 44% vs. FL1/2: 9%) (Table 1). Of interest in this context, the GEP of 3 FL3B (MPI-661, MPI721 and MPI-817; 3 of 6, 50%) was more closely related to DLBCL than to FL1/2, while the remaining 3 FL3B (50%) were clearly classified as FL1/2 (Figure 3A). However, there was no significant difference in genetic features of the 2 FL3B clusters, especially not with respect to the occurrence of the t(14;18). With special regard to immunohistochemical markers, no differences were detected that could possibly separate the ‘core’ group from the outliers. The majority of the DLBCL/FL3B samples (8 of 9, 89%) were classified as DLBCL, apart from MPI-650 which was classified as FL1/2 in more than 70% of all iterations (Figure 3A). The only striking finding in this sample was the lack of a BCL6-translocation, while 44% of either DLBCL/FL3B or DLBCL, respectively, harbored the rearrangement. To summarize, gene expression profiling and classification analysis showed FL1/2 to be clearly separated from DLBCL, indicating a distinct FL-specific and DLBCL-specific gene expression pattern, as would have been expected. FL3B turned out to be closely related to FL3A, not categorizing within a separate gene expression cluster, and haematologica | 2018; 103(7)

both FL3A and FL3B showed overlapping GEPs in between FL1/2 and DLBCL. Finally, based upon their expression pattern, DLBCL/FL3B did indeed seem to represent a composite form of FL3B and DLBCL, with the majority of samples more closely resembling DLBCL (Figure 3B).

Gene expression profiles of FL with or without the t(14;18) do not differ significantly in the various histological subtypes Presence or absence of the t(14;18) was one of the factors most distinguishing FL1-3A from FL3B, DLBCL/FL3B and GCB-DLBCL (see above). We, therefore, asked whether GEP might be different in lymphomas with or without the t(14;18). Interestingly, such a difference was not seen either in the entire FL cohort, nor within different histological subtypes. Only 2 genes (FAM30A and IL17RB) were differentially expressed between t(14;18)-positive and t(14;18)-negative FL1/2, FL3A and FL3B (n=33 and n=11, respectively). Upon classification analysis, trying to separate t(14;18)-positive from negative FL1-3B samples, GEP of t(14;18)-negative FL1-3B was quite homogenous, while the classification profiles of some t(14;18)-positive cases fluctuated quite heavily, with 6 FL with t(14;18) (MPI-600, MPI-604, MPI-640, MPI-659, MPI-667 and MPI668) (6 of 33, 18%) showing a GEP more similar to the t(14;18)-negative cohort (Figure 4A). When testing for the differential expression of single genes that varied between 1187


H. Horn et al. A

B

C

Figure 4. Gene expression profiles and overall survival of follicular lymphoma (FL) with and without t(14;18). Classification analysis of t(14;18)-positive and t(14;18)-negative FL revealed that the expression profile of t(14;18)-negative samples was homogeneous, while the profile of t(14;18)-positive FL fluctuates, with 6 t(14;18)-positive FL that were mainly classified as t(14;18)-negative (A). Gene expression of CD10 (=MME) was only significantly different when comparing the entire group of t(14;18)-positive and t(14;18)-negative lymphomas [including germinal center B-cell diffuse large B-cell lymphoma (GCB-DLBCL), first two boxplots from the left], but not when comparing FLs only (boxplots 3 and 4). Only a trend towards lowered CD10 expression in t(14;18)-negative FLs was observed. In the boxplot, the mean expression of CD10 is represented by a diamond symbol in addition to its median value (B). Gene expression of BCL2 showed no significant differences at all, neither in the entire group of t(14;18)-positive and t(14;18)-negative lymphomas, nor in the FL group only (C).

t(14;18)-positive and t(14;18)-negative FL samples in another cohort (e.g. BCL2, BCL6, CD10, IRF4/MUM1, IKBKE),13 a significant difference was detected only for CD10 (=MME) (Figure 4B), while the median expression of the other genes was almost identical in both groups (e.g. BCL2 expression in Figure 4C).

Clinical outcome of different histological FL subtypes In this small series, no significant differences were noted in the overall survival (OS) of FL1/2, FL3A, FL3B and FL3B/DLBCL, which ranged from 55 to 98 months (Table 1 and Figure 5A). In agreement with previous studies,13,17 no difference in OS was observed when comparing t(14;18)-positive and t(14;18)-negative FL of all subtypes (FL1/2, FL3A and FL3B) (Figure 5B).

Discussion The up-dated WHO classification for lymphoid neoplasms categorizes FL into three histological grades according to the number of centroblasts and the presence or absence of centrocytes within the tumor follicles.1,18 The vast majority of FL are FL1/2, and hence have formed the backbone of a plethora of reports defining genetic features underlying FL pathogenesis.19 In contrast, data on grade 3 FL are scarce, obviously due to the limited number of 1188

these cases available for studies. In particular, few studies have been conducted to elucidate the molecular mechanisms of the pathogenesis of FL3B. With respect to available immunophenotypic and genetic data, the two subgroups of FL3 are regarded as discrete entities, with FL3A more closely related to FL1/2, while FL3B, in contrast, to a greater extent resembles DLBCL.2,6,8,20,21 This hypothesis is supported by the fact that FL3A frequently harbor FL1/2 follicles in a given tumor specimen and harbor the t(14;18) in roughly 60% of cases, while FL3B only show infrequent BCL2 translocations, while they are often CD10 negative and IRF4/MUM1 positive.2,6,8,20,21 These differences were also evident in the present study, showing a linear decrease in the frequency of BCL2 rearrangements (although not significant), as well as an increase in positive stainings for IRF4/MUM1 with the number of centroblasts. A CD10+IRF4/MUM1â&#x20AC;&#x201C; immunophenotype has been described as typical for FL1/2, while an increased frequency of CD10IRF4/MUM1+ cells was reported in FL3B and DLBCL/FL3B.6,9 In the present study, a characteristic CD10-negative phenotype was not observed in FL3B, obviously due to the fact that only 2 of the 6 FL3B analyzed had CD10 stainings available. Nevertheless, present data support the concept that FL3A and FL3B are characterized by a different spectrum of underlying genethaematologica | 2018; 103(7)


Gene expression profiling of FL subtypes

A

B

Figure 5. Overall survival (OS) of histological follicular lymphoma (FL) subtypes and FL with and without t(14;18). Kaplan-Meier-plot of the different FL subtypes FL1/2, FL3A, FL3B and diffuse large B-cell lymphoma (DLBCL)/FL3B (A), as well as the comparison of t(14;18)-positive and t(14;18)-negative FL (B) revealed no significant differences in OS.

ic events, suggesting that they are biologically distinct. In order to obtain a more detailed insight into the pathogenesis of FL3A and FL3B, GEP had been performed by Piccaluga et al.8 However, they had not integrated markers for immunohistochemistry and FISH into their study. The main finding of this study was a relatively homogeneous GEP of different FL subtypes. In a supervised analysis approach, Piccaluga et al.8 found that FL1/2 and FL3A formed one cluster, while FL3B formed a separate group distinguishable from FL1/2/3A based on the differential expression of 30 genes. In contrast to these findings, we failed to observe a significant difference in the gene expression patterns of FL1/2 and FL3A on the one hand and of FL3B on the other hand. In contrast, from our data set, a significantly differential gene expression emerged between FL1/2 and FL3A, while FL3B profiles more closely resembled those of FL3A. Despite these different findings concerning the relationship of FL1/2, FL3A and FL3B, we and Piccaluga et al. identified similar pathways affected in FL1/2 and FL3, mainly targeting cellular metabolism, cell cycle, and cell growth. Applying previously published proliferation signatures14,15 to our FL samples, however, revealed a highly heterogeneous spectrum of proliferation indices in FL1/2 cases in the present study. In keeping with the fact that almost no overlap was observed between the genes from known proliferation signatures and the 643 differentially expressed genes between FL1/2 and FL3A, we provide evidence that proliferation is not the only explanation for the difference between the GEP in FL1/2 and FL3A. An increased proliferation, in particular in FL1/2 samples, had already been described by using miRNA profiling, further substantiating the finding of a wide proliferation spectrum even in FL1/2.22,23 Of 12 genes distinguishing FL1/2 from FL3A/B, all were over-expressed in FL3A/B. Intriguingly, 3 genes, MRE11A, TXN and TOP2A, had already been associated with the pathogenesis of lymphoma and, therefore, targeting their expression might be beneficial for tailored therapy (Online Supplementary Table S3).24-26 Based on the classification approach performed in the present study, FL1/2 were clearly distinguishable from GCB-DLBCL, while FL3A and FL3B, in contrast, showed haematologica | 2018; 103(7)

highly similar gene expression patterns; moreover, they both formed an expression cluster intermediate between FL and DLBCL. Piccaluga et al. found that FL3A and FL3B were clearly distinguishable in their gene expression patterns in supervised analysis and that both subgroups resembled more closely FL than DLBCL.8 Nevertheless, in their study, they identified 2 FL3B that clustered within the DLBCL group, similar to the findings in the present study.8 These data might indicate that a distinct classification of FL subtypes based on their GEP is not possible in all cases. Furthermore, these results underline the fact that the grouping of FL3B within FL1-3A might not be biologically justified in all cases. In the Kiel classification system,27 FL3B had been regarded as a follicular variant of DLBCL (â&#x20AC;&#x153;follicular centroblastic lymphomaâ&#x20AC;?). Since so far only two GEP global studies on FL are available, each one analyzing only a limited number of samples, it is difficult to draw universal conclusions from these investigations and, therefore, additional validation studies are clearly needed. In fact, considering only gene expression data, no clear-cut pattern that might be useful to distinguish tumors with a different follicular component can be obtained. Although only a limited number of FL3B were available to be investigated within the present study, the fact that GEP clearly separated FL1/2 and FL3A/FL3B suggests a close relationship between FL3A and FL3B. This notion, however, is in contrast to immunohistochemical and genetic profiles of the different histological FL subtypes that point to a closer relationship between FL1/2 and FL3A, and separating them from FL3B. This phenomenon could possibly be explained by the different methological approaches used, focusing on the examination of tumor cells by immunohistochemistry and FISH, while both tumor and non-malignant bystander cells are simultaneously interrogated by gene expression profiling. Finally, the therapeutic implication of a diagnosis of FL3A is still a subject of debate. Many hemato-oncologists regard FL3A as belonging to the spectrum of conventional FL1/2. On the other hand, a recent retrospective analysis of FL3A cases enrolled in the German low- and high-grade lymphoma trials failed to observe any difference in survival between FL3A and FL3B and, most intriguingly, 1189


H. Horn et al.

found a plateau formation in FL3A after six years without late progression-free survival events, in contrast to FL1/2.3 This observation may well relate to the fact that the present study found GEPs to be distinct between FL1/2 and FL3A, and supports a hypothesis of a differential molecular background that might influence the clinical outcome of patients with FL. Furthermore, recent clinical data showed that patients with DLBCL/FL3B had clinical features comparable to either FL or DLBCL,28 a notion supported by the GEP findings of the present study, pointing to the fact that DLBCL/FL3B may represent a composite form of FL3B and DLBCL with similar gene expression patterns. Since standardized treatment decisions and distinct therapy strategies for patients with FL still have to be defined, the translation of histological grading into differ-

References 1. Swerdlow S, Campo E, Harris N, et al. World Health Organization Classification of Tu-mours of Haematopoietic and Lymphoid tissues. Lyon: IARC Press; 2008. 2. Ott G, Katzenberger T, Lohr A, et al. Cytomorphologic, immunohistochemical, and cyto-genetic profiles of follicular lymphoma: 2 types of follicular lymphoma grade 3. Blood. 2002;99(10):3806-3812. 3. Koch K, Hoster E, Ziepert M, et al. Clinical, pathological and genetic features of follicular lymphoma grade 3A: A joint analysis of the German low-grade and high-grade lymphoma study groups GLSG and DSHNHL. Ann Oncol. 2016;27(7):1323-1329. 4. A clinical evaluation of the International Lymphoma Study Group classification of non-Hodgkin's lymphoma. The NonHodgkin's Lymphoma Classification Project. Blood. 1997;89(11):3909-3918. 5. Xerri L, Dirnhofer S, Quintanilla-Martinez L, et al. The heterogeneity of follicular lympho-mas: From early development to transformation. Virchows Arch. 2016; 468(2):127-139. 6. Horn H, Schmelter C, Leich E, et al. Follicular lymphoma grade 3B is a distinct neoplasm according to cytogenetic and immunohistochemical profiles. Haematologica. 2011;96(9):1327-1334. 7. Katzenberger T, Ott G, Klein T, et al. Cytogenetic alterations affecting BCL6 are pre-dominantly found in follicular lymphomas grade 3B with a diffuse large B-cell component. Am J Pathol. 2004; 165(2):481490. 8. Piccaluga PP, Califano A, Klein U, et al. Gene expression analysis provides a potential rationale for revising the histological grading of follicular lymphomas. Haematologica. 2008;93(7):1033-1038. 9. Karube K, Guo Y, Suzumiya J, et al. CD10MUM1+ follicular lymphoma lacks BCL2 gene translocation and shows characteristic biologic and clinical features. Blood. 2007; 109(7):3076-3079. 10. Hummel M, Bentink S, Berger H, et al. A biologic definition of Burkitt's lymphoma from transcriptional and genomic profiling.

1190

ent treatment modalities for the moment remains unclear. However, it had previously been postulated that an individualized approach might be warranted in FL treatment since specific molecular signatures have been shown to reflect a clinical aggressiveness of FL that is independent of histological grade.29 Acknowledgments We gratefully acknowledge Petra Hitschke, Katja Bräutigam, Daniela Pumm and Thomas Hees for excellent technical assistance. Funding This study was supported by the Deutsche Krebshilfe and the Robert Bosch Stiftung, Germany. The authors declare no competing interests.

N Engl J Med. 2006;354(23):2419-2430. 11. Aukema SM, Kreuz M, Kohler CW, et al. Biological characterization of adult MYCtranslocation-positive mature B-cell lymphomas other than molecular Burkitt lymphoma. Haematologica. 2014;99(4):726735. 12. Horn H, Bausinger J, Staiger AM, et al. Numerical and structural genomic aberrations are reliably detectable in tissue microarrays of formalin-fixed paraffinembedded tumor samples by fluorescence in-situ hybridization. PloS One. 2014; 9(4):e95047. 13. Leich E, Salaverria I, Bea S, et al. Follicular lymphomas with and without translocation t(14;18) differ in gene expression profiles and genetic alterations. Blood. 2009; 114(4):826-834. 14. Rosenwald A, Wright G, Chan WC, et al. The use of molecular profiling to predict sur-vival after chemotherapy for diffuse large-B-cell lymphoma. N Engl J Med. 2002; 346(25):1937-1947. 15. Su AI, Wiltshire T, Batalov S, et al. A gene atlas of the mouse and human proteinencoding transcriptomes. Proc Natl Acad Sci USA. 2004;101(16):6062-6067. 16. Lenz G, Wright G, Dave SS, et al. Stromal gene signatures in large-B-cell lymphomas. N Engl J Med. 2008;359(22):2313-2323. 17. Leich E, Hoster E, Wartenberg M, et al. Similar clinical features in follicular lymphomas with and without breaks in the BCL2 locus. Leukemia. 2016;30(4):854-860. 18. Swerdlow SH, Campo E, Pileri SA, et al. The 2016 revision of the World Health Organi-zation classification of lymphoid neoplasms. Blood. 2016;127(20):2375-2390. 19. Pastore A, Jurinovic V, Kridel R, et al. Integration of gene mutations in risk prognostica-tion for patients receiving first-line immunochemotherapy for follicular lymphoma: A ret-rospective analysis of a prospective clinical trial and validation in a population-based reg-istry. Lancet Oncol. 2015;16(9):1111-1122. 20. Bosga-Bouwer AG, van Imhoff GW, Boonstra R, et al. Follicular lymphoma grade 3B includes 3 cytogenetically defined subgroups with primary t(14;18), 3q27, or

21.

22.

23.

24.

25.

26.

27. 28.

29.

other trans-locations: T(14;18) and 3q27 are mutually exclusive. Blood. 2003; 101(3):1149-1154. Bosga-Bouwer AG, van den Berg A, Haralambieva E, et al. Molecular, cytogenetic, and immunophenotypic characterization of follicular lymphoma grade 3B; a separate entity or part of the spectrum of diffuse large B-cell lymphoma or follicular lymphoma?. Hum Pathol. 2006; 37(5):528533. Wang W, Corrigan-Cummins M, Hudson J, et al. MicroRNA profiling of follicular lymphoma identifies microRNAs related to cell proliferation and tumor response. Haematologica. 2012;97(4):586-594. Leich E, Zamo A, Horn H, et al. MicroRNA profiles of t(14;18)-negative follicular lymphoma support a late germinal center B-cell phenotype. Blood. 2011;118(20):55505558. Spehalski E, Capper KM, Smith CJ, et al. MRE11 Promotes Tumorigenesis by Facilitat-ing Resistance to OncogeneInduced Replication Stress. Cancer Res. 2017;77(19):5327-5338. Sewastianik T, Szydlowski M, Jablonska E, et al. FOXO1 is a TXN- and p300-dependent sensor and effector of oxidative stress in diffuse large B-cell lymphomas characterized by increased oxidative metabolism. Oncogene. 2016;35(46):5989-6000. Schrader C, Meusers P, Brittinger G, et al. Topoisomerase IIalpha expression in mantle cell lymphoma: A marker of cell proliferation and a prognostic factor for clinical outcome. Leukemia. 2004;18(7):1200–1206. Stansfeld AG, Diebold J, Noel H, et al. Updated Kiel classification for lymphomas. Lancet. 1988;1(8580):292-293. Magnano L, Balagué O, Dlouhy I, et al. Clinicobiological features and prognostic impact of diffuse large B-cell lymphoma component in the outcome of patients with previously untreated follicular lymphoma. Ann Oncol. 2017;28(11):2799-2805. Glas AM, Kersten MJ, Delahaye LJMJ, et al. Gene expression profiling in follicular lymphoma to assess clinical aggressiveness and to guide the choice of treatment. Blood. 2005;105(1):301-307.

haematologica | 2018; 103(7)


ARTICLE

Non-Hodgkin Lymphoma

The outcome of peripheral T-cell lymphoma patients failing first-line therapy: a report from the prospective, International T-Cell Project

Monica Bellei,1 Francine M. Foss,2 Andrei R. Shustov,3 Steven M. Horwitz,4 Luigi Marcheselli,1 Won Seog Kim,5 Maria E. Cabrera,6 Ivan Dlouhy,7 Arnon Nagler,8 Ranjana H. Advani,9 Emanuela A. Pesce,1 Young-Hyeh Ko,10 Virginia Martinez,6 Silvia Montoto,11 Carlos Chiattone,12 Alison Moskowitz,4 Michele Spina,13 Irene Biasoli,14 Martina Manni1 and Massimo Federico;1 on behalf of the International T-cell Project Network Department of Diagnostic, Clinical and Public Health Medicine, University of Modena and Reggio Emilia, Modena, Italy; 2Yale University School of Medicine, New Haven, CT, USA; 3 University of Washington School of Medicine, Seattle, WA, USA; 4Memorial Sloan-Kettering Cancer Center, New York, NY, USA; 5Hematology-Oncology Samsung Medical Center, Seoul, South Korea; 6Hospital del Salvador, Universidad de Chile, Santiago, Chile; 7Department of Hematology, Hospital Clinic, Barcelona, Spain; 8Sheba Medical Center, Tel Hashomer, Israel; 9Stanford University, Stanford, CA, USA; 10Department of Pathology, Samsung Medical Center, Seoul, South Korea; 11Department of Haemato-Oncology, St. Bartholomew's Hospital, Barts Health NHS Trust, London, UK; 12Departamento de Clínica Médica, FCM da Santa Casa de São Paulo, Brazil; 13Medical Oncology A, National Cancer Institute, Aviano, Italy and 14Department of Medicine, University Hospital and School of Medicine, Universidade Federal do Rio de Janeiro, Brazil. 1

Ferrata Storti Foundation

Haematologica 2018 Volume 103(7):1191-1197

ABSTRACT

T

his analysis explored factors influencing survival of patients with primary refractory and relapsed peripheral T-cell lymphomas enrolled in the prospective International T-cell Project. We analyzed data from 1020 patients with newly diagnosed disease, enrolled between September 2006 and December 2015. Out of 937 patients who received first-line treatment, 436 (47%) were identified as refractory and 197 (21%) as relapsed. Median time from the end of treatment to relapse was 8 months (range 2-73). Overall, 75 patients (8%) were consolidated with bone marrow transplantation, including 12 refractory and 22 relapsed patients. After a median follow up of 38 months (range 1-96 months) from documentation of refractory/relapsed disease, 440 patients had died. The median overall survival (OS) was 5.8 months; 3-year overall survival rates were 21% and 28% for refractory and relapsed patients, respectively (P<0.001). Patients receiving or not salvage bone marrow transplantation had a 3-year survival of 48% and 18%, respectively (P<0.001). In a univariate Cox regression analysis, refractory disease was associated with a higher risk of death (HR=1.43, P=0.001), whereas late relapse (>12 months, HR 0.57, P=0.001) and salvage therapy with transplantation (HR=0.36, P<0.001) were associated with a better OS. No difference was found in OS with respect to histology. This study accurately reflects outcomes for patients treated according to standards of care worldwide. Results confirm that peripheral T-cell lymphomas patients had dismal outcome after relapse or progression. Patients with chemotherapy sensitive disease who relapsed after more than 12 months might benefit from consolidation bone marrow transplantation. (Registered at clinicaltrials.gov identifier: 01142674).

haematologica | 2018; 103(7)

Correspondence: monica.bellei@unimore.it

Received: December 18, 2017. Accepted: March 26, 2018. Pre-published: March 29, 2018.

doi:10.3324/haematol.2017.186577 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/103/7/1191 ©2018 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.

1191


M. Bellei et al.

Introduction The mature or peripheral T-cell lymphomas (PTCL) encompass a biologically and clinically heterogeneous group of rare neoplasia arising from post-thymic lymphocytes. They represent 10-15% of all lymphomas in the Western hemisphere.1 Peripheral T-cell lymphoma patients, except for anaplastic large cell lymphoma (ALCL), anaplastic lymphoma kinase (ALK)-positive, have a poor prognosis.2,3 Current treatment strategies are largely unsatisfactory both in firstline and in the refractory/relapsed settings. First-line therapy relies on CHOP (cyclophosphamide, doxorubicin, vincristine, prednisone) and CHOP-like regimens, with a remission rate of 50-65%.4-6 Phase II studies demonstrated that early consolidation with high-dose chemotherapy and stem cell rescue could improve outcome, but this approach is restricted to good performance status patients and chemotherapy responsive disease. For the majority of patients, risk of relapse remains quite high, and relapsed or refractory patients have been shown to have a dismal outcome.7,8 In recent years, there have been several studies testing novel therapies in this subset of patients.9 Two recent observational, population-based studies focusing on the outcome of relapsed or refractory PTCL patients have been published.7,8 The first, conducted by the British Columbia Cancer Agency, Canada, included 208 refractory or relapsed patients diagnosed between 1976 and 2010. The second study included 53 patients identified from the Modena Cancer Registry, Italy, with diagnosis confirmed between 1997 and 2010. Both showed extremely poor outcome with short remissions (median survival after relapse of 5.5 months and 2.5 months, respectively), and they confirmed that the outcome was superior in patients able to go forward for transplant. The International T-cell Project is an international prospective cohort study that enrolled patients at 74 academic centers on four continents. Data on epidemiology, clinical features, treatments and outcomes were collected. The purpose of the present study was to analyze clinical features and explore factors influencing survival of patients with primary refractory or relapsed PTCL.

Methods The T-Cell Project (TCP; registered at clinicaltrials.gov identifier: 01142674), sponsored by the International T-Cell Lymphoma Project (ITCLP), was set up in 2006, and builds on the retrospective study carried on by the network.2 Patients with different PTCL subtypes according to World Health Organization (WHO) 2001 or 2008 classifications1,10 were registered in the TCP at initial diagnosis. The T-Cell Project is a prospective cohort study that collected clinical and diagnostic information to better define clinical characteristics, therapies and prognosis for the most frequent subtypes of PTCL: PTCL not otherwise specified (PTCL-NOS) and angioimmunoblastic T-cell lymphoma (AITL). Further aims were to better outline clinical characteristics and outcome of the less common PTCL subtypes: extranodal NK/T-cell lymphoma, enteropathy-type T-cell lymphoma, hepatosplenic T-cell lymphoma, peripheral gamma-delta T-cell lymphoma, subcutaneous panniculitis-like T-cell lymphoma.11 Data were collected on frontline treatment, response evaluation and up-dated follow up for at least five years. Patients who did not receive any kind of treatment 1192

were also to be registered on the study. Data management was performed at the Trial Office in Modena, Italy (Department of Diagnostic, Clinical and Public Health Medicine, University of Modena and Reggio Emilia). Registration was based on locally established histological diagnosis. A panel of expert hematopathologists reviewed the diagnosis of 70% of all patients. Approximately 4% could not be adequately classified by central reviewers and were retained in the study with the diagnosis made by a local pathologist. Finally, for 26% of cases, samples were not centralized and these cases were evaluated on the basis of the local diagnosis. The TCP was conducted in compliance with the Declaration of Helsinki and was approved by the appropriate research ethics committees or institutional review boards at each participating institution. Each patient was required to provide written informed consent before registration.

End points The principal end point of the analysis was survival after relapse (SAR) for patients with primary refractoriness and those who relapsed, measured from the date refractoriness was documented/date of relapse until last follow up or death from any cause. Conventional response assessment after the first treatment has been adapted from the Standardized Response Criteria for NonHodgkin’s Lymphoma and from Recommendations for Revised Response Criteria for Malignant Lymphoma.12,13 Assessments were made by computed tomography (CT) scan or positron emission tomography (PET) scan according to physician’s discretion; responses were determined by the treating physician. For the present analysis, primary refractory disease was defined as no response or progression to initial treatment within one month from the end of initial therapy or unsatisfactory partial remission (PR), i.e. a PR that according to the physician’s judgment was considered to be inadequate for the patient, and thus requiring salvage therapy immediately after completion of front-line treatment. Relapsed disease was defined as progression at least one month from completion of front-line therapy in patients who achieved a complete remission (CR) or a satisfactory PR.

Statistical analysis Standard descriptive analyses were carried out. For a crude association analysis, categorical data were analyzed using the χ2 or Fisher’s exact test (two-sided) for data analysis. Survival curves were estimated using the Kaplan-Meier method, and compared using the log-rank test. Univariate regression analyses were conducted to identify prognostic factors associated with SAR. Odds ratios with their 95% confidence intervals (95%CI) were computed. Two-tailed P<0.05 was considered statistically significant. The Stata software, version 14·0 or greater (StataCorp, LLC, College Station, TX, USA; www.stata.com) was used for data analysis.

Results From September 2006 to July 2016, 1451 patients have been registered by 74 institutions worldwide. Among them, 1020 had baseline clinical data, information on firstline treatment, response to initial therapy, time to relapse and salvage treatment available for evaluation. At the time of diagnosis, 83 patients (8%) received only best supportive care and were excluded from this analysis. Out of 937 patients who received an active treatment, 633 (68%) were identified as refractory or relapsed patients, while 304 (32%) patients remained in complete remission. haematologica | 2018; 103(7)


Dismal outcome of refractory/relapsed PTCL

Among the 633 refractory/relapsed patients, 436 (69%) were classified as refractory and 197 (31%) as relapsed patients. The median time to relapse was 8 months (range 2-73 months). Among the relapsed patients, 125 (63%) presented with an early relapse (â&#x2030;¤12 months) and 72 (37%) presented with late relapse (>12 months). Main baseline patientsâ&#x20AC;&#x2122; characteristics of refractory/relapsed patients and all of the analyzed subset are shown in Table 1. The median age at diagnosis of refractory patients was 59 years (range 18-89 years) and that of relapsed patients was 58 years (range 21-88 years). Thirty-three percent of refractory patients and 16% of relapsed patients had ECOG performance status over 1 at diagnosis, respectively. A similar number of patients with AITL and PTCL had refractory (AITL: 16%, PTCL-NOS: 42%) or relapsed (AITL: 21%, PTCL-NOS: 42%) disease (Table 1). Patients with ALCL ALK- were more likely to have refractory disease than ALCL ALK+ (14% vs. 5%), but the frequency of relapsed disease was similar between both groups (11% vs. 7%). The majority of patients (n=844, 90%) received chemotherapy +/- radiotherapy as first-line treatment and 75 (8%) were consolidated with high-dose therapy (HDT) and hematopoietic cell transplantation (HCT) (Table 1). HCT was considered to be part of first-line therapy when it was given within six weeks from the end of the induction chemotherapy; in addition, patients who received HCT after six weeks from the end of initial chemotherapy,

Figure 1. Flow chart of patients included in the analysis. CR: complete remission.

Table 1. Main characteristics at diagnosis of 436 refractory and 197 relapsed patients, and of all 937 patients analyzed.

Parameter

Median age (range), years Age, >60 years, N, % Sex, male, N, % ECOG-PS, >1, N, % Serum LDH, > ULN, N, % [578] Ann Arbor staging, III-IV, N, % [591] ENS, >1, N, % [568] Serum albumin, <3.5 g/dL, N, % [562] NLR, >6.5, N, % [598] PIT, high-risk (2-4), N, % [503] IPI, high-risk (3-5), N, % [551] Histology subtype PTCL, NOS AITL ALCL, ALK ALCL, ALK + NKTCL Other 1st line therapy CHT +/- RT RT alone CHT/consolidation HCT Response to 1st line therapy CR PR <PR

Refractory (n=436)

Relapsed (n=197)

All (n=937)

N

%

N

%

N

%

59 (18-89) 203 273 143 234 325 148 204 125 172 184

58 (21-88) 47 63 33 54 75 34 47 29 39 42

56 (18-89) 90 136 31 82 128 43 70 48 52 53

46 69 16 42 65 22 36 24 26 27

401 579 214 404 605 250 359 231 283 294

43 62 23 43 65 27 38 25 30 31

185 70 60 21 35 65

42 16 14 5 8 15

83 42 22 14 21 15

42 21 11 7 11 8

346 154 140 77 109 80

37 16 15 8 12 9

422 2 12

97 <1 3

174 1 22

88 <1 11

844 18 75

90 2 8

137 299

31 69

170 27 -

86 14 -

474 164 299

51 18 32

ECOG-PS: perfomance status according to the Eastern Cooperative Oncology Group (ECOG) definition; LDH: lactate dehydrogenase; ULN: upper limit of normal; ENS: number of extra-nodal sites involved; NLR: neutrophil-to-lymphocyte ratio; PIT: Prognostic Index for T-cell lymphoma; IPI: International Prognostic Index; PTCL, NOS: peripheral T-cell lymphoma not otherwise specified; AITL: angioimmunoblastic T-cell lymphoma; ALCL: anaplastic large cell lymphoma; ALK: anaplastic lymphoma kinase; NKTCL: extra-nodal NK/T-cell lymphoma; CHT: chemotherapy; RT: radiotherapy; HCT: hematopoietic cell transplantation; CR: complete remission; PR: partial remission.

haematologica | 2018; 103(7)

1193


M. Bellei et al.

A

B

Figure 2. Survival after relapse (SAR). (A) SAR curve of 633 refractory/relapsed patients. (B) SAR by status: refractory versus relapse. Refractory patients are those with primary refractoriness.

Table 2. Details of treatment and events for the refractory/relapsed patients (n=633).

Parameter

Type of event Relapse after CR Relapse after PR Unsatisfactory PR Refractory (<PR) Timing of events Refractory Early relapse (â&#x2030;¤ 12 months) Late relapse (>12 months) HCT as Salvage HCT, yes No HCT (eligible for HCT) No HCT (CR/PR not eligible for HCT) No HCT (<PR not eligible for HCT) No HCT (whichever the reason)

Refractory (n=436)

Relapsed (n=197)

All (n=633)

N

%

N

%

N

%

137 299

31 69

170 27 -

86 14 -

170 27 137 299

27 4 22 47

436 -

69 -

125 72

20 11

436 125 72

69 20 11

42 125 269 394

9 29 62 91

57 124 16 140

29 63 8 71

99

16

534

84

CR: complete remission; PR: partial remission; Unsatisfactory PR: PR requiring immediate treatment after initial therapy; HCT: hematopoietic cell transplantation.

1194

haematologica | 2018; 103(7)


Dismal outcome of refractory/relapsed PTCL

but for whom the clinician specified in the planned treatment schedule that HCT was going to be given as consolidation, and who did not receive additional salvage therapies, were also considered to have received HCT as part of first-line therapy. Of those who received HCT, 41 patients were in first remission, 12 (3%) had refractory and 22 (11%) had relapsed disease. Details of salvage treatments are shown in Table 2. Overall, 99 patients (16%) received HCT as part of salvage treatment. Of those with refractory disease, 62% did not achieve at least a PR with salvage therapy and were therefore not eligible to undergo transplantation. Twenty-nine percent responded well to salvage therapy but were not considered candidates for transplantation. In the relapsed group, likewise, most of the patients (71%) did not undergo transplantation. Data on the reason why patients eligible for transplant were not referred to HCT consolidation was not collected; the choice as to whether the patient should go forward for transplant was at the physicianâ&#x20AC;&#x2122;s discretion.

As expected, patients responding to salvage therapy who proceeded to HCT had a better outcome compared to patients with no response (and therefore, ineligible for HCT) and to patients in CR/PR not eligible for HCT (for any reason), with 3-year survival rates of 48%, 7% and 30%, respectively. Similarly, patients proceeding to HCT had significantly better outcome than patients who were eligible but did not undergo HCT for any reason (3-year SAR 48% and 27%).Overall, patients who received HCT had a better outcome with respect to the subset of patients who did not (3-year SAR 48% and 18%, respectively) (P<0.001) (Table 3 and Figure 4). In a univariate Cox regression analysis, refractory disease was associated with a higher risk of death compared to relapsed patients (HR 1.43, 95%CI: 1.16-1.76; P=0.001), whereas late relapse compared to early relapse (HR 0.57, 95%CI: 0.41-0.79; P=0.001) and salvage therapy with HCT compared to no HCT (HR 0.36, 95%CI: 0.26-0.48; P<0.001) were associated with a longer SAR (Table 3).

Survival after relapse After a median follow up of 38 months (range 1-96 months) from documentation of refractory/relapsed disease, 440 (70%) patients had died. The median survival after relapse (SAR) was 5.8 months (95%CI: 4.9-7.2 months) and 3-year SAR was 23% (95%CI: 19-27) (Figure 2A). Median SAR for refractory and relapsed patients were 5 and 11 months, respectively, and 3-year SAR rates were similar for both groups at 21% (95%CI: 17-25) and 28% (95%CI: 21-35), respectively (Figure 2B). Univariate analysis showed that in the first 24 months refractory patients had a poorer outcome with respect to relapsed patients [Hazard Ratio (HR) HR 1.50, 95%CI: 1.12-1.86; P<0.001], while after 24 months their outcome became similar to that of the relapsed group (HR 0.75, 95%CI: 0.34-1.64; P=0.470). (Figure 2B). No difference was found in outcomes for refractory/relapsed patients with respect to PTCL subtype, with the exception of ALCL ALK+ (Figure 3).

Table 3. Univariate Cox regression analysis for SAR.

Status Relapse Refractory Early relapse (â&#x2030;¤ 12 months) Late relapse (> 12 months) Not eligible to HCT <PR Not eligible to HCT (CR/PR) Eligible HCT (CR/PR) HCT No HCT at salvage HCT at salvage

3-year SAR%(95%CI)

HR (95%CI)

28 (21-35) 21 (17-25) 23 (16-32) 34 (21-48) 7 (3-11) 30 (21-38) 27 (19-36) 48 (37-58) 18 (14-22) 48 (37-58)

1.00 1.43 (1.16-1.76) 1.00 0.57 (0.41-0.79) 1.00 0.43 (0.34-0.55) 0.45 (0.35-0.58) 0.22 (0.16-0.30) 1.00 0.36 (0.26-0.48)

CR: complete remission; PR: partial remission; SAR: survival after relapse; HCT: hematopoietic cell transplantation; HR: Hazard Ratio; CI: Confidence Interval

Figure 3. Outcomes for refractory/relapsed patients depending on histological subtypes. PTCL-NOS: peripheral T-cell lymphoma not otherwise specified; AITL: angioimmunoblastic T-cell lymphoma; ALCL (-): anaplastic large cell lymphoma, anaplastic lymphoma kinase negative; ALCL (+): anaplastic large cell lymphoma, anaplastic lymphoma kinase positive; NKTCL: extranodal NK/T-cell lymphoma.

haematologica | 2018; 103(7)

1195


M. Bellei et al.

Figure 4. Survival after relapse (SAR) by salvage therapy including or not hematopoietic cell transplantation (HCT). Elig: eligible; Not elig: not eligible; CR: complete remission; PR: partial remission.

Discussion The International TCP represents the largest cohort of prospectively collected data on patients with aggressive Tcell lymphomas and accurately reflects outcomes for patients treated according to standards of care around the world. In the present study, we sought to analyze the outcomes of patients with relapsed and refractory disease after failure of first-line therapy and to explore potential prognostic factors influencing survival, retrieved from this database. We demonstrated that the outcomes are worse for patients with refractory disease and that the SAR at three years for these patients was only 21%. We also found that late relapse and consolidation with HCT were associated with a longer survival in chemotherapy sensitive patients. In our analysis, refractory/relapsed patients presented a poor risk profile. Sixty-nine percent of failing patients were refractory to first-line treatment and 80% of refractory and 70% of relapsed patients had advanced stage disease at diagnosis. Fifty percent of refractory and 33% of relapsed patients were at high risk according to the Prognostic Index for T-cell lymphoma (PIT); 50% and 32% of relapsed and refractory patients, respectively, were at high-risk according to the International Prognostic Index (IPI). Survival was poor in our cohort, with a median SAR for refractory and relapsed patients of only 5 and 11 months, respectively. The results from this prospective cohort confirm findings from other reports that refractory disease is a poor prognostic factor.7,8 While late relapse occurring after 12 months versus early relapse at less than 12 months from front-line therapy was associated with a longer survival, only 11% of patients were in the late relapse category as most relapses occurred within one year from frontline therapy. Surprisingly, there was no difference in outcomes for refractory/relapsed patients with respect to PTCL subtype, suggesting that significant improvements are needed in treatment strategies for all subtypes of 1196

PTCL. We were surprised to find similar survival rates between relapsed and refractory PTCL patients at three years post completion of therapy. However, further analysis showed that within the first 24 months of follow up relapsed patients had superior survival, and it is only past that time point that the advantage disappeared. These results suggest that within a category of relapsed patients there is a subgroup with biologically refractory disease and current definitions based on clinical responses are not sensitive enough to identify individuals that would benefit from alternative approaches rather than standard salvage protocols. Furthermore, only about half of the refractory PTCL patients exhibited clinical high risk based on IPI or PIT scores at diagnosis. Emerging genome-wide analysis at diagnosis and/or relapse might overcome these restrictions and provide a better guide for initial and salvage therapy in the near future. In the relapsed and refractory setting, the best chance of long-term remission and best outcomes occurred in patients with late (>12 months) relapse who were able to undergo HDT followed by HCT, with SAR at three years of 48%. However, a major problem remains: only 16% of the patients could proceed to this strategy as part of the salvage treatment due to refractoriness to induction therapy, early relapses, ineffective salvage therapies, and overall poor performance status and patient-specific factors. Two recent population-based retrospective studies focusing on the outcome of relapsed or refractory PTCL patients have been published and they reported a similar poor outcome (median survival after relapse of 5.5 months and 2.5 months).7,8 Likewise, the outcome was far superior in patients able to receive a transplant. Taken together, the current challenge remains to increase the response rates of induction therapies to raise the number of eligible patients for the most effective available intent-to-treat treatment. Moreover, these results highlight the urgent need for novel agents and more effective salvage therapies. Advances in understanding the biology and genetics of T-cell lymphomas have led to the identification of several haematologica | 2018; 103(7)


Dismal outcome of refractory/relapsed PTCL

potential novel targets.14-17 Recently, four new-generation drugs have been approved in the USA in refractory/relapsed TCL: pralatrexate (antifolate), romidepsin and belinostat (histone deacetylase inhibitor), and brentuximab vedotin. In addition to these approved drugs, a number of novel drugs with different mechanisms are under investigation: crizotinib (oral tyrosine kinase inhibitor of ALK), mogamulizumab, duvelisib, plitidepsin, and selinexor.9,18 Hopefully, these agents will have an impact both when combined with front-line chemotherapy as well as in the relapsed and refractory setting. Although the patient cohort could be not completely homogeneous (Investigators were requested to register consecutive cases satisfying the inclusion criteria without selection), the amplitude of the TCP reflects the realworld scenario, obtained through a multi-national database describing the distribution of PTCL subtypes and therapeutic outcomes with standard therapies. Our results complement those of the COMPLETE registry (clinicaltrials.gov identifier: 01110733), a similar prospective

References 1. Swerdlow S, Campo E, Harris N, et al. WHO Classification of Tumours of Haematopoietic and Lymphoid Tissues. 4th ed. Lyon, France: IARC Press; 2008. 2. Vose J, Armitage J, Weisenburger D. International peripheral T-cell and natural killer/T-cell lymphoma study: pathology findings and clinical outcomes. J Clin Oncol. 2008;26(25):4124-4130. 3. Adams SV, Newcomb PA, Shustov AR. Racial Patterns of Peripheral T-Cell Lymphoma Incidence and Survival in the United States. J Clin Oncol. 2016;34(9):963-971. 4. Melnyk A, Rodriguez A, Pugh WC, Cabannillas F. Evaluation of the Revised European-American Lymphoma classification confirms the clinical relevance of immunophenotype in 560 cases of aggressive non-Hodgkin’s lymphoma. Blood. 1997;89(12):4514-4520. 5. Lopez-Guillermo A, Cid J, Salar A, et al. Peripheral T-cell lymphomas: initial features, natural history, and prognostic factors in a series of 174 patients diagnosed according to the R.E.A.L. Classification. Ann Oncol. 1998;9(8):849-855. 6. Gisselbrecht C, Gaulard P, Lepage E, et al. Prognostic significance of T-cell phenotype in aggressive non-Hodgkin’s lymphomas.

haematologica | 2018; 103(7)

7.

8.

9.

10.

11.

12.

study of PTCL patients in the US. These results will provide a useful baseline on which to assess the efficacy of novel agents and therapies for refractory/relapsed patients with T-cell lymphomas. Clinical trials are underway exploring the activity of novel agents in combination with chemotherapy to improve overall response in the front line, and single agent and combination studies of novel agents are underway for patients with refractory/relapsed disease. Acknowledgments For supporting this study, the authors would like to thank the Fondazione Cassa di Risparmio di Modena, Modena, Italy, the Associazione Angela Serra per la Ricerca sul Cancro, Modena, Italy, the Fondazione Italiana Linfomi (FIL), Alessandria, Italy, Allos Therapeutics, Inc., Westminster, CO, USA, and Spectrum Pharmaceuticals Inc., Henderson, NV, USA, AIRC (Associazione Italiana per la Ricerca sul Cancro) 5x1000 (grant n. 10007 to Stefano Pileri), the NIH/NCI CCSG P30 CA008748 (grant to Steven Horwitz).

Groupe d’Etudes des Lymphomes de l’Adulte (GELA). Blood. 1998;92(1):76-82. Mak V, Hamm J, Chhanabhai M, et al. Survival of Patients With Peripheral T-Cell Lymphoma After First Relapse or Progression: Spectrum of Disease and Rare Long-Term Survivors. J Clin Oncol. 2013;31(16):1970-1976. Biasoli I, Cesaretti M, Bellei M, et al. Dismal outcome of T-cell lymphoma patients failing first-line treatment: results of a population-based study from the Modena Cancer Registry. Hematol Oncol. 2015;33(3):147-151. Zinzani PL, Bonthapally V, Huebner D, Lutes R, Chi A, Pileri S. Panoptic clinical review of the current and future treatment of relapsed/refractory T-cell lymphomas: Peripheral T-cell lymphomas. Crit Rev Oncol Hematol. 2016;99214-227. Jaffe ES, Harris NL, Stein H, Vardiman JW. WHO classification of Tumours of Haematopoietic and Lymphoid Tissues: Pathology and Genetics. Lyon, France: IARC Press; 2001. Federico M, Bellei M, Pesce E, et al. T-Cell Project: an international, longitudinal, observational study of patients with aggressive peripheral T-cell lymphoma. Rev Bras Hematol Hemoter. 2009;31(2):21-25. Cheson BD, Horning SJ, Coiffier B, et al. Report of an international workshop to

13. 14.

15.

16.

17.

18.

standardize response criteria for nonHodgkin’s lymphomas. NCI Sponsored International Working Group. J Clin Oncol. 1999;17(4):1244. Cheson BD, Pfistner B, Juweid ME, et al. Revised response criteria for malignant lymphoma. J Clin Oncol. 2007;25(5):579-586. Piccaluga PP, Fuligni F, De Leo A, et al. Molecular profiling improves classification and prognostication of nodal peripheral Tcell lymphomas: results of a phase III diagnostic accuracy study. J Clin Oncol. 2013;31(24):3019-3025. Iqbal J, Wright G, Wang C, et al. Gene expression signatures delineate biological and prognostic subgroups in peripheral Tcell lymphoma. Blood. 2014;123(19):29152923. O’Connor OA, Bhagat G, Ganapathi K, et al. Changing the paradigms of treatment in peripheral T-cell lymphoma: from biology to clinical practice. Clin Cancer Res. 2014;20(20):5240-5254. Inghirami G, Chan WC, Pileri S. Peripheral T-cell and NK cell lymphoproliferative disorders: Cell of origin, clinical and pathological implications. Immunol Rev 2015;263 (1):124–159. Coiffier B, Federico M, Caballero D, et al. Therapeutic options in relapsed or refractory peripheral T-cell lymphoma. Cancer Treat Rev. 2014;40(9):1080-1088.

1197


ARTICLE

Chronic Lymphocytic Leukemia

Ferrata Storti Foundation

Haematologica 2018 Volume 103(7):1198-1208

Low-count monoclonal B-cell lymphocytosis persists after seven years of follow up and is associated with a poorer outcome

Ignacio Criado,1 Arancha Rodríguez-Caballero,1 M. Laura Gutiérrez,1 Carlos E. Pedreira,2 Miguel Alcoceba,3 Wendy Nieto,1 Cristina Teodosio,1 Paloma Bárcena,1 Alfonso Romero,4 Paulino Fernández-Navarro,5 Marcos González,3 Julia Almeida,1* Alberto Orfao1* and The Primary Health Care Group of Salamanca for the Study of MBL

Cancer Research Centre (IBMCC, USAL-CSIC), Department of Medicine and Cytometry Service (NUCLEUS), University of Salamanca, IBSAL and CIBERONC, Spain; 2Systems and Computing Department (PESC), COPPE, Federal University of Rio de Janeiro (UFRJ), Brazil; 3Hematology Service, University Hospital of Salamanca, IBMCC, IBSAL, CIBERONC and Department of Nursery and Physiotherapy, University of Salamanca, Spain; 4Centro de Atención Primaria de Salud Miguel Armijo, Salamanca, Sanidad de Castilla y León (SACYL), Spain and 5Centro de Atención Primaria de Salud de Ledesma, Salamanca, Sanidad de Castilla y León (SACYL), Spain 1

*AO and JA contributed equally to this work.

ABSTRACT

L

Correspondence: orfao@usal.es

Received: November 3, 2017. Accepted: March 15, 2018. Pre-published: March 22, 2018. doi:10.3324/haematol.2017.183954 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/103/7/1198 ©2018 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.

1198

ow-count monoclonal B-cell lymphocytosis is defined by the presence of very low numbers of circulating clonal B cells, usually phenotypically similar to chronic lymphocytic leukemia cells, whose biological and clinical significance remains elusive. Herein, we re-evaluated 65/91 low-count monoclonal B-cell lymphocytosis cases (54 chronic lymphocytic leukemia-like and 11 non-chronic lymphocytic leukemialike) followed-up for a median of seven years, using high-sensitivity flow cytometry and interphase fluorescence in situ hybridization. Overall, the clone size significantly increased in 69% of low-count monoclonal B-cell lymphocytosis cases, but only one subject progressed to high-count monoclonal B-cell lymphocytosis. In parallel, the frequency of cytogenetic alterations increased over time (32% vs. 61% of cases, respectively). The absolute number of the major T-cell and natural killer cell populations also increased, but only among chronic lymphocytic leukemia-like cases with increased clone size vs. age- and sex-matched controls. Although progression to chronic lymphocytic leukemia was not observed, the overall survival of low-count monoclonal B-cell lymphocytosis individuals was significantly reduced vs. non-monoclonal Bcell lymphocytosis controls (P=0.03) plus the general population from the same region (P≤0.001), particularly among females (P=0.01); infection and cancer were the main causes of death in low-count monoclonal B-cell lymphocytosis. In summary, despite the fact that mid-term progression from low-count monoclonal B-cell lymphocytosis to high-count monoclonal B-cell lymphocytosis and chronic lymphocytic leukemia appears to be unlikely, these clones persist at increased numbers, usually carrying more genetic alterations, and might thus be a marker of an impaired immune system indirectly associated with a poorer outcome, particularly among females.

Introduction Chronic lymphocytic leukemia (CLL) is the most common leukemia in adults in the Western world, typically affecting older patients, particularly males, with a median age at diagnosis of 70 years (y) old.1 It is characterized by the accumulation of mature B cells in peripheral blood (PB), bone marrow (BM) and also secondary lymphoid tissues, with a uniquely aberrant CD19+ CD20+lo CD5+/++ CD23+ sIgM-/+lo phenotype and restricted immunoglobulin (Ig) light chain usage.2,3 Typically, CLL haematologica | 2018; 103(7)


7y follow-up of monoclonal B-cell lymphocytosis

shows a heterogeneous clinical outcome; thus, whereas in some patients the disease remains stable and they will never require treatment, in around 70% of cases treatment is required and results in variable outcomes, from complete response and prolonged survival to refractory disease and death.3–5 Currently, it is well established that virtually every CLL case is preceded by monoclonal B-cell lymphocytosis (MBL) defined by smaller numbers of circulating PB clonal CLL-like B-cells (<5,000 clonal B-cells/mL) in the absence of any clinical symptoms or signs of disease.6 In 2010, MBL was further subdivided into low-count (MBLlo) and high-count MBL (MBLhi), depending on the number of PB clonal B cells (lower vs. higher than 0.5x109/L, respectively).7 While MBLhi has been reported to progress to overt CLL requiring treatment at a rate of 1–2% cases per year,8,9 no information is available at present regarding the ≥5year risk of progression of MBLlo to MBLhi and CLL.10 The detection of MBLlo has become routinely feasible due to the use of highly sensitive flow cytometry (FCM) approaches for the screening of subjects from the general population who present normal blood cell counts. Of note, the prevalence of MBLlo is significantly higher than that of MBLhi and CLL, with a frequency that ranges between 3% and 14% of the general adult (≥40y) population, depending on the sensitivity of the FCM technique used.11 Independently of the method, it is well-established that the incidence of MBLlo progressively increases with age, with a prevalence >20% among individuals of more than 70 years of age.12 Whether MBLlo represents the normal counterpart of CLL (e.g., some studies suggest that MBLlo clones are more likely related to immunosenescence)13 or a very early stage of development of CLL, remains an open question. This is partially because, in contrast to MBLhi, long-term follow-up studies in large series of MBLlo cases have not been reported thus far, which limits our understanding of the biological and clinical significance of very low numbers of circulating CLLlike clones, as well as those factors and mechanisms involved in potential long-term progression of (conceivably) a minor proportion of all MBLlo cases to MBLhi and CLL; likewise, little information is available about the evolution of non CLL-like MBL. Such information is critical to a better understanding of the ontogenesis of CLL from the very early stages of the disease, and to better identify MBL patients with stable vs. progressive B-cell lymphocytosis who might benefit from a closer clinical follow-up. Herein, we report on a cohort of 91 MBLlo (CLL-like and non CLL-like) subjects identified in a population-based screening study and followed for a minimum of five years (median >seven years). Our primary goal was to determine the rate of medium-term progression of MBLlo to MBLhi and CLL, and to identify the most relevant clinical and biological characteristics of PB lymphocytes associated with progression.

sensitive FCM.12,14 At inclusion, all subjects had normal PB cell counts and did not suffer from any hematological/immunological disease, as described elsewhere.1,6 In 91/639 subjects studied (14.2%), ≥1 PB clonal B-cell population was detected at recruitment; in the vast majority of them (80/91; 88%) clonal B cells were consistent with CLL-like MBLlo (<0.5x109clonal B cells/L showing a CLL-like phenotype), whereas the remaining 11 individuals (12%) were classified as non CLL-like MBLlo.12,14 MBLlo subjects were re-evaluated at a median time of seven years after recruitment (range: 61 to 95 months). All subjects gave their written informed consent at baseline for both the initial and the follow-up studies, and they filled out an epidemiological questionnaire with demographic and (self-reported) medical information, under the supervision of his/her primary care doctor.15 The study was approved by the Ethics Committee of the University Hospital of Salamanca (Spain).

Flow cytometry immunophenotypic studies Overall, 1-4 mL of ethylenediamine tetraacetic acid (EDTA)anticoagulated PB was collected per case and follow-up timepoint; subsequently it was processed and analyzed using previously reported highly sensitive FCM approaches12,14,16,17 (Online Supplementary Methods and Online Supplementary Table S1).

Interphase fluorescence in situ hybridization (iFISH) studies studies The most common CLL - i.e., del(13q14), trisomy 12, del(11q)(ATM) and del(17p)(TP53) - along with other B-cell chronic lymphoproliferative disorders (B-CLPD)-associated cytogenetic alterations were investigated by iFISH on fluorescence-activated cell sorting (FACS)-purified (sorted) single clonal B cells (≥95% purity), as previously described18 (Online Supplementary Table S2). A total of 31/91 PB samples studied at baseline and 56/65 at follow-up (year +7) were analyzed by iFISH; in 21 cases (18 CLL-like and three non CLL-like MBLlo) paired samples were analyzed by iFISH at both baseline and year +7. The potential presence of del(13q14) was also tested in non-clonal B-cells from 5/7 MBLlo cases found to have del(13q14)+ MBL cells.

Statistical analyses All conventional statistical analyses (i.e., descriptive statistics, univariate analyses, including overall survival (OS) analysis, as well as multivariate analyses to predict the variables independently associated with a greater/lower risk of death), were performed with SPSS 19.0 software (SPSS-IBM, Armonk, NY, USA), using the tests, databases and statistical significance values detailed in Online Supplementary Methods. Appropriate tests were further used to objectively evaluate real changes in the size of the B-cell clones studied during follow-up (resampling bootstrap method)19 and to build a predictive linear regression model to estimate the time CLL-like MBLlo clones might potentially take to progress to MBLhi and CLL, using MATLAB R2015a (Mathworks, Natick, MA, USA) (Online Supplementary Methods).

Results Methods Subjects and samples The baseline study was conducted from December 2007 to October 2009, when PB samples from 639 healthy adult (≥40y) volunteers (54% females/46% males) from the general population of the same geographical area (Salamanca, Northwest of Spain) were screened for the presence of small B-cell clones, using highly haematologica | 2018; 103(7)

Follow-up of the MBLlo cohort From those 91 MBLlo individuals identified in the screening study performed in the general population of Salamanca between 2007 and 2009,12,14 65 -71% of MBLlo cases from the original series; 29 males and 36 females; median age at baseline 70 (range: 43-84 years old)-; were re-evaluated after a median follow-up of seven years (range: 61 to 95 months) (Table 1). These 65 individuals 1199


I. Criado et al.

In all 65 individuals who were evaluated after five years, ≥1 clonal B-cell population was reliably identified in PB at follow-up. In 22/65 (34%) cases ≥2 clones were detected (vs. 32% at baseline), resulting in a total of 86 MBLlo clones detected (Table 1 and Table 2). All MBLlo clones showed an identical phenotype at both time-points (Table 2). Thus, 74/86 B-cell clones (86%) showed a classical CLL-like phenotype and 12 (14%) were classified as non CLL-like MBL clones. At year +7, 35/74 CLL-like clones (47%) corresponded to monoclonal cases and the remaining 39 (53%), to 19 subjects with bi(multi)clonal CLL-like MBLlo (Table 2); in two subjects, CLL-like and non CLL-like clones

were representative of the original MBLlo cohort for all variables analyzed, except for a significantly lower age (P=0.02) vs. those 26 individuals that could not be followed - median age of 75 (range: 48-95 years)-. These later subjects could only be re-evaluated for their death vs. alive status at the end of the study because of: i) 12/26 (46%) died before the fifth year of follow-up; ii) 2 subjects declined continuing their participation in the study; and iii) the remaining 12 cases were lost to follow-up after >5y from recruitment. Eight of 65 cases followed for >5y (12%) died afterward, making a total of 21 (26%) deaths among MBLlo cases included in OS analyses.

Table 1. Clinical and biological characteristics of MBLlo subjects at baseline and after follow-up (year +7).

All subjects (n=65)

Follow-up time (months) Male/Female* Age, years Leukocytosis (>10x109/L)* Lymphocytosis (>4x109/L)* N. total T cells/mL N. CD4+ T cells/mL N. CD8+ T cells/mL N. CD4+/CD8+ T cells/mL N. CD4–/CD8– T cells/mL N. NK cells/mL N. total B cells/mL N. normal B cells/mL N. clonal B cells/ mL Subjects with ≥2 MBL clones* Progression * (to MBLhi) Deaths*

Baseline

Follow-up

0

84 (61-95)

29/36 (45%/55%) 70 (43-84) 0 (0%) 0 (0%) 1261 (341-2428) 687 (253-1572) 449 (71-1154) 4.3 (0.19-38) 56 (8.1-254) 304 (76-1138) 133 (26-1173) 119 (23-478) 0.99 (0.03-1101) 21 (32%) NA NA

75 (49-91) 2 (3%) 3 (5%) 1448 (276-3753) 840 (184-2045) 491 (66-1742) 8.2 (1.1-147) 62 (1.9-407) 373 (87-3415) 155 (22-1218) 116 (21-536) 2.0 (0.05-1149) 22 (34%) 1 (2%) 8 (12%)

CLL-like MBLlo subjects (n=54) Baseline Follow-up 0 22/32 (41%/59%) 68 (43-84) 0 (0%) 0 (0%) 1290 (341-2428) 684 (253-1572) 446 (71-1154) 4.5 (0.55-37) 58 (8.0-214) 297 (76-1138) 132 (41-478) 126 (37-478) 0.75 (0.03-66) 18 (33%) NA NA

84 (61-95)

75 (49-91) 2 (3%) 2 (4%) 1508 (460-3753) 898 (227-2045) 479 (96-1742) 8.2 (1.3-147) 64 (1.9-338) 373 (89-3415) 150 (28-1218) 140 (26-536) 1.7 (0.05-808) 19 (35%) 1 (2%) 7 (13%)

Non CLL-like MBL subjects (n=11) Baseline Follow-up 0 7/4 (64%/36%) 76 (58-81) 0 (0%) 0 (0%) 1111 (796-1965) 732 (351-1395) 453 (237-750) 4.3 (0.19-27) 36 (11-254) 394 (150-848) 137 (26-1173) 72 (23-136) 56 (0.62-1101) 3 (27%) NA NA

83 (63-87)

P

NA NA

83 (65-88) 0 (0%) 1 (9%) 1206 (276-2907) 629 (184-1995) 617 (66-848) 8.2 (1.1-29) 34 (8.1-406) 372 (178-937) 190 (22-1207) 45 (21-190) 90 (1.3-1149) 3 (27%) 0 (0%) 1 (9%)

<0.01a,b,c NS NS <0.01a,b 0.015a,b <0.03a,b <0.02a,b,c <0.05a,b 0.001a,b <0.01a,b 0.08b <0.001a,b NS NA NA

CLL-like or non CLL-like with ≥1 B-cell clone with different phenotypes were classified depending on the phenotype of the larger clone. Results expressed as median (range) or *as number of cases (percentage). aBaseline vs. follow-up (year +7) for all cases. bBaseline vs. follow-up (year +7) for CLL-like MBL cases. cBaseline vs. follow-up (year +7) for non CLL-like MBL cases. CLL: chronic lymphocytic leukemia; MBLhi: high-count monoclonal B-cell lymphocytosis; MBLlo: low-count monoclonal B-cell lymphocytosis; N: number; NA: not applicable; NK: natural killer; NS: not statistically significantly different (P>0.05).

1200

haematologica | 2018; 103(7)


7y follow-up of monoclonal B-cell lymphocytosis

coexisted. Of note, two individuals carrying two CLL-like B-cell clones became “monoclonal” while a second clone emerged in one monoclonal CLL-like MBLlo case at seven years follow-up. In turn, non CLL-like clones (n=12) showed phenotypic profiles identical to those observed at baseline and comparable to those of different B-CLPD, as detailed in Online Supplementary Table S3.6

Clonal B-cell load in PB at re-evaluation (year +7). Overall, a significant (P≤0.001) increase in the median size of MBLlo clones was found at follow-up, both for CLL-like (≈2-fold median increase) and for non CLL-like MBLlo clones (≈3-fold median increase) (Table 2 and Figure 1A,B). Such increased absolute number of clonal B-cells over time was associated with a significantly increased (P≤0.001) percentage of clonal B cells from all PB B cells (Table 2). In detail, most MBLlo clones (59/86; 69%) showed significantly increased numbers at re-evaluation vs. baseline, while the remaining 27 B-cell clones persisted at similar (16%) or lower levels (15%); this behavior was very similar for CLL-like and non CLL-like clones (Table 2). Of note, 30/35 (86%) CLL-like clones from (mono)clonal cases increased in size at follow-up vs. only 21/39 (54%) clones from bi(multi)clonal cases (P=0.004). Interestingly, among non CLL-like clones, most marginal zone lymphoma-like clones increased (5/6; 83%), while the two mantle cell lymphoma-like B-cell clones decreased significantly in number (Online Supplementary Table S3).

Cytogenetic alterations of MBLlo clonal B cells at baseline and follow-up The overall frequency of CLL-like MBLlo cases carrying CLL-associated cytogenetic alterations, for example del(13q14), trisomy 12, del(11q)(ATM) and del(17p)(TP53), at baseline was of 29% (7/24 cases tested). At recruitment del(13q14)(D13S25) was found in 56%±34% cells from 6/20 cases evaluated (30%), the RB1 gene was additionally involved in 3 of them, and trisomy 12 was present in the remaining case (59% of cells), both as single alterations. After seven years of follow-up, the percentage of cytogenetic altered cases augmented to 62% of MBLlo cases (31/50 cases, including 15 cases studied at baseline). Interestingly, all cytogenetic alterations observed at base-

line also remained at follow-up; in addition, 4/15 (27%) individuals studied at both time-points further acquired del(13q14)(D13S25) (Online Supplementary Table S4). Overall, del(13q14)(D13S25) remained the most frequent alteration at follow-up (27/48; 56%), affecting 32±27% of CLL-like cells. Of note, in five cases in which clonal B cells showed del(13q14)(D13S25), non-clonal B cells were also studied for this alteration, and was found to be absent in all of them. RB1 gene involvement was identified in only 1/7 cases tested; furthermore, trisomy 12 was restricted to one patient who had the same abnormality at baseline (Table 3). Clonal B cells from one individual in whom del(17p)(TP53) was not investigated at baseline was found to carry this cytogenetic alteration in 10% of cells at follow-up. Alterations involving 14q32 were investigated only at follow-up in a subset of 20 CLL-like MBLlo cases, being found in five (20%) patients (Table 3). Regarding non CLL-like clones, t(11;14)(q13-q32) was detected in 100% of clonal B cells from one of the two MCL-like cases studied, while del(7q32) was detected in 2/5 splenic marginal zone lymphoma (SMZL)-like cases (Table 3). None of the cases investigated showed t(14;18) (data not shown).

Distribution of normal residual T-, B- and NK-cell populations The PB counts of total T cells and their CD4+CD8–, CD8+CD4– and CD4–CD8-/lo subsets, as well as NK cells and normal residual polyclonal B cells was significantly increased (P<0.05) in CLL-like MBLlo at follow-up vs. baseline (Table 1). In contrast, among non CLL-like MBL cases, CD4+CD8+ T cells were the only lymphoid subset significantly increased (P=0.02) at the seven year follow-up. To rule out a potential age-related bias and further confirm these findings, we compared the number of PB normal lymphocyte subsets at seven years follow-up vs. a large series of non-MBL healthy donors matched per age and sex distribution to the CLL-like MBLlo cases at seven years (Online Supplementary Table S5) and the same differences were found, ruling out an impact of sex or more advanced age on the increased PB residual lymphocyte counts. No significant correlation (P>0.05) was revealed between the absolute number of clonal B cells and any of the normal residual PB lymphocyte subsets analyzed (data not shown).

Table 2. Biological characteristics of MBLlo clones at baseline and at follow-up (year +7).

All clones (n=86) Baseline Follow-up N. of clones from monoclonal/Bi(multi)clonal subjects* N. of clones that increased* N. clonal B cells/mL % clonal B cells (from total B cells)

44/42 (51%/49%) NA 0.06 (0.03-1101) 0.48% (0.02%-94%)

42/44 (49%/51%) 59 (69%) 1.3 (0.05-1146) 0.95% (0.02%-97%)

CLL-like MBLlo clones (n=74) Baseline Follow-up 36/38 (49%/51%) NA 0.46 (0.03-66) 0.35% (0.02%-21%)

35/39 (47%/53%) 51 (69%) 0.85 (0.05-789) 0.73% (0.02%-65%)

Non CLL-like MBL clones (n=12) Baseline Follow-up 8/4 (67%/33%) NA 37 (0.57-1101) 30% (0.46%-94%)

7/5 (54%/46%) 8 (67%) 68 (1.3-1146) 60% (1.4%-97%)

P

NS NA <0.001a,b <0.03a,b,c

Results expressed as median (range) or as * number of cases (percentage). aBaseline vs. follow-up (year +7) for all cases. bBaseline vs. follow-up (year +7) for CLL-like MBL clones. cBaseline vs. follow-up (year +7) for non CLL-like MBL clones. CLL: chronic lymphocytic leukemia; MBLlo: low-count monoclonal B-cell lymphocytosis; N: number; NA: not applicable; NS: not statistically significantly different (P>0.05).

haematologica | 2018; 103(7)

1201


I. Criado et al.

Clinical and biological characteristics of CLL-like MBLlo at baseline vs. follow-up, according to the kinetics of the B-cell clone Upon comparing CLL-like MBLlo cases with increased vs. stable/decreased clonal B-cell numbers at seven years follow-up, the former had a similar male/female distribution, but they were significantly younger (median age: 68y vs. 78y; Table 4). Strikingly, MBLlo cases who showed larger CLL-like clone sizes over time also showed significantly higher (P<0.05) numbers of the distinct normal residual T-, B- and NK-cell subsets at follow-up (vs. baseline) (Table 4). Moreover, in these subjects a direct correlation was observed between the absolute number of clonal B cells and CD4+CD8â&#x20AC;&#x201C; T cells (r2=0.5; P=0.001). In contrast, no significant (P>0.05) association was found between higher numbers of clonal CLL-like B cells in PB over time, and an increased frequency of cytogenetic alterations. Interestingly, del(13q14) was the sole genetic alteration detected at the seven year follow-up within cases with stable/decreased CLL-like B-cell clones, while those cases with increased CLL-like B-cell clones at year +7 showed cytogenetic alterations other than del(13q14), e.g., trisomy 12 (1/40), del(17p)(TP53) (1/39) and t(14q32) (5/20 cases tested) (Table 4).

Clinical outcome of MBLlo cases Three subjects developed absolute lymphocytosis after seven years of follow-up (median: 5.3x109 lymphocytes/L; range: 4.1x109-5.9x109/L) in the absence of signs of disease. Two had CLL-like B-cell clones carrying del(13q14), while the remaining case had a non CLL-like clone. In one of the two CLL-like MBLlo cases, the size of the B-cell clone increased over the threshold for MBLhi (>500 clonal B cells/mL), while the other two cases remained as MBLlo. Remarkably, these three subjects displayed the highest increase in clone size at re-evaluation: this translated into a significantly lower (estimated) time to progression into CLL (median: 95y; range: 54-128y) according to the predictive mathematical model used. In turn, the estimated time to progression to CLL for the other MBLlo individuals was far beyond a normal life expectancy (median: 54,767y; range: 54->63 million years).

Overall survival of MBL vs. non-MBL individuals At the end of the study (January 2017), the clinical records and epidemiological questionnaires from all individuals recruited at baseline were reviewed. During follow-up, 21/89 (24%) MBLlo cases and 41/290 (14%) age- and sexmatched non-MBLlo subjects from the original cohort had died (P=0.03). Though the median OS for the two groups

A

B

1202

Figure 1. Changes in the number of clonal B cells during follow-up. Panel A shows the absolute number of PB clonal B cells/mL detected in MBLlo individuals at baseline and at follow-up, according to the phenotype of the clonal population. Panel B represents the foldchange in the number of clonal B cells/mL from baseline, which is represented by the horizontal light gray box. Notched boxes represent 25th and 75th percentile values; the lines in the middle correspond to median values (50th percentile) and vertical lines represent the highest and lowest values that are neither outliers nor extreme values, which are represented as single dots. ***P-value <0.001. N: number; CLL: chronic lymphocytic leukemia.

haematologica | 2018; 103(7)


7y follow-up of monoclonal B-cell lymphocytosis

had not been reached yet, a significantly shorter OS was observed for MBLlo individuals vs. age- and sex-matched non-MBL controls from the same cohort and geographical area (10y OS rates of 76% vs. 86%, respectively; P=0.03) (Figure 2A,B). Moreover, MBLlo subjects also showed a significantly shortened survival vs. age-matched individuals of the general population from the same geographical region (8.0% vs. 1.8% in the period 2015-2016, respectively; P<0.001) (Online Supplementary Figure S1). Interestingly, such differences in OS were at the expense of a lower OS of CLL-like MBLlo females, who showed a significantly (P=0.01) higher risk of death (hazard ratio (HR) of 2.5; 95% confidence interval (CI) of 1.2-5.4) than non-MBL females of the same age (Figure 2C,F). Infections (21%; mostly respiratory infections and sepsis), cancer (36%; all solid tumors except for an essential thrombocythemia) and cardiovascular diseases (29%; i.e., myocardial infarction and acute ischemic stroke) were the main causes of death among MBLlo subjects. Overall, infections were overrepresented among the MBLlo cohort vs. age- and sex-matched subjects from the general population of the same geographical area (21% vs. 1.4%, respectively; P≤0.001). In contrast, the proportion of deaths caused by tumors (36% vs. 26%, respectively; P>0.05) and by cardiovascular diseases (29% vs. 33%, respectively; P>0.05) were similar in both groups. In

turn, no MBLlo subjects died as a cause of non-infectious respiratory tract diseases or genitourinary diseases, diabetes, dementia or other nervous system disorders, which accounted for ≈30% of deaths in the age- and sex-matched general population cohort living in the same geographical area. In order to identify those variables independently associated with OS, a multivariate Cox regression analysis, including laboratory, epidemiological and medical information, was carried out. Advanced age- HR of 5.1; 95% CI: 1.5-17.5; P=0.01-, co-existing cardiovascular diseases (HR: 2.7; 95%CI: 1.3-5.4; P=0.01), solid tumors (HR: 2.9; 95%CI: 1.3-6.5; P=0.007) and, to a lesser extent, the presence of MBLlo clones (HR: 2.1, 95%CI: 0.97-4.7; P=0.06), were independently associated with a shorter OS in the whole cohort (Table 5 and Online Supplementary Table S6).

Discussion Several preceding studies have shown that virtually all CLL cases are preceded by MBLhi;8,20,21 in contrast, such a relationship has not been demonstrated for MBLlo cases, its role as a preleukemic condition still remaining to be confirmed.9,21 In fact, there exist very few studies with short-term follow-up (i.e., ≤3y) which have investigated

Table 3. Frequency of cases with CLL-associated cytogenetic alterations and percentage of cells affected by each genetic abnormality.

N. of cases with cytogenetic alterations (%) Chromosomal region del(13q14)(D13S25) % altered cells del(13q14)(RB1) % altered cells Trisomy 12 % altered cells del(11q)(ATM) % altered cells del(17p)(TP53) % altered cells t(14q32)* % altered cells t(11;14)(q13-q32) % altered cells del(7q32) % altered cells 3q27 (BCL6) % altered cells 18q21 (MALT1) % altered cells

CLL-like MBLlo cases Non CLL-like MBLlo cases Baseline Follow-up # Baseline Follow-up (n=24) (n=50) (n=7) (n=6)

P

All MBLlo cases Baseline (n=31)

Follow-up (n=56)

10/31 (32%)

34/56 (61%)

7/24 (29%)

31/50 (62%)

3/7 (43%)

3/6 (50%)

0.01a,b

7/22 (32%) 49±36% 3/15 (20%) 14±3% 2/21 (10%) 34±35% 2/12 (17%) 39±44% 1/10 (10%) 13% 0/5 (0%) NA 1/2 (50%) 100% 0/1 (0%) NA 0/1 (0%) NA 0/2 (0%) NA

28/54 (52%) 31±27% 1/7 (14%) 47% 2/55 (3.6%) 45±35% 1/54 (1.9%) 50% 1/54 (1.9%) 10% 7/27 (26%) 33±30% NA

6/20 (30%) 56±34% 3/15 (20%) 14±3% 1/19 (5.3%)¥ 59% 0/10 (0%) NA 0/8 (0%) NA NA

1/2 (50%) 8% NA

1/6 (17%) 7% NA

0.06b NA NS

1/6 (17%) 20% 1/6 (17%) 50% 0/6 (0%) NA 2/4 (50%) 38±30% NA

2/5 (40%) 20±2.1% 0/5 (0%) NA 0/4 (0%) NA

NA

NA NA

NA

NA

2/5 (40%) 20±2.1% 0/5 (0%) NA 0/4 (0%) NA

NS

NA

1/2 (50%) 9% 2/2 (100%) 39±44% 1/2 (50%) 13% 0/4 (0%) NA 1/2 (50%) 100% 0/1 (0%) NA 0/1 (0%) NA 0/2 (0%) NA

NS

NA

27/48 (56%) 32±27% 1/7 (14%) 47% 1/49 (2%)¥ 70% 0/48 (0%) NA 1/48 (2.1%) 10% 5/23 (22%) 31±33% NA

NS NS NS NA

NS NA

Results expressed as number of cases (percentage of cases) and mean ± SD of percentage of cells affected by each specific genetic alteration. aBaseline vs. follow-up (year +7) for all cases. bBaseline vs. follow-up (year +7) for CLL-like MBL cases. cBaseline vs. follow-up (year +7) for non CLL-like MBL cases. # 2/50 individuals carried a clonal MBLlo CLLlike population along with at least one MBLlo non CLL-like clone. ¥The same case at baseline and follow-up. *Other than t(11;14). CLL: chronic lymphocytic leukemia; MBLlo: lowcount monoclonal B-cell lymphocytosis; N: number; NA: not applicable; NS: not statistically significantly different (P>0.05).

haematologica | 2018; 103(7)

1203


I. Criado et al.

the progression rate from MBLlo to MBLhi and CLL thus far.10,22,23 Hence, Fazi et al. showed persistent MBLlo clones over time in 90% of CLL-like MBLlo and only 67% of non CLL-like clones, after a median follow-up of ≈3y.10 Herein, we demonstrate the systematic persistence of both CLLlike and non CLL-like MBLlo B-cell clones with an identical phenotype to baseline after seven years follow-up in 65/65 MBLlo cases, confirming that MBLlo is not a transient condition. Similarly, Matos et al. also found the persistence of B-cell clones in their limited series of CLL-like MBLlo cases (n=5) after a median follow-up of ≈7y.23 Interestingly, in 3/56 CLL-like MBLlo cases, the number of clones identi-

1204

A

B

C

D

E

F

fied at seven years follow-up changed, which might suggest the emergence of MBLlo from an oligoclonal background that mirrors competition and natural selection among multiple coexisting clones.24 Changes observed in the VDJ sequences of the expanded B cells from most of these cases (data not shown), together with the progressively decreasing rate of oligoclonality from MBLlo (12-19%) to MBLhi (2.9-13%) and CLL (0.7-3.4%), would further support this hypothesis.9,12,25–27 The significance of such oligoclonal B-cell expansions in MBLlo remains unknown, but might be the consequence of the early stages of altered oligoclonal immune responses against multiple antigens,

Figure 2. Overall survival from baseline (mortality rates) of MBLlo individuals vs. age- and sex-matched non-MBL controls. Left column panels represent comparisons of overall survival curves from MBLlo subjects (black) and age- and sex-matched non-MBL controls (dotted gray). The same comparison is depicted for all individuals (Panel A) and separately for males (Panel C) and females (Panel E). In the right column, overall survival curves comparing all MBLlo subjects with a CLL-like phenotype vs. all ageand sex-matched non-MBL controls (Panel B). The same subjects distributed according to sex are shown in Panel D (males) and in Panel F (females). MBLlo: low-count monoclonal B-cell lymphocytosis.

haematologica | 2018; 103(7)


7y follow-up of monoclonal B-cell lymphocytosis

Table 4. Clinical and biological characteristics of CLL-like MBLlo subjects at baseline and at follow-up (+7 years) according to the kinetics of the MBL clone in PB (decreased/stable vs. increased size).

CLL-like MBLlo subjects (n=56) Decrease/stable B-cell clones (n=9) Increased B-cell clones (n=47)# Male/Female* Age at baseline, years Monoclonal at follow-up* Leukocytosis (>10x109/L) at follow-up * Lymphocytosis (>4x109/L) at follow-up * N. total T cells/µL Baseline follow-up P N. CD4+ T cells/mL Baseline follow-up P N. CD8+ T cells/mL Baseline follow-up P N. CD4+/CD8+ T cells/mL Baseline follow-up P N. CD4–/CD8– T cells/mL Baseline follow-up P N. total B cells/mL Baseline follow-up P N. normal B cells/mL Baseline follow-up P N. clonal B cells/mL Baseline follow-up P % clonal B cells Baseline follow-up P N. NK cells/mL Baseline follow-up P Cytogenetic alterations Baseline follow-up P del(13q14)(D13S25) Baseline follow-up

P

4/5 (44%/56%) 78 (55-84) 5/9 (56%) 0 (0%) 0 (0%)

18/29 (38%/62%) 68 (43-81) 30/47 (64%) 2 (4%) 2 (4%)

NS 0.03 NS NS NS

1471 (1105-2035) 1406 (711-2313) NS

1285 (341-2428) 1520 (460-3753) <0.01

NS NS

821 (461-1186) 792 (295-1327) NS

448 (253-1572) 908 (227-2045) <0.01

NS NS

452 (374-900) 491 (245-1469) NS

448 (72-1154) 467 (96-1742) 0.02

NS NS

4.3 (0.97-17) 4.7 (2.3-13) NS

4.6 (0.55-37) 8.6 (1.3-147) <0.001

NS 0.03

70 (8.2-214) 60 (7.2-272) NS

58 (8.0-190) 65 (1.9-338) 0.02

NS NS

110 (41-263) 80 (29-390) NS

139 (50-1066) 175 (28-1218) <0.01

NS 0.02

94 (37-256) 79 (26-389) NS

122 (50-478) 140 (27-536) 0.03

NS NS

0.80 (0.13-23) 0.60 (0.05-3.2) 0.02

0.71 (0.03-66) 2.0 (0.10-808) <0.001

NS 0.03

0.92 (0.10-20) 0.44 (0.04-10) 0.05

0.53 (0.02-21) 1.0 (0.06-66) <0.001

NS NS

304 (167-874) 492 (310-1066) NS

292 (76-1138) 361 (87-3415) <0.01

NS NS

0/4 (0%) 6/9 (67%) 0.03

6/15 (38%) 26/41 (63%) 0.14

NS NS

0/2 (0%) 6/9 (67%); 17±9%

5/13 (39%); 57±38% 21/39 (54%); 37±29%

NS NS continued in the next page

haematologica | 2018; 103(7)

1205


I. Criado et al.

del(13q14)(RB1) Baseline follow-up Trisomy 12 Baseline follow-up del(11q)(ATM) Baseline follow-up del(17p)(TP53) Baseline follow-up t(14q32) Baseline follow-up

0/2 (0%) NA

3/13 (23%); 14±3.2% 1/7 (14%); 47%

NS NA

0/3 (0%) 0/9(0%)

1/11 (9%); 59% 1/40 (3%); 70%

NS NS

0/2 (0%) 0/9 (0%)

0/6 (0%) 0/39 (0%)

NS NS

0/2 (0%) 0/9 (0%)

0/5 (0%) 1/39 (3%); 10%

NS NS

NA 0/3 (0%)

NA 5/20 (25%); 31±33%

NA NS

# 2/56 individuals carried a clonal MBLlo CLL-like population along with at least one MBLlo non CLL-like clone. Results expressed as median (range) or as *number of cases (percentage). Cytogenetic alterations are expressed as percentage of cases and mean percentage of cells affected ± SD. P-values shown in the right column refer to comparisons between MBLlo subjects who showed decreased/stable vs. increase clone sizes, while P-values shown in rows represent differences among subjects within each group at baseline and after seven years of follow-up. CLL: chronic lymphocytic leukemia; MBLlo: low-count monoclonal B-cell lymphocytosis; N.: number; NA: not applicable; NK: natural killer; NS: not statistically significantly different (P>0.05).

in which a single clone had not yet emerged as dominant vs. the others, as might occur at the latter, e.g., CLL stage. Most importantly, over two thirds of all CLL-like MBLlo clones showed a significantly increased size in PB after seven years, while for non CLL-like clones more variable kinetics were observed, depending on the specific phenotype of clonal B-cells. Interestingly, we also observed a significant increase in the frequency of cytogenetic alterations over time, evidencing that B-cell clones are not only dynamic in terms of clone size, but also regarding their capacity to acquire new cytogenetic alterations. Of note, del(13q14), which has been found to be a common mosaicism in the general population,28,29 was absent in non-clonal B cells from 5/5 cases investigated in which CLL-like clonal cells did carry this alteration, indicating that the emergence of this alteration in MBLlo is specific for the clonal population. Altogether, these findings suggest that cytogenetic alterations are a relatively early, but not primary, event in the natural history of MBL/CLL, and might have a potential role in the progression of MBLlo to MBLhi and CLL. The presence and type of cytogenetic lesions, the IGHV mutational status, or the presence of stereotyped receptors are some of the most important prognostic factors in CLL, which also define the outcome of MBLhi individuals; furthermore, it might identify a subset of cases in whom the presence of the B-cell clonal population influences OS.30–33 Unfortunately, in the present study, the mutational status and VDJ rearrangements were only assessed (both baseline and follow-up) in 8/65 MBLlo individuals (data not shown), making it impossible to validate solid conclusions regarding the potential association with the risk for progression into MBLhi and CLL. To the best of our knowledge, the frequency and impact on disease progression of recurrent mutations (i.e., NOTCH1, SF3B1, MYD88, etc.) found in CLL, and also in MBLhi, to a lesser extent, has not been elucidated for MBLlo.34–36 Therefore, analysis of these CLL-related mutations in MBLlo cases might further contribute to an improvement in better delineating intrinsic tumor cell factors associated to disease progression. In addition, the environment in which CLL-like MBLlo 1206

Table 5. Variables studied in the Cox regression multivariate analysis showing an independent impact (P<0.1) on OS for the whole MBLlo plus non-MBL cohort.

Variables Whole cohort Cardiovascular disease Age (<65y vs. ≥65y) Solid tumor MBLlo clones

HR (95%CI)

P

2.65 (1.30 - 5.41) 5.08 (1.48 - 17.49) 2.86 (1.26 - 6.46) 2.14 (0.97 - 4.72)

0.007 0.01 0.01 0.06

CI: confidence interval; HR: hazard ratio; MBLlo: low-count monoclonal B-cell lymphocytosis; N: number; OS: overall survival; PB: peripheral blood. The complete list of variables analyzed in the Cox regression model is provided in Online Supplementary Table S6.

clones develop might be influenced by chronic immune responses against e.g., host viruses, that might play a critical role in the expansion of clonal B cells, as recently suggested.37 In line with this hypothesis, herein we also show that the expansion of CLL-like MBLlo clones after seven years of follow-up (vs. baseline) is accompanied by a significant increase of all T-cell (but CD4+CD8+cytotoxic Tcells) and NK-cell populations in PB. Controversial results have been reported regarding PB Tcell numbers in MBLlo. Hence, while te Raa et al. found normal CD4+ and CD8+ T-cell counts in PB of MBLhi,38 other studies have demonstrated that around half of the MBLlo individuals show ≥1 clonal/oligoclonal CD4+CD8+ T-cell population, with an overall increased frequency of clonal T-cell populations vs. age-matched individuals from the general population.10,39 However, the presence of clonal (CD4+CD8+ and other) T-cell expansions has also been described as a common event in older individuals, and has been associated with the ageing of the immune system.39 In this respect, we demonstrate herein that changes in the number of circulating PB T-cell and NK-cell populations among our CLL-like MBLlo subjects were not age-related, via a parallel analysis of a large group of 250 age- and sexmatched non-MBL controls (Online Supplementary Table S5). From a pathophysiological point of view, the increase in most PB T- and NK-cell populations could be associated haematologica | 2018; 103(7)


7y follow-up of monoclonal B-cell lymphocytosis

with either a potentially protective or activating effect of these cellular components of the immune system (microenvironment) on the expanded clonal B-cells.40,41 Therefore, on one hand, increased numbers of (functionally impaired) T cells have been described in CLL38,42,43 while on the other hand, we have recently shown increased titers of plasma antibodies against CMV and EBV in MBLhi and CLL patients vs. MBLlo and non-MBL controls, despite their antibody (immune)deficient state.37 Taken together, these latter findings might further support the existence of additional signals coming from immune cells other than clonal B cells, that could already contribute to the expansion of (cyto)genetically altered CLL-like clones at the earliest stages of disease, by promoting activation, proliferation and/or survival of specific B-cell clones. A major goal of our study was to investigate the medium-term rate of progression of MBLlo to MBLhi and (potentially also) CLL. Overall, only one subject evolved from MBLlo to MBLhi, and none transformed to CLL, which would translate into a progression rate from MBLlo to MBLhi of 1.8% after seven years of follow-up. Despite the fact that the rate of progression of MBLlo to MBLhi and CLL appears to be extremely low, one of the most astonishing findings of our follow-up study was the significantly higher frequency of deaths among MBLlo subjects, associated with a significant adverse impact on OS vs. both non-MBL controls, particularly among females, and the general population (of similar age and sex distribution) living in the same region in Spain. However, comparisons with the general population must be considered with care, since the conditions of this population might differ from that of non-MBL individuals recruited at the Primary Health Services. Multivariate analysis showed a borderline significant association between the presence of MBLlo clones and a shorter survival. Despite this, the specific mechanisms responsible for the higher frequency of infections and deaths observed, particularly among women, are unknown, and further studies are required to validate and clarify these results. In this regard, controversial results have been reported on MBLhi subjects in the literature. Thus, while Shanafelt et al. showed no differences in OS of MBLhi vs. the general population,33 Shim et al. pointed out a higher frequency of deaths in their MBLlo cohort (4/11; 36%), albeit no statistically significant differences were found vs. non-MBL controls in the latter study, probably due to the small sample size.22 In addition, Fazi et al. also reported that 16/137 (12%) CLL-like MBLlo subjects died before re-evaluation after a median time of three years, which is a high proportion of their whole cohort.10 However, in the aforementioned report no information about the age of the deceased subjects is provided, and therefore, if it is the case they were older (than those subjects remaining alive) such high mortality rates might have been expected. Even more strikingly is the overrepresenta-

References : 1. Rai KR, Jain P. Chronic lymphocytic leukemia (CLL)-Then and now. Am J Hematol. 2016;91(3):330–340. 2. Hallek M, Cheson BD, Catovsky D, et al. Guidelines for the diagnosis and treatment of chronic lymphocytic leukemia : a report

haematologica | 2018; 103(7)

tion of infections as causes of death in MBLlo compared to that of our non-MBL cohort. Impaired immune responses and higher frequencies of infection have been recurrently reported in both MBLhi and CLL,44–47 but so far very little information exists in MBLlo, and such an association deserves further investigations. Several groups pointed out that the frequency of clonal hematopoiesis dramatically increases with age in the general population, especially among the elderly, in a similar way to the increased frequency of MBLlo, reflecting a clear relationship between clonal hematopoiesis and a higher risk of death.48,49 Whether or not both phenomena are related to MBLlo deserves future investigations. In summary, we show herein that although MBLlo is a persistent and dynamic condition with a progressive acquisition of cytogenetic alterations, usually associated with an increased clone size and higher T- and NK-cell numbers in PB over time, progression of MBLlo to MBLhi and CLL is extremely rare in the medium-term. Despite this, the MBLlo subjects analyzed herein, particularly women, showed a shortened OS associated with an increased risk of death, particularly due to infections, further supporting the notion that MBLlo could be a marker of an impaired immune system, indirectly associated with a poorer outcome. Additional studies are necessary to confirm these findings and shed light onto the specific immune defects and microenvironmental factors involved in MBLlo. Funding This work was supported by the RD06/0020/0035 and RD12/0036/0048 grants from Red Temática de Investigación Cooperativa en Cáncer (RTICC), Instituto de Salud Carlos III, Ministerio de Economía y Competitividad, (Madrid, Spain and FONDOS FEDER); CB16/12/00400 grant (CIBERONC, Instituto de Salud Carlos III, Ministerio de Economía y Competitividad, (Madrid, Spain and FONDOS FEDER); the FIS PI06/0824-FEDER, PS09/02430-FEDER, PI12/00905FEDER, DTS15/00119-FEDER, PI16/00787-FEDER and PI17/00399-FEDER grants, from the Fondo de Investigación Sanitaria of Instituto de Salud Carlos III; the GRS206/A/08 grant, (Ayuda al Grupo GR37 de Excelencia, SAN/1778/2009) from the Gerencia Regional de Salud (Consejería de Educación and Consejería de Sanidad of Castilla y León, Valladolid, Spain) and the SA079U14 grant (Consejería de Educación and Consejería de Sanidad of Castilla y León, Valladolid, Spain). ML Gutiérrez is supported by grant PTA2014-09963-I from the Instituto de Salud Carlos III. Acknowledgments The authors would like to thank “The Primary Health Care Group of Salamanca for the Study of MBL” for their contribution to the study; a complete list of members is included in the Online Supplementary Information.

from the International Workshop on Chronic Lymphocytic Leukemia updating the National Cancer Institute – Working Group 1996 guidelines. Blood. 2008;111(12):5446–5456. 3. Scarfo L, Ferreri AJM, Ghia P. Chronic lymphocytic leukaemia. Crit Rev Oncol Hematol. 2016;104:169–182. 4. Bachow SH, Lamanna N. Evolving strate-

gies for the treatment of chronic lymphocytic leukemia in the upfront setting. Curr Hematol Malig Rep. 2016;11(1):61–70. 5. Chiorazzi N. Implications of new prognostic markers in chronic lymphocytic leukemia. Hematol Am Soc Hematol Educ Progr. 2012;2012:76–87. 6. Swerdlow SH, Campo E, Harris NL, et al. WHO Classification of Tumours of

1207


I. Criado et al.

7.

8.

9.

10.

11.

12.

13. 14.

15.

16.

17.

18.

19.

20.

21.

1208

Haematopoietic and Lymphoid Tissues. 4th ed. Lyon (France): IARC; 2017. Rawstron AC, Shanafelt T, Lanasa MC, et al. Different biology and clinical outcome according to the absolute numbers of clonal B-cells in monoclonal B-cell lymphocytosis (MBL). Cytometry B Clin Cytom. 2010;78(Suppl. 1):19–23. Rawstron AC, Bennett FL, O’Connor SJM, et al. Monoclonal B-cell lymphocytosis and chronic lymphocytic leukemia. N Engl J Med. 2008;359(6):575–583. Vardi A, Dagklis A, Scarfo L, et al. Immunogenetics shows that not all MBL are equal: the larger the clone, the more similar to CLL. Blood. 2013;121(22):4521–4528. Fazi C, Scarfo L, Pecciarini L, et al. General population low-count CLL-like MBL persists over time without clinical progression, although carrying the same cytogenetic abnormalities of CLL. Blood. 2011;118(25):6618–6625. Shim YK, Middleton DC, Caporaso NE, et al. Prevalence of monoclonal B-cell lymphocytosis: a systematic review. Cytometry B Clin Cytom. 2010;78(Supp.l 1):10-18. Nieto WG, Almeida J, Romero A, et al. Increased frequency (12%) of circulating chronic lymphocytic leukemia-like B-cell clones in healthy subjects using a highly sensitive multicolor flow cytometry approach. Blood. 2009;114(1):33–37. Scarfo L, Fazi C, Ghia P. MBL versus CLL: how important is the distinction? Hematol Oncol Clin North Am. 2013;27(2):251–265. Nieto WG, Teodosio C, López A, et al. Non-CLL-like monoclonal B-Cell lymphocytosis in the general population: Prevalence and phenotypic/genetic characteristics. Cytometry B Clin Cytom. 2010;78(Suppl. 1):24–34. Casabonne D, Almeida J, Nieto WG, et al. Common infectious agents and monoclonal B-cell lymphocytosis: a cross-sectional epidemiological study among healthy adults. PLoS One. 2012;7(12):e52808. Kalina T, Flores-Montero J, van der Velden VHJ, et al. EuroFlow standardization of flow cytometer instrument settings and immunophenotyping protocols. Leukemia. 2012;26(9):1986–2010. Flores-Montero J, Sanoja-Flores L, Paiva B, et al. Next Generation Flow for highly sensitive and standardized detection of minimal residual disease in multiple myeloma. Leukemia. 2017;31(10):2094–2103. Quijano S, López A, Rasillo A, et al. Impact of trisomy 12, del(13q), del(17p), and del(11q) on the immunophenotype, DNA ploidy status, and proliferative rate of leukemic B-cells in chronic lymphocytic leukemia. Cytometry B Clin Cytom. 2008;74(3):139–149. Efron B, Tibshirani R. Bootstrap methods for standard errors, confidence intervals, and other measures of statistial accuracy. Stat Sci. 1986;1(1):54-77. Landgren O, Albitar M, Ma W, et al. B-cell clones as early markers for chronic lymphocytic leukemia. N Engl J Med. 2009; 360(7):659-667. Shanafelt TD, Ghia P, Lanasa MC, Landgren O, Rawstron AC. Monoclonal Bcell lymphocytosis (MBL): biology, natural history and clinical management.

Leukemia. 2010;24(3):512-520. 22. Shim YK, Vogt RF, Middleton D, et al. Prevalence and natural history of monoclonal and polyclonal B-cell lymphocytosis in a residential adult population. Cytometry B Clin Cytom. 2007;72(5):344-353. 23. Matos DM, Furtado FM, Falcao RP. Monoclonal B-cell lymphocytosis in individuals from sporadic (non-familial) chronic lymphocytic leukemia families persists over time, but does not progress to chronic B-cell lymphoproliferative diseases. Rev Bras Hematol Hemoter. 2015;37(5):292295. 24. Greaves M, Maley CC. Clonal evolution in cancer. Nature. 2012;481(7381):306-313. 25. Henriques A, Rodriguez-Caballero A, Nieto WG, et al. Combined patterns of IGHV repertoire and cytogenetic/molecular alterations in monoclonal B lymphocytosis versus chronic lymphocytic leukemia. PLoS One. 2013;8(7):e67751. 26. Sanchez M-L, Almeida J, Gonzalez D, et al. Incidence and clinicobiologic characteristics of leukemic B-cell chronic lymphoproliferative disorders with more than one Bcell clone. Blood. 2003;102(8):2994-3002. 27. Kern W, Bacher U, Schnittger S, et al. Flow cytometric identification of 76 patients with biclonal disease among 5523 patients with chronic lymphocytic leukaemia (BCLL) and its genetic characterization. Br J Haematol. 2014;164(4):565-569. 28. Jacobs KB, Yeager M, Zhou W, et al. Detectable clonal mosaicism and its relationship to aging and cancer. Nat Genet. 2012;44(6):651-658. 29. Machiela MJ, Zhou W, Caporaso N, et al. Mosaic 13q14 deletions in peripheral leukocytes of non-hematologic cancer cases and healthy controls. J Hum Genet. 2016;61(5):411-418. 30. Shanafelt TD, Kay NE, Rabe KG, et al. Brief report: natural history of individuals with clinically recognized monoclonal B-cell lymphocytosis compared with patients with Rai 0 chronic lymphocytic leukemia. J Clin Oncol. 2009;27(24):3959-3963. 31. Kern W, Bacher U, Haferlach C, et al. Monoclonal B-cell lymphocytosis is closely related to chronic lymphocytic leukaemia and may be better classified as early-stage CLL. Br J Haematol. 2012;157(1):86-96. 32. Strati P, Shanafelt TD. Monoclonal B-cell lymphocytosis and early-stage chronic lymphocytic leukemia: diagnosis, natural history, and risk stratification. Blood. 2015; 126(4):454-462. 33. Shanafelt TD, Kay NE, Rabe KG, et al. Survival of patients with clinically identified monoclonal B-cell lymphocytosis (MBL) relative to the age- and sex-matched general population. Leukemia. 2012; 26(2):373-376. 34. Landau DA, Tausch E, Taylor-Weiner AN, et al. Mutations driving CLL and their evolution in progression and relapse. Nature. 2015;526(7574):525-530. 35. Barrio S, Shanafelt TD, Ojha J, et al. Genomic characterization of high-count MBL cases indicates that early detection of driver mutations and subclonal expansion are predictors of adverse clinical outcome. Leukemia. 2017;31(1):170-176. 36. Agathangelidis A, Ljungström V, Scarfò L, et al. Highly similar genomic landscapes in

37.

38.

39.

40.

41.

42.

43.

44.

45.

46. 47.

48.

49.

monoclonal B-cell lymphocytosis and ultrastable chronic lymphocytic leukemia with low frequency of driver mutations. Haematologica. 2018 Feb 15. [Epub ahead of print]. Criado I, Muñoz-Criado S, RodríguezCaballero A, et al. Host virus and pneumococcus-specific immune responses in highcount monoclonal B-cell lymphocytosis and chronic lymphocytic leukemia: implications for disease progression. Haematologica. 2017;102(7):1238-1246. te Raa GD, Tonino SH, Remmerswaal EBM, et al. Chronic lymphocytic leukemia specific T-cell subset alterations are clonesize dependent and not present in monoclonal B lymphocytosis. Leuk Lymphoma. 2012;53(11):2321-2325. Ghia P, Prato G, Stella S, Scielzo C, Geuna M, Caligaris-Cappio F. Age-dependent accumulation of monoclonal CD4+CD8+ double positive T lymphocytes in the peripheral blood of the elderly. Br J Haematol. 2007;139(5):780-790. Purroy N, Wu CJ. Coevolution of leukemia and host immune cells in chronic lymphocytic leukemia. Cold Spring Harb Perspect Med. 2017;7(4):a026740. Os A, Bürgler S, Ribes AP, et al. Chronic lymphocytic leukemia cells are activated and proliferate in response to specific T helper cells. Cell Rep. 2013;4(3):566-77. Ramsay AG, Clear AJ, Fatah R, Gribben JG. Multiple inhibitory ligands induce impaired T-cell immunologic synapse function in chronic lymphocytic leukemia that can be blocked with lenalidomide: establishing a reversible immune evasion mechanism in human cancer. Blood. 2012; 120(7):14121421. Riches JC, Davies JK, McClanahan F, et al. T cells from CLL patients exhibit features of T-cell exhaustion but retain capacity for cytokine production. Blood. 2013; 121(9):1612-1621. Morrison VA. Infectious complications of chronic lymphocytic leukaemia: pathogenesis, spectrum of infection, preventive approaches. Best Pract Res Clin Haematol. 2010;23(1):145-153. Moreira J, Rabe KG, Cerhan JR, et al. Infectious complications among individuals with clinical monoclonal B-cell lymphocytosis (MBL): a cohort study of newly diagnosed cases compared to controls. Leukemia. 2013;27(1):136-141. Forconi F, Moss P. Perturbation of the normal immune system in patients with CLL. Blood. 2015;126(5):573-81. Solomon BM, Chaffee KG, Moreira J, et al. Risk of non-hematologic cancer in individuals with high-count monoclonal B-cell lymphocytosis. Leukemia. 2016; 30(2):331336. Jaiswal S, Fontanillas P, Flannick J, et al. Age-related clonal hematopoiesis associated with adverse outcomes. N Engl J Med. 2014;371(26):2488-2498. Genovese G, Kahler AK, Handsaker RE, et al. Clonal hematopoiesis and blood-cancer risk inferred from blood DNA sequence. N Engl J Med. 2014;371(26):2477-2487.

haematologica | 2018; 103(7)


ARTICLE

Chronic Lymphocytic Leukemia

Efficacy of bendamustine and rituximab as first salvage treatment in chronic lymphocytic leukemia and indirect comparison with ibrutinib: a GIMEMA, ERIC and UK CLL FORUM study

Antonio Cuneo,1 George Follows,2 Gian Matteo Rigolin,1 Alfonso Piciocchi,3 Alessandra Tedeschi,4 Livio Trentin,5 Angeles Medina Perez,6 Marta Coscia,7 Luca Laurenti,8 Gerardo Musuraca,9 Lucia Farina,10 Alfredo Rivas Delgado,11 Ester Maria Orlandi,12 Piero Galieni,13 Francesca Romana Mauro,14 Carlo Visco,15 Angela Amendola,16 Atto Billio,17 Roberto Marasca,18 Annalisa Chiarenza,19 Vittorio Meneghini,20 Fiorella Ilariucci,21 Monia Marchetti,22 Stefano Molica,23 Francesca Re,24 Gianluca Gaidano,25 Marcos Gonzalez,26 Francesco Forconi,27 Stefania Ciolli,28 Agostino Cortelezzi,29 Marco Montillo,4 Lukas Smolej,30 Anna Schuh,31 Toby A. Eyre,32 Ben Kennedy,33 Kris M. Bowles,34 Marco Vignetti,3 Javier de la Serna,35 Carol Moreno,36 Robin Foà14 and Paolo Ghia37 on behalf of the GIMEMA, European Research Initiative on CLL (ERIC) and UK CLL forum

Hematology, Department of Medical Sciences, St. Anna University Hospital, Ferrara, Italy; 2UK CLL Forum, Cambridge University Hospitals NHS Foundation Trust, UK; 3Italian Group for Adult Hematologic Diseases (GIMEMA), Data Center and Health Outcomes Research Unit, Rome, Italy; 4Hematology, Niguarda Cancer Center, ASST Grande Ospedale Metropolitano Niguarda, Milan, Italy; 5Hematology and Clinical Immunology, Department of Medicine, University of Padua, Italy; 6Hospital Costa del Sol, Marbella, Málaga, Spain; 7Hematology Unit, Città della Salute e della Scienza, University of Turin, Italy; 8Hematology, Università Cattolica del Sacro Cuore, Policlinico A. Gemelli, Rome, Italy; 9Hematology Unit, Istituto Scientifico Romagnolo per lo Studio e la Cura dei Tumori (IRST) IRCCS, Meldola, Italy; 10Hematology Department, Fondazione IRCCS Istituto Nazionale Tumori, Milano, Italy; 11Department of Hematology, Hospital Clinic, Institut d'Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS), Barcelona, Spain; 12 Hematology Unit, Fondazione IRCCS Policlinico San Matteo, Pavia, Italy; 13Hematology and Cellular Therapy, “Ospedale C. e G. Mazzoni”, Ascoli Piceno, Italy; 14Hematology, Department of Biomedical Sciences and Hematology, "Sapienza" University, Rome, Italy; 15 Hematology, San Bortolo Hospital, Vicenza, Italy; 16Hematology, San Carlo Hospital, Potenza, Italy; 17Hematology and Transplant Unit, San Maurizio Hospital, Azienda Sanitaria dell'Alto Adige, Bolzano, Italy; 18Hematology Unit, University Hospital, Modena, Italy; 19Hematology Unit, Azienda Universitaria Ospedaliera Policlinico Vittorio Emanuele, Catania, Italy; 20Hematology, Department of Cell Therapy and Hematology, University Hospital, Verona, Italy; 21Hematology Unit, Arcispedale S. Maria Nuova, Reggio Emilia, Italy; 22Oncology Unit, Cardinal Massaia Hospital, Asti, Italy; 23Hematology Unit, A. Pugliese Hospital, Azienda Ospedaliera Pugliese Ciaccio, Catanzaro, Italy; 24Hematology, University Hospital, Parma, Italy; 25Hematology, DIMECS e Dipartimento Oncologico, Università del Piemonte Orientale Amedeo Avogadro, Novara, Italy; 26Hematology, University Hospital-IBSAL and CIBERONC, Salamanca, Spain; 27Haematology Department, University Hospital National Health Service Trust, Southampton, UK; 28 Hematology Unit, Careggi Hospital, Florence, Italy; 29Hematology Unit, Fondazione IRCCS Ca' Granda Ospedale Maggiore Policlinico, University of Milan, Italy; 30Department of Hematology, University Hospital, Hradec Kralove, Czech Republic; 31UK CLL Forum, Oxford University Hospitals NHS Foundation Trust, UK; 32Oxford University Hospitals NHS Foundation Trust, UK; 33University Hospitals of Leicester NHS Trust, UK; 34Norwich Medical School, UK; 35Hematology Unit, Hospital Universitario 12 de Octubre, Madrid, Spain; 36Hospital de la Santa Creu i Sant Pau, Barcellona, Spain and 37Division of Experimental Oncology, IRCCS San Raffaele Scientific Institute and Vita-Salute San Raffaele University, Milan, Italy

Ferrata Storti Foundation

Haematologica 2018 Volume 103(7):1209-1217

1

AC and GF contributed equally to this work as first authors. RF and PG contributed equally to this work as last authors

ABSTRACT

W

e performed an observational study on the efficacy of bendamustine and rituximab (BR) as first salvage regimen in chronic lymphocytic leukemia (CLL). In an intention-to-treat analysis including 237 patients, the median progression-free survival (PFS) was 25 months. The presence of del(17p), unmutated IGHV and advanced stage were associated with a shorter PFS at multivariate analysis. The haematologica | 2018; 103(7)

Correspondence: cut@unife.it

Received: February 6, 2018. Accepted: April 18, 2018. Pre-published: April 19, 2018. doi:10.3324/haematol.2018.189837 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/103/7/1209 ©2018 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


A. Cuneo et al.

median time-to-next treatment was 31.3 months. Front-line treatment with a chemoimmunotherapy regimen was the only predictive factor for a shorter time to next treatment at multivariate analysis. The median overall survival (OS) was 74.5 months. Advanced disease stage (i.e. Rai stage III-IV or Binet stage C) and resistant disease were the only parameters significantly associated with a shorter OS. Grade 3-5 infections were recorded in 6.3% of patients. A matched-adjusted indirect comparison with ibrutinib given second-line within Named Patient Programs in the United Kingdom and in Italy was carried out with OS as objective end point. When restricting the analysis to patients with intact 17p who had received chemoimmunotherapy in first line, there was no difference in OS between patients treated with ibrutinib (63% alive at 36 months) and patients treated with BR (74.4% alive at 36 months). BR is an efficacious first salvage regimen in CLL in a real-life population, including the elderly and unfit patients. BR and ibrutinib may be equally effective in terms of OS when used as first salvage treatment in patients without 17p deletion. (Registered at clinicaltrials.gov identifier: 02491398)

Introduction Treatment of chronic lymphocytic leukemia (CLL) has dramatically changed over the last years. Chemotherapy and anti-CD20 monoclonal antibodies produce high overall response rates (ORR), including complete remissions (CR) with negative minimal residual disease, and prolonged progression-free-survival (PFS) and overall survival (OS), both in fit1,2 and unfit patients.3 In patients with TP53 disruption and/or with relapsed/refractory (R/R) disease, who represent a difficult-to-treat patient population, mechanism-driven drugs targeting the Bruton tyrosine kinase (BTK), the phosphoinositide 3-kinase delta (PI3K d) or the BCL2 protein can induce durable responses.4-7 In the absence of TP53 disruption, a chemoimmunotherapy (CIT) regimen is recommended as front-line and second-line treatment in those patients who attained a long progression-free survival (PFS) with the previous regimen.8,9 On the other hand, the National Comprehensive Cancer Network recommends one of the new agents, ibrutinib, idelalisib with rituximab or venetoclax, as alternatives to CIT for patients with relapsed or refractory disease.10 Uncertainty in the recommendations on first salvage treatment may partly derive from the consideration that the majority of studies on R/R CLL report efficacy data in an aggregate fashion, analyzing patients who had previously received one or more lines of treatment all together. Consequently, little information is currently available on the outcome of second-line treatment. Bendamustine and rituximab (BR) is one of the most widely adopted CIT regimens, both as front-line11 and second-line treatment.12-14 The BR regimen was followed by a median PFS of 18 months when used as first salvage treatment after fludarabine, cyclophosphamide and rituximab (FCR) in 62 patients regardless of TP53 aberrations and/or refractoriness to prior therapy.13 In 78 CLL patients who had received 1-3 previous lines of treatment, the BR combination was associated with a 59% ORR with a median PFS of 15.2-months.15 Bendamustine and ofatumumab produced a 23.6-month median PFS in 47 patients, 61% and 29% of whom had received 1 or 2 prior lines of treatment, respectively.16 The oral agent ibrutinib represents an effective therapy in the R/R setting.17 In a recent analysis describing a 5-year experience, a median PFS of 52 months was reported in R/R CLL treated with ibrutinib after 4 or more previous lines of treatment in more than 50% of patients.18 In a recent update of the phase III Resonate study comparing 1210

ibrutinib and ofatumumab, the PFS rate appeared to be better in patients treated with ibrutinib in second line compared to patients who had received 2 or more previous lines of treatment.19 Recent experiences with ibrutinib in a real-world setting have reported a higher rate of discontinuation compared to clinical trials,20,21 possibly due to older age and worse Performance Status (PS) of the patient population treated in the day-to-day clinical practice.22 On these grounds, we performed a retrospective observational study within the Gruppo Italiano Malattie Ematologiche dell'Adulto (GIMEMA) and European Research Initiative on CLL (ERIC) networks to collect data on the efficacy and safety profile of the BR regimen used as second-line treatment in a real-world setting. We then set out to perform an indirect comparison with ibrutinib given as first salvage treatment in the UK and the Italian Named Patient Programs (NPP).

Methods Patients Patients treated between 2008 and 2014 at GIMEMA and ERIC centers were eligible. The inclusion criteria were: i) diagnosis of CLL according to the National Cancer Institute (NCI);23 ii) age â&#x2030;Ľ18 years; iii) one previous treatment using alkylating agents and/or purine analogs with or without monoclonal antibodies; iv) progression requiring therapy (NCI criteria);23 v) second-line treatment with BR at the conventional dose of 70 mg/m2, as described.15 Patients were excluded if they had Richterâ&#x20AC;&#x2122;s syndrome transformation, HIV infection, active HCV or HBV infection. The study was registered at clinicaltrials.gov identifier: 02491398. The study was approved by the local ethics committees.

Study design and end points Data were obtained from the medical files and entered into case record forms (CRF) by treating physicians. Computerized and manual consistency checks were performed by the data manager of the GIMEMA data center. Evaluation of bone marrow response and radiographic imaging at baseline and at response were performed according to local guidelines. Treatment response and disease progression were assessed according to the NCI criteria.23

Primary end point The primary end point was PFS at 12 months from treatment start. Subjects who were withdrawn from the study without progression were censored at the date of the last assessment. Subjects haematologica | 2018; 103(7)


Efficacy of BR as first salvage treatment in CLL

without post-baseline assessments but known to be alive were censored at the time of first dose of study drug.

Secondary end points

Differences in terms of PFS, TTNT and OS were evaluated by Log-Rank test in univariate analysis and Cox regression model in multivariate analysis. Cumulative Incidence curves were estimated using the proper non-parametric method. The Gray test was applied for significance tests on cumulative incidence curves. All the analyses were performed using the SAS software (v.9.4 or later); all tests were two-sided. P=0.05 was considered statistically significant. Confidence intervals were calculated at 95% (95%CI).

The ORR was assessed in all the patients who started treatment (intention to treat). Time to next anti-leukemic treatment (TTNT) was calculated using the cumulative incidence method, from the date of the first dose of the study drugs until the date of retreatment. OS was calculated from the date of the first dose of the study drug until the date of death. Patients without follow-up assessment were censored at the day of the last treatment administration. Evaluation of safety was reported according to NCI Common Terminology Criteria for Adverse Events version 4.0.

Results

Indirect comparison with ibrutinib

Patients’ characteristics

Data from patients treated in second line with single agent ibrutinib in the UK and Italy within the NPP were retrospectively retrieved. Patients with R/R CLL treated in the UK have been reported previously.20 Patients treated in Italy were extracted from the GIMEMA LLC1415 trial (clinicaltrials.gov identifier: 02582320). The end point for this analysis was OS.

A total of 237 patients treated at 37 centers (28 centers belonging to the GIMEMA group and 9 centers affiliated with the ERIC group) were enrolled (Online Supplementary Table S1). Baseline patients’ characteristics are outlined in Table 1: median age was 70.4 years, range 39.4-87.8; 70.9% of patients were over 65 years old; 58.3% had 2 or more comorbidities; 46.9% had a creatinine clearance ≤70 ml/min; and 21.4% had an advanced disease stage (i.e. Rai III-IV or Binet C). Seventy-three percent (data available in 61.6% of the patients) had an unmutated tumor immunoglobulin gene heavy chain variable region configuration (U-IGHV) and 33.4% had 11q- and/or 17p13 deletion (data available in 79.3% of the patients). These patients were representative of the entire study population in terms of baseline characteristics and outcome (data not shown).

Statistical analysis Statistical analysis was performed following the intention-totreat principle. Non-parametric tests were applied for comparisons between groups (χ2 and Fisher Exact test for categorical variables or response rate, Mann-Whitney and Kruskal-Wallis test for continuous variables) and logistic regression were applied in multivariate analysis. Survival distributions were estimated using the Kaplan-Meier Product Limit estimator.

Table 1. Patients' characteristics.

Frequency (%) Variable Age, years [median, range] Age, years ≤65/>65 Sex, M/F ECOG PS (%) 0-1/≥2 Stage, Rai III/IV or Binet C no/yes Bulky lymph nodes (> 5cm) no/yes Comorbidities 0-1/≥2 Creatinine clearance (mL/min) ≤ 70/>70 CD38 (>20%) neg/pos 17p- yes/no FISH 13q-/+12/11q-/17p-/no aberrations IGHV Mutated/unmutated Months between 1st and 2nd treatment <36/≥ 36 Previous treatment ORR rate to 1st line treatment (%) yes/no Refractory no/yes CIT no/yes Chemo Chl/FL-based/bendamustine CIT Chl/F-based/bendamustine <6 cycles and/or dose reductions yes/no

Benda + R n=237

Ibrutinib n=95

P

70.4 [39.4-87.8] 69 (29.1)/168(70.9) 168 (70.9)/69 (29.1) 198 (90.0)/22 (10.0) 165 (78.6)/45 (21.4) 20 (8.9)/204 (91.1) 98 (41.7)/137 (58.3) 100 (46.9)/113 (53.1) 52 (47.3)/58 (52.7) 23 (12.6)/ 160 (87.4) 45 (24.6)/32 (17.5)/38 (20.8)/23 (12.6)/45 (24.6) 40 (27.4)/106 (72.6) 124 (52.3)/113 (47.7)

69.3 [27.5-85.3] 32 (34.0)/62 (66.0) 60 (63.2)/35 (36.8) 75 (83.3)/15 (16.7) 33 (39.8)/ 50 (60.2) 14 (38.9)/22 (61.1) 54 (75.0)/18 (25.0)

195 (82.3)/42 (17.7) 174 (90.6)/18 (9.4) 95 (41.0)/137 (59.0) 39 (41.1)/42 (44.2)/14 (14.7) 19 (13.9)/77 (56.2)/41 (29.9) 140 (59.6)/95 (40.4)

56 (78.9)/15 (21.1) 22 (23.7)/71 (76.3) -

0.344 0.427 0.215 0.147 <0.001 0.251 0.001 0.636 0.005 -

n: number; ECOG: Eastern Cooperative Oncology Group; neg: negative; pos: positive; ORR: overall response rates; chemo: chemotherapy; Chl: chlorambucil; CIT: chemoimmunotherapy; M: male; F: female; FL: fludarabine.

haematologica | 2018; 103(7)

1211


A. Cuneo et al.

First-line treatment included CIT regimens combining rituximab with fludarabine (with or without cyclophosphamide), bendamustine or chlorambucil in 59% of patients; 41% of patients received chemotherapy or, in 2 cases, single agent treatment with rituximab or alemtuzumab. No patient received ibrutinib or other novel oral agents front line. The use of chemotherapy alone front line was more frequent before 2010 (52.8% of patients) than from 2011 onwards (27.9% of patients). Eighteen patients (9.4%) were refractory to first-line treatment.

Treatment with BR One hundred and sixty-five of the 237 patients (69.6%) received the planned number of cycles; treatment was discontinued early in 72 patients as a result of toxicity (n=39), withdrawal of consent (n=7), progressive disease (n=6), or for other reasons (n=20). The number of cycles actually administered to patients who discontinued treatment was â&#x2030;Ľ4 in 52.8% (n=38) of cases. Dose reduction of over 10% of the planned dose of bendamustine (i.e. <70 mg/m2) was recorded in 28.9% of cases; a treatment delay occurred in 22.5% of patients. Overall, 95 patients (40.1%) received 6 cycles without dose reduction. The median dose administered to the patients who discontinued treatment or received a reduced dose was 350 mg/m2.

Efficacy The 12-month PFS rate was 78.6% (95%CI: 73.584.1%). The estimated PFS at 30 and 60 months was 30.9% (95%CI: 24.8-38.5%) and 16.2% (95%CI: 10.624.6%), respectively, with a median overall PFS of 25

months (Figure 1) (median follow up 37.1 months, range 0.4-98.5). Factors predicting for a shorter PFS at univariate analysis (Table 2) were 17p deletion (median 14.5 months vs. 25.5 months), U-IGHV (median 20.7 months vs. 32.1) and a less than 36-month interval between first- and second-line treatment (median 21.1 months vs. 26.8), whereas an advanced stage was of borderline significance (median 20.6 vs. 25.8 months). Age (cutoff 65 years), creatinine clearance [cutoff 70 mL/minute (min)] and the presence or absence of 2 or more comorbidities had no impact on PFS. The presence of 17p-, U-IGHV and Binet/Rai stage C/IIIIV were associated with a shorter PFS at multivariate analysis (Table 2). Patients with a low-risk profile, i.e. without del(17p), with M-IGHV and Rai stage 0-2 (12.2% of the total patient population), had a median PFS of 40.4 months compared to 20.7 months in the remaining patients (P=0.003) (Online Supplementary Figure S1). The ORR was 82.3% and the probability of attaining a response was significantly lower in patients with del(17p) (69.6%) compared to patients with del(11q) (73.7%), del(13q) (82.2%), no aberrations (86.7%) or +12 (96.9%) (P=0.04). The other clinico-biological variables had no significant impact on the ORR (Online Supplementary Table S2). The TTNT at 12 months was 18.1% (95%CI: 12.6-22.2) (median 31.3 months) (Online Supplementary Figure S2). A shorter TTNT was associated with del(17p) (median 20.2 months vs. 34.6) and with the group of patients who received previous CIT vs. chemotherapy as front-line regimen (27.2 months vs. 40.4) (Table 3). An U-IGHV status (29.3 months vs. 45.7) and the presence of 2 or more

A

B

C

D

Figure 1. Progression-free survival (PFS) of patients treated with bendamustine and rituximab (BR) second-line. PFS of all 237 patients (A), by fluorescence in situ hybridization (B), IGHV status (C), and interval between first-line and second-line treatments (D).

1212

haematologica | 2018; 103(7)


Efficacy of BR as first salvage treatment in CLL

comorbidities (27.2 months vs. 39.1) were of borderline significance, whereas age and the creatinine clearance had no impact on TTNT. First-line treatment with a CIT regimen was the only predictive factor for a shorter TTNT at multivariate analysis. Seventy-three patients died due to CLL (n=14), infection with or without active CLL (n=27), second primary tumors (n=7), Richter’s syndrome (n=4). In 12 patients, the cause of death was not reported. Other causes of death in single patients (n=9) are listed in Online Supplementary Table S3. Overall survival at 12, 36 and 60 months was 92.7%, 72.2% and 54%, respectively, with a median OS of 74.5 months (Figure 2). Fifty-eight percent of patients in advanced stage were alive at 36 months compared to 75.8% in stage 0-II; 42.4% of patients who did not respond to BR were alive at 36 months compared to 78.2% of those who responded. An advanced stage (i.e. Rai stage III-IV or Binet stage C) and resistant disease were the only parameters significantly associated with a shorter OS at univariate and multivariate analysis (Table 4 and Figure 2).

Safety A detailed report of grade 3-5 adverse events (AE) is shown in Online Supplementary Table S4. Thirty-three percent of patients (n=79) reported at least one grade 3-4 AE. Overall, cytopenia was recorded in 24.9% of patients. Grade 3-4 neutropenia (including febrile neutropenia) occurred in 20.7% of cases, thrombocytopenia in 6 patients (2.5%), anemia in 3 patients (1.2%), 2 of whom had autoimmune hemolytic anemia. Grade 3-5 infections were recorded in 16 patients including 4 with febrile neutropenia (6.7%), 8 of whom (3.4%) had a lung infection. One case of fatal infection was reported (encephalitis). Rash and/or dermatitis were reported in 2 patients (0.8%).

Efficacy of ibrutinib in the UK CLL forum and in the Italian Named Patient Program Ninety-five patients were treated in 2014-2015 with single agent ibrutinib in second-line within the NPP (73 in the UK and 22 in Italy). Median follow up in the UK cohort was 3.1 years. These 95 patients were heterogeneous in baseline risk factors (Table 1), with an Eastern Cooperative Oncology Group (ECOG) PS≥2 being the

Table 2. Progression-free survival (PFS) with bendamustine and rituximab (BR) in second-line: univariate and multivariate analysis.

Variable

Age, years ≤65 vs. >65 Sex, F vs. M Stage others vs. Rai III/IV or Binet C Bulky lymph nodes (>5cm) yes vs. no Comorbidities 0-1 vs. ≥2 Creatinine clearance (mL/min) ≤70 vs. > 70 CD38 (>20%) neg vs. pos FISH 17p- vs. others IGHV mutated vs. unmutated Months between 1st and 2nd treatment < 36 vs. ≥36 First-line chemo vs. CIT <6 cycles and/or dose reductions no vs. yes

Univariate HR (95% CI)

P

0.899 (0.636-1.271) 1.110 (0.787-1.566) 0.676 (0.454-1.005) 1.643 (0.959-2.815) 1.159 (0.844-1.592) 1.179 (0.846-1.643) 0.841 (0.531-1.330) 1.965 (1.214-3.180) 0.484 (0.297-0.787) 1.398 (1.018-1.921) 0.846 (0.612-1.168) 0.752 (0.547-1.034)

0.5467 0.5519 0.0529 0.0705 0.3625 0.3312 0.4587 0.0060 0.0035 0.0387 0.3088 0.0794

Multivariate HR (95% CI)

0.536 (0.319-0.903) 2.92 (1.61-5.296) 0.53 (0.299-)0.94 -

P

0.0192 0.0004 0.0299 -

HR: Hazard Ratio; CI: Confidence Interval; Chemo: chemotherapy; CIT: chemoimmunotherapy; F: female; M: male.

Table 3. Time to next anti-leukemic treatment with bendamustine and rituximab second-line: univariate and multivariate analysis.

Variable

Age (years) ≤65 vs. >65 Sex, F vs. M Stage others vs. Rai III/IV or Binet C Bulky lymph nodes (>5 cm) yes vs. no Comorbidities 0-1 vs. ≥2 Creatinine clearance (mL/min) ≤ 70 vs. > 70 CD38 (>20%) neg vs. pos FISH 17p- vs. others IGHV mutated vs. unmutated Months between 1st and 2nd treatment < 36 vs. ≥36 First-line chemo vs. CIT <6 cycles and/or dose reductions no vs. yes

Univariate HR (95% CI)

P

1.299 (0.918-1.838) 1.040 (0.716-1.511) 0.881 (0.557-1.393) 1.598 (0.877-2.912) 1.372 (0.978-1.923) 0.833 (0.579-1.199) 1.079 (0.650-1.793) 1.863 (1.096-3.166) 0.597 (0.345-1.033) 1.044 (0.741-1.469) 0.586 (0.407-0.843) 0.776 (0.546-1.104)

0.1400 0.8349 0.5873 0.1256 0.0671 0.3261 0.7678 0.0215 0.0653 0.8062 0.0040 0.1593

Multivariate HR (95% CI)

0.59 (0.41-0.84) -

P

0.0040 -

HR: Hazard Ratio; CI: Confidence Interval; Chemo: chemotherapy; CIT: chemoimmunotherapy; F: female; M: male.

haematologica | 2018; 103(7)

1213


A. Cuneo et al. A

B

C

Figure 2. Overall survival (OS) of patients treated with bendamustine and rituximab (BR) second-line. OS of all 237 patients (A), by stage (B) and by response to BR (C).

only predictive factor with borderline statistical significance of shorter survival (Online Supplementary Table S5). When restricting the analysis to patients who had received CIT front-line (Table 5), the ibrutinib cohort and the BR cohort were comparable in terms of median age, ECOG PS, ORR rate to first-line treatment, and frequency of UIGHV (available in a proportion of cases), although with a slightly shorter interval between first- and second-line treatment in the ibrutinib cohort (interval <36 months in 76.1% vs. 59.1% of patients) and a higher number of patients with 17p deletion in the ibrutinib cohort (36.1% vs. 14.8%). When excluding patients with del(17p) from the analysis, there was no significant difference in OS between the 39 patients treated with ibrutinib (63% alive at 36 months, 95%CI: 48.8-81.6) and the 92 patients treated with BR (74.4% alive at 36 months, 95%CI: 64.7-85.5 (Figure 3). A subanalysis of the OS in patients with intact 17p and with a less than 36-month interval between firstline and first salvage treatment in the BR cohort (n=55) and in the ibrutinib cohort (n=33) showed no significant difference, with 72.6% of patients alive at three years with BR (95%CI: 60.1-87.7) and 59.8% alive at three years with ibrutinib (95%CI: 44.2-80.7) (P=0.19).

Discussion Accepting the limitations of retrospective analyses, we set out to collect data on the efficacy of BR, one of the most widely utilized CIT regimens in CLL. We elected to include in this study only patients who received secondline treatment with BR given the limited availability of published data in this setting in order to contribute new information that may assist clinicians in the selection of the most appropriate first salvage treatment in CLL. To minimize possible selection biases and imprecise reporting of data: i) we encouraged clinicians to report all 1214

P=0.146

Months from treatment start

Figure 3. Indirect comparison of overall survival in 39 patients treated secondline with ibrutinib and in 92 patients treated with bendamustine and rituximab (Benda+RTX). All patients had intact 17p and received chemoimmunotherapy as front-line therapy.

patients who initiated BR treatment; ii) we analyzed the reported data according to the intention-to-treat principle; and iii) we performed computerized and manual consistency checks on each case report form. Besides PFS, we included objective efficacy measures of the BR regimen, such as OS and the TTNT. Keeping in mind that response assessment may vary among centers and that bone biopsy was not routinely performed, we agreed to record as “response” what each treating clinician graded as “partial” or “complete” remission. The patient population who received BR included in this study closely resembled the typical CLL patient seen in daily clinical practice in terms of age, PS and comorbidities.14 The number of patients who completed the planned treatment (69.6%) was in line with a previous prospective phase-II GIMEMA study, where 76% of R/R CLL patients completed treatment.16 This finding suggests that there haematologica | 2018; 103(7)


Efficacy of BR as first salvage treatment in CLL

Table 4. Overall survival after univariate and multivariate analysis. Age, years ≤65 vs. >65 Sex, M vs. F Stage, others vs. Rai III/IV or Binet C Bulky lymph nodes (>5 cm) yes vs. no Comorbidities 0-1 vs. ≥2 Creatinine clearance (mL/min) ≤70 vs. >70 CD38 (>20%) neg vs. pos FISH 17p- vs. others IGHV mutated vs. unmutated Months between 1st and 2nd treatment <36 vs. ≥36 First-line chemo vs. CIT <6 cycles and/or dose reductions no vs. yes ORR CR; Cri; PR; nPR/vs. PD; SD; NR

Univariate HR (95% CI)

P

Multivariate HR (95% CI)

0.741 (0.439-1.250) 0.836 (0.491-1.425) 0.501 (0.296-0.846) 1.161 (0.464-2.905) 1.069 (0.671-1.702) 1.401 (0.850-2.308) 0.722 (0.356-1.465) 1.500 (0.734-3.064) 0.604 (0.290-1.254) 1.496 (0.934-2.398) 0.977 (0.609-1.565) 0.706 (0.444-1.123) 0.330 8 (0.197-0.552)

0.2612 0.5107 0.0098 0.7492 0.7797 0.1855 0.3665 0.2663 0.1761 0.0941 0.9216 0.1419 <.0001

0.547 (0.320-0.935) 0.344 (0.198-0.595)

P

0.0276 0.0001

HR: Hazard Ratios; CI: Confidence Interval; Chemo: chemotherapy; CIT: chemoimmunotherapy; ORR: overall response rate; CR: complete remission with incomplete marrow recovery; nPR: nodular partial response; F: female; M: male; NR: no remission; PD: progressive disease; PR: partial remission; SD: stable disease.

Table 5. Baseline characteristics of patients treated with chemoimmunotherapy in first-line in the bendamustine and rituximab (BR) and in the ibrutinib cohorts (UK + NPP GIMEMA).

Variable Median age, years (range) Age, years (%) ≤65/>65 Sex, (%) M/F ECOG PS (%) 0-1/≥2 Months between 1st line and 2nd line Median (range) n. <36/≥36 (%) Response to 1st line treatment (%) no/yes IGHV (%) mutated/unmutated 17p- (%) yes/no

BR (n=137)

Ibrutinib (n=71)

P

68.2 (39.4-84.6) 39 (34.5)/74 (65.5) 91 (66.4)/46 (33.6) 113 (91.9)/10 (8.1)

67.1 (27.5-85.3) 27 (38.6)/43 (61.4) 45 (63.4)/ 26 (36.6) 57 (82.6)/ 12 (17.4)

0.603 0.691 0.777 0.090

30.60 (0.40, 79.40) 81 (59.1)/56 (40.9) 28 (20.4)/109 (79.6) 17 (19.5)/70 (80.5) 16 (14.8)/92 (85.2)

19.40 (1.80, 77.60) 54 (76.1)/17 (23.9) 8 (15.1)/45 (84.9) 8 (32.0)/17 (68.0) 22 (36.1)/39 (63.9)

0.001 0.023 0.524 0.295 0.003

NPP GIMEMA: Named Patient Program-Gruppo Italiano Malattie Ematologiche dell'Adulto; ECOG: Eastern Cooperative Oncology Group; n: number. F: female; M: male; PS: performance status.

was minimal, if any, patient selection bias in our study. The number of grade 3-4 infections (6.7%) is similar to the 4.2% incidence of severe infections in a trial using bendamustine and ofatumumab in patients who had received 1-2 previous lines of treatment.16 In another trial using the BR regimen, the incidence of grade 3 infections was 12.8% in patients who had received 1-5 previous lines of treatment.15 Thus, our data show that BR is a relatively safe second-line regimen in terms of infectious complications in a real-life population. The lower incidence of grade 3-4 cytopenias in this study compared to other prospective studies showing a 50-78% incidence of grade 3-4 cytopenias15,16 reflects the policy not to perform a blood count in the routine practice at the nadir time point at many centers. With a 78.6% PFS rate at 12 months (median 25 months), a 31.3-month median TTNT, and a 92.7% OS rate at 12 months (median 74.5 months), our data show that the BR regimen is an effective first salvage regimen. Interestingly, the efficacy of this regimen in terms of PFS, TTNT and OS was not influenced significantly by age, creatinine clearance, by the presence of 2 or more comorbidities. PFS was negatively influenced by advanced stage, haematologica | 2018; 103(7)

del(17p) and U-IGHV, confirming the strong prognostic significance of these parameters24 also in the second-line setting. Patients without any of these unfavorable characteristics experienced a prolonged median PFS (40.4 months). Although PFS estimation should be interpreted with caution in a retrospective analysis, our data are similar to those observed in a prospective phase II GIMEMA trial16 that reported a median PFS of 23.6 months with bendamustine and ofatumumab in 49 R/R CLL (61% with 1 previous treatment, 39% with 2 previous lines). In another analysis of BR in patient who had received a median number of 2 previous treatments (range 1-5), the median PFS was 15.2 months (95%CI: 12.5-17.9 months).15 A 18month median PFS was reported with BR as first salvage after fludarabine, cyclophosphamide and rituximab in 62 patients.13 Time to next treatment was longer in those patients who had received chemotherapy as first-line treatment. It is worth noting that even though published guidelines proposed the preferential usage of CIT, 27.9% of our patients who started treatment after 2010 had received only chemotherapy as initial treatment. 1215


A. Cuneo et al.

The survival data in our analysis (92.7% alive at 12 months and 72.2% at 36 months) reflect previous experiences with bendamustine and anti-CD20 in clinical trials,16,25 showing that this combination is equally effective in clinical practice across many centers. Negative predictive factors on OS were represented by advanced stage and chemorefractory disease, whereas the presence of del(17p) was not associated in our analysis with a significantly shorter OS, possibly due to the relatively low number of patients (n=23) and to the use of effective salvage regimens in subsequent lines of treatment. Accordingly, the survival in the BR arm of the Helios trial after adjusting for crossover to BR and ibrutinib was close to 90% at 12 months and more than 80% at 24 months.26 Thus, the data presented here show that BR is an efficacious first salvage regimen in CLL in a real-life population, including elderly patients, patients with 2 or more comorbidities and a creatinine clearance less than 70 mL/min. The outcome was better in patients with favorable genetic features and with early/intermediate disease stage. Importantly, no significant differences in terms of OS were noted in this real-life report with respect to the survival data recently observed in clinical trials. Because no direct comparison was performed between CIT and new oral agents in first relapse, we elected to compare our data with ibrutinib used in a real-life patient population treated in the UK and in Italy using OS as an objective end point. We restricted our comparative analysis to patients who had previously received CIT because this is the recommended initial treatment in CLL. The BR and ibrutinib cohorts had similar baseline risk factors and, when excluding patients with del(17p) from the analysis as nowadays they would no longer be exposed to secondline CIT, there was no difference in OS (Figure 3). Interestingly, no difference in OS was found with BR or ibrutinib when including in the subanalysis patients with a less than 36-month interval between front-line and first salvage treatment. The survival curve showed an excess of early deaths in the ibrutinib cohort compared to the BR cohort. Due to the small size of the patient population included in this analysis, there is no recurrent pattern or obvious explanation for this observation; severe infection and Richterâ&#x20AC;&#x2122;s syndrome in 2 patients each were the only recurrent causes of death in the first 12 months. It remains unclear as to whether ibrutinib directly contributed to any of these deaths. The observed outcome for ibrutinib-treated patients in this observational study could be due to premature interruption of ibrutinib exposure. It is noteworthy that in the UK and Ireland data on the overall cohort of R/R CLL treated with ibrutinib, discontinuations during the first year were due to AEs (54%), Richterâ&#x20AC;&#x2122;s transformation (26%), and progressive CLL (17%). Beyond the first year, the rate of discontinuations due to progressive CLL increased to 29%.27 When comparing CIT and novel inhibitors, one also has

References 1. Fischer K, Bahlo J, Fink AM, et al. Longterm remissions after FCR chemoimmunotherapy in previously untreated

1216

to consider the long-term detrimental effects due to the clonal selection and DNA damage occurring with repetitive lines of chemotherapy-based treatments, resulting in second tumors, acute leukemias/myelodysplastic syndromes, that cannot be evaluated in the short follow up of our retrospective study. On the other hand, elegant in vitro studies have shown that treatment of mouse B cells with idelalisib or duvelisib, and to a lesser extent ibrutinib, increased somatic hypermutation through enhanced expression of activation-induced cytidine deaminase.28 In another analysis including R/R CLL in second and subsequent lines of treatment,29 an OS advantage was noted when comparing ibrutinib (with or without BR) and BR alone.30 In other studies arriving at the same conclusion,31,32 chemotherapy +/- anti CD20 regimens used in a real-world setting were compared with ibrutinib data of the clinical trials Resonate and Helios. However, there is now evidence that adherence to treatment with ibrutinib in the real-world population did not reflect the data obtained in clinical trials,22,33 possibly due to the heterogeneity of the patient populations or to a more limited experience of physicians in managing side effects occurring on treatment. Furthermore, the UK real-life data show that duration of ibrutinib therapy and OS seem very similar when ibrutinib is used at first or subsequent relapses, suggesting that the relative benefit for ibrutinib compared with chemotherapy is more evident in patients with multiple relapse where re-treatment with further chemotherapy results in progressively worse response rates and remission duration. It is noteworthy that a highly significant PFS advantage with the BCL2 inhibitor venetoclax plus rituximab compared to BR has been recently reported in the planned interim analysis of the randomized phase III Murano study, where 42.8% of patients had received 2 or more lines of therapy.34 However, in this trial, an OS benefit has not yet been shown according to the predefined statistical model. Although, obviously, data derived from different series must be treated with caution, these data suggest that BR and ibrutinib may be equally effective in terms of OS when used as first salvage treatment in CLL patients without 17p deletion managed in the real-life setting. Whether this is due to limited compliance of patients and/or suboptimal management of side effects with the novel therapies remains to be established. Funding AIL-Ferrara, Italy, Ricerca Finalizzata project RF-201102349712 Ministero della Salute, Rome, Italy to AC, RF, GG; MIUR-PRIN 2015ZMRFEA_004, Rome, Italy to AC, RF, PG, GG; Special Program Molecular Clinical Oncology 5 x 1000 N. 10007, Associazione Italiana per la Ricerca sul Cancro (AIRC), Milan, Italy; Bloodwise grant 16003 to FF. AS is supported by the United Kingdomâ&#x20AC;&#x2122;s National Institute for Health Research (NIHR). The views expressed in this paper are those of the authors and not necessarily those of the NIHR.

patients with CLL: updated results of the CLL8 trial. Blood. 2016;127(2):208-215. 2. Eichhorst B, Fink AM, Bahlo J, et al. Firstline chemoimmunotherapy with bendamustine and rituximab versus fludarabine, cyclophosphamide, and rituximab in

patients with advanced chronic lymphocytic leukaemia (CLL10): an international, open-label, randomised, phase 3, non-inferiority trial. Lancet Oncol. 2016;17(7):928942. 3. Goede V, Fischer K, Engelke A, et al.

haematologica | 2018; 103(7)


Efficacy of BR as first salvage treatment in CLL

Obinutuzumab as frontline treatment of chronic lymphocytic leukemia: updated results of the CLL11 study. Leukemia. 2015;29(7):1602-1604. 4. Foà R, Guarini A. A mechanism-driven treatment for chronic lymphocytic leukemia? N Engl J Med. 2013;369(1):85-87. 5. Byrd JC, Furman RR, Coutre SE, et al. Targeting BTK with ibrutinib in relapsed chronic lymphocytic leukemia. N Engl J Med. 2013;369(1):32-42. 6. Furman RR, Sharman JP, Coutre SE, et al. Idelalisib and rituximab in relapsed chronic lymphocytic leukemia. N Engl J Med. 2014;370(11):997-1007. 7. Roberts AW, Davids MS, Pagel JM, et al. Targeting BCL2 with Venetoclax in Relapsed Chronic Lymphocytic Leukemia. N Engl J Med. 2016;374(4):311-322. 8. Hallek M. Chronic lymphocytic leukemia: 2017 update on diagnosis, risk stratification, and treatment. Am J Hematol. 2017;92(9):946-965. 9. Buske C, Hutchings M, Ladetto M, et al. ESMO Consensus Conference on malignant lymphoma: general perspectives and recommendations for the clinical management of the elderly patient with malignant lymphoma. Ann Oncol. 2018;29(3):544562. 10. Wierda WG, Zelenetz AD, Gordon LI, et al. NCCN Guidelines Insights: chronic lymphocytic leukemia/small lymphocytic leukemia, Version 1.2017. J Natl Compr Canc Netw. 2017;15(3):293-311. 11. Green MR, Williams ME, Willey J, Buettner A, Neely D, Lankford M. First-Line Prescribing Preferences of U.S. Hematology-Oncology Physicians for Patients with CLL: Impact of Novel Agents [abstract]. Blood. 2014;124(21):4676. 12. Cramer P, Fink AM, Busch R, et al. Secondline therapies of patients initially treated with fludarabine and cyclophosphamide or fludarabine, cyclophosphamide and rituximab for chronic lymphocytic leukemia within the CLL8 protocol of the German CLL Study Group. Leuk Lymphoma. 2013;54(8):1821-1822. 13. Fornecker LM, Aurran-Schleinitz T, Michallet AS, et al. Salvage outcomes in patients with first relapse after fludarabine, cyclophosphamide, and rituximab for chronic lymphocytic leukemia: the French intergroup experience. Am J Hematol. 2015;90(6):511-514. 14. Knauf W, Abenhardt W, Dörfel S, et al. Routine treatment of patients with chronic lymphocytic leukaemia by office-based haematologists in Germany-data from the Prospective Tumour Registry Lymphatic Neoplasms. Hematol Oncol. 2015;33(1):1522.

haematologica | 2018; 103(7)

15. Fischer K, Cramer P, Busch R, et al. Bendamustine combined with rituximab in patients with relapsed and/or refractory chronic lymphocytic leukemia: a multicenter phase II trial of the German Chronic Lymphocytic Leukemia Study Group. J Clin Oncol. 2011;29(26):3559-3566. 16. Cortelezzi A, Sciumè M, Liberati AM, et al. Bendamustine in combination with ofatumumab in relapsed or refractory chronic lymphocytic leukemia: a GIMEMA Multicenter Phase II Trial. Leukemia. 2014;28(3):642-648. 17. Coutré SE, Furman RR, Flinn IW, et al. Extended treatment with single-agent Ibrutinib at the 420 mg dose leads to durable responses in chronic lymphocytic leukemia/small lymphocytic lymphoma. Clin Cancer Res. 2017;23(5):1149-1155. 18. O'Brien SM, Furman RR, Coutre SE, et al. Five-Year Experience with Single-Agent Ibrutinib in Patients with Previously Untreated and Relapsed/Refractory Chronic Lymphocytic Leukemia/Small Lymphocytic Leukemia. Blood. 2016; 128:233. 19. Brown JR, Hillmen P, O'Brien S, et al. Extended follow-up and impact of highrisk prognostic factors from the phase 3 RESONATE study in patients with previously treated CLL/SLL. Leukemia. 2018 32(1):83-91. 20. UK CLL Forum. Ibrutinib for relapsed/refractory chronic lymphocytic leukemia: a UK and Ireland analysis of outcomes in 315 patients. Haematologica. 2016;101(12):1563-1572. 21. Winqvist M, Asklid A, Andersson PO, et al. Real-world results of ibrutinib in patients with relapsed or refractory chronic lymphocytic leukemia: data from 95 consecutive patients treated in a compassionate use program. A study from the Swedish Chronic Lymphocytic Leukemia Group. Haematologica. 2016;101(12):1573-1580. 22. Ghia P, Cuneo A. Ibrutinib in the real world patient: many lights and some shades. Haematologica. 2016;101(12):1448-1450. 23. Hallek M, Cheson BD, Catovsky D, et al. Guidelines for the diagnosis and treatment of chronic lymphocytic leukemia: a report from the International Workshop on Chronic Lymphocytic Leukemia updating the National Cancer Institute-Working Group 1996 guidelines. Blood. 2008; 111(12):5446-5456. 24. International CLL-IPI working group. An international prognostic index for patients with chronic lymphocytic leukaemia (CLLIPI): a meta-analysis of individual patient data. Lancet Oncol. 2016;17(6):779-790. 25. Zelenetz AD, Barrientos JC, Brown JR, et al. Idelalisib or placebo in combination

26.

27.

28.

29.

30.

31.

32.

33.

34.

with bendamustine and rituximab in patients with relapsed or refractory chronic lymphocytic leukaemia: interim results from a phase 3, randomised, double-blind, placebo-controlled trial. Lancet Oncol. 2017;18(3):297-311. Chanan-Khan A, Cramer P, Demirkan F, et al. Ibrutinib combined with bendamustine and rituximab compared with placebo, bendamustine, and rituximab for previously treated chronic lymphocytic leukaemia or small lymphocytic lymphoma (HELIOS): a randomised, double-blind, phase 3 study. Lancet Oncol. 2016;17(2):200-211. Follows GA and CLL Forum UK. Outcomes of patients post ibrutinib treatment for relapsed/refractory CLL: a UK and Ireland analysis [abstract]. Hematol Oncol. 2017;35:237-238. Compagno M, Wang Q, Pighi C, et al. Phosphatidylinositol 3-kinase blockade increases genomic instability in B cells. Nature. 2017;542(7642):489-493. Hillmen P, Fraser G, Jones J, et al. Comparing Single-Agent Ibrutinib, Bendamustine Plus Rituximab (BR) and Ibrutinib Plus BR in Patients with Previously Treated Chronic Lymphocytic Leukemia/Small Lymphocytic Lymphoma (CLL/SLL): An Indirect Comparison of the RESONATE and HELIOS Trials. Blood. 2015;126(23):2944. Byrd JC, Brown JR, O'Brien S, et al. Ibrutinib versus ofatumumab in previously treated chronic lymphoid leukemia. N Engl J Med. 2014;371(3):213-223. Salles GA, Baseggio L, Bachy E, et al. Single-Agent Ibrutinib Vs Standard of Care for Patients with Relapsed/Refractory (R/R) and Treatment-Naive (TN) Chronic Lymphocytic Leukemia (CLL): An Adjusted Comparison of RESONATETM and RESONATE-2TM with the French Lyon-Sud Database. Blood. 2016;128(22):2039. Hansson L, Asklid A, Diels J, et al. Ibrutinib versus previous standard of care: an adjusted comparison in patients with relapsed/refractory chronic lymphocytic leukaemia. Ann Hematol. 2017; 96(10):1681-1691. Mato AR, Lamanna N, Ujjani CS, et al. Toxicities and Outcomes of IbrutinibTreated Patients in the United States: Large Retrospective Analysis of 621 Real World Patients. Blood. 2016;128(22):3222. Seymour JF, Kipps TJ, Eichhorst B, et al. Venetoclax Plus Rituximab Is Superior to Bendamustine Plus Rituximab in Patients with Relapsed/ Refractory Chronic Lymphocytic Leukemia - Results from PrePlanned Interim Analysis of the Randomized Phase 3 Murano Study [abstract]. Blood. 2017;130(Suppl 1):LBA-2.

1217


ARTICLE

Plasma Cell Disorders

Ferrata Storti Foundation

Repurposing tofacitinib as an anti-myeloma therapeutic to reverse growth-promoting effects of the bone marrow microenvironment Christine Lam,1,2 Ian D. Ferguson,1,2 Margarette C. Mariano,1,2 Yu-Hsiu T. Lin,1,2 Megan Murnane,2,3 Hui Liu,1,2 Geoffrey A. Smith,4 Sandy W. Wong,2,3 Jack Taunton,4 Jun O. Liu,5 Constantine S. Mitsiades,6 Byron C. Hann,2 Blake T. Aftab2,3 and Arun P. Wiita1,2,*

Haematologica 2018 Volume 103(7):1218-1228

Department of Laboratory Medicine, University of California, San Francisco, CA; 2Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA; 3 Department of Medicine, University of California, San Francisco, CA; 4Department of Cellular and Molecular Pharmacology, University of California, San Francisco, CA; 5 Department of Pharmacology and Molecular Sciences, Johns Hopkins School of Medicine, Baltimore, MD and 6Department of Medical Oncology, Dana-Farber Cancer Institute, Boston, MA, USA 1

ABSTRACT

T

Correspondence: arun.wiita@ucsf.edu

Received: June 12, 2017. Accepted: March 15, 2018. Pre-published: April 5, 2018.

he myeloma bone marrow microenvironment promotes proliferation of malignant plasma cells and resistance to therapy. Activation of JAK/STAT signaling is thought to be a central component of these microenvironment-induced phenotypes. In a prior drug repurposing screen, we identified tofacitinib, a pan-JAK inhibitor Food and Drug Administration (FDA) approved for rheumatoid arthritis, as an agent that may reverse the tumor-stimulating effects of bone marrow mesenchymal stromal cells. Herein, we validated in vitro, in stromal-responsive human myeloma cell lines, and in vivo, in orthotopic disseminated xenograft models of myeloma, that tofacitinib showed efficacy in myeloma models. Furthermore, tofacitinib strongly synergized with venetoclax in coculture with bone marrow stromal cells but not in monoculture. Surprisingly, we found that ruxolitinib, an FDA approved agent targeting JAK1 and JAK2, did not lead to the same anti-myeloma effects. Combination with a novel irreversible JAK3-selective inhibitor also did not enhance ruxolitinib effects. Transcriptome analysis and unbiased phosphoproteomics revealed that bone marrow stromal cells stimulate a JAK/STAT-mediated proliferative program in myeloma cells, and tofacitinib reversed the large majority of these pro-growth signals. Taken together, our results suggest that tofacitinib reverses the growthpromoting effects of the tumor microenvironment. As tofacitinib is already FDA approved, these results can be rapidly translated into potential clinical benefits for myeloma patients.

doi:10.3324/haematol.2017.174482 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/103/7/1218 Š2018 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.

1218

Introduction Multiple myeloma (MM) is the second most common hematologic malignancy in the United States of America and still has no known cure. Years of research have revealed that a major driver of malignant plasma cell proliferation, as well as therapeutic resistance, is signaling to the tumor cells from the bone marrow (BM) microenvironment.1-3 Cell types within the BM that influence myeloma cells include mesenchymal stromal cells, osteoblasts, osteoclasts, and multiple classes of immune cells.1-3 Overcoming the growth-promoting phenotype of the BM microenvironment is thought to be a promising therapeutic strategy in MM. One approach to identifying new therapeutic agents for many diseases is drug repurposing. In this context, a large library of drugs, all of which are either Food and Drug Administration (FDA)-approved, or at the minimum shown to be safe in humans, is screened against the biological system of interest.4,5 The premise behind these screens is that small molecules, initially designed for one indication, may actually have beneficial effects across other diseases. In fact, the use of thalidomide in MM is one of the most impactful examples of successful drug repurposing. If new indications are found for already existing drugs, clinical develhaematologica | 2018; 103(7)


Repurposing tofacitinib as anti-myeloma therapy

opment times and associated costs can be drastically reduced, thus accelerating the potential benefits to patients.6,7 To identify agents which may reverse the tumor-promoting effects of the MM BM microenvironment, we recently reported a repurposing screen of 2684 compounds against three MM cell lines, either grown alone (monoculture) or in coculture with MM patientderived BM mesenchymal stromal cells.8 From that screen, we identified tofacitinib citrate, an FDA-approved small molecule for the treatment of rheumatoid arthritis (RA), as an agent which may reverse stromal-induced growth proliferation of malignant plasma cells. Tofacitinib citrate is a potent inhibitor of all four members of the Janus kinase (JAK) family, with preferential inhibition of JAK1 and JAK3 over JAK2 and TYK2 in cellular assays.9 JAK signaling, mediated by downstream STAT transcription factors, is necessary for lymphocyte stimulation in response to encountered antigens.10 Therefore, JAK inhibition holds promise for the treatment of autoimmune diseases like RA.11 In parallel, the JAKs have gained interest as therapeutic targets in MM as they mediate signaling via interleukin-6 (IL-6). IL-6 is secreted by many cell types within the BM microenvironment, as well as by malignant plasma cells themselves, and it is thought that proliferation of malignant plasma cells within the human BM is dependent on this cytokine.12 This dependence on IL-6 was underscored by the recent development of a patientderived xenograft model of MM, where primary plasma cell growth only occurred in immunocompromised mouse BM after knock-in of human IL-6.13 In fact, a number of groups have variously targeted JAK1/JAK2,14,15 JAK2,16-18 or all four JAKs19 with reported preclinical therapeutic efficacy in MM. However, per the registry at clinicaltrials.gov, none of these experimental JAK inhibitors have entered into MM clinical trials. Therefore, all of these agents are very far from use in MM patients, if they ever become available. Herein, we demonstrate that the already FDA-approved agent tofacitinib has robust preclinical activity in MM models. We further use ribonucleic acid sequencing (RNA-seq) and unbiased mass spectrometry-based phosphoproteomics to delineate pro-proliferative signals from the BM stroma and show that they are largely reversed by tofacitinib treatment. Furthermore, we find that an alternate repurposing candidate, the FDA-approved JAK1/2 inhibitor ruxolitinib, surprisingly does not show the same antimyeloma properties. Therefore, our results support the rapid repurposing of tofacitinib as an anti-myeloma therapeutic to reverse the pro-growth effects of the BM microenvironment and potentiate the effects of existing myeloma therapies.

Methods Cell culture conditions All cell lines were authenticated by DNA genotyping at ATCC. All cells, including patient BM mononuclear cells, were maintained in complete media with roswell park memorial institute1640 (RPMI-1640; Gibco) supplemented with 10% fetal bone serum (FBS; Gemini), 1% penicillin-streptomycin University of California San Francisco (UCSF), and 2 mM L-Glutamine (UCSF) with 5% CO2. INA-6 media was supplemented with 50 ng/mL recombinant human IL-6 (ProSpec). Additional details are provided in the Online Supplementary Methods. haematologica | 2018; 103(7)

MM and bone marrow stromal cells (BMSC) coculture and viability testing Cocultures were seeded into 384 well plates (Corning) with the Multidrop Combi (Thermo Scientific). 800 stromal cells were seeded and incubated overnight. 17 hours later, 700 myeloma cells were added on top of the stromal cells. On the third day, 24 hours after the addition of myeloma cells, cocultures were treated with tofacitinib (LC Laboratories), ruxolitinib (Selleck Chemicals), JAK3i,20 or IL-6 blocking antibody (R&D Systems). For drug combination studies, on the fourth day, melphalan (Sigma Aldrich), carfilzomib (Selleck Chemicals), or venetoclax (Selleck Chemicals) were additionally added to cocultures. On the fifth day, myeloma cell viability was detected with the addition of luciferin (Gold Biotechnology) and read for luminescence on Glomax Explorer plate reader (Promega) as previously described.21 For monoculture studies cell viability was measured using CellTiter-Glo reagent (Promega). All measurements were performed in quadruplicate. All viability data are reported as normalized to dimethyl sulfoxide (DMSO)-treated cell line in monoculture.

RNA-seq For coculture RNA-seq, 5x106 MM.1S cells were grown with 3x106 HS5 cells for 24 hours. CD138+ enrichment to >95% was verified by flow cytometry for mCherry expression (Online Supplementary Figure S1A,B). MM.1S harvested from coculture, as well as MM.1S and HS5 grown in monoculture, were then processed for RNA-seq as previously described.22 Significantly upregulated transcripts were identified by DESeq23 and bioinformatic analysis was performed using Enrichr.24 Raw sequencing data are available at the Gene Expression Omnibus (GEO) repository (Accession number GSE99293). Additional details are provided in the Online Supplementary Methods.

Western Blot Analysis Described in the Online Supplementary Methods.

Liquid chromatographyâ&#x20AC;&#x201C;tandem mass spectrometry phosphoproteomics For coculture experiments, 5x106 HS5 cells were seeded into a T75 flask. Seventeen hours later, cultures were washed with phosphate buffered saline (PBS), before the addition of 107 MM.1S mC/Luc cells. Twenty-four hours later, cocultures were treated with 1 mM tofacitinib for 1.5 hours and 24 hours. MM.1S cells in suspension were harvested by aspiration, centrifuged, washed with PBS, and flash-frozen prior to analysis. For untreated MM.1S monoculture or HS5 monoculture experiments, 107 cells were used. For sample preparation, frozen cell pellets were lysed in 8 M urea. 1 mg of total protein was then reduced in tris(2-carboxyethyl)phosphine (TCEP) and free cysteines alkylated with iodoacetamide. Proteins were then digested at room temperature for 18 hours with trypsin. Peptides were desalted, lyophilized, and enriched for phosphopeptides using immobilized-metal affinity column (IMAC) with Fe-NTA loaded beads.25 Phosphopeptides were analyzed on a Thermo Q-Exactive Plus mass spectrometer coupled to a Dionex Ultimate 3000 NanoRSLC liquid chromatography instrument with 3.5 hour linear gradient. Raw proteomic data files are available at the ProteomXchange PRIDE repository (Accession number PXD006581). Additional details are provided in the Online Supplementary Methods.

Xenograft mouse model NOD.Cg-Prkdcscid Il2rgtm1Wjl/SzJ (NSG) mice were obtained from the Jackson laboratory. 106 MM.1S mC/Luc cells, stably expressing luciferase, were transplanted via tail vein injection into each mouse. Tumor burden was assessed through weekly biolumines1219


C. Lam et al.

cent imaging, beginning 13 days after implantation and on the same day as treatment initiation. Mice were treated for four weeks with vehicle or tofacitinib as indicated (five mice/arm.) Tofacitinib was formulated in 50% DMSO, 10% (polyethylene glycol 400) PEG 400, and 40% water and administered at 21.5 mg/kg daily by continuous subcutaneous infusion. All mouse studies were performed according to UCSF Institutional Animal Care and Use Committee-approved protocols.

Patient samples De-identified primary MM BM samples were obtained from the UCSF Hematologic Malignancy Tissue Bank in accordance with the UCSF Committee on Human Research-approved protocols and the Declaration of Helsinki. BM mononuclear cells were isolated by density gradient centrifugation Histopaque-1077 (Sigma Aldrich), then adjusted to 2 x 105/well in a 96 well plate. Primary cells were stimulated with 50 ng/ml recombinant human IL-6 (ProsPec) for 17 hours before treatment with tofacitinib for 24 hours. Cells were then stained with Alexa-Fluor 647 mouse antihuman CD138 antibody (BD Pharmingen) and SYTOX Green (Thermo) and analyzed on a CytoFLEX instrument (BD).

strongly increased compared to monoculture growth after 24 hours, confirming stromal-induced proliferative signaling in this cell line. Tofacitinib treatment reduced MM.1S cell numbers in a dose-dependent manner, such that at >1 mM tofacitinib, MM.1S cell numbers in coculture return to approximately monoculture levels (Figure 1A). Tofacitinib has no effect on MM.1S cell viability alone nor on stromal cells alone (Figure 1B). We further studied the effect of tofacitinib on several other MM cell lines. In monoculture we found that tofacitinib only demonstrates strong anti-MM activity in the IL-6 dependent cell line INA-6, with limited effect on the AMO-1 cell line and minimal to no effect on the other MM cell lines (Figure 1C). We further evaluated four myeloma cell lines (MM.1S, RPMI-8226, JJN-3, AMO-1) in which luciferase was stably expressed, allowing for the distinction of MM cell viability versus stromal cell viability in coculture.21 Only the stromal-responsive cell lines exhibit any sensitivity to tofacitinib treatment (Figure 1D). Taken together, these results suggest that tofacitinib selectively targets the growth-promoting interaction between MM cells and the stromal microenvironment known to occur in patients.

Results BMSC-mediated plasma cell proliferation is through a mechanism partially dependent on IL-6

Tofacitinib targets the BM microenvironment and reverses BMSC-mediated growth promotion To initially validate findings from our drug repurposing screen, we cocultured the human MM cell line MM.1S, which was included in the screen,8 with the immortalized BMSC lines HS5 and HS27A (Figure 1A) and low-passage stromal cells derived from primary myeloma patient BM (Online Supplementary Figure S1C). MM.1S cell numbers

A

B

C

D

We next focused on the MM.1S cell line as it showed the unique phenotype of responsiveness to tofacitinib only in the context of stromal stimulation. To further characterize the nature of pro-growth signaling, we performed RNA-seq on MM.1S cells grown alone or in coculture with HS5 stromal cells. We first noticed that the most significantly upregulated transcript in MM.1S in the cocul-

Figure 1. Tofacitinib inhibits stromal cell-mediated proliferation in MM cells. A. Tofacitinib has no effect vs. MM.1S MM cells in monoculture, but instead reverses proliferation induced by BMSC lines HS5 and HS27A. B. Tofacitinib has no viability effect vs. BMSC C. Tofacitinib has limited or minimal effects vs. most MM cell lines in monoculture, except the IL-6 dependent line INA-6. D. In stromal cell coculture, tofacitinib does not have anti-MM effects vs. JJN-3 and RPMI-8226 cell lines, which do not proliferate in response to stroma, but shows strong reversal of proliferation in both MM.1S and AMO1 lines. All error bars represent +/- S.D. from CellTiter-Glo assay performed in quadruplicate in 384-well plates. PBS: phosphate buffered saline.

1220

haematologica | 2018; 103(7)


Repurposing tofacitinib as anti-myeloma therapy

ture setting was SOCS3, part of a well-characterized negative feedback mechanism strongly induced by JAK-STAT activation10 (Figure 2A). We further examined all 67 transcripts significantly upregulated in MM.1S in the coculture vs. monoculture setting (P< 0.005 per DESeq tool;23 listed in Online Supplementary Dataset S1). Using the Enrichr tool,24 ChIP-X enrichment analysis (ChEA) of chromatin immunoprecipitation sequence (ChIP-seq) datasets26 found the most significant enrichment of STAT3-binding sites at the promoter of these upregulated transcripts, among all transcription factors (Figure 2B). Furthermore, Protein ANalysis THrough Evolutionary Relationships (Panther)27 pathway analysis found the only two significantly enriched pathways to be related to JAK/STAT signaling and interleukin signaling (Figure 2C). Taken together, these RNA-seq findings suggest that factors secreted from stromal cells mediate proliferation by activating JAK/STAT signaling, with STAT3 playing a central role.28 Notably, similar activation of SOCS3 and related JAK/STAT genes were previously found after exposure to IL-6 in the INA-6 cell line.29 However, using recombinant cytokines, even at high concentrations of IL-6, we did not observe the same degree of growth promotion as coculture with stromal cells (Figure 2D). A recent study identified other cytokines besides IL-6 highly secreted from HS5 stromal cells.30 We

A

D

tested four of these: macrophage inflammatory protein 3A (MIP-3A), interleukin-8 (IL-8), granulocyte colony stimulating factor (G-CSF), granulocyte-macrophage colony stimulating factor (GM-CSF), and none showed any growth promotion (Figure 2D). Furthermore, an IL-6 blocking antibody could only partially reverse stromal cell effects at the highest achievable concentration (Figure 2E). These findings support the role of IL-6 in this system, but also indicate that other factors likely play a role in MM growth promotion.

Tofacitinib inhibits JAK/STAT signaling Given our results above, we chose to further evaluate downstream effects of tofacitinib inhibition of the JAK/STAT pathway. Two of the primary downstream mediators of IL-6 receptor and JAK activation are thought to be pro-proliferation signaling by STAT3 and inhibitory signaling by STAT1.28 We found that STAT3 and STAT1 phosphorylation in MM.1S dramatically increases when in coculture with HS5 (Figure 3A,B). 1 mM tofacitinib inhibits STAT3 phosphorylation in MM.1S cells in coculture almost to monoculture level by two hours of treatment. In addition, phosphorylation of JAK1, JAK2, and TYK2, which can also activate STAT3, were studied. We found evidence of a â&#x20AC;&#x153;reboundâ&#x20AC;? effect by 24 hours of treatment, mediated by well-characterized feedback mecha-

C

B

E

Figure 2. Stromal-induced signatures in MM.1S identified by transcriptome analysis. A. Examples of significantly upregulated genes in MM.1S cells cocultured with HS5 stromal cells in comparison to MM.1S grown in monoculture. B. ChEA analysis of 67 significantly upregulated transcripts from untreated MM.1S in HS5 coculture vs. monoculture (P<0.005 based on DESeq analysis) demonstrates a significant enrichment of STAT3 transcription-factor binding sites based on ChIP-seq data. C. Panther pathway analysis of this gene list demonstrates significant upregulation of interleukin signaling and JAK-STAT signaling. D. Recombinant cytokines known to be secreted from HS5 stromal cells30 were tested for their ability to promote MM.1S proliferation. Up to the maximum achievable concentration, neither IL-6 alone nor a combination of all tested cytokines could recapitulate the growth promotion induced by BMSC. E. An IL-6 neutralizing antibody could partially reverse stromalinduced proliferation of MM.1S. All error bars represent +/- S.D. from CellTiter-Glo assay performed in quadruplicate in 384-well plates. PBS: phosphate buffered saline; RNA-seq: ribonucleic acid sequencing; ChIP-seq: chromatin immunoprecipitation sequencing.

haematologica | 2018; 103(7)

1221


C. Lam et al. nisms,10 serving to phosphorylate the JAKs and subsequently re-activate STAT3. Despite this rebound of STAT3 activation, however, MM growth continues to be inhibited based on the dose-response results of tofacitinib. In INA-6 cells we found similar effects of decreased STAT3 phosphorylation after tofacitinib treatment and a rebound at 24 hours (Online Supplementary Figure S2A). As tofacitinib is known to potently inhibit JAK3, of additional interest was an increase in JAK3 expression in MM.1S in coculture versus monoculture, found both by RNA-seq (Figure 2A) and Western blot (Figure 3C). However, tofacitinib did not lead to any significant decrease in JAK3 phosphorylation (Figure 3C). We also evaluated signaling through JAK3â&#x20AC;&#x2122;s primary pro-proliferative downstream mediator STAT5.31 We found no evidence of STAT5 phosphorylation in either monoculture or coculture (Online Supplementary Figure S2B). These results suggest that JAK3/STAT5 signaling is less central to stroma-supported MM growth. We further confirmed this result using JAK3i, a newly-described, highly-specific, irreversible JAK3 inhibitor.31 JAK3i had no effect on MM.1S in mono- or coculture (Figure 4A), nor did it show any synergy with carfilzomib treatment (Online Supplementary Figure S2C,D). Taken together, these results suggest a mechanism whereby stromal cell-induced MM proliferation is mediated through STAT3 transcriptional effects. These signaling pathways are inhibited by tofacitinib, ultimately leading to the reversal of the proliferation phenotype.

Ruxolinitib has less anti-MM activity than tofacitinib Given that our results above suggest a more important role for JAK1 and/or JAK2 in stromal-induced MM proliferation than JAK3, we turned to an alternate candidate for drug repurposing, ruxolinitib. This agent is FDA-approved for use in myeloproliferative neoplasms and has much

A

higher affinity for JAK1 and JAK2 over JAK3 or TYK2.32 Ruxolitinib had minimal effects versus RPMI-8226 or JJN-3 cells, and, surprisingly, treatment of MM.1S actually showed a promotion of growth at higher concentrations, both in the monoculture and coculture settings (Figure 4B,C). Western blotting demonstrated that 1 mM ruxolitinib was unable to inhibit STAT3 activation in MM.1S in coculture (Figure 4D). In fact, STAT3 phosphorylation increased after two hours of treatment, consistent with the pro-proliferative effect seen in Figure 4B,C. We found that combining ruxolitinib with JAK3i was also insufficient to recapitulate tofacitinibâ&#x20AC;&#x2122;s effects in mono- or coculture (Figure 4E,F). Taken together, these findings suggest that ruxolitinib is unable to inhibit pro-proliferative STAT3 signaling in MM.1S cells, thereby supporting tofacitinib as having greater potential as a repurposed anti-MM therapy.

Unbiased phosphoproteomics demonstrates that tofacitinib broadly reverses pro-growth signaling induced by bone marrow stroma Our targeted investigations above specifically focused on the JAK/STAT pathway. To further elucidate the mechanism of tofacitinib in this system, as well as gain a broader view of stromal-induced proliferation and signaling in MM cells, we pursued unbiased mass spectrometry (MS)based phosphoproteomics. We studied four samples, performed in biological replicate: 1) MM.1S cells in monoculture, MM.1S cells in coculture, either 2) untreated (DMSO control), 3) treated with 1 mM tofacitinib for 1.5 hours, or 4) treated with 1 mM tofacitinib for 24 hours. We harvested MM.1S cells in suspension, enriched for phosphorylated peptides using immobilized metal chromatography, and analyzed by liquid chromatography-tandem mass spectrometry (LC/MS-MS) with peptide quantification performed using MaxQuant.33

B

C

Figure 3. Tofacitinib inhibits JAK/STAT signaling in MM cells in a time-dependent manner. A. Western blotting demonstrates a marked increase in STAT3 phosphorylation in untreated coculture vs. monoculture in MM.1S cells. After treatment with 1 mM tofacitinib there is a rapid decrease in STAT3 phosphorylation with rebound by 24 hours. B. STAT1 phosphorylation is also increased in response to coculture and rapidly reversed by tofacitinib treatment. C. While coculture increases JAK3 protein expression, there does not seem to be any significant change in signaling via JAK3 after tofacitinib based on phosphorylation status. Western blots are representative of assays performed in biological duplicate. Monocult: monoculture; Cocult: coculture.

1222

haematologica | 2018; 103(7)


Repurposing tofacitinib as anti-myeloma therapy

In total, 4862 phosphopeptides had intensity data in all four samples and were used for further analysis (listed in Online Supplementary Dataset S2). >99% of these sites are serine and threonine phosphorylation events, consistent with other phosphoproteomic studies using this enrichment method.34 We first evaluated for phosphosites with >4-fold intensity increases in untreated coculture vs. monoculture. Using our RNA-seq data as a proxy for protein-level changes, we verified these phospho-site changes were largely driven by changes in signaling and not protein abundance (Online Supplementary Figure S3A). Panther pathway analysis revealed the only significantly enriched pathway among the 544 upregulated phosphosites to be JAK/STAT signaling (Figure 5A). However, based on

kinase enrichment analysis,35 we found enriched signatures not only of JAK1 substrates, but also of other kinases driving proliferation and the cell cycle, such as mTOR, CDK1, and CDK2 (Figure 5B). These findings demonstrate that unbiased phosphoproteomics can uncover broad signaling effects of the BM microenvironment even downstream of JAK/STAT. Next, for validation of the effects of tofacitinib treatment, we first examined Ser727 on STAT3 (Figure 5C), a known JAK-responsive phosphosite.28 Our quantitative MS results were remarkably in line with Western blotting for another JAK-responsive phosphosite on STAT3, Tyr707 (Figure 3A), with a very large increase in both phosphosites in untreated coculture compared to baseline,

A

B

C

D

E

F

Figure 4. Ruxolitinib demonstrates less anti-MM activity than tofacitinib. A. A highly selective, irreversible inhibitor of JAK3, JAK3i, does not have any effects on MM.1S either in monoculture or in coculture with HS5. B. The JAK1/2 inhibitor ruxolitinib has minimal anti-MM effects in monoculture vs. three MM cell lines, and in fact appears to promote growth of MM.1S at higher concentrations. C. A similar phenomenon is noted in HS5 coculture. D. Ruxolitinib does not inhibit, but in fact increases signaling via STAT3 in MM.1S grown in HS5 coculture. E.-F. Combination of ruxolitinib with JAK3i, to achieve simultaneous JAK1/2/3 inhibition, does not recapitulate effects of tofacitinib in MM.1S. All error bars represent +/- S.D. from CellTiter-Glo assay performed in quadruplicate in 384-well plates. Monocult: monoculture; Cocult: coculture; PBS: phosphate buffered saline.

haematologica | 2018; 103(7)

1223


C. Lam et al.

a >2-fold decrease in phosphorylation after short-term tofacitinib treatment, and a rebound in phosphorylation at 24 hours. This finding both serves to validate our phosphoproteomic data as well as help us define a signature of tofacitinib-responsive phosphosites. Remarkably, 336 of the 544 up-regulated phosphosites in untreated coculture vs. monoculture (62%) met the same criteria of being tofacitinib-responsive (Figure 5D,E). Furthermore, examination across all measured phosphosites demonstrated that while phosphorylation is broadly increased in the coculture setting, noted as a general shift toward positive phosphosite intensities in the untreated sample, treatment with tofacitinib largely reverses this finding, recreating a normal distribution around the intensity values found in monoculture (Figure 5F). Taken together, these findings demonstrate that tofacitinib broadly reverses the signaling pathways driving stromal-induced proliferation in MM cells, both at the level of direct JAK targets as well as downstream proliferative signals, informing our mechanistic

understanding of this treatment beyond targeted Western blots alone. As a comparator, we also performed unbiased phosphoproteomics of HS5 stromal cells after 1.5 hours tofacitinib treatment. We found many fewer significantly changed phosphosites than identically-treated MM.1S in HS5 coculture (Online Supplementary Figure S3B,C), consistent with the lack of viability change after tofacitinib against HS5 alone (Figure 1B).

Phosphoproteomics does not reveal specific off-target activity for tofacitinib Another advantage of unbiased phosphoproteomics is potentially detecting additional off-target effects mediating tofacitinib response. We filtered for peptides that appeared unaffected by stromal-induced signaling (less than +/- 50% intensity change in untreated coculture vs. monoculture) that decreased in intensity >4-fold after 1.5 hours tofacitinib treatment (Figure 6A). We identified only 54 peptides that fit this filter, and neither Panther nor

A

B

C

D

E

F

Figure 5. Unbiased phosphoproteomics reveals that tofacitinib broadly reverses pro-growth signaling from stroma to MM cells. A. Analysis of 4862 phosphorylation sites quantified by LC-MS/MS on biological replicate samples revealed 544 phosphosites to be upregulated 4-fold in untreated MM.1S in coculture with HS5 vs. monoculture. Of these upregulated phosphopeptides, Panther pathway analysis showed JAK/STAT signaling to be the only significantly enriched pathway (P=0.036). B. Kinase enrichment analysis demonstrated that many of the upregulated phosphopeptides derive from known substrates of other kinases related to proliferation, such as mTOR, CDK1, and CDK2, as well as JAK1. Phosphorylated protein substrates are on the left and enriched kinases across the top. Length of red bar in kinase name is indicative of strength of enrichment. C. Quantitative phosphoproteomic intensity of the JAK-responsive phosphosite Ser727 on STAT3, both in untreated coculture vs. monoculture, and after tofacitinib treatment, is very similar to the pattern found by Western blotting for the other known JAK responsive STAT3 phosphosite Tyr707 (Figure 3A), serving to validate this proteomics approach. D. Dynamics of all phosphosites found to be responsive to tofacitinib based on the criteria: 1) increased 4-fold in coculture vs. monoculture, 2) decreased at least-2 fold from untreated coculture after 1.5 hours of 1 mM tofacitinib treatment, and 3) phosphosite intensity remains below the untreated coculture level after 24 hours of tofacitinib treatment. E. Of 544 upregulated phosphopeptides in coculture, 336 (62%) were defined as being tofacitinib-responsive using the criteria in D. F. Examining the global phosphosite intensity across all quantified phosphopeptides demonstrates a general increase in phosphorylation of MM.1S proteins in untreated coculture with HS5 (â&#x20AC;&#x153;Cocx DMSOâ&#x20AC;?; note shift of distribution maximum to log2-fold change vs. monoculture of ~1) which is then broadly reversed by tofacitinib treatment. Tof: tofacitinib; monocx: monoculture; cocx: coculture.

1224

haematologica | 2018; 103(7)


Repurposing tofacitinib as anti-myeloma therapy

kinase enrichment analysis identified any enriched signatures (data not shown). Given the small number of peptides and lack of any biological signatures, it appears most likely these 54 peptides are background noise in the data and do not reveal any specific off-target effects of tofacitinib. To further investigate the possibility of any off-target effects of tofacitinib, we downloaded available data from the LINCS KINOMEscan database and plotted versus the human kinase phylogenetic tree (Online Supplementary Figure S4). These results demonstrate that, at least in cellfree assays, tofacitinib shows much greater specificity for JAK-family kinases compared to ruxolitinib. Together, these findings suggest that tofacitinibâ&#x20AC;&#x2122;s effects in this coculture model appear to be due to on-target activity. However, these results do not rule out off-target effects of tofacitinib not detected by these analyses herein.

Tofacitinib synergizes strongly with venetoclax only in the coculture setting To begin to identify potential rational combinations with myeloma therapies, we first studied tofacitinib in combination with carfilzomib in MM.1S cells. Based on zero-inflated Poisson regression (ZIP) model scoring,30,36

A

B

C

D

we found very mild synergy with this proteasome inhibitor in both monoculture or HS5 coculture (Online Supplementary Figure S5). Subsequently, inspired by recent findings in acute myeloid leukemia (AML) primary samples,30 we tested the combination of tofacitinib and the Bcl-2 inhibitor venetoclax, a promising investigational agent in MM.37 Intriguingly, similar to the findings in AML, we found strong synergy of these two agents only in the coculture context but not in monoculture (Figure 6).

Tofacitinib has anti-MM activity in the BM microenvironment in vivo Toward the goal of repurposing tofacitinib as an antiMM therapy in patients, we next examined the efficacy of tofacitinib in vivo. For this orthotopic disseminated xenograft model, we used a luciferase-labeled MM.1S cell line, which specifically homes to the murine BM after intravenous implantation in NOD scid g (NSG) mice. Treatment was initiated after two weeks of tumor growth and continued for four weeks at ~2/3 of the maximal tolerated dose of tofacitinib (21.5 mg/kg/day by subcutaneous infusion).38 Encouragingly, we found significantly increased murine survival in this cell line model (Figure

Figure 6. Tofacitinib shows synergy with venetoclax only in the coculture setting. A-B. Treatment was performed sequentially with 24 hours of tofacitinib followed by 24 hours of venetoclax at specified doses. C-D. For combination matrices, the interaction landscapes are shown in 2D plots. The ZIP method is used to calculate synergy across the landscape (red = positive score, synergistic; green = negative score, antagonistic) as well as to calculate an overall synergy score d, the difference in percentage inhibition compared with the expected additive compound effect. The coculture results demonstrate synergy across all combinations with the strongest synergy at high doses of venetoclax, as well as an overall very high synergy score of ~10.5 across the landscape. The monoculture results demonstrate mild antagonism at high venetoclax doses and overall antagonistic score of ~-1.5. Note different scale bars on each 2D landscape output from ZIP. Error bars represent +/- S.D. from CellTiter-Glo assay performed in quadruplicate in 384-well plates. ZIP: zero-inflated Poisson regression.

haematologica | 2018; 103(7)

1225


C. Lam et al.

7A) as well as decreased tumor burden based on bioluminescent imaging quantification (Figure 7B,C). We further tested tofacitinib versus two primary MM BM samples treated ex vivo with the co-addition of 50 ng/mL IL-6. We found modest viability effects against malignant plasma cells (Figure 7D and Online Supplementary Figure S6). This result appears consistent with our in vitro and phosphoproteomic results, which suggest that plasma cell proliferation is necessary for tofacitinib to have significant effects. Primary MM plasma cells isolated ex vivo in 2D culture are known to mini-

A

mally proliferate even in the presence of cytokines or stromal stimulation.39 Therefore, these results may not fully reflect the potential therapeutic efficacy of tofacitinib in MM patients, where plasma cells are constantly proliferating within the BM.

Discussion Given the importance of the BM microenvironment for MM pathogenesis, the JAK/STAT signaling axis has gener-

B

C

D

Figure 7. Tofacitinib has anti-MM activity in vivo and versus primary samples ex vivo. Luciferase labeled-MM.1S MM cells were implanted intravenously into NSG mice and tumors allowed to grow for 13 days. Mice were randomized (n=5 mice per arm) and drug dosing was then begun for four weeks. Tofacitinib dose = 21.5 mg/kg/day by subcutaneous pump. A. Tofacitinib significantly increased survival of NSG mice in these aggressive mouse models of MM by log-rank test. B. Example bioluminescent images from MM.1S study showing prominent localization of tumor cells to hind limb BM. C. Bioluminescence imaging quantification of tumor burden. Error bars represent +/- S.D. D. Primary CD138+ plasma cells from two patients show modest response to 24 hour tofacitinib treatment when cultured ex vivo (with addition of 50 ng/mL IL-6 to the media) and measured by flow cytometry (n = 2 technical replicates). Veh: vehicle; Tof: tofacitinib; DMSO: dimethyl sulfoxide.

1226

haematologica | 2018; 103(7)


Repurposing tofacitinib as anti-myeloma therapy

ated significant interest as a therapeutic target in MM. JAK inhibition has already been validated in a number of preclinical studies as a way to target this pathway.14-19 However, the studied compounds are not yet available clinically, and may never be. Herein, following the results of a large-scale repurposing screen, we validated tofacitinib as a potential therapy that can be rapidly translated into MM patients. Using a combination of mechanistic pharmacology and unbiased mass spectrometry-based phosphoproteomics, we found that tofacitinib appears to reverse stromalinduced proliferation of MM plasma cells by inhibiting JAK/STAT signaling. Our results also support the use of unbiased phosphoproteomics both in kinase inhibitor evaluation and more broadly in MM biology. While others have shown that IL-6 can lead to increased signaling through STAT3,40 our results suggest that additional factors derived from BMSC can lead to plasma cell proliferation. While we were unable to identify these factors, we did rule out a number of cytokines highly secreted by HS5 cells.30 Furthermore, given that murine IL-6 does not crossreact with the human IL-6 receptor,41 our in vivo results also suggest that other factors in the murine marrow microenvironment can stimulate JAK-STAT signaling, which is then reversed by tofacitinib. Candidates include leukemia inhibitory factor (LIF), which cross-reacts between mouse and humans,41,42 or the complex between murine IL-6 and soluble murine IL-6 receptor, both of which can stimulate JAK/STAT signaling in human cells.43 Surprisingly, in our studies, we found that the FDAapproved JAK1/2 inhibitor ruxolitinib did not lead to the same anti-MM effects as tofacitinib. We do note that ruxolitinib was previously evaluated in a small trial of 13 MM patients in combination with dexamethasone (clinicaltrials.gov Identifier: 00639002), and no significant anti-MM effects were noted in this small study. Our in vitro results presented herein may provide a partial explanation for the lack of ruxolitinib efficacy in that trial. Another recent study also found minimal anti-MM cell line effects of ruxolitinib unless used at extremely high doses and exposure times.44 However, it remains mechanistically unclear why ruxolitinib is unable to block JAK-STAT signaling in these myeloma models. We note that our studies are undoubtedly limited in that

References 5. 1. Manier S, Kawano Y, Bianchi G, Roccaro AM, Ghobrial IM. Cell autonomous and microenvironmental regulation of tumor progression in precursor states of multiple myeloma. Curr Opin Hematol. 2016; 23(4):426-433. 2. Kuehl WM, Bergsagel PL. Molecular pathogenesis of multiple myeloma and its premalignant precursor. J Clin Invest. 2012; 122(10):3456-3463. 3. Bianchi G, Munshi NC. Pathogenesis beyond the cancer clone(s) in multiple myeloma. Blood. 2015;125(20):3049-3058. 4. Gupta SC, Sung B, Prasad S, Webb LJ, Aggarwal BB. Cancer drug discovery by repurposing: teaching new tricks to old

haematologica | 2018; 103(7)

6.

7. 8.

9.

they are performed in MM cell lines. While we primarily focused our analysis on stromal-responsive lines, which appear to better mimic the malignant plasma cell phenotype found in MM patients, focused clinical trials in MM will be necessary at this point to truly evaluate whether tofacitinib has anti-MM effects. Toward this goal, the value of drug repurposing becomes readily apparent. Tofacitinib can be quickly moved into Phase I/II studies in MM as the tolerated doses and adverse event profiles of this drug are well-characterized.11 Intriguingly, a patient population with both MM and RA could readily serve as the basis of a multi-center trial. Alternatively, a patient population with early-stage disease, perhaps smoldering myeloma, may be the optimal setting for clinical use, when plasma cells may be most dependent on microenvironmental cues. Our findings of strong synergy between tofacitinib and venetoclax in the context of the BM microenvironment may be relevant given the exciting progress of venetoclax in the clinic.37 This finding may also reflect a general vulnerability specific to the BM niche, given the effectiveness of this combination both in MM and AML.30 In conclusion, tofacitinib is a promising agent to reverse the tumor-proliferative effects of the BM microenvironment that can be rapidly repurposed to benefit MM patients. Funding This work was supported by the UCSF Stephen and Nancy Grand Multiple Myeloma Translational Initiative and the Myeloma Research Fund of the Silicon Valley Community Foundation (to BTA and APW), an NCI Cancer Center Support Grant (P30 CA082103) (to BCH), and an NCI Clinical Scientist Development Award (K08 CA184116), a Dale F. Frey Breakthrough Award from the Damon Runyon Cancer Research Foundation (DFS 14-15), and an American Cancer Society Individual Research Award (IRG-97-150-13) (to APW). Acknowledgments We thank Drs. Jeffrey Wolf, Tom Martin, Nina Shah, and Cammie Edwards for discussions, advice, and insight. We thank the staff of the UCSF Helen Diller Family Cancer Center Preclinical Therapeutic Core facility for completion of murine studies. We thank Dr. Diego Acosta-Alvear for providing luciferase-labeled MM cell lines.

dogs. Trends Pharmacol Sci. 2013; 34(9):508-517. Corsello SM, Bittker JA, Liu Z, et al. The Drug Repurposing Hub: a next-generation drug library and information resource. Nat Med. 2017;23(4):405-408. Shim JS, Liu JO. Recent advances in drug repositioning for the discovery of new anticancer drugs. Int J Biol Sci. 2014;10(7):654663. Nosengo N. Can you teach old drugs new tricks? Nature. 2016;534(7607):314-316. Murnane M, Dhimolea E, Li R, et al. Defining primary marrow microenvironment-induced synthetic lethality and resistance for 2,684 approved drugs across molecularly distinct forms of multiple myeloma. Blood. 2015;126:503. Meyer DM, Jesson MI, Li X, et al. Anti-

10. 11.

12. 13.

inflammatory activity and neutrophil reductions mediated by the JAK1/JAK3 inhibitor, CP-690,550, in rat adjuvantinduced arthritis. J Inflamm. 2010;7:41. Shuai K, Liu B. Regulation of JAK-STAT signalling in the immune system. Nat Rev Immunol. 2003;3(11):900-911. Fleischmann R, Kremer J, Cush J, et al. Placebo-controlled trial of tofacitinib monotherapy in rheumatoid arthritis. N Engl J Med. 2012;367(6):495-507. Klein B, Zhang XG, Lu ZY, Bataille R. Interleukin-6 in human multiple myeloma. Blood. 1995;85(4):863-872. Das R, Strowig T, Verma R, et al. Microenvironment-dependent growth of preneoplastic and malignant plasma cells in humanized mice. Nat Med. 2016; 22(11):1351-1357.

1227


C. Lam et al. 14. Monaghan KA, Khong T, Burns CJ, Spencer A. The novel JAK inhibitor CYT387 suppresses multiple signalling pathways, prevents proliferation and induces apoptosis in phenotypically diverse myeloma cells. Leukemia. 2011;25(12):1891-1899. 15. Li J, Favata M, Kelley JA, et al. INCB16562, a JAK1/2 selective inhibitor, is efficacious against multiple myeloma cells and reverses the protective effects of cytokine and stromal cell support. Neoplasia. 2010; 12(1):28-38. 16. De Vos J, Jourdan M, Tarte K, Jasmin C, Klein B. JAK2 tyrosine kinase inhibitor tyrphostin AG490 downregulates the mitogen-activated protein kinase (MAPK) and signal transducer and activator of transcription (STAT) pathways and induces apoptosis in myeloma cells. Br J Hematol. 2000;109(4):823-828. 17. Ramakrishnan V, Kimlinger T, Haug J, et al. TG101209, a novel JAK2 inhibitor, has significant in vitro activity in multiple myeloma and displays preferential cytotoxicity for CD45+ myeloma cells. Am J Hematol. 2010;85(9):675-686. 18. Scuto A, Krejci P, Popplewell L, et al. The novel JAK inhibitor AZD1480 blocks STAT3 and FGFR3 signaling, resulting in suppression of human myeloma cell growth and survival. Leukemia. 2011; 25(3):538-550. 19. Burger R, Le Gouill S, Tai YT, et al. Janus kinase inhibitor INCB20 has antiproliferative and apoptotic effects on human myeloma cells in vitro and in vivo. Mol Cancer Therap. 2009;8(1):26-35. 20. Smith EJ, Olson K, Haber LJ, et al. A novel, native-format bispecific antibody triggering T-cell killing of B-cells is robustly active in mouse tumor models and cynomolgus monkeys. Sci Rep. 2015;5:17943. 21. McMillin DW, Delmore J, Weisberg E, et al. Tumor cell-specific bioluminescence platform to identify stroma-induced changes to anticancer drug activity. Nat Med. 2010; 16(4):483-489. 22. Wiita AP, Ziv E, Wiita PJ, et al. Global cellular response to chemotherapy-induced apoptosis. eLife. 2013;2:e01236. 23. Anders S, Huber W. Differential expression analysis for sequence count data. Genome Biol. 2010;11(10):R106.

1228

24. Kuleshov MV, Jones MR, Rouillard AD, et al. Enrichr: a comprehensive gene set enrichment analysis web server 2016 update. Nucl Acids Res. 2016;44(W1):W9097. 25. Fila J, Honys D. Enrichment techniques employed in phosphoproteomics. Amino acids. 2012;43(3):1025-1047. 26. Lachmann A, Xu H, Krishnan J, Berger SI, Mazloom AR, Ma'ayan A. ChEA: transcription factor regulation inferred from integrating genome-wide ChIP-X experiments. Bioinformatics. 2010;26(19):2438-2444. 27. Mi H, Huang X, Muruganujan A, et al. PANTHER version 11: expanded annotation data from Gene Ontology and Reactome pathways, and data analysis tool enhancements. Nucl Acids Res. 2017; 45(D1):D183-D189. 28. Nelson EA, Walker SR, Frank DA. Jak/STAT signaling in the pathogenesis and treatment of multiple myeloma. In: Anderson KC, ed. Advances in biology and therapy of multiple myeloma: Volume 1: Basic Science. New York: Springer, 2013: 117-138. 29. Brocke-Heidrich K, Kretzschmar AK, Pfeifer G, et al. Interleukin-6-dependent gene expression profiles in multiple myeloma INA-6 cells reveal a Bcl-2 family-independent survival pathway closely associated with Stat3 activation. Blood. 2004; 103(1):242-251. 30. Karjalainen R, Pemovska T, Popa M, et al. JAK1/2 and BCL2 inhibitors synergize to counteract bone marrow stromal cellinduced protection of AML. Blood. 2017; 130(6):789-802. 31. Smith GA, Uchida K, Weiss A, Taunton J. Essential biphasic role for JAK3 catalytic activity in IL-2 receptor signaling. Nat Chem Biol. 2016;12(5):373-379. 32. Quintas-Cardama A, Vaddi K, Liu P, et al. Preclinical characterization of the selective JAK1/2 inhibitor INCB018424: therapeutic implications for the treatment of myeloproliferative neoplasms. Blood. 2010; 115(15):3109-3117. 33. Cox J, Mann M. MaxQuant enables high peptide identification rates, individualized p.p.b.-range mass accuracies and proteome-wide protein quantification. Nat Biotechnol. 2008;26(12):1367-1372.

34. Leitner A. Enrichment strategies in phosphoproteomics. Meth Mol Biol. 2016; 1355:105-121. 35. Lachmann A, Ma'ayan A. KEA: kinase enrichment analysis. Bioinformatics. 2009; 25(5):684-686. 36. Yadav B, Wennerberg K, Aittokallio T, Tang J. Searching for drug synergy in complex dose-response landscapes using an interaction potency model. Comp Struct Biotechnol J. 2015;13:504-513. 37. Moreau P, Chanan-Khan A, Roberts AW, et al. Promising efficacy and acceptable safety of venetoclax plus bortezomib and dexamethasone in relapsed/refractory MM. Blood. 2017;130(22):2392-2400. 38. Yokoyama S, Perera PY, Terawaki S, et al. Janus kinase inhibitor tofacitinib shows potent efficacy in a mouse model of autoimmune lymphoproliferative syndrome (ALPS). J Clin Immunol. 2015; 35(7):661-667. 39. Zlei M, Egert S, Wider D, Ihorst G, Wasch R, Engelhardt M. Characterization of in vitro growth of multiple myeloma cells. Exp Hemaol. 2007;35(10):1550-1561. 40. Shain KH, Yarde DN, Meads MB, et al. Beta1 integrin adhesion enhances IL-6mediated STAT3 signaling in myeloma cells: implications for microenvironment influence on tumor survival and proliferation. Cancer Res. 2009;69(3):1009-1015. 41. Burger R, Gunther A, Klausz K, et al. Due to interleukin-6 type cytokine redundancy only glycoprotein 130 receptor blockade efficiently inhibits myeloma growth. Haematologica. 2017;102(2):381-390. 42. Burger R, Guenther A, Bakker F, et al. Gp130 and ras mediated signaling in human plasma cell line INA-6: a cytokineregulated tumor model for plasmacytoma. Hematol J. 2001;2(1):42-53. 43. Tenhumberg S, Waetzig GH, Chalaris A, et al. Structure-guided optimization of the interleukin-6 trans-signaling antagonist sgp130. J Biol Chem. 2008;283(40):2720027207. 44. de Oliveira MB, Fook-Alves VL, Eugenio AIP, et al. Anti-myeloma effects of ruxolitinib combined with bortezomib and lenalidomide: A rationale for JAK/STAT pathway inhibition in myeloma patients. Cancer Lett. 2017;403:206-215.

haematologica | 2018; 103(7)


ARTICLE

Plasma Cell Disorders

Plasma cell proliferative index predicts outcome in immunoglobulin light chain amyloidosis treated with stem cell transplantation

M. Hasib Sidiqi,1 Mohammed A. Aljama,1 Dragan Jevremovic,2 William G. Morice,2 Michael Timm,2 Francis K. Buadi,1 Rahma Warsame,1 Martha Q. Lacy,1 Angela Dispenzieri,1 David Dingli,1 Wilson I. Gonsalves,1 Shaji Kumar,1 Prashant Kapoor,1 Taxiarchis Kourelis,1 Nelson Leung,3 William J. Hogan1 and Morie Gertz1

1 Division of Hematology, Department of Internal Medicine; 2Department of Laboratory Medicine and Pathology and 3Division of Nephrology and Hypertension, Mayo Clinic, Rochester, MN, USA

Ferrata Storti Foundation

Haematologica 2018 Volume 103(7):1229-1234

ABSTRACT

T

he plasma cell proliferative index provides an insight into plasma cell biology in plasma cell disorders and is an important prognostic marker in myeloma and smoldering myeloma. We analyzed the prognostic impact of the plasma cell proliferative index in 513 patients with systemic immunoglobulin light chain (AL) amyloidosis undergoing stem cell transplantation at the Mayo Clinic between 1st January 2003 and 31st August 2016. Two cohorts were identified according to Low or Elevated plasma cell proliferative index. Patients with an Elevated plasma cell proliferative index had more cardiac involvement (56% vs. 44%; P=0.01), less renal involvement (55% vs. 70%; P=0.001), and were more likely to have 10% or over bone marrow plasma cells (58% vs. 32%; P<0.0001) compared to those with a Low plasma cell proliferative index. Both progression-free survival and overall survival were lower in patients with an Elevated compared to Low plasma cell proliferative index: median progression-free survival 44 vs. 95 months (P<0.0001) and median overall survival 102 vs. 143 months (P=0.0003). All-cause mortality at 100 days was higher in patients with an Elevated plasma cell proliferative index (elevated 10.3% vs. low 4.3%; P=0.008). On multivariate analysis Elevated plasma cell proliferative index was an independent prognostic factor for overall survival (Hazard Ratio 1.5, 95%CI: 1.1-2.1; P=0.021). The plasma cell proliferative index is an important prognostic tool in patients with AL amyloidosis undergoing stem cell transplant.

Introduction Immunoglobulin light chain (AL) amyloidosis is a multi-system disorder characterized by a plasma cell or B-cell clone producing misfolded light chain proteins that deposit in tissues causing tissue damage and organ dysfunction.1 Patients with AL amyloidosis typically have a low tumor burden with the majority of patients having bone marrow plasma cells (BMPC) of less than 10% at diagnosis.2 The plasma cell proliferative index (PCPI) provides an insight into plasma cell biology in plasma cell disorders. It recognizes cells that are actively synthesizing DNA and gives an indication of the proliferative rate of the malignant plasma cells. The bone marrow PCPI has been identified as a strong prognostic marker in both active multiple myeloma and smoldering myeloma.3-5 Data regarding its utility in AL amyloidosis are scant, with an early report suggesting patients with an elevated PCPI had a worse overall survival.6 However, this pre-dated the use of autologous stem cell transplant (ASCT) therapy for AL amyloidosis and the advent of novel therapies as therapeutic options for patients with AL amyloidosis. haematologica | 2018; 103(7)

Correspondence: gertz.morie@mayo.edu

Received: January 30, 2018. Accepted: April 16, 2018. Pre-published: April 19, 2018. doi:10.3324/haematol.2018.189985 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/103/71229 Š2018 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.

1229


M.H. Sidiqi et al.

We review the prognostic impact of the plasma cell proliferative index in a large cohort of patients with AL amyloidosis undergoing stem cell transplant at the Mayo Clinic, and assess patients’ and disease characteristics as well as the prognostic impact of PCPI on outcome and survival.

Methods After approval by the Mayo Clinic Institutional Review Board, data were reviewed on all patients with biopsy proven systemic AL amyloidosis who underwent autologous stem cell transplant between 1st January 2003 and 31st August 2016 and had a plasma

Table 1. Patients' baseline characteristics.

Characteristic Age, years, median (IQR) Male % Organs involved, n (%) Cardiac Renal Hepatic Neurological Other Organs involved, n (%) 1 2 ≥3 Light chain, n (%) Lambda Kappa Pre-ASCT light chain mg/dL, median (IQR) Bone marrow plasma cells, n (%) <10% ≥10% Creatinine, median, mg/dL Urine 24-hour protein, median, g NT-Pro BNP, median, pg/mL (IQR) NT-Pro BNP >8500, n (%) Troponin T, median, ng/mL (IQR) Mayo Stage 2004, n (%) I II III Missing Mayo Stage 2012, n (%) I II III IV Missing Conditioning, n (%) Melphalan 200 mg/m2 Melphalan <200 mg/m2 Pre-ASCT chemotherapy n (%) Untreated Corticosteroid only Melphalan based IMID based Bortezomib based Other Time Period of Transplant, n (%) < 2010 ≥ 2010

Low PCPI (n=348)

Elevated PCPI (n=165)

P

59 (53-65) 64

58 (52-63) 63

0.26 0.77

153 (44) 243 (70) 35 (10) 49 (14) 76 (22)

93 (56) 90 (55) 19 (11) 25 (15) 52 (31)

0.01 0.001 0.65 0.79 0.02 0.12

164 (47) 124 (36) 60 (17)

62 (37) 69 (42) 34 (21)

258 (74) 90 (26) 12.4 (5.8-34.4)

118 (72) 47 (28) 20.8 (8.7-56.3)

236 (68) 112 (32) 1 (0.9-1.3) 3.3 (0.2-7.0) 417 (150-1757) 13 (4%) 0.01 (0.01-0.02)

70 (42) 95 (58) 1 (0.9-1.2) 0.98 (0.2-6.1) 901 (197-2499) 3 (2%) 0.01 (0.01-0.03)

200 (68) 46 (16) 48 (16) 54

91 (61) 29 (20) 28 (19) 17

139 (48) 79 (27) 45 (16) 26 (9) 59

51 (35) 54 (37) 22 (15) 20 (13) 18

256 (74) 89 (26)

104 (63) 60 (37)

221 (64) 35 (10) 14 (4) 19 (5) 48 (14) 11 (3)

78 (47) 19 (12) 15 (9) 16 (10) 35 (21) 2 (1)

207 (59) 141 (41)

93 (56) 72 (44)

0.52

0.0004 <0.0001

0.88 0.049 0.02 0.28 0.13 0.38

0.03

0.02

0.0023 0.0006

0.5

ASCT,: autologous stem cell transplant; PCPI: plasma cell proliferation index; IMiD: immunomodulatory drug; IQR: interquartile range; n: number.

1230

haematologica | 2018; 103(7)


Labeling index in light chain amyloidosis

Figure 1. Overall survival and progression-free survival. PCPI: plasma cell proliferation index; ASCT: autologous stem cell transplant; PFS: progression-free survival; OS: overall survival.

Table 2. Hematologic response.

Low PCPI n (%) CR VGPR PR NR NA

150 (44) 101 (29) 48 (14) 44 (13) 5

Elevated PCPI n (%) 63 (40) 49 (31) 25 (16) 21 (13) 7

P 0.86 0.44

CR: complete response; VGPR: very good partial response; PR: partial response; NR: no response; NA: not available; PCPI: plasma cell proliferation index.

cell proliferative index (PCPI) performed on bone marrow samples at diagnosis. The PCPI is expressed as a percentage of plasma cells in S phase. Patients with insufficient plasma cells identified for an accurate assessment of the PCPI were excluded. Prior to May of 2012, the plasma cell proliferative index was measured using the bromodeoxyuridine (BrdU) method described previously.4 From May 2012 onwards, the BrdU method was replaced by the DNA content measurement using a flow cytometry technique and a summary of this method is available in the Online Supplementary Appendix. We initially analyzed all patients with a PCPI performed using the BrdU method (n=390). The median PCPI was 0% with at least two-thirds (69%) of patients having a PCPI of 0%. Amongst the patients with a PCPI performed using the flow cytometric method (n=123), the median PCPI was 0.3% with approximately two-thirds (65%) of patients having a PCPI of 0.5% or under. To account for differences in the sensitivity of the methods, we established “Low” PCPI as 0% by the BrdU method and 0.5% or under by the flow cytometry method. “Elevated” PCPI was defined as over 0% by the BrdU method and over 0.5% by the flow cytometric method. This established two cohorts of patients: those with a Low PCPI (n=348) and those with an Elevated PCPI (n=165). Patients’, disease and treatment characteristics as well as outcomes were assessed between the two groups. haematologica | 2018; 103(7)

Organ involvement and hematologic response were assessed according to consensus criteria.7,8 Risk stratification was according to the Mayo 2012 and 2004 staging systems.9,10 Patients were selected for ASCT using available criteria at the time of transplant. Patients were mobilized, conditioned and transplanted according to previously published institutional protocols. Response was measured at approximately 100 days post ASCT according to up-dated consensus criteria. Overall response rate (ORR) was defined as a partial response (PR) or greater. Statistical analysis was performed on JMP software (SAS, Cary, NC, USA). Patients’ and disease related factors were compared using the χ2 test for categorical variables and the Wilcoxon signed rank test for continuous variables. Survival analysis was performed using the Kaplan-Meier method. Overall survival (OS) was calculated from day 0 of bone marrow transplant to death from any cause. All-cause mortality at 100 days was defined as death from any cause within 100 days post ASCT. The Cox proportional hazards model was used to assess for predictors of OS. The variables included in the univariate analysis were age, sex, number of organs involved, bone marrow plasma cells, Mayo Stage, conditioning dose, pre-transplantation chemotherapy, and PCPI. Variables reaching a P<0.1 were included in the multivariate analysis.

1231


M.H. Sidiqi et al.

Figure 2. Landmark survival analysis from 100 days post stem cell transplant by hematologic response. CR: complete response; VGPR: very good partial response; PR: partial response; NR: no response; PCPI: plasma cell proliferation index; OS: overall survival; N: number.

Table 3. Prognostic factors for survival.

Variable Age â&#x2030;Ľ65 Male BMPCs â&#x2030;Ľ10 PCPI elevated Mayo stage 2012 III/IV vs. I/II Mayo Stage 2004 III vs. I/II Conditioning dose 200 vs. <200

Univariate P

HR

0.0074 0.014 0.0033 0.0005 <0.0001 <0.0001 <0.0001

1.6 (1.1-2.1) 1.4 (1.1-2.0) 1.5 (1.2-2.1) 1.7 (1.3-2.3) 3.5 (2.5-4.8) 0.3 (0.2-0.4) 0.3 (0.3-0.5)

Multivariate Model A P HR NS NS NS 0.021 <0.0001 Not included <0.0001

1.5 (1.1-2.1) 2.3 (1.6-3.3) 0.4 (0.3-0.6)

Multivariate Model B P HR NS NS NS 0.019 Not included 0.0006 <0.0001

1.5 (1.1-2.1) 2.0 (1.3-2.8) 0.4 (0.3-0.6)

BMPCs: bone marrow plasma cells; PCPI: plasma cell proliferation index; NS: not significant; HR: Hazard Ratio.

Results Between 1st January 2003 and 31st of August 2016 548 patients with systemic AL amyloidosis underwent ASCT at the Mayo Clinic in Rochester, for 94% of whom (n=513) data were available on bone marrow PCPI performed at diagnosis and 6% (n=35) had insufficient plasma cells on bone marrow biopsy for accurate assessment of PCPI. Of those with a reported PCPI, 68% (n=348) had a Low PCPI and 32% (n=165) had an Elevated PCPI. Table 1 summarizes the baseline characteristics for each group of patients. Median age and proportion of males in the Low PCPI group was similar to those with an Elevated PCPI: 59 vs. 58 years (P=0.26) and 64% vs. 63% (P=0.77), respectively. Patients with an Elevated PCPI had more cardiac involvement (56% vs. 44%; P=0.01) and less renal 1232

involvement (55% vs. 70%; P=0.001). The number of organs involved was similar between the two groups. Bone marrow plasma cell burden was higher in the Elevated PCPI group with 58% of patients having BMPCs of 10% or over compared to 32% in the Low PCPI group (P<0.0001). A higher plasma cell percentage would be expected in the former group given the higher proliferative capacity of the plasma cell clone. More patients had a Mayo Stage of II or more in the Elevated PCPI group (65% elevated PCPI vs. 52% low PCPI; P=0.008). Fewer patients in the Elevated PCPI group received full intensity (melphalan 200 mg/m2) conditioning (63% Elevated PCPI vs. 74% Low PCPI; P=0.02). Reduced intensity conditioning (melphalan <200 mg/m2) was given to patients aged 70 years or older and those with creatinine over 1.8 mg/dL. More patients received pre-transplant chemotherapy in the Elevated PCPI group haematologica | 2018; 103(7)


Labeling index in light chain amyloidosis

Figure 3. Mayo Stage 2012. Survival by Mayo Stage 2012. PCPI: plasma cell proliferation index; OS: overall survival; N: number.

compared to the Low PCPI cohort (53% vs. 36%; P=0.0006). This may reflect our policy of giving pretransplantation chemotherapy induction in patients with more than 10% plasma cells in their bone marrow at diagnosis. Data regarding response were available in 501 of 513 (98%) of patients and are summarized in Table 2. Overall response rate was similar in both groups (87% in Elevated PCPI and 87% in Low PCPI; P=0.86) as were rates of complete response (40% in Elevated PCPI and 44% in Low PCPI; P=0.44). In the Low PCPI group, the ORR was not affected by pre-transplantation chemotherapy: 86% in untreated patients and 90% in treated patients (P=0.4). For patients with an Elevated PCPI, the ORR was significantly higher in patients receiving pretransplantation chemotherapy than those who did not (92% vs. 74%, respectively; P=0.003). Median PFS and OS for the entire cohort were 84 and 126 months, respectively. Both PFS and OS were lower in patients with an Elevated PCPI compared to Low PCPI (median PFS 44 vs. 95 months, P<0.0001, and median OS 102 vs. 143 months, P=0.0003, respectively) (Figure 1). Hematologic response (Figure 2) and Mayo Stage (Figure 3) predicted survival. In the Elevated PCPI cohort, those with Mayo Stage IV appeared to do better than those with Mayo Stage III; however, we believe this is likely to be due to the small numbers in the two groups (n=20 and 22, respectively). Chemotherapy prior to transplantation did not impact OS in either cohort (Low PCPI Pre-ASCT chemotherapy median OS 143 months vs. Untreated median OS 142 months, P=0.65, and PCL over 0 PreASCT chemotherapy median OS 102 months vs. Untreated median OS 114 months, P=0.86). All-cause mortality at 100 days was significantly higher in patients haematologica | 2018; 103(7)

Figure 4. Day 100 all-cause mortality. PCPI: plasma cell proliferation index; ASCT: autologous stem cell transplant.

with an Elevated PCPI compared to those with a Low PCPI (10.3% vs. 4.3%; P=0.008) (Figure 4). Table 3 shows the results for predictors of OS by univariate and multivariate analysis. On univariate analysis age, male sex, BMPCs 10% or over, Elevated PCPI, Mayo Stage, and conditioning dose were all prognostic factors for OS. On multivariate analysis the only independent prognostic factors for OS were Mayo Stage, conditioning dose (melphalan 200 vs. melphalan less than 200) and an Elevated PCPI (Hazard Ratio 1.5, 95%CI: 1.1-2.1; P=0.02). 1233


M.H. Sidiqi et al.

Discussion Increased plasma cell proliferation rate has been recognized as a predictor of a more aggressive clinical course in patients with myeloma and amyloidosis.3-6,11 The early studies of plasma cell proliferation rate were carried out with BrDU incorporation and microscopic review of plasma cells on the slide. However, flow cytometric evaluation of proliferation index is much more robust, precise and reproducible.11 In addition, since flow cytometry immunophenotying is already the standard of care for the diagnosis of plasma cell neoplasms, incorporation of proliferation measurement by DAPI staining is an easy addition to currently run assays, and therefore readily available for any clinical laboratory to perform. Since all patients with light chain amyloidosis require a bone marrow for assessment, ample material should be available for direct measurement of the proliferative index at time of diagnosis. The proliferative index is an independent predictor of survival and adds importantly to what is currently available based on the pre-transplantation variable of Mayo Stage as well as the post hoc variables of depth of response and conditioning dose of chemotherapy. We note that patients with an elevated PCPI had more cardiac involvement and a higher pre-transplant light chain level. Despite this, elevated PCPI remained an independent predictor of survival in a multivariate model that included the powerful prognostic factor of Mayo Stage, a variable that incorporates both cardiac biomarkers and light chain level in staging patients with AL amyloidosis. Pre-transplantation chemotherapy was associated with an improved overall response rate in patients with an elevated PCPI but not

References 1. Merlini G, Seldin DC, Gertz MA. Amyloidosis: pathogenesis and new therapeutic options. J Clin Oncol. 2011;29(14):1924-1933. 2. Kourelis TV, Kumar SK, Gertz MA, et al. Coexistent multiple myeloma or increased bone marrow plasma cells define equally high-risk populations in patients with immunoglobulin light chain amyloidosis. J Clin Oncol. 2013;31(34):4319-4324. 3. Madan S, Kyle RA, Greipp PR. Plasma cell labeling index in the evaluation of smoldering (asymptomatic) multiple myeloma. Mayo Clin Proc. 2010;85(3):300. 4. Kumar S, Rajkumar SV, Greipp PR, Witzig TE. Cell proliferation of myeloma plasma cells: comparison of the blood and marrow compartments. Am J Hematol. 2004; 77(1):7-11. 5. Greipp PR, Lust JA, O'Fallon WM, Katzmann JA, Witzig TE, Kyle RA. Plasma cell labeling index and beta 2-microglobu-

1234

6.

7.

8.

9.

those with a low PCPI. Therefore the plasma cell proliferative index may be used as a potential guide to identify patients that may benefit most from pre-transplantation chemotherapy, suggesting that patients with an Elevated labeling index should be strongly considered for induction chemotherapy prior to transplantation. In addition, an Elevated PCPI predicts earlier progression, suggesting that consolidation chemotherapy or maintenance post transplant should be considered for this patient population. Given that hematologic response is a strong predictor of survival, achieving deeper responses in this cohort may help improve outcomes. Our data also reveal higher 100-day mortality in patients with an Elevated PCPI. Patients with an Elevated PCPI were more likely to have cardiac involvement and a higher Mayo Stage; both have been associated with a worse prognosis.10,12 In addition, patients with an Elevated PCPI were more likely to receive chemotherapy, with potential toxicity, prior to transplantation. Our study is limited by its retrospective nature and potentially by a change in the method of testing for the plasma cell proliferation index during the study time period. We have tried to account for the difference in assay sensitivity by establishing unique cut offs for elevated PCPI for the two methods. In addition, our study includes patients over a time period of almost 14 years and selection criteria for stem cell transplant eligibility in AL amyloidosis have significantly changed over this time. Despite these limitations, the significance of the labeling index on survival independent of previously known factors is an important addition to the prognosis of patients with this serious disorder.

lin predict survival independent of thymidine kinase and C-reactive protein in multiple myeloma. Blood. 1993;81(12):33823387. Gertz MA, Kyle RA, Greipp PR. The plasma cell labeling index: a valuable tool in primary systemic amyloidosis. Blood. 1989;74(3):1108-1111. Gertz MA, Comenzo R, Falk RH, et al. Definition of organ involvement and treatment response in immunoglobulin light chain amyloidosis (AL): a consensus opinion from the 10th International Symposium on Amyloid and Amyloidosis, Tours, France, 18-22 April 2004. Am J Hematol. 2005;79(4):319-328. Palladini G, Dispenzieri A, Gertz MA, et al. New criteria for response to treatment in immunoglobulin light chain amyloidosis based on free light chain measurement and cardiac biomarkers: impact on survival outcomes. J Clin Oncol. 2012;30(36):45414549. Dispenzieri A, Gertz MA, Kyle RA, et al.

Serum cardiac troponins and N-terminal pro-brain natriuretic peptide: a staging system for primary systemic amyloidosis. J Clin Oncol. 2004;22(18):3751-3757. 10. Kumar S, Dispenzieri A, Lacy MQ, et al. Revised prognostic staging system for light chain amyloidosis incorporating cardiac biomarkers and serum free light chain measurements. J Clin Oncol. 2012; 30(9):989-995. 11. Paiva B, Vidriales MB, Montalban MA, et al. Multiparameter flow cytometry evaluation of plasma cell DNA content and proliferation in 595 transplant-eligible patients with myeloma included in the Spanish GEM2000 and GEM2005<65y trials. Am J Pathol. 2012;181(5):1870-1878. 12. Dispenzieri A, Gertz MA, Kyle RA, et al. Prognostication of survival using cardiac troponins and N-terminal pro-brain natriuretic peptide in patients with primary systemic amyloidosis undergoing peripheral blood stem cell transplantation. Blood. 2004;104(6):1881-1887.

haematologica | 2018; 103(7)


ARTICLE

Platelet Biology & its Disorders

Platelet Munc13-4 regulates hemostasis, thrombosis and airway inflammation

Ferrata Storti Foundation

Eduardo I. Cardenas,1,2 Keegan Breaux,1 Qi Da,3,4 Jose R. Flores,1 Marco A. Ramos,1 Michael J. Tuvim,1 Alan R. Burns,5 Rolando E. Rumbaut3,4 and Roberto Adachi1

Department of Pulmonary Medicine, The University of Texas MD Anderson Cancer Center, Houston, TX, USA; 2Tecnologico de Monterrey, Escuela de Ingenieria y Ciencias, Monterrey, Mexico; 3Center for Translational Research on Inflammatory Diseases (CTRID), Michael E. DeBakey Veterans Affairs Medical Center, Houston, TX, USA; 4 Department of Medicine, Baylor College of Medicine, Houston, TX, USA and 5College of Optometry, University of Houston, TX, USA 1

Haematologica 2018 Volume 103(7):1235-1244

ABSTRACT

P

latelet degranulation is crucial for hemostasis and may participate in inflammation. Exocytosis in platelets is mediated by SNARE proteins and should be controlled by Munc13 proteins. We found that platelets express Munc13-2 and -4. We assessed platelet granule exocytosis in Munc13-2 and -4 global and conditional knockout (KO) mice, and observed that deletion of Munc13-4 ablates dense granule release and indirectly impairs alpha granule exocytosis. We found no exocytic role for Munc13-2 in platelets, not even in the absence of Munc13-4. In vitro, Munc13-4-deficient platelets exhibited defective aggregation at low doses of collagen. In a flow chamber assay, we observed that Munc13-4 acted as a rate-limiting factor in the formation of thrombi. In vivo, we observed a dose-dependency between Munc13-4 expression in platelets and both venous bleeding time and time to arterial thrombosis. Finally, in a model of allergic airway inflammation, we found that platelet-specific Munc13-4 KO mice had a reduction in airway hyper-responsiveness and eosinophilic inflammation. Taken together, our results indicate that Munc13-4-dependent platelet dense granule release plays essential roles in hemostasis, thrombosis and allergic inflammation.

Introduction A key effector response from platelets is exocytosis of their alpha, dense and lysosomal granules. Alpha granules are the most abundant, and contain soluble molecules and receptors that propagate platelet activation and aggregation.1 Dense granules are secreted at a faster rate and store ADP, an autocrine agonist for platelet activation.2 Lysosomal granules contain membrane-associated proteins and acidhydrolases, and may contribute to thrombus remodeling.3 During exocytosis, the membrane of a platelet granule fuses with the plasma membrane. This requires the formation of a SNARE (soluble N-ethylmaleimidesensitive factor attachment protein receptor) complex by proteins localized on both membranes.4 Prior to fusion, granules are brought close to the plasma membrane by tethering and docking processes, but this proximity is not sufficient to drive fusion, it also requires priming.5 Fundamental to priming is the interaction of Munc (mammalian homolog of C. elegans uncoordinated gene) 13 with Munc18, which allows Syntaxin to interact with the other exocytic SNARE proteins.6 Among the four paralogs of Munc13, only Munc13-4 has been studied in mouse platelets.7 Different groups have agreed that deletion of Munc13-4 inhibits primarily dense granule release,8,9 which may affect alpha granule exocytosis9,10 or integrin αIIbβ3 activation11 depending on experimental variables.12 In addition, global deficiency of Munc13-4 affects hemostasis, probably due to defective platelet exocytosis,10,11,13 a difficult conclusion to reach because Munc13-4 is also expressed in other tissues important for hemostasis (e.g. endothelial cells).14 In humans, mutations in the gene encoding Munc13-4 cause familial hemophagocytic lymphohistiocytosis type 3 (FHL3), an autosomal recessive disorder characterized by defective secretion in cytotoxic T lymphocytes and natural killer cells, multisystemic inflamhaematologica | 2018; 103(7)

Correspondence: radachi@mdanderson.org

Received: November 30, 2017. Accepted: April 12, 2018. Pre-published: April 19, 2018. doi:10.3324/haematol.2017.185637 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/103/7/xxx ©2018 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.

1235


E.I. Cardenas et al.

mation, and organ infiltration by CD8+ T cells and macrophages.15 Platelets from FHL3 patients present defective degranulation,16 and bleeding diathesis has been reported in some patients with FHL3.17 There is mounting evidence that platelets participate in inflammation, including allergic airway disease.18 Patients with asthma have lower platelet counts19 and increased levels of markers of platelet activation20 after allergen exposure. Platelets have been found extravascularly in the airways,21 and platelet products have been measured in bronchoalveolar lavage (BAL) fluid of asthmatic patients.22 This suggests that platelets migrate to the lungs during asthma, which has been experimentally confirmed in mice.23 In animal models of asthma, platelet depletion has been shown to decrease the number of leukocytes infiltrating the airways24 and bronchoconstriction induced by allergen.25 Serotonin and/or ADP released from platelet dense granules may be responsible for these findings. In a mouse model of asthma, deletion of the enzyme that synthesizes serotonin in peripheral tissues caused a significant reduction in asthmatic symptoms.26 Multiple studies in mouse models of asthma have targeted the ADP receptor P2Y12 with mostly favorable outcomes,27,28 but there is evidence that the pro-inflammatory effects of platelets may be mediated by P2Y1 (another ADP receptor) and P2Y14 (a UDP-glucose receptor), and not P2Y12.29,30 Given that platelets participate in highly distinct physiological responses (hemostasis and airway inflammation), we investigated whether exocytosis of dense granules, so crucial for hemostasis, is also important for the development of asthma. We manipulated the expression of Munc13-2 and -4 in platelets in vivo. While absence of Munc13-2 had no significant effect, absence or reduced levels of Munc13-4 altered platelet dense granule secretion directly and alpha granule exocytosis indirectly, impairing platelet aggregation and thrombus formation. By using platelet-specific knockout (KO) mice, we proved that Munc13-4 from platelets, and not from other tissues, is required for venous and arterial hemostasis, and for arterial thrombosis in vivo. Finally, we observed that Munc13-4dependent platelet exocytosis is essential for the full development of allergic airway inflammation.

our Munc13-4 global KO line (Munc13-4−/−). This line was propagated by heterozygote (Munc13-4+/−) crossings to generate Munc13-4−/−, Munc13-4+/− and control Munc13-4+/+ littermates. We also generated megakaryocyte/platelet-specific KO mice (Munc13-4Δ/Δ) by crossing Munc13-4F/F mice with C57BL/6– Tg(Pf4–icre)Q3Rsko/J mice (Munc13-4+/+ Cre+; The Jackson Laboratory #008535), which selectively express Cre recombinase in megakaryocytes. Munc13-4+/+ Cre+ mice were also used as additional controls. All lines were on a C57BL/6J background, as confirmed by speed-congenics scanning for 105 SNPs. Mice of both sexes were used in all experiments. Mice were kept in a pathogen-free facility and handled in accordance with the Institutional Animal Care and Use Committees of the University of Texas MD Anderson Cancer Center and Baylor College of Medicine.

Bleeding time tests We used mice (20±2 weeks old) of the same weight (30±3 g) anesthetized with Avertin (tribromoethanol in tert-amyl alcohol) 0.4 mg/g intraperitoneally (i.p.). In the transection model, tails were cut 5 mm from the tip with a razor blade, and bleeding depended mainly on the tail artery. For the incision model, we created a device to make a reproducible transversal dorsal tail incision 0.8 mm in depth at a point where the tail has a diameter of 3.8 mm (Online Supplementary Figure S1), sectioning only the dorsal tail venous plexus. In both models, the tails were immediately immersed in 37°C saline and the time to cessation of bleeding was recorded. All animals were euthanized after bleeding stopped or at 20 minutes (min).

Asthma model Mice (9±1 weeks old) were sensitized i.p. on days 0 and 7 with 10 mg ovalbumin (OVA; grade V) adsorbed to 1 mg of aluminum potassium sulfate dodecahydrate (both from Sigma-Aldrich) in 100 mL of saline. They were challenged once a day on days 19-21 in a nebulization chamber with 1% OVA in PBS for 30 min using an Aerotech II jet nebulizer (Biodex) at 10 L/min. Then, they were studied on day 22. A detailed description of the airway mechanics assessment, airway mucin quantification, histology and BALs is provided in the Online Supplementary Methods.

Results Methods

Expression and targeting of Munc13 proteins in platelets

A detailed description of the blood collection and platelet isolation, expression studies, secretion and activation assays, electron microscopy and stereology, aggregometry and flow-chamber assays, in vivo thrombosis model, and statistical analysis is provided in the Online Supplementary Methods.

By qPCR, we found that C57BL/6J platelets express Munc13-1, -2 and -4 (Figure 1A). We had created Munc13-4 global and conditional KO lines,32 and we obtained Munc13-2 KO mice (Dr. Christian Rosenmund, Charité Universitaetsmedizin).31 We did not study Munc13-1 because its global deletion is perinatally lethal34 and a conditional KO line was not available. We confirmed lack of Munc13-2 and normal expression levels of Munc13-1 and -4 in Munc13-2−/− platelets (Figure 1B). Immunoblots of tissues and platelets from all Munc13-4 mutants confirmed the global reduction and absence of Munc13-4 expression in Munc13-4+/− and Munc13-4−/−, respectively, and the specific deletion of Munc13-4 in platelets in Munc13-4Δ/Δ mice (Figure 1C). The decreased expression in Munc13-4F/F mice to approximately 20% has been reported,32 and we used it to interrogate dose-response relationships between Munc13-4 expression (0, 20, 50, 100%) and the outcomes of our experiments.

Mice Munc13-2 KO and Munc13-4 global and conditional KO mice have been described previously.31,32 In short, we flanked exon 3 of the mouse Munc13-4 gene (Unc13d) with two loxP sites (“floxed” or F allele). Due to this genetic manipulation, mice homozygous for the F allele (Munc13-4F/F) had a reduced expression of Munc13-4 globally (i.e. Munc13-4F/F mice are hypomorphs). Exon 3 contains the start codon of Unc13d, its sequence is present in all described splice variants of mouse Munc13-4, and its removal by Cre-mediated recombination eliminates Munc13-4 expression.32 We crossed Munc13-4F/F mice with B6.C-Tg(CMV-cre)1Cgn/J mice (The Jackson Laboratory #006054), which express Cre recombinase ubiquitously, to delete Unc13d in the germ line33 and generate 1236

haematologica | 2018; 103(7)


Munc13-4 in hemostasis and airway inflammation

A

B

C Figure 1. Munc13 expression and deletion. (A) RT-qPCR of all Munc13 proteins relative to β-actin in C57BL/6J platelets. N=3. (B) RT-qPCR from Munc13-2−/− platelets. All values are relative to β-actin and Munc13 isoform levels in Munc13-2+/+ platelets (WT). N=3. Bar: mean; error bar: Standard Error of Mean. (C) Representative immunoblots of platelet and spleen lysates from all Munc13-4 mutant mice probed with anti-mouse Munc13-4 antibody. β-actin was used as loading control. Numbers below blots represent densitometry results relative to β-actin and Munc13-4+/+. ND: not detected.

Munc13-4 regulates platelet dense granule exocytosis To test how different Munc13 isoforms contribute to platelet granule secretion, we assessed exocytosis of each type of platelet granule. We measured the translocation of P-selectin (alpha granules) and LAMP-1 (lysosomal granules) to the plasma membrane, and the release of ATP (dense granules). To interrogate potential agonist-dependent differences, we activated platelets with collagen (Figure 2) and thrombin (Figure 3). Based on dose-response curves (Online Supplementary Figure S2), we chose a high and a low dose for each agonist. In agreement with previous reports, the concentration of collagen necessary to induce translocation of membrane-associated proteins (P-selectin and LAMP-1) was higher than that required to induce release of soluble mediators (ATP).35 There were no differences among all the wild-type (WT) controls from our Munc13-2, Munc13-4 and double KO (DKO) colonies, therefore we pooled them in a single control group (WT in Figures 2 and 3). We observed that dense granule exocytosis was completely abolished in platelets lacking Munc13-4 regardless of the agonist used (Figures 2A and B, and 3A and B). With collagen, we documented a clear correlation between the level of expression of Munc13-4 (nil in Δ/Δ, −/− and DKO, approx. 20% in F/F, approx. 50% in +/−, and 100% in WT) and the secretion of ATP (Figure 2B). Although we observed a significant decrease in P-selectin translocation only at the lower dose of thrombin (Figures 2C and D, and 3C and D), we found a complete thrombin dependent defect in lysosomal granule exocytosis in platelets lacking Munc13-4 (Figures 2E and F, and 3E and F). Munc13-2−/− platelets showed no defects in any of these tests. To assess if the consequences of removing Munc13-2 would manifest only in the absence of Munc13-4, we tested platelets from DKO mice and did not observe any additional exocytic defects (Figures 2 and 3). haematologica | 2018; 103(7)

Given the secondary effect of Munc13-4 on alpha granule release,9,10 we decided to further interrogate alpha granule exocytosis by measuring the release of platelet factor 4 (PF4). Interestingly, Munc13-4−/− platelets had a significant reduction in PF4 release when stimulated with thrombin (Figure 3G). To test if the defects observed in alpha and lysosomal granule exocytosis could be indirectly caused by an impaired release of ADP from dense granules, we added exogenous ADP to stimulated platelets and showed that it was capable of rescuing PF4 release (Figure 3G) and P-selectin translocation (Figure 3H), but not LAMP-1 translocation (Figure 3I). Once again, Munc13-2−/− platelets had no exocytic defect and DKO platelets failed to show any additional phenotype (Figure 3G); we therefore decided to drop these two lines from subsequent experiments.

Platelet granule biogenesis and platelet activation are Munc13-4-independent The exocytic defects observed in platelets lacking Munc13-4 could be due to defective platelet granule biogenesis. Therefore, we analyzed resting platelets by electron microscopy (EM) (Figure 4A). Stereological analysis of platelet profiles before stimulation revealed no difference in the volume densities (Vv) of alpha or dense granules among all genotypes (Figure 4B and C). EM also provided an additional method to assess platelet exocytosis. After stimulation, Vv of alpha and dense granules in Munc13-4+/+ platelets decreased due to loss of granules through exocytosis. However, Munc13-4-deficient (Munc13-4−/− and Munc13-4Δ/Δ) platelets lost no dense granules (Figure 4C) and only a fraction of alpha granules (Figure 4B). Munc13-4F/F platelets showed an intermediate phenotype. To rule out the possibility that lack of Munc13-4 might interfere with inside-out platelet activation, we measured integrin αIIbβ3 activation, and found no differences (Figure 4D). 1237


E.I. Cardenas et al.

Munc13-4-regulates platelet aggregation and thrombus stability

bination as an explanation for the findings in the following in vivo studies.

To determine how the observed exocytic defects affect platelet function, we first studied platelet aggregation. When using a low concentration of collagen, platelets not expressing Munc13-4 did not undergo shape change (as evidenced by the flat aggregometry recordings) and were unable to form aggregates (Figure 5A and B). While hypomorphic Munc13-4F/F platelets underwent some shape change, they were still unable to form aggregates. Platelets from all genotypes underwent shape change and formed aggregates at similar levels when using a high dose of collagen (Figure 5A and B). We then assessed platelet thrombus formation under shear stress in a collagen-coated flow chamber and documented a severe defect in Munc13-4-deficient platelets (Figure 5C). The severity of this defect increased at a higher shear stress (Figure 5D), which suggests that platelets lacking Munc13-4 are unable to form stable thrombi. At high shear, platelets with approximately 50% expression of Munc13-4 (Munc13-4+/−) behaved normally, but those with approximately 20% expression (Munc13-4F/F) showed an intermediate defect. In all our in vitro studies (Figures 2-5) platelets from Munc13-4Δ/Δ mice behaved almost identically to those from Munc13-4−/− mice, eliminating inefficient Cre recom-

A

B

C

D

E

1238

F

Lack of Munc13-4 disrupts hemostasis and thrombosis We used two tail-bleeding tests to study hemostasis. In the classical transection model, Munc13-4 deletant mice were unable to stop bleeding and were euthanized after 20 min (Figure 6A). Furthermore, the hypomorphic Munc13-4F/F mice displayed an identical bleeding diathesis. Because we did not observe a defect in thrombus formation ex vivo at low shear stress, we hypothesized that the Munc13-4F/F mice would stop bleeding if we avoided sectioning the tail ventral artery and induced only venous bleeding. With this in mind, we developed a device to make a reproducible incision on the tail dorsal venous plexus only (Online Supplementary Figure S1). Munc13-4−/− and Munc13-4Δ/Δ mice still showed prolonged bleeding, but Munc13-4F/F mice behaved similarly to their WT counterparts (Figure 6B). This phenotype was not due to an abnormal number of platelets in the mutant mice (Table 1). Finally, we assessed thrombosis in vivo with the FeCl3 model of carotid thrombosis. In accordance with our in vitro results, we did not observe vessel occlusion in Munc13-4Δ/Δ mice, while WT arteries occluded in less than 5 min, and hypomorphic Munc13-4F/F mice presented an

Figure 2. Deletion of Munc13-4 impairs dense granule release in collagen-stimulated platelets. Samples from wild-type mice for Munc13-2 and -4 (WT), Munc13-4 +/−, −/−, F/F, Δ/Δ, Munc13-2 −/−, and Munc13-2 and 4 double KO (DKO) mice were stimulated with collagen. Representative tracings (A) and mean release of ATP (dense granules) (B) measured by luminometry in whole blood. N=5-14. Representative tracings (C and E) and mean fluorescence intensity over baseline (ΔMFI) (D and F) of P-selectin (alpha granules) (C and D) and LAMP-1 (lysosomal granules) (E and F) translocated to the surface of washed platelets measured by flow cytometry. N=8-16. Color legend in (A) applies to all panels. Bar: mean; error bar: Standard Error of Mean. #P<0.05; †P<0.01; *P<0.001; comparisons are with WT unless otherwise specified.

haematologica | 2018; 103(7)


Munc13-4 in hemostasis and airway inflammation

intermediate phenotype (Figure 6C). The proportion of vessels occluded at 30 min was 100% for Munc13-4+/+ mice, 50% for Munc13-4F/F mice and 0% for Munc13-4Δ/Δ mice.

baseline values were used to normalize the dose-response curves (Figure 7A). Given that the results in mice expressing Cre recombinase in their platelets (Munc13-4+/+ Cre+ mice) were indistinguishable from those of Munc13-4+/+ mice, we concluded that this effect was not a Cre-induced artifact. We found that, although there was no difference between Munc13-4+/+ and Munc13-4F/F mice, Munc13-4Δ/Δ mice completely mimicked the blunted AHR observed in Munc13-4−/− mice (Figure 7A and B). Analysis of methacholine PC1000 (data not shown) confirmed the same differences. The protection observed in Munc13-4−/− and Munc13-4Δ/Δ mice was not related to changes in airway mucous metaplasia, as we observed no significant differences of intracellular epithelial mucin content in periodic acid-fluorescent Schiff (PAFS)-stained sections (Figure 7C

Platelets contribute to allergic airway inflammation in a Munc13-4-dependent manner We observed that Munc13-4−/− mice were partially protected from developing AHR in a model of asthma. While looking for the cell responsible, we found that megakaryocyte/platelet-specific Munc13-4 KO mice presented a similar phenotype. We then studied our Munc13-4 mutant mice under an OVA-dependent acute model of allergic asthma. As expected, all OVA-exposed animals had an elevated baseline Rrs (total respiratory system resistance) compared to the naïve controls (data not shown), and these

A

B

C

D

E

F

G

haematologica | 2018; 103(7)

H

I

Figure 3. Deletion of Munc13-4 impairs dense and lysosomal granule release directly and alpha granule release indirectly in thrombin-stimulated platelets. Samples from wild-type mice for Munc-13-2 and -4 (WT), Munc13-4 +/−, −/−, F/F, Δ/Δ, Munc13-2 −/−, and Munc13-2 and -4 double knockout (DKO) mice were with thrombin (Thr) or thrombin and ADP (10 mM). Representative tracings (A) and mean release of ATP (dense granules) (B) measured by luminometry in whole blood. N=5-9. Representative tracings (C and E) and mean fluorescence intensity over baseline (ΔMFI) (D, F, H and I) of P-selectin (alpha granules) (C, D and H) and LAMP-1 (lysosomal granules) (E, F and I) translocated to the surface of washed platelets measured by flow cytometry. N=8-11 (C-F), N=3 (H and I). (G) Mean release of PF4 (alpha granules) measured by ELISA. N=3-9. Color legend in (A) applies to all panels. Bar: mean; error bar: Standard Error of Mean. # P<0.05; †P<0.01; *P<0.001; comparisons are with WT unless otherwise specified.

1239


E.I. Cardenas et al. A

B

C

Figure 4. Deletion of Munc13-4 impairs dense granule release. Washed platelets from Munc13-4 mutant mice were activated with thrombin 0.1 U/mL. (A) Representative electron microscopy (EM) cell profiles. Red arrows: dense granules. Scale bar: 1 mm. Mean volume density (Vv) of alpha granules (AG) (B) and dense granules (DG) (C) obtained by stereology. (D) Mean fluorescence intensity over baseline (ΔMFI) of Jon/A antibody binding to activated integrin αIIbβ3 on stimulated washed platelets. N=5. Color legend in (B) applies to all panels. Bar: mean; error bar: Standard Error of # † P<0.05; P<0.01; Mean. *P<0.001; comparisons are with +/+ unless otherwise Munc13-4 specified.

D

Table 1. Blood cell counts from Munc13-4 mutant mice. Red blood cells (x1012/L) White blood cells (x109/L) Platelets (x109/L)

Munc13-4+/+

Munc13-4−/−

Munc13-4F/F

Munc13-4Δ/Δ

8.3 ± 0.3 2.2 ± 0.3 613 ± 28

9.1 ± 0.5 2.8 ± 0.6 611 ± 42

8.5 ± 0.6 3.1 ± 0.4 668 ± 47

8.0 ± 0.3 2.6 ± 0.3 645 ± 50

Mean ± Standard Error of Mean. N=7-12.

and D). However, there was a significant decrease in eosinophilic inflammation, which was also independent of Cre expression. Qualitatively, Munc13-4Δ/Δ and Munc13-4−/− mice had reduced histological evidence of tissue eosinophilia (Figure 7E), and quantitatively, a decreased number of eosinophils in BAL fluid (Figure 7F).

Discussion Platelet exocytosis Exocytosis occurs constitutively in all eukaryotic cells, but some cell types possess a parallel regulated process to release pre-made mediators upon stimulation in a spatiotemporally-controlled manner. All known forms of regulated exocytosis require a Munc13 protein.31,34 Platelets and mast cells rely on regulated exocytosis as their main effector mechanism, and both express Munc13-2 and -4. Like in mast cells, we found that Munc13-4 plays an important role in granule release, but in contrast to mast cells we documented no role for Munc13-2 in platelet exocytosis (Figures 2 and 3). An alternative explanation is that a minor role for Munc13-2 could not be resolved by our methods, which cannot yield the high resolution attained by single-cell membrane capacitance recordings in mast cells.32 1240

Previous studies, as well as our current findings in secretion assays and morphometry (Figures 2-4), indicate that Munc13-4 regulates exocytosis of platelet dense granules directly.8,9 No ATP release could be detected after Munc13-4 was deleted, and platelets with decreased expression of Munc13-4 released proportionally less ATP than their WT counterparts regardless of the agonist used. By studying the relationship between Munc13-4 expression levels (0, 20, 50, 100%) and dense granule release (Figures 2 and 3), we have shown that Munc13-4 is an agonist-independent, rate-limiting factor in dense granule exocytosis. Interestingly, it has been proposed that, besides its known role in priming, Munc13-4 participates in dense granule tethering and conveys Ca2+ sensitivity to platelet exocytosis.7 We found no involvement of Munc13-4 in alpha or lysosomal granule exocytosis when we used collagen as agonist. With thrombin, we detected a partial exocytic defect in both types of granules, but only alpha granule release could be rescued by exogenous ADP. This confirms that deletion of Munc13-4 indirectly impairs alpha granule exocytosis,10 but that the alpha granule exocytic machinery is intact. There is evidence that the release of the three types of granules from platelets is differentially regulated,2 and it is possible that each requires different exocytic components. haematologica | 2018; 103(7)


Munc13-4 in hemostasis and airway inflammation

B

A

Figure 5. Deletion of Munc13-4 impedes platelet aggregation and interferes with thrombus formation in vitro. Samples from Munc13-4 mutant mice were stimulated with collagen. Representative tracings of platelet aggregation (A) and average maximum aggregation measured by light transmittance (B) on plateletrich plasma. N=3-6. Whole blood was fluorescently-labeled and perfused over collagen-coated plates at low (C) or high (D) sheer stress. Thrombus buildup was monitored by fluorescence intensity and compared at 200 seconds (s). N=5-10. Color legend in (A) applies to all panels. Bar or circle: mean; error bar: Standard Error of Mean. #P<0.05; â&#x20AC; P<0.01; *P<0.001; comparisons are with Munc13-4+/+ unless otherwise specified.

D

C

A

B

C

Figure 6. Deletion of Munc13-4 disrupts hemostasis and thrombus formation in vivo. Arterial (A) or venous (B) bleeding was measured in tails from Munc13-4 mutant mice. N=8-10. (C) Time to carotid flow occlusion after applying FeCl3 for 3 minutes (min). N=6-9. Circle: individual mouse; horizontal line or bar: mean; error bar: Standard Error of Mean. â&#x20AC; P<0.01; *P<0.001; comparisons are with Munc13-4+/+ unless otherwise specified.

While in mast cells the residual exocytosis observed after removing Munc13-4 was mediated by Munc13-2,32 the priming component that mediates alpha granule exocytosis in platelets remains unknown. Though generally thought of as a neuronal protein, Munc13-1 has been shown to regulate exocytosis in cytotoxic T lymphocytes,36 and it could be involved in platelet alpha granule release. In contrast to Munc13-4, Munc13-1 contains a regulatory diacylglycerol (DAG)-binding C1 domain.37 Collagen stimulation of platelets causes more sustained DAG formation than thrombin.38 It is possible that Munc13-1 plays a compensatory role in lysosomal granule haematologica | 2018; 103(7)

release, which would explain why we only observed a thrombin-dependent defect in the absence of Munc13-4.

Hemostasis and thrombosis Munc13-4 global KO mice have a severe hemostatic defect, and the hypothesis is that this is due to impaired ADP release from platelets.11,13 This cannot be tested with a regular global KO mouse given that Munc13-4 is expressed in other tissues relevant to hemostasis (e.g. endothelium).14 Our finding that the bleeding diathesis is identical in global and conditional KO mice (Figure 6) supports the notion that this Munc13-4-dependent phenome1241


E.I. Cardenas et al.

non relies exclusively on platelets. ADP secreted by platelets, acting via P2Y1 and P2Y12 receptors on the platelet surface, is known to amplify the primary response provided by collagen or thrombin, and to participate in platelet aggregation, thromboxane A2 generation, and thrombus formation under shear stress.39 The release of platelet nucleotides has been found to be critical for in vitro aggregation at lower doses of agonist.9

A

We have shown that platelets unable to release nucleotides (Munc13-4-deficient) fail to aggregate at a low dose of collagen, and that this defect causes the formation of unstable thrombi (Figure 5), which results in disruption of hemostasis and thrombosis (Figure 6). It has been proposed that platelets forming a thrombus can be grouped into two populations: a “core” of contact-dependent highly activated platelets and a “shell” of less activated ADP-

B

C

D

E

F

1242

Figure 7. Platelet-specific deletion of Munc13-4 reduces excessive airway constriction to non-specific stimuli (AHR) and lung eosinophilia in a model of asthma. Shown are results from Munc13-4 mutant mice one day after the last airway challenge of an acute allergic airway inflammation model. Naïve: sensitized but not challenged mice. (A) Total respiratory system resistance (Rrs) at increasing doses of nebulized methacholine (Mch). Circle: mean; error bar: Standard Error of Mean. (B) Rrs at the highest dose of Mch. N: number inside boxes. (C) Representative airway sections stained with PAFS (mucin in red). Scale bar: 20 mm. (D) Mucin volume density at the left lobar bronchus. N=9. (E) Representative lung sections (hematoxylin & eosin) and identification of eosinophils in tissues (insets). Scale bars=20 μm. (F) Eosinophil counts in bronchoalveolar lavage (BAL) fluid. N=9. Color legend in (A) applies to all panels. White line: mean; box: 25th-75th percentile; whiskers: 5th-95th # † percentile. P<0.05; P<0.01; *P<0.001; comparisons are with sensitized and challenged Munc13-4+/+ unless otherwise specified.

haematologica | 2018; 103(7)


Munc13-4 in hemostasis and airway inflammation

dependent platelets.40 Therefore, partial defects in ADP release would be expected to translate into thrombi that weaken and break as shear stress increases, and this is exactly what we observed in a perfusion chamber in hypomorphic Munc13-4F/F platelets (Figure 5). Moreover, we obtained a strong hemostatic defect in Munc13-4F/F mice when we transected the tail artery, but no prolonged bleeding time when we sectioned only the tail dorsal vein (Figure 6). Taken together, these results show that by controlling platelet dense granule release, Munc13-4-dependent exocytosis is a limiting factor for platelet aggregation, thrombus stabilization and hemostasis.

Allergic airway inflammation Asthma is a chronic airway disease characterized by excessive airway constriction to non-specific stimuli (also known as AHR), airway eosinophilic inflammation, and epithelial mucous metaplasia.41 Although hyperproduction of mucins and secretion of mucus by airway epithelial cells have been shown to be a component of AHR,42 we found that platelet exocytosis mediated by Munc13-4 did not influence airway mucous metaplasia. Instead, the decrease in AHR in animals with platelets deficient in Munc13-4 correlated with a reduction in the number of eosinophils recruited to the airways (Figure 6). Platelets store in their alpha and dense granules many mediators and membrane-bound proteins linked to allergic airway inflammation. Platelet P-selectin mediates the recruitment of eosinophils and lymphocytes into the airways of mice subjected to a model of asthma,43 and recent evidence suggests a similar function in humans.44,45 Additionally, exocytosis of alpha granule contents may influence inflammation through many chemokines; for example, PF4 can induce AHR in rats.46 Platelets actively take up serotonin from plasma and concentrate it in their dense granules, becoming the main source of serotonin outside the central nervous system.47 Serotonin modulates adhesion, migration, and cytokine and chemokine production in inflammatory cells and lung epithelial cells.48 Serotonin also induces bronchoconstriction, and mice unable to synthesize serotonin in peripheral tissues had a marked decrease in platelet serotonin and AHR.26 ADP can induce changes on the endothelium that result in leukocyte transendothelial migration via the P2Y1 and P2Y2 receptors,49,50 and global deletion of the ADP receptor P2Y12 in mice has been shown to partially protect from asthma.27 Even though treatment with the P2Y12 inhibitor clopidogrel had contradictory outcomes in mouse models of asthma,27-29 it is currently being tested in asthmatic patients (clinicaltrials.gov identifier: 01955512). On the other hand, the PRINA trial failed to show a significant decrease in mannitol-induced AHR in patients with asthma treated with prasugrel.51 The mechanism of action of anti-P2Y12 thienopyridines in allergic inflammation needs to be clarified, because it may be acting on eosinophils and

References 1. Blair P, Flaumenhaft R. Platelet alpha-granules: basic biology and clinical correlates. Blood Rev. 2009;23(4):177-189. 2. Jonnalagadda D, Izu LT, Whiteheart SW. Platelet secretion is kinetically heterogeneous in an agonist-responsive manner.

haematologica | 2018; 103(7)

not necessarily on platelets.28 Other molecules present in dense granules such as ATP,52 glutamate53 and polyphosphates54 can influence different steps of the allergic inflammatory response. Therefore, the lack of Munc13-4 in platelets could significantly affect the development of asthma by directly impeding dense granule exocytosis and indirectly hindering alpha granule release. Up until now all the evidence supporting a role for platelets in asthma has come from in vitro studies,55 in vivo experimentation on platelet-depleted animals24,25,56 or global KO mice,23,26,27,43 and from pharmacological studies that targeted receptors expressed in multiple cell types.28-30 Here we show in mice with normal numbers of circulating platelets that selective deletion of Munc13-4 in platelets results in reduced AHR and lung eosinophilia, most likely because of impaired dense granule release. It has been proposed that platelet activation could lead to either a hemostatic or an inflammatory response independently of each other.18 A recent publication suggests that low concentrations of collagen are capable of inducing platelet secretion of soluble factors, without causing aggregation or any of the other characteristic features of platelet activation.35 Studies on mouse models of allergic and bacterial airway inflammation suggest that platelet P2Y12 participates mostly in hemostasis, P2Y14 in inflammation, and P2Y1 in both.29,30 The ligand of P2Y1 (ADP) is stored in platelet dense granules, but it is unclear whether the ligand for P2Y14 (UDP-glucose) comes from platelets57 or from other cells.58 Hence, the defect in dense granule exocytosis we noted in platelets lacking Munc13-4 could result in failure to activate P2Y1, P2Y12 and perhaps P2Y14. The simultaneous failure to activate all these receptors could explain why our mutant animals have both abnormal hemostasis and decreased allergic airway inflammation. Acknowledgments The authors would like to thank Margaret M. Gondo (University of Houston) for her assistance with the electron microscopy, Kimberly Langlois and Ngoc-Anh Bui-Thanh (Baylor College of Medicine) for their assistance with the in vivo thrombosis assay, and Terry Hoppe (Texas A&M Institute of Biosciences and Technology) for his assistance manufacturing the tail-bleeding device. Funding This project was supported by the National Institutes of Health AI093533A, CA016672 and HL116524, and the Department of Veterans Affairs Merit Review Award I01 BX002551. EIC received support from Instituto TecnolĂłgico y de Estudios Superiores de Monterrey and Consejo Nacional de Ciencia y TecnologĂ­a (CONACyT, PhD grant scholarship # 352566). The contents of this manuscript do not represent the views of the Department of Veterans Affairs or the United States Government.

Blood. 2012;120(26):5209-5216. 3. Ciferri S, Emiliani C, Guglielmini G, Orlacchio A, Nenci GG, Gresele P. Platelets release their lysosomal content in vivo in humans upon activation. Thromb Haemost. 2000;83(1):157-164. 4. Chen YA, Scheller RH. SNARE-mediated membrane fusion. Nat Rev Mol Cell Biol. 2001;2(2):98-106.

5. Klenchin VA, Martin TF. Priming in exocytosis: attaining fusion-competence after vesicle docking. Biochimie. 2000;82(5):399-407. 6. Ma C, Li W, Xu Y, Rizo J. Munc13 mediates the transition from the closed syntaxinMunc18 complex to the SNARE complex. Nat Struct Mol Biol. 2011;18(5):542-549. 7. Chicka MC, Ren Q, Richards D, et al. Role of Munc13-4 as a Ca2+-dependent tether

1243


E.I. Cardenas et al.

8.

9.

10.

11.

12. 13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

1244

during platelet secretion. Biochem J. 2016; 473(5):627-639. Schumacher D, Strilic B, Sivaraj KK, Wettschureck N, Offermanns S. Plateletderived nucleotides promote tumor-cell transendothelial migration and metastasis via P2Y2 receptor. Cancer Cell. 2013;24(1):130-137. Ren Q, Wimmer C, Chicka MC, et al. Munc13-4 is a limiting factor in the pathway required for platelet granule release and hemostasis. Blood. 2010;116(6):869-877. Harper MT, van den Bosch MT, Hers I, Poole AW. Platelet dense granule secretion defects may obscure alpha-granule secretion mechanisms: evidence from Munc13-4-deficient platelets. Blood. 2015; 125(19):30343036. Savage JS, Williams CM, Konopatskaya O, Hers I, Harper MT, Poole AW. Munc13-4 is critical for thrombosis through regulating release of ADP from platelets. J Thromb Haemost. 2013;11(4):771-775. Schumacher D, Strilic B, Sivaraj KK, Wettschureck N, Offermanns S. Response to Harper et al. Cancer Cell. 2013; 24(3):288. Stegner D, Deppermann C, Kraft P, et al. Munc13-4-mediated secretion is essential for infarct progression but not intracranial hemostasis in acute stroke. J Thromb Haemost. 2013;11(7):1430-1433. Chehab T, Santos NC, Holthenrich A, et al. A novel Munc13-4/S100A10/annexin A2 complex promotes Weibel-Palade body exocytosis in endothelial cells. Mol Biol Cell. 2017;28(12):1688-1700. Feldmann J, Callebaut I, Raposo G, et al. Munc13-4 is essential for cytolytic granules fusion and is mutated in a form of familial hemophagocytic lymphohistiocytosis (FHL3). Cell. 2003;115(4):461-473. Nakamura L, Bertling A, Brodde MF, et al. First characterization of platelet secretion defect in patients with familial hemophagocytic lymphohistiocytosis type 3 (FHL-3). Blood. 2015;125(2):412-414. Chen Y, Wang Z, Cheng Y, Tang Y. Novel mutations in the UNC13D gene carried by a Chinese neonate with hemophagocytic lymphohistiocytosis. Yonsei Med J. 2013; 54(4):1053-1057. Idzko M, Pitchford S, Page C. Role of platelets in allergic airway inflammation. J Allergy Clin Immunol. 2015;135(6):14161423. Sullivan PJ, Jafar ZH, Harbinson PL, Restrick LJ, Costello JF, Page CP. Platelet dynamics following allergen challenge in allergic asthmatics. Respiration. 2000; 67(5):514-517. Kowal K, Pampuch A, Kowal-Bielecka O, DuBuske LM, Bodzenta-Lukaszyk A. Platelet activation in allergic asthma patients during allergen challenge with Dermatophagoides pteronyssinus. Clin Exp Allergy. 2006;36(4):426-432. Jeffery PK, Wardlaw AJ, Nelson FC, Collins JV, Kay AB. Bronchial biopsies in asthma. An ultrastructural, quantitative study and correlation with hyperreactivity. Am Rev Respir Dis. 1989;140(6):1745-1753. Metzger WJ, Sjoerdsma K, Richerson HB, et al. Platelets in bronchoalveolar lavage from asthmatic patients and allergic rabbits with allergen-induced late phase responses. Agents Actions Suppl. 1987;21:151-159. Pitchford SC, Momi S, Baglioni S, et al. Allergen induces the migration of platelets to lung tissue in allergic asthma. Am J Resp Crit Care. 2008;177(6):604-612. Pitchford SC, Yano H, Lever R, et al. Platelets are essential for leukocyte recruitment in allergic inflammation. J Allergy Clin

Immunol. 2003;112(1):109-118. 25. Coyle AJ, Page CP, Atkinson L, Flanagan R, Metzger WJ. The requirement for platelets in allergen-induced late asthmatic airway obstruction. Eosinophil infiltration and heightened airway responsiveness in allergic rabbits. Am Rev Respir Dis. 1990;142(3):587-593. 26. Durk T, Duerschmied D, Muller T, et al. Production of serotonin by tryptophan hydroxylase 1 and release via platelets contribute to allergic airway inflammation. Am J Resp Crit Care. 2013;187(5):476-485. 27. Paruchuri S, Tashimo H, Feng C, et al. Leukotriene E4-induced pulmonary inflammation is mediated by the P2Y12 receptor. J Exp Med. 2009;206(11):2543-2555. 28. Suh DH, Trinh HK, Liu JN, et al. P2Y12 antagonist attenuates eosinophilic inflammation and airway hyperresponsiveness in a mouse model of asthma. J Cell Mol Med. 2016;20(2):333-341. 29. Amison RT, Momi S, Morris A, et al. RhoA signaling through platelet P2Y(1) receptor controls leukocyte recruitment in allergic mice. J Allergy Clin Immunol. 2015;135(2):528-538. 30. Amison RT, Arnold S, O'Shaughnessy BG, et al. Lipopolysaccharide (LPS) induced pulmonary neutrophil recruitment and platelet activation is mediated via the P2Y1 and P2Y14 receptors in mice. Pulm Pharmacol Ther. 2017;45:62-68. 31. Varoqueaux F, Sigler A, Rhee JS, et al. Total arrest of spontaneous and evoked synaptic transmission but normal synaptogenesis in the absence of Munc13-mediated vesicle priming. Proc Natl Acad Sci USA. 2002; 99(13):9037-9042. 32. Rodarte EM, Ramos MA, Davalos AJ, et al. Munc13 proteins control regulated exocytosis in mast cells. J Biol Chem. 2018; 293(1):345-358. 33. Schwenk F, Baron U, Rajewsky K. A cretransgenic mouse strain for the ubiquitous deletion of loxP-flanked gene segments including deletion in germ cells. Nucleic Acids Res. 1995;23(24):5080-5081. 34. Augustin I, Rosenmund C, Sudhof TC, Brose N. Munc13-1 is essential for fusion competence of glutamatergic synaptic vesicles. Nature. 1999;400(6743):457-461. 35. Ollivier V, Syvannarath V, Gros A, et al. Collagen can selectively trigger a platelet secretory phenotype via glycoprotein VI. PLoS One. 2014;9(8):e104712. 36. Dudenhoffer-Pfeifer M, Schirra C, Pattu V, et al. Different Munc13 isoforms function as priming factors in lytic granule release from murine cytotoxic T lymphocytes. Traffic. 2013;14(7):798-809. 37. Koch H, Hofmann K, Brose N. Definition of Munc13-homology-domains and characterization of a novel ubiquitously expressed Munc13 isoform. Biochem J. 2000;349(Pt 1):247-253. 38. Werner MH, Hannun YA. Delayed accumulation of diacylglycerol in platelets as a mechanism for regulation of onset of aggregation and secretion. Blood. 1991; 78(2):435444. 39. Li Z, Delaney MK, O'Brien KA, Du X. Signaling during platelet adhesion and activation. Arterioscler Thromb Vasc Biol. 2010;30(12):2341-2349. 40. Stalker TJ, Traxler EA, Wu J, et al. Hierarchical organization in the hemostatic response and its relationship to the plateletsignaling network. Blood. 2013; 121(10):1875-1885. 41. Broide DH. Molecular and cellular mechanisms of allergic disease. J Allergy Clin

Immunol. 2001;108(2 Suppl):S65-71. 42. Evans CM, Raclawska DS, Ttofali F, et al. The polymeric mucin Muc5ac is required for allergic airway hyperreactivity. Nat Commun. 2015;6:6281. 43. Pitchford SC, Momi S, Giannini S, et al. Platelet P-selectin is required for pulmonary eosinophil and lymphocyte recruitment in a murine model of allergic inflammation. Blood. 2005;105(5):2074-2081. 44. Johansson MW, Mosher DF. Activation of beta1 integrins on blood eosinophils by Pselectin. Am J Resp Cell Mol. 2011; 45(4):889-897. 45. Johansson MW, Han ST, Gunderson KA, Busse WW, Jarjour NN, Mosher DF. Platelet activation, P-selectin, and eosinophil beta1integrin activation in asthma. Am J Resp Crit Care. 2012;185(5):498-507. 46. Coyle AJ, Uchida D, Ackerman SJ, Mitzner W, Irvin CG. Role of cationic proteins in the airway. Hyperresponsiveness due to airway inflammation. Am J Resp Crit Care. 1994;150(5 Pt 2):S63-71. 47. Mercado CP, Kilic F. Molecular mechanisms of SERT in platelets: regulation of plasma serotonin levels. Mol Interv. 2010; 10(4):231241. 48. Bayer H, Muller T, Myrtek D, et al. Serotoninergic receptors on human airway epithelial cells. Am J Resp Cell Mol. 2007; 36(1):85-93. 49. Kukulski F, Ben Yebdri F, Bahrami F, Fausther M, Tremblay A, Sevigny J. Endothelial P2Y2 receptor regulates LPS-induced neutrophil transendothelial migration in vitro. Mol Immunol. 2010;47(5):991-999. 50. Shen J, DiCorleto PE. ADP stimulates human endothelial cell migration via P2Y1 nucleotide receptor-mediated mitogen-activated protein kinase pathways. Circ Res. 2008;102(4):448-456. 51. Lussana F, Di Marco F, Terraneo S, et al. Effect of prasugrel in patients with asthma: results of PRINA, a randomized, doubleblind, placebo-controlled, cross-over study. J Thromb Haemost. 2015;13(1):136-141. 52. Younas M, Hue S, Lacabaratz C, et al. IL-7 modulates in vitro and in vivo human memory T regulatory cell functions through the CD39/ATP axis. J Immunol. 2013;191(6):3161-3168. 53. Ganor Y, Besser M, Ben-Zakay N, Unger T, Levite M. Human T cells express a functional ionotropic glutamate receptor GluR3, and glutamate by itself triggers integrin-mediated adhesion to laminin and fibronectin and chemotactic migration. J Immunol. 2003;170(8):4362-4372. 54. Dinarvand P, Hassanian SM, Qureshi SH, et al. Polyphosphate amplifies proinflammatory responses of nuclear proteins through interaction with receptor for advanced glycation end products and P2Y1 purinergic receptor. Blood. 2014;123(6):935-945. 55. Palma-Carlos AG, Palma-Carlos ML, Santos MC, de Sousa JR. Platelet aggregation in allergic reactions. Int Arch Aller a Imm. 1991;94(1-4):251-253. 56. Pitchford SC, Riffo-Vasquez Y, Sousa A, et al. Platelets are necessary for airway wall remodeling in a murine model of chronic allergic inflammation. Blood. 2004; 103(2):639-647. 57. Wandall HH, Rumjantseva V, Sorensen AL, et al. The origin and function of platelet glycosyltransferases. Blood. 2012; 120(3):626635. 58. Kreda SM, Okada SF, van Heusden CA, et al. Coordinated release of nucleotides and mucin from human airway epithelial Calu-3 cells. J Physiol. 2007;584(Pt 1):245-259.

haematologica | 2018; 103(7)


ARTICLE

Coagulation & its Disorders

C-reactive protein and risk of venous thromboembolism: results from a population-based case-crossover study

Ferrata Storti Foundation

Gro Grimnes,1,2 Trond Isaksen,1,2 Ynse Ieuwe Gerardus Vladimir Tichelaar,1,3 Jan Brox,1,2 Sigrid Kufaas Brækkan1,2 and John-Bjarne Hansen1,2

K.G. Jebsen Thrombosis Research and Expertise Center (TREC), Department of Clinical Medicine, UiT The Arctic University of Norway, Tromsø, Norway; 2Division of Internal Medicine, University Hospital of North Norway, Tromsø, Norway and 3Department of Vascular Medicine, Academic Medical Center, University of Amsterdam, the Netherlands

1

Haematologica 2018 Volume 103(7):1245-1250

ABSTRACT

L

ong-term, low-grade inflammation does not seem to be a risk factor for venous thromboembolism. The impact of acute inflammation, regardless of cause, on risk of venous thromboembolism is scarcely studied. We aimed to investigate the impact of acute inflammation, assessed by C-reactive protein, on short-term risk of venous thromboembolism. We conducted a case-crossover study of patients with venous thromboembolism (n=707) recruited from a general population. Information on triggers and C-reactive protein levels were retrieved from hospital records during the 90 days before the event (hazard period) and in four preceding 90-day control periods. Conditional logistic regression was used to obtain β coefficients for change in natural log (ln) transformed C-reactive protein from control to hazard periods and to determine corresponding odds ratios for venous thromboembolism. Median C-reactive protein was 107 mg/L in the hazard period, and ranged from 7 mg/L to 16 mg/L in the control periods. The level of C-reactive protein was 58% (95% CI 39-77%) higher in the hazard period than in the control periods. A one-unit increase in ln-C-reactive protein was associated with increased risk of venous thromboembolism (OR 1.79, 95% CI 1.48-2.16). The risk estimates were only slightly attenuated after adjustment for immobilization and infection. In stratified analyses, ln-C-reactive protein was associated with increased risk of venous thromboembolism in cases with (OR 1.55, 95% CI 1.01-2.38) and without infection (OR 1.77, 95% CI 1.22-2.57). In conclusion, we found that acute inflammation, assessed by C-reactive protein, was a trigger for venous thromboembolism.

Introduction Venous thromboembolism (VTE), consisting of deep vein thrombosis (DVT) and pulmonary embolism (PE), is a multicausal disease associated with substantial morbidity and mortality.1 Contrary to arterial thrombotic disease, there has been no decline in the incidence of VTE during the last decades.2,3 Thus, there is an unmet need for improved risk stratification and prevention of VTE. Chronic inflammation is recognized as part of the pathophysiological process in arterial thrombosis,4 but its role in venous thrombosis has been less clear.5,6 Inflammatory biomarkers such as high-sensitivity C-reactive protein (hs-CRP) can predict long-term risk of arterial cardiovascular disease, but have not been associated with risk of VTE in prospective studies with long-term follow up.7-9 However, in studies with shorter follow up time, inflammatory markers such as hs-CRP and neutrophil to lymphocyte ratio were associated with increased risk of VTE.10,11 Several conditions associated with increased risk of VTE, including cancer, acute infections, autoimmune diseases and obesity, share the feature of inflammation.12haematologica | 2018; 103(7)

Correspondence: gro.grimnes@uit.no

Received: December 21, 2017. Accepted: April 18, 2018. Pre-published: April 19, 2018. doi:10.3324/haematol.2017.186957 Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/103/7/1245 ©2018 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.

1245


G. Grimnes et al.

Figure 1. Case-crossover study design. Relevant risk factors and levels of C-reactive protein were recorded for each case of venous thromboembolism, in the 90-day hazard period prior to the event and in four preceding 90-day control periods, separated by a 90-day washout period.

15

Even though chronic diseases such as autoimmune diseases carry an increased VTE risk, the risk of VTE is more pronounced during disease flare-ups, where inflammation is predominant.14 Acute infection triggers an acute inflammatory response, and several studies have found an increased risk of VTE associated with infections.16,17 We have previously investigated the role of infection during hospitalization in a case-crossover design, and found that acute infection was a frequent and strong trigger for VTE, also after adjustment for immobilization and other transient risk factors.18 Taken together, this points towards an association between inflammation and VTE which is dependent on the degree of the inflammatory response within a shorter time perspective than as observed in arterial thrombotic disease. The objective of this study was to investigate the role of acute inflammation, assessed by CRP, as a trigger for VTE using a case-crossover design. In this study design, each case serves as his or her own control, and the design is therefore well suited for studying transient risk factors.19 We hypothesized that increased CRP, independent of cause, was a trigger of VTE.

Methods For an extensive description of the methods, please see the Online Supplementary Methods. We conducted a case-crossover study including all incident VTE cases (n=707) diagnosed among the participants of the fourth Tromsø Study during 1994-2012. The study was approved by the regional ethics committee, and all participants provided informed written consent. A hazard period of 90 days preceding the incident VTE was compared to four preceding 90-day control periods. To avoid carry-over effects, we included a 90-day washout period between the hazard and control periods (Figure 1). For every VTE case, trained medical personnel searched the hospital medical records for relevant risk factors, diagnostic procedures, surgical and medical treatment, laboratory test results and diagnoses during hospital admissions, day care and outpatient clinic visits in any of the hazard or control periods. We did not have access to medical records from general practice. A transient risk factor, or trigger, was defined by its presence during the defined 90-day period. If 1246

an exposure occurred over several days, it was considered to have occurred if any of the days of exposure fell within the specified 90-day time period. CRP was analyzed in serum with a particle-enhanced immunoturbidimetric assay. CRP measurements from the last two days before the date of VTE were not included in the analyses to avoid reverse causation, as CRP in these cases could be caused by an inflammatory response to the VTE itself. If a participant had several CRP-measurements during a control or hazard period, the maximum CRP value for each period was used. Statistical analyses were carried out using STATA version 14.0 (Stata corporation, College station, Texas, USA). Natural log (ln) transformation was used for CRP to achieve normal distributions. Only cases who had their CRP measured in both the hazard and a control period were included in the main analyses. Since CRP was measured upon request, this would yield the most conservative risk estimate. We used conditional logistic regression to obtain β coefficients with 95% confidence intervals (CI) for change in ln-CRP from control to hazard periods, and to calculate odds ratios (ORs) with 95% CI per one-unit change in ln-CRP. The analyses were adjusted for immobilization and infection in two different models. Further, we performed analyses comparing CRP in the hazard period with each individual control period, to investigate whether time to event influenced the association between acute inflammation and VTE. In the main analyses, we included only hazard and control periods in which CRP had been measured. The risk estimates from this conservative approach could be underestimations, as subjects with no hospital contact during a hazard or control period, or with a hospital contact without a CRP measurement, most likely had a low CRP at that time. To address this concern, we performed sensitivity analyses where missing CRP values were set at the lower reported cut-off level of 5 mg/L. We also performed sensitivity analyses where we included only those CRP-measurements performed more than seven days before the date of VTE, to address potential bias due to reverse causation.

Results In total, 707 incident VTEs were identified, of which there were 408 DVTs and 299 PEs (with or without concurrent DVT). The median age at time of VTE-diagnosis haematologica | 2018; 103(7)


CRP as a trigger for venous thromboembolism

was 71 years, and 53.6% were women. Moreover, 416 (58.8%) VTEs were not related to recent hospitalization, 135 (19.1%) occurred during hospitalization, and 156 (22.1%) were diagnosed within 30 days after hospitalization (Table 1). Infection was the most common risk factor in the hazard period, recorded in 267 (37.8%) of the periods, followed by immobilization (31.4%), and cancer (24.3%). The distribution of VTE risk factors and triggers in the hazard- and control periods are shown in Table 2. Prophylactic treatment with low-molecular weight heparin was prescribed in 138 (19.5%) of the 707 hazard periods, and in 78 (2.8%) of the 2828 control periods. In total, after exclusion of hospital contacts the last two days before VTE-diagnosis, 1283 hospital contacts were registered during the hazard period and control periods (Table 3). The number of hospital contacts was higher in the periods closest to the VTE, increasing from 165, 172, 187 and 199, respectively, in the control periods, to 560 contacts in the hazard period. CRP was measured in 298 cases during the hazard period, and in 75, 72, 86 and 96 cases during the four control periods. Median CRP was highest in the hazard period (CRP 107 mg/L), and ranged from 7 mg/L to 16 mg/L in the control periods (Table 3). Based on β coefficients for ln-CRP obtained from logistic regression analyses, the mean CRP level was 58% (95% CI 39-77%) higher in hazard than in control periods. After adjustment for immobilization, the CRP level was 51% (95% CI 31-70%) higher, and after adjustment for infection the CRP level was 40% (95% CI 19-61%) higher in the hazard than in the control periods (Table 4). This corresponded to a 1.8-fold increased estimated VTE risk per unit increase in ln-CRP (OR 1.79, 95% CI 1.482.16), which were only slightly attenuated after adjustment for immobilization (OR 1.66, 95% CI 1.37-2.02) and for infection (OR 1.50, 95% CI 1.21-1.85). In analyses stratified for infection, the mean CRP level was 57% (95% CI 20-94%) higher in the hazard versus control periods in those without infection and 44% (95% CI 1-87%) higher in those with infection. Estimated increase in VTE risk according to a one-unit increase in lnCRP was 1.8-fold (OR 1.77, 95% CI 1.22-2.57) in those without infection and 1.6-fold in those with infection (OR 1.55, 95% CI 1.01-2.38). Adjustment for immobilization revealed similar results (Table 4). Sensitivity analyses restricted to CRP measurements that were conducted more than 7 days before the date of VTE-diagnosis yielded essentially similar results (Online Supplementary Table S1). In sensitivity analyses where

missing CRP values were set to 5 mg/L, the estimated VTE risk per unit increase in ln-CRP was 2.4-fold increased (OR 2.36, 95% CI 2.14-2.61) (Online Supplementary Table S2). When comparing CRP levels in the hazard period to each control period (C1-4) separately, there was no trend for change in CRP level according to time between control and hazard period (the CRP level was increased by 65%, 77%, 42% and 60%, respectively) (Table 5). The estimated risk of VTE by ln-CRP was 1.9-fold increased when comparing the hazard period with C1, 2.2-fold increased when compared with C2, 1.5-fold increased when compared with C3, and 1.8-fold increased when compared with C4.

Discussion In this case-crossover study including 707 incident VTEs, we found that acute inflammation, assessed by increase in CRP, was a trigger for VTE. The association remained after adjustment for immobilization and for infection. In stratified analyses, inflammation assessed by CRP was associated with increased risk of VTE also in cases without infection. The strength of the estimated risk of VTE by CRP remained similar when separately compared to the different control periods. Clinically, DVT often presents with the cardinal signs of inflammation; i.e., redness, swelling, heat, pain and disturbance of function. In a case-control study investigating inflammatory markers, patients with DVT had significantly higher levels of inflammatory markers, including CRP, than controls.20 Based on these results the authors suggested that inflammation was a consequence rather than a cause of VTE. However, the fact that inflammation is a consequence of VTE does not exclude the possibility that inflammation can be a cause of the disease. Previous prospective studies on the association between inflammation and VTE have shown conflicting results. In a cohort of healthy men followed for more than 8 years, baseline hs-CRP was associated with increased risk of arterial but not venous thrombosis.9 Similarly, two long-term population-based cohorts reported no association between inflammatory markers and VTE.7,8 In contrast, studies with shorter follow-up time have shown an association between CRP and incident VTE. In a population-based case-cohort study with 515 VTE-cases and 1505 controls, an association between

Table 2. Triggers and risk factors for venous thromboembolism.

Triggers/risk factors Table 1. Characteristics of study participants at the time of venous thromboembolism (VTE) diagnosis. Median age; years ± SD 71 ±14 Female sex; n (%) 379 (53.6) DVT only; n (%) 408 (57.7) PE+/-DVT; n (%) 299 (42.3) VTE during hospitalization; n (%) 135 (19.1) VTE within 30 days after hospitalization; n (%) 156 (22.1) No hospitalization the last 30 days before VTE; n (%) 416 (58.8) DVT: deep vein thrombosis, PE: pulmonary embolism.

haematologica | 2018; 103(7)

Infection; n (%) Immobilizationb; n (%) Cancer; n (%) Surgery; n (%) Trauma; n (%) Central venous catheter; n (%)

Hazard period Control periods (n=707) (n=2828)a 267 (37.8) 222 (31.4) 172 (24.3) 118 (16.7) 71 (10.0) 56 (7.9)

107 (3.8) 57 (2.0) 375 (13.2)c 88 (3.1) 25 (0.9) 17 (0.6)

a 707 cases, four control periods for each case; bbedrest ≥3 days, ECOG 4, other immobilizing factor (e.g., wheelchair use); cbased on 106 unique individuals with cancer in one or more of the control periods.

1247


G. Grimnes et al. H

Table 3. Hospital contacts and measurements of C-reactive protein (CRP) in hazard (H) and control periods (C1-C4).

C2

C3 (3-0 months) N=707

C4 (18-15 months) N=707

(15-12 months) N=707

(12-9 months) N=707

(9-6 months) N=707

Number of hospital contacts Number of cases with CRP measurements CRPa (mg/L), median (25-perc, 75-perc)

560 298 107 (25, 195)

165 75 8 (5, 61)

172 72 7 (5, 23)

187 86 15 (5, 94)

199 96 16 (5, 85.5)

a

C1

Maximum CRP, with measurements the last two days before date of VTE excluded.

Table 4. Association of C-reactive proteina with risk of venous thromboembolism

βb (95% CI)

Hazard period compared to control periods Adjusted for immobilization βb (95% CI)

Adjusted for infection βb (95% CI)

All cases Cases with infection Cases without infection

0.58 (0.39-0.77) 0.44 (0.01-0.87) 0.57 (0.20-0.94)

0.51 (0.31-0.70) 0.45 (-0.02-0.92) 0.57 (0.18-0.96)

0.40 (0.19-0.61) -

OR (95% CI)

OR (95% CI)

OR (95% CI)

All cases Cases with infection Cases without infection

1.79 (1.48-2.16) 1.55 (1.01-2.38) 1.77 (1.22-2.57)

1.66 (1.37-2.02) 1.57 (0.98-2.51) 1.77 (1.20-2.60)

1.50 (1.21-1.85) -

OR: odds ratio, CI: confidence interval. aNatural log transformed C-reactive protein. bWhen multiplied by 100, β coefficients can be interpreted as percentage difference compared with the reference group.

baseline CRP and VTE was only present in cases suffering a VTE within the first year after baseline.10 We found a similar time-dependent pattern between another inflammation marker, the neutrophil to lymphocyte ratio (NLR), and VTE risk in the Tromsø study cohort.11 There was no association between NLR and risk of VTE after a median follow-up time of 17.7 years, but when follow-up time was restricted to the first 3 years, those with the highest baseline NLR had a 2.4-fold increased risk of VTE. Taken together, these studies suggest that acute and augmented inflammation rather than longstanding, low-grade inflammation is associated with VTE risk. Accordingly, in this case-crossover study, we found that acute inflammation assessed by CRP was associated with increased risk of VTE. Acute infection, a strong trigger of inflammation, is a risk factor for VTE, and higher CRPlevels are expected in patients with acute infections.16,17,21 After adjustment for infection, and in analyses stratified for infection, increased serum levels of CRP were still associated with increased VTE risk, also in cases without infection. Thus, our findings suggest that inflammatory responses caused by non-infectious conditions, such as cancer, surgery, acute medical conditions and trauma, can partly explain the VTE risk related to these conditions. Immobilization may accompany these conditions, and thereby act as a confounder for the observed association. However, in our study the risk estimates remained essentially similar after adjustment for immobilization. The risk of VTE has been shown to be highest the first two weeks following an infection, and to gradually decline thereafter.16,17 In our study, there was no trend of a change in risk estimates according to time between the hazard and control periods. This further supports that acute inflammation of short duration is more important for the VTE risk. Some chronic inflammatory conditions, 1248

such as autoimmune disorders and rheumatic diseases, also carry increased risk of VTE. However, the risk of VTE in patients with inflammatory bowel disease, for example, is especially high during disease flare-ups, where acute inflammation dominates.22 Furthermore, in a populationbased cohort study on VTE risk in patients with psoriasis and rheumatoid arthritis (RA), patients with severe psoriasis and RA-patients in need of a disease modifying antirheumatic drug (DMARD) had higher estimated risk of VTE than those not prescribed DMARDs.23 Inflammation and coagulation are closely linked.24 Inflammation can be triggered by infection, tissue injury or tissue stress and malfunction.25 Of these triggers, inflammation induced by infection has been best characterized. Extensive crosstalk exists between the coagulation and the complement cascades, and complement activation enhances coagulation through increased tissue factor (TF) expression and by inhibition of fibrinolysis.26 Anticoagulant activity by the protein C- pathway is down-regulated by inflammatory cytokines.27 TF expression increases in response to inflammatory cytokines and through recruitment from microvesicles and monocytes induced by P-selectin.28 Activated neutrophils secrete neutrophil extracellular traps (NETs), composed of proteins and decondensed chromatin.29 In addition to an important role in neutralizing and killing microbes, NETs also contribute to coagulation and platelet aggregation.30 NET formation occurs not only in response to infection, but also in sterile inflammatory processes and in metastatic cancer.30 CRP is an acute-phase protein rapidly synthesized mainly in the liver under control by inflammatory cytokines, and CRP levels cease rapidly when the stimuli for production is diminished.31 CRP has no diurnal variation, is unaffected by eating, and drugs reducing CRP typhaematologica | 2018; 103(7)


CRP as a trigger for venous thromboembolism

Table 5. Association of C-reactive proteina with risk of venous thromboembolism.

Hazard period (H) compared to individual control periods (C1-C4) H versus C1 H versus C2 H versus C3 βb (95% CI) βb (95% CI) βb (95% CI) All cases All cases

0.65 (0.22-1.08) OR (95% CI) 1.92 (1.26-2.95)

0.77 (0.33-1.20) OR (95% CI) 2.15 (1.39-3.33)

0.42 (0.11-0.73) OR (95% CI) 1.52 (1.11-2.08)

H versus C4 βb (95% CI) 0.60 (0.30-0.90) OR (95% CI) 1.82 (1.36-2.45)

OR: odds ratio; CI: confidence interval. aNatural log transformed C-reactive protein. bWhen multiplied by 100, β coefficients can be interpreted as percentage difference compared with the reference group.

ically also affect the underlying acute-phase stimulus.31 CRP is commonly used as a marker of inflammation in clinical practice. Since our study was based on clinical data, other markers of inflammation and coagulation were not available as these were only occasionally measured. CRP is therefore well suited to serve as a marker of inflammation, a process linked to coagulation through several pathways as described above. In addition to its role in innate immunity and complement activation,31 CRP has been found to have prothrombotic effects in some studies.32-34 Due to methodological issues, especially the possibility of contamination of CRP preparations with bacterial lipopolysaccharides, controversy regarding a direct role of CRP in thrombosis still exists.35 Our study has both strengths and limitations. The casecrossover design is suitable for studying transient risk factors, as potential fixed confounders are mainly controlled for through the design. Further, the VTE-cases were derived from a large, population-based cohort with high attendance rate, and all VTE events were symptomatic and validated. All hospital care in the region is provided by a single hospital, facilitating the completeness of the VTE registry. However, some VTE cases might have been clinically diagnosed and treated without hospital contact, and some cases of PE presenting as sudden death might have been misclassified. As each subject serves as his or her own control, such potential cases would most likely not affect our results. Our study was limited to information from hospital records, as we did not have access to data from general practice. In most cases, a high CRP level measured in general practice will increase the likeli-

References 1. Rosendaal FR. Venous thrombosis: a multicausal disease. Lancet. 1999; 353(9159): 1167-1173. 2. Mannsverk J, Wilsgaard T, Mathiesen EB, et al. Trends in modifiable risk factors are associated with declining incidence of hospitalized and nonhospitalized acute coronary heart disease in a population. Circulation. 2016;133(1):74-81. 3. Arshad N, Isaksen T, Hansen JB, Braekkan SK. Time trends in incidence rates of venous thromboembolism in a large cohort recruited from the general population. Eur J Epidemiol. 2017;32(4):299-305. 4. Libby P, Ridker PM, Hansson GK. Inflammation in atherosclerosis: from pathophysiology to practice. J Am Coll Cardiol. 2009;54(23):2129-2138. 5. Zacho J, Tybjaerg-Hansen A, Nordestgaard

haematologica | 2018; 103(7)

hood of the patient to be referred to hospital. We cannot find any reason why hospital referral praxis should differ in hazard versus control periods. Our data source was the hospital medical records of VTE-cases, and CRP measurements were made according to each clinician’s preference. As each case serves as his or her own control in the case-crossover design, only those who had a CRP measurement in both the hazard period and in one or more control periods were included in the analyses. We can assume that CRP levels were lower in periods where the clinician did not find a reason to have CRP measured than in periods where CRP was measured. As more CRP measurements were done in hazard than in control periods, this might have diluted the risk estimates. Sensitivity analyses where missing CRP values were set to 5 mg/L resulted in higher risk estimates than in the main analyses. We did not adjust for therapeutic agents that could affect CRP-levels, such as statins, steroids and other immunosuppressive drugs, as we wanted to investigate the impact of inflammation on VTE regardless of the cause of the inflammation. CRP levels measured the last two days before the date of VTE were omitted to avoid reverse causation. However, we cannot be certain that elevated CRP was not caused by the incident VTE more than two days before the date of VTE diagnosis. To address this, we performed sensitivity analyses excluding CRP measurements the last seven days before the date of VTE-diagnosis, and the results were essentially similar. In conclusion, acute inflammation assessed by CRP was a trigger for VTE in this case-crossover study, also in cases with inflammatory triggers other than infection.

BG. C-reactive protein and risk of venous thromboembolism in the general population. Arterioscler Thromb Vasc Biol. 2010; 30(8):1672-1678. 6. Riva N, Donadini MP, Ageno W. Epidemiology and pathophysiology of venous thromboembolism: similarities with atherothrombosis and the role of inflammation. Thromb Haemost. 2015; 113(6):1176-1183. 7. Hald EM, Braekkan SK, Mathiesen EB, et al. High-sensitivity C-reactive protein is not a risk factor for venous thromboembolism: the Tromso study. Haematologica. 2011;96(8):1189-1194. 8. Tsai AW, Cushman M, Rosamond WD, et al. Coagulation factors, inflammation markers, and venous thromboembolism: the longitudinal investigation of thromboembolism etiology (LITE). Am J Med. 2002;113(8):636-642.

9. Ridker PM, Cushman M, Stampfer MJ, Tracy RP, Hennekens CH. Inflammation, aspirin, and the risk of cardiovascular disease in apparently healthy men. N Engl J Med. 1997;336(14):973-979. 10. Quist-Paulsen P, Naess IA, Cannegieter SC, et al. Arterial cardiovascular risk factors and venous thrombosis: results from a population-based, prospective study (the HUNT 2). Haematologica. 2010;95(1):119125. 11. Grimnes G, Horvei LD, Tichelaar V, Braekkan SK, Hansen JB. Neutrophil to lymphocyte ratio and future risk of venous thromboembolism and mortality-The Tromso Study. Haematologica. 2016; 101(10):e401-e404. 12. Heit JA. Epidemiology of venous thromboembolism. Nat Rev Cardiol. 2015; 12(8):464-474. 13. Mantovani A, Allavena P, Sica A, Balkwill F.

1249


G. Grimnes et al.

14.

15. 16.

17.

18.

19.

20.

1250

Cancer-related inflammation. Nature. 2008;454(7203):436-444. Tichelaar YI, Kluin-Nelemans HJ, Meijer K. Infections and inflammatory diseases as risk factors for venous thrombosis. A systematic review. Thromb Haemost. 2012; 107(5):827-837. Samad F, Ruf W. Inflammation, obesity, and thrombosis. Blood. 2013;122(20):3415-3422. Schmidt M, Horvath-Puho E, Thomsen RW, Smeeth L, Sorensen HT. Acute infections and venous thromboembolism. J Intern Med. 2012;271(6):608-618. Smeeth L, Cook C, Thomas S, Hall AJ, Hubbard R, Vallance P. Risk of deep vein thrombosis and pulmonary embolism after acute infection in a community setting. Lancet. 2006;367(9516):1075-1079. Grimnes G, Isaksen T, Tichelaar YIGV, Braekkan SK, Hansen J-B. Acute infection as a trigger for incident venous thromboembolism: Results from a populationbased case-crossover study. Res Pract Thromb Haemost. 2018;2(1):85-92. Maclure M. The case-crossover design: a method for studying transient effects on the risk of acute events. Am J Epidemiol. 1991;133(2):144-153. Roumen-Klappe EM, den Heijer M, van Uum SH, van der Ven-Jongekrijg J, van der Graaf F, Wollersheim H. Inflammatory

21.

22.

23.

24. 25. 26. 27.

response in the acute phase of deep vein thrombosis. J Vasc Surg. 2002;35(4):701706. Rogers MA, Levine DA, Blumberg N, Flanders SA, Chopra V, Langa KM. Triggers of hospitalization for venous thromboembolism. Circulation. 2012;125(17):20922099. Grainge MJ, West J, Card TR. Venous thromboembolism during active disease and remission in inflammatory bowel disease: a cohort study. Lancet. 2010; 375(9715):657-663. Ogdie A, Kay McGill N, Shin DB, et al. Risk of venous thromboembolism in patients with psoriatic arthritis, psoriasis and rheumatoid arthritis: a general populationbased cohort study. Eur Heart J. 2017 Apr 20. [Epub ahead of print] Esmon CT, Xu J, Lupu F. Innate immunity and coagulation. J Thromb Haemost. 2011; 9 Suppl 1:182-188. Medzhitov R. Origin and physiological roles of inflammation. Nature. 2008; 454(7203):428-435. Rittirsch D, Flierl MA, Ward PA. Harmful molecular mechanisms in sepsis. Nat Rev Immunol. 2008;8(10):776-787. Esmon CT. The impact of the inflammatory response on coagulation. Thromb Res. 2004;114(5-6):321-327.

28. Esmon CT. Molecular circuits in thrombosis and inflammation. Thromb Haemost. 2013;109(3):416-420. 29. Papayannopoulos V. Neutrophil extracellular traps in immunity and disease. Nat Rev Immunol. 2018;18(2):134-147. 30. Martinod K, Wagner DD. Thrombosis: tangled up in NETs. Blood. 2014;123(18):27682776. 31. Pepys MB, Hirschfield GM. C-reactive protein: a critical update. J Clin Invest. 2003; 111(12):1805-1812. 32. Singh U, Devaraj S, Jialal I. C-reactive protein decreases tissue plasminogen activator activity in human aortic endothelial cells: evidence that C-reactive protein is a procoagulant. Arterioscler Thromb Vasc Biol. 2005;25(10):2216-2221. 33. Devaraj S, Yun JM, Adamson G, Galvez J, Jialal I. C-reactive protein impairs the endothelial glycocalyx resulting in endothelial dysfunction. Cardiovasc Res. 2009;84(3):479-484. 34. Pasceri V, Willerson JT, Yeh ET. Direct proinflammatory effect of C-reactive protein on human endothelial cells. Circulation. 2000;102(18):2165-2168. 35. Ridker PM. From C-reactive protein to interleukin-6 to interleukin-1: moving upstream to identify novel targets for atheroprotection. Circ Res. 2016;118(1): 145-156.

haematologica | 2018; 103(7)


Haematologica, Volume 103, Issue 7  
Haematologica, Volume 103, Issue 7