Haematologica, Volume 101, issue 1 by Haematologica

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

Editor-in-Chief Jan Cools (Leuven)

Deputy Editor Luca Malcovati (Pavia)

Managing Director Antonio Majocchi (Pavia)

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

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

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

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

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

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

Information for readers, authors and subscribers Haematologica (print edition, pISSN 0390-6078, eISSN 1592-8721) publishes peer-reviewed papers on all areas of experimental and clinical hematology. The journal is owned by a non-profit organization, the Ferrata Storti Foundation, and serves the scientific community following the recommendations of the World Association of Medical Editors (www.wame.org) and the International Committee of Medical Journal Editors (www.icmje.org). Haematologica publishes editorials, research articles, review articles, guideline articles and letters. Manuscripts should be prepared according to our guidelines (www.haematologica.org/information-for-authors), and the Uniform Requirements for Manuscripts Submitted to Biomedical Journals, prepared by the International Committee of Medical Journal Editors (www.icmje.org). Manuscripts should be submitted online at http://www.haematologica.org/. Conflict of interests. According to the International Committee of Medical Journal Editors (http://www.icmje.org/#conflicts), “Public trust in the peer review process and the credibility of published articles depend in part on how well conflict of interest is handled during writing, peer review, and editorial decision making”. The ad hoc journal’s policy is reported in detail online (www.haematologica.org/content/policies). Transfer of Copyright and Permission to Reproduce Parts of Published Papers. Authors will grant copyright of their articles to the Ferrata Storti Foundation. No formal permission will be required to reproduce parts (tables or illustrations) of published papers, provided the source is quoted appropriately and reproduction has no commercial intent. Reproductions with commercial intent will require written permission and payment of royalties. Detailed information about subscriptions is available online at www.haematologica.org. Haematologica is an open access journal. Access to the online journal is free. Use of the Haematologica App (available on the App Store and on Google Play) is free. For subscriptions to the printed issue of the journal, please contact: Haematologica Office, via Giuseppe Belli 4, 27100 Pavia, Italy (phone +39.0382.27129, fax +39.0382.394705, E-mail: info@haematologica.org). Rates of the International edition for the year 2015 are as following: Print edition

Institutional Euro 500

Personal Euro 150

Advertisements. Contact the Advertising Manager, Haematologica Office, via Giuseppe Belli 4, 27100 Pavia, Italy (phone +39.0382.27129, fax +39.0382.394705, e-mail: marketing@haematologica.org). Disclaimer. Whilst every effort is made by the publishers and the editorial board to see that no inaccurate or misleading data, opinion or statement appears in this journal, they wish to make it clear that the data and opinions appearing in the articles or advertisements herein are the responsibility of the contributor or advisor concerned. Accordingly, the publisher, the editorial board and their respective employees, officers and agents accept no liability whatsoever for the consequences of any inaccurate or misleading data, opinion or statement. Whilst all due care is taken to ensure that drug doses and other quantities are presented accurately, readers are advised that new methods and techniques involving drug usage, and described within this journal, should only be followed in conjunction with the drug manufacturer’s own published literature. Direttore responsabile: Prof. Edoardo Ascari; Autorizzazione del Tribunale di Pavia n. 63 del 5 marzo 1955. Printing: Tipografia PI-ME, via Vigentina 136, Pavia, Italy. Printed in December 2015.

haematologica calendar of events

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

ESH-ENERCA Training Course on Diagnosis and Management of Very Rare Red Cell and Iron Disorders European School of Haematology Chairs: P Aguilar-Martinez, P Bianchi, A Iolascon, R Van Wijk, A Zanella January 29-30, 2016 Lisbon, Portugal

Hematology Tutorial on Myeloid Malignancies & Multiple Myelomas Chairs: MB Argawal, R Foà, S McCann January 29-31, 2016 Mumbai, India

The 9th Annual Congress of the European Association for Haemophilia and Allied Disorders 2016 Chairs: P de Moerloose, C Hermans, J Astermark February 3 – 5, 2016 Malmö, Sweden

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

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

Calendar of Events updated on December 11, 2015

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

Table of Contents Volume 101, Issue 1: January 2016 Cover Figure Humanized mice as novel tools in hematology research. (Image created by www.somersault1824.com)

Editorials 1

Resolutions for the NewYear Jan Cools

Innovation in the field of thrombocytopenias: achievements since the beginning of the century and promises for the future Carlo L. Balduini and Patrizia Noris

Review Articles 5

Humanized hemato-lymphoid system mice - Leaders in Hematology review series Alexandre P.A. Theocharides, et al.

Risk factors for relapse after allogeneic transplantation in acute myeloid leukemia Gert J. Ossenkoppele, et al.

Articles Hematopoiesis

The role of CD44 in fetal and adult hematopoietic stem cell regulation Huimin Cao, et al.

Iron Metabolism & Its Disorders

Second international round robin for the quantification of serum non-transferrin-bound iron and labile plasma iron in patients with iron-overload disorders Louise de Swart, et al.

Platelet Biology & Its Disorders

Cytoskeletal perturbation leads to platelet dysfunction and thrombocytopenia in variant forms of Glanzmann thrombasthenia Loredana Bury, et al.

Bone Marrow Failure

PPARg antagonist attenuates mouse immune-mediated bone marrow failure by inhibition of T cell function Kazuya Sato, et al.

Acute Lymphoblastic Leukemia

Relapsed childhood acute lymphoblastic leukemia in the Nordic countries: prognostic factors, treatment and outcome Trausti Oskarsson, et al.

Haematologica 2016; vol. 101 no. 1 - January 2016 http://www.haematologica.org/

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

Clinical significance of bax/bcl-2 ratio in chronic lymphocytic leukemia Maria Ilaria Del Principe, et al.

Plasma Cell Disorders

Risk factors for venous thromboembolism in immunoglobulin light chain amyloidosis Katherine M. Bever, et al.

Cell theraphy & Immunotherapy

Impact of immunosuppressive drugs on the therapeutic efficacy of ex vivo expanded human regulatory T cells Cristiano ScottĂ , et al.

Letters to the Editor Letters are available online only at www.haematologica.org/content/101/1.toc

Ineffective erythropoiesis caused by binucleated late-stage erythroblasts in mDia2 hematopoietic specific knockout mice Yang Mei, et al. http://www.haematologica.org/content/101/1/e1

Iron deficiency anemia treatment response to oral iron therapy: a pooled analysis of five randomized controlled trials Maureen M. Okam, et al. http://www.haematologica.org/content/101/1/e6

Combination of Tmprss6-ASO and the iron chelator deferiprone improves erythropoiesis and reduces iron overload in a mouse model of beta-thalassemia intermedia Carla Casu, et al. http://www.haematologica.org/content/101/1/e8

e12

Risk of recurrent venous thromboembolism on progestin-only contraception: a cohort study Emmanuelle Le Moigne, et al. http://www.haematologica.org/content/101/1/e12

e15

STAT1 activation in association with JAK2 exon 12 mutations Anna L. Godfrey, et al. http://www.haematologica.org/content/101/1/e15

e20

PAX5-ESRRB is a recurrent fusion gene in B-cell precursor pediatric acute lymphoblastic leukemia Yanara Marincevic-Zuniga, et al. http://www.haematologica.org/content/101/1/e20

e24

NUDT15 c.415C>T increases risk of 6-mercaptopurine induced myelosuppression during maintenance therapy in children with acute lymphoblastic leukemia Kanhatai Chiengthong, et al. http://www.haematologica.org/content/101/1/e24

Haematologica 2016; vol. 101 no. 1 - January 2016 http://www.haematologica.org/

haematologica Journal of the European Hematology Association Published by the Ferrata Storti Foundation e27

Phase II trial of dose-adjusted EPOCH in untreated systemic anaplastic large cell lymphoma Kieron Dunleavy, et al. http://www.haematologica.org/content/101/1/e27

e30

Presensitization to HY antigens in female donors prior to transplant is not associated with male recipient post-transplant HY antibody development nor with clinical outcomes Hideki Nakasone, et al. http://www.haematologica.org/content/101/1/e30

e34

Allogeneic hematopoietic cell transplantation in acute myeloid leukemia with normal karyotype and isolated Nucleophosmin-1 (NPM1) mutation: outcome strongly correlates with disease status Ali Bazarbachi, et al. http://www.haematologica.org/content/101/1/e34

Comments Comments are available online only at www.haematologica.org/content/101/1.toc

e38

Comment on: ‘Homozygous knockout of the piezo1 gene in the zebrafish is not associated with anemia’ Adèle Faucherre, et al. http://www.haematologica.org/content/101/1/e38

e39

Authors’ response to "Comment on: 'Homozygous knockout of the piezo1 gene in the zebrafish is not associated with anemia'" Boris E. Shmukler, et al. http://www.haematologica.org/content/101/1/e39

Haematologica 2016; vol. 101 no. 1 - January 2016 http://www.haematologica.org/

EDITORIALS Resolutions for the NewYear Jan Cools Editor-in-Chief E-mail: jan.cools@med.kuleuven.be

doi:10.3324/haematol.2015.139352

A focus on excellence With 68 hematology journals available today, and more being launched each year, Haematologica aspires to stand out from the crowd, not by publishing huge numbers of articles, but by focusing on scientific quality, open access and constructive peer review. Just as both the European Hematology Association and the Ferrata Storti Foundation aim to promote excellence in science, Haematologica supports their vision by providing an open access forum to publish high-quality research. Over the past years, the editorial team has used stricter selection criteria based on the novelty of the results and the importance of the data in their field. In 2015, we have accepted 142 original research articles for publication out of the 752 articles submitted (acceptance rate: 19%). Over the next years we will continue to carry out a strict selection procedure, while at the same time we aim to provide fast and fair peer review. On average, we communicate a first decision after 15 days, and we can reduce that to one week for selected articles under our fast-track review system. Haematologica will remain an open access journal, without the need to pay an extra fee, as some other journals do.

The 'Leaders in Hematology' review series Haematologica can count on a history going back 95 years and in 2015 we celebrated the publication of Volume 100 of the journal. And although it’s good to look back, we are also looking forward to a bright and exciting future. As part of our new activities, we have initiated the review series ‘Leaders in Hematology’. These authoritative review articles by leaders in the field provide a comprehensive overview of the latest developments in one particular aspect of hematology accompanied

by high-quality figures and illustrations. In 2015, we covered various topics, including β-thalassemia, T-cell and natural killer cell therapies, the bone marrow niche, and B-cell receptor signaling inhibitors in CLL.1-4 We will continue this review series in 2016, and we will also link it to a ‘Question & Answer’ session where you as a reader will have the opportunity to put questions to the authors. I hope the review articles published in Haematologica are of value to you and your colleagues and students, and that they help to address your need to stay upto-date on new developments in hematology. Feel free to share these articles, and use the figures in your presentations and teaching activities. Haematologica is an open access journal and we like to see that our articles are useful! We are determined to remain on this path of excellence, and to continue to publish top review articles and high-quality research articles in 2016. I wish you a successful year and hope you will consider publishing your work in Haematologica! Financial and other disclosures provided by the author using the ICMJE (www.icmje.org) Uniform Format for Disclosure of Competing Interests are available with the full text of this paper at www.haematologica.org.

References 1. Rivella S. β-thalassemias: paradigmatic diseases for scientific discoveries and development of innovative therapies. Haematologica. 2015;100(4):41830. 2. Cruz CR, Bollard CM. T-cell and natural killer cell therapies for hematologic malignancies after hematopoietic stem cell transplantation: enhancing the graft-versus-leukemia effect. Haematologica. 2015;100(6):709-19. 3. Krause DS, Scadden DT. A hostel for the hostile: the bone marrow niche in hematologic neoplasms. Haematologica. 2015;100(11):1376-87. 4. Wiestner A. The role of B-cell receptor inhibitors in the treatment of patients with chronic lymphocytic leukemia. Haematologica. 2015;100(12):1495-507.

Figure 1. An overview of the activities of Haematologica during 2015. haematologica | 2016; 101(1)

Editorials

Innovation in the field of thrombocytopenias: achievements since the beginning of the century and promises for the future Carlo L. Balduini and Patrizia Noris Department of Internal Medicine, IRCCS Policlinico San Matteo Foundation and University of Pavia, Pavia, Italy E-mail: c.balduini@smatteo.pv.it

doi:10.3324/haematol.2015.138149

nnovation is defined as â&#x20AC;&#x2DC;the act or process of introducing new ideas, devices, or methods.â&#x20AC;&#x2122; According to this definition, there is no doubt that, in the field of hematology, it is in the area of thrombocytopenias that most innovation has been seen. The new idea of inherited thrombocytopenias (ITs) that we developed in the last few years is one of the most remarkable innovations. Until the end of the last century, only a few diseases were known; all are characterized by severe bleeding diathesis. The most common and best known IT was the classical, biallelic form of BernardSoulier syndrome, a disorder affecting 1:1,000,000 subjects and characterized by recurrent, life-threatening hemorrhages.1 With the beginning of the new century, the techniques for gene sequencing have become gradually more and more efficient, faster and cheaper, and they have been used in large series of patients with unknown disorders. Reports of this experience have been collected thanks to wide international collaborative efforts that have identified new genes responsible for ITs and increased the number of well characterized diseases to 25. Interestingly, 14 novel disorders have been unveiled since 2010 (Figure 1).2 This explosion of knowledge has revealed that ITs are actually much less rare than previously thought, affecting at least 2.7:100,000 subjects.3 Moreover, we realized that the vast majority of patients have a bleeding risk that is mild or even not significantly different from that of healthy people. This is the reason why thrombocytopenia is often discovered incidentally in adult life and many patients are initially misdiagnosed with immune thrombocytopenia (ITP).4 Although clinically relevant bleeding affects only a minority of patients, ITs still have to be taken seriously because we now know that many affected subjects will acquire additional, life-threatening illnesses. This is the case of MYH9related disease, the most frequent form of IT. After MYH9 was identified as the causative gene in 2000, the creation of a large international registry (www.registromyh9.org) showed that patients usually do not bleed spontaneously, and that their main problem is the high risk during infancy or adult life of developing deafness, pre-senile cataracts and, most importantly, a proteinuric nephropathy that evolves into end stage renal failure and requires dialysis or kidney transplantation.5 Similarly, the identification in 2013 of ANKRD26 as the gene responsible for a mild form of IT stimulated a collaborative, international study revealing that this disorder exposes subjects to the risk of myelodysplastic syndromes and acute myeloid leukemia.6 A similar risk was shown to affect also subjects with familial platelet disorder with predisposition to acute myeloid leukemia due to mutations in RUNX1, while, quite recently, it has been suggested that subjects with ETV6-related thrombocytopenia are at risk of lymphoid malignancies.7 Thus, recognizing patients with these ITs predisposing to additional disorders is mandatory in order to personalize follow up 2

and be ready to give appropriate treatments if new illnesses develop. The use of innovative technologies and international collaborative efforts have provided the basis for this new insight into ITs. Another important innovation in the field of platelet disorders is represented by the development of techniques for in vitro culture of megakaryocytes (Mks) starting from hemopoietic progenitors or multipotent stem cells. The discovery of thrombopoietin (TPO) at the end of the last century was the turning point that made the achievement of this goal possible. Thereafter, further technical innovation resulted in the availability of devices for Mk culture that mimic bone marrow in terms of 3-dimensional structure, cellular composition, extracellular matrix and blood supply.8 In vitro culture of Mks made a big contribution to our understanding of the enigmatic biology of these cells. In brief, it has been shown that Mks originating from

Figure 1. The evolving view of inherited thrombocytopenias. Since the beginning of this century, a number of new inherited thrombocytopenias have been discovered and this advancement of knowledge changed our view of these diseases. A few patients have clinically relevant bleeding, but many subjects are at risk of developing additional, life-threatening disorders (names of diseases in bold). Of note, slightly less than 50% of patients remain without a diagnosis because they do not fit the criteria for any known disease, this indicating that a further effort should be made to improve the knowledge of these diseases. The reported figure is based on the database of our institution comprising 274 consecutive families and 566 patients. This case series does not include all known disorders, since many of them are exceedingly rare. BSS: Bernard-Soulier syndrome; WAS: Wiskott-Aldrich syndrome; GPS: gray platelet syndrome; ACTN1-RT: ACTN1-related thrombocytopenia; ITGA2B/B3-RT: ITGA2B/ITGB3-related thrombocytopenia; CYCS-RT: CYCS-related thrombocytopenia; FLNA-RT: FLNA-related thrombocytopenia; GATA1-RT: GATA1-related thrombocytopenia.

haematologica | 2016; 101(1)

Editorials

Figure 2. The increasing indications for thrombopoietin mimetics. Thrombopoietin mimetics (eltrombopag and romiplostim) were approved in 2008 for adults with chronic immune thrombocytopenia who did not respond to other treatments. Eltrombopag was also later approved for thrombocytopenia due to hepatitis C, refractory aplastic anemia and refractory chronic immune thrombocytopenia in children. Preliminary evidence indicated that thrombopoietin mimetics increase platelet count also in inherited thrombocytopenias (MYH9-related disease and Wiskott-Aldrich syndrome/X-linked thrombocytopenia) and myelodysplastic syndromes. Moreover, it is expected that these drugs might reduce the thrombocytopenia induced by non-myeloablative chemotherapy, but there are still no conclusive data available.

hemopoietic progenitors in the stem cell niche migrate to the vascular niche. Here they extend pro-platelets into the lumen of marrow sinusoids and release pre-platelets directly into the circulating blood, where platelets are finally formed.9 Moreover, the development of techniques for culturing Mks from circulating progenitors of patients with different forms of IT made it possible, not only to define the pathogenetic mechanisms of these disorders, but also to begin to decipher the sequence of molecular events that are required for effective platelet production.10 Platelets formed by cultured Mks resemble blood platelets in terms of morphology and function,8 and this originated the idea of using them for transfusion purposes. In principle, an amount of human MKs appropriate to the in vitro production of platelet concentrates might reasonably be obtained from CD34+ embryonic stem cells, CD34+ umbilical cord blood stem cells, and induced pluripotent stem cells.11 The most attractive source are human-induced pluripotent stem cells, in that they can be expanded and potentially maintained in culture indefinitely. Besides, they are generated from adult cells and therefore do not give rise to ethical concerns. An attractive feature of platelets derived from stem cell culture is the prospect of genetically manipulating progenitor cells to generate platelets devoid of HLA antigens.12 This would eliminate the risk of alloimmunization against HLA class I antigens and the resulting refractoriness to subsequent platelet transfusions, which is still a serious clinical problem. Clearly, further progress must be made for this platelet source to become a clinical reality. Particularly relevant challenges remain the need to generate huge numbers of stem cells, ensure that platelets produced with this methodology are safe, and verify whether they are hemostatically effective through clinical trials. There is still a long way to go but the arduous journey has begun. The discovery of TPO not only represented a turning point for in vitro culture of megakaryocytes, but also opened a new chapter in the treatment of various forms of thrombocytopenia. The initial attempts to stimulate in vivo haematologica | 2016; 101(1)

platelet formation with molecules structurally similar to native TPO gave promising results in patients with chemotherapy-induced thrombocytopenia. However, clinical trials were stopped in 2001 after a few healthy volunteers developed thrombocytopenia as a result of the formation of alloantibodies cross-reacting with endogenous TPO.13 Research then shifted to the development of novel agents with the capacity to stimulate the TPO receptor but with no structural homology with human TPO. Two of these molecules, eltrombopag and romiplostim, demonstrated good efficacy and safety in patients with chronic ITP, and were approved in 2008 and 2015 for adults and children, respectively, not responsive to traditional treatments. The experience gained so far indicates that TPO mimetics are the most effective drugs for ITP and questions the current classification of these molecules as second- or third-line agents. The interest in TPO mimetics continued to grow after the recent demonstration that many ITP patients who responded to this treatment can discontinue therapy without relapsing.14 Investigation into this unexpected phenomenon could provide new insights into the pathogenesis of ITP and the mechanisms of action of TPO mimetics. Efficacy of TPO mimetics is not limited to ITP (Figure 2). In 2012, eltrombopag was approved to allow patients with chronic hepatitis C and low platelet count to receive interferon and ribavirin therapy. Furthermore, in 2014 eltrombopag was approved for patients with severe aplastic anemia refractory to immunosuppressive therapy after evidence that it was able to produce a hematologic response in at least one cell lineage in 40% of patients, and a trilineage response in nearly half of them. It is interesting to note that, as in ITP, stable blood counts were often maintained after treatment discontinuation.15 This observation further supports the idea that not all the mechanisms of action of TPO mimetics have been identified. Finally, eltrombopag was proven to be effective in increasing platelet count in MYH9-related disease16 and WiskottAldrich syndrome/X-linked thrombocytopenia.17 3

Editorials

Interestingly, a few, severely thrombocytopenic patients have already successfully received this drug instead of platelet transfusion as preparation for surgery.18 An increase in platelet count has been obtained with the TPO mimetics also in myelodysplastic syndromes, but it remains unclear if these drugs favor the progression to leukemia in this clinical setting.19 On the basis of this success, there is no doubt that the introduction of TPO mimetics represents one the most relevant innovations in hematology since the beginning of this century. One final, important innovation in the area of thrombocytopenias has been the successful application of gene therapy to Wiskott-Aldrich syndrome, a form of IT associated with immunodeficiency, which is usually associated with death before adulthood due to hemorrhages or infections. Nine of 10 patients receiving lentiviral-transduced hematopoietic stem cells had a stable engraftment resulting in increased platelet count, less bleeding and improved immune function.20 After this proof of principle, it is desirable that this approach is made available also for a few other life-threatening forms of ITs to provide an alternative to those patients who cannot receive bone marrow transplantation due to the lack of compatible donors. In conclusion, exciting innovations in the area of thrombocytopenias characterized the beginning of this century, and this is the best promise we have that much more progress will be achieved in the next few years. Financial and other disclosures provided by the author using the ICMJE (www.icmje.org) Uniform Format for Disclosure of Competing Interests are available with the full text of this paper at www.haematologica.org.

References 1. Lรณpez JA, Andrews RK, Afshar-Kharghan V, Berndt MC. BernardSoulier syndrome. Blood. 1998;91(12):4397-4418. 2. Savoia A. Molecular basis of inherited thrombocytopenias. Clin Genet.

2015 May 7. [Epub ahead of print] 3. Balduini CL, Pecci A, Noris P. Inherited thrombocytopenias: the evolving spectrum. Hamostaseologie. 2012;32(4):259-270. 4. Balduini CL, Savoia A, Seri M. Inherited thrombocytopenias frequently diagnosed in adults. J Thromb Haemost. 2013;11(6):1006-1019. 5. Pecci A, Klersy C, Gresele P, et al. MYH9-related disease: a novel prognostic model to predict the clinical evolution of the disease based on genotype-phenotype correlations. Hum Mutat. 2014;35(2):236-247. 6. Noris P, Favier R, Alessi MC, et al. ANKRD26-related thrombocytopenia and myeloid malignancies. Blood. 2013;122(11):1987-1989. 7. Noetzli L, Lo RW, Lee-Sherick AB, et al. Germline mutations in ETV6 are associated with thrombocytopenia, red cell macrocytosis and predisposition to lymphoblastic leukemia. Nat Genet. 2015;47(5):535-538. 8. Di Buduo CA, Wray LS, Tozzi L, et al. Programmable 3D silk bone marrow niche for platelet generation ex vivo and modeling of megakaryopoiesis pathologies. Blood. 2015;125(14):2254-2264. 9. Kaushansky K. Thrombopoiesis. Semin Hematol. 2015;52(1):4-11. 10. Pecci A, Balduini CL. Lessons in platelet production from inherited thrombocytopenias. Br J Haematol. 2014;165(2):179-192 11. Thon JN, Medvetz DA, Karlsson SM, Italiano JE Jr. Road blocks in making platelets for transfusion. J Thromb Haemost. 2015;13 Suppl 1:S55-62. 12. Feng Q, Shabrani N, Thon JN, et al. Scalable generation of universal platelets from human induced pluripotent stem cells. Stem Cell Reports. 2014;3(5):817-831. 13. Li J, Yang C, Xia Y, et al. Thrombocytopenia caused by the development of antibodies to thrombopoietin. Blood. 2001;98(12):3241-3248. 14. Carpenedo M, Cantoni S, Coccini V, Fedele M, Morra E, Pogliani EM. Feasibility of romiplostim discontinuation in adult thrombopoietinreceptor agonist responsive patients with primary immune thrombocytopenia: an observational retrospective report in real life clinical practice. Hematol Rep. 2015;7(1):5673. 15. McCormack PL. Eltrombopag: a review of its use in patients with severe aplastic anaemia. Drugs. 2015;75(5):525-31. 16. Pecci A, Gresele P, Klersy C, et al. Eltrombopag for the treatment of the inherited thrombocytopenia deriving from MYH9 mutations. Blood. 2010;116(26):5832-5837. 17. Gerrits AJ, Leven EA, Frelinger AL 3rd, et al. Effects of eltrombopag on platelet count and platelet activation in Wiskott-Aldrich syndrome/Xlinked thrombocytopenia. Blood. 2015;126(11):1367-1378. 18. Favier R, Feriel J, Favier M, Denoyelle F, Martignetti JA. First successful use of eltrombopag before surgery in a child with MYH9-related thrombocytopenia. Pediatrics. 2013;132(3):e793-795. 19. Prica A, Sholzberg M, Buckstein R. Safety and efficacy of thrombopoietin-receptor agonists in myelodysplastic syndromes: a systematic review and meta-analysis of randomized controlled trials. Br J Haematol. 2014;167(5):626-638. 20. Hacein-Bey Abina S, Gaspar HB, Blondeau J, et al. Outcomes following gene therapy in patients with severe Wiskott-Aldrich syndrome. JAMA. 2015;313(15):1550-1563.

haematologica | 2016; 101(1)

REVIEW ARTICLE

Leaders in Hematology review series

Humanized hemato-lymphoid system mice EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Alexandre P.A. Theocharides,1 Anthony Rongvaux,2 Kristin Fritsch,1 Richard A. Flavell,2 and Markus G. Manz1

Hematology, University Hospital Zurich and University of Zurich, 8091 Zurich, Switzerland; and 2Department of Immunobiology and Howard Hughes Medical Institute, Yale University School of Medicine, New Haven, CT, USA 1

Haematologica 2016 Volume 101(1):5-19

ABSTRACT

Correspondence: markus.manz@usz.ch

Received: 21/9/2015. Accepted: 23/10/2015.

doi:10.3324/haematol.2014.115212

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/1/5

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

haematologica | 2016; 101(1)

ver the last decades, incrementally improved xenograft mouse models, supporting the engraftment and development of a human hemato-lymphoid system, have been developed and now represent an important research tool in the field. The most significant contributions made by means of humanized mice are the identification of normal and leukemic hematopoietic stem cells, the characterization of the human hematopoietic hierarchy, and their use as preclinical therapy models for malignant hematopoietic disorders. Successful xenotransplantation depends on three major factors: tolerance by the mouse host, correct spatial location, and appropriately cross-reactive support and interaction factors such as cytokines and major histocompatibility complex molecules. Each of these can be modified. Experimental approaches include the genetic modification of mice to faithfully express human support factors as non-cross-reactive cytokines, to create free niche space, the co-transplantation of human mesenchymal stem cells, the implantation of humanized ossicles or other stroma, and the implantation of human thymic tissue. Besides the source of hematopoietic cells, the conditioning regimen and the route of transplantation also significantly affect human hematopoietic development in vivo. We review here the achievements, most recent developments, and the remaining challenges in the generation of pre-clinicallypredictive systems for human hematology and immunology, closely resembling the human situation in a xenogeneic mouse environment.

Introduction Over the last decades humanized mouse models have become an important tool in human hematopoiesis research. The information extracted from experimentation on primary human cells that have been maintained or generated in vivo in a mouse host can be used to understand the physiology and pathophysiology of human hematopoiesis (Figure 1). While the characterization of hematopoietic stem cells (HSC) by multicolor flow cytometry and in depth genetic analysis improves steadily, the xenograft model remains the only broadly accessible assay to functionally define HSC and their malignant counterparts, leukemic stem cells (LSC). The significant differences in cell surface marker expression between mouse and human stem and progenitor cells (HSPC) highlights the importance of assessing the development of human hematopoietic cells in vivo and shows that both murine and humanized models are complementary. It is, therefore, essential to assess the advantages and limitations of each experimental system in the light of the scientific and clinical question being addressed. Although humanized mice have contributed extensively to the characterization of the physiology of human hematopoiesis, one of the biggest contributions of the model lies in the understanding of malignant hematopoiesis and the leukemic hierarchy. Today, primary human leukemia research relies heavily on the leukemia 5

A.P.A. Theocharides et al.

xenograft model because of several key advantages over murine models. Humanized mouse models, unlike genetically modified mouse models of leukemia development, can reflect the disease heterogeneity observed in patients. Disease heterogeneity can be assessed at the clonal level and has led to the identification of the sub-clonal architecture of leukemia. Humanized mouse models are also used extensively as preclinical models to test novel therapeutic approaches. This allows assessment of therapeutic response in a heterogeneous disease such as leukemia. However, several limitations remain; the composition of the hematopoietic system does not fully recapitulate human hematopoiesis, and the long-term maintenance of human cells is limited. In addition, the functionality of human hematopoietic cells (both healthy and diseased) may be altered in a mouse environment, possibly because of the lack of cross-reactivity of specific factors between the mouse host and the human graft. Finally, less aggressive HSC/HSPC neoplasms do not develop efficiently in mouse xenograft models. This review focuses on the most recent developments in humanized mouse models for healthy hematopoiesis and primary hematopoietic malignancies. We highlight the differences and advantages of next-generation recipient mouse strains over previous models and their potential application in future research.

Key requirements for successful xenotransplantation models Humanized mice are generated by transplantation of human hematopoietic cells into recipient mice.1,2 An ideal model would be a humanized mouse in which the human graft fully replaces the endogenous mouse hematopoietic and immune systems, in both space and function. To

achieve, this goal, three basic requirements need to be met (Figure 2). First, the recipient strain of mouse needs to be deficient in both its adaptive and innate immune compartments to prevent the xeno-rejection of the human graft by the endogenous mouse immune system. Second, a niche

Figure 1. The principle of humanized mouse models. Normal (healthy) or neoplastic hematopoietic cells are transplanted into immuno-compromised mice to develop humanized mouse models. The models help to learn and understand the physiology and pathophysiology of human hematopoiesis. The knowledge gained with these models can then be translated to humans.

Figure 2. Prerequisites for successful xenotransplantion of human hematopoiesis. The (mouse) host must be tolerant for the transplanted human hematopoietic cells and provide a supportive niche, including cross-reactive growth factors, cytokines and MHC molecules. Novel models carry multiple genetic modifications in the mouse host to meet these three criteria and support successful human hematopoietic engraftment.

haematologica | 2016; 101(1)

Humanized hemato-lymphoid system mice

Table 1. History of the development of humanized mouse models.

Year

Strain

Acronym

Characteristics

References

1988 1988-1992 1995-1996

Nude-beige-xid CB17-Scid NOD-Scid

bg/nu/xid SCID

3 4, 5, 6 7, 8, 9

2002-2005

NOD-Scid Il2rγ -/-

NOG/NSG

BALB/c-Rag2-/-Il2rγ -/Tg(hSIRPA)Rag2-/-Il2rγ -/-

BRG SRG

T and NK cell deficiency T and B cell deficiency T and B cell deficiency Phagocytic tolerance T, B and NK cell deficiency Phagocytic tolerance T, B and NK cell deficiency T, B and NK cell deficiency Phagocytic tolerance

2004 2011

needs to be generated to accommodate the human cells. Third, the human cells need to obtain supportive crossreactive signals from the created niche.

Mouse immune-deficiency The development of mouse strains that support human hematopoiesis in vivo started in the late 1980s with bg/nu/xid (beige-nude-xid) mice, which are athymic and have a reduced number of natural killer (NK) and so-called lymphokine activated killer cells (Table 1).3 This was followed by the generation of CB17-SCID mice, which are deficient in mouse T and B cell development.4-6 Today, immunodeficient mice lacking T lymphocytes, B lymphocytes and NK cells are used as recipients to prevent the immune rejection of the human graft. Defective V(D)J recombination results in the absence of mature B and T cells and is found in mice harboring the severe combined immunodeficiency (SCID) mutation in the gene encoding the DNA-PKcs enzyme or in mice deficient for either of the two recombination activating genes (RAG1 or RAG2).514 Deficiency of NK cells is achieved by disrupting the signaling downstream of interleukin (IL)-15, a critical cytokine required for NK-cell development and maturation. The common gamma chain of cytokine receptors (IL2Rγ, CD122) is shared by multiple members of the IL-2 family of cytokines, including IL-15.15 Hence, mutation or deficiency in the Il2rγ gene results in the complete absence of NK cells in the mouse host.12,13,16-18

Macrophage tolerance The development of mice with phagocytic tolerance against human cells was a major improvement and was accomplished by the generation of SCID mice on a nonobese diabetic (NOD)-background.7-9 The CD47-SIRPα axis is a signaling pathway by which endogenous CD47expressing cells, i.e. most, if not all, healthy cells of the body, prevent their phagocytosis by providing an inhibitory signal (the so-called “don’t eat me” signal) to macrophages.19 Human CD47 and mouse SIRPα do not cross-react in most mouse strains and consequently mouse macrophages engulf the transplanted human cells.20 The Sirpa gene is highly polymorphic between different mouse strains. In particular, the Sirpa allele of NOD mice encodes for a protein with a 20 amino acid difference and a distinct glycosylation pattern in the extracellular domain of SIRPα compared to the C57Bl/6 allele. In fact, the NOD allele of Sirpa is very similar to the human SIRPA allele and NODSIRPα binds to human CD47, whereas C57Bl/6-SIRPα does not.20 As a consequence, immunodeficient mice on haematologica | 2016; 101(1)

16, 17, 18 13 22

the NOD background are significantly more supportive of human hematopoietic development than any other mouse strain in which SIRPα does not cross-react with human CD47.20 A method used to induce mouse-to-human phagocytic tolerance consists in introducing the human SIRPA gene into the mouse genome of “non-NOD” mice, either by BAC-transgenesis or by knock-in humanization of the mouse gene.21,22 Alternatively, murine CD47 expressed in transplanted human cells can induce tolerance.23,24 NOD/ShiLtJ-Prkdcscid (NOD-SCID), NOD.CgPrkdcscidIl2rgtm1wjl/SzJ (NSG), and NOD.Cg-Prkdcscid Il2rgtm1Sug/Jic (NOG) mice, the three most commonly used recipient mouse strains for humanized mouse models, combine murine T-, B-lymphocyte and NK-cell deficiency (NOD-SCID mice can be depleted of NK cells by anti-CD122 antibody treatment) with macrophage tolerance towards human cells (Table 1).7-9,16-18 Furthermore the lack of the IL-2Rγ limits the spontaneous development of murine thymoma, a phenomenon commonly observed in NOD-SCID mice.25

Niche space In addition to immunodeficiency, the ablation of mouse cells can create open niches in which transplanted human cells can home and develop. This was recently demonstrated in mice harboring mutations in the gene encoding c-kit (CD117), which is important for HSC maintenance and function.26,27 In c-kit mutant mice, mouse HSC are reduced in number and function (see below). The reduction in mouse HSC confers a competitive advantage to human HSC homing, maintenance and differentiation, and eliminates the need for irradiation pre-conditioning of the host mice prior to transplantation. A similar observation was made in mice with knock-in replacement of mouse cytokine-encoding genes by the human ortholog.2830 In particular, the knock-in humanization of THPO, encoding thrombopoietin (TPO) and CSF2 [encoding granulocyte-macrophage colony-stimulating factor (GMCSF)] results in deficiency of the respective mouse cytokines and functional impairment of mouse HSC and alveolar macrophages, respectively.

Cytokine support for human hematopoietic development and immune function While human HSPC development and differentiation are well recapitulated in NOD-SCID and NSG/NOG mice, several developmental and functional defects remain.31-33 For example, the differentiation of human HSC into func7

A.P.A. Theocharides et al.

tional myeloid cells and NK cells is limited,34,35 B cells do not undergo sufficient maturation to become memory and antibody-producing cells,36,37 and, unless a fetal thymus is co-transplanted, the selection and maturation of T cells in adult animals is disturbed due to the lack of thymic support, and the ratio of B- to T-lymphocytes is strongly biased towards B-cell differentiation and does not mimic the human system.38,39 It has been hypothesized that, despite the issue of MHC restriction, the limitations are due to a lack of cross-reactivity of mouse cytokines, and possibly chemotactic factors.1,32 To circumvent some of these limitations, several methods have been developed to deliver human cytokines in vivo (Figure 2). Exogenous recombinant cytokines were first injected into SCID mice transplanted with human bone marrow in the early 1990s.4 However although this approach is simple, it is expensive and not suited for long-term experiments. The hydrodynamic injection of cytokine-encoding plasmids, which results in transient, systemic expression of candi-

date proteins, is more convenient and results in significant support of human NK cells and monocytes when human IL-15 or macrophage colony-stimulating factor (M-CSF) is expressed, respectively.40 Similarly, the transgenic expression of cytokine-encoding genes under the control of a strong constitutive promoter provides support for human hematopoietic cell development.41-44 However, these approaches all result in the expression of systemic supraphysiological concentrations of human cytokines. In the case of transgenic mice expressing human IL-3, GM-CSF and stem cell factor (SCF) this induces the mobilization of HSPC and limits long-term engraftment.44 This example highlights the importance of achieving physiological expression of human cytokines and other factors, which can be achieved either by BAC-transgenesis or by knockin replacement of the mouse gene.29 We have reported individual gene knock-in humanization for four cytokines (Table 2): TPO, IL-3/GM-CSF (the two contiguous genes were replaced simultaneously) and M-CSF, conferring sup-

Table 2. Comparison of next-generation humanized mouse models in their support of normal hematopoiesis.

Strain Background MISTRG

BRG

TPO

BRG

Genetic modification

Conditioning Cells Compared regimen injected to

Human Sublethal FL, knockin: M-CSF, irradiation intrahepatic IL-3, GM-CSF, TPO or no irradiation BAC transgene: human SIRPÎą

Key features

Functional assays

BRG NSG

- Median human BM CD45+ (85%) similar to NSG (80%) at 10-12 weeks - Increased human myeloid (CD33+) and monocytic (CD14+/CD16+) differentiation in the PB and BM - Supports human macrophage and NK-cell development

- Monocytes (cytokines, phagocytosis) - Human response to Listeria, Influenza infection - Human NK-cellmediated lysis - Human tumor-associated macrophages - CD34+ cell secondary transplantation

- Median human BM CD45+ (80%) significantly higher compared to BRG (45%) at 3-4 months - Increased myeloid (CD33+CD66+) differentiation - Increase in human CD34+CD38-, increase in human CD34+CD38-CD90+CD45RA- Median human BM CD45+ in - Secondary non-irradiated BRgWV: 80%, transplantation similar to irradiated NSG at 17-24 weeks - Increased myelo-monocytic differentiation - Increased neutrophils and neutrophil precursors - Significant increase in human CD34+CD38- and CD90+CD45RA(frequency equivalent to human BM) - Similar results in neonates and adults - Supports human lymphopoiesis, block in B-cell differentiation

Human TPO knockin

Sublethal irradiation

FL and CB

BRG

BRgWv BALBc (also some neonates)

Mouse c-kit mutation Kit Wv

No irradiation

NSG (irradiated)

NBSGW

Mouse c-kit mutations Kit W41

+/- irradiation

NSG

NSG irradiated - Median human PB CD45+ similar in NSG non-irradiated NBSGW and irradiated non-irradiated NSG at 4 (30-40%), 8 (50-60%), and 12 (50-60%) weeks - Median human BM CD45+ significantly higher (97%) in non-irradiated NBSGW compared to non-irradiated NSG (30%) at 12 weeks - Signifcantly more human CD34+, CD33+, CD45-GlyA+

Reference(s) 46

- Secondary transplantation

continued on the next page

haematologica | 2016; 101(1)

Humanized hemato-lymphoid system mice

port for human HSC, alveolar macrophages and monocytes/macrophages, respectively.29,30,45 Finally, hematopoiesis is a complex process in which successive developmental steps are regulated by multiple cytokines.1,31 To optimize multi-lineage human hematopoiesis, it is, therefore, necessary to provide multiple cytokines that support the successive differentiation steps, from the HSC to terminally differentiated mature cells, and their subsequent maintenance and function.40,46 This principle is best illustrated in the MISTRG mouse, in which four cytokines (M-CSF, IL-3, GM-CSF and TPO) are genetically “humanized” in one mouse strain.46 As a consequence, MISTRG mice strongly support the development of human myelo-monocytic cells and show a strong human innate immune response in response to viral and bacterial infections. An additional and indirect consequence of improved myelopoiesis in MISTRG recipients is the expression of the IL15/IL15Rα complex on human monocytes, which in turn supports the development and function of human NK cells.46 The MISTRG model thus demonstrates the necessity of combining multiple factors that synergize, directly and indirectly, to support the development of a functional human immune system.

Generation of a human microenvironment in humanized mice To create a microenvironment that is more similar to the human bone marrow (BM) and thymic niche, additional human tissue can be co-transplanted along with the human hematopoietic cells.

Bone marrow microenvironment Human hematopoiesis is regulated by a specialized microenvironment, the BM niche. Besides hematopoietic cells, the BM niche is composed of adipocytes, osteoblasts/osteoclasts, megakaryocytes, endothelial cells, macrophages and Schwann cells, as well as specialized stromal cells including Nestin-positive mesenchymal stromal cells (MSC), CXCL12-abundant reticular (CAR) cells and leptin receptor-positive cells.47 The BM niche cells provide survival and maintenance signals to HSPC and, in hematologic cancers, to leukemia-initiating cells and malignant plasma cells, for example. Friedenstein and co-workers showed that MSC can be separated from other cells in the BM by their tendency to adhere to tissue culture plastic.48 Furthermore they found that MSC in culture can differentiate into several lineages such as osteoblasts, chondroblasts and adipocytes.49-51 As

continued from the previous page

Strain Background

Genetic modification

Conditioning regimen

hSCFTgNSG NSG

Human transgene: Membranebound SCF

Sublethal irradiation

hu-mSCF, NSG Human transgene: +/- irradiation neonates Membrane(also some adults) bound SCF

Cells Compared injected to CB

NSG

- Median human BM CD45+ significantly higher (97%) compared to NSG (63%) at 8-35 weeks - Increased myeloid (CD33+) differentiation - Supports human mast cell development

None

- Rejection of human skin graft

- Rescue of pulmonary alveolar proteinosis Lung influenza infection

- Inducible cutaneous anaphylaxis

- Human cytokine response to LPS - Human phagocytosis - Enhanced chemotaxis

NSG irradiated - Non-irradiated: trend towards higher NSG human BM CD45+ (57%) compared non-irradiated to NSG (37%) at 12 weeks - Irradiated: trend towards higher human BM CD45+ (72%) compared to NSG (51%) at 12 weeks - Increased support of CD71+/CD235a+ development FL and CB BRG - Deficiency in mouse lung alveolar macrophages - Development and function of human lung alveolar macrophages

Human IL3, GM-CSF knockin

Sublethal irradiation

IL-3/GM-CSF NOG

Human transgene: IL3, GM-CSF

No irradiation

NOG

CSF1 neonates

Human CSF1 knockin

Sublethal irradiation

BRG

BAC transgene: human SIRPα

Sublethal irradiation

BRG NSG

hSIRPαTg BRG (neonates)

Functional assays

IL-3/GM-CSF BRG

BRG

Key features

- BM: 3-4x more human CD34+CD38- Increased myeloid (CD33+) differentiation - Increased support of dendritic cell, basophil and mast cell differentiation. - Increased myelo-monocytic differentiation

Reference(s)

- Median human BM CD45+ significantly - Improved humoral higher in hSIRPαTg (60%) compared immune response to to BRG (35%) at 12-14 weeks ovalbumin - Median human BM CD45+ similar in hSIRPαTg and NSG - Increased human CD34+, CD34+CD38engraftment - Similar human CD45+ engraftment - Similar human CD34+, CD34+CD38- engraftment

FL: fetal liver; CB: umbilical cord blood; BM: bone marrow; PB: peripheral blood; GMP: granulocyte macrophage progenitor; CMP: common myeloid progenitor; LPS: lipopolysaccharide.

haematologica | 2016; 101(1)

A.P.A. Theocharides et al.

MSC are able to differentiate into chondrocytes through a process that is called endochondral ossification, they can be used as a source to form a BM microenvironment in vivo (Figure 3).52 Additionally, in vitro maturation (chondrogenic differentiation) of human BM-derived MSC results in accelerated formation of larger, hematopoiesis-supporting bone tissue.53 Furthermore, several studies have shown that upon subcutaneous implantation of human MSC into immune-deficient mice a supportive BM cavity can be formed through a vascularized cartilage intermediate which is replaced by hematopoietic tissue and bone and can attract and support murine hematopoiesis.54-57 Reinisch et al. showed that only BM-derived MSC are able to form a BM cavity through a vascularized cartilage intermediate.58 Holzapfel et al. developed an engineered humanized bone construct by seeding MSC on a tubular medical-grade polycaprolactone scaffold followed by culture in a rotating bioreactor before implantation into NSG mice.59 By inducing differentiation of MSC in vitro they were able to recapitulate morphological features and biological functions of the human HSC niche. Several studies now suggest that the BM niche can actively participate in the initiation of leukemia and consequently novel therapeutic approaches aim at targeting

not only the leukemia clone, but also the supportive microenvironment.60,61 Indeed, Chen et al. developed an ectopic human BM niche that supports human normal and leukemic engraftment (Table 3).62 In this model, knockdown of hypoxia-inducible factor 1Îą in human MSC leads to a significant reduction of human leukemic engraftment and demonstrates that the human BM niche can be genetically manipulated. In a recent myelodysplastic syndrome (MDS) model, patient-derived MSC were co-injected along with CD34+ cells into NSG mice and NSG mice expressing human SCF, GM-CSF and IL-3 (NSG-SGM3).63 Interestingly, myelodysplasia was only observed in mice co-injected with human MSC in this study. Moreover, Groen et al. demonstrated the engraftment of primary human multiple myeloma cells into a scaffold-based ectopic BM niche.64 By using luciferase gene marking of multiple myeloma cells they were able to visualize myeloma growth and therapeutic response. From these interesting observations it can be inferred that stromal cells, through their interaction with diseased hematopoietic cells, contribute actively to disease development and that a functional human BM niche may be necessary to fully recapitulate human disease in vivo.

Figure 3. Engineering of heterotopic human niches in a xenograft mouse model. Human adult BM-derived mesenchymal stromal cells (MSC) are ex vivo-differentiated into hypertrophic cartilage, and implanted into immune-compromised mice. CD34+ human umbilical cord blood cells are transplanted into sublethally irradiated mice with implanted ossicles. Upon in vivo implantation the hypertrophic cartilage develops into a fully mature bone organ through endochondral ossification.

haematologica | 2016; 101(1)

Humanized hemato-lymphoid system mice

Sources of human hematopoietic cells, routes of transplantation and mouse age at transplantation

Thymic microenvironment The idea behind co-transplanting human thymus is to provide a human microenvironment that supports the development of human T cells, as well as their selection on human MHC molecules.5 In the so-called “BLT” (bone marrow-liver-thymus) model, small fragments of human fetal liver and thymus are co-transplanted under the mouse kidney capsule.65,66 Mice transplanted according to the BLT protocol sustain efficient human thymic lymphopoiesis, and T cells are the main component of the human graft.39 Although the BLT model was initially reported as a system in which human adaptive immune responses would be improved compared to those of other models, the evidence for improved adaptive immunity in BLT mice, in comparison to transplantation of human CD34+ cells without cotransplantation of thymic tissue, is still scant.67,68 Recently, it was demonstrated that the transplantation of human embryonic stem cell-derived thymic epithelial progenitors into thymus-deficient nude mice supports the development of functional human T cells.69 Neither approach, however, is accessible to most investigators due to the limited availability of embryonic and fetal tissues. Moreover, the biology of human thymus educated T cells and their reactivity against the xenogeneic environment needs to be further characterized. In addition, serum immunoglobulin levels and antigen-specific human antibody responses (particularly IgG responses) in BLT mice remain several orders of magnitude lower than in humans.37,39,68 Nevertheless, despite the remaining functional deficiencies, BLT mice are particularly useful for studies of human immunodeficiency virus infection because of the high frequency of human T cells in lymphoid and mucosal tissues.70-72

Human HSPC are the standard source of cells for transplantation into humanized mice (Figure 4). HSC can differentiate into all hematopoietic lineages, self-renew and sustain hematopoiesis for the entire life of the organism. The frequency of HSC is highest in the CD34+ cell population, although a lower number of HSC might be found in the CD34– fraction as well.31,73-75 CD34+ cells can be isolated from different sources: fetal liver, cord blood, adult BM or G-CSF-mobilized peripheral blood cells. The frequency of HSC among CD34+ cells declines from fetal life to adulthood.46 The number of cells that needs to be injected and the engraftment that can be expected are, therefore, highly variable depending on the source of CD34+ cells used. CD34+ cells are transplanted into adult mice by either intravenous or intrafemoral injection (Figure 4). Transplantation into newborn mice (up to 3 days after birth) results in efficient engraftment and multilineage human hematopoietic differentiation, including T-cell development.13 Human CD34+ cells can be engrafted into newborn mice by intravenous, intracardiac or intrahepatic injection, with the liver being a natural site of hematopoiesis during fetal development and the first few days of the life. It has been demonstrated that female NSG mice are more supportive of human engraftment than male NSG mice when HSC are transplanted at a limiting dose.76 It is, therefore, crucial that the appropriate transplantation protocol is chosen in order to enable the experimental question to be answered.

Table 3. Comparison of ectopic human BM niches in xenograft models.

Source

Matrix Number of cells seeded

Culture condition

Additions

Mouse Irradiation model

Human cells transplanted

Number of transplanted cells

Timepoint Functional Reference(s) of readout transplantation

1.5x106 MSC

NSG

Sublethal

MNC from CB, human acute myeloid leukemia cell line MOLM13

2x106 MNC or 2x106 MOLM13

After ossicle implantation

Human 4 Hybrid MSC scaffold from BM consisting of biphasic calcium phosphate particles Human Matrigel2x106 MSC equivalent MSC from BM, matrix WAT, UCB, Skin Human Tubular 3x105 MSC medicalMSC from BM grade polycaprolactone scaffold

BRG

CD34+ from CB

1-5x105 CD34+ or 1-5x106 MNC

8 weeks after ossicle implantation

NSG

Sublethal

CD34+ from CB

1x104 to 2x105

NSG

Sublethal

CD34+ from BM

1.3x105 CD34+ and

Human Matrigel MSC from BM

MSC expanded, Human mixed with BM-derived Matrigel, MSC injected mixed with subcutaneously 1.5x106 ECFC 5 2x10 MSC expanded, No MSC loaded on scaffold, 7 days in vitro culture, implanted subcutaneously

MSCs expanded, mixed with Matrigel, injected subcutaneously 4 weeks static culture followed by 4 weeks dynamic culture in a bi-axial rotating bioreactor

8-10 weeks Serial before transplantation ossicle transplantation 10 weeks after No ossicle implantation 1x106 CD34-

BM: bone marrow; ECFC: endothelial colony-forming cells; MNC: mononuclear cells; MSC: mesenchymal stroma cells; CB: umbilical cord blood; WAT: white adipocyte tissue.

haematologica | 2016; 101(1)

A.P.A. Theocharides et al.

Figure 4. Parameters that influence the engraftment of human hematopoietic cells in vivo. Depending on the experimental conditions, different strains of mice, methods of pre-conditioning and routes of injection can be used to transplant human hematopoietic cells obtained from various sources.

Normal human hematopoiesis in humanized mice The development of human hematopoiesis in mice was pioneered and recently reviewed by John Dick’s group.31 HSC are very rare with an estimated frequency of 1 in 106 human BM cells.77 To be defined a HSC, the cell must have the potential to self-renew and differentiate into all hematopoietic lineages. Most investigators assess these properties 12-16 weeks after transplantation in mouse xenografts. The time-point of lineage engraftment assessment is particularly critical since the engraftment of certain lineages is transient and limited to a specific time window. In this regard engraftment of myelo-eythroid cells is observed early after transplantation, while the engraftment of T cells can increase significantly beyond 12 weeks.78 The CD34+ population remains a heterogeneous mixture of cells with a minority of cells being bona fide HSC. Considerable effort has, therefore, been taken to identify markers that may reliably help to isolate HSC. Most publications define the human HSC population as CD34+CD38–Thy1+CD45RA–.31,79 The closest progenitors to HSC are multipotent progenitors (MPP) and it was suggested that these intermediates may be distinguished from HSC by the lack of Thy1 expression.80 However some 12

MPP retain serial repopulation capacity. Additional markers are, therefore, needed to discriminate MPP from HSC. One such marker is CD49f, as demonstrated by the lack of long-term engraftment of CD49f – cells.81 Interestingly, single-cell transplants of CD49f + cells generated long-term engraftment independently of Thy1 expression. Thus, while HSC express CD49f, MPP should be defined as CD49f – cells. Importantly, gene expression analysis of HSC and MPP based on CD49f expression revealed distinct profiles of these two populations.

Hematopoietic stem and progenitor cell engraftment in next-generation humanized mice The engraftment of HSC is a prerequisite to ensure multilineage engraftment and one approach to increase human HSC engraftment is to reduce the mouse HSC compartment. This has been successfully achieved in NSG and BALB/cJ mice (NBSGW and BRgWv mice, respectively) that carry mutations in c-Kit (Table 2).26,27 Strikingly, NBSGW and BRgWv mice are highly supportive of human hematopoiesis even without irradiation. The level of human engraftment at non-limiting doses in non-irradiated NBSGW and BRgWv mice is comparable to that in irradiated NSG mice and dramatically higher in NBSGW than haematologica | 2016; 101(1)

Humanized hemato-lymphoid system mice

Figure 5. The expression of certain surface markers is significantly increased on LSC and allows these cells to be distinguished from their normal counterparts. LSC express high levels of CD47 and evade innate immune attacks by macrophages. The expression of TIM-3 has only been reported on LSC, but not on HSC. The expression of CD44 and the v6 isoform of CD44 are increased on LSC. These and other markers such as CLL-1, CD96 and CD123 allow LSC to be distinguished and targeted by monoclonal antibodies.

in non-irradiated NSG mice. In BRgWv mice, the frequency of human bona fida HSC (CD45RA–CD90+) 17-24 weeks after transplantation is similar to that in human BM. As the HSPC pool in next-generation humanized mice has not so far been characterized extensively, it will be interesting to see how these compare to BRgWv mice in this respect. Clearly, TPO knock-in mice demonstrate enhanced HSC maintenance and promote granulocytemacrophage progenitor (GMP) differentiation, and the increased support of myelopoiesis in TPO and MISTRG mice most likely results from an increased GMP pool (see below).

Myeloid development in humanized mice Human peripheral blood contains a significant proportion of myeloid cells, a situation not reflected in xenografted NOD-SCID mice and NSG mice.1,82 Several approaches have, therefore, been taken to develop models that more faithfully mimic human hematopoiesis with a focus on myeloid lineage development in vivo.1 Very early in the development of humanized mouse models sublethally irradiated SCID mice were injected with the “myeloid” cytokines IL-3, GM-CSF and SCF to assess myeloid development in vivo.3,4 After 4 months myeloid cells were still detectable in the host, demonstrating the multilineage potential of human HSC. Several approaches have been taken to develop mouse models that promote myeloid differentiation of human HSC (Table 2). Mice that express the haematologica | 2016; 101(1)

classic human “myeloid” cytokines IL-3, GM-CSF or human M-CSF (CSF1) show increased support of myeloid differentiation.42,45 This confirms that the introduction of human genes into mice leads to the production of functional proteins. Interestingly, however, the support of myeloid differentiation is not confined to the presence of classical myeloid cytokines. In fact, most next-generation mouse models have been reported to promote improved human myeloid differentiation.26-28,46 As described above, increased support of myelopoiesis is observed in mice that express human TPO.28,46 Furthermore, the number of murine platelets is significantly reduced in TPO mice compared to control mice. Nevertheless, no significant increase in human megakaryocytes or platelets is observed, suggesting that human TPO is insufficient to promote full human megakaryopoiesis in vivo. Finally, NBSGW mice seem to be more supportive of human erythropoiesis. In summary, next-generation humanized mice promote myeloid differentiation, in part likely because they favor the development of myeloid progenitors.

Development of functional hematopoietic cells in mice One particular focus in xenograft models has been the assessment of functional properties of engrafted human cells. MISTRG mice support development of functional monocytes characterized by the production of human tumor necrosis factor-α and IL-6 as a response to stimulation with lipopolysaccharides or pathogens, and by phagocytic activity against E. coli.46 MISTRG mice also 13

A.P.A. Theocharides et al.

support the development of functional NK cells and the development of CD163+ tumor-infiltrating macrophages in a melanoma model. Strikingly, tumor growth was significantly higher in MISTRG mice than in NSG mice, which were both engrafted with human hematopoiesis, probably as a result of increased numbers of human tumorassociated macrophages, a process reversible by anti-vascular endothelial growth factor antibody therapy. A human cytokine response was also observed in CSF1 mice injected with lipopolysaccharide, and human monocytic cells purified from mice showed increased phagocytic activity against bacteria.45 Furthermore the cells showed enhanced chemotaxis to MIP3β. In addition to models that develop a human immune response against pathogens, humanized models of human allergic reactions and immune-rejection have been established; mature human basophils and mast cells were detected in IL-3/GM-CSF transgenic mice and when exposed to serum from patients with cedar pollinosis mice developed passive cutaneous anaphylactic reactions.42 Moreover, mice that express membrane-bound human SCF rejected HLA-mismatched human skin grafts.41 These observations demonstrate that next-generation humanized mouse models support the development of functional human immune cells in vivo. For more detailed information on human immune responses in humanized mice, we refer to the respective review articles.

Malignant human hematopoiesis in humanized mice Murine, surrogate genetically modified models of HSC malignancies and cell lines might not reflect the complex-

ity of human disease. In contrast, humanized mouse models can reproduce disease characteristics of individual patients and allow assessment of disease heterogeneity, which is essential for the genetic characterization of neoplastic clones and to test therapeutic responses. Humanized mouse models for acute myeloid (AML) and lymphoid leukemia are the most widely used xenograft models. The engraftment of less aggressive, chronic diseases such as MDS, myeloproliferative neoplasms and mature B-cell (including multiple myeloma) and T-cell lymphomas remains a challenge and, as for healthy hematopoiesis, novel approaches aim at generating robust and more consistent models. The identification of LSC in 1994 was the foundation for the cancer stem cell hypothesis.82,83 Human LSC are functionally defined as cells that can be engrafted into mice and propagated into secondary recipients. While it is well accepted that the frequency of LSC is highest in the CD34+CD38– fraction in most AML, LSC are also found outside this population.84,85 The search for additional markers that may allow more stringent identification of LSC is, therefore, ongoing. In contrast to healthy hematopoiesis, leukemia shows a very heterogeneous behavior and the surface phenotype of LSC varies between leukemia subtypes. Additional markers that have been used to characterize AML stem cells include CD123, CD44, CD47, CLL1 (in AML with fms-related tyrosine kinase 3 mutations), CD96 and TIM3 (Figure 5).82,86-90 These markers are expressed at a significantly higher level on the surface of AML stem cells compared to normal HSC. This allows the distinction of

Table 4. Comparison of next-generation humanized mouse models in their support of malignant hematopoiesis.

Disease

Injected population

Route of injection

Number of samples compared

Mouse strain

Conditioning

Additional treatment/ scaffold

Timepoint of analysis

Primary AML

Bulk AML cells

Tail vein

NOD.SCID N.S-β2m-/N/S-S/GM/3

Sublethal irradiation

None

8-12 weeks

CB-MA9Nras

CB-derived line (leukemic)

Tail vein

NOD.SCID None NSS (NOD.SCIDSGM3) NSG NSG-SGM3 NSG Sublethal NSG-SGM3 irradiation

None

CBCB-derived Intrafemoral AML1-ETO line CB-CBFβ- (pre-leukemic) MYH11 Primary Bulk Tail vein AML AML cells Primary Bulk AML Tail AML cells vein

Primary MDS

NSG NSG-SGM3 NSG NSG-SGM3

Sublethal irradiation None

MDS CD34+ Intrafemoral 24 (low risk, NSG n=5; NSG-SGM3 intermediate-1 risk, n=19

Sublethal irradiation

Features

Reference(s)

- BM engraftment: N.S-β2m-/- and N/S-S/GM/3 > NOD.SCID - only 1/8 with high and stable BM engraftment in N/S-S/GM/3 mice When sick - Increased homing in NSG and (d+40 or later) NSG-SGM3 - BM engraftment: NSG-SGM3 > NSS > NSG > NS

None

16 weeks

None

12-16 weeks

None

12-16 weeks

5E+5 patient- 16-28 weeks derived MSC co-transplanted

- BM engraftment: NSG-SGM3 > NSG

108

- 3/5 samples with higher BM engraftment 108 level in NSG-SGM3 (2 NSG non-engrafters) - 4/6 samples with higher BM engraftment 106 level in NSG-SGM3 - Higher CD34+ expression in NSG-SGM3 - Distinct clonal architecture in NSG and NSG-SGM3 - 1/7 samples engrafted in NSG without MSC 63 - 14/20 samples engrafted in NSG-MSC - 4 samples compared side-by-side - increased engraftment in NSG-GM3 mice - Dysplasia only observed in mice transplanted with MSC

AML: acute myeloid leukemia; MDS: myelodysplastic syndrome; CB: umbilical cord blood; MSC: mesenchymal stem cells.

107

haematologica | 2016; 101(1)

Humanized hemato-lymphoid system mice

LSC from HSC, and has been used as a rationale to develop targeted therapies.91 Differential expression of “LSCspecific” antigens provides a therapeutic window for novel therapeutic approaches. Indeed, antibodies directed against these antigens have been assessed in the leukemia xenograft model and have shown promising results, which are now translated into clinical trials. LSC have also been detected in the CD34– fraction, particularly in AML with mutated Nucleophosmin 1.92 An interesting phenomenon is that the leukemia-initiating potential may also reside in more mature progenitors that have acquired LSC properties. This has been demonstrated in an AML xenograft model in which lymphoid multipotent progenitor (LMPP)-like (CD34+CD38–CD45RA+) and GMP-like (CD34+CD38+CD45RA+) populations function as putative LSC.93 Furthermore, recent studies have demonstrated pre-leukemic stem cells upstream of LSC.94,95 The characterization of LSC is most advanced in AML, while the identification of disease-initiating cells in other HSC neoplasms remains a challenge. Although disease heterogeneity may explain the lack of suitable markers to some extent, one of the main reasons for the difficulties in characterization is related to the low engraftment potential of the less aggressive disease in humanized mouse models. A recent study, however, identified MDS 5q- stem cells in the Thy1+CD45RA– fraction of CD34+CD38– cells.96

The identification of LSC in B- and T-lymphoblastic leukemia has been more difficult since there does not seem to be a consistent phenotype among disease subtypes. Although some studies found that CD34+CD38– LSC lack CD19 expression in B-lymphoblastic leukemia other studies have shown that putative LSC express CD19.97-100 Similarly the expression of T cell markers and CD34 may vary in Tlymphoblastic leukemia initiating cells.100-102 Finally, a seminal study identified chronic lymphocytic leukemia initiating cells in the CD34+CD38–CD90+ fraction.103 These studies demonstrate that humanized mouse models can help to identify premalignant and disease-initiating cells in a wide variety of HSC neoplasms and nextgeneration humanized mouse models may further increase the spectrum of hematopoietic malignancies that can be investigated in such models.

Leukemia engraftment and clonal heterogeneity in humanized mouse models Gene expression profiling of a functionally validated LSC enriched population identified a LSC signature that correlates with clinical outcome.104 This demonstrates that LSC and the AML xenograft are not artifacts, but reproduce human disease. However in the AML xenograft model, robust and reproducible engraftment is limited to 40-50% of patients’ samples. A recent correlative study

Figure 6. Possible disease-specific, tailored mouse models in future research. Different humanized mouse models that might serve specific research questions.

haematologica | 2016; 101(1)

A.P.A. Theocharides et al.

with 307 AML samples showed that only 44% generated a graft that reproduced the human correlate, while 28% did not engraft and 29% generated a T-cell or multi-lineage graft.105 The same study demonstrated that a higher proportion of relapse samples compared to diagnostic samples (66% versus 44%) engrafted the mice and the engraftment potential of AML correlated with the aggressiveness of disease. Caution is, therefore, needed when interpreting data and conclusions may only be drawn for a specific disease subtype. Indeed, most published AML xenograft models do not mention the efficacy (engrafting versus non-engrafting samples) of AML engraftment. Consequently, there is an unmet need for mouse strains that support engraftment not only of high-risk, but also of lower-risk disease.

Development of malignant hematopoiesis in next-generation humanized mice. Four studies demonstrated that engraftment of AML and MDS was significantly improved in NSG-SGM3 mice (Table 4).63,106-108 Larger comparative studies are required to determine whether next-generation humanized mice can replace the standard mouse strains in some AML and other less aggressive myeloid neoplasms. However, the AML study by Klco et al. shows that NSG-SGM3 mice preferentially support certain AML sub-clones that are probably more sensitive to regulation and support by human cytokines produced in vivo from the transgene.106 Interestingly, in that study, the most abundant clone found in NSG-SGM3 mice was not the most abundant clone found in the AML sample injected. This suggests that minor sub-clones present in the injected sample can become dominant depending on the different driving or selecting milieu present in the mouse. Moreover NSGSGM3 mice also seem to affect the differentiation process since greater CD34 expression was found in AML cells in NSG-SGM3 mice than in NSG mice. Importantly, in this model none of the xenografts had the same sub-clonal architecture as the one of the injected AML. Preclinical AML therapy models have also revealed heterogeneity in response to treatment between mice transplanted with the same AML sample, suggesting that certain AML subclones may be more or less susceptible to a specific therapeutic intervention – an observation that has also been made in humans. In line with the data outlined above, AML sub-clones with strong dependence on myeloid cytokines may be more susceptible to therapies that target these pathways. Thus, xenograft models for AML not only define LSC and are preclinical therapy models, but also represent a tool allowing dissection of the clonal architecture of AML and the functional properties of AML sub-clones that could become targets for therapy.

Conclusions Humanized mouse models have contributed significantly to the understanding of healthy and diseased human hematopoiesis and have led to the translation of promising therapeutic compounds into clinical applications. However there is room for improvement as several limitations remain. Human adaptive immune responses remain limited in

humanized mice, even when the human T-cell repertoire is selected on human fetal thymic tissue. Therefore, further improvement is needed to strengthen the development of human B and T cells, their education and maturation, the homeostasis of those cells in the periphery, the appropriate formation of secondary lymphoid structures (i.e. lymph nodes with functional germinal centers), and the functional interaction between human B cells and mouse follicular dendritic cells. Several strains of mice that express human HLA molecules have been reported in the past few years, conferring some improvement in the amplitude and/or in the repertoire of the adaptive response.109-112 However, a combination of multiple factors (human cytokines, human HLA molecules, human lymphoid tissues) will be required to sustain an adaptive immune response comparable to that observed in humans, a critical step to make these models suitable for e.g. vaccine testing. Furthermore, with the increasing functionality of human immune cells in the new models of humanized mice developed in the past few years, a new challenge emerges: the human immune system is not entirely tolerant for the mouse host. This was first observed in the BLT model in which human T cells selected on human HLA in the human fetal thymus mediate xenogeneic graft-versushost disease, as peripheral tissues exclusively express mouse MHC molecules.39,113 Moreover human phagocytic cells in mice that support increased myelopoiesis can engulf mouse red blood cells and platelets.43,46 Thus, while previous efforts in the field of humanized mice consisted in generating strains in which the mouse immune system would tolerate the human graft, the development of newer models with a better functional human immune system will also need to focus on inducing tolerance of the human immune system for the mouse host. While equally substantial improvements have been made in the development of humanized mice that support diseased hematopoiesis, significant limitations remain. Future models will need to support the engraftment of the majority of patient samples for a specific disease to allow solid conclusions regarding disease biology and response to therapy. Although mice that produce “myeloid” cytokines seem to move the field in the right direction, new challenges such as the selective and preferential support of hematopoietic sub-clones need to be taken into consideration. Finally, while humanized mouse models for HSC malignancies have led the field of xenograft models in the past, models for human solid organ malignancies and for autoimmune disorders will contribute significantly to disease understanding and development of novel therapeutic approaches in the future. Thus, one can envision that tailored humanized mouse models might be used for specific disease questions in the future (Figure 6). Funding This work was supported by the Bill and Melinda Gates Foundation and US National Institutes of of Health CA156689 (to RAF and MGM) and the University of Zurich Clinical Research Priority Program (to MGM). Funding was also provided by the Swiss Cancer League (to APAT, KLS-3298-08-2013). APAT is a recipient of a Cloëtta medical research position.

haematologica | 2016; 101(1)

Humanized hemato-lymphoid system mice

References 1. Rongvaux A, Takizawa H, Strowig T, et al. Human hemato-lymphoid system mice: current use and future potential for medicine. Annu Rev Immunol. 2013;31:635-674. 2. Shultz LD, Brehm MA, Garcia-Martinez JV, Greiner DL. Humanized mice for immune system investigation: progress, promise and challenges. Nat Rev Immunol. 2012;12 (11):786-798. 3. Kamel-Reid S, Dick JE. Engraftment of immune-deficient mice with human hematopoietic stem cells. Science. 1988;242(4886):1706-1709. 4. Lapidot T, Pflumio F, Doedens M, Murdoch B, Williams DE, Dick JE. Cytokine stimulation of multilineage hematopoiesis from immature human cells engrafted in SCID mice. Science. 1992;255(5048):1137-1141. 5. McCune JM, Namikawa R, Kaneshima H, Shultz LD, Lieberman M, Weissman IL. The SCID-hu mouse: murine model for the analysis of human hematolymphoid differentiation and function. Science. 1988;241 (4873):1632-1639. 6. Mosier DE, Gulizia RJ, Baird SM, Wilson DB. Transfer of a functional human immune system to mice with severe combined immunodeficiency. Nature. 1988;335(6187): 256-259. 7. Hesselton RM, Greiner DL, Mordes JP, Rajan TV, Sullivan JL, Shultz LD. High levels of human peripheral blood mononuclear cell engraftment and enhanced susceptibility to human immunodeficiency virus type 1 infection in NOD/LtSz-scid/scid mice. J Infect Dis. 1995;172(4):974-982. 8. Lowry PA, Shultz LD, Greiner DL, et al. Improved engraftment of human cord blood stem cells in NOD/LtSz-scid/scid mice after irradiation or multiple-day injections into unirradiated recipients. Biol Blood Marrow Transplant. 1996;2(1):15-23. 9. Pflumio F, Izac B, Katz A, Shultz LD, Vainchenker W, Coulombel L. Phenotype and function of human hematopoietic cells engrafting immune-deficient CB17-severe combined immunodeficiency mice and nonobese diabetic-severe combined immunodeficiency mice after transplantation of human cord blood mononuclear cells. Blood. 1996;88(10):3731-3740. 10. Mombaerts P, Iacomini J, Johnson RS, Herrup K, Tonegawa S, Papaioannou VE. RAG-1-deficient mice have no mature B and T lymphocytes. Cell. 1992;68(5):869-877. 11. Shinkai Y, Rathbun G, Lam KP, et al. RAG-2deficient mice lack mature lymphocytes owing to inability to initiate V(D)J rearrangement. Cell. 1992;68(5):855-867. 12. Mazurier F, Fontanellas A, Salesse S, et al. A novel immunodeficient mouse model-RAG2 x common cytokine receptor gamma chain double mutants--requiring exogenous cytokine administration for human hematopoietic stem cell engraftment. J Interferon Cytokine Res. 1999;19(5):533541. 13. Traggiai E, Chicha L, Mazzucchelli L, et al. Development of a human adaptive immune system in cord blood cell-transplanted mice. Science. 2004;304(5667):104-107. 14. Bosma GC, Custer RP, Bosma MJ. A severe combined immunodeficiency mutation in the mouse. Nature. 1983;301(5900):527-530. 15. Ma A, Koka R, Burkett P. Diverse functions of IL-2, IL-15, and IL-7 in lymphoid homeostasis. Annu Rev Immunol. 2006;24:657679.

haematologica | 2016; 101(1)

16. Ito M, Hiramatsu H, Kobayashi K, et al. NOD/SCID/gamma(c)(null) mouse: an excellent recipient mouse model for engraftment of human cells. Blood. 2002;100(9): 3175-3182. 17. Ishikawa F, Yasukawa M, Lyons B, et al. Development of functional human blood and immune systems in NOD/SCID/IL2 receptor {gamma} chain(null) mice. Blood. 2005;106(5):1565-1573. 18. Shultz LD, Lyons BL, Burzenski LM, et al. Human lymphoid and myeloid cell development in NOD/LtSz-scid IL2R gamma null mice engrafted with mobilized human hemopoietic stem cells. J Immunol. 2005;174(10):6477-6489. 19. Barclay AN, Van den Berg TK. The interaction between signal regulatory protein alpha (SIRPalpha) and CD47: structure, function, and therapeutic target. Annu Rev Immunol. 2014;32:25-50. 20. Takenaka K, Prasolava TK, Wang JC, et al. Polymorphism in Sirpa modulates engraftment of human hematopoietic stem cells. Nat Immunol. 2007;8(12):1313-1323. 21. Deng K, Pertea M, Rongvaux A, et al. Broad CTL response is required to clear latent HIV1 due to dominance of escape mutations. Nature. 2015;517(7534):381-385. 22. Strowig T, Rongvaux A, Rathinam C, et al. Transgenic expression of human signal regulatory protein alpha in Rag2-/-gamma(c)-/mice improves engraftment of human hematopoietic cells in humanized mice. Proc Natl Acad Sci USA. 2011;108(32):1321813223. 23. Legrand N, Huntington ND, Nagasawa M, et al. Functional CD47/signal regulatory protein alpha (SIRP(alpha)) interaction is required for optimal human T- and natural killer- (NK) cell homeostasis in vivo. Proc Natl Acad Sci USA. 2011;108(32):1322413229. 24. Waern JM, Yuan Q, Rudrich U, et al. Ectopic expression of murine CD47 minimizes macrophage rejection of human hepatocyte xenografts in immunodeficient mice. Hepatology. 2012;56(4):1479-1488. 25. Prochazka M, Gaskins HR, Shultz LD, Leiter EH. The nonobese diabetic scid mouse: model for spontaneous thymomagenesis associated with immunodeficiency. Proc Natl Acad Sci USA. 1992;89(8):3290-3294. 26. Cosgun KN, Rahmig S, Mende N, et al. Kit regulates HSC engraftment across the human-mouse species barrier. Cell Stem Cell. 2014;15(2):227-238. 27. McIntosh BE, Brown ME, Duffin BM, et al. Nonirradiated NOD,B6.SCID Il2rgamma-/Kit(W41/W41) (NBSGW) mice support multilineage engraftment of human hematopoietic cells. Stem Cell Reports. 2015;4(2):171180. 28. Rongvaux A, Willinger T, Takizawa H, et al. Human thrombopoietin knockin mice efficiently support human hematopoiesis in vivo. Proc Natl Acad Sci USA. 2011;108(6): 2378-2383. 29. Willinger T, Rongvaux A, Strowig T, Manz MG, Flavell RA. Improving human hematolymphoid-system mice by cytokine knockin gene replacement. Trends Immunol. 2011;32(7):321-327. 30. Willinger T, Rongvaux A, Takizawa H, et al. Human IL-3/GM-CSF knock-in mice support human alveolar macrophage development and human immune responses in the lung. Proc Natl Acad Sci USA. 2011;108(6): 2390-2395. 31. Doulatov S, Notta F, Laurenti E, Dick JE. Hematopoiesis: a human perspective. Cell

Stem Cell. 2012;10(2):120-136. 32. Manz MG. Human-hemato-lymphoid-system mice: opportunities and challenges. Immunity. 2007;26(5):537-541. 33. Rongvaux A, Takizawa H, Strowig T, et al. Human hemato-lymphoid system mice: current use and future potential for medicine. Annu Rev Immunol. 2013;31:635-674. 34. Gille C, Orlikowsky TW, Spring B, et al. Monocytes derived from humanized neonatal NOD/SCID/IL2Rgamma(null) mice are phenotypically immature and exhibit functional impairments. Human Immunol. 2012;73(4):346-354. 35. Tanaka S, Saito Y, Kunisawa J, et al. Development of mature and functional human myeloid subsets in hematopoietic stem cell-engrafted NOD/SCID/ IL2rgammaKO mice. J Immunol. 2012;188 (12):6145-6155. 36. Lang J, Kelly M, Freed BM, et al. Studies of lymphocyte reconstitution in a humanized mouse model reveal a requirement of T cells for human B cell maturation. J Immunol. 2013;190(5):2090-2101. 37. Martinez-Torres F, Nochi T, Wahl A, Garcia JV, Denton PW. Hypogammaglobulinemia in BLT humanized mice--an animal model of primary antibody deficiency. PLoS One. 2014;9(10):e108663. 38. Brehm MA, Cuthbert A, Yang C, et al. Parameters for establishing humanized mouse models to study human immunity: analysis of human hematopoietic stem cell engraftment in three immunodeficient strains of mice bearing the IL2rgamma(null) mutation. Clin Immunol. 2010;135(1):84-98. 39. Covassin L, Jangalwe S, Jouvet N, et al. Human immune system development and survival of NOD-scid IL2rgamma (NSG) mice engrafted with human thymus and autologous hematopoietic stem cells. Clin Exp Immunol. 2013;174(3):372-388. 40. Chen Q, Khoury M, Chen J. Expression of human cytokines dramatically improves reconstitution of specific human-blood lineage cells in humanized mice. Proc Natl Acad Sci USA. 2009;106(51):21783-21788. 41. Brehm MA, Racki WJ, Leif J, et al. Engraftment of human HSCs in nonirradiated newborn NOD-scid IL2rgamma null mice is enhanced by transgenic expression of membrane-bound human SCF. Blood. 2012;119(12):2778-2788. 42. Ito R, Takahashi T, Katano I, et al. Establishment of a human allergy model using human IL-3/GM-CSF-transgenic NOG mice. J Immunol. 2013;191(6):2890-2899. 43. Takagi S, Saito Y, Hijikata A, et al. Membrane-bound human SCF/KL promotes in vivo human hematopoietic engraftment and myeloid differentiation. Blood. 2012;119(12):2768-2777. 44. Nicolini FE, Cashman JD, Hogge DE, Humphries RK, Eaves CJ. NOD/SCID mice engineered to express human IL-3, GM-CSF and Steel factor constitutively mobilize engrafted human progenitors and compromise human stem cell regeneration. Leukemia. 2004;18(2):341-347. 45. Rathinam C, Poueymirou WT, Rojas J, et al. Efficient differentiation and function of human macrophages in humanized CSF-1 mice. Blood. 2011;118(11):3119-3128. 46. Rongvaux A, Willinger T, Martinek J, et al. Development and function of human innate immune cells in a humanized mouse model. Nat Biotechnol. 2014;32(4):364-372. 47. Mendelson A, Frenette PS. Hematopoietic stem cell niche maintenance during homeostasis and regeneration. Nat Med.

A.P.A. Theocharides et al. 2014;20(8):833-846. 48. Friedenstein AJ, Gorskaja JF, Kulagina NN. Fibroblast precursors in normal and irradiated mouse hematopoietic organs. Exp Hematol. 1976;4(5):267-274. 49. Luria EA, Owen ME, Friedenstein AJ, Morris JF, Kuznetsow SA. Bone formation in organ cultures of bone marrow. Cell Tissue Res. 1987;248(2):449-454. 50. Luriia EA, Kuznetsov SA, Genkina EN, Fridenshtein A. [Bone tissue formation in organ cultures of human bone marrow]. Biull Eksp Biol Med. 1989;107(5):593-595. 51. Wakitani S, Saito T, Caplan AI. Myogenic cells derived from rat bone marrow mesenchymal stem cells exposed to 5-azacytidine. Muscle Nerve. 1995;18(12):1417-1426. 52. Park D, Spencer JA, Koh BI, et al. Endogenous bone marrow MSCs are dynamic, fate-restricted participants in bone maintenance and regeneration. Cell Stem Cell. 2012;10(3):259-272. 53. Scotti C, Tonnarelli B, Papadimitropoulos A, et al. Recapitulation of endochondral bone formation using human adult mesenchymal stem cells as a paradigm for developmental engineering. Proc Natl Acad Sci USA. 2010;107(16):7251-7256. 54. Chan CK, Chen CC, Luppen CA, et al. Endochondral ossification is required for haematopoietic stem-cell niche formation. Nature. 2009;457(7228):490-494. 55. Sacchetti B, Funari A, Michienzi S, et al. Selfrenewing osteoprogenitors in bone marrow sinusoids can organize a hematopoietic microenvironment. Cell. 2007;131(2):324336. 56. Kuznetsov SA, Krebsbach PH, Satomura K, et al. Single-colony derived strains of human marrow stromal fibroblasts form bone after transplantation in vivo. J Bone Miner Res. 1997;12(9):1335-1347. 57. Scotti C, Piccinini E, Takizawa H, et al. Engineering of a functional bone organ through endochondral ossification. Proc Natl Acad Sci USA. 2013;110(10):3997-4002. 58. Reinisch A, Etchart N, Thomas D, et al. Epigenetic and in vivo comparison of diverse MSC sources reveals an endochondral signature for human hematopoietic niche formation. Blood. 2015;125(2):249-260. 59. Holzapfel BM, Hutmacher DW, Nowlan B, et al. Tissue engineered humanized bone supports human hematopoiesis in vivo. Biomaterials. 2015;61:103-114. 60. Lane SW, Scadden DT, Gilliland DG. The leukemic stem cell niche: current concepts and therapeutic opportunities. Blood. 2009;114(6):1150-1157. 61. Konopleva MY, Jordan CT. Leukemia stem cells and microenvironment: biology and therapeutic targeting. J Clin Oncol. 2011;29(5):591-599. 62. Chen Y, Jacamo R, Shi YX, et al. Human extramedullary bone marrow in mice: a novel in vivo model of genetically controlled hematopoietic microenvironment. Blood. 2012;119(21):4971-4980. 63. Medyouf H, Mossner M, Jann JC, et al. Myelodysplastic cells in patients reprogram mesenchymal stromal cells to establish a transplantable stem cell niche disease unit. Cell Stem Cell. 2014;14(6):824-837. 64. Groen RW, Noort WA, Raymakers RA, et al. Reconstructing the human hematopoietic niche in immunodeficient mice: opportunities for studying primary multiple myeloma. Blood. 2012;120(3):e9-e16. 65. Melkus MW, Estes JD, Padgett-Thomas A, et al. Humanized mice mount specific adaptive and innate immune responses to EBV and

TSST-1. Nat Med. 2006;12(11):1316-1322. 66. Lan P, Tonomura N, Shimizu A, Wang S, Yang YG. Reconstitution of a functional human immune system in immunodeficient mice through combined human fetal thymus/liver and CD34+ cell transplantation. Blood. 2006;108(2):487-492. 67. Jaiswal S, Pazoles P, Woda M, et al. Enhanced humoral and HLA-A2-restricted dengue virus-specific T-cell responses in humanized BLT NSG mice. Immunology. 2012;136(3):334-343. 68. Seung E, Tager AM. Humoral immunity in humanized mice: a work in progress. J Infect Dis. 2013;208 (Suppl 2):S155-159. 69. Parent AV, Russ HA, Khan IS, et al. Generation of functional thymic epithelium from human embryonic stem cells that supports host T cell development. Cell Stem Cell. 2013;13(2):219-229. 70. Denton PW, Garcia JV. Humanized mouse models of HIV infection. AIDS Rev. 2011;13(3):135-148. 71. Denton PW, Nochi T, Lim A, et al. IL-2 receptor gamma-chain molecule is critical for intestinal T-cell reconstitution in humanized mice. Mucosal Immunol. 2012;5(5):555-566. 72. Nochi T, Denton PW, Wahl A, Garcia JV. Cryptopatches are essential for the development of human GALT. Cell Rep. 2013;3(6):1874-1884. 73. Kondo M, Wagers AJ, Manz MG, et al. Biology of hematopoietic stem cells and progenitors: implications for clinical application. Annu Rev Immunol. 2003;21:759-806. 74. Bhatia M, Bonnet D, Murdoch B, Gan OI, Dick JE. A newly discovered class of human hematopoietic cells with SCID-repopulating activity. Nat Med. 1998;4(9):1038-1045. 75. Ishii M, Matsuoka Y, Sasaki Y, et al. Development of a high-resolution purification method for precise functional characterization of primitive human cord bloodderived CD34-negative SCID-repopulating cells. Exp Hematol. 2011;39(2):203-213.e1. 76. McDermott SP, Eppert K, Lechman ER, Doedens M, Dick JE. Comparison of human cord blood engraftment between immunocompromised mouse strains. Blood. 2010;116(2):193-200. 77. Wang JC, Doedens M, Dick JE. Primitive human hematopoietic cells are enriched in cord blood compared with adult bone marrow or mobilized peripheral blood as measured by the quantitative in vivo SCIDrepopulating cell assay. Blood. 1997;89(11): 3919-3924. 78. Mazurier F, Doedens M, Gan OI, Dick JE. Rapid myeloerythroid repopulation after intrafemoral transplantation of NOD-SCID mice reveals a new class of human stem cells. Nat Med. 2003;9(7):959-963. 79. Manz MG, Miyamoto T, Akashi K, Weissman IL. Prospective isolation of human clonogenic common myeloid progenitors. Proc Natl Acad Sci USA. 2002;99(18):11872-11877. 80. Majeti R, Park CY, Weissman IL. Identification of a hierarchy of multipotent hematopoietic progenitors in human cord blood. Cell Stem Cell. 2007;1(6):635-645. 81. Notta F, Doulatov S, Laurenti E, Poeppl A, Jurisica I, Dick JE. Isolation of single human hematopoietic stem cells capable of longterm multilineage engraftment. Science. 2011;333(6039):218-221. 82. Goyama S, Wunderlich M, Mulloy JC. Xenograft models for normal and malignant stem cells. Blood. 2015;125(17):2630-2640. 83. 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. 84. Taussig DC, Miraki-Moud F, Anjos-Afonso F, et al. Anti-CD38 antibody-mediated clearance of human repopulating cells masks the heterogeneity of leukemia-initiating cells. Blood. 2008;112(3):568-575. 85. Eppert K, Takenaka K, Lechman ER, et al. Stem cell gene expression programs influence clinical outcome in human leukemia. Nat Med. 2011;17(9):1086-1093. 86. Majeti R, Chao MP, Alizadeh AA, et al. CD47 is an adverse prognostic factor and therapeutic antibody target on human acute myeloid leukemia stem cells. Cell. 2009; 138(2):286-299. 87. Theocharides AP, Jin L, Cheng PY, et al. Disruption of SIRPalpha signaling in macrophages eliminates human acute myeloid leukemia stem cells in xenografts. J Exp Med. 2012;209(10):1883-1899. 88. Jan M, Chao MP, Cha AC, et al. Prospective separation of normal and leukemic stem cells based on differential expression of TIM3, a human acute myeloid leukemia stem cell marker. Proc Natl Acad Sci USA. 2011;108(12):5009-5014. 89. Hosen N, Park CY, Tatsumi N, et al. CD96 is a leukemic stem cell-specific marker in human acute myeloid leukemia. Proc Natl Acad Sci USA. 2007;104(26):11008-11013. 90. van Rhenen A, van Dongen GA, Kelder A, et al. The novel AML stem cell associated antigen CLL-1 aids in discrimination between normal and leukemic stem cells. Blood. 2007;110(7):2659-2666. 91. Pollyea DA, Gutman JA, Gore L, Smith CA, Jordan CT. Targeting acute myeloid leukemia stem cells: a review and principles for the development of clinical trials. Haematologica. 2014;99(8):1277-1284. 92. Taussig DC, Vargaftig J, Miraki-Moud F, et al. Leukemia-initiating cells from some acute myeloid leukemia patients with mutated nucleophosmin reside in the CD34(-) fraction. Blood. 2010;115(10):1976-1984. 93. Goardon N, Marchi E, Atzberger A, et al. Coexistence of LMPP-like and GMP-like leukemia stem cells in acute myeloid leukemia. Cancer Cell. 2011;19(1):138-152. 94. Jan M, Snyder TM, Corces-Zimmerman MR, et al. Clonal evolution of preleukemic hematopoietic stem cells precedes human acute myeloid leukemia. Sci Transl Med. 2012;4(149):149ra118. 95. Shlush LI, Zandi S, Mitchell A, et al. Identification of pre-leukaemic haematopoietic stem cells in acute leukaemia. Nature. 2014;506(7488):328-333. 96. Woll PS, Kjallquist U, Chowdhury O, et al. Myelodysplastic syndromes are propagated by rare and distinct human cancer stem cells in vivo. Cancer Cell. 2014;25(6):794-808. 97. Cox CV, Evely RS, Oakhill A, Pamphilon DH, Goulden NJ, Blair A. Characterization of acute lymphoblastic leukemia progenitor cells. Blood. 2004;104(9):2919-2925. 98. Castor A, Nilsson L, Astrand-Grundstrom I, et al. Distinct patterns of hematopoietic stem cell involvement in acute lymphoblastic leukemia. Nat Med. 2005;11(6):630-637. 99. Hong D, Gupta R, Ancliff P, et al. Initiating and cancer-propagating cells in TEL-AML1associated childhood leukemia. Science. 2008;319(5861):336-339. 100. Gerby B, Clappier E, Armstrong F, et al. Expression of CD34 and CD7 on human Tcell acute lymphoblastic leukemia discriminates functionally heterogeneous cell populations. Leukemia. 2011;25(8):1249-1258. 101. Cox CV, Martin HM, Kearns PR, Virgo P,

haematologica | 2016; 101(1)

Humanized hemato-lymphoid system mice Evely RS, Blair A. Characterization of a progenitor cell population in childhood T-cell acute lymphoblastic leukemia. Blood. 2007;109(2):674-682. 102. Chiu PP, Jiang H, Dick JE. Leukemia-initiating cells in human T-lymphoblastic leukemia exhibit glucocorticoid resistance. Blood. 2010;116(24):5268-5279. 103. Kikushige Y, Ishikawa F, Miyamoto T, et al. Self-renewing hematopoietic stem cell is the primary target in pathogenesis of human chronic lymphocytic leukemia. Cancer Cell. 2011;20(2):246-259. 104. Eppert K, Takenaka K, Lechman ER, et al. Stem cell gene expression programs influence clinical outcome in human leukemia. Nat Med. 2011;17(9):1086-1093. 105. Kennedy JA, Mitchell A, Chen WC, et al. Leukemic Engraftment in NOD.SCID Mice is correlated with clinical parameters and predicts outcome in human AML. Blood. 2013;122(21):50.

haematologica | 2016; 101(1)

106. Klco JM, Spencer DH, Miller CA, et al. Functional heterogeneity of genetically defined subclones in acute myeloid leukemia. Cancer Cell. 2014;25(3):379-392. 107. Feuring-Buske M, Gerhard B, Cashman J, Humphries RK, Eaves CJ, Hogge DE. Improved engraftment of human acute myeloid leukemia progenitor cells in beta 2microglobulin-deficient NOD/SCID mice and in NOD/SCID mice transgenic for human growth factors. Leukemia. 2003; 17(4):760-763. 108. Wunderlich M, Chou FS, Link KA, et al. AML xenograft efficiency is significantly improved in NOD/SCID-IL2RG mice constitutively expressing human SCF, GM-CSF and IL-3. Leukemia. 2010;24(10):1785-1788. 109. Strowig T, Gurer C, Ploss A, et al. Priming of protective T cell responses against virusinduced tumors in mice with human immune system components. J Exp Med. 2009;206(6):1423-1434.

110. Shultz LD, Saito Y, Najima Y, et al. Generation of functional human T-cell subsets with HLA-restricted immune responses in HLA class I expressing NOD/SCID/IL2r gamma(null) humanized mice. Proc Natl Acad Sci USA. 2010;107(29):13022-13027. 111. Danner R, Chaudhari SN, Rosenberger J, et al. Expression of HLA class II molecules in humanized NOD.Rag1KO.IL2RgcKO mice is critical for development and function of human T and B cells. PLoS One. 2011;6(5):e19826. 112. Suzuki M, Takahashi T, Katano I, et al. Induction of human humoral immune responses in a novel HLA-DR-expressing transgenic NOD/Shi-scid/gammacnull mouse. Int Immunol. 2012;24(4):243-252. 113. Greenblatt MB, Vrbanac V, Tivey T, Tsang K, Tager AM, Aliprantis AO. Graft versus host disease in the bone marrow, liver and thymus humanized mouse model. PLoS One. 2012;7(9):e44664.

REVIEW ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Risk factors for relapse after allogeneic transplantation in acute myeloid leukemia

Gert J. Ossenkoppele, Jeroen J.W.M. Janssen, and Arjan A. van de Loosdrecht VU University Medical Center, Amsterdam, the Netherlands

ABSTRACT

Haematologica 2016 Volume 101(1):20-25

T his review ar ticle is based on the educational manuscript from the EHA20 meeting (Vienna, June 2015).

Correspondence: g.ossenkoppele@vumc.nl

Received: 20/11/2015. Accepted: 24/11/2015.

doi:10.3324/haematol.2015.139105

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/1/20

cute myeloid leukemia is a clonal neoplasm derived from myeloid progenitor cells with a varying outcome. The initial goal of treatment is the achievement of complete remission, defined for over 40 years by morphology. However, without additional postremission treatment the majority of patients relapse. In many cases of acute myeloid leukemia, allogeneic stem cell transplantation offers the best prospects of cure. In 2013, 5608 stem cell transplantations in acute myeloid leukemia were performed in Europe (5228 allogeneic and 380 autologous stem cell transplantations). Most stem cell transplantations are performed in first complete remission. However, despite a considerable reduction in the chance of relapse, in most studies, overall survival benefit of allogeneic stem cell transplantation is modest due to substantial non-relapse mortality. Here we discuss the many factors related to the risk of relapse after allogeneic stem cell transplantation.

Introduction Acute myeloid leukemia (AML) is a clonal neoplasm derived from myeloid progenitor cells with a varying outcome. The initial goal of treatment is the achievement of complete remission (CR), defined for over 40 years by morphology.1 Without additional post-remission treatment, however, the majority of patients relapse. In many cases of AML, allogeneic stem cell transplantation (alloSCT) offers the best prospects of cure. Apart from the conditioning regimen, the anti-leukemic potential is mainly based on the immunological graft-versus-leukemia effect. Indeed, AML is the most frequent indication for alloSCT, as indicated by data from both the European Group for Blood and Marrow Transplantation (EBMT) and the International Bone Marrow Transplant Registry (IBMTR), and the number of patients transplanted for this indication is growing year by year.2 In 2013, 5608 stem cell transplants in AML in Europe were performed [alloSCT: 5228 and autologous (auto) SCT: 380]. The rise in frequency in recent years is due to the application of reduced intensity conditioning (RIC) regimens and the expansion of alternative donor stem cell sources derived from mismatched relatives and unrelated volunteers. The total number of family donors was 2354 (1913 HLA identical and 441 non-identical) while 2863 unrelated transplants were performed, among which there were 211 cord blood transplantations. The majority of patients were transplanted in first complete remission (CR1). A meta-analysis of prospective trials and trials reporting relapse-free survival (RFS) and/or overall survival (OS) outcomes after assigning adult patients with AML in CR1 to undergo alloSCT versus nonalloSCT treatment, based on donor availability (donor vs. no-donor comparisons), showed that, except for good risk AML in CR1, alloSCT gives a significant survival benefit for intermediate and poor risk AML.3 However, despite a considerable reduction in the possibility of relapse, in most studies, OS benefit of alloSCT is modest due to substantial non-relapse mortality (Table 1).4-9 Now we are approaching a situation in which we can identify a suitable donor, either matched unrelated, haplo-identical or cord, for nearly every patient, This means that both disease-related and transplant-related factors should be carefully balanced before proceeding to an allogeneic or non-allogeneic approach after achieving CR.10 haematologica | 2016; 101(1)

Risk factors for relapse after allogeneic transplantation in AML

Unfortunately, even after alloSCT, a substantial number of AML patients will ultimately relapse, and in these cases survival is very poor.11 Hematologists are now facing the question â&#x20AC;&#x153;which patient should get a transplant in first remission?â&#x20AC;? Predicting relapse in an individual patient still remains a challenge. Here, we will mainly focus on factors predicting for relapse after allogeneic transplant in AML.

Transplant-related factors in relation to relapse risk after transplantation Dose intensity is a main determinant for relapse. With increasing dosage, even at the myeloablative (MA) level, the chance of relapse decreases.12 RIC have become popular and these are being increasingly adopted. Although no prospective randomized trials have been completed, these lower intensity conditioning regimens seem to be associated with a higher rate of relapse. A retrospective EBMT comparison of conditioning regimens showed an increase in relapse rate of 23%-39% after MA versus RIC.13 Nevertheless OS benefits for MA are, at best, modest due to the increased non-relapse mortality (NRM) in comparison with RIC. Although the only prospective randomized study comparing RIC versus MA was stopped early because of slow accrual of patients, there was no significant difference in RFS, NRM and OS.14 In general, it is clear that increasing the intensity of the conditioning regimen results in more acute graft-versus-host disease (GvHD) and increased NRM. In several, but not all, studies, more intensive GvHD prophylaxis, for example by pre-transplant administration of anti-thymocyte globulin or T-cell depletion of the graft, resulted in a higher relapse rate, probably due to less GvL. Whether early donor lymphocyte infusion after T-cell depleting strategies or high-dose cyclophosphamide post transplant counterbalance NRM versus relapse is the subject of much debate and investigation.15,16 Chronic GvHD reduced the chance of relapse in many studies; however, in a recent registry study of the Center for International Blood and Marrow Transplant Research (CIBMTR), its impact seemed only clinically relevant for CML after myeloablative alloSCT and not for AML.17 In this study, cGvHD was primarily associated with a higher transplant-related mortality (TRM).

Disease-related characteristics in relation to relapse risk after transplantation Patient-specific biological factors associated with risk of relapse after transplantation are basically the same as those in patients treated with chemotherapy only. Based

on cytogenetic analysis, three risk groups for relapse can be identified, as described by Grimwade et al.18 New categories such as the monosomal karyotype associated with a very poor outcome are also used to estimate the possibility of relapse by various AML trial groups.19 Novel genomic technology has paved the way for the detection of numerous new molecular aberrations, further underlining the heterogeneity of AML, and these are also helpful to establish a prognosis. Some of these molecular aberrancies (mutations in the NPM1, CEBPA and FLT3 genes) have already been incorporated into the widely applied recommendations for standardized reporting of genetic abnormalities by the European LeukemiaNet panel of experts published in 2010 (Table 2).20 Since then, several new molecular aberrancies have been identified, some of which are clearly associated with outcome, while others are less so. The assumption that most of these aberrancies are equally important for risk of relapse after chemotherapy, as well as after transplantation, still has to be proved. Some of these novel genetic data will be discussed below.

Mutations of CCAAT/enhancer binding protein alpha Mutations of CCAAT/enhancer binding protein alpha (CEBPA) encodes a transcription factor essential for differentiation along the neutrophil lineage. Although already incorporated into the ELN recommendations, it has become clear that only double and not single mutated CEBPA is associated with a favorable outcome.21

Mutations of the nucleophosmin gene The nucleophosmin gene (NPM1) gene encodes for a protein that shuttles between nucleus and cytoplasm, acting, amongst other things, as a molecular chaperone of histone proteins. In addition, it is involved in other critical cellular functions, such as ribosome biogenesis and transport, and centrosome duplication during the cell cycle. Mutations consist of a 4-base pair insertion that alters the C-terminal end of the protein, leading to a nuclear export signal of the encoded protein, and thus to loss of shuttling activity. Several studies have shown a favorable impact of the NPM1 mutation on outcome in the absence of FLT3 gene mutations. In a donor versus no donor analysis, Rollig et al. recently showed that alloSCT resulted in a significantly prolonged RFS in patients with NPM1+ mutated AML. However, OS was not improved, most probably due to the fact that relapsed NPM1+ AML patients responded well to salvage treatment.22 NPM1+FLT3 ITDAMLs are now grouped together with the core binding

Table 1. Allogeneic stem cell transplantation for acute myeloid leukemia in first complete remission in comparison to autologous stem cell transplantation and chemotherapy.

AML Trial Group EORTC/AML-8 GOELAM ECOG/CALGB/SWOG EORTC AML-10 UK MRC AML-10 HOVON-SAKK

Relapse rate Allo Auto 24% 37% 29% 30% 36% 32%

Chemo

41% 45% 48% 52%

57% 55% 61% 52%* 59%*

Overall survival Allo Auto 59% 55% 46% 58% 55% 54%

Chemo

56% 52% 43% 50%

46% 58% 52% 42%* 46%*

AML: acute myeloid leukemia; ALLO: allogeneic stem cell transplantation; AUTO: autologous stem cell transplantation; CHEMO: chemotherapy.Table modified from Kanate et al.51 *No separate data of autoSCT and chemotherapy available.

haematologica | 2016; 101(1)

G.J. Ossenkoppele et al.

factor (CBF) leukemias and the CEBPA double mutants and have a good prognosis.

Internal tandem duplications of the FMS-like tyrosine kinase 3 gene Internal tandem duplications (ITDs) of the FMS-like tyrosine kinase 3 gene occur in 20%-30% of AML cases, predominantly in those with normal cytogenetics. They are associated with a poor outcome, especially when the ratio between mutated and non-mutated FLT3 gene is more than 0.51. This is not only due to an increased risk of relapse, but also due to refractoriness to induction treatment.23,24

Expression of the ecotropic viral integration-1 oncogene Ten percent of AML cases show high ecotropic viral integration-1 (EVI-1) oncogene expression, which predicts for a particularly poor outcome. Besides inv(3)/t(3;3), EVI1(+) is significantly associated with the chromosome abnormalities monosomy 7 and t(11q23). EVI1+ is virtually absent in favorable-risk AML and AML with NPM1 mutations.25

Mutations in DNA-methyltransferase-3A The DNA-methyltransferase-3A (DNMT3A) enzyme plays a role in DNA methylation by transferring methylgroups to DNA CpG islands. Mutations in this gene are detected in approximately 20% of AML cases and enriched in normal karyotype AML (CN-AML) with an incidence of between 30%-37%. Various studies have proved to be inconclusive as to the prognostic value for outcome. A recent meta-analysis showed a slightly poor prognostic impact on OS.26,27

Mutations in the additional sex combs-like 1 gene The additional sex combs-like (ASXL)1 gene is an epigenetic scaffolding protein that assembles epigenetic regulators and transcription factors to specific genomic loci with histone modifications. The incidence of truncating mutations in ASXL1 is between 5%-12% and is always associated with a poor outcome.28

Mutations in the ten-eleven translocation-2 gene The ten-eleven translocation-2 (TET2) gene is a key

enzyme for DNA demethylation and a critical regulator for hematopoietic stem cell homeostasis, whose functional impairment leads to hematologic malignancies. Mutations are frequently found in AML. Although in most studies it is associated with a poor outcome, the prognostic relevance has not been definitively established.29,30

TP53 mutations and 17p abnormalities The Study Alliance Leukemia (SAL) and other groups reported poor survival data in patients with abn(17p) or TP53-mutated AML.31 In a large retrospective repositorybased analysis published by Grossmann et al., patients with TP53-mutated AML had the worst prognosis of all molecularly defined risk groups, with a median OS of 4.6 months and event-free survival (EFS) of 0% at three years.32

Isocitrate dehydrogenase1 and -2 mutations Isocitrate dehydrogenase1 and -2 (IDH1/2) mutations are found in AML as well in gliomas, both with an incidence of approximately 15%-20%. IDH1 and IDH 2 are critical enzymes in the citric acid cyclus, converting isocitrate to α-ketoglutarate. Due to the mutation, 2 hydroxyglutarate (2-HG) instead of α-ketoglutarate is formed. 2-HG inhibits TET enzymes, which results in hypermethylation. This is thought to suppress expression of tumor suppressor genes. These mutations are mutually exclusive with TET2 mutations. Although the prognostic significance has still not been established, some studies have shown a poor outcome.33,34

RUNX1 mutations RUNX1 mutations are considered to be associated with a poor prognosis. The incidence (8%-16%) is higher in elderly AML and in secondary AML.35

Mutations in other genes C-KIT mutations especially have poor prognostic implications in t(8;21) AML and predict for higher relapse rates than unmutated cases; however, MRD status before alloSCT was more predictive than c-kit status.36 The clinical impact of expression levels of numerous other genes such as BAALC, ERG, MN1 etc., still has to be determined. How these new molecular aberrancies could be integrated

Table 2. Standardized reporting for correlation of cytogenetic and molecular genetic data in acute myeloid leukemia with clinical data.2

Genetics group

Subsets

Favorable

t(8;21)(q22;q22); RUNX1-RUNX1T1 inv(16)(p13.1q22) or t(16;16)(p13.1;q22); CBFB-MYH11 Mutated NPM1 without FLT3-ITD (normal karyotype) Mutated CEBPA (normal karyotype) Mutated NPM1 and FLT3-ITD (normal karyotype) Wild-type NPM1 and FLT3-ITD (normal karyotype) Wild-type NPM1 without FLT3-ITD (normal karyotype) t(9;11)(p22;q23); MLLT3-MLL Cytogenetic abnormalities not classified as favorable or adverse† inv(3)(q21q26.2) or t(3;3)(q21;q26.2); RPN1-EVI1 t(6;9)(p23;q34); DEK-NUP214 t(v;11)(v;q23); MLL rearranged −5 or del(5q); −7; abnl(17p); complex karyotype‡

Intermediate-I*

Intermediate-II Adverse

*Includes all AMLs with normal karyotype except for those included in the favorable subgroup; most of these cases are associated with poor prognosis, but they should be reported separately because of the potential different response to treatment. †For most abnormalities, adequate numbers have not been studied to draw firm conclusions regarding their prognostic significance. ‡Three or more chromosome abnormalities in the absence of one of the WHO designated recurring translocations or inversions, that is, t(15;17), t(8;21), inv(16) or t(16;16), t(9;11), t(v;11)(v;q23), t(6;9), inv(3) or t(3;3), indicate how many complex karyotype cases have involvement of chromosome arms 5q, 7q, and 17p.

haematologica | 2016; 101(1)

Risk factors for relapse after allogeneic transplantation in AML Table 3. Recommendation for allogeneic SCT in AML CR1 based on integrated risk profiles. Adapted by HOVON-SAKK from the ELN recommendation by adding new molecular markers and MRD.10

AML risk group‡

AML risk assessment criteria at diagnosis and early/late CR

Good

t(8;21) or AML1-ETO, WBC≤20 +orinv16/t(16;16) or CBFB-MYH11 CEBPA-biallelic mutant+ FLT3TD-/NMP1+, CN –X –Y, WBC≤100, CRe t(8;21) or AML1-ETO, plus WBC>20 or mutant KIT CN –X –Y, WBC≤100, CRe + t(8;21) or AML1-ETO, WBC>20 + and/or mutant KIT + CN –X –Y, WBC≤100, not CRe +orCN –X –Y, WBC>100, _ CA, but non-CBF, MK-, no abn3q26 _ CN –X –Y, WBC>100 CA, but non CBF, MK-, no abn3q26, EVI1-neg MK+ abn3q26 Non CBF, EVI1+ +orNon CBF with mutant p53, or mutant RUNX1, or mutant ASXL1 or bi-allelic FLT3-ITD with FLT3-ITD/FLT3wt ratio of >0.6

Intermediate

Poor

Very poor

MRD after cycle2 +/-

Risk of relapse following consolidation approach Chemotherapy or Allogeneic autologous HSCT (%) HSCT (%)

Prognostic scores for nonrelapse mortality that would indicate allogeneic HSCT as preferred consolidation EBMT HCT–CI Nonrelapse score52 score53 mortality risk (%)

35–40

15–20

NA (≤1)

NA (<1)

10–15

50–55

20–25

≤2

<20–25

70–80

30–40

≤3–4

<30

>90

40–50

≤5

<40

The proposed patient-specific application of allogeneic HSCT in AML CR1 integrates the individual risks for relapse and non-relapse mortality and aims for a DFS benefit of at least 10% for the individual patient compared with consolidation by a non-allogeneic HSCT approach.The categorization of AML is based on cytogenetic, molecular and clinical parameters (including WBC) into good, intermediate and (very) poor subcategories as currently applied by the Dutch–Belgian Cooperative Trial Group for Hematology Oncology and Swiss Group for Clinical Cancer Research (HOVON–SAKK) consortium.

into a new prognostic algorithm was illustrated by Patel et al. who applied high throughput sequencing of TET2, ASXL1, DNMT3A, CEBPA, PHF6, WT1, TP53, EZH2, RUNX1, PTEN, FLT3, NPM1, HRAS, KRAS, NRAS, KIT, IDH1 and IDH2 in over 500 patients from the ECOG E1900 trial. They showed that the mutational analysis of 9 of these genes could be used to retrospectively classify patients into more precise subgroups with favorable-risk, intermediate-risk, or unfavorable-risk profiles, with marked differences in the overall outcome.37 Another example of how molecular profiling may be helpful has been shown by Grosmann et al.38 In a large cohort of AML patients for whom cytogenetic data was available, they investigated the following molecular alterations: PMLRARA, RUNX1-RUNX1T1, CBFB-MYH11, FLT3-ITD, and MLL-PTD, as well as mutations in NPM1, CEPBA, RUNX1, ASXL1, and TP53. Five distinct prognostic subgroups were identified: 1) very favorable: PML-RARA rearrangement or CEPBA double mutations (OS at 3 years: 82.9%); 2) favorable: RUNX1-RUNX1T1, CBFB-MYH11, or NPM1 mutation without FLT3-ITD (OS at 3 years: 62.6%); 3) intermediate: none of the mutations leading to assignment into groups 1, 2, 4, or 5 (OS at 3 years: 44.2%); 4) unfavorable: MLL-PTD and/or RUNX1 mutation and/or ASXL1 mutation (OS at 3 years: 21.9%); and 5) very unfavorable: TP53 mutation (OS at 3 years: 0%). This profile, based only on molecular abnormalities, provided a more haematologica | 2016; 101(1)

powerful model for prognostication than cytogenetics.38 Taken together, these data indicate that more detailed genetic analysis may lead to improved risk stratification

Clinical features in relation to risk of relapse after transplantation Apart from cytogenetic and molecular prognostic markers identified at diagnosis, a number of clinical variables that can be assessed either at diagnosis or during induction and consolidation treatment might offer additional prognostic information. These include high WBC count, time to complete remission, day 15 presence of blasts, extramedullary disease, and quantifiable levels of minimal residual disease (MRD) after induction or consolidation therapy. Here we will focus on the prognostic value of MRD.

Minimal residual disease Despite a multitude of prognostic factors at diagnosis, the outcome of patients is still highly variable and not individually predictable. It thus seems that prognosticators at diagnosis will not enable clinicians to reach the ultimate goal of truly individualized risk assessment. One important issue is that the prognostic impact of these factors in the present risk groups does not take into account the contribution of several cellular resistance mechanisms at diagnosis, and also post-diagnosis factors, which include 23

G.J. Ossenkoppele et al.

dosage, compliance, pharmacological resistance, and probably other unknown features. These factors, which corroborate proper risk classification, are only partly covered by inclusion of CR status. However, the ability to define residual disease far below the level of 5% blast cells is changing the landscape of risk classification. This so-called minimal residual disease (MRD) approach currently enables detection of leukemia cells down to levels of 1:1,000–1:106 white blood cells, compared to only 1:20 for morphology.39,40

Methods for detection of minimal residual disease The most sensitive method at present is real-time quantitative polymerase chain reaction (RQ-PCR). Chimeric fusion genes like PML-RARA, RUNX1-RUNXT1, and CBFB-MYH11 are reliable markers for MRD evaluation; however these genetic abnormalities are only present in approximately 20% of AML cases. NPM1 mutation is an abnormality that occurs more frequently for which mutation-specific RQ-PCR assays have been developed to monitor MRD.41 Another approach for assessment of MRD has gained increasing attention: the use of aberrant (i.e. leukemia-associated) immunophenotypes (LAIP) with flow cytometry. Currently, with this approach, aberrancies may be detected in over 90% of AML cases at diagnosis. These LAIPs consist of normally occurring markers, present in aberrant combinations in AML bone marrow (BM), while only at very low frequencies, or even absent, in normal and regenerating BM.42

Minimal residual disease in clinical studies During the last 15 years, numerous single institute studies have been carried out, both in adult and pediatric AML, which have established the independent prognostic value of “Immunophenotypic MRD.” The results of the first multicenter (31 sites) multinational prospective MRD study in adult AML (aged 18-60 years) were recently reported by the HOVON/SAKK investigators. In this study, BM MRD was assessed at five different sites in samples obtained after one and 2 cycles of induction therapy and after consolidation therapy.43 MRD measurements were performed in a blinded fashion, i.e. without knowing the patients’ performance, while patients were treated according to protocol also without knowledge of

References 1. Bishel HF. Criteria for the evaluation of response for treatment of AML. Blood 1956;11:676-681. 2. Passweg JR, Baldomero H, Bader P, et al. Hematopoietic SCT in Europe 2013: recent trends in the use of alternative donors showing more haploidentical donors but fewer cord blood transplants. Bone Marrow Transplant. 2015;50(4):476-482. 3. Koreth J, Schlenk R, Kopecky KJ, et al. Allogeneic stem cell transplantation for acute myeloid leukemia in first complete remission: systematic review and metaanalysis of prospective clinical trials. JAMA. 2009;301(22):2349-2361. 4. Cassileth PA, Harrington DP, Appelbaum FR, et al. Chemotherapy compared with autologous or allogeneic bone marrow

MRD-related data. After all treatment cycles, low MRD values distinguished patients with relatively favorable outcome from those with adverse RFS and OS. In the whole patient group, and in the clinically most interesting subgroup with intermediate risk cytogenetics, MRD was an independent prognostic. After all treatment courses, low MRD values distinguished patients with relatively favorable outcome from those with adverse RFS and OS. Multivariate analysis after cycle 2I, when decisions about consolidation treatment have to be made, confirmed that high MRD values (>0.1% of WBC) was associated with a considerably higher risk of relapse even after adjustment for consolidation treatment time-dependent co-variate risk score and early or later CR. Similar results were obtained by Freeman et al. who used MRD assessments in a group of elderly patients. In addition, MRD determination by means of quantitative PCR of PML-RARA, AML1ETO, CBFB-MYH11 transcripts and of mutations in NPM1 has also proven to be of value in the clinical arena.44-46 Before transplantation, MRD status in CR1 AML patients was demonstrated to be a highly valuable marker of disease recurrence and shorter OS after transplant. Several studies in patients undergoing myeloablative conditioning all show that the presence of MRD has a negative impact on post-transplant relapse risk.47-49 Recently, Walter et al. showed that the same holds true for non-myeloablative conditioning.50 An integrated riskadapted approach on allogeneic transplantation for patients with AML is given in Table 3. This is currently being followed in the HOVON-SAKK trials and follows the ELN consensus statement, now further refined by introducing new molecular markers and MRD determination after cycle 2.

Conclusion Risk of relapse after allogeneic stem cell transplantation is a composite of many factors, as described here. An estimation of this risk together with an estimation of expected treatment-related mortality should be used as a guide to determine which patient should be offered an alloSCT as post-remission treatment. It is clear that this must be an individualized, well-balanced clinical decision, which cannot be made from a single recommendation that fits all.

transplantation in the management of acute myeloid leukemia in first remission. N Engl J Med. 1998;339(23):1649–1656. 5. Zittoun RA, Mandelli F, Willemze R, et al. Autologous or allogeneic bone marrow transplantation compared with intensive chemotherapy in acute myelogenous leukemia. European Organization for Research and Treatment of Cancer (EORTC) and the Gruppo Italiano Malattie Ematologiche Maligne dell’Adulto (GIMEMA) Leukemia Cooperative Groups. N Engl J Med. 1995;332(4):217–223. 6. Harousseau JL, Cahn JY, Pignon B, et al. Comparison of autologous bone marrow transplantation and intensive chemotherapy as postremission therapy in adult acute myeloid leukemia. The Groupe Ouest Est Leucémies Aiguës Myéloblastiques (GOELAM) Blood. 1997;90(8):2978–2986. 7. Burnett AK, Wheatley K, Goldstone AH, et

al. The value of allogeneic bone marrow transplant in patients with acute myeloid leukaemia at differing risk of relapse: results of the UK MRC AML 10 trial. Br J Haematol. 2002;118(2):385–400. 8. Suciu S, Mandelli F, de Witte T, et al. Allogeneic compared with autologous stem cell transplantation in the treatment of patients younger than 46 years with acute myeloid leukemia (AML) in first complete remission (CR1): an intention-to-treat analysis of the EORTC/GIMEMAAML-10 trial. Blood. 2003;102(4):1232–1240. 9. Cornelissen JJ, van Putten WL, Verdonck LF, et al. Results of a HOVON/SAKK donor versus no-donor analysis of myeloablative HLA-identical sibling stem cell transplantation in first remission acute myeloid leukemia in young and middle-aged adults: benefits for whom? Blood. 2007;109(9): 3658–3666.

haematologica | 2016; 101(1)

Risk factors for relapse after allogeneic transplantation in AML 10. Cornelissen JJ, Gratwohl A, Schlenk RF, et al. The European LeukemiaNet AML Working Party consensus statement on allogeneic HSCT for patients with AML in remission: an integrated-risk adapted approach. Nat Rev Clin Oncol. 2012;9(10):579-590. 11. Fathi AT, Chen YB. Treatment of relapse of acute myeloid leukemia after allogeneic hematopoietic stem cell transplantation. Curr Hematol Malig Rep. 2014;9(2):186-192. 12. Appelbaum FR. Dose intensity of preparative regimens for acute myeloid leukemiaone size fits all- or tailor- made? Best Pract Res Clin Hematol. 2010;23(4):509-517. 13. Martino R, de Wreede L, Fiocco M, et al. Comparison of conditioning regimens of various intensities for allogeneic hematopoietic SCT using HLA-identical sibling donors in AML and MDS with <10% BM blasts: a report from EBMT. Bone Marrow Transplant. 2013;48(6):761-770. 14. Bornhäuser M, Kienast J, Trenschel R, et al. Reduced-intensity conditioning versus standard conditioning before allogeneic haemopoietic cell transplantation in patients with acute myeloid leukaemia in first complete remission: a prospective, open-label randomised phase 3 trial. Lancet Oncol. 2012;13(10):1035-1044. 15. Blaise D, Castagna L. Do different conditioning regimens really make a difference? Hematology Am Soc Hematol Educ Program. 2012;2012:237-245. 16. Al-Homsi AS, Roy TS, Cole K, Feng Y, Duffner U. Post-Transplant High-Dose Cyclophosphamide for the Prevention of Graft-versus-Host Disease. Biol Blood Marrow Transplant. 2015;21(4):604-611. 17. Boyiadzis M, Arora M, Klein J, et al. Impact of chronic graft-versus-host disease on late relapse and survival on 7489 patients after myeloablative allogeneic hematopoietic cell transplantation for leukemia. Clin Cancer Res. 2015;21(9):2020-2028. 18. Grimwade D, Hills RK, Moorman AV, et al. Refinement of cytogenetic classification in acute myeloid leukemia: determination of prognostic significance of rare recurring chromosomal abnormalities among 5876 younger adult patients treated in the United Kingdom Medical Research Council trials. Blood. 2010;116(3):354–365. 19. Breems DA, Van Putten WL, De Greef GE, et al. Monosomal karyotype in acute myeloid leukemia: a better indicator of poor prognosis than a complex karyotype. J Clin Oncol. 2008;26(29):4791-4797. 20. Döhner H, Estey EH, Amadori S, et al. Diagnosis and management of acute myeloid leukemia in adults: recommendations from an international expert panel, on behalf of the European LeukemiaNet. Blood. 2010;115(3):453-474. 21. Taskesen E, Bullinger L, Corbacioglu A, et al. Prognostic impact, concurrent genetic mutations, and gene expression features of AML with CEBPA mutations in a cohort of 1182 cytogenetically normal AML patients: further evidence for CEBPA double mutant AML as a distinctive disease entity. Blood. 2011;117(8):2469-2475. 22. Röllig C, Bornhäuser M, Kramer M, et al. Allogeneic Stem-Cell Transplantation in Patients With NPM1-Mutated Acute Myeloid Leukemia: Results From a Prospective Donor Versus No-Donor Analysis of Patients After Upfront HLA Typing Within the SAL-AML 2003 Trial. J Clin Oncol. 2015;33(5):403-410. 23. Kiyoi H, Yanada M, Ozekia K. Clinical sig-

haematologica | 2016; 101(1)

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

nificance of FLT3 in leukemia. Int J Hematol. 2005;82(2):85-92. Schlenk RF, Kayser S, Bullinger L, et al. Differential impact of allelic ratio and insertion site in FLT3-ITD-positive AML with respect to allogeneic transplantation. Blood. 2014;124(23):3441-3449. Gröschel S, Lugthart S, Schlenk RF, et al. High EVI1 expression predicts outcome in younger adult patients with acute myeloid leukemia and is associated with distinct cytogenetic abnormalities. J Clin Oncol. 2010;28(12):2101-2107. Gaidzik VI, Schlenk RF, Paschka P, et al. Clinical impact of DNMT3A mutations in younger adult patients with acute myeloid leukemia: results of the AML Study Group (AMLSG). Blood. 2013;121(23):4769-4777. Tie R, Zhang T, Fu H, et al. Association between DNMT3A mutations and prognosis of adults with de novo acute myeloid leukemia: a systematic review and metaanalysis. PLoS One. 2014;9(6):e93353. Schnittger S, Eder C, Jeromin S, et al. ASXL1 exon 12 mutations are frequent in AML with intermediate risk karyotype and are independently associated with an adverse outcome. Leukemia. 2013;27(1):82-91. Damm F, Markus B, Thol F, et al. TET2 mutations in cytogenetically normal acute myeloid leukemia: clinical implications and evolutionary patterns. Genes Chromosomes Cancer. 2014;53(10):824-832. Gaidzik VI, Paschka P, Späth D, et al. TET2 mutations in acute myeloid leukemia (AML): results from a comprehensive genetic and clinical analysis of the AML study group. J Clin Oncol. 2012;30(12):1350-1357. Seifert H, Mohr B, Thiede C, et al. The prognostic impact of 17p (p53) deletion in 2272 adults with acute myeloid leukemia. Leukemia. 2009;23(4):656-663. Middeke JM, Beelen D, Stadler M, et al. Outcome of high-risk acute myeloid leukemia after allogeneic hematopoietic cell transplantation: negative impact of abnl(17p) and -5/5q- Blood. 2012;120(12): 2521–2528. Kats LM, Reschke M, Taulli R, et al. Protooncogenic role of mutant IDH2 in leukemia initiation and maintenance.Cell Stem Cell. 2014;14(3):329-341. Nomdedéu J, Hoyos M, Carricondo M, et al. Adverse impact of IDH1 and IDH2 mutations in primary AML: experience of the Spanish CETLAM group.Leuk Res. 2012;36(8):990-997. Gaidzik VI, Bullinger L, Schlenk RF, et al. RUNX1 mutations in acute myeloid leukemia: results from a comprehensive genetic and clinical analysis from the AML study group. J Clin Oncol. 2011;29(10): 1364-1372. Wang Y, Wu DP, Liu QF, et al. In adults with t(8;21)AML, posttransplant RUNX1/RUNX1T1-based MRD monitoring, rather than c-KIT mutations, allows further risk stratification. Blood. 2014;124(12): 1880-1886. Patel JP, Gönen M, Figueroa ME, et al. Prognostic relevance of integrated genetic profiling in acute myeloid leukemia. N Engl J Med. 2012;366(12):1079-1089. Grossmann V, Schnittger S, Kohlmann A, et al. A novel hierarchical prognostic model of AML solely based on molecular mutations. Blood. 2012;120(15):2963-2972. Grimwade D, Freeman SD. Defining minimal residual disease in acute myeloid leukemia: which platforms are ready for

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

"prime time"? Blood. 2014;124(23):33453355. Ossenkoppele GJ, Schuurhuis GJ. MRD in AML: it is time to change the definition of remission. Best Pract Res Clin Haematol. 2014;27(3-4):265-271. Gorello P, Cazzaniga G, Alberti F, et al. Quantitative assessment of minimal residual disease in acute myeloid leukemia carrying nucleophosmin (NPM1) gene mutations. Leukemia. 2006;20(6):1103-1108. Ossenkoppele GJ, van de Loosdrecht AA, Schuurhuis GJ. Review of the relevance of aberrant antigen expression by flow cytometry in myeloid neoplasms. Br J Haematol. 2011;153(4):421-436. Terwijn M, van Putten WL, Kelder A, et al. High prognostic impact of flow cytometric minimal residual disease detection in acute myeloid leukemia: data from the HOVON/SAKK AML 42A study. J Clin Oncol. 2013;31(31):3889-3897. Gabert J, Beillard E, van der Velden VH, et al. Standardization and quality control studies of ‘real-time’ quantitative reverse transcriptase polymerase chain reaction of fusion gene transcripts for residual disease detection in leukemia - a Europe Against Cancer program. Leukemia. 2003;17(12):2318-2357. Jourdan E, Boissel N, Chevret S, et al; French AML Intergroup. Prospective evaluation of gene mutations and minimal residual disease in patients with core binding factor acute myeloid leukemia. Blood 2013;121 (12):2213-2223. Grimwade D, Jovanovic JV, Hills RK, et al. Prospective minimal residual disease monitoring to predict relapse of acute promyelocytic leukemia and to direct pre-emptive arsenic trioxide therapy. J Clin Oncol. 2009;27(22):3650-3658. Walter RB, Buckley SA, Pagel JM, et al. Significance of minimal residual disease before myeloablative allogeneic hematopoietic cell transplantation for AML in first and second complete remission. Blood. 2013;122(10):1813-1821. Walter RB, Gooley TA, Wood BL, et al. Impact of pretransplantation minimal residual disease, as detected by multiparametric flow cytometry, on outcome of myeloablative hematopoietic cell transplantation for acute myeloid leukemia. J Clin Oncol. 2011;29(9):1190-1197. Buckley SA, Appelbaum FR, Walter RB. Prognostic and therapeutic implications of minimal residual disease at the time of transplantation in acute leukemia. Bone Marrow Transplant. 2013;48(5):630-641. Walter RB, Gyurkocza B, Storer BE, et al. Comparison of minimal residual disease as outcome predictor for AML patients in first complete remission undergoing myeloablative or nonmyeloablative allogeneic hematopoietic cell transplantation. Leukemia. 2015;29(1):137-144. Kanate AS, Pasquini MC, Hari PN, Hamadani M. Allogeneic hematopoietic cell transplant for acute myeloid leukemia: Current state in 2013 and future directions. World J Stem Cells. 2014;6(2):69-81. Gratwohl A, Stern M, Brand R, et al. Risk score for outcome after allogeneic hematopoietic stem cell transplantation. Cancer. 2009;115(20):4715–4726. Sorror ML, Maris MB, Storb R, et al. Hematopoietic cell transplantation (HCT)specific comorbidity index: a new tool for risk assessment before allogeneic HCT. Blood. 2005;106(8):2912-2919.

ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Hematopoiesis

Ferrata Storti Foundation

The role of CD44 in fetal and adult hematopoietic stem cell regulation

Huimin Cao,1,2 Shen Y. Heazlewood,1,2 Brenda Williams,1,2 Daniela Cardozo,1,2 Julie Nigro,1 Ana Oteiza,3 and Susan K. Nilsson1,2

Manufacturing, Commonwealth Scientific and Industrial Research Organization (CSIRO), Melbourne, Australia; 2Australian Regenerative Medicine Institute, Monash University, Melbourne, Australia; and 3Department of Medical Biology, Faculty of Health Sciences, University of Tromsø, Norway

Haematologica 2016 Volume 101(1):26-37

ABSTRACT

Correspondence: susie.nilsson@csiro.au

Received: 1/9/2015. Accepted: 4/11/2015. Pre-published: 6/11/2015. doi:10.3324/haematol.2015.135921

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/1/26

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

hroughout development, hematopoietic stem cells migrate to specific microenvironments, where their fate is, in part, extrinsically controlled. CD44 standard as a member of the cell adhesion molecule family is extensively expressed within adult bone marrow and has been previously reported to play important roles in adult hematopoietic regulation via CD44 standard-ligand interactions. In this manuscript, CD44 expression and function are further assessed and characterized on both fetal and adult hematopoietic stem cells. Using a CD44–/– mouse model, conserved functional roles of CD44 are revealed throughout development. CD44 is critical in the maintenance of hematopoietic stem and progenitor pools, as well as in hematopoietic stem cell migration. CD44 expression on hematopoietic stem cells as well as other hematopoietic cells within the bone marrow microenvironment is important in the homing and lodgment of adult hematopoietic stem cells isolated from the bone/bone marrow interface. CD44 is also involved in fetal hematopoietic stem cell migration out of the liver, via a process involving stromal cell-derived factor-1α. The absence of CD44 in neonatal bone marrow has no impact on the size of the longterm reconstituting hematopoietic stem cell pool, but results in an enhanced long-term engraftment potential of hematopoietic stem cells.

Introduction CD44, a cell adhesion molecule (CAM) encoded by one gene, but with more than 20 isoforms, is produced by alternative mRNA splicing and/or post-translational modifications.1 CD44 standard (CD44s) is the most abundant isoform, expressed by most mammalian cells2 and involved in hyaluronan uptake and degradation,3 angiogenesis,4 wound healing,5 tissue formation and patterning.2 CD44s is also a critical signaling receptor.6 In contrast, CD44 variants (CD44v) are up-regulated in neoplasia,7 and involved in tumor metastasis8 and aggression.9 Via different binding domains and regulated by post-translational modifications, CD44s binds the extracellular matrix (ECM) components hyaluronan,10 collagen,11 fibronectin,12 as well as transmembrane receptors such as E-selectin.13Among these, hyaluronan is the most common ligand. Although many cells express high levels of CD44s, they do not constitutively bind hyaluronan, with N-glycosylation14 and carbohydrate-sulfation15 each independently modulating binding capacity. Hyaluronan is synthesized by three hyaluronan synthases: HAS1, HAS2 and HAS3 at the inner face of the plasma membrane and extruded to the outer surface, where it is secreted. CD44 is the most common receptor for hyaluronan, and hyaluronan uptake and degradation occurs in a CD44-dependent manner.16,17 Hyaluronan is an important component of the hematopoietic stem cell (HSC) niche,18,19 participating in HSC lodgment in the endosteal region as well as functioning in HSC proliferation and differentiation. haematologica | 2016; 101(1)

The role of CD44 in HSC regulation

Figure 1. CD44 expression by stem and progenitor cells during development. (A) Flow cytometric analysis of LSK (black), LSKCD150+CD48- (red) and LSKCD150+CD48+ (blue) cells. (B) Representative histograms of CD44 expression on LSK, LSKCD150+CD48- and LSKCD150+CD48+ cells from (A) and unstained (gray) control. (C) The level of CD44 expression on LSK, LSKCD150+CD48- and LSKCD150+CD48+ cells during ontogeny. Each dot is an individual adult animal or a pool of 2 or more embryos or newborn pups. Data are representative from 2 biological repeats. MFI: mean fluorescence intensity; E: endosteal; C: central. *P<0.05, **P<0.01. Data show mean±SEM.

Furthermore, it has been reported that CD44, hyaluronan and stromal cell-derived factor-1 (SDF-1) interaction affects HSC and progenitor cell trafficking.20 Hence, CD44s is anticipated to be involved in hematopoietic regulation. Previously, important roles for CD44s in adult hematopoiesis had been identified. In mice, blocking CD44s inhibits lymphopoiesis in long-term bone marrow (BM) cultures,21 T-precursor trafficking to the thymus and lymph nodes,22 and memory cell activation.23 Furthermore, CD44s expression is down-regulated during myeloid and erythroid development24 and involved in the retention of hematopoietic progenitors in BM and spleen.25 Similarly, CD44s plays a role in human hematopoietic regulation, including lymphocyte migration22,26,27 and activation,28-30 as well as progenitor cell proliferation and homing to BM.31 CD44 has been shown to regulate HSC and their BM microenvironment by influencing: 1) matrix assembly; 2) cytokine/chemokine capture and/or release; 3) cytoskeletal linker protein binding (eg. ankyrin, ezrin, radixin and moesin) and signal transduction; and 4) matrix degradation via protease production to influence HSC adhesion, homing, migration, quiescence, resistance to oxidative stress as well as mobilization (reviewed by Zoller32). These networks are extremely complex and depend on CD44 post-translational modifications. In this manuscript, we further investigated the role of CD44 on highly enriched HSC (Lineage-Sca-1+cKit+CD150+CD48– cells) and extended our studies to embryonic hematopoiesis, as CD44s is known to be extensively expressed in fetal hematopoietic organs33,34 but their roles are poorly understood.

wild-type (WT) controls were bred on the most common genetic mouse strain (C57BL/6J, CD45.2) at Monash Animal Research Platform (MARP) (Monash University, Australia). Protein tyrosine phosphatase receptor type C (PTPRCA, CD45.1) mice were purchased from the Animal Resources Centre (Perth, Australia). C57BL/6J and PTPRCA mice are congenic strains, genetically identical except for the CD45 locus. Therefore, when transplanted together, specific CD45.1 and CD45.2 antibodies allow the specific contribution of each donor population to be determined. Adult mice were 6-8 weeks old and sex-matched. Non-ablated recipients and WT BM carrier cells were used in homing and spatial distribution assays, while irradiated recipients and BM carrier cells were used for long-term transplants. Recipients were irradiated using a split dose (4.5Gy each), 4-5 h apart, 24 h prior to transplant and carrier cells exposed to a single dose of 15Gy on the day of transplant. The MARP ethics committee approved all experiments and institutional and national guidelines for the care and use of laboratory animals were followed. Timed pregnancies were set up late afternoon and separated early the following morning with vaginal plugs designated as 0.5 days (E0.5). Pups were harvested from embryonic day 14.5 (E14.5) until newborn day 9 (d9), as previously described.37

Hematopoietic cell isolation Endosteally and centrally located adult BM HSC and fetal HSC (Lineage–Sca-1+c-Kit+ (LSK) CD150+CD48–) were isolated and analyzed as previously described.37,38 Cell counts were performed on a Sysmex KX-21N automated cell counter (Sysmex, Japan).

Flow cytometry Methods Mice CD44–/– mice, devoid of all CD44 isoforms,35 (Tak Mak, Amgen Institute, Canada), red fluorescent (RFP)36 mice and haematologica | 2016; 101(1)

Flow cytometric analysis used an LSRII and cell sorting used an Influx (Becton Dickinson, NJ, USA) equipped as previously described.39 HSC cells were sorted at approximately 20,000 cells per second and re-analyzed to confirm purity (>95%). CD44 expression was analyzed on hematopoietic stem and progenitor cells using anti-CD44-biotin-streptavidin-V500. 27

H. Cao et al.

Figure 2. The incidence and content of HSC and progenitors in adult bone marrow. Data represent the cell content after normalization using individual mouse weights. Each dot represents an individual animal. WT: closed circle; CD44-/-: open circle; E: endosteal, C: central. *P<0.05, **P<0.01. Data show mean±SEM, n=9 from 3 independent experiments.

Homing and spatial distribution assays Sorted fetal liver HSC or adult BM HSC were stained with CFDA-SE (CFSE when incorporated) or SNARF at 0.5 mM and 1 mM, respectively, as previously described.40 The homing efficiency of donor cells was determined by co-transplanting 1-3x103 CFSE+ CD44–/– HSC and SNARF+ WT HSC with 2x105 unlabeled carrier cells into non-ablated recipients as previously described.39 Homing efficiency was calculated as the number of donor HSC homed to the BM divided by the number of donor HSC transplanted, assuming one iliac crest, femur and tibia contain 15% of total BM cells.41 A homing assay in a non-ablated mouse model42,43 allows for assessment in steady-state hematopoiesis, with ablation disrupting BM integrity44,45 and causing increased cytokine activity46 that drives donor cells to cycle.47 Sequential transplant spatial distribution assays were performed as previously described.40 Specifically, irradiated WT and CD44–/– mice were initially transplanted with either WT or CD44–/– BM cells and allowed to reconstitute for 20 weeks (>95% chimerism). WT or CD44–/– CFSE+ LSK cells with 2x105 unlabeled carrier cells were then transplanted into non-ablated pre-reconstituted mice to assess their lodgment 15 h post-transplant both via flow cytometry as well as histologically in paraffin embedded femur sections.48 The number of cells located in the endosteal (within 12-cell diameters from the bone/BM interface in the diaphysis or in the metaphysis) and central BM regions were counted manually.40

Cell lysis and SDF-1α quantification via ELISA

Whole embryonic livers were dissociated in PBS and separated into supernatant and pellet fractions by centrifugation and the pellet was lysed as previously described.39 E16.5 fetal bones were dissected and cut with fine scissors before lysis, as for the method above. SDF-1α expression was assayed in duplicates from 3 biological repeats at each time point using an ELISA kit (R&D) according to the manufacturer’s instructions and normalized for total protein.

and progenitor was assessed using limiting dilution transplant analysis in RFP recipients with 2x105 irradiated RFP carrier cells (carrier cells were irradiated to provide short-term radioprotection before donor HSC reconstituted the recipient BM). The number of mice failing to reconstitute at each dilution (10, 30, 100 or 300 HSC) was measured and the frequency of long-term multi-lineage reconstituting cells calculated using L-Calc (Stem Cell Technologies, Vancouver, Canada).38,49,50 Multi-lineage reconstitution was detected by staining with Gr-1, Mac-1 (myeloid), B220 and CD3 (lymphoid) antibodies as previously described.38 Mice with more than 1% donor multi-lineage reconstitution were designated positive.

Competitive limiting dilution reconstitution assay The multi-lineage reconstitution potential of CD44–/– d8 BM HSC was assessed using competitive limiting dilution transplant analysis, which compares the hematopoietic potential of one population with another.38,51,52 A serial dilution of 10, 30, 100, 300 CD44–/– (CD45.2) LSKCD150+CD48– cells were competed with 100 PTPRCA (CD45.1) WT LSKCD150+CD48– cells and 2x105 irradiated RFP (red fluorescent CD45.2) carrier cells in irradiated RFP recipients were used to compare the hematopoietic potential of CD44–/– HSC in relation to WT HSC. Analysis was assessed as described above, except with the addition of anti-CD45.2 and anti-CD45.1 to discriminate donor populations.

Statistical analysis and data presentation Flow cytometric analysis was performed using FlowJo and Sigma Plot for data comparison. Student’s t-test, Mann-Whitney rank sum test, one-way ANOVA or one-way ANOVA on ranks was used to determine statistical significance as appropriate. A Kaplan-Meier and log rank test were used to assess survival.

Results

Limiting dilution reconstitution assay

HSC express CD44 throughout development, but at lower levels than progenitors

The frequency of long-term multi-lineage reconstituting cells in d8 CD44–/– and WT BM HSC or adult CD44–/– and WT BM stem

Fetal and adult BM LSK cells, HSC (LSKCD150+CD48–) and restricted progenitors (LSKCD150+CD48+) (Figure 1A)

haematologica | 2016; 101(1)

The role of CD44 in HSC regulation

Figure 3. The incidence and content of HSC in fetal and newborn hematopoietic organs. Each dot represents a pool of 2 or more embryos or newborn pups for each time point different. *P<0.05, **P<0.01. Line shows mean value, n≥3.

were analyzed for the presence of CD44 (Figure 1B). CD44 expression was significantly reduced on HSC (1.52-fold) compared to LSK and LSKCD150+CD48+ cells (P<0.05) (Figure 1C). Furthermore, CD44 expression on adult BM HSC was significantly higher when they were located in the endosteal region compared to those in the central BM region (P<0.01) (Figure 1C, inset). Previously we have shown that these two populations are identical in their LSKCD150+CD48– phenotype, but those isolated from the endosteal region have a higher hematopoietic potential.38 haematologica | 2016; 101(1)

CD44–/– mice are consistently smaller than WT, but the absence of CD44 does not alter their lineage composition From E16.5 onwards, including during adulthood, CD44–/– mice consistently weighed significantly less than age- and sex-matched WT controls (Online Supplementary Figure S1A). After normalizing for weight, CD44–/– fetal liver (FL) cellularity was significantly higher at E14.5 and E16.5 (Online Supplementary Figure S1B). In contrast, normalized fetal and adult BM as well as fetal spleen cellularities were comparable between CD44–/– and WT mice 29

H. Cao et al. A

Figure 4. The role of CD44 in FL HSC homing and SDF-1α expression in fetal tissue. (A) Homing efficiency of FL HSC to adult BM. WT: black bar; CD44-/-: white bar. (B) SDF-1α in FL and (C) E16.5 fetal bones. WT: closed circle; CD44-/-: open circle. Data are the mean±SEM, n=3. **P<0.01. NS: not significant.

(Online Supplementary Figure S1B). Normalized values were used for all subsequent calculations of cell content. Furthermore, lineage analysis of fetal and adult hematopoietic organs revealed no significant differences in the proportions of granulocytes (Gr-1+ in fetal53 and Gr-1+/Mac-1+ in adult), T cells (CD3+) or B cells (B220+) in CD44–/– and WT mice (Online Supplementary Figure S1C).

The absence of CD44 significantly alters HSC and progenitor pools in fetal and adult hematopoietic organs in a developmental age dependent manner The incidence and number of HSC and progenitors were assessed in the absence of CD44, providing important but different information; with a number of examples occurring where one is altered but the other is not affected or is altered in the opposite direction due to changes in total cellularity.54,55 In adult BM, CD44 acts as a negative regulator with a significant increase in both the proportion and number of LSK, LSKCD150+CD48– and LSKCD150+CD48+ cells in CD44–/– mice compared to WT controls (P<0.05) (Figure 2). Furthermore, cell cycle analysis of adult CD44–/– BM LSK cells revealed no changes in cell cycle (n=3) (data not shown). In FL, the absence of CD44 resulted in increased frequencies and numbers of HSC (Figure 3A), LSK (Online Supplementary Figure S2A) and LSKCD150+CD48+ cells (Online Supplementary Figure S3A) at E16.5 up until d3 after birth. Furthermore, the increased CD44–/– FL HSC pool was accompanied by a delayed depletion of HSC from the liver, with a HSC content of zero being detected at d2 in WT pups, in contrast to d4 in CD44–/– pups (Figure 3A). Similarly to that evident in the adult, cell-cycle analysis revealed no changes in E16.5 CD44–/– FL LSK cells compared to their WT counterparts (n=3) (data not shown). In CD44–/– pups, fetal BM had a significantly lower incidence and number of HSC from d0-d4 (Figure 3B). In contrast, the absence of CD44 had limited impact on BM progenitors (LSK and LSKCD150+CD48+), with a significant difference in their content only detected at d2 (Online Supplementary Figures S2B and S3B). The absence of CD44 in the spleen resulted in a significantly lower incidence and number of HSC at d0 and d2 (Figure 3C). Collectively, the absence of CD44 significantly altered the HSC and progenitor pools in fetal and adult hematopoietic organs in a developmental age dependent manner.

CD44 is implicated in the migration of fetal HSC from liver to BM via changes in SDF-1α expression

With no evident changes to cell cycle, a proliferative

defect of CD44–/– FL HSC is unlikely to explain the observed augmented HSC pools or the delayed HSC depletion in FL (Figure 3A). However, the accompanying lower numbers of HSC in the BM and spleen at the same developmental ages (Figure 3B and C) suggest CD44 is involved in fetal HSC migration. As a consequence, the absence of CD44 on the homing efficiency of fetal FL HSC was assessed. No differences were observed between CD44–/– and WT FL HSC homing to adult BM (Figure 4A), suggesting the absence of CD44 does not result in an intrinsically altered homing ability. Alternatively, the impact on HSC migration in the absence of CD44 could be the consequence of changes in the FL microenvironment. As the CXCR4/SDF-1α pathway has been demonstrated to be important in HSC migration during development,56 CXCR4 and SDF-1α expression was examined. Analysis of the expression of CXCR4 on CD44–/– and WT FL HSC revealed no differences at any of the time points assessed (data not shown). In FL, SDF-1α was exclusively detected in the cellular fraction. In WT FL, the concentration of SDF-1α significantly decreased from E14.5 to E17.5. In contrast, no significant changes in FL SDF-1α concentration was evident over this period in the absence of CD44, resulting in a significantly greater amount of SDF-1α in CD44–/– FL at E16.5 and E17.5 (Figure 4B). However, analysis of the concentration of SDF-1α in E16.5 fetal bones found no difference between CD44–/– and WT animals (Figure 4C). This organ and developmental age specific change in SDF-1α levels directly correlated with the delayed HSC migration from FL.

CD44 is important for adult HSC homing and lodgment As the presence of CD44 was important for the migration of FL HSC, a potential role in the homing of adult BM HSC was also assessed (Figure 5A). The presence of CD44 on both donor endosteal HSC and in the recipient microenvironment was critical for homing of endosteal HSC, with its absence from either resulting in a significant decrease in homing efficiency (Figure 5B). However, the presence of CD44 on donor HSC or in the microenvironment had no impact on the ability of central HSC to home to the BM (Figure 5B). Importantly, significantly increased homing of WT endosteal HSC into a WT animal was observed compared to the homing of HSC isolated from the central BM region (Figure 5B), which is in agreement with that previously published.38 After homing to the BM, HSC and progenitors lodge and reside in specific BM microenvironments. The role of CD44 on hematopoietic progenitors lodging in the haematologica | 2016; 101(1)

The role of CD44 in HSC regulation

Figure 5. CD44 is critical in adult HSC homing to BM and lodgment in the endosteal region. (A) Schematic presentation of homing assay. (B) Homing efficiency. Each dot represents an individual recipient and different colors represent independent experiments. *P<0.05. Data show mean±SEM, n≥3. (C) Schematic presentation of spatial distribution assay into primary reconstituted recipients with different microenvironments (HM) and hematopoietic cells (H). The spatial distribution of total BM LSK cells was assessed. (D) Centrally and endosteally located CFSE+ donor cells in recipient BM sections. (E) Proportion of donor cells located in the endosteal BM region. *P<0.05, **P<0.01. Data show mean±SEM, n=5.

endosteal BM region was assessed using a sequential transplant spatial distribution assay which allows the evaluation of the role of CD44, either expressed in the microenvironment (HM) or on hematopoietic cells (H) (Figure 5C and D). Consistent with our previous findings,40 approximately 55% of WT donor cells lodged in the endosteal region 15 h post transplant (Figure 5E). The absence of CD44 only in the recipient HM did not significantly change the spatial distribution of donor WT cells. When WT or CD44–/– LSK cells were transplanted into WT haematologica | 2016; 101(1)

or CD44–/– mice reconstituted with CD44–/– BM (H), the proportion of cells homing to the endosteal region was significantly reduced (≥25% of WT cells into WT mice) (Figure 5E). Furthermore, the absence of CD44 on both the recipient BM and recipient HM resulted in an even greater reduction (approx. 70%) of WT cells lodging in the endosteal region (Figure 5E). Together the data suggest that CD44 expressed by adult HSC and recipient BM cells, rather than in the recipient HM, is more critical in the lodgment of transplanted hematopoietic progenitors. 31

H. Cao et al.

Figure 6. CD44 on adult BM LSK cells has a functional role in supporting stem and progenitor engraftment. A serial dilution of 300, 1000, 3000 and 10000 WT or CD44-/- LSK cells isolated from total BM were transplanted into WT or CD44-/- recipients and assessed 20 weeks post transplant for multi-lineage reconstitution. Data show mean±SEM, n=5.

CD44 in adult BM has a functional role in supporting stem and progenitor engraftment but does not change the frequency of long-term multi-lineage reconstituting HSC Analysis of adult stem and progenitor potential in a limiting dilution assay in vivo demonstrated a critical requirement for CD44 on both the donor LSK cells as well as in the recipient microenvironment (Figure 6). When CD44 was absent from the donor cells, less than 50% chimerism was obtained 20 weeks following a transplant of 1000 cells compared to 100% donor reconstitution following a transplant of equivalent numbers of WT cells. Furthermore, when WT LSK cells were transplanted into a CD44–/– recipient, significantly more donor cells were required to obtain 100% donor reconstitution 20 weeks post transplant compared to WT cells transplanted into a WT recipient. Although there was a functional role for CD44 in the regulation of HSC potential, the absence of CD44 did not change the frequency of long-term multi-lineage reconstituting HSC in adult LSK cells. Following a transplant of CD44–/– LSK cells into a WT recipient, the frequency of long-term multi-lineage reconstituting HSC was 1:181 (95% confidence interval 1 in 108 to 1 in 304, n=5), equivalent to transplants of WT LSK cells into a WT recipient (1:277, 95% confidence interval 1 in 176 to 1 in 437, n=5) or WT cells into a CD44–/– recipient (1:391, 95% confidence interval 1 in 249 to 1 in 615, n=5).

The absence of CD44 in neonatal BM also has no impact on the number of the long-term multi-lineage reconstituting HSC, but results in an enhanced long-term engraftment potential We examined the impact of the absence of CD44 on the number of HSC with long-term reconstitution capability in vivo (Figure 7). To assess the frequency of long-term multi-lineage reconstituting HSC, a limiting dilution longterm transplant was performed using CD44–/– and WT d8 newborn BM HSC (Figure 7A). No differences in the survival (Figure 7B), or donor chimerism (Figure 7C) were evident between mice transplanted with limiting numbers of 32

WT or CD44–/– HSC. In addition, an equivalent frequency of long-term multi-lineage reconstituting HSC was detected in CD44–/– (1:17, 95% confidence interval 1 in 8 to 1 in 38) and WT (1:17, 95% confidence interval 1 in 7 to 1 in 39) newborn d8 HSC, which was also equivalent to adult endosteal HSC (1:12, 95% confidence interval 1 in 7 to 1 in 20) (Figure 7D). Together with the observed equivalent number of HSC observed in CD44–/– and WT d8 BM (Figure 3B), the data suggest the absence of CD44 does not alter the long-term multi-lineage reconstituting HSC pool in newborn BM. The role of CD44 in long-term multi-lineage reconstituting HSC potential was then investigated by competing CD44–/– HSC with WT HSC in a competitive long-term limiting dilution transplant (Figure 7E). Analysis of recipient peripheral blood (PB) at 6 and 12 weeks post transplant revealed equivalent hematopoietic potential, with the expected proportions of donor contribution (Figure 7F). However, at 20 weeks, the proportion of donor CD44–/– cells was significantly higher than expected when equivalent numbers of CD44–/– and WT cells were transplanted (61±5 and 39±5%, respectively) (Figure 7F). Importantly, analysis of BM also revealed an increased contribution from CD44–/– cells (Figure 7F). Furthermore, lineage commitment analysis demonstrated that when equivalent numbers of CD44–/– and WT cells were transplanted, CD44–/– donor cells resulted in a significantly higher proportion of B cells at all time points and more myeloid cells at 20 weeks in both PB and BM (Figure 7G). Together, the data suggest CD44–/– HSC have higher long-term engraftment potential than their WT counterparts.

Discussion Previous studies had shown CD44 is highly expressed on adult HSC57 and early erythroid progenitors, but down-regulated during lineage commitment.24 In addition, adult BM cells with colony-forming-unit granulocytes-macrophage and burst-forming unit-erythroid are haematologica | 2016; 101(1)

The role of CD44 in HSC regulation

predominantly found in the CD44bright population.58 The current analysis extended these findings to reveal CD44 expression on HSC is significantly reduced compared to that detected on progenitor cells throughout development. Furthermore, in adult BM, CD44 is expressed at a higher level on HSC isolated from the endosteal BM region compared to those isolated from the central BM. Although there is no difference in the incidence of HSC

between endosteal and central BM, due to the higher total cellularity in central BM, the total number of HSC in this region is significantly higher than in the endosteal region.38 CD44–/– mice are born healthy and fertile with no apparent developmental disorders.35 Previous studies have reported CD44–/– mice develop normal organ morphology with normal nucleated cell counts in hematopoi-

Figure 7. Neonatal CD44-/- stem cell pool has equivalent numbers of long-term multi-lineage reconstituting HSC but higher long-term engraftment potential compared to WT counterparts. (A) Limiting dilution transplant analysis of D8 BM LSKCD150+CD48– cells. (B) Survival analysis: WT 10 (red), CD44-/- 10 (pink), WT 30 (purple), CD44-/- 30 (green), WT 100 (dark blue), CD44-/- 100 (light blue), WT 300 (black), and CD44-/- 300 (orange). (C) Peripheral blood (PB) and BM reconstitution from CD44-/- (white bars) and WT (black bars) donor HSC. (D) Frequency of long-term multi-lineage reconstituting HSC. Each symbol represents a group of 5 mice. (E) Competitive limiting dilution transplant analysis of LSKCD150+CD48– cells. (F) PB and BM reconstitution from CD44-/- donor HSC. (G) CD44-/- (white bars) and WT (black bars) donor lymphoid and myeloid reconstitution following a transplant of 100 CD44-/- and 100 WT donor cells. Data are mean±SEM, n=5. *P<0.05.

haematologica | 2016; 101(1)

H. Cao et al. etic tissues.35,59 However, our CD44–/– mice are consistently smaller than their WT counterparts. These differences may be due to differences in mouse strains, with Schmits et al. using heterozygous CD44+/– mice as controls35 and Oostendorp et al. using CD44–/– mice on a PTPRCA background,59 whilst our CD44–/– mice were bred on a C57/Bl6 background. However, despite the reduced body weight of CD44–/– mice, normalization for body weight revealed equivalent organ cellularities from E16.5. Furthermore, analysis of lineage commitment in CD44–/– mice did not

show any evidence of lineage skewing in fetal or adult hematopoietic organs. However, the data demonstrate CD44 functions in maintaining the HSC and progenitor pools in a developmental age and location dependent manner. In adulthood, CD44–/– BM had significantly increased stem and progenitor cell pools. Together the data suggest CD44 acts as a negative regulator of HSC proliferation, which is potentially regulated through the CD44/hyaluronan interaction. Previously, hyaluronan had been shown to be highly

SDF-1 /CXCR4 CD44/SDF-1/HA

SDF-1 /CXCR4 SDF-1/HA

Figure 8. Schematic presentation summarizing the role of CD44 in fetal and adult hematopoiesis. (A) Proposed model for the mechanism of the role of CD44 in the migration of HSC and progenitors out of the FL. (B) Observed changes in fetal and adult hematopoiesis in the absence of CD44.

haematologica | 2016; 101(1)

The role of CD44 in HSC regulation

expressed at the endosteal BM surface as well as acting as a negative regulator of HSC proliferation.18,48 This could also explain why endosteal HSC have significantly higher CD44 expression compared to centrally isolated HSC. Previous analysis of HSC devoid of CD44 revealed no functional changes, with CD44–/– adult BM and FL cells having equivalent capacity to initiate long-term marrow cultures in vitro59 and form d12 spleen colonies in vivo.35 These findings are consistent with our data, which showed an equivalent frequency of long-term multi-lineage reconstituting HSC in CD44–/– and WT newborn BM HSC as well as CD44–/– and WT adult LSK cells. However, the use of a HSC competitive limiting dilution transplant assay revealed CD44–/– neonatal (d8) HSC have higher engraftment potential than WT controls. This finding is in contrast to that previously published by Oostendorp et al.59 who showed equivalent frequency of competitive repopulating units in adult CD44–/– BM. The difference may reflect the roles of CD44 in fetal and adult HSC regulation, although it is known that while fetal HSC expand from E14.560 they have acquired adult characteristics by one week after birth (d8).61 Alternatively, these differences may be due to the assay conditions used. In this study, purified HSC were assessed, while Oostendorp et al.59 used total BM. In addition, Waterstrat et al.62 have suggested that, due to superior homing ability, CD45.2 cells naturally have increased engraftment potential compared to CD45.1 cells; a possible contributing explanation to our results. However, this difference was not evident until 30 weeks post transplant, whereas we detected a significant difference in engraftment potential by 20 weeks post transplant, making this unlikely to be the sole explanation. Furthermore, we demonstrate CD44 is an important molecule in adult HSC homing. The reduced homing was evident when CD44 was absent on endosteal HSC or in the recipient microenvironment and could be due to the interaction between CD44, hyaluronan and SDF-1 in anchoring homed hematopoietic progenitor cells.20 We have previously demonstrated that HSC isolated from the endosteal region home to BM better than their central counterparts as well as preferentially home back to the endosteal region,38 and that hyaluronan is highly expressed in the endosteal BM region compared to central BM.18,48 Therefore, without CD44, HSC and progenitors may have impaired interactions with their microenvironment during the homing process. Furthermore, hyaluronan has been shown to act as a chemoattractant via CD4463 and we did not find any differences in the expression of HAS3, critical in the generation of hyaluronan, in CD44–/– mouse BM (data not shown). Together, these data suggest an extrinsic effect on HSC in the absence of CD44. While Khaldoyanidi et al.31 had previously demonstrated a key role for CD44 in the homing of progenitor cells, Oostendorp et al.59 did not. The discrepancy between results could be due to assay conditions: Oostendorp et al.59 analyzed the homing of whole BM cells transplanted into ablated recipients, with altered cytokine release,46 cell cycling,47 and vascular permeability post irradiation,44,45 while we used purified HSC separated based on their anatomical location and demonstrated a significant reduced homing of both endosteal CD44–/– HSC into WT steady-state, non-ablated recipients as well as WT HSC into CD44–/– steady-state, non-ablated recipients. Because haematologica | 2016; 101(1)

only one-quarter to one-third of HSC are endosteal in origin,38,64 simply assaying total BM HSC, let alone total BM, could miss the effect we identified. In further support of a role for CD44 in progenitor cell homing and migration, Protin and colleagues65 found CD44–/– T-cell progenitors showed a homing defect to the thymus and lymph nodes as well as a dysregulation of progenitor cell mobilization with a significant decrease in the number of CFU-GM in the spleen following G-CSF administration.35 The current data extend our understanding of fetal hematopoiesis and highlight a role for CD44 in fetal HSC anchoring and migration, with a proposed model of the underpinning mechanisms described in Figure 8A. During embryogenesis, HSC initially migrate to FL before colonizing the spleen and BM later in development. Migration is essential for fetal hematopoiesis as multiple sites allow different inductive signals from the microenvironments to support hematopoietic development.66 We show that, in WT mice, FL SDF-1α concentration progressively decreased from E14.5, with HSC migrating from the FL to the BM after E16.5. The accumulated data in other in vivo and in vitro contexts, including the homing of HSC and progenitors to BM,20 suggest this occurs due to a co-ordinated sequence of events including: 1) a polarized source of SDF-1 developing; 2) the cells producing projections extending towards the SDF-1, with cell-associated CD44 migrating into the projections; and 3) CD44 binding to an external source of hyaluronan and the cells undergoing trans-endothelial migration. However, in the absence of CD44, the concentration of FL SDF-1α remained unchanged from E14.5 to E17.5 and HSC and progenitors were not released from the FL microenvironment to migrate to the BM (Figure 8A). These data are supported by Petit et al.67 who demonstrated that decreased SDF-1 levels result in HSC detachment and migration from their microenvironment. The major source of SDF-1 in E15.5 mouse FL is hepatic stem/progenitor (stromal) cells68 and it has been shown that SDF-1–/– HSC are not retained in the FL but migrate into the PB.56 Furthermore, genetic silencing of CD44 expressed by endothelial cells increased SDF-1 mRNA expression via NFκB.69 This suggests the accumulation of SDF-1α in CD44–/– FL is the direct result of the absence of CD44, with a resulting increase in HSC and progenitors that remain firmly attached to the FL microenvironment, delaying their migration to the BM. Our data demonstrate that this increase in HSC and progenitors was not due to changes in cell cycle or due to an intrinsic homing defect. In addition, examination of CXCR4 on WT and CD44–/– FL HSC and progenitors at E16.5 revealed no differences. This finding is also in agreement with previous studies reporting that while human HSC and progenitors pre-incubated with CD44 blocking antibody had impaired homing ability in vivo as well as decreased migration towards a SDF-1 gradient in vitro, this was not due to the downregulation of CXCR4 expression.20 In summary, the current study demonstrates a number of roles of CD44 in hematopoiesis throughout development (Figure 8B). CD44 is highly expressed on HSC and progenitors and regulates their pools in situ. Furthermore, CD44 participates in HSC anchoring and migration via hyaluronan and SDF-1α, with roles in adult HSC homing and lodgment in the BM post transplant as well as in fetal HSC migration out of the liver. As an in depth study of the roles of key molecular interactions during hematopoietic 35

H. Cao et al.

development, the current study is critical for both our understanding and the treatment of hematologic disorders, as well as the translation of this to the derivation of HSC from human embryonic stem cells and/or induced pluripotent stem cells. Acknowledgments The authors would like to thank Andrea Reitsma, Jessica Hatwell-Humble, Rebecca Neaves, the staff at Monash Animal Research Platform and FlowCore for excellent technical help and

References 1. Goodison S, Urquidi V, Tarin D. CD44 cell adhesion molecules. Mol Pathol. 1999;52(4):189-196. 2. Wheatley SC, Isacke CM, Crossley PH. Restricted expression of the hyaluronan receptor, CD44, during postimplantation mouse embryogenesis suggests key roles in tissue formation and patterning. Development. 1993;119(2):295-306. 3. Culty M, Nguyen HA, Underhill CB. The hyaluronan receptor (CD44) participates in the uptake and degradation of hyaluronan. J Cell Biol. 1992;116(4):1055-1062. 4. Trochon V, Mabilat C, Bertrand P, et al. Evidence of involvement of CD44 in endothelial cell proliferation, migration and angiogenesis in vitro. Int J Cancer. 1996;66(5):664-668. 5. Jain M, He Q, Lee WS, et al. Role of CD44 in the reaction of vascular smooth muscle cells to arterial wall injury. J Clin Invest. 1996;98(3):877. 6. McKee CM, Penno MB, Cowman M, et al. Hyaluronan (HA) fragments induce chemokine gene expression in alveolar macrophages. The role of HA size and CD44. J Clin Invest. 1996;98(10):2403-2413. 7. Khaldoyanidi S, Achtnich M, Hehlmann R, et al. Expression of CD44 variant isoforms in peripheral blood leukocytes in malignant lymphoma and leukemia: inverse correlation between expression and tumor progression. Leuk Res. 1996;20(10):839-851. 8. Gunthert U, Hofmann M, Rudy W, et al. A new variant of glycoprotein CD44 confers metastatic potential to rat carcinoma cells. Cell. 1991;65(1):13-24. 9. Birch M, Mitchell S, Hart IR. Isolation and characterization of human melanoma cell variants expressing high and low levels of CD44. Cancer Res. 1991;51(24):6660-6667. 10. Aruffo A, Stamenkovic I, Melnick M, et al. CD44 is the principal cell surface receptor for hyaluronate. Cell. 1990;61(7):1303-1313. 11. Faassen AE, Schrager JA, Klein DJ, et al. A cell surface chondroitin sulfate proteoglycan, immunologically related to CD44, is involved in type I collagen-mediated melanoma cell motility and invasion. J Cell Biol. 1992;116(2):521-531. 12. Verfaillie CM, Benis A, Iida J, et al. Adhesion of committed human hematopoietic progenitors to synthetic peptides from the C-terminal heparin-binding domain of fibronectin: cooperation between the integrin alpha 4 beta 1 and the CD44 adhesion receptor. Blood. 1994;84(6):1802-1811. 13. Dimitroff CJ, Lee JY, Rafii S, et al. CD44 is a major E-selectin ligand on human hematopoietic progenitor cells. J Cell Biol. 2001;153(6):1277-1286. 14. Lesley J, English N, Perschl A, et al. Variant

15.

16.

17. 18.

19.

20.

21.

22. 23.

24.

25.

26.

27.

support. The authors also would like to acknowledge Shannon L. McKinney-Freeman for help with harvesting embryonic organs, Patrick Tam for providing RFP mice and Ivan Bertoncello for critically assessing the manuscript. Funding This work is in part funded by an SIEF grant to SN. The Australian Regenerative Medicine Institute is supported by grants from the State Government of Victoria and the Australian Government.

cell lines selected for alterations in the function of the hyaluronan receptor CD44 show differences in glycosylation. J Exp Med. 1995;182(2):431-437. Brown KL, Maiti A, Johnson P. Role of sulfation in CD44-mediated hyaluronan binding induced by inflammatory mediators in human CD14(+) peripheral blood monocytes. J Immunol. 2001;167(9):5367-5374. Harada H, Takahashi M. CD44-dependent intracellular and extracellular catabolism of hyaluronic acid by hyaluronidase-1 and -2. J Biol Chem. 2007;282(8):5597-5607. Weigel PH, Hascall VC, Tammi M. Hyaluronan synthases. J Biol Chem. 1997;272(22):13997-14000. Nilsson SK, Haylock DN, Johnston HM, et al. Hyaluronan is synthesized by primitive hemopoietic cells, participates in their lodgment at the endosteum following transplantation, and is involved in the regulation of their proliferation and differentiation in vitro. Blood. 2003;101(3):856-862. Legras S, Levesque JP, Charrad R, et al. CD44-mediated adhesiveness of human hematopoietic progenitors to hyaluronan is modulated by cytokines. Blood. 1997;89(6):1905-1914. Avigdor A, Goichberg P, Shivtiel S, et al. CD44 and hyaluronic acid cooperate with SDF-1 in the trafficking of human CD34+ stem/progenitor cells to bone marrow. Blood. 2004;103(8):2981-2989. Miyake K, Medina KL, Hayashi S, et al. Monoclonal antibodies to Pgp-1/CD44 block lympho-hemopoiesis in long-term bone marrow cultures. J Exp Med. 1990;171(2):477-488. O'Neill HC. Antibody which defines a subset of bone marrow cells that can migrate to thymus. Immunology. 1989;68(1):59-65. Zeng H, Chen Y, Yu M, et al. T cell receptor-mediated activation of CD4+CD44hi T cells bypasses Bcl10: an implication of differential NF-kappaB dependence of naive and memory T cells during T cell receptormediated responses. J Biol Chem. 2008;283(36):24392-24399. Kansas GS, Muirhead MJ, Dailey MO. Expression of the CD11/CD18, leukocyte adhesion molecule 1, and CD44 adhesion molecules during normal myeloid and erythroid differentiation in humans. Blood. 1990;76(12):2483-2492. Vermeulen M, Le Pesteur F, Gagnerault MC, et al. Role of adhesion molecules in the homing and mobilization of murine hematopoietic stem and progenitor cells. Blood. 1998;92(3):894-900. Jalkanen S, Reichert RA, Gallatin WM, et al. Homing receptors and the control of lymphocyte migration. Immunol Rev. 1986;91:39-60. Jalkanen S, Bargatze RF, de los Toyos J, et al. Lymphocyte recognition of high endotheli-

28. 29.

30.

31.

32. 33.

34.

35.

36.

37.

38.

39.

40.

um: antibodies to distinct epitopes of an 85-95-kD glycoprotein antigen differentially inhibit lymphocyte binding to lymph node, mucosal, or synovial endothelial cells. J Cell Biol. 1987;105(2):983-990. Huet S, Groux H, Caillou B, et al. CD44 contributes to T cell activation. J Immunol. 1989;143(3):798-801. Shimizu Y, Van Seventer GA, Siraganian R, et al. Dual role of the CD44 molecule in T cell adhesion and activation. J Immunol. 1989;143(8):2457-2463. Denning SM, Le PT, Singer KH, et al. Antibodies against the CD44 p80, lymphocyte homing receptor molecule augment human peripheral blood T cell activation. J Immunol. 1990;144(1):7-15. Khaldoyanidi S, Denzel A, Zoller M. Requirement for CD44 in proliferation and homing of hematopoietic precursor cells. J Leukoc Biol. 1996;60(5):579-592. Zoller M. CD44, Hyaluronan, the Hematopoietic Stem Cell, and LeukemiaInitiating Cells. Front Immunol. 2015;6:235. Ren AH, Zhang Y, Zhao YL, et al. [Comparative study of differentially expressed genes in mouse bone marrow and fetal liver cells]. Zhongguo Shi Yan Xue Ye Xue Za Zhi. 2003;11(5):444-449. Huang H, Auerbach R. Identification and characterization of hematopoietic stem cells from the yolk sac of the early mouse embryo. Proc Natl Acad Sci USA. 1993;90(21):10110-10114. Schmits R, Filmus J, Gerwin N, et al. CD44 regulates hematopoietic progenitor distribution, granuloma formation, and tumorigenicity. Blood. 1997;90(6):2217-2233. Ellis SL, Heazlewood SY, Williams B, et al. The role of Tenascin C in the lymphoid progenitor cell niche. Exp Hematol. 2013;41(12):1050-1061. Cao H, Williams B, Nilsson SK. Investigating the Interaction Between Hemopoietic Stem Cells and Their Niche During Embryonic Development: Optimizing the Isolation of Fetal and Newborn Stem Cells From Liver, Spleen, and Bone Marrow. In: Bunting K, ed. Hematopoietic Stem Cell Protocols. USA: Humana Press, 2014. Grassinger J, Haylock DN, Williams B, et al. Phenotypically identical hemopoietic stem cells isolated from different regions of bone marrow have different biologic potential. Blood. 2010;116(17):3185-3196. Grassinger J, Haylock DN, Storan MJ, et al. Thrombin-cleaved osteopontin regulates hemopoietic stem and progenitor cell functions through interactions with alpha9beta1 and alpha4beta1 integrins. Blood. 2009;114(1):49-59. Nilsson SK, Johnston HM, Coverdale JA. Spatial localization of transplanted hemopoietic stem cells: inferences for the local-

haematologica | 2016; 101(1)

The role of CD44 in HSC regulation

41. 42.

43.

44.

45.

46.

47.

48.

49.

50.

ization of stem cell niches. Blood. 2001; 97(8):2293-2299. Smith LH, Clayton ML. Distribution of injected 59Fe in mice. Exp Hematol. 1970;20:82-86. Shaaban AF, Kim HB, Milner R, et al. A kinetic model for the homing and migration of prenatally transplanted marrow. Blood. 1999;94(9):3251-3257. Nilsson SK, Dooner MS, Tiarks CY, et al. Potential and distribution of transplanted hematopoietic stem cells in a nonablated mouse model. Blood. 1997;89(11):40134020. Daldrup-Link HE, Link TM, Rummeny EJ, et al. Assessing permeability alterations of the blood-bone marrow barrier due to total body irradiation: in vivo quantification with contrast enhanced magnetic resonance imaging. Bone Marrow Transplant. 2000;25(1):71-78. Shirota T, Tavassoli M. Alterations of bone marrow sinus endothelium induced by ionizing irradiation: implications in the homing of intravenously transplanted marrow cells. Blood Cells. 1992;18(2):197-214. Gaugler MH, Squiban C, Mouthon MA, et al. Irradiation enhances the support of haemopoietic cell transmigration, proliferation and differentiation by endothelial cells. Br J Haematol. 2001;113(4):940-950. Cui J, Wahl RL, Shen T, et al. Bone marrow cell trafficking following intravenous administration. Br J Haematol. 1999;107(4): 895-902. Ellis SL, Grassinger J, Jones A, et al. The relationship between bone, hemopoietic stem cells, and vasculature. Blood. 2011;118(6):1516-1524. Taswell C. Limiting dilution assays for the determination of immunocompetent cell frequencies. I. Data analysis. J Immunol. 1981;126(4):1614-1619. Szilvassy SJ, Fraser CC, Eaves CJ, et al. Retrovirus-mediated gene transfer to purified hemopoietic stem cells with long-term lympho-myelopoietic repopulating ability.

haematologica | 2016; 101(1)

51. 52.

53.

54.

55.

56.

57.

58.

59.

60.

Proc Natl Acad Sci USA. 1989;86(22):87988802. Harrison DE. Competitive repopulation: a new assay for long-term stem cell functional capacity. Blood. 1980;55(1):77-81. Harrison DE, Jordan CT, Zhong RK, et al. Primitive hemopoietic stem cells: direct assay of most productive populations by competitive repopulation with simple binomial, correlation and covariance calculations. Exp Hematol. 1993;21(2):206-219. Morrison SJ, Hemmati HD, Wandycz AM, et al. The purification and characterization of fetal liver hematopoietic stem cells. Proc Natl Acad Sci USA. 1995;92(22):1030210306. Kajiume T, Ninomiya Y, Ishihara H, et al. Polycomb group gene mel-18 modulates the self-renewal activity and cell cycle status of hematopoietic stem cells. Exp Hematol. 2004;32(6):571-578. Qian H, Buza-Vidas N, Hyland CD, et al. Critical role of thrombopoietin in maintaining adult quiescent hematopoietic stem cells. Cell Stem Cell. 2007;1(6):671-684. Ara T, Tokoyoda K, Sugiyama T, et al. Long-term hematopoietic stem cells require stromal cell-derived factor-1 for colonizing bone marrow during ontogeny. Immunity. 2003;19(2):257-267. Nabors LK, Wang LD, Wagers AJ, et al. Overlapping roles for endothelial selectins in murine hematopoietic stem/progenitor cell homing to bone marrow. Exp Hematol. 2013;41(7):588-596. Lewinsohn DM, Nagler A, Ginzton N, et al. Hematopoietic progenitor cell expression of the H-CAM (CD44) homing-associated adhesion molecule. Blood. 1990;75(3):589595. Oostendorp RA, Ghaffari S, Eaves CJ. Kinetics of in vivo homing and recruitment into cycle of hematopoietic cells are organspecific but CD44-independent. Bone Marrow Transplant. 2000;26(5):559-566. Ema H, Nakauchi H. Expansion of hematopoietic stem cells in the developing

61.

62.

63.

64.

65.

66.

67.

68.

69.

liver of a mouse embryo. Blood. 2000;95(7): 2284-2288. Kikuchi K, Kondo M. Developmental switch of mouse hematopoietic stem cells from fetal to adult type occurs in bone marrow after birth. Proc Natl Acad Sci USA. 2006;103(47):17852-17857. Waterstrat A, Liang Y, Swiderski CF, et al. Congenic interval of CD45/Ly-5 congenic mice contains multiple genes that may influence hematopoietic stem cell engraftment. Blood. 2010;115(2):408-417. Tzircotis G, Thorne RF, Isacke CM. Chemotaxis towards hyaluronan is dependent on CD44 expression and modulated by cell type variation in CD44hyaluronan binding. J Cell Sci. 2005;118(Pt 21):5119-5128. Haylock DN, Williams B, Johnston HM, et al. Hemopoietic stem cells with higher hemopoietic potential reside at the bone marrow endosteum. Stem Cells. 2007;25(4):1062-1069. Protin U, Schweighoffer T, Jochum W, et al. CD44-deficient mice develop normally with changes in subpopulations and recirculation of lymphocyte subsets. J Immunol. 1999;163(9):4917-4923. Cao H, Oteiza A, Nilsson SK. Understanding the role of the microenvironment during definitive hemopoietic development. Exp Hematol. 2013; 41(9):761-768. Petit I, Szyper-Kravitz M, Nagler A, et al. GCSF induces stem cell mobilization by decreasing bone marrow SDF-1 and up-regulating CXCR4. Nat Immunol. 2002;3(7): 687-694. Chou S, Lodish HF. Fetal liver hepatic progenitors are supportive stromal cells for hematopoietic stem cells. Proc Natl Acad Sci USA. 2010;107(17):7799-7804. Olofsson B, Porsch H, Heldin P. Knockdown of CD44 regulates endothelial cell differentiation via NFkappaB-mediated chemokine production. PLoS One. 2014;9 (3):e90921.

ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Iron Metabolism & Its Disorders

Ferrata Storti Foundation

Haematologica 2016 Volume 101(1):38-45

Second international round robin for the quantification of serum non-transferrin-bound iron and labile plasma iron in patients with iron-overload disorders

Louise de Swart,1 Jan C.M. Hendriks, 2 Lisa N. van der Vorm, 3 Z. Ioav Cabantchik, 4 Patricia J. Evans, 5 Eldad A. Hod, 6 Gary M. Brittenham, 7 Yael Furman, 8 Boguslaw Wojczyk, 6 Mirian C.H. Janssen, 9 John B. Porter,5 Vera E.J.M. Mattijssen,10 Bart J. Biemond,11 Marius A. MacKenzie,1 Raffaella Origa,12 Renzo Galanello,12* Robert C. Hider,13 and Dorine W. Swinkels3

Departments of Hematology, 2Health Evidence and 3Department of Laboratory Medicine, Radboud University Medical Center, Nijmegen, The Netherlands; 4Department of Biochemical Chemistry, Hebrew University of Jerusalem, Israel; 5Department of Haematology, University College London, UK; 6Department of Pathology and Cell Biology, Columbia University Medical Center, New York, NY, USA; 7Department of Pediatrics, Columbia University Medical Center, New York, NY, USA; 8Aferrix Ltd., Tel-Aviv, Israel; 9 Department of Internal Medicine, Radboud University Medical Center, Nijmegen, The Netherlands; 10Department of Hematology, Rijnstate Hospital, Arnhem, The Netherlands; 11 Department of Hematology, Academic Medical Center, Amsterdam, The Netherlands; 12 Department of Biomedical Science and Biotechnology, Regional Microcythemia Hospital, University of Cagliari, Italy; and 13Institute of Pharmaceutical Science, King’s College London, UK 1

*This work is dedicated to the memory and in honor of Renzo Galanello, who was instrumental for the study until the end of his career

ABSTRACT

Correspondence: dorine.swinkels@radboudumc.nl

Received: 23/7/2015. Accepted: 18/9/2015. Pre-published: 18/9/2015. doi:10.3324/haematol.2015.133983

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/1/38

on-transferrin-bound iron and its labile (redox active) plasma iron component are thought to be potentially toxic forms of iron originally identified in the serum of patients with iron overload. We compared ten worldwide leading assays (6 for non-transferrinbound iron and 4 for labile plasma iron) as part of an international interlaboratory study. Serum samples from 60 patients with four different iron-overload disorders in various treatment phases were coded and sent in duplicate for analysis to five different laboratories worldwide. Some laboratories provided multiple assays. Overall, highest assay levels were observed for patients with untreated hereditary hemochromatosis and β-thalassemia intermedia, patients with transfusion-dependent myelodysplastic syndromes and patients with transfusion-dependent and chelated β-thalassemia major. Absolute levels differed considerably between assays and were lower for labile plasma iron than for non-transferrin-bound iron. Four assays also reported negative values. Assays were reproducible with high between-sample and low withinsample variation. Assays correlated and correlations were highest within the group of non-transferrin-bound iron assays and within that of labile plasma iron assays. Increased transferrin saturation, but not ferritin, was a good indicator of the presence of forms of circulating nontransferrin-bound iron. The possibility of using non-transferrin-bound iron and labile plasma iron measures as clinical indicators of overt iron overload and/or of treatment efficacy would largely depend on the rigorous validation and standardization of assays.

haematologica | 2016; 101(1)

Round robin for NTBI and LPI

Introduction The predominant forms of iron present in living entities are associated with proteins such as transferrin in the major circulating fluid and heme and ferritin in cells. However, the iron binding capacity of the iron transport protein transferrin can be exceeded when a substantial amount of iron enters the circulation due to excessive iron absorption from the diet or release of iron from cell stores. In these conditions non-transferrin-bound iron (NTBI; circulating iron not bound to transferrin, ferritin or heme) appears in the plasma.1-5 Plasma NTBI apparently comprises several subspecies, which may be classified by their chemical composition, chemical reactivity or susceptibility to chelation.6-15 The chemical composition of NTBI is heterogeneous and it is thought that there are several circulating isoforms, that is Fe(III) bound to albumin and citrate and potentially to acetate, malate and phosphate.6-8,14 Of these, citrate has the highest affinity for Fe(III), and under physiological conditions two isoforms dominate, i.e. monomeric and oligomeric Fe(III) complexes.8,16 The fraction of plasma NTBI that is redox active and can be chelated is designated labile plasma iron (LPI).12,13 Iron complexes assumed to represent NTBI have been shown experimentally to be taken up by susceptible cell types, including hepatocytes, cardiomyocytes and pancreatic islet cells, with consequent oxidant injury.5 Imbalances in iron homeostasis are responsible for a variety of disorders. Excess iron accumulates in the circulation and tissues of patients with hereditary hemochromatosis (HH), iron-loading anemias (Î˛-thalassemia major and intermedia), myelodysplastic syndromes (MDS) and sickle-cell disease (SCD) after transfusion.2,3,5 To prevent iron-induced tissue damage, impending iron toxicity must be detected before complications develop and become irreversible. Currently, the most widely adopted method for the detection of iron overload is the measurement of serum ferritin, occasionally combined with transferrin saturation (TSAT). However, it is well known that as an acute phase reactant, serum ferritin levels are influenced by factors such as inflammation and liver disease and are not, therefore, exclusively indicative of toxic parenchymal iron overload.17 Moreover, with the introduction of T2*-weighted magnetic resonance imaging for the assessment of tissue iron overload,18 it became clear that organs such as the heart and endocrine glands load iron differently compared to the liver and non-commensurately with serum ferritin. Studies of plasma NTBI in patients with thalassemia major suggest that NTBI may be an important early indicator of extra-hepatic iron toxicity.19 Studies in patients with various iron-loading disorders have shown reductions in NTBI and LPI upon phlebotomy and chelation therapy and that these reductions are associated with an improved prognosis.2,3,20-23 The results of NTBI and LPI assays are, therefore, promising as therapeutic targets and for the evaluation of iron overload and the efficacy of and compliance with iron-lowering therapies.5,13,24 Due to the complexity and potential clinical importance of NTBI, several assays have been developed for its detection.6,10-12,15,25-29 In our previous round robin 1 we found that NTBI values of patients with HFE-related hemochromatosis differed considerably depending on which of various assays was used.30 We concluded that NTBI assays were insufficiently standardized and the most pertinent assay haematologica | 2016; 101(1)

for clinical applications was uncertain. Since that round robin 1 for NTBI, novel assays have been published and made available for use.12,15,29 Therefore, and as a stepping stone in the path to defining the clinical utility of these assays, the aims of our study were to update round robin 1 and to increase our understanding of the various NTBI and LPI levels measured by the current leading analytical assays in four different groups of iron-overloaded patients (those with HH, thalassemia, MDS, and SCD) undergoing various treatments (phlebotomy, iron chelation, red blood cell transfusion). More specifically, in these populations of patients, we aimed to: (i) establish correlations between assays, (ii) establish levels of reproducibility of each of the assays, (iii) assess levels of the NTBI fraction consisting of iron citrate, and (iv) correlate assays with other established parameters of iron overload (e.g. TSAT and serum ferritin).

Methods Study design and participants We compared ten different assays (5 NTBI, 1 NTBI isoform-specific and 4 LPI) performed in five different laboratories worldwide. Serum samples (n=60) were collected from patients with four iron overload disorders in different treatment phases (transfusiondependent; receiving iron chelation) making a total of ten different patient and treatment groups. Samples were collected with approval from local Institutional Review Boards and in conformity with the code for proper secondary use of human tissues. More detailed information on the collection of samples and laboratory analyses are given in the Online Supplementary Information.

Non-transferrin-bound iron and labile plasma iron assays The NTBI/LPI assays were divided into three different groups: (i) five NTBI assays (N1-N5); (ii) one NTBI isoform-specific assay (N6); and (iii) four LPI assays (L1-L4) (Online Supplementary Table S1). The test principles of the NTBI assays can be divided into two subgroups. The first subgroup (N2-4) consists of a chelation-ultrafiltration-detection approach based on the prior mobilization of serum NTBI by weak iron-mobilizing chelators such as nitrilotriacetate at 80 mM. The chelated NTBI is separated from transferrin-bound iron by ultrafiltration and detected by colorimetry or high performance liquid chromatography.26-28 The second subgroup of NTBI assays (N1 and N5) comprises the measurement of directly chelatable iron. In these assays NTBI is mobilized and detected in the same reaction mixture by iron-binding probes such as fluorescently-labeled desferrioxamine and the hexadentate pyridinone chelator (CP851) attached to a fluorescent probe.6,10,15 In the latter assay the CP851 was bound to beads and the fluorescent signal from the bead-bound chelator was differentiated from the non-specific plasma signals by flow cytometry.6,15 None of the NTBI assays used a cobalt compound to block unsaturated transferrin. The NTBI isoform-specific method N6 measures the sum of the oligomeric and monomeric iron (III)-citrate complexes of NTBI. This methodology consists of gel filtration chromatography by high performance liquid chromatography to separate the individual components of NTBI by mass, followed by quantification of iron by inductively coupled-plasma mass spectrometry (LC-ICP-MS).6 The LPI assays (L1-4) measure the redox active component of NTBI, a potentially toxic species. Assay L1 used bleomycin, which forms a complex with DNA and redox active ferrous iron. When ascorbate is added, reactive oxygen species are generated which degrade the deoxyribose moieties of the DNA in a site-specific reaction. These degradation products bind to thiobarbituric acid to 39

L. de Swart et al.

thalassemia major, MDS or SCD, in thalassemia major patients receiving iron chelation therapy and in HH patients at presentation. The ferritin levels in HH patients decreased after phlebotomy. TSAT levels were similar for the two different TSAT methodologies used and were highest in thalassemia major patients who were transfusion-dependent or receiving iron chelation (median 100%). The median TSAT was also very high (>94%) in patients with HH or thalassemia intermedia at presentation, transfusion-dependent MDS and less increased (>51%) in patients with thalassemia major at presentation, in the depletion phase of phlebotomies in HH patients, in MDS subtypes refractory anemia with ringed sideroblasts/refractory cytopenia with multilineage dysplasia with ringed sideroblasts and in transfusion-dependent SCD. The median TSAT was within the reference range (<45%) for most SCD patients at presentation and HH patients in the maintenance phase.

give a pink chromogen for quantification.25 The assays L2, L3 and L4 used a reducing agent (ascorbate) and an oxidizing agent (atmospheric O2) to generate reactive oxygen species from endogenous oxidants and labile catalytic iron. Reactive oxygen species were detected by an oxidation-sensitive probe (dihydrorhodamine). Comparison of the generated fluorescence in the presence and absence of an iron chelator (deferiprone or deferrioxamine) makes the assay specific for iron, which can then be measured from a standard curve generated from known iron concentrations.12,29 For the extraction of iron from NTBI, assay L4 used 0.1 mM nitrilotriacetate to capture the mobilizer-dependent form of LPI (enhanced LPI).11, 29 The statistical analysis is described in detail in the Online Supplementary Information.

Results Patients’ characteristics

Assay levels and limits of detection

The characteristics of the patients, divided by disease and treatment group, are shown in Table 1. Serum ferritin levels were highest in patients with transfusion-dependent

Absolute NTBI and LPI levels differed considerably between assays (Table 2). Overall, the NTBI assays (N1-N4,

Table 1. Patients’ characteristics by disease and treatment groups. Median (range) are given for five to seven measurements in each disease and treatment group.

Age years

Hb mmol/La

MCV fL

CRP mg/L

Diseasec

Hemochromatosis 1. Naïve

58 (42-66)

9.5 (8.4-10.0) 91 (82-96)

2. Depletion

59 (42-66)

8.5 (7.9-9.0)

93 (85-104)

3. Maintenance

59 (43-67)

9.2 (8.3-9.6)

90 (82-97)

β-Thalassemia 4. Major Naïve

(0.5-2)

5.7 (4.6-6.6)

81 (73-84)

5. Major TD & CHb

31 (28-35)

6.0 (5.4-6.3)

84 (82-86)

6. Intermedia Naïve

39 (22-60)

5.5 (4.7-6.3)

85 (64-94)

MDS 7. RARS/RCMD-RS

69 (26-77)

6.7 (5.4-7.4)

97 (79-99)

8. TD

71 (68-85)

4.6 (4.0-5.6)

92 (87-97)

Sickle cell disease 9. Naïve

34 (18-57)

5.2 (4.1-7.1)

87 (76-114)

10. TDd

23 (19-41)

5.9 (4.2-6.5)

85 (81-90)

Reference range Male Female

8.5-11.0 80-100 7.5-10.0 80-100

<10 <10

LDHe U/L

Iron mmol/L

Ferritin TSAT immuno TSAT gel mg/L % %

(0-0)

157 (128-238) (0-0) 121 (105-140) (0-0) 126 (112-144)

37 (33-44) 31 (18-39) 23 (10-36)

1389 (948-2049) 631 (570-786) 58 (30-166)

97 100 (76-100) (80-100) 66 58 (35-89) (31-95) 44 43 (21-95) (24-100)

(0-5)

252 (233-478) (0-0) 127 (91-141) (0-6) 228 (150-550)

24 (17-29) 38 (17-46) 35 (16-42)

912 (639-2060) 2507 (623-5451) 487 (306-940)

52 52 (43-94) (46-100) 100 100 (45-103) (47-100) 95 100 (42-103) (48-100)

(0-0)

35 474 (17-39) (289-826) 30 2633 (19-45) (1379-11834)

71 52 (31-90) (31-88) 94 100 (93-106) (92-100)

(0-6)

19 (11-27) 25 (7-38)

164 (79-996) 2091 (1451-5353)

37 (18-58) 62 (19-95)

37 (28-48) 51 (17-100)

14-35 10-25

25-250 20-250

<45 <45

190 (105-292) (0-275) 181 (161-290) 437 (290-864) (0-30) 448 (231-649) <250 <250

Naïve: before start of any therapy; depletion: during the depletion phase of phlebotomy treatment; maintenance: during maintenance phase of phlebotomy. TD: transfusion-dependent; CH: receiving iron chelation therapy; RARS: refractory anemia with ringed sideroblasts; RCMD: refractory cytopenia with multilineage dysplasia; Hb: hemoglobin; MCV: mean corpuscular volume; CRP: C-reactive protein; LDH: lactate dehydrogenase; iron: serum iron; ferritin: serum ferritin; TSAT immuno, transferrin saturation, transferrin measured immunochemically and iron with a colorimetric method, both on an automated analyzer; TSAT gel: transferrin saturation measured with gel electrophoresis; NA: not available; Hb and MCV in thalassemia patients were from the first sampling due to pooling afterwards. aHb (g/dL) = mmol/L x 1.6206; bchelation consisted of deferiprone (Ferriprox, DFP, n=2) and deferasirox (Exjade, DFX, n=4); cthe disease groups have been given numbers as identities; d: including five patients on iron chelation therapy that consisted of DFP (n=3) and DFX (=2); e: the hemolytic index (reflecting in vivo and ex vivo hemolysis) levels were within the reference range (<20 mg/dL) for all samples (data not shown). Patients with SCD were of African descent (African-Dutch) and the patients with β-thalassemia were all Italian-Caucasian (Italians). All other patients were Dutch-Caucasian.

haematologica | 2016; 101(1)

Round robin for NTBI and LPI

N6) measured ~2-30 times higher concentrations compared to the LPI assays (L1-L4), in agreement with the concept that NTBI assays measure total circulating NTBI, whereas LPI assays specifically measure the redox active fraction. Nonetheless, major differences in absolute values were found within the group of NTBI and LPI assays. For example, both N1 and N5 measure directly chelatable iron but levels were much higher for N1 than for N5. In fact, N5 levels were greater than the cut-off of zero in only 8% of the samples (n=5). This assay was, therefore, excluded from further analysis. As expected, levels of assay L4, which measures the forms of LPI that are detectable in the presence of nitrilotri-

acetate (designated “enhanced LPI”), were higher than those obtained by assay L3 that uses a methodology that is identical to L4 except that nitrilotriacetate is not added. Assays differed widely in the use of predefined detection limits (Online Supplementary Table S1). Four out of ten methods studied reported negative values: i.e. N1, N2, L1 and L2 in 20%, 60%, 53% and 25% of the samples, respectively. For assays N5, L3 and L4, 92%, 70%, and 28% of the samples, respectively, were measured as ≤0 but were reported as 0 (zero) NTBI or LPI. For two of these assays, L3 and L4, an upper limit of detection (LOD) of 2.2 mmol/L was also employed since the assay is non-linear above this concentration. Three assays (N3, N4, and N6)

Table 2. Mean (SD, in µmol/L) of the NTBI and LPI values by assay, disease and treatment.

N Assay ID Assay subgroup

N1 DCI

Disease/Total 60 Hemochromatosis 1. Naïve 6 2. Phlebotomy 6 3. Maintenance 6 β-Thalassemia 4. Major Naïve 6 5. Major TD & CH 6 6. Intermedia Naïve 7 MDS 7. RARS/RCMD-RS 6 8. TD 5 Sickle cell disease 9. Naïve 6 10. TD & CH 6

N2e NTBI

NTBI assays N3a NTBI

LPI assays N4b NTBI

N6 Isoform

L1 LPI

L2 LPI

L3c,d LPI

L4c,d eLPI

1.60 (0.37) -0.73 (0.20) 1.32 (0.14)

2.19 (0.64) 1.10 (0.15)

0.09 (0.07) 0.24 (0.17)

0.14 (0.04)

0.46 (0.11)

2.61 (0.43) 0.42 (0.19) 2.08 (0.16) 0.36 (0.30) -1.32 (0.20) 0.66 (0.13) 0.25 (0.26) -1.52 (0.15) 0.57 (0.02)

3.25 (0.73) 1.58 (0.22) 0.21 (0.11) 0.39 (0.13) 1.15 (0.20) 1.36 (0.15) -0.04 (0.05) -0.09 (0.14) 0.91 (0.35) 1.19 (0.19) -0.07 (0.03) 0.02 (0.15)

0.10 (0.06) 0.00 (0.00) 0.00 (0.00)

1.34 (0.31) 0.03 (0.04) 0.05 (0.03)

0.61 (0.25) -1.19 (0.17) 0.76 (0.05) 5.79 (0.65) 0.38 (0.27) 2.46 (0.12) 3.32 (0.44) 0.04 (0.17) 2.46 (0.17)

1.68 (1.07) 1.08 (0.36) -0.01 (0.04) 4.00 (0.67) 0.87 (0.11) 0.43 (0.12) 3.74 (2.25) 1.51 (0.06) 0.28 (0.10)

0.08 (0.09) 0.68 (0.24) 0.64 (0.25)

0.01 (0.01) 0.54 (0.15) 0.52 (0.06)

0.17 (0.05) 1.43 (0.17) 0.83 (0.14)

0.14 (0.16) -1.38 (0.25) 0.61 (0.03) 2.97 (0.81) 0.53 (0.15) 2.33 (0.43)

1.37 (0.20) 1.11 (0.05) -0.05 (0.02) 3.65 (0.31) 0.60 (0.04) 0.22 (0.10)

0.04 (0.17) 0.62 (0.23)

0.00 (0.00) 0.20 (0.11)

0.02 (0.03) 0.66 (0.29)

-0.10 (0.24) -2.01 (0.15) 0.44 (0.22) -0.05 (0.19) -1.12 (0.32) 0.86 (0.14)

0.49 (0.12) 0.94 (0.06) -0.06 (0.04) -0.01 (0.12) 1.69 (0.20) 0.57 (0.21) 0.00 (0.04) -0.01 (0.15)

0.00 (0.00) 0.00 (0.00)

0.01 (0.01) 0.06 (0.06)

Assay ID: random numbering within assay group. Each mean was calculated as the outcome of the means of a total of four NTBI or LPI measurements (2 days, 2 measurements with exception of 2 methods, 1 which performed assays over 4 days, 1 measurement and 1 which performed 1 day, 2 measurements), according to the number of patients in each disease category (n=5-7). NTBI assays: measurement of total NTBI using chelators such as nitrilotriacetate or desferrioxamine as scavenging agent; LPI assays: measurement of redox active fraction of NTBI using reducing agents such as ascorbate or bleomycin to induce redox cycling and generate reactive oxygen species; NTBI isoform-specific measurement of the iron-citrate fraction in the serum samples; DCI: direct chelatable iron; eLPI: enhanced LPI. Method N5 (DCI) has been omitted from this table since NTBI was only present in 8% of the samples (n=5: 3 thalassemia major, transfusion-dependent or receiving iron chelation therapy, 1 untreated thalassemia intermedia, 1 transfusion-dependent MDS); a,b: for statistical calculations, results below the LLOD of 0.60 and 0.87 were included as 0.30 and 0.40, respectively; c: results were recommended to be interpreted as follows: X <0.1 =negative, 0.1 ≥ X <0.2 = undetermined, X ≥0.2 = positive; d: for statistical calculations, results above the upper LOD of 2.2 were included as 2.4 mmol/L; e: range of levels assessed in healthy individuals (23-61 years, 50% men; TSAT 16-56%): -2.29 to 0.35 µmol/L.

Table 3. The between-sample and within-sample variation of the assay concentrations by method.

Method

Total

Betweensample

SD (µmol/L) Within-sample Within-day

N1 N2 N3 N4 N6 L1 L2 L3 L4

3.20 1.23 1.18 2.07 1.08 0.26 0.51 0.40 0.69

3.15 1.19 1.15 1.70 1.05 0.24 0.47 0.38 0.67

0.00 0.00 NA 0.07 0.00 0.00 0.00 0.00 0.00

Residual

Between-day 0.15 0.18 0.06 0.00 0.00 0.03 0.00 0.00 0.00

Variance (%) Within-sample

Betweensample

Within-day 0.51 0.21 0.25 1.19 0.26 0.08 0.21 0.10 0.18

97.2 94.8 95.3 67.1 94.3 87.7 83.5 94.0 93.5

0.0 0.0 NA 0.1 0.0 0.0 0.0 0.0 0.0

Residual

Between-day 0.2 2.2 0.2 0.0 0.0 1.3 0.0 0.0 0.0

2.6 3.0 4.5 32.8 5.8 11.0 16.5 6.0 6.5

SD: standard deviation, absolute error; between-sample SD: segment due to variation between samples; within-sample within-day: segment due to variation in repeated measurements on the same day; within-sample between-day: segment due to variation in repeated measurements on two different days. NA: not available, measurements were on four different days instead of twice on two different days. For N6 only a duplicate measurement on day 2 is available because of iron contamination issues on day 1.

haematologica | 2016; 101(1)

L. de Swart et al.

measured positive values only. Two of them (N3 and N4) applied a lower LOD of 0.60 mmol/L and 0.87 mmol/L with reported results below the lower LOD in 42% and 32% of the samples, respectively.

able to identify different NTBI/LPI concentrations in the various groups of patients (Table 3). The relatively low betweensample variance of assay N4 could be attributed to the relatively high residual variance (33%) due to some outliers. For all assays, the contribution of the within-sample variance (within day and between-day) to the total variance was relatively low (0-2.2%), indicating good reproducibility.

Reproducibility Assays gave a large contribution of the between-sample variance to the total variance indicating that the assays are

Table 4. Spearman correlation coefficients between the assays, using the mean of the duplicate measurements on day 2.

N1 N2 N3 N4 N6 L1 L2 L3 L4

0.60 0.57 0.57 -0.06 0.50 0.59 0.75 0.62

0.88 0.85 -0.06 0.89 0.54 0.73 0.71

0.86 -0.06 0.85 0.48 0.69 0.63

-0.11 0.83 0.44 0.62 0.58

0.05 0.05 0.03 0.13

0.54 0.72 0.67

0.74 0.57

0.77

Bland-Altman plots of the bold numbers are presented in Figure 1. Correlation plots of all cells are provided in Online Supplementary Figure S1.

Figure 1. Bland-Altman plots showing differences in mean NTBI or LPI levels (mmol/L) between two methods versus the average of the mean values of these two methods. Solid lines represent the mean difference between the two methods. Thick dashed lines represent the 95% confidence interval. Thin dashed lines at 0 refer to no difference between methods. Dots represent the means of duplicates on day 2. Note that absolute values on both axes differ between plots.

haematologica | 2016; 101(1)

Round robin for NTBI and LPI

Comparison between assays Overall, assays correlated well and correlations were highest within the same group of NTBI or LPI assays (Table 4, Online Supplementary Figure S1). The NTBI and LPI assays with the highest mutual Spearman correlation were N2 and N3 (rs=0.88), N3 and N4 (rs=0.86), N2 and N4 (rs=0.85), L3 and L4 (rs=0.77), and L2 and L3 (rs=0.74). Spearman correlations of N6 with the other assays were all <0.10. Remarkably, the Spearman correlation of LPI assay L1, using bleomycin, with NTBI assays was higher than with the LPI assays. In contrast, NTBI assay N1 showed a better Spearman correlation with LPI assays than with NTBI assays (Table 4). Figure 1 shows the Blandâ&#x20AC;&#x201C;Altman plots of a selection of nine graphs with the highest mutual correlation or a specific correlation of interest. Two patterns can be observed in the plots: (i) if the scales of one variable are small compared to the other one, there is a straight line pattern (e.g. in the N2-L1 plot); and (ii) if cut-off values are used in one variable and not in the other variable, a straight line occurs in combination with a scattered pattern (e.g. the N2-N3 plot).

load diseases in various treatment phases. Nine out of the ten assays reported an assay value above zero in â&#x2030;Ľ30% of the samples. These nine methods exhibited a high between-sample variation and low within-sample variation indicating their suitability to identify different serum NTBI and LPI concentrations in patients with ironoverload. Overall, results correlated both within the group of five NTBI assays and in the group of the four LPI assays, but between group correlations were also present. Nonetheless, absolute assay levels differed con-

Discrimination of disorder and treatment groups by the assays Table 2 shows the results obtained by the ten assays in the ten different disease and treatment groups. Using a linear mixed model, we found that all methods were suitable to discriminate, at least on a group level, between some different types of iron-overload disease (P<0.05), except for the NTBI isoform-specific assay N6 (P=0.268) (Online Supplementary Table S2). Overall, assays were best in discriminating LPI and NTBI from patients with highest TSAT levels (thalassemia major patients who are transfusion-dependent or receiving iron chelation therapy, thalassemia intermedia patients, transfusion-dependent MDS patients and untreated HH patients) from those of the diseases with lower TSAT (untreated and transfusion-dependent SCD, HH maintenance; Table 2 and Online Supplementary Table S2).

Comparison of assay results with transferrin saturation and ferritin levels The assays show either a hyperbolic relationship or a threshold effect with TSAT, regardless of the underlying etiological condition, NTBI concentrations increase sharply when TSAT levels exceed 70%, NTBI is sporadically observed in samples with TSAT lower than 70% and LPI is found essentially when TSAT exceeds 90% (Figure 2). More specifically, at TSAT >90%, the presence of LPI and NTBI may be taken for granted for N1, N2, N3, N4, N6, L1 and L4 since for these assays the proportion of samples in which NTBI or LPI is found is 95.7% or 100%. At TSAT >70%, these proportions are somewhat lower, but for N3 and N6 already 100% (Online Supplementary Table S3). There is no relationship between serum ferritin and either LPI or NTBI measured in any of the assays (data not shown).

Discussion We compared the performance of ten different NTBI and LPI assays in patients with four different iron-overhaematologica | 2016; 101(1)

Figure 2. Relationship between the assay concentrations and transferrin saturation (TSAT). Assay results are given for day 2 as duplicate measurements (circles and triangles).

L. de Swart et al.

siderably between assays, with lowest levels for LPI methods. Overall, assay levels correlated with TSAT but not with ferritin concentrations and iron-overload diseases with the highest TSAT also had the highest levels of NTBI and LPI. Some of the differences in absolute values between the assays can be attributed to differences in the assay principles that result in the measurement of different NTBI entities, e.g. NTBI directly chelatable iron, NTBI-iron citrate, LPI and enhanced LPI.3,11 In addition, minor variations in analytical procedures between methods aimed at the measurement of the same entities (N2-N4 or N1 and N5 or L2 and L3) may also have contributed to betweenassay variation in absolute values. For instance, it has recently been shown that, for nitrilotriacetate-based assays that use filtration, centrifugation at extremely high speed can lead to overestimation of the NTBI concentration due to passage of transferrin-bound iron through the filter membrane. Conversely, when centrifuging at lower speed, iron-nitrilotriacetate complexes do not pass the membrane, leading to underestimation.31 The LPI assays also differ with regards to reagents and buffers. The presence of residual chelator or chelates could contribute to higher values in some assays, such as the bleomycinbased LPI (L1) assay and the nitrilotriacetate-filtration assays (N2-N4), but not in others.11 Furthermore, iron contamination may also be a source of variation in absolute assay concentrations.32 It could contribute to the reported negative values by mimicking free circulating iron. Rather than registering as an increase in NTBI, this contaminant free iron can shuttle from the iron chelator nitrilotriacetate to vacant binding sites on transferrin.27,31,33 Thus samples containing unsaturated transferrin may have lower NTBI values compared to standards in which no transferrin is present. In principle, this can result in underestimation of NTBI and negative values. This iron shuttling can be prevented by saturating transferrin with cobalt(III) prior to addition of the nitrilotriacetate.34 Still, others have reported that the addition of cobalt (III) may result in substantial iron-contamination that could account for a rise in NTBI even in normal individuals.32 Finally, with regards to the iron chelator nitrilotriacetate, applied in three NTBI assays, the use of 80 mM was shown by some authors to remove 1-8% of the iron bound to transferrin in the case of elevated TSAT, thus leading to overestimation of NTBI measurements.27,31 By contrast, at physiological concentrations of serum iron in control subjects others reported no mobilization of iron from transferrin under these conditions.28 NTBI is thought to consist of Fe(III) bound to several ligands. Since iron bound to citrate is likely to be the most prevalent isoform of NTBI in most conditions, a LC-ICPMS assay for the sum of these Fe(III)-citrate complexes of NTBI was developed and included as N6 in the current round robin.6 The assay was reproducible and discriminated between samples, but assay levels did not correlate with other assays and could not discriminate between any of the disorders. Clearly, this method may need optimization, but if successful, we anticipate that quantification of NTBI subspecies could provide the basis for further studies on their toxicity. Different methods may not be equivalent at measuring these subspecies. It has been postulated that the abundance of each subspecies varies with the type and degree of iron-overload disease.7,11 44

Although this possibility is still under investigation, it implies that specific assays might be required for different chemical species of NTBI.6,7,11,13 Overall the assays correlated well. Mutual correlations between the assays were found to be highest for methods based on a similar principle (i.e. in the group of NTBI assays and the group of LPI assays). We observed two unexpected correlations: first, despite the chelator-based principle of N1, the correlation of N1 with LPI assays exceeded its correlation with other NTBI assays. A possible explanation could be the kinetics of access of CP851 to redox active NTBI.6,12,15,27 Secondly, even though assay L1 is based on binding of the antibiotic bleomycin to redox active iron and DNA, its results correlated slightly better with most NTBI assays than with LPI assays. Bleomycin-detectable NTBI might represent a larger redox active fraction of NTBI than LPI, in equilibrium with a larger, still uncharacterized pool of NTBI. Differences in the range of reported values, as well as in absolute values, have highlighted the urgent need for thorough method-validation protocols, standardization and consensus on how best to report assays before their introduction in clinical use. Four out of ten assays report negative values and as such provide information on the residual iron-binding capacity of transferrin. Other assays report 0 (zero) for results â&#x2030;¤0, or below the lower LOD for results that do not significantly differ from 0. It remains to be determined which reporting strategy is clinically more useful. Importantly, both LPI and NTBI assays showed a relationship with TSAT, irrespective of the type of iron-overload disease. NTBI and LPI were essentially found above certain thresholds of TSAT, lower for NTBI (~70%) than for LPI (~90%). A comparable pattern was previously reported for LPI assays that were similar to assays L1-L3 for patients with various conditions.2,3,35,36 Altogether, our findings corroborate previous reports in HH and thalassemia patients that TSAT levels of around 70% can be considered as a threshold for the presence of potentially toxic iron species.2,3 Nonetheless, because a single measurement only reflects the labile iron species present during the previous 24-48 hours, repeated measurements are imperative to assess a potential risk of impending NTBI toxicity.18 The majority of the assays in the current round robin have given useful insights into the efficacy of iron chelation in transfusional siderotic patients and of phlebotomy in patients with HH.2,3,9,13,19-22,37-40 Despite this finding, in the absence of studies that assess the independent relation of assay levels to clinical outcome in these various categories of patients, the clinically most relevant assay formats and their decision limits remain unknown. In conclusion, several novel methods have been developed since the previous round robin for NTBI, including LPI and directly chelatable iron assays and first attempts to specifically determine the NTBI iron-citrate isoforms. While NTBI and LPI values of various assays were well correlated and were reproducible, absolute values differed considerably. The assay(s) that best represent the clinically relevant species and are most relevant for diagnostic and therapeutic purposes could not be determined from this study. Before NTBI and LPI assays can be introduced into clinical practice, rigorous validation and standardization of assays, consensus on how to report the results, haematologica | 2016; 101(1)

Round robin for NTBI and LPI

and clinical outcome studies to determine clinically relevant assay formats and their toxic thresholds are needed. Acknowledgments We would like to thank all round robin 2 NTBI laboratory participants and sample collectors for their contributions and the members of the advisory committee: Prof. Dr. Theo de Witte and Prof.

References 1. Hershko C, Graham G, Bates GW, Rachmilewitz EA. Non-specific serum iron in thalassaemia: an abnormal serum iron fraction of potential toxicity. Br J Haematol. 1978;40(2):255-263. 2. Le Lan C, Loreal O, Cohen T, et al. Redox active plasma iron in C282Y/C282Y hemochromatosis. Blood. 2005;105(11): 4527-4531. 3. Pootrakul P, Breuer W, Sametband M, Sirankapracha P, Hershko C, Cabantchik ZI. Labile plasma iron (LPI) as an indicator of chelatable plasma redox activity in ironoverloaded beta-thalassemia/HbE patients treated with an oral chelator. Blood. 2004;104(5):1504-1510. 4. Hod EA, Brittenham GM, Billote GB, et al. Transfusion of human volunteers with older, stored red blood cells produces extravascular hemolysis and circulating non-transferrinbound iron. Blood. 2011;118(25): 6675-6682. 5. Brissot P, Ropert M, Le Lan C, Loreal O. Non-transferrin bound iron: a key role in iron overload and iron toxicity. Biochim Biophys Acta. 2012;1820(3):403-410. 6. Hider RC, Silva AM, Podinovskaia M, Ma Y. Monitoring the efficiency of iron chelation therapy: the potential of nontransferrinbound iron. Ann NY Acad Sci. 2010; 1202:94-99. 7. Silva AM, Hider RC. Influence of non-enzymatic post-translation modifications on the ability of human serum albumin to bind iron. Implications for non-transferrin-bound iron speciation. Biochim Biophys Acta. 2009;1794(10):1449-1458. 8. Silva AM, Kong X, Parkin MC, Cammack R, Hider RC. Iron(III) citrate speciation in aqueous solution. Dalton Trans. 2009;(40):86168625. 9. Grootveld M, Bell JD, Halliwell B, Aruoma OI, Bomford A, Sadler PJ. Non-transferrinbound iron in plasma or serum from patients with idiopathic hemochromatosis. Characterization by high performance liquid chromatography and nuclear magnetic resonance spectroscopy. J Biol Chem. 1989;264(8):4417-4422. 10. Breuer W, Ermers MJ, Pootrakul P, Abramov A, Hershko C, Cabantchik ZI. Desferrioxamine-chelatable iron, a component of serum non-transferrin-bound iron, used for assessing chelation therapy. Blood. 2001;97(3):792-798. 11. Breuer W, Ghoti H, Shattat A, et al. Nontransferrin bound iron in thalassemia: differential detection of redox active forms in children and older patients. Am J Hematol. 2012;87(1):55-61. 12. Esposito BP, Breuer W, Sirankapracha P, Pootrakul P, Hershko C, Cabantchik ZI.

haematologica | 2016; 101(1)

13. 14.

15.

16.

17.

18. 19.

20.

21.

22.

23.

24. 25.

26.

27.

Dr. Norbert Gattermann. Special thanks also to Siem Klaver for his support in logistics and Rian Roelofs and Erwin Wiegerinck for their technical assistance and valuable discussions. Funding This work was supported by an educational grant from Novartis Pharmacy B.V. Europe/The Netherlands to DS and LdS.

Labile plasma iron in iron overload: redox activity and susceptibility to chelation. Blood. 2003;102(7):2670-2677. Cabantchik ZI. Labile iron in cells and body fluids: physiology, pathology, and pharmacology. Front Pharmacol. 2014;5(45):1-11. Evans RW, Rafique R, Zarea A, et al. Nature of non-transferrin-bound iron: studies on iron citrate complexes and thalassemic sera. J Biol Inorg Chem. 2008;13(1):57-74. Ma Y, Podinovskaia M, Evans PJ, et al. A novel method for non-transferrin-bound iron quantification by chelatable fluorescent beads based on flow cytometry. Biochem J. 2014;463(3):351-362. Pierre JL, Gautier-Luneau I. Iron and citric acid: a fuzzy chemistry of ubiquitous biological relevance. Biometals. 2000;13(1):9196. Wang W, Knovich MA, Coffman LG, Torti FM, Torti SV. Serum ferritin: past, present and future. Biochim Biophys Acta. 2010; 1800(8):760-769. Wood JC. Guidelines for quantifying iron overload. Hematology Am Soc Hematol Educ Program. 2014;2014 (1):210-215. Piga A, Longo F, Duca L, et al. High nontransferrin bound iron levels and heart disease in thalassemia major. Am J Hematol. 2009;84(1):29-33. 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. Gattermann N, Finelli C, Porta MD, et al. Deferasirox in iron-overloaded patients with transfusion-dependent myelodysplastic syndromes: results from the large 1-year EPIC study. Leuk Res. 2010;34(9):1143-1150. Zanninelli G, Breuer W, Cabantchik ZI. Daily labile plasma iron as an indicator of chelator activity in thalassaemia major patients. Br J Haematol. 2009;147(5):744751. Wood JC, Glynos T, Thompson A, et al. Relationship between labile plasma iron, liver iron concentration and cardiac response in a deferasirox monotherapy trial. Haematologica. 2011;96(7):1055-1058. Hershko C. Pathogenesis and management of iron toxicity in thalassemia. Ann NY Acad Sci. 2010;1202:1-9. Evans PJ, Halliwell B. Measurement of iron and copper in biological systems: bleomycin and copper-phenanthroline assays. Methods Enzymol. 1994;233:82-92. Singh S, Hider RC, Porter JB. A direct method for quantification of non-transferrin-bound iron. Anal Biochem. 1990;186 (2):320-323. Gosriwatana I, Loreal O, Lu S, Brissot P, Porter J, Hider RC. Quantification of non-

28.

29. 30.

31.

32.

33. 34.

35.

36.

37.

38.

39.

40.

transferrin-bound iron in the presence of unsaturated transferrin. Anal Biochem. 1999;273(2):212-220. Zhang D, Okada S, Kawabata T, Yasuda T. An improved simple colorimetric method for quantitation of non-transferrin-bound iron in serum. Biochem Mol Biol Int. 1995;35(3):635-641. Aferrix. www.aferrix.com. July 2015. Jacobs EM, Hendriks JC, van Tits BL, et al. Results of an international round robin for the quantification of serum non-transferrinbound iron: need for defining standardization and a clinically relevant isoform. Anal Biochem. 2005;341(2):241-250. Makino T, Nakamura K, Takahara K. Potential problems in the determination of serum non-transferrin-bound iron using nitrilotriacetic acid and ultrafiltration. Clin Chim Acta. 2014;429:12-13. Sasaki K, Ikuta K, Tanaka H, et al. Improved quantification for non-transferrin-bound iron measurement using high-performance liquid chromatography by reducing iron contamination. Mol Med Rep. 2011;4(5): 913-918. Hider RC. Nature of nontransferrin-bound iron. Eur J Clin Invest. 2002;32 (Suppl 1): 50-54. Breuer W, Ronson A, Slotki IN, Abramov A, Hershko C, Cabantchik ZI. The assessment of serum nontransferrin-bound iron in chelation therapy and iron supplementation. Blood. 2000;95(9):2975-2982. Itkonen O, Vaahtera L, Parkkinen J. Comparison of bleomycin-detectable iron and labile plasma iron assays. Clin Chem. 2013;59(8):1271-1273. Danjou F, Cabantchik ZI, Origa R, et al. A decisional algorithm to start iron chelation in patients with beta thalassemia. Haematologica. 2014;99(3):e38-40. Porter J, Galanello R, Saglio G, et al. Relative response of patients with myelodysplastic syndromes and other transfusion-dependent anaemias to deferasirox (ICL670): a 1-yr prospective study. Eur J Haematol. 2008;80 (2):168-176. Domenica Cappellini M, Tavazzi D, Duca L, Marelli S, Fiorelli G. Non-transferrin-bound iron, iron-related oxidative stress and lipid peroxidation in beta-thalassemia intermedia. Transfus Sci. 2000;23(3):245-246. Inati A, Musallam KM, Cappellini MD, Duca L, Taher AT. Nontransferrin-bound iron in transfused patients with sickle cell disease. Int J Lab Hematol. 2011;33(2):133137. Taher A, Musallam KM, El Rassi F, et al. Levels of non-transferrin-bound iron as an index of iron overload in patients with thalassaemia intermedia. Br J Haematol. 2009;146(5):569-572.

ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Platelet Biology & Its Disorders

Ferrata Storti Foundation

Cytoskeletal perturbation leads to platelet dysfunction and thrombocytopenia in variant forms of Glanzmann thrombasthenia Loredana Bury,1 Emanuela Falcinelli, 1 Davide Chiasserini, 2 Timothy A. Springer, 3 Joseph E. Italiano Jr., 4 and Paolo Gresele 1

Department of Medicine, Section of Internal and Cardiovascular Medicine, University of Perugia, Italy; 2Department of Medicine, Section of Neurology, University of Perugia, Italy; 3Department of Biological Chemistry and Molecular Pharmacology, Harvard Medical School and Program in Cellular and Molecular Medicine, Children’s Hospital, Boston, MA, USA; and 4Hematology Division, Department of Medicine, Brigham and Women's Hospital, Vascular Biology Program, Boston Children’s Hospital and Harvard Medical School, Boston, MA, USA 1

Haematologica 2016 Volume 101(1):46-56

ABSTRACT

Correspondence: paolo.gresele@unipg.it

Received: 18/5/2015. Accepted: 1/10/2015. Pre-published: 9/10/2015. doi:10.3324/haematol.2015.130849

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/1/46

everal patients have been reported to have variant dominant forms of Glanzmann thrombasthenia, associated with macrothrombocytopenia and caused by gain-of-function mutations of ITGB3 or ITGA2B leading to reduced surface expression and constitutive activation of integrin αIIbβ3. The mechanisms leading to a bleeding phenotype of these patients have never been addressed. The aim of this study was to unravel the mechanism by which ITGB3 mutations causing activation of αIIbβ3 lead to platelet dysfunction and macrothrombocytopenia. Using platelets from two patients carrying the β3 del647-686 mutation and Chinese hamster ovary cells expressing different αIIbβ3-activating mutations, we showed that reduced surface expression of αIIbβ3 is due to receptor internalization. Moreover, we demonstrated that permanent triggering of αIIbβ3-mediated outside-in signaling causes an impairment of cytoskeletal reorganization arresting actin turnover at the stage of polymerization. The induction of actin polymerization by jasplakinolide, a natural toxin that promotes actin nucleation and prevents depolymerization of stress fibers, in control platelets produced an impairment of platelet function similar to that of patients with variant forms of dominant Glanzmann thrombasthenia. del647-686β3-transduced murine megakaryocytes generated proplatelets with a reduced number of large tips and asymmetric barbell-proplatelets, suggesting that impaired cytoskeletal rearrangement is the cause of macrothrombocytopenia. These data show that impaired cytoskeletal remodeling caused by a constitutively activated αIIbβ3 is the main effector of platelet dysfunction and macrothrombocytopenia, and thus of bleeding, in variant forms of dominant Glanzmann thrombasthenia.

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

Integrin αIIbβ3 (GPIIb/IIIa), the main platelet receptor, is a heterodimeric calciumdependent cell-surface glycoprotein expressed on platelets and megakaryocytes that plays a central role in platelet aggregation and thrombus formation.1 Recent observations suggest that αIIbβ3 is also involved in proplatelet formation.2,3 Integrin αIIbβ3 function depends on two different signal transduction pathways: inside-out and outside-in signaling. Under resting conditions, αIIbβ3 is expressed on platelets in a bent, inactive conformation. Upon platelet activation, inside-out signaling causes αIIbβ3 extension and headpiece opening, leading the receptor to assume an active conformation, to acquire the ability to bind its ligands, mainly fibrinogen, and to haematologica | 2016; 101(1)

Mechanism of platelet dysfunction in dominant GT

initiate platelet aggregation.4 On the other hand, ligand binding leads αIIbβ3 complexes to cluster, thus triggering outside-in signaling, with phosphorylation of the β3 cytoplasmic tail, activation of signaling proteins (e.g. Src-family kinases and focal adhesion kinase), and ultimately reorganization of the actin cytoskeleton.5 Cytoskeletal remodeling, with actin polymerization and depolymerization, is a finely regulated event,6 and when altered it may lead to impaired platelet function and formation, as observed in patients with mutations of filamin A or with MYH9-RD.7,8 Mutations of ITGA2B and ITGB3, the genes coding for integrins αIIb and β3, generate Glanzmann thrombasthenia (GT), an autosomal recessive bleeding disorder characterized by absent platelet aggregation and a normal platelet count and volume, due to quantitative or qualitative defects of αIIbβ3. Heterozygous carriers of GT are usually asymptomatic because 50% of normal αIIbβ3 is sufficient for platelet aggregation,9 but rare autosomal dominant variants of GT, with platelet dysfunction and macrothrombocytopenia, have been associated with gain-of-function mutations of ITGA2B or ITGB3 leading to reduced expression and constitutive activation of αIIbβ3.3,10 We have previously described a hereditary Glanzmannlike platelet disorder transmitted in an autosomal dominant way, associated with macrothrombocytopenia and a bleeding diathesis, due to a heterozygous G>C transversion (c.2134+1G>C) of ITGB3 leading to the deletion of exon 13 and to the loss of 40 amino acids (del647-686) of integrin β3.11 This mutation was the first-described, naturally-occurring deletion of the β-tail domain (β-TD) of integrin β3, the membrane-proximal portion of the extracellular domain of the protein. β-TD contributes to maintain the bent, inactive conformation of αIIbβ3;12 in fact, the loss of its disulfide bonds causes constitutive activation of αIIbβ3,13-15 but no information is available on its role in outside-in signaling. Recently, a Japanese family carrying a heterozygous ITGB3 c.2134+1G>A transversion leading to the same integrin β3 deletion and a similar phenotype was reported.16 We have previously reported that del647-686 integrin β3 leads to constitutive activation of αIIbβ3 in patients’ megakaryocytes,3 a finding recently confirmed in transfected 293T cells.16 A few other heterozygous patients with gain-of-function mutations of ITGB3, mucocutaneous bleeding and macrothrombocytopenia have been reported17-19 suggesting that, independently of the mutation, constitutive activation of αIIbβ3 leads to platelet dysfunction and macrothrombocytopenia by a common mechanism that, however, has never been addressed. Here we show that constitutive activation of integrin αIIbβ3 decreases surface expression of the complex through receptor internalization and permanently triggers outsidein signaling, which leads to altered cytoskeletal reorganization that is the main effector of platelet dysfunction and macrothrombocytopenia, and thus of bleeding, in autosomal dominant GT variant forms.

Declaration of Helsinki. The study was approved by the ethical committee of the University of Perugia.

Construction of the expression vectors, mutagenesis and transfection Expression vectors were obtained20 and Chinese hamster ovary (CHO) cells were transfected as described in the Online Supplementary Data. All experiments were performed 2 days after transfection.

Surface biotinylation, β3 immunoprecipitation and western blotting

Surface proteins on CHO cells and platelets were biotinylated, and integrin β3 was immunoprecipitated and analyzed by western blotting as described in the Online Supplementary Data.

αIIbβ3 expression and internalization, and αvβ3 expression

αIIbβ3 expression on resting or activated platelets or CHO cells21 was assessed by flow cytometry22 or by western blotting after biotinylation of membrane proteins and β3 immunoprecipitation. αvβ3 expression was assessed by flow cytometry.22 Internalization of αIIbβ3 was assessed by flow cytometry as described elsewhere.23 Platelet fibrinogen content was quantified by enzyme-linked immunosorbent assay (GenWay Biotech. San Diego, CA, USA). Details are given in the Online Supplementary Data.

Adhesion assay CHO cells and platelets were layered on human fibrinogen. Platelets were also layered on human von Willebrand factor after treatment with 0.1 U/mL human α-thrombin.18,24 Details are given in the Online Supplementary Data.

Protein phosphorylation CHO cells and platelets were plated on human fibrinogen and protein phosphorylation was assessed as described in the Online Supplementary Data.

Clot retraction Clot retraction was assessed with CHO cells and platelet-rich plasma25 as described in the Online Supplementary Data.

Actin polymerization Actin polymerization in CHO cells and platelets was assessed by flow cytometry as described in the Online Supplementary Data.

Analysis of cytoskeletal proteins The cytoskeleton was extracted from either resting platelets or platelets that had been stimulated with thrombin as described elsewhere26. Cytoskeletal proteins were separated on an acrylamide-gel and stained with Comassie-blue26. For details see the Online Supplementary Data.

Mass spectrometry Protein digestion and peptide analysis were performed as reported previously.27 Databases were searched using MASCOT v.2.2 and MyriMatch v.2.1.87.28 Protein assembly for MyriMatch analysis was made using IDPicker v.3.0.29 Spectral-counts were used for semiquantitative comparisons between controls and patients. For details see the Online Supplementary Data.

Methods Blood samples were taken from healthy volunteers and from two patients carrying the ITGB3c.2134+1G>C mutation.11 All subjects gave written informed consent in accordance with the haematologica | 2016; 101(1)

Effect of the perturbation of actin polymerization on platelet function

αIIbβ3 activation, platelet aggregation,30 clot retraction and spreading on fibrinogen30 were assessed in platelets treated with 47

L. Bury et al.

jasplakinolide to induce actin polymerization31 as described in the Online Supplementary Data.

Effect of the perturbation of actin polymerization on proplatelet formation

Construction of retroviral vectors, retrovirus production and megakaryocyte infection

Human megakaryocyte cultures3 were treated with jasplakinolide (1 mM) for 10 min and then cultured for 16 h. Proplatelet formation, spreading on fibrinogen, proplatelet tip number and diameter were assessed as described above.

The bicistronic pMYs-IRES-GFP/αIIb, pRetroX-IRESDsRedExpress/β3 and pRetroX-IRES-DsRedExpress/β3del647-686 expression vectors were obtained, retroviruses produced and megakaryocytes double-infected32 as described in the Online Supplementary Data.

Megakaryocyte spreading, proplatelet formation and morphology Spreading and proplatelet formation in murine megakaryocytes32 were evaluated by immunofluorescence.3 Proplatelet morphology in human blood was analyzed by immunofluorescence as described elsewhere.3,33 For details see the Online Supplementary Data.

Structural consequences of del647-686 To unravel the possible structural consequences of del647-686 the three-dimensional structure of the β-TD was derived from the αvβ3 structure (PDBID-code 4G1E) and visualized using PyMOL (DeLano Scientific, San Carlos, CA, USA).34

Statistical analysis Data are expressed as means ± standard deviation. An unpaired t-test or the two-way ANOVA with the Bonferroni post-test was applied, where appropriate, using GraphPad Prism version 5.00 (GraphPad Software,

Figure 1. αIIbβ3 surface expression. (A) Western blotting of biotinylated (surface) and notbiotinylated (total) integrin β3 of CHO cells expressing wild-type β3 (WT), del647-686 β3 (HOM) or both (HET) and of control and heterozygous patients’ platelets. Biotinylated proteins were identified using HRP-conjugated avidin. The first band is wild-type β3 and the second is mutant β3. Relative quantification (arbitrary units), CHO cells: WT=1, HOM=0.18±0.02, HET=0.24±0.09 (n=5, P<0.05 vs. WT); Platelets: control=1, Patient=0.60±0.01 (n=5, P<0.05 vs. control). Total integrin β3 expression was measured in not-biotinylated CHO cell and platelet lysates, probing membranes with a mouse anti humanβ3 monoclonal antibody. The total amount of integrin β3 was comparable. Relative quantification (arbitrary units) CHO cells: WT=1, HOM=0.89±0.06, HET=0.97±0.1 (n=5, P=ns vs. WT); Platelets: Control=1, Patients=1.2±0.2 (n=5, P=ns vs. control). (B) αIIbβ3 mean fluorescence intensity (MFI) in CHO cells transfected with empty vectors (mock) or expressing wild type (WT), del647-686 (HOM) and both (HET) integrin β3 (*P<0.01 vs. WT, #P<0.01 vs. mock). αIIbβ3 MFI in CHO cells is comparable with αIIbβ3 MFI in platelets (22.1±3.2 in control platelets and 11.9±4.1 in patients’ platelets). (C) Internalized αIIbβ3 in wild type and mutant CHO cells, activated or not with DTT (*P<0.05 vs. resting WT). (D) Internalized αIIbβ3 in control and patients’ platelets, stimulated or not with ADP (*P<0.05 vs. resting control). The same experiments were conducted using TRAP-6 obtaining comparable results (data not shown). (E) Fibrinogen content in patients’ platelets compared with controls as measured by enzyme-liked immunosorbent assay (*P<0.05 vs. control). (F) Western blotting of biotinylated β3 and flow cytometry of αIIbβ3 in control and patients’ platelets, resting or stimulated with ADP or TRAP-6. Biotinylated proteins were identified using HRP-conjugated avidin.

haematologica | 2016; 101(1)

Mechanism of platelet dysfunction in dominant GT

San Diego, CA, USA). Differences were considered statistically significant when P<0.05.

Results

β3 mutations and αIIbβ3 activation

CHO cells transfected with normal αIIb and with four different mutant β3 subunits express a constitutively activated αIIbβ3 receptor on their surface, i.e. a receptor able to bind PAC-1 and fibrinogen under resting conditions (Online Supplementary Figure S1A,B), confirming previous findings.3,17-19

Gain-of-function mutant β3 reduces surface expression of αIIbβ3 by enhancing its internalization Western

blotting

and

haematologica | 2016; 101(1)

flow-cytometry

showed

decreased surface expression of β3 but a normal amount of β3 in cell lysates, both in CHO cells expressing del647-686 integrin β3 and in heterozygous patients’ platelets (Figures 1A,B). Confocal microscopy of integrin β3-transfected CHO cells showed that mutant β3 is localized predominantly in the cytoplasm, while normal β3 is localized predominantly on the cell surface (Online Supplementary Figure S2). Mutant αIIbβ3 co-localized with concavalin A and with WGA as shown by fluorescence microscopy, confirming normal synthesis and maturation (Online Supplementary Figure S3). Moreover, in patients’ platelets, cytoplasmic αIIbβ3 co-localized with P-selectin in α-granules but also localized in other cytoplasmic structures, probably the open canalicular and dense tubular systems35 (Online Supplementary Figure S4). The fraction of internalized αIIbβ3 was significantly high-

Figure 2. Spreading. (A) Spreading of wild type (WT), mutant homozygous (HOM) and heterozygous (HET) integrin β3-expressing CHO cells after 30 and 60 min of adhesion to fibrinogen (n=3, *P<0.05 vs. WT). Representative images of WT and HET integrin β3 -expressing CHO cells after 60 min of adhesion to fibrinogen. The total surface covered was measured in 20 different high magnification microscopic fields. Mocktransfected CHO cells adhered to fibrinogen but did not spread. CHO cell adhesion on fibrinogen was comparable using the different cell lines: WT 30 min: 53.4±6.8 n. of adherent cells; WT 60 min: 43.5±5.4 n. of adherent cells; HOM 30 min: 52.5±7.9 n. of adherent cells; HOM 60 min: 51.7±5.3 n. of adherent cells; HET 30 min: 47.6±5.4 n. of adherent cells; HET 60 min: 51.6±8.4 n. of adherent cells. (B) Spreading of control (Ctrl) and patients’ (Pat) platelets after 10, 30 and 60 min of adhesion to fibrinogen (n=5, *P<0.05, #P<0.01 vs. control). Representative images of control and one patient’s platelets after 10 min of deposition on fibrinogen. Control and patient’s platelet adhesion on fibrinogen was comparable: control 10 min: 9.5±2.0 n. of adherent platelets; control 30 min: 51.3±11.4 n. of adherent platelets; control 60 min: 69.9±7.2 n. of adherent platelets; Patient’s 10 min: 11.5±3.1 n. of adherent platelets; Patient’s 30 min: 48.7±5.3 n. of adherent platelets; Patient’s 60 min: 59.6±8.7 n. of adherent platelets. (C) Representative images of control and patient’s platelets after 30 min of deposition on von Willebrand factor under resting conditions or stimulated with human α-thrombin (+Thr). Both adhesion and spreading of patient’s platelets under resting conditions were increased with respect to control: control: 52.5±6.2 n. of adherent platelets; 5.2±2.6% of spreading (% of covered surface); Patient’s: 116.0±9.8 n. of adherent platelets; 55.2±10.6% of spreading (% of covered surface). After stimulation with thrombin both adhesion and spreading of patient’s platelets were comparable with those of control platelets: control Thr: 198.1±11.0 n. of adherent platelets; 73.5±12.4% of spreading (% of covered surface); Patient’s Thr: 214.8±12.9 n. of adherent platelets; 69.7±15.6 % of spreading (% of covered surface). Platelets were stained with FITC-conjugated phalloidin. Every microscopic field covers 0.14 mm2.

L. Bury et al. er in unstimulated CHO cells expressing mutant β3 than in cells expressing wild-type β3 (Figure 1C). Similar findings were observed with resting patients’ platelets as compared with control platelets (Figure 1D). In accordance, patients’ platelets contained more fibrinogen than control platelets (Figure 1E) and mutant αIIbβ3-expressing CHO cells internalized fibrinogen under resting conditions differently from CHO cells expressing normal αIIbβ3 that required activation with DTT to internalize it (Online Supplementary Figure S1C). However, upon platelet stimulation with ADP or TRAP-6, the internal pool of αIIbβ3 of patient’s platelets externalized regularly showing that it is correctly recycled after internalization (Figure 1F). αvβ3 expression on patient’s platelets was comparable to that on control platelets (control 34.9±8.0% versus patients 37.6±7.1%, P=ns).

Constitutively activated αIIbβ3 triggers outside-in signaling

CHO cells expressing mutant β3 spread faster on fibrinogen than wild-type cells (Figure 2A). Spreading of patients’ platelets was initially faster (Figure 2B), but after 30 min became defective11. Moreover, patients’ platelets spread

spontaneously on von Willebrand factor, unlike control platelets that required αIIbβ3 activation to undergo full spreading18,24 (Figure 2C). Integrin β3 Tyr773 and Tyr785, as well as focal adhesion kinase, were constitutively phosphorylated in mutant CHO cells, while they were phosphorylated only after spreading on fibrinogen in control cells. Similarly, constitutive β3 and focal adhesion kinase phosphorylation was observed in resting patients’ platelets but not in control platelets (Figure 3A-C). Finally, clot retraction induced by CHO cells expressing the mutant receptor was decreased (Figure 3D) as well as clot retraction of patients’ PRP (Figure 3E).

Constitutive αIIbβ3 activation leads to impaired cytoskeletal reorganization

The content of polymerized actin (F-actin) of CHO cells expressing each of the heterozygous, β3-activating mutants described so far11,17-19 together with normal αIIb was higher than that of CHO cells expressing normal αIIbβ3, and it did not increase after stimulation with DTT (Figure 4A). Similarly, F-actin content was significantly higher in resting patients’ platelets as compared with con-

Figure 3. αIIbβ3 –triggered outside-in signaling. (A) β3 Tyr773 phosphorylation of wild-type (WT) and mutant (MUT) integrin β3 -expressing CHO cells and of control and one patient’s platelets in suspension (susp) and after adhesion to fibrinogen (Fbg). Relative quantification (arbitrary units), CHO cells: WT susp=1, WT Fbg=29.6±4.7, MUT susp=55.3±6.8, MUT Fbg,=69.6±12.3 (n=5, P<0.05 vs. WT); Platelets: control susp=1, control Fbg=5.1±1.5, patient susp=7.1±1.8, patient Fbg=8.6±1.3 (n=5, P<0.05 vs. control). (B) β3 Tyr785 phosphorylation of wild-type and mutant integrin β3-expressing CHO cells and of control and one patient’s platelets in suspension and after adhesion to fibrinogen. Relative quantification (arbitrary units), CHO cells: WT susp=1, WT Fbg=41.3±2, MUT susp=62.3±7.5, MUT Fbg=65.7±5.5 (n=5, P<0.05 vs. WT); Platelets: control susp=1, control Fbg=5.7±1.6, patient susp=2.4±1.1, patient Fbg= 2.9±0.6 (n=5, P<0.05 vs. control). (C) Focal adhesion kinase (FAK) phosphorylation of wild-type and mutant integrin β3 expressing CHO cells and of control and one patient’s platelets in suspension and after adhesion to fibrinogen. Relative quantification (arbitrary units), CHO cells: WT susp=1, WT Fbg=7.9±2.5, MUT susp=5.2±0.7, MUT Fbg=5.1±0.3 (n=5, P<0.05 vs. WT); Platelets: control susp=1, control Fbg=3.9±0.8, patients susp=2.1±0.7, patients Fbg=4.5±0.9 (n=5, P<0.05 vs. control). (D) Clot retraction induced by wild-type (WT) and the different mutant integrin β3-expressing CHO cells after 60 and 90 min of incubation at 37°C (*P<0.01 vs. WT). (E) Clot retraction induced by control platelets (Ctrl), Glanzmann thrombasthenia platelets (GT) and patients’ (Pat) platelets after 60 and 90 min of incubation at 37°C (*p<0.01 vs. control). (F) Representative images of WT and the different mutant integrin β3-expressing CHO cells after 90 min of incubation at 37°C. (G) Representative images of platelets from a control, a subject with Glanzmann thrombasthenia and one patient after 90 min of incubation at 37°C.

haematologica | 2016; 101(1)

Mechanism of platelet dysfunction in dominant GT

trol platelets. Consistently, stimulation with ADP did not significantly increase F-actin content of patients’ platelets while it doubled it in control platelets (Figure 4B). Cytoskeletal protein content was higher in resting patients’ platelets than in resting control platelets. Moreover, electrophoresis showed three overexpressed bands in patients’ platelets when compared to resting control platelets. One had a molecular weight of about 70 kDa (band A) while the other two bands migrated at 55 kDa (bands B and C) (Figure 4C). Mass spectrometry showed that band A was mainly composed of fibrinogen α-chain, band B of fibrinogen β-chain, and band C of a fragment of fibrinogen γ-chain (Online Supplementary Tables S1 and S2, Online Supplementary Figure S5).

Perturbed cytoskeletal reorganization causes platelet dysfunction To induce actin polymerization in control platelets we used jasplakinolide, a natural cyclic peptide that induces actin polymerization and stabilizes actin filaments.31 We first determined that jasplakinolide is able to induce actin polymerization in control platelets (Online Supplementary Figure S6). Incubation with jasplakinolide did not affect αIIbβ3 surface expression (Figure 5A), but significantly impaired its activation, induced by ADP (Figure 5B). Pre-incubation with jasplakinolide impaired, dosedependently, ADP-induced platelet aggregation (Figure 5C), clot retraction (Figure 5D) and spreading on fibrinogen (Figure 5E). Platelets treated with jasplakinolide (1 mM) resembled patients’ platelets with regards to their spreading morphology11 (Figure 5F). These effects were not a consequence of cell death because platelets were still able to expose P-selectin on their surface upon stimulation with 10 mM ADP (resting platelets 2.92±1.33%; ADPstimulated platelets 55.03±3.1%; jasplakinolide-treated resting platelets 4.52±1.8%; jasplakinolide-treated ADPstimulated platelets 57.21±7.9%).

Perturbed actin polymerization impairs proplatelet formation The proportion of jasplakinolide-treated human megakaryocytes extending proplatelets in suspension was not different from that of control megakaryocytes (vehicle=26.3±2.6%; jasplakinolide=26.9±9.3%, n=3, P=ns); however, proplatelet formation was abnormal, with most megakaryocytes displaying a reduced number of proplatelet tips (vehicle=13.9±6.1 tips; jasplakinolide=1.4±1.4 tips, n=3, P<0.01) and a reduced tip diameter (vehicle=2.9±0.9 mm; jasplakinolide=1.7±0.8 mm, n=3, P<0.01) (Figure 7). Adhesion to fibrinogen of jasplakinolide-treated megakaryocytes was strongly impaired (vehicle=41.4±5.7%; jasplakinolide=4.6±0.5%, n=3, P<0.01) and the few adhering megakaryocytes did not spread (Figure 7).

Structural consequences of del647-686

The β-TD connects the lower β-leg of integrin β3 to the transmembrane portion and consists of an amphipathic α-

Constitutively activated αIIbβ3 impairs proplatelet formation

The proportion of mutant-β3-transduced megakaryocytes extending proplatelets in suspension was not different from that of wild-type megakaryocytes (wild-type megakaryocytes=73.8±19.3%, mutant megakaryocytes=61.4±20.7%; n=5, P=ns). However, proplatelet number was reduced and tip diameter was larger in mutant megakaryocytes (Figure 6A,B). Moreover, barbellproplatelets were significantly more asymmetrical (difference between tips: wild-type megakaryocytes=1±0.9 mm, mutant megakaryocytes=2.5±1.1 mm; n=5, P<0.01). Similarly, barbell-proplatelets circulating in peripheral blood of our patient were more asymmetrical than in normal controls (difference between tips: controls=0.5±0.4 µm; patient=1.2±0.8 mm, n=5, P<0.05) (Figure 6C). Mutant β3-transduced megakaryocytes adhered normally to fibrinogen (wild-type megakaryocytes=51.2±15.3%, mutant megakaryocytes=44.6±20.1%; n=5, P=ns) but displayed abnormal spreading, reminiscent of what was previously observed with patient’s megakaryocytes3 (Figure 6D). In contrast, when megakaryocytes were plated on von Willebrand factor, spreading was not different from that of controls (wild-type megakaryocytes=47.9±19.3%, mutant megakaryocytes=50.2±22.6%; n=5, P=ns), as already observed with patient’s megakaryocytes.3

Figure 4. Cytoskeleton dynamics. (A) Polymerized actin (F-actin) content of controls and patients’ platelets measured by flow cytometry before and after stimulation with ADP (20 mM) (n=4, *P<0.05 vs. resting, #P<0.05 vs. control). (B) Polymerized actin (F-actin) content of wild-type (WT) and of different mutant β3expressing CHO cells measured by flow cytometry before and after stimulation with DTT (10 mM) (n=6, *P<0.05 vs. resting, #P<0.05 vs. control). (C) Representative electrophoresis of cytoskeletal proteins from resting control platelets (Ctrl1, Ctrl2), control platelets stimulated with thrombin (thr) and platelets from the two different del647-686 β3 patients (Pat1, Pat2). Relative quantification (arbitrary units), Control=1; thr=2±0.8, n=3, P<0.01; Patients=1.2±0.1 (n=3, P<0.01 vs. ctrl). Molecular weight marker on the left, identity of the main bands from mass spectrometry analysis on the right. All the proteins found in each band are listed in Online Supplementary Table S1.

haematologica | 2016; 101(1)

L. Bury et al. helix lying across a single β-sheet with four β-strands and four disulfide bonds; it non-covalently associates with the calf-2 domain of αIIb (Figure 8). Residues spanning from Lys532 through Gly690 in the lower β-leg stabilize the interactions with integrins αv and αIIb in the bent conformation, and mutations of these residues activate αIIbβ3 and trigger fibrinogen binding.36 Del647-686 removes the 2, 3, and 4 β-strands, which represent a large interaction interface with the calf-2 domain in both αIIbβ312 and αvβ3,34 thus destroying the β-TD structure (Figure 8), hindering the adoption of the bent conformation. The observation that, despite a 40-amino acid deletion, αIIbβ3 is still synthesized and expressed on the surface, confirms that β-TD is a domain important for integrin function/activation but not for β3 synthesis or dimerization with αIIb.12 Del647-686 also removes Cys-655 and 663, partners of Cys-608 and 687 in the formation of disulfide bonds. Whether the remaining cysteines would then disulfidelink to one another, exchange with the two α1-helix-loop disulfides, or remain as free sulfhydryls is unknown. However, previous site-directed mutagenesis of Cys-655

and 663 leading to the loss of disulfide bonds, resulted in constitutive activation of αIIbβ3,13-15 suggesting that the same mechanism is responsible for the constitutive αIIbβ3 activation associated with del647-686.

Discussion Our results show that constitutive activation of αIIbβ3 due to gain-of-function mutations of ITGB3, and the consequent permanent triggering of αIIbβ3-mediated outside-in signaling, induce a perturbation of cytoskeletal remodeling that leads to platelet dysfunction and impaired proplatelet formation. Starting from the study of platelets from two patients carrying the heterozygous ITGB3 del647-686 mutation11 we extended our observations to three additional ITGB3 gain-of-function mutations17-19 in order to describe a general mechanism that leads to bleeding in patients with dominant variants of GT. We show that constitutive activation of αIIbβ3 leads to permanent triggering of αIIbβ3-mediated outside-in signal-

Figure 5. Effect of jasplakinolide on platelet function. (A) αIIβ3 expression in platelets treated with jasplakinolide 5 mM (JAS) or its vehicle (CTRL) as assessed by flow-cytometry (n=4, P=ns). Data show mean fluorescence intensity. (B) PAC-1 binding to platelets induced by ADP (10 mM) after preincubation with jasplakinolide 5 mM (JAS) or its vehicle (CTRL) (n=4, *P<0.05 vs. resting, #P<0.05 vs. ctrl ADP). Data show mean fluorescence intensity. (C) ADP-induced aggregation of platelets treated with jasplakinolide 5 μM (JAS) or its vehicle (CTRL) (n=4, *P<0.01 vs. ctrl). (D) Clot retraction mediated by platelets treated with jasplakinolide 5 μM (JAS) or its vehicle (CTRL) (n=4, *P<0.01 vs. ctrl). (E) Platelet spreading on fibrinogen after treatment with 1 μM jasplakinolide. Platelets were stained with the CD41 clone P2 monoclonal antibody. (F) Spreading on fibrinogen of platelets treated with jasplakinolide (JAS) or its vehicle (CTRL) after 60 min of deposition (n=4, *P<0.01 vs. control).

haematologica | 2016; 101(1)

Mechanism of platelet dysfunction in dominant GT ing, as shown by the phosphorylation of β3 Tyr773, Tyr785 and focal adhesion kinase in patients’ resting platelets and in CHO cells expressing del647-686 β3, by faster spreading on fibrinogen and spontaneous spreading on von Willebrand factor of patients’ platelets, contrarily to control platelets that require αIIbβ3 activation for spreading on von Willebrand factor.18,24 Activation of αIIbβ3 is physiologically followed by receptor internalization, a way for limiting platelet aggregation.37,38 We show here that enhanced αIIbβ3 internalization is the common mechanism leading to reduced surface expression of αIIbβ3 in patients with GT-like syndromes associated with constitutive activation of αIIbβ3 .10 In fact, we observed enhanced αIIbβ3 internalization in CHO cells expressing several different ITGB3 gain-of-function muta-

haematologica | 2016; 101(1)

tions17-19 as well as in platelets of del647-686 β3 patients. We have previously shown, by flow cytometry using a set of clones directed toward different epitopes of αIIbβ3, that our patients’ platelets express on average 40% residual αIIbβ3 on their surface if compared with control platelets.11 We show here that in patients’ platelets the ratio between normal and mutant β3 is maintained on the platelet surface (Figure 1A). The same is observed in CHO cells transfected with both wild-type and mutant β3 (heterozygous cell model) which express an approximately equal amount of αIIbβ3 to that of CHO cells transfected with only mutant β3 (homozygous cell model) (Figure 1A,B). It can, therefore, be speculated that normal αIIbβ3 is passively internalized along with the mutant one. Confocal microscopy revealed that in patients’ platelets

Figure 6. Megakaryocyte spreading and proplatelet formation. (A) Number and diameter of proplatelet tips generated by murine megakaryocytes transduced with the wild-type (WT) or the mutant (MUT) integrin β3 (*P<0.01 vs. WT). Only cells showing both green and red fluorescence (i.e. expressing αIIb and β3) were analyzed. (B) Representative image of proplatelet formation by wild type- or mutant β3–expressing murine megakaryocytes. The number of megakaryocytes expressing normal or mutant αIIbβ3 was comparable (GFP expression in WT 20.2±1.7% vs. mutant 19.6±1.3%; Ds-red expression in WT 18.6±1.2% vs. mutant 17.7±1.9%; n=5, P=ns). Megakaryocytes were stained with a rabbit anti-mouse β1-tubulin antibody; nuclei were stained with Hoechst. (C) An asymmetric barbellproplatelet circulating in a patient’s peripheral blood is indicated by a white arrow (left). Symmetric barbell-proplatelet generated by wildmurine type β3-expressing megakaryocytes compared with an asymmetric barbell-proplatelet generated by mutant β3-expressing murine megakaryocytes (right). Barbell-proplatelets were stained with a rabbit anti-human β1-tubulin antibody. (D) Representative picture of wild type- or mutant β3–expressing murine megakaryocytes spreading on fibrinogen. β1-tubulin was stained with a rabbit anti-human β1tubulin, actin was stained by rhodamine-phalloidin, nuclei were stained with Hoechst.

L. Bury et al. the internal pool of integrin β3 localizes in α-granules and in other cytoplasmic compartments, probably corresponding to the open canalicular and dense tubular systems,35 similar to control platelets. Moreover, we show that upon platelet activation, mutant αIIbβ3 externalizes as well as normal αIIbβ3, and is, therefore, presumably correctly recycled after internalization. Enhanced αIIbβ3 internalization was associated with an increased platelet content of fibrinogen. The intra-platelet fibrinogen pool is largely generated by its internalization mediated by activated αIIbβ3 ,39,40 although internalization by inactive αIIbβ3 has also been reported.41 Increased fibrinogen content in the patients’ platelets is, therefore, in accordance with constitutively activated αIIbβ3. Given that integrin β3 is a component of the vitronectin

receptor αvβ3, we measured αvβ3 expression on patients’ platelets and found it to be comparable to that on control platelets. Therefore, our mutation, like others previously reported,9 influences αvβ3 and αIIbβ3 expression differently. In normal conditions, αIIbβ3 -mediated outside-in signaling leads to a reorganization of the cytoskeleton that is required for platelet aggregation, clot retraction and spreading. Cytoskeletal actin reorganization is a finely regulated process with consecutive phases including actin polymerization, which enhances the cytoskeletal rigidity required for platelet shape change, and then depolymerization, which restores the cytoskeletal plasticity required for aggregation and clot retraction.42 An alteration of actin dynamics and cytoskeletal remodeling can, therefore, impair platelet function.7,8

C Figure 7. Effect of jasplakinolide on megakaryocyte spreading and proplatelet formation. Representative pictures of proplatelet formation (top and center) and spreading on fibrinogen (bottom) by megakaryocytes treated with vehicle (Ctrl) or 1 mM jasplakinolide (JAS). Megakaryocytes treated with jasplakinolide display a reduced number of proplatelet tips of altered diameter (arrows) and impaired spreading on fibrinogen.

haematologica | 2016; 101(1)

Mechanism of platelet dysfunction in dominant GT

Figure 8. Structural consequences of del647-686. Cartoon diagram showing the structure of the β-tail domain, with each domain in a different color. The deleted region in del647-686 is colored magenta and the remainder of the βtail domain is colored green. Disulfides and Val-664 side chain are shown as sticks, with sulfur atoms in gold. Ectodomain subunit termini are marked with α and β. The β-tail domain secondary structure elements are variable in length in different β3 integrin ectodomain structures, and are long in the αVβ3 structure depicted here (PDB ID code 4G1E).36

We show here that CHO cells expressing ITGB3 gainof-function mutations, as well as del647-686 β3-bearing platelets, have an increased F-actin content under resting conditions and show impaired clot retraction, suggesting that actin turnover is arrested at the stage of polymerization due to permanently triggered outside-in signaling. This is in line with the earlier spreading on fibrinogen of mutant platelets becoming defective at later time points, because actin turnover is arrested at the stage of polymerization and further cytoskeletal remodeling is no longer possible. Moreover, patients’ platelets had an enhanced cytoskeletal-associated protein content as compared with control platelets, and fibrinogen α, β and γ chains were found to be associated with the cytoskeleton. In order to assess whether impaired cytoskeletal remodeling recapitulates the platelet features associated with αIIbβ3 -activating ITGB3 mutations we used jasplakinolide, a natural cyclic peptide that induces actin polymerization by promoting actin nucleation and by preventing depolymerization of stress fibers.31 Jasplakinolide-treated control platelets showed impaired αIIbβ3 activation, reduced aggregation, altered spreading on fibrinogen and clot retraction, a phenotype resembling that of del647-686 platelets.11 Altogether, these data are compatible with a model in which αIIbβ3 constitutive activation causes the arrest of cytoskeletal remodeling at haematologica | 2016; 101(1)

the stage of polymerization, thus impairing platelet aggregation, clot retraction and spreading. On the other hand, reduced αIIbβ3 surface expression does not seem to contribute significantly to platelet dysfunction because incubation of normal platelets with jasplakinolide induced the same platelet dysfunction observed in patients with variant GT without affecting αIIbβ3 surface expression. Moreover, αIIbβ3 internalization is not triggered by actin polymerization but depends on the constitutive activation of the receptor. Cytoskeletal remodeling is also crucial for proplatelet formation: in fact actin polymerization leads to the amplification of proplatelet ends, thus increasing tip number,43 and regulates platelet size.35,36 Impaired actin remodeling in megakaryocytes may thus perturb platelet formation, and in fact defects in actin-related proteins were shown to lead to thrombocytopenia and to the production of platelets with altered dimensions.44-47 We induced actin polymerization by incubating control megakaryocytes with jasplakinolide and indeed we observed impaired proplatelet formation. Murine megakaryocytes transduced with del647-686 human β3 and wild-type human αIIb showed impaired proplatelet formation, with a reduced number of abnormally large proplatelet tips. These data confirm our previous observations with peripheral blood CD34+-derived patients’ megakaryocytes3 and prove that the 2134+1 G>C mutation is the cause of macrothrombocytopenia. We also observed that barbell-proplatelets produced by transduced murine megakaryocytes are asymmetrical, similar to those found in patients’ peripheral blood. This is therefore probably responsible for the platelet anisocytosis observed in patients with the del647-686 mutation.11 Our observations confirm that αIIbβ3 plays a role in proplatelet formation3,18,48 and show that when outside-in signaling is constitutively triggered, cytoskeletal reorganization is disturbed and altered proplatelet formation and preplatelet maturation occur. In conclusion, we show that gain-of-function mutations of ITGB3 generating constitutive activation of αIIbβ3 lead to receptor internalization and to the arrest of cytoskeletal remodeling in platelets and megakaryocytes which, in turn, generate platelet dysfunction and perturb the finely regulated process of proplatelet formation leading to macrothrombocytopenia. Reduced platelet number and platelet dysfunction in patients with dominant variants of GT are both consequent to the cytoskeletal perturbation induced by the constitutive αIIbβ3 -mediated outside-in signaling. Funding This work was supported by a Telethon grant (GGP10155) to PG. Acknowledgments The authors thank Ildo Nicoletti and Alessandra Balduini for help with confocal microscopy, and Silvia Giannini, Teresa Corazzi, Luca Cecchetti, Anna Maria Mezzasoma, Giuseppe Guglielmini and Viviana Appolloni for help with some experiments and discussion of results. W. Vainchenker (Université Paris-Sud, Villejuif, France) kindly gave the GP+E-86 TPOsecreting cell line, and F. Grignani (Department of Experimental Medicine, University of Perugia) the Phoenix cell line. We also thank Francesca for her kind and continuous collaboration. 55

L. Bury et al.

References 1. Kasirer-Friede A, Kahn ML, Shattil SJ. Platelet integrins and immunoreceptors. Immunol Rev. 2000;218:247-264. 2. Larson MK, Watson SP. Regulation of proplatelet formation and platelet release by integrin αIIβ3. Blood. 2006;108(5):1509-1514. 3. Bury L, Malara A, Gresele P, Balduini A. Outside-in signalling generated by a constitutively activated integrin αIIβ3 impairs proplatelet formation in human megakaryocytes. PLoS One. 2012;7(4):e34449. 4. Springer TA, Dustin ML. Integrin inside-out signaling and the immunological synapse. Curr Opin Cell Biol. 2012;24(1):107-115. 5. Shattil SJ, Kim C, Ginsberg MH. The final steps of integrin activation: the end game. Nature Rev. 2010;1(4):288-300. 6. Fox JE. Cytoskeletal proteins and platelet signaling. Thromb Haemost. 2001; 86(1):198-213. 7. Berrou E, Adam F, Lebret M, et al. Heterogeneity of platelet functional alterations in patients with filamin A mutations. Arterioscler Thromb Vasc Biol. 2013; 33(1):e11-18. 8. Canobbio I, Noris P, Pecci A, Balduini A, Balduini CL, Torti M. Altered cytoskeleton organization in platelets from patients with MYH9-related disease. J Thromb Haemost. 2005;3(5):1026-1035. 9. Nurden AT, Fiore M, Nurden P, Pillois X. Glanzmann thrombasthenia: a review of ITGA2B and ITGB3 defects with emphasis on variants, phenotypic variability, and mouse models. Blood. 2011;118(23):59966005. 10. Nurden AT, Pillois X, Fiore M, Heilig R, Nurden P. Glanzmann thrombasthenia-like syndromes associated with macrothrombocytopenias and mutations in the genes encoding the αIIβ3 integrin. Semin Thromb Hemost. 2011;37(6):698-706. 11. Gresele P, Falcinelli E, Giannini S, et al. Dominant inheritance of a novel integrin β3 mutation associated with a hereditary macrothrombocytopenia and platelet dysfunction in two Italian families. Haematologica. 2009;94(5):663-669. 12. Zhu J, Luo BH, Xiao T, Zhang C, Nishida N, Springer TA. Structure of a complete integrin ectodomain in a physiologic resting state and activation and deactivation by applied forces. Mol Cell. 2008;32(6):849-861. 13. Butta N, Arias-Salgado EG, GonzálezManchón C, et al. Disruption of the β3 663687 disulfide bridge confers constitutive activity to β3 integrins. Blood. 2003;102 (7):2491-2497. 14. Mor-Cohen R, Rosenberg N, Landau M, Lahav J, Seligsohn U. Specific cysteines in β3 are involved in disulfide bond exchangedependent and -independent activation of αIIbβ3. J Biol Chem. 2008; 283(28):1923519244. 15. Mor-Cohen R, Rosenberg N, Einav Y, et al. Unique disulfide bonds in epidermal growth factor (EGF) domains of β3 affect structure and function of αIIbβ3 and αvβ3 integrins in different manner. J Biol Chem. 2012;287(12): 8879-8891. 16. Kashiwagi H, Kunishima S, Kiyomizu K, et al. Demonstration of novel gain-of-function mutations of αIIbβ3: association with macrothrombocytopenia and Glanzmann

17.

18.

19.

20.

21. 22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

thrombasthenia-like phenotype. Mol Genet Genomic Med. 2013;1(2):77-86. Vanhoorelbeke K, De Meyer SF, Pareyn I, et al. The novel S527F mutation in the integrin β3 chain induces a high affinity αIIbβ3 receptor by hindering adoption of the bent conformation. J Biol Chem. 2009;284(22): 14914-14920. Ghevaert C, Salsmann A, Watkins NA, et al. A nonsynonymous SNP in the ITGB3 gene disrupts the conserved membrane-proximal cytoplasmic salt bridge in the αIIbβ3 integrin and cosegregates dominantly with abnormal proplatelet formation and macrothrombocytopenia. Blood. 2008;111(7):3407-3414. Jayo A, Conde I, Lastres P, et al. L718P mutation in the membrane-proximal cytoplasmic tail of β3 promotes abnormal αIIbβ3 clustering and lipid microdomain coalescence, and associates with a thrombasthenia-like phenotype. Haematologica. 2010;95(7):1158-1166. Edelheit O, Hanukoglu A, Hanukoglu I. Simple and efficient site-directed mutagenesis using two single-primer reactions in parallel to generate mutants for protein structure-function studies. BMC Biotechnol. 2009;9:61. Yan B, Smith JW. Mechanism of integrin activation by disulfide bond reduction. Biochemistry. 2001;40(30):8861-8867. Giannini S, Mezzasoma A, Guglielmini G, Rossi R, Falcinelli E, Gresele P. A new case of acquired Glanzmann’s thrombasthenia: diagnostic value of flow cytometry. Cytometry B Clin Cytom. 2008;74(3):194-199. Schober JM, Lam SC, Wencel-Drake JD. Effect of cellular and receptor activation on the extent of integrin αIIbβ3 internalization. J Thromb Haemost. 2003;1(11):2404-2410. Kieffer N, Fitzgerald LA, Wolf D, Cheresh DA, Phillips DR. Adhesive properties of the β3 integrins: comparison of GPIIb-IIIa and the vitronectin receptor individually expressed in human melanoma cells. J Cell Biol. 1991;113(2):451-461. Flevaris P, Stojanovic A, Gong H, Chishti A, Welch E, Du X. A molecular switch that controls cell spreading and retraction. J Cell Biol. 2007;179(3):553-565. Falcinelli E, Guglielmini G, Torti M, Gresele P. Intraplatelet signaling mechanisms of the priming effect of matrix metalloproteinase-2 on platelet aggregation. J Thromb Haemost. 2005;3(11):2526-2535. Susta F, Chiasserini D, Fettucciari K, et al. Protein expression changes induced in murine peritoneal macrophages by group B Streptococcus. Proteomics. 2010;10(11): 2099-2112. Tabb DL, Fernando CG, Chambers MC. MyriMatch: highly accurate tandem mass spectral peptide identification by multivariate hypergeometric analysis. J Proteome Res. 2007;6(2):654-661. Ma ZQ, Dasari S, Chambers MC, et al. IDPicker 2.0: improved protein assembly with high discrimination peptide identification filtering. J Proteome Res. 2009;8(8): 3872-3881. Momi S, Falcinelli E, Giannini S, et al. Loss of matrix metalloproteinase 2 in platelets reduces arterial thrombosis in vivo. J Exp Med. 2009;206(11):2365-2379. Bubb MR, Senderowicz AM, Sausville EA, Duncan KL, Korn ED. Jasplakinolide, a cytotoxic natural product, induces actin polymerization and competitively inhibits the

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42. 43. 44.

45.

46.

47. 48.

binding of phalloidin to F-actin. J Biol Chem. 1994;269(21):14869-14871. Thon JN, Montalvo A, Patel-Hett S, et al. Cytoskeletal mechanics of proplatelet maturation and platelet release. J Cell Biol. 2010;191(4):861-874. Thon JN, Macleod H, Begonja AJ, et al. Microtubule and cortical forces determine platelet size during vascular platelet production. Nat Commun. 2012;3:852. Dong X, Mi LZ, Zhu J, et al. αvβ3 integrin crystal structures and their functional implications. Biochemistry. 2012;51(44):88148828. Cramer ER, Savidge GF, Vainchenker W, et al. Alpha-granule pool of glycoprotein IIbIIIa in normal and pathologic platelets and megakaryocytes. Blood. 1990;75(6):12201227. Donald JE, Zhu H, Litvinov RI, DeGrado WF, Bennett JS. Identification of interacting hot spots in the β3 integrin stalk using comprehensive interface design. J Biol Chem. 2010;285(49):38658-38665. Bennett JS, Zigmond S, Vilaire G, Cunningham ME, Bednar B. The platelet cytoskeleton regulates the affinity of the integrin αIIbβ3 for fibrinogen. J Biol Chem. 1999;274(36):25301-25307. Wencel-Drake JD, Boudignon-Proudhon C, Dieter MG, Criss AB, Parise LV. Internalization of bound fibrinogen modulates platelet aggregation. Blood. 1996;87(2): 602-612. Hung WS, Huang CL, Fan JT, et al. The endocytic adaptor protein Disabled-2 is required for cellular uptake of fibrinogen. Biochim Biophys Acta. 2012;1823(10):1778-1788. Hato T, Pampori N, Shattil SJ. Complementary roles for receptor clustering and conformational change in the adhesive and signaling functions of integrin αIIbβ3. J Cell Biol. 1998;141(7):1685-1695. Handagama P, Scarborough RM, Shuman MA, Bainton DF. Endocytosis of fibrinogen into megakaryocyte and platelet alpha-granules is mediated by alpha IIb beta 3 (glycoprotein IIb-IIIa). Blood. 1993;82(1):135-138. Bearer EL, Prakash JM, Li Z. Actin dynamics in platelets. Int Rev Cytol. 2002;217:137182. Hartwig JH, Italiano JE Jr. Cytoskeletal mechanisms for platelet production. Blood Cells Mol Dis. 2006;36(2):99-103. Chen Z, Shivdasani RA. Regulation of platelet biogenesis: insights from the MayHegglin anomaly and other MYH9-related disorders. J Thromb Haemost. 2009;7 (Suppl 1):272-276. Nurden P, Debili N, Coupry I, et al. Thrombocytopenia resulting from mutations in filamin A can be expressed as an isolated syndrome. Blood. 2011;118(22):59285937. Kunishima S, Okuno Y, Yoshida K, et al. ACTN1 mutations cause congenital macrothrombocytopenia. Am J Hum Genet. 2013;92(3):431-438. Thon JN, Italiano JE Jr. Does size matter in platelet production? Blood. 2012;120(8): 1552-1561. Kunishima S, Kashiwagi H, Otsu M, et al. Heterozygous ITGA2B R995W mutation inducing constitutive activation of the αIIbβ3 receptor affects proplatelet formation and causes congenital macrothrombocytopenia. Blood. 2011;117(20):5479-5484.

haematologica | 2016; 101(1)

ARTICLE

Bone Marrow Failure

PPARγ antagonist attenuates mouse immunemediated bone marrow failure by inhibition of T cell function

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Kazuya Sato,1* Xingmin Feng,1* Jichun Chen,1 Jungang Li,1Pawel Muranski,1 Marie J. Desierto,1 Keyvan Keyvanfar,1 Daniela Malide,2 Sachiko Kajigaya,1 and Neal S. Young1 *Authors contributed equally Hematology Branch, 2Light Microscopy Core Facility, National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, MD, USA 1

Haematologica 2016 Volume 101(1):57-67

ABSTRACT

cquired aplastic anemia is an immune-mediated disease, in which T cells target hematopoietic cells; at presentation, the bone marrow is replaced by fat. It was reported that bone marrow adipocytes were negative regulators of hematopoietic microenvironment. To examine the role of adipocytes in bone marrow failure, we investigated peroxisomal proliferator-activated receptor gamma, a key transcription factor in adipogenesis, utilizing an antagonist of this factor called bisphenol-A-diglycidyl-ether. While bisphenol-A-diglycidyl-ether inhibited adipogenesis as expected, it also suppressed T cell infiltration of bone marrow, reduced plasma inflammatory cytokines, decreased expression of multiple inflammasome genes, and ameliorated marrow failure. In vitro, bisphenol-A-diglycidyl-ether suppressed activation and proliferation, and reduced phospholipase C gamma 1 and nuclear factor of activated T-cells 1 expression, as well as inhibiting calcium flux in T cells. The in vivo effect of bisphenol-A-diglycidyl-ether on T cells was confirmed in a second immune-mediated bone marrow failure model, using different strains and non-major histocompatibility antigen mismatched: bisphenol-A-diglycidyl-ether ameliorated marrow failure by inhibition of T cell infiltration of bone marrow. Our data indicate that peroxisomal proliferator-activated receptor gamma antagonists may attenuate murine immune-mediated bone marrow failure, at least in part, by suppression of T cell activation, which might hold implications in the application of peroxisomal proliferator-activated receptor gamma antagonists in immune-mediated pathophysiologies, both in the laboratory and in the clinic. Genetically “fatless” mice developed bone marrow failure with accumulation of marrow adipocytes in our model, even in the absence of body fat, suggesting different mechanisms of systematic and marrow adipogenesis and physiologic versus pathophysiologic fat accumulation.

Introduction Aplastic anemia (AA) is the paradigmatic bone marrow (BM) failure syndrome in humans.1,2 AA behaves as an immune-mediated disease in most patients: activated cytotoxic T cells and type I cytokines destroy hematopoietic stem and progenitor cells, resulting in pancytopenia and absence of hematopoietic precursors in the BM.1,2 The BM of patients with AA is typically described as “empty”, but in reality the hypocellular marrow space is occupied by fat, and specifically increased numbers of large adipocytes.3 BM adipocytes in AA have been assumed to passively haematologica | 2016; 101(1)

Correspondence: fengx2@nhlbi.nih.gov

Received: 2/12/2014. Accepted: 17/11/2015. Pre-published: 20/11/2015. doi:10.3324/haematol.2014.121632

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/1/57

K. Sato et al.

occupy marrow and to be metabolically inert under most physiological conditions.4 Recently, evidence has been presented to support the notion that BM adipocytes might play a central function in regulating hematopoiesis.4-6 Gene expression profiles suggest that mouse BM adipocytes

possess a phenotype functionally distinct from extramedullary fat cells:6 for example, inflammatory response genes, such as Tnfa, Il6, and Tgf1, are highly expressed, while expression of adipose-specific genes is low.6 In one prominent report, pharmacological inhibition

B E

Figure 1. PPARγ antagonists ameliorate immune-mediated BM failure in AA mice. (A) Effects of PPARγ antagonists on BM adipocytes and cellularity in AA mice. In confocal imaging (left panels), adipocytes and nuclei in sternums were stained with BODIPY (493/503 nm, green) and DAPI (405 nm, blue). Right panels show H&E staining of mouse femurs in different groups. Representative images of AA (n=16), BADGE-treated (n=16), GW9662-treated (n=10) are shown. Blood counts (B) and BM total nucleated cell numbers of AA (n=16), BADGE-treated (n=16), and GW9662-treated (n=10), frequency of BM lineage negative cells (Lin–) and absolute number of Lin–Sca1+c-kit+ (LSK) cells (C) of AA (n=7), BADGE-treated mice (n=7). Dose response to BADGE on BM adipogenesis in confocal imaging (D) and on BM total nucleated cell numbers (E) in AA (n=6), low dose BADGE (15 mg/kg, n=6), standard dose BADGE (30 mg/kg, n=8), and high dose BADGE (60 mg/kg, n = 6) treated mice. All results shown are at 2 weeks. All error bars indicate SEM. *P<0.05; **P<0.01; ***P<0.001. BM hematopoietic stem and progenitor cells (HSPCs, CD117Texas Red) (F), megakaryocytes (CD41-PE) (G), and adipocytes (BODIPY, green) in TBI only, AA control mice, and BADGE-treated mice were visualized by confocal imaging of sternums at 2 weeks in representative images. Original magnification, x100 (A left, D, F, and G); x40 (A right).

haematologica | 2016; 101(1)

PPARγ antagonist in bone marrow failure

of peroxisomal proliferator-activated receptor gamma (PPARγ), a master regulator of adipogenesis, enhanced BM engraftment in mice after BM transplantation.7 PPARγ is a family of ligand-activated nuclear receptor transcription factors. There are two distinct isoforms, PPARγ1 and PPARγ2.8-10 PPARγ2 is exclusively expressed in adipose tissue, while PPARγ1 is ubiquitously expressed at lower levels in other tissues, such as skeletal muscle, liver, breast, prostate, and colon.8-12 PPARγ plays a critical role in lipid metabolism and also in immune function, cell growth, differentiation, and apoptosis.8-15 In particular, cells of the immune system, such as normal macrophages, dendritic cells, eosinophils, T cells and B cells, also express PPARγ. Collectively, these findings prompted us to revisit the role of adipogenesis in BM failure and the possibility of PPARγ as a therapeutic target in AA. In this study, we investigated the roles of bisphenol-Adiglycidyl-ether (BADGE, a PPARγ antagonist) in our AA

mouse model. BADGE not only suppressed BM adipogenesis, but also inhibited T cell infiltration of BM and improved the marrow inflammatory environment, eventually ameliorating pancytopenia and BM destruction. In vitro BADGE suppressed T cell activation and proliferation, and reduced T-cell cytokine secretion. We also tested the antagonist in a second immune-mediated BM failure murine model, using different strains and non-major histocompatibility (non-MHC) mismatched. Unexpectedly, we observed the accumulation of BM adipocytes in genetically “fatless” mice in our marrow failure model.

Methods Mice Inbred C57BL/6 (B6, H2b/b), DBA/1J (DBA/1, H2q/q), FVB/NJ (FVB, H2q/q) mice and congenic C.B10- H2b/b/LilMcd (CB10, H2b/b) mice

haematologica | 2016; 101(1)

Figure 2. Alterations of inflammation and adipogenesis by PPARγ antagonists. (A) Inflammation, Th1 and adipogenesis related cytokines in AA mice treated with PPARγ antagonists. Plasma samples were collected at 2 weeks, and were divided for each of two different MILLIPLEX MAP Mouse Magnetic Bead Panels in order to measure cytokine concentrations. All error bars indicate SEM of TBI alone (n=5), AA (n=8), BADGE-treated (n=8), GW9662-treated (n=5) mice. *P<0.05; **P<0.01; ***P<0.001. Quantitative analysis of mRNA expression of adipogenesis- (B), and inflammasome- (C) related genes in whole femurs showing more than 2-fold changes in BADGE-treated AA mice (n=4), compared with untreated AA mice (n=4) at 2 weeks. Blue represents low expression, red indicates high expression. Each column of the heat map represents one individual mouse. (D) Immunoblot to validate protein levels of selected adipogenesis-related genes PPARγ and AGT obtained from PCR arrays in whole femurs. (E) Protein levels of PPARγ in T cells from BM of AA mice and BADGEtreated AA mice. β-actin was used as a loading control. Representative images of at least three independent experiments are shown.

K. Sato et al.

were purchased from The Jackson Laboratory (Bar Harbor, ME, USA). FVB.A-ZIP/F1 (A-ZIP/F) “fatless” mice and their wild-type littermates were kindly provided by Dr. Charles Vinson (NCI, NIH). All mice were bred and maintained at NIH animal facilities under standard care and nutrition. Female mice aged from 8 to 12 weeks were used in each experiment. All animal studies were approved by the National Heart, Lung, and Blood Institute’s Animal Care and Use Committee.

AA mouse models AA mice were generated through immune-mediated BM failure as previously described.16,17 Briefly, recipient CB10 mice were exposed to 5 Gy of total body irradiation (TBI). Lymph node (LN) cells were obtained from B6 donors, and infused into recipients by tail vein injection at 5.0 x 106 cells per mouse. Wild-type FVB and A-ZIP/F1 “fatless” AA mice were created as was the CB10 AA model, except that recipient FVB or A-ZIP/F “fatless” mice were irradiated at 6.8 Gy and injected with LN cells from MHC H2matched donor DBA/1 mice at 10 x 106 per mouse.

Treatment of mice PPARγ antagonists, BADGE or GW9662 (Sigma, St. Louis, MO, USA), were dissolved in dimethyl sulfoxide (DMSO) and stored at -20°C. The aliquots were diluted with PBS to a final concentration of 10% DMSO and administrated by daily intraperitoneal injec-

tion at 30 mg/kg for BADGE, or at 1 mg/kg for GW9662, from one day prior to the experiment and continued for up to 2 weeks. In the FVB AA model, some mice were injected with cyclosporine A (CsA, 50 mg/kg/day, Sigma) starting 1 hour after the LN injection, and continued for 5 days as immunosuppression. At the end of the experiments, the mice were euthanized by CO2 inhalation. Methods, including peripheral blood (PB) and BM cell counting, flow cytometry, RNA isolation and gene expression analysis by PCR array, protein extraction and immunoblotting, cytokine measurement, histology using confocal microscopy, calcium flux assay and cell culture are detailed in the Online Supplementary Methods.

Statistics To compare gene expression in adipogenesis and inflammasome pathways between AA and BADGE-treated AA mice, twoway hierarchical cluster analysis was performed for genes that had more than 2-fold changes, based on PCR array data using Ward’s method (JMP version 10.0.0, SAS Institute, Cary, NC, USA). Differences of CBCs, total BM cell number, cytokines, and flow cytometry data between different groups were analyzed by student's t-test, Mann-Whitney test, or one-way ANOVA test using Prism software (GraphPad Software, La Jolla, CA, USA). Significant difference was set at P < 0.05 for all the statistical tests. Data were expressed as mean ± SEM.

Figure 3. T cell activation in BADGEtreated AA mice. Altered CD4+ and CD8+ T cells in BM (A) and PB (B) in BADGE-treated AA mice. Representative data at 2 weeks are shown. Data were gated on all nucleated cells. Results of percentage and absolute numbers are summarized beside the flow cytometry graphs. AA (n=15), BADGE-treated (n=16). (C) Comparison of CD150 expression on BM T cells in BADGEtreated and control AA mice. The histograms were gated on viable CD3+ T cells. Representative data of at least 3 separate experiments are shown. (D) The effect of delayed BADGE treatment on PB cell count and BM hematopoiesis. AA (n=6), BADGE-treated (n=7). *P<0.05; **P<0.01.

haematologica | 2016; 101(1)

PPARγ antagonist in bone marrow failure

Results

matched at MHC H2 antigens but differed in multiple minor histocompatibility antigens (miHAs). In this adaptation of “runt” disease, all mice uniformly develop progressive and fatal pancytopenia, accumulating a large number of adipocytes in the BM - closely resembling human AA, and without evidence of graft-versus-host disease.16,17 BM destruction in this model is mediated by miHA H60-reactive T cells.17 We treated recipient mice with PPARγ antagonists BADGE or GW9662, or control vehicle. On day 14, mice were sacrificed and evaluated in PB by cell counts

PPARγ antagonists ameliorated pancytopenia and BM destruction in AA mice Naveiras et al.’s study suggested that adipocytes were negative regulators of hematopoiesis.7 We speculated that PPARγ antagonists could ameliorate the immune-mediated marrow failure model by inhibiting adipogenesis. We induced BM failure by the injection of B6 LN cells into sublethally irradiated CB10 recipients, which were

Figure 4. T cell activation and proliferation are inhibited by BADGE. Dose response of BADGE on CD25 (A) and CD150 (B) expression of CD4+ and CD8+ T cells. Mouse lymph node T cells were stimulated with PMA (5 ng/ml) and ionomycin (500 nM) in the presence of BADGE (50 or 100 mM) or control vehicle DMSO for 16 hours. (C) Dose response of BADGE on proliferation of CD3+ T cells. Mouse T cells were labeled with CFSE and stimulated with PMA (5 ng/ml) and ionomycin (500 nM), and then incubated with BADGE (50 or 100 mM) or control vehicle DMSO for 3 days. Statistic analysis of flow cytometry data is shown on the right. All the data were gated on viable cells (7AAD-). (D) T cell related cytokines in the supernatants of PMA-stimulated T cell culture in the presence of BADGE (50 or 100 mM) or control vehicle DMSO for 16 hours (n=5 for each group). (E) Calcium influx is altered by BADGE. Mouse LN cells were cultured with or without BADGE (50 or 100 mM) for 4 hours, and were then loaded with Fluo-4 at 37°C for 30 minutes. The cells were added to biotinylated anti-CD3 antibodies, and crosslinked with streptavidin at 20 sec of the analysis (arrow) to initiate calcium flux. (F) Protein levels of PPARγ, PLCγ1, and NFAT1 in PMA-stimulated T cells in the presence or absence of BADGE for 16 hours. β-actin was used as a loading control. Representative results (A, B, C, E) or images (F) of at least three experiments are shown.

haematologica | 2016; 101(1)

K. Sato et al.

and in BM by estimating cellularity and morphologic examination of marrow adipocytes. Confocal microscopy showed massive expansion of adipocytes in the BM of AA mice. Overall BM cellularity, estimated by the density of 4',6-diamidino-2-phenylindole (DAPI) staining (nuclear, blue color), was very low. In contrast, mice treated with PPARγ antagonists had many fewer adipocytes (green color, boron-dipyrromethene dye -BODIPY) in the BM and much higher marrow cellularity (Figure 1A, left). By conventional staining, the BM structure of AA mice showed extensive disruption and replacement by adipocytes, which occupied the “empty” space,

whereas the BM of BADGE- or GW9662-treated mice showed less empty space and more hematopoietic cellularity (Figure 1A, right). Blood leukocyte and platelet counts were higher in BADGE-treated mice compared with AA mice (Figure 1B). BM nucleated cell counts in both BADGE- and GW9662-treated mice were significantly higher than counts in the AA group. Using flow cytometry, we found the frequency of Lin- cells and the absolute number of Lin-Sca1+c-kit+ (LSK) stem cells in BM were higher in BADGE-treated mice than in controls (Figure 1C). To test for a dose effect of BADGE in the AA mouse

Figure 5. In vivo effect of BADGE on T cells in FVB BM failure model. (A) CBCs and BM cell numbers of FVB BM failure mice. (B) H&E staining of liver, spleen, gut, and skin in BM failure and control mice. Original magnification, x100. (C) Frequencies and absolute numbers of CD8+ and CD4+ T cells in the BM of LN-infused FVB mice. (D) Impact of BADGE and CsA on plasma cytokine levels in FVB BM failure model. TBI control, LN-infused, CsA-treated, and BADGE-treated mice (n=10 for each group) at 2 weeks. Error bars indicate SEM. *P<0.05; **P<0.01; ***P<0.001. (E) H&E staining and confocal microscopic imaging show BM structure and adipocytes in the sternums of FVB TBI control, LN-infused, CsA-treated, and BADGE-treated mice at 2 weeks. Representative pictures of each group of at least three separate experiments are shown. Original magnification, x 200 (H&E), x 100 (confocal).

haematologica | 2016; 101(1)

PPARγ antagonist in bone marrow failure

model, we compared BADGE at low dose (15 mg/kg), standard dose (30 mg/kg), and high dose (60 mg/kg). Low dose BADGE had no effect, and high dose BADGE appeared to produce no added activity for the promotion of hematopoiesis and inhibition of BM adipogenesis, as compared to standard dose BADGE (Figure 1D,E), suggesting that an optimal dose of PPARγ antagonists is needed to inhibit BM adipogenesis. To visualize hematopoietic stem and progenitor cells (HSPCs, Figure 1F) and megakaryocytes (Figure 1G) in the BM, we stained fixed sternums with anti-CD117 and antiCD41 antibodies, respectively, followed by confocal imaging. In TBI alone mice, HSPC, megakaryocytes, and a few adipocytes were present in BM; in AA mice, the numbers of HSPCs and megakaryocytes were markedly reduced and adipocytes were abundant. In BADGE-treated mice, HSPCs and megakaryocytes were well preserved but with many fewer adipocytes, consistent with platelet counts on CBC and LSK cells in flow cytometry data. In both TBI only and BADGE-treated mice, HSPCs were distributed along the edge of the bones (Figure 1F).

Altered immunity of PPARγ antagonist-treated AA mice

In our AA mice model, T cell immunity is the cause of BM destruction. In order to determine if treatment with PPARγ antagonists altered the immunological status of AA mice, we measured plasma cytokine levels in a multiplex assay. Insulin, an adipogenesis-related hormone, was higher in AA mice than in TBI controls; BADGE- or GW9662treatment corrected insulin levels to normal. AA mice showed higher levels of the inflammatory cytokine monocyte chemoattractant protein-1 (MCP-1) and Th1 cytokines such as IFNγ, IFNγ-induced protein 10 (IP-10), and TNFα than did TBI control mice. BADGE- or GW9662-treatment reduced plasma MCP-1 as compared to levels in AA mice; BADGE reduced IFNγ and IP-10 levels, and GW9662-treatment reduced TNFα levels as well. BADGE- or GW9662treatment tended to decrease IL-6 levels (Figure 2A). We extracted RNA from whole femurs and performed PCR-based arrays, focusing on adipogenesis (Figure 2B) and inflammasome (Figure 2C) pathways (gene expression levels are presented as heat maps). Compared with AA mice, expression levels of many adipogenesis genes decreased by

BM# (x106)

WBC# (x103/uL)

RBC# (x106/uL)

PLT# (x103/uL)

C E

Figure 6. Marrow and body fat in “fatless” mice with bone marrow failure. (A) CBCs and BM cell numbers of A-ZIP/F “fatless” and wild-type (WT) mice with bone marrow failure (BMF) induced by injection of DBA lymph node (LN) cells post total body irradiation (TBI, 6.8 Gy). N=5 for each group. Error bars indicate SEM. *P<0.05; **P<0.01; ***P<0.001. (B) Confocal images and H&E staining of BM fat in immune-mediated BMF with “fatless” and wild-type mice. (C) Lack of body fat in “fatless” mice compared with wild-type mice. (D) Genotyping of “fatless” and wild-type mice. PCR results show upper band (332 bp, A-ZIP/F transgene) and lower band (200 bp, internal positive control). (E) Confocal images and H&E staining of BM fat in irradiation (7.5 Gy)-mediated BMF with “fatless” and wild-type mice. All data shown are at 2 weeks post BMF. Original magnification, x 200 (H&E), x 100 (confocal).

haematologica | 2016; 101(1)

K. Sato et al.

3- to 20-fold in BADGE-treated mice. The affected genes were known PPARγ targets, pro-adipogenesis, and adipokine-related genes. For example, expression levels of adipose angiotensinogen (Agt), Nrob2, Sirt1, Ncoa2, Adig, Bmp2, and Acacb were more than 5-fold lower in BADGEtreated mice. Expression of cell cycle- and proliferationrelated genes E2f1 and Mapk14 increased 6- and 3-fold, respectively, perhaps reflecting active hematopoietic cell repopulation in the BM of BADGE-treated mice (Figure 2B). Inflammasome genes include four family members (Aim2, Ipaf, Nlrp1, and Nlrp3); BADGE only affected Ipaf, as expression levels of its components Naip1, Naip5, and Nlrc4, were reduced more than 3-fold, and no changes were observed in the genes of other inflammasome family members. Expression of inflammation-related genes including Il6, Tnfsf4, and Tnf was also markedly decreased in the BADGE-treated group compared with control AA mice, while expression of the anti-inflammation related gene Nlrp12 (10-fold) was elevated (Figure 2C). Decreased Il6 and Tnf expression at mRNA levels in treated mice was concordant with plasma protein levels (Figure 2A). Gene expression levels, as determined by PCR array, were validated by immunoblot in order to confirm protein levels of PPARγ and AGT in BM. PPARγ isoform 2, an adipocyte-specific master regulator, was highly expressed in the BM of AA mice; both PPARγ isoforms 1 and 2 were greatly reduced in BADGE-treated mice, confirming that BADGE inhibited PPARγ expression in the model. AGT, one of the adipogenesis regulatory hormones and a PPARγ target protein, was not visible in TBI control CB10 mice, but was present at high levels in AA mice. BADGE treatment reduced AGT protein levels (Figure 2D), consistent with PCR array data (Figure 2B). Furthermore, immunoblotting results demonstrated that T cells isolated from the BM of BADGE-treated mice had decreased PPARγ protein levels compared to that from control AA mice (Figure 2E). In order to determine whether PPARγ antagonist affected T cell populations, and especially T cells infiltrating the marrow of AA mice, we performed flow cytometry of PB and nucleated cells manually flushed from the BM. In AA mice, there was massive expansion of CD8+ and CD4+ T cells in the BM as expected; in contrast, BADGE reduced the frequencies of both CD8+ and CD4+ T cells significantly, while the absolute numbers of CD8+ and CD4+ T cells were not reduced (Figure 3A). In the PB of AA mice, CD8+ T cells were markedly expanded, and BADGE largely reduced CD8+ T cell frequencies, with less effect on CD4+ T cells (Figure 3B). We measured expression of CD150, an activation marker for T cells, on residual BM CD3+ T cells by flow cytometry: BADGE suppressed CD150 expression compared with control AA mice (Figure 3C), suggesting inhibition of T-cell activation by BADGE. In these experiments, PPARγ was administered beginning one day prior to LN cell infusion. To exclude the possibility that the drug might affect donor T cells homing to recipient BM, we delayed BADGE administration to one day after LN cell infusion. Amelioration of BM failure (as well as blockade of adipogenesis) was still observed (Figure 3D), suggesting that the effect of BADGE on T cells might be the inhibition of T cells directly, rather than altering T cell homing.

PPARγ antagonist inhibited T cell activation and proliferation in vitro

In our AA mice, BM adipogenesis was a consequence of marrow destruction mediated by T cells. In vivo data pro64

vided evidence of T cell inhibition by PPARγ antagonist. To test whether immunosuppression was a direct effect of the antagonist, we assessed the action of BADGE on T cell function in vitro. We stimulated mouse LN T cells with phorbol myristate acetate (PMA) and ionomycin and then incubated cells with different concentrations of BADGE for 16 hours, followed by flow cytometry analysis. CD25, an activation marker on both CD4+ and CD8+ T cells, was greatly decreased in the presence of BADGE, in a dose dependent manner (Figure 4A). Similarly, expression of T cell activation marker CD150 was also inhibited by BADGE (Figure 4B), parallel to the results of in vivo experiments (Figure 3C). Higher concentrations of BADGE (> 100 mM) were cytotoxic, as determined by flow cytometry with 7-amino-actinomycin D viability staining, as previously reported.14 PPARγ antagonists not only inhibited T cell activation but also suppressed T cell proliferation (Figure 4C): PMA plus ionomycin promoted T cell proliferation, reflected by the multiple divisions seen using carboxyfluorescein succinimidyl ester (CFSE) signal, while BADGE prevented T cells from dividing. To confirm the early effect of BADGE on T cell activation, we collected culture supernatants and measured cytokines. BADGE decreased levels of PMA-induced secretion of soluble CD137, IL-2, and IL-6 by T cells (Figure 4D). Cytoplasmic calcium changes in T cells accompany activation by multiple signaling pathways. To investigate if the inhibition of T cell activation by BADGE is related to this process, we performed calcium flux assay. Stimulation with CD3 antibody crosslinked with streptavidin induced marked calcium flux in mouse LN cells, but pre-incubation with BADGE was inhibitory of cation shift in a dosedependent fashion (Figure 4E). The effects of BADGE on T cell activation, proliferation, and cytokine secretion were related to PPARγ expression levels. As shown in immunoblotting, BADGE treatment decreased PPARγ in T cells in a dose-dependent fashion (Figure 4F). Similarly, BADGE also suppressed phospholipase C gamma 1 (PLCγ1) and the nuclear factor of activated T-cells 1 (NFAT1) expression in activated T cells (Figure 4F).

In vivo effect of BADGE on T cells in FVB BM failure model To test if the effect of BADGE could be replicated in another mouse model (MHC-matched but non-MHC mismatched), we created a second immune-mediated BM failure model using FVB strain mice. As described in Naveiras et al.,7 transplantation of BM cells from donor DBA/1 mice rescued lethally irradiated FVB mice. Because the donor and recipient mice are MHC-matched (H2q) but non-MHC mismatched, these strains have been reported to differ in patterns of expressed genes, among which some are quantitative trait loci related to immune diseases.18,19 We tested whether the injection of LN cells (mainly T cells) from donor DBA mice could induce T cell-mediated BM destruction of FVB mice, as in the CB10 AA model.17 We infused 10 x 106 DBA LN cells into sublethally irradiated (6.8 Gy) FVB mice. At 2 weeks, mice that had received LN cells developed pancytopenia and BMs were hypocelluar (Figure 5A) without apparent graft-versus-host disease (Figure 5B). When mice were treated with the immunosuppressive drug CsA, BM failure was prevented; when FVB haematologica | 2016; 101(1)

PPARγ antagonist in bone marrow failure

BM failure mice were treated with BADGE, marrow cellularity improved (Figure 5A). On flow cytometry, there was an expansion of T cells, especially CD8+ T cells, in the BM of control LN-infused mice (Figure 5C). CsA or BADGE significantly decreased the frequency of infiltrating T cells in the BM, although the absolute numbers of CD8+ and CD4+ T cells in BM remained elevated (Figure 5C). Plasma levels of insulin, MCP-1, TNFα, IL-6, and IFNγ were increased when BM failure was induced, suggesting an inflammatory microenvironment. Both BADGE and CsA decreased levels of IFNγ significantly, while BADGE also corrected insulin and IL-6 levels and tended to decrease other inflammatory cytokines such as TNFα and MCP-1 (Figure 5D). Low dose irradiation was associated with the appearance of only a few adipocytes in the BM of TBI control mice (Figure 5E). We anticipated adipocytes to be increased in the BM of LN-injected recipient mice due to BM destruction, but surprisingly this did not occur (Figure 5E). In contrast, CsA-treated mice had much higher BM cellularity and also many more adipocytes in their BM compared with marrow from TBI control mice and untreated BM failure mice, while almost no change in BM adipocytes was observed in BADGE-treated mice (Figure 5E). Lack of correlation between marrow fat cell content and BM failure indicated that BADGE might act by inhibiting T cell function rather than by suppressing BM adipogenesis in this model.

BM and body adipogenesis in “fatless”mice Ideally, mice genetically incapable of forming adipocytes would be useful in our BM failure model, in order to separate the mechanisms of anti-adipogenesis and immunosuppression by PPARγ antagonists. We attempted such experiments with inbred lipoatrophic A-ZIP/F1 “fatless” transgenic mice.20 The injection of DBA/1 LN cells into A-ZIP/F “fatless” mice indeed induced BM failure at 2 weeks, producing the expected hypocellularity of BM and decreased peripheral blood counts (Figure 6A). To our surprise we observed adipocytes in the marrow (Figure 6B) with the induction of BM failure, although we confirmed the absence of body fat in these mice (Figure 6C) and the genotype as previously described20 (Figure 6D). BM fat was similarly present in non-immunemediated marrow failure induced by irradiation (Figure 6E). The different distribution of body fat and marrow fat in both immune- and irradiation-mediated BM failure models suggests different mechanisms between BM and systemic adipogenesis under physiologic and stressed conditions.

Discussion We observed that PPARγ antagonists ameliorated the severity of BM failure in an immune-mediated AA mouse model, and we hypothesized that PPARγ antagonists might modulate T cell function in addition to suppressing BM adipogenesis under these circumstances. In support of our in vivo data, in vitro experiments showed that T cell activation and proliferation, as well as cytokines, were significantly diminished in the presence of PPARγ antagonists. Treatment with BADGE in a FVB BM failure model, in which very few adipocytes accumulated in the BM, further confirmed the effect of PPARγ antagonists on T cell haematologica | 2016; 101(1)

functions. BM failure is a disorder of hematopoiesis. Cells of the perivascular niche (vascular endothelial cells) and endosteal niche (osteoblasts and their progenitors) are known to play active roles in the regulation of hematopoiesis.21-23 Osteoblasts, an important component of the endosteal niche, originate from mesenchymal stromal cells (MSCs), the common pluripotent precursor shared with adipocytes. Differentiation of MSCs to osteoblasts or to adipocytes is regulated by intracellular and extracellular cytokines and transcription factors, including PPARγ.22,23 An inverse reciprocal relationship exists between adipogenesis and osteoblastosis. Most growth factors and cytokines that promote osteoblastosis negatively regulate adipogenesis, and vice versa. BM failure is often accompanied by active adipogenesis. A clinical study has shown that an increased volume of BM adipocytes is strongly correlated with reduced bone formation.24 In recently published work, the numbers of HSPCs were reduced by up to 3-fold in adipocyte-rich tail vertebrae, as compared to adipocyte-poor thoracic vertebrae.7 Lipoatrophic A-ZIP/F1 “fatless” transgenic mice, which are genetically incapable of forming adipocytes, showed markedly stimulated osteogenesis that resulted in bone formation after lethal irradiation.7 Thus, it was attractive to postulate that BM adipocytes might have a negative role in the regulation of hematopoiesis, in contrast with osteoblasts’ supportive function in hematopoiesis. Adipogenesis involves a complex network of transcription factors, PPARγ and C/EBPs are the master regulators of these processes. In particular, PPARγ signaling is understood to be both necessary and sufficient for fat cell formation.25 In this current work, we show that expression of PPARγ (particularly the PPARγ2 isoform) is increased in our CB10 AA mouse model with the accumulation of fat cells in BM. Thus, the inhibition of PPARγ and suppression of adipogenesis are implicated functionally. However, there are controversies concerning the effects of PPARγ antagonists on adipogenesis. Some studies have concluded that BADGE blocks adipogenesis from a mouse preadipocyte cell line and from normal isolated MSCs (IC50 of approximately 100 mM).26 Conversely, in other experiments, BADGE and GW9662 did not antagonize PPARγ signaling nor prevent adipogenesis in vitro and in vivo.27-30 In one study, BADGE actually promoted adipogenesis at nanomolar levels, while higher concentrations of BADGE were cytotoxic.28 We found that low dose BADGE had either no therapeutic effect or even worsened BM failure, whereas high dose BADGE did not add to hematological improvement. Both PPARγ agonists and antagonists have been reported to be cytotoxic at high doses.28,31 For example, thiazolidinediones, which are clinically used as specific high affinity ligands for PPARγ, have shown a series of unexpectedly severe adverse effects, including fatal hepatitis, heart failure due to water retention, increased risk of cardiovascular events, and a higher incidence of bladder cancer.31 These cytotoxic effects may restrict their utility as well as hinder evaluation of molecular mechanisms. Overall, the effects of PPARγ antagonists vary greatly depending on drug concentrations, as well as on biological or culture conditions. Our observation suggests that fresh preparation, shortterm (< 2 weeks) preservation, and timely administration are critical for efficacy of BADGE or GW9662, probably due to their short active chemical half-life. 65

K. Sato et al.

BM adipogenesis often has an inverse correlation with hematopoiesis, but this is not a consistent relationship. For example, adipose-derived MSCs efficiently support hematopoiesis in vivo and in vitro.32,33 Herein, we demonstrate that the injection of DBA/1 LN cells into FVB mice (different at non-MHC) induced severe BM destruction without the accumulation of large numbers of adipocytes in the BM, different from other BM failure models; possibly, more severe BM damage was the result of a higher irradiation dose, a larger LN cell inoculum, and/or a stronger bystander effect of injected T cells, which may have led to a more efficient elimination of HSC and stromal cells, including adipocyte precursors. CsA suppressed T cells effectively, and allowed for the recovery of HSC and stromal cells including adipocyte precursors, leading to the coexistence of improved BM cellularity and the expansion of BM adipocytes. Although BADGE also suppressed T cells, the efficacy was not as potent as CsA, as evidenced by residual T cells in BM, documented by flow cytometry. We attempted to separate the mechanisms of anti-adipogenesis and immunosuppression by PPARγ antagonists by inducing BM failure with inbred lipoatrophic A-ZIP/F1 “fatless” transgenic mice.20 However, we noted that for both immune- and irradiation-mediated BM failure, adipocytes still accumulated in BM despite the phenotype of missing body fat and confirmed genotype, suggesting that subcutaneous and BM adipogenesis are critically different. The underlying molecular biology in these circumstances is under further investigation. Although an immunomodulatory role of PPARγ signaling has been inferred from experiments with agonists or antagonists in vitro, applications to animal models have been limited.7,9-12,15,34 Most rodent studies have explored extramedullary adipocytes, focusing on obesity or metabolic syndrome rather than immune models.15,34 PPARγ plays pleiotropic roles in inflammation, and the inflammatory reaction may be stimulated or suppressed by the presence of PPAR ligands35 and PPARγ antagonists. Competitive radioligand binding studies showed BADGE to be a ligand for PPARγ with micromolar affinity, and functionally

References 1. Scheinberg P, Young NS. How I treat acquired aplastic anemia. Blood. 2012;120(6):1185-1196. 2. Young NS, Calado RT, Scheinberg P. Current concepts in the pathophysiology and treatment of aplastic anemia. Blood. 2006;108(8): 2509-2519. 3. Takaku T, Malide D, Chen J, Calado RT, Kajigaya S, Young NS. Hematopoiesis in 3 dimensions: human and murine bone marrow architecture visualized by confocal microscopy. Blood. 2010;116(15):e41-55. 4. Gimble JM, Robinson CE, Wu X, Kelly KA. The function of adipocytes in the bone marrow stroma: an update. Bone. 1996;19(5): 421-428. 5. Krings A, Rahman S, Huang S, Lu Y, Czernik PJ, Lecka-Czernik B. Bone marrow fat has brown adipose tissue characteristics, which are attenuated with aging and diabetes. Bone. 2012;50(2):546-552. 6. Liu LF, Shen WJ, Ueno M, Patel S, Kraemer FB. Characterization of age-related gene

10. 11.

BADGE is a pure antagonist for this receptor.26 Therefore, the effects of BADGE on adipogenesis or immunomodulation are complicated. The study by Dworzanski et al. used very high concentrations of BADGE (120 mg/kg),35 at which the toxic and inflammatory effects might be dominated. In our BM failure models, BADGE at 30-50 mg/kg seems optimal to suppress immune reaction. Our initial intention in these experiments was to clarify the role of BM adipocytes in AA by inhibiting PPARγ signaling. However, we unexpectedly found that, while alloreactive T cell activation in response to multiple miHAs appeared to be ameliorated by PPARγ antagonism, the inhibition of adipocyte formation was not a generalizable activity when related to marrow failure. Previous studies showed that PPARγ agonists inhibited T cell proliferative response and IL-2 secretion by activated T cells.9,10,12,13 However, our results indicated that PPARγ antagonists also had similar effects on T cell activation and IL-2 secretion; the antagonists directly inhibited calcium flux and suppressed expression of PPARγ and NFAT1. In addition, PPARγ antagonists were reported to have potent antiproliferative effects on hematopoietic and cancer cell lines by induction of apoptosis, in caspase-dependent or –independent mechanisms.8,11,14 It is noteworthy that both agonists and antagonists induced apoptosis in these cells, effects not necessarily related to PPARγ expression levels.11 Based on our results, we conclude that PPARγ antagonists act as negative regulators of T cells in addition to their inhibition of BM adipogenesis. Our findings hold implications for the application of PPARγ antagonists in immune-mediated pathophysiologies both in the laboratory and in the clinic. Funding This work was supported by NIH Intramural Research Program. Acknowledgments The authors would like to thank Dr. Haiming Cao (NHLBI) for careful reading of our manuscript and helpful suggestions, and Eric Nimako (NCI) for technique support for A-ZIP/F “fatless” mice.

expression profiling in bone marrow and epididymal adipocytes. BMC Genomics. 2011;12:212. Naveiras O, Nardi V, Wenzel PL, Hauschka PV, Fahey F, Daley GQ. Bone-marrow adipocytes as negative regulators of the haematopoietic microenvironment. Nature. 2009;460(7252):259-263. Harris SG, Phipps RP. The nuclear receptor PPAR gamma is expressed by mouse T lymphocytes and PPAR gamma agonists induce apoptosis. Eur J Immunol. 2001;31 (4):1098-1105. Clark RB, Bishop-Bailey D, EstradaHernandez T, Hla T, Puddington L, Padula SJ. The nuclear receptor PPAR gamma and immunoregulation: PPAR gamma mediates inhibition of helper T cell responses. J Immunol. 2000;164(3):1364-1371. Clark RB. The role of PPARs in inflammation and immunity. J Leukoc Biol. 2002;71 (3):388-400. Burton JD, Castillo ME, Goldenberg DM, Blumenthal RD. Peroxisome proliferatoractivated receptor-gamma antagonists

12.

13.

14.

15.

exhibit potent antiproliferative effects versus many hematopoietic and epithelial cancer cell lines. Anticancer Drugs. 2007;18 (5):525-534. Szeles L, Torocsik D, Nagy L. PPARgamma in immunity and inflammation: cell types and diseases. Biochim Biophys Acta. 2007;1771(8):1014-1030. Yang XY, Wang LH, Chen T, et al. Activation of human T lymphocytes is inhibited by peroxisome proliferator-activated receptor gamma (PPARgamma) agonists. PPARgamma co-association with transcription factor NFAT. J Biol Chem. 2000;275 (7):4541-4544. Fehlberg S, Trautwein S, Goke A, Goke R. Bisphenol A diglycidyl ether induces apoptosis in tumour cells independently of peroxisome proliferator-activated receptorgamma, in caspase-dependent and -independent manners. Biochem J. 2002;362(Pt 3):573-578. Botolin S, McCabe LR. Inhibition of PPARgamma prevents type I diabetic bone marrow adiposity but not bone loss. J Cell

haematologica | 2016; 101(1)

PPARÎł antagonist in bone marrow failure

Physiol. 2006;209(3):967-976. 16. Bloom ML, Wolk AG, Simon-Stoos KL, Bard JS, Chen J, Young NS. A mouse model of lymphocyte infusion-induced bone marrow failure. Exp Hematol. 2004;32(12):1163-1172. 17. Chen J, Ellison FM, Eckhaus MA, et al. Minor antigen h60-mediated aplastic anemia is ameliorated by immunosuppression and the infusion of regulatory T cells. J Immunol. 2007;178(7):4159-4168. 18. Bauer K, Yu X, Wernhoff P, Koczan D, Thiesen HJ, Ibrahim SM. Identification of new quantitative trait loci in mice with collagen-induced arthritis. Arthritis and rheumatism. 2004;50(11):3721-3728. 19. Yu X, Bauer K, Wernhoff P, et al. Fine mapping of collagen-induced arthritis quantitative trait loci in an advanced intercross line. Journal of immunology. 2006;177(10): 7042-7049. 20. Moitra J, Mason MM, Olive M, et al. Life without white fat: a transgenic mouse. Genes Dev. 1998;12(20):3168-3181. 21. Ema H, Suda T. Two anatomically distinct niches regulate stem cell activity. Blood. 2012;120(11):2174-2181. 22. Takada I, Kouzmenko AP, Kato S. Molecular switching of osteoblastogenesis versus adipogenesis: implications for targeted therapies. Expert Opin Ther Targets. 2009;13(5): 593-603. 23. Takada I, Kouzmenko AP, Kato S. Wnt and

haematologica | 2016; 101(1)

24.

25. 26.

27.

28.

29.

PPARgamma signaling in osteoblastogenesis and adipogenesis. Nat Rev Rheumatol. 2009;5(8):442-447. Wren TA, Chung SA, Dorey FJ, Bluml S, Adams GB, Gilsanz V. Bone marrow fat is inversely related to cortical bone in young and old subjects. J Clin Endocrinol Metab. 2011;96(3):782-786. Lefterova MI, Lazar MA. New developments in adipogenesis. Trends Endocrinol Metab. 2009;20(3):107-114. Wright HM, Clish CB, Mikami T, et al. A synthetic antagonist for the peroxisome proliferator-activated receptor gamma inhibits adipocyte differentiation. J Biol Chem. 2000;275(3):1873-1877. Bishop-Bailey D, Hla T, Warner TD. Bisphenol A diglycidyl ether (BADGE) is a PPARgamma agonist in an ECV304 cell line. Br J Pharmacol. 2000;131(4):651-654. Chamorro-Garcia R, Kirchner S, Li X, et al. Bisphenol A diglycidyl ether induces adipogenic differentiation of multipotent stromal stem cells through a peroxisome proliferator-activated receptor gamma-independent mechanism. Environ Health Perspect. 2012;120(7):984-989. Hung SH, Yeh CH, Huang HT, et al. Pioglitazone and dexamethasone induce adipogenesis in D1 bone marrow stromal cell line, but not through the peroxisome proliferator-activated receptor-gamma pathway. Life Sci. 2008;82(11-12):561-569.

30. Zhu M, Flynt L, Ghosh S, et al. Antiinflammatory effects of thiazolidinediones in human airway smooth muscle cells. Am J Respir Cell Mol Biol. 2011;45(1):111-119. 31. Riddle MC. Therapy: What evidence should guide the use of thiazolidinediones? Nat Rev Endocrinol. 2010;6(11): 600-602. 32. Nakao N, Nakayama T, Yahata T, et al. Adipose tissue-derived mesenchymal stem cells facilitate hematopoiesis in vitro and in vivo: advantages over bone marrowderived mesenchymal stem cells. Am J Pathol;177(2): 547-554. 33. Nishiwaki S, Nakayama T, Saito S, et al. Efficacy and safety of human adipose tissue-derived mesenchymal stem cells for supporting hematopoiesis. Int J Hematol; 96(3):295-300. 34. Cuzzocrea S, Pisano B, Dugo L, et al. Rosiglitazone, a ligand of the peroxisome proliferator-activated receptor-gamma, reduces acute pancreatitis induced by cerulein. Intensive Care Med. 2004;30(5): 951-956. 35. Dworzanski T, Celinski K, Korolczuk A, et al. Influence of the peroxisome proliferator-activated receptor gamma (PPARgamma) agonist, rosiglitazone and antagonist, biphenol-A-diglicydyl ether (BADGE) on the course of inflammation in the experimental model of colitis in rats. J Physiol Pharmacol. 2010;61(6):683-693.

ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Acute Lymphoblastic Leukemia

Ferrata Storti Foundation

Haematologica 2016 Volume 101(1):68-76

Relapsed childhood acute lymphoblastic leukemia in the Nordic countries: prognostic factors, treatment and outcome

Trausti Oskarsson,1,2 Stefan Söderhäll,1,2 Johan Arvidson,3 Erik Forestier,4 Scott Montgomery,5-7 Matteo Bottai,8 Birgitte Lausen,9 Niels Carlsen,10 Marit Hellebostad,11 Päivi Lähteenmäki,12 Ulla M. Saarinen-Pihkala,13 Ólafur G.Jónsson,14 and Mats Heyman1,2 on behalf of the Nordic Society of Paediatric Haematology and Oncology (NOPHO) ALL relapse working group

Department of Pediatric Oncology, Astrid Lindgren Children´s Hospital, Stockholm, Sweden; 2Childhood Cancer Research Unit, Department of Women´s and Children´s Health, Karolinska Institutet, Stockholm, Sweden; 3Department of Pediatric Oncology, Uppsala University Hospital, Sweden; 4Department of Pediatrics, Umeå University Hospital, Sweden; 5Clinical Epidemiology and Biostatistics, Faculty of Medicine and Health, Örebro University, Sweden; 6Clinical Epidemiology Unit, Karolinska University Hospital, Karolinska Institutet, Stockholm, Sweden; 7Department of Epidemiology and Public Health, University College London, UK; 8Unit of Biostatistics, IMM, Karolinska Institutet, Stockholm, Sweden; 9Department of Pediatric Oncology, Rigshospitalet University Hospital, Copenhagen, Denmark; 10Department of Pediatrics, Odense University Hospital, Denmark; 11Department of Pediatrics, Ullevål Hospital, Oslo, Norway; 12 Department of Pediatrics, Turku University Hospital, Turku, Finland; 13Children's Hospital, University of Helsinki and Helsinki University Central Hospital, Finland; and 14Children’s Hospital, Landspitali University Hospital, Reykjavik, Iceland 1

ABSTRACT

Correspondence: trausti.oskarsson@ki.se

Received: 4/6/2015. Accepted: 20/10/2015. Pre-published: 22/10/2015. doi:10.3324/haematol.2015.131680

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/1/68

elapse is the main reason for treatment failure in childhood acute lymphoblastic leukemia. Despite improvements in the up-front therapy, survival after relapse is still relatively poor, especially for high-risk relapses. The aims of this study were to assess outcomes following acute lymphoblastic leukemia relapse after common initial Nordic Society of Paediatric Haematology and Oncology protocol treatment; to validate currently used risk stratifications, and identify additional prognostic factors for overall survival. Altogether, 516 of 2735 patients (18.9%) relapsed between 1992 and 2011 and were included in the study. There were no statistically significant differences in outcome between the up-front protocols or between the relapse protocols used, but an improvement over time was observed. The 5-year overall survival for patients relapsing in the period 2002-2011 was 57.5±3.4%, but 44.7±3.2% (P<0.001) if relapse occurred in the period 1992-2001. Factors independently predicting mortality after relapse included short duration of first remission, bone marrow involvement, age ten years or over, unfavorable cytogenetics, and Down syndrome. T-cell immunophenotype was not an independent prognostic factor unless in combination with hyperleukocytosis at diagnosis. The outcome for early combined pre-B relapses was unexpectedly poor (5-year overall survival 38.0±10.6%), which supports the notion that these patients need further risk adjustment. Although survival outcomes have improved over time, the development of novel approaches is urgently needed to increase survival in relapsed childhood acute lymphoblastic leukemia.

Introduction With advances in chemotherapy, hematopoietic stem cell transplantation (HSCT) and supportive care, long-term survival in childhood acute lymphoblastic leukemia (ALL) is now 85-90%.1,2 Despite increasing concerns regarding treatment-related haematologica | 2016; 101(1)

Relapsed childhood ALL in the Nordic countries

mortality and second malignancies, the main reason for treatment failure is still relapse.3 In the Nordic countries, the relapse rate was close to 40% between 1981 and 1993 and only 30% remained in long-term second remission.4 Over the last two decades, the reported relapse rates have been 15-20%1-3,5,6 in the developed countries and the overall survival after relapse approximately 40-70%, depending on the follow-up time and the risk groups involved.7-12 Since 1992, all children aged one year and over diagnosed with pre-B and T-cell ALL in the Nordic countries (Denmark, Finland, Iceland, Norway, Sweden) have been treated according to a common Nordic Society of Paediatric Haematology and Oncology (NOPHO) ALL protocol. Children with relapsed ALL, on the other hand, have been treated heterogeneously since there has not been a common NOPHO ALL relapse protocol. The most commonly used relapse protocols have been the high-risk (HR) arms of the NOPHO ALL-92 or ALL-2000 front-line protocols, the German Berlin Frankfurt Münster (BFM) ALL-REZ relapse protocols, and the Finnish Relapse in Acute Lymphoblastic Leukemia (RALLE) pilot protocol,13 which was used mainly in Finland between 2004 and 2010. After 2009, the British Children´s Cancer and Leukemia Group (CCLG) ALLR3 relapse protocol has also been used in the Nordic countries, but the International study for treatment of childhood Relapsed ALL (IntReALL) trial is expected to replace other relapse protocols in the near future. Risk stratification at relapse is based on the time from initial diagnosis to relapse, the anatomic site of relapse, and immunophenotype.10-14 In addition, the most recent relapse protocols have integrated therapy response with minimal residual disease (MRD) for further treatment adjustments.9,11,15,16 But unlike risk stratification at primary diagnosis, cytogenetic aberrations, including MLL rearrangements17,18 and hypodiploidy,19,20 currently do not directly modify relapse treatment intensity. Patients meeting high-risk criteria at relapse are recommended to undergo allogeneic HSCT in CR2, but the indication for HSCT in patients with lower risk is still under dispute, although in most centers patients with high MRD after conventional re-induction are candidates for allogeneic HSCT.8,21 Rigorous selection of patients for the most appropriate treatment intensity is important not only to minimize the

risk of subsequent relapses but also to minimize treatment-related toxicity and mortality.22-25 Relatively few studies of the long-term outcome after relapsed childhood ALL have been published. Our study cohort is population-based and includes over 500 patients treated according to common up-front protocols, making it, to our best knowledge, the largest of its kind. We hypothesize that thorough analysis of prognostic factors, validation of the current risk stratification and comparison of treatment modalities could be helpful in improving treatment for relapsed childhood ALL.

Methods Study population and data collection Information on all children aged 1.0-14.9 years at diagnosis (n=2735) treated according to the NOPHO ALL-92 (n=1644) and ALL-2000 (n=1091) protocols were extracted from the NOPHO ALL registry and patients that relapsed before January 1st 2012 were identified. In 516 (18.9%) patients, relapse occurred as a primary event, in 339 (20.6%) after ALL-92 treatment, and in 177 (16.2%) following ALL-2000 treatment. In total, 130 (4.8%) patients underwent allogeneic HSCT in CR1 of which 31 (23.8%) relapsed. Since patients that relapse after HSCT in CR1 differ substantially from patients treated with chemotherapy only, in baseline characteristics, treatment and outcome, they were excluded from all outcome analysis. Thus this report includes the 485 relapse patients who did not receive HCST in CR1. The database was frozen on the 31st of December 2013 and this dataset was used for outcome analysis. In 95 patients for whom the registration of relapse treatment, therapy response, outcome or follow-up status was incomplete, data were acquired from the treating hospitals to supplement the registration. Data concerning genetic aberrations were centrally reviewed by the NOPHO cytogenetic group. Cytogenetic findings were divided into four groups. Unfavorable: hypodiploidy (modal chromosomal number <45), MLL rearrangements, t(9;22) BCR/ABL and t(1;19); Favorable: high hyperdiploidy (modal chromosomal number >50) and t(12;21); Other: iAMP21, dic(9;20), unspecified chromosomal abnormalities; and Normal/missing: 46XX/XY or missing values. See Online Supplementary Appendix for the definition of relapse and second complete remission. This study was approved by the Ethical Review Board in Stockholm and was conducted in accordance with the Declaration of Helsinski.

Table 1. Risk stratification by immunophenotype, the time from diagnosis to relapse and the anatomic site of relapse.

5y-OS ± s.e.% or alive/total Very early Early Late

iEM

Pre-B combined

iBM

iEM

T-cell combined

iBM

n=9 6/9 n=44 76.0 ± 6.6% n=24 82.0 ± 8.3%

n=4 0/4 n=21 38.0 ± 10.6% n=43 77.4 ± 6.7%

n=50 22.0 ± 5.9% n=67 36.6 ± 6.0% n=155 60.3 ± 4.1%

n=18 27.8 ± 10.6% n=3 1/3 n=2 1/2

n=10 30.0 ± 14.5% n=3 0/3 n=1 1/1

n=12 8.3 ± 8.0% n=8 4/8 n=3 1/3

Standard-risk group (in bold) and high-risk group (in italics) according to the IntReALL risk classification. The boxes include the total number of patients and the overall survival for each subgroup. For subgroups involving less than 10 patients, survival is presented as the proportion of patients alive within the subgroup at the end of the follow-up period instead of 5-year overall survival (OS) (± standard error) (s.e.). Isolated extramedullary relapses (iEM): relapses not involving the bone marrow, such as the central nervous system (CNS), testis, lymph nodes, mediastinum and skin. Combined relapses: co-existent bone marrow and extramedullary involvement. Isolated bone marrow relapses (iBM): bone marrow relapses without any extramedullary involvement.Very early relapses: occurring <18 months from primary diagnosis. Early relapses: occurring ≥18 months from diagnosis and <6 months after completion of primary therapy. Late relapses: occurring ≥6 months after completion of primary therapy. Eight patients with unknown immunophenotype were excluded from the survival analysis; very early iBM = 1, early iBM = 2, early iEM = 1, late iBM = 2, late iEM = 2.

haematologica | 2016; 101(1)

T. Oskarsson et al.

Treatment A detailed description of the risk groups and treatment used in the NOPHO ALL-92 and ALL-2000 protocols, as well as a comparison of the long-term results of these up-front ALL treatments have been published by Schmiegelow et al.1 We categorized the relapse treatment into four groups: ALL-REZ BFM protocols (90, 95/96 and

2002), NOPHO ALL-92 and ALL-2000 HR arms used as relapse therapy, RALLE pilot and “other treatment”. The “other treatment” group included patients treated with combinations of protocols, the CCLG ALLR3 relapse protocol, Children´s Cancer Group (CCG) relapse protocols and non-protocol chemotherapy. None of the 5 patients with t(9;22) were treated with tyrosine kinase inhibitors.

Table 2. Cox’s proportional hazards regression analysis of risk factors for overall survival after ALL relapse.

Prognostic factors

Unadjusted model HR (95% CI)

Adjusted model 1 HR (95% CI)

Adjusted model 2 HR (95% CI)

3.67 (2.56 – 5.26)*** 2.33 (1.70 – 3.19)*** 2.87 (1.97 - 4.19)*** 2.22 (1.39 - 3.52)** 1.43 (0.97 - 2.11) 1.27 (0.69 - 1.79)

3.80 (2.64 – 5.48)*** 2.36 (1.73 – 3.24)*** 2.98 (2.04 – 4.35)*** 2.31 (1.45 – 3.68)*** -

314 97

3.84 (2.82 – 5.24)*** 1.85 (1.37 – 2.51)*** 1.91(1.33 – 2.73)*** 1.47 (0.93 – 2.31) 2.25 (1.61 – 3.13)*** 1.70 (1.24 – 2.34)** 0.96(0.74 - 1.24) 1.76 (1.32 - 2.34)***

1.07 (0.82 - 1.41) 1.73 (1.28 - 2.34)**

1.08 (0.82 – 1.42) 1.80 (1.33 – 2.44)***

28 173 123 17 27 33 45 201 154

2.55 (1.61 – 4.06)*** 0.80 (0.58 – 1.09) 1.18 (0.85 – 1.62) 2.07 (1.21 - 3.56)** 3.46 (2.23 – 5.38)*** 1.66 (1.05 – 2.64)* 1.22 (0.79 – 1.86) 1.91(1.38 - 2.64)*** 1.10 (0.77 - 1.57)

1.86 (1.14 – 3.03)* 1.05 (0.74 – 1.48) 1.08 (0.77 – 1.50) 2.70 (1.51 – 4.82)** -

1.83 (1.12 – 2.98)* 1.05 (0.75 – 1.49) 1.15 (0.82 – 1.60) 2.63 (1.48 – 4.70)** 2.38 (1.43 – 3.96)** 1.01 (0.60 – 1.69) 0.95 (0.61 – 1.49) -

N. Very early relapse1 Early relapse1 Isolated bone marrow relapse2 Combined relapse2 T-cell ALL3 WBC ≥ 100 x 109/L at primary diagnosis4 Male5 Age ≥ 10 years at primary diagnosis6 Unfavorable cytogenetics7,* Favorable cytogenetics7,* Other cytogenetics7,* Down syndrome8 T-cell + WBC ≥100 x 109/L9 T-cell + WBC <100 x 109/L9 Pre-B + WBC ≥100 x 109/L9 Initial risk group High Risk10 Initial risk group Intermediate Risk10

104 149 300 82 60 72

Hazard ratio (HR) for death. Reference groups: 1late relapse, 2isolated extramedullary relapse, 3pre-B ALL, 4white blood cell count (WBC) <100 x 109/L at primary diagnosis, 5female, 6 age <10 years at primary diagnosis, 7no detected or missing data on chromosomal abnormalities, 8not Down syndrome, 9pre-B + WBC <100 x 109/L at primary diagnosis, 10Standard Risk patients according to the NOPHO ALL-92 and ALL-2000 protocols. *Unfavorable cytogenetics: MLL rearrangements n=7, hypodiploidy (modal chromosomal number <45) n=10, t(9;22) n=5, t(1;19) n=6; Favorable cytogenetics: high hyperdiploidy (modal chromosomal number >50) n=106, t(12;21) n=67. ***P<0.001, **P<0.01, *P<0.05.

Table 3. Relapse treatment for patients initially treated according to the NOPHO ALL-92 and NOPHO ALL-2000 protocols.

Relapse protocol NOPHO HR arms Standard-risk High-risk ALL-REZ BFM Standard-risk High-risk RALLE*** Standard-risk High-risk Other treatment Standard-risk High-risk Total number

Total n. (%)

NOPHO ALL-92 (%)*

NOPHO ALL-2000 (%)*

CR2 rate (%)

HSCT in CR2** (%)

5-year OS ± s.e %

10-year OS ± s.e %

91 57 (63) 34 (37) 289 187 (65) 102 (35) 41 18 (44) 23 (56) 64 31 (48) 33 (52) 485

83 (26) 53 30 198 (61) 128 70 4 (1) 3 1 39 (12) 17 22 324

8 (5) 4 4 91 (56) 59 32 37 (23) 15 22 25 (16) 14 11 161

89 (98) 57 (100) 32 (94) 261 (90) 180 (96) 81 (79) 38 (93) 18 (100) 20 (87) 53 (83) 28 (90) 25 (81) 441 (91)

43 (47) 27 (47) 16 (47) 113 (39) 59 (32) 54 (53) 23 (56) 7 (39) 16 (70) 28 (44) 15 (45) 13 (41) 207 (43)

0.52 ± 0.05 0.70 ± 0.06 0.21 ± 0.07 0.57 ± 0.03 0.67 ± 0.04 0.37 ± 0.05 0.38 ± 0.08 0.49 ± 0.12 0.30 ± 0.10 0.37 ± 0.06 0.55 ± 0.10 0.21 ± 0.07 0.52 ± 0.02

0.45 ± 0.05 0.59 ± 0.07 0.21 ± 0.07 0.52 ± 0.03 0.63 ± 0.04 0.33 ± 0.05 0.38 ± 0.08 0.49 ± 0.12 0.30 ± 0.10 0.37 ± 0.06 0.55 ± 0.10 0.21 ± 0.07 0.47 ± 0.02

NOPHO High Risk (HR) arms, ALL-REZ Berlin Frankfurt Münster (BFM) relapse protocols, Relapse in Acute Lymphoblastic Leukemia (RALLE) pilot protocol,“Other treatment”; combinations of protocols, the Children´s Cancer and Leukemia Group (CCLG) ALLR3 relapse protocol, Children´s Cancer Group (CCG) relapse protocols and non-protocol treatment. Second complete remission (CR2), hematopoietic stem cell transplantation (HSCT). OS: overall survival; s.e.: standard error. *Proportion of relapsed patients within the primary protocol. ** Proportion of patients undergoing HSCT in CR2 within each relapse protocol. *** Patients with isolated extramedullary relapses were not enrolled in the study.

haematologica | 2016; 101(1)

Relapsed childhood ALL in the Nordic countries

Statistical analysis

Results

Institute (NCI) risk groups]. The only statistically significant differences in the relapse pattern between the upfront protocols were the distribution of cytogenetic findings (which can largely be explained by improvements in the detection methods), and the distribution of second events (which can partly be explained by the shorter follow-up time for the ALL-2000 patients). Subsequent relapse was the most common second event, occurring in 38% of patients. Second malignancies occurred in 11 patients of which 7 had undergone HSCT in CR2. To validate the commonly used risk classification systems, we retrospectively assigned relapse risk groups according to the criteria in the new IntReALL trial and compared the overall survival between the groups (Table 1). This risk-grouping is based on the CCLG and the BFM risk categories. Patients with early combined pre-B ALL relapses were stratified as standard-risk but the 5 year-OS was only 38.0±10.6% (standard eror, s.e.). Fifteen of these 21 patients underwent HSCT in CR2. In total, 57 patient with Down syndrome were treated according to the NOPHO ALL-92 and ALL-2000 protocols of which 18 relapsed (32%): 17 after chemotherapy only and one after HSCT in CR1. Twelve were initially treated with the NOPHO ALL-92 protocol and 6 with the ALL2000 protocol. Fifteen patients were categorized as standard-risk at relapse, but even though 14 of them reached CR2 and 3 subsequently underwent HSCT in CR2, only 3 survived long term. All deaths occured after second relapse.

Patients' characteristics and second events

Prognostic factors

The characteristics and second events among patients with ALL relapse are listed in Online Supplementary Tables S1 and S2. Of the 103 patients with isolated extramedullary relapses, 72 had isolated CNS relapses, 17 isolated testicular, three combined CNS and testicular, and 11 included other extramedullary sites (Online Supplementary Tables S2). In total, 134 patients (28%) had CNS involvement at relapse of which 30 (22%) were Tcell lineage. Of the 104 patients with very early relapses, 89 (86%) were initially stratified as high or greater [68% if classified as high-risk according to the National Cancer

Since we were studying a large cohort, we were able to include a number of variables in the regression analysis. The results of the Cox's proportional hazards regression analysis for overall survival are presented in Table 2. Primary risk groups were not included in the adjusted models since we adjusted for base-line characteristics at diagnosis. In the univariate analyses, age ten years or over at primary diagnosis, T-cell immunophenotype, short time in CR1, hyperleukocytosis at primary diagnosis, isolated bone marrow relapse, and unfavorable cytogenetics were adverse prognostic factors. In the first adjusted model,

Base-line variables were compared between the two up-front protocols using Fisher´s exact tests for categorical variables and the Mann-Whitney U test for continuous variables. In all survival analyses, the time-scale was defined by the time of diagnosis of relapse. Ten patients were lost to follow up, all in CR2 at the time of last contact and with a median time of follow up of 8.2 years (range 1.1-12.2 years). The Kaplan-Meier method was applied for generating survival curves for event-free survival (EFS) and overall survival (OS). Log rank test was used for comparing survival across groups. Cox's proportional hazards regression models were used for analysis of prognostic factors, estimating hazard ratios (HR) with 95% confidence intervals. OS was defined as the period from relapse diagnosis to death from any cause and censoring occurred at the date of last known follow up. In the analysis of EFS, patients were censored at the time of occurrence of the following second events: last known follow up for patients alive in CR2, second relapse, second malignancy (SMN), death caused by resistant disease or re-induction failure or death of undefined cause for patients in CR2. Cumulative incidence curves of second events were estimated by accounting for the competing nature of the second events.26 The methods used for analyzing HSCT patients are described in the Online Supplementary Appendix. All tests were twosided. P<0.05 were considered statistically significant. SPSS Statistics software version 21.0 and STATA version 13 were used for all statistical analyses.

Table 4. Cox’s proportional hazards regression analysis of risk factors for overall survival in patients stratified as standard-risk at acute lymphoblastic leukemia relapse with HSCT in CR2 as a time-dependent covariate.

Prognostic factors

HSCT in CR2 /total (%)

Unadjusted model HR (95% CI)

Adjusted model HR (95% CI)

HSCT in CR21 Male2 Age ≥ 10 years at primary diagnosis3 WBC ≥ 100 x 109/L at primary diagnosis4 Unfavorable cytogenetics5 Favorable cytogenetics5 Other cytogenetics5

108/283 (38%) 68/179 (38%)

2.94 (1.90 – 4.53)*** 0.70 (0.47 – 1.04)

2.82 (1.80 – 4.53)*** 1.29 (0.85 – 1.94)

28/46 (61%)

1.76 (1.10 – 2.81)*

1.39 (0.85 – 2.29)

11/21 (52%) 2/8 (25%) 33/117 (28%) 27/63 (43%)

1.11(0.54 – 2.30) 2.17 (0.91 – 5.20) 0.89 (0.56 – 1.43) 1.23 (0.73 – 2.10)

1.02 (0.49 – 2.15) 2.15 (0.88 – 5.25) 1.11 (0.68 - 1.83) 1.27 (0.74 – 2.16)

Hazard ratio (HR) for death. Number of patients included in regression models n=283, number of observations n=417. Patients that did not achieve CR2 were excluded from the model (n=10). Reference groups: 1chemotherapy only, 2female, 3age <10 years at primary diagnosis, 4white blood cell count (WBC) <100 x 109/L at primary diagnosis, 5missing data or no detected chromosomal abnormalities. Unfavorable cytogenetics: MLL rearrangements, hypodiploidy (modal chromosomal number <45), t(9;22), t(1;19). Favorable cytogenetics: high hyperdiploidy (modal chromosomal number >50) and t(12;21).

haematologica | 2016; 101(1)

T. Oskarsson et al.

Table 5. Reported outcomes of trials and cohorts in relapsed childhood acute lymphoblastic leukemia.

Relapse treatment

Relapse period

Risk group or type of relapse

N. of patients

Survival ± s.e.

Non-uniform relapse treatment NOPHO4

1981-1993

All risk groups

315

NOPHO

1992-2001

All risk groups

246

NOPHO

2002-2011

All risk groups

239

COG29

1988-2002

All risk groups

1961

11-year OS 33% ± 3 11-year EFS 28% ± 3 5-year OS 45% ± 3 5-year EFS 36% ± 3 5-year OS 58% ± 3 5-year EFS 51% ± 3 5-year OS 36% ± 2

COG28

1996-2003

All risk groups

347

3-year OS 56% ± 3 3-year EFS 45% ±3

Non-randomized trials MRC UKALL R149

1991-1995

All risk groups

256

5-year EFS 46% ± 3

CCLG ALL R27

1995-2002

All risk groups

150

RALLE Pilot13

2004-2010

BM involving

5-year OS 56% ± 4 5-year EFS 47% ± 4 5-year OS 37% ± 8 5-year EFS 37% ± 8

1997-2002

Pre-B iEM relapses

COPRALL-9750 Time in CR1 <24 months

Time in CR1 ≥24 months

ALL-REZ BFM 95/969

1995-2001

Intermediate risk

MRD <10-3 after induction

MRD ≥10-3 after induction

ALL-REZ BFM 20028 Continuation chemotherapy if MRD <10-3 after induction HSCT in CR2 if MRD ≥10-3 after induction Trials with randomizations ALL-REZ BFM 8710

2002-2009

Intermediate risk

Timing of ARA-C and MTX during induction

10-year OS 91% ± 4 10-year EFS 76% ± 6 10-year OS 32% ± 8 10-year EFS 18% ± 7

208 109

1987-1990

4-year OS 40% ± 8 4-year EFS 31% ± 8 4-year OS 76% ± 7 4-year EFS 61% ±8

8-year OS 73% ± 7 8-year EFS 70% ± 5 8-year OS 68% ± 5 8-year EFS 64% ± 5

All risk groups

183

15-year OS 37% ± 3 15-years EFS 30% ± 3

Very early and early BM involving relapses

All risk groups

525

10-year OS 36% ± 2 10-year EFS 30% ± 2 NS

ALL-REZ BFM 9012

1990-1995

MTX 1g/m2/36 hours vs. MTX 5g/m2/24 hours CCLG ALL R311

Pre-B iEM, pre-B early and late relapses 2003-2009 All risk groups

269 212

Idarubicin in induction

All risk groups

109

Mitoxantrone in induction

All risk groups

103

3-year OS 57% ± 4 3-year PFS 50% ± 4 3-year OS 45% ± 5 3-year PFS 36% ± 5 3-year OS 69% ± 5 3-year PFS 65% ± 5

Nordic Society of Paediatric Haematology and Oncology (NOPHO), Children´s Oncology Group (COG), Medical Research Council (MRC). Children’s Cancer and Leukemia Group (CCLG), Relapse in Acute Lymphoblastic Leukemia (RALLE) pilot, Berlin Frankfurt Münster (BFM). CR1: first complete remission; MRD: minimal residual disease; HSCT in CR2: hematopoietic stem cell transplantation in second complete remission; ARA-C: cytarabine; MTX: methotrexate; BM: bone marrow; iEM: isolated extramedullary.Very early relapse: occurring <18 months from primary diagnosis. Early relapse: occurring ≥18 months from diagnosis and <6 months after completion of primary therapy. Late relapse: occurring ≥ 6 months after completion of primary therapy. OS: overall survival; EFS: event-free survival; pFS: progression-free survival; NS: no statistically significant difference. If standard error (s.e.) was not available it was derived from the reported 95% confidence interval. aIncludes patients who underwent HSCT in CR1 (n=65, 45 allogenic and 19 autologous). In 1993, 6 NOPHO patients relapsed and are were included in both studies involving the 1981-1993 and 1992-2001 periods. bPatients enrolled in Children’s Cancer Group (CCG) trials 19882002 for initial treatment but non-uniform relapse treatment. cPatients enrolled in the CCG-1952 study for initial treatment, only standard risk ALL (WBC <50x109/L and age 1 to 9 years) but non-uniform relapse treatment. dPatients with early relapses (time in CR1 <24 months) were treated with VANDA induction and block therapy followed by HSCT in CR2 if donor available but patients with late relapses (time in CR1 ≥24 months were treated with block therapy followed by radiotherapy and maintenance. eIntermediate risk (S2): preB early or late combined bone marrow relapse, pre-B late isolated bone marrow relapse, pre-B or T-cell very early and early isolated extramedullary relapse.

haematologica | 2016; 101(1)

Relapsed childhood ALL in the Nordic countries

time to relapse (worse if earlier), site of relapse (worse if involving the bone marrow), age ten years or over at primary diagnosis, unfavorable cytogenetics and Down syndrome were all statistically significant independent prognostic factors. Adding up-front or relapse protocol to the adjusted model did not generate significant HRs or result in any notable change in the HRs of the other co-variates in the models. Immunophenotype is commonly used in the risk assessment at relapse, but although immunophenotype was a risk factor in the univariate analysis, it was not in the multivariate analysis. In the second adjusted model, we added an interaction variable combining immunophenotype and WBC at diagnosis and found that T-cell lineage disease with hyperleukocytosis at primary diagnosis (n=27; 24 high-risk, 3 standard risk) was a notable independent risk factor for survival after relapse, HR 2.38 (95%CI: 1.43-3.96; P=0.001). We analyzed separately the standard-risk group and adjusted for sex, age, Down syndrome and cytogenetics (data not shown). For the standard-risk group, age ten years or over HR 1.99 (1.243.21; P=0.004) and Down syndrome, HR 4.70 (2.46-8.94; P<0.001) were both independent prognostic factors for overall survival.

bone marrow relapses, but only 71% for very early bone marrow relapses, compared to 97% for late bone marrow relapses. There was no significante difference in CR2 rate between the two primary protocols, relapse periods or specific relapse protocols. We did not observe a statistically significant difference in overall survival between the relapse protocols used during the study period (Table 3).

Hematopoietic stem cell transplantation Hematopoietic stem cell transplantation in CR2 was performed in 207 of the 485 patients (43%) included in the study: 137 of 324 ALL-92 patients (42%) and 70 of 161 ALL-2000 patients (44%). The allocation to HSCT was 43% (102 of 239) during 1992-2001 and 43% (105 of 246) during 2002-2011. The proportion of standard-risk patients allocated to HSCT was slightly higher in the period 1992-2001 (41% vs. 34%) but the proportion of high-risk patients allocated to HSCT increased from 45%

Survival analysis In the whole study population, the 5-year EFS was 43.7±2.3% and the 5-year OS was 51.5±2.3%. At five years, the EFS for the ALL-92 patients was 42.1±2.7% and the OS was 49.8±2.8% but 46.8±4.2 and 52.7±4.4% for the ALL-2000 patients, respectively. To investigate if there was a generally improved prognosis over time, the patients were divided into two relapse periods, 1992-2001 (n=239) and 2002-2011 (n=246), approximately corresponding to the timing of the introduction of more general MRD measurements in the Nordic countries. Both OS and EFS were markedly higher for patients who relapsed between the years 2002-2011 compared to 1992-2001 (Figure 1A and B). HR for death was 0.62 (0.480.80; P<0.001) for 2002-2011 but for second events the HR was 0.64 (0.51-0.82; P<0.001). We compared the cumulative incidence of second events between the two relapse periods and found a reduction of second relapses in the later period (Figure 1C). There were no statistically significant differences between the time periods for the other second events. Looking for a possible explanation for these differences, we compared the pattern of relapse between the two time periods and observed a difference in the time distribution of relapses (Online Supplementary S3). In 1992-2001, 26.8% of relapses occurred very early, 32.6% early, and 40.6% late, but between 2002 and 2011, 16.3% occurred very early, 28.9% early, and 54.9% late (P=0.002).

Relapse treatment The ALL-REZ BFM protocols were the most commonly used treatment for ALL relapse (60%) (Table 3). The proportion of patients receiving BFM treatments was relatively stable over the whole study period but the proportion of patients receiving NOPHO treatment was much lower in the later part of the study period (no patient after 2005). In addition, 8 patients were treated with the CCLG’s ALLR3 protocol 2009-2011. The CR2 rate for the whole study period was 91%: 97% for standard-risk relapses and 82% for high-risk relapses. The CR2 rate for isolated extramedullary relapses was 95% and 90% for haematologica | 2016; 101(1)

Figure 1. Comparison of relapse periods 1992-2001 and 2002-2011. (A) Overall survival after relapse. (B) Event-free survival after relapse. (C) Cumulative incidence of second relapse using competing risks.

T. Oskarsson et al.

during the period 1992-2001 to 61% during the period 2002-2011. As expected, OS for high-risk patients was markedly higher if HSCT was performed in CR2, 46.7±5.1% compared to 25.0±6.0% (P<0.001). Interestingly, we observed the opposite association in the standard-risk group in which the overall survival was 61.1±4.8% for patients allocated to HSCT in CR2 compared to 74.5±3.6% (P=0.02) if continuation chemotherapy was used for consolidation. We investigated the effect of HSCT in CR2 on survival in the standard-risk group further in a time-dependent regression model and found a HR for the HSCT group of 2.94 (95%CI: 1.90-4.53) (Table 4). Adjusting for other non-stratifying base-line variables neither changed the HRs significantly (2.82; 95%CI: 1.80-4.41) nor yielded additional co-variates with a statistically significant and independent association with prognosis. Since there were only 3 patients with T-cell immunophenotype plus hyperleukocytosis in the standard-risk group, this variable was not included in the model as a single co-variate. Of the 27 patients with T-cell disease and hyperleukocytosis, 13 underwent HSCT in CR2 but only 4 survived (15%).

Discussion Relapse of ALL is still one of the most common childhood cancer subgroups with an incidence similar to rhabdomyosarcoma and nephroblastoma (Wilms tumor).27 Despite the vast improvement in outcome after up-front ALL treatment, the increase in survival after relapse has not been as pronounced.4,7,10-12,28 In this study, an improvement in OS over time was observed and is now close to 60%. Table 5 summarizes the reported results in relapsed childhood ALL from multicenter trials and cohorts over the last three decades. At present, patients with relapsed ALL are allocated to different risk groups based on the immunophenotype, the time from primary diagnosis to relapse and the anatomic site of relapse. But unlike the risk stratification at primary diagnosis, cytogenetics are not used to individualize the treatment intensity. We demonstrate that unfavorable cytogenetics, age ten years or over, T-cell immunophenotype with hyperleukocytosis and Down syndrome were all additional individual prognostic factors in relapsed ALL. The time from diagnosis to relapse is the strongest known individual risk factor for overall survival.4,29 Nearly 90% of patients with very early relapses were stratified as high-risk or greater at initial diagnosis indicating that clinical and genetic features present at diagnosis affect survival after relapse.30,31 In this study, two-thirds of the T-cell relapses occurred within 18 months, but contrary to previous findings, immunophenotype was not an individual prognostic factor for OS since it was over-ruled by other co-variates in the adjusted regression analysis. Interestingly, the interaction between WBC at diagnosis and T-cell immunophenotype created a strong prognostic variable. Only 4 out of 27 (15%) patients with T-cell immunophenotype and hyperleukocytosis survived longterm despite the use of HSCT in CR2 in 13 (48%) of them. However, this risk factor may be of limited additional value since with the current risk stratification, the majority of these patients are categorized as high-risk at relapse. Children with Down syndrome have both an increased risk of developing ALL32 and an increased risk of treat74

ment-related toxicity during primary ALL treatment.23-25,33 Historically, Down syndrome ALL (DS-ALL) has been associated with inferior outcome, both with regard to OS and EFS.34,35 In this study, DS-ALL was associated with very poor outcome, irrespective of the time period (early vs. late period) and the fact that most of these patients were stratified as standard-risk. Second relapses were the most common reason for treatment failure, indicating that patients with relapsed DS-ALL might have been treated with less intensive post-induction regimens to minimize the risk of treatment toxicity but subsequently failed to remain in long-term second remission.36 In a study by Meyr et al., children with DS had worse outcome after relapse mainly because of increased toxicity rather than subsequent relapse, but if the relapse occurred after the year 2000 this difference was not maintained.35 Adverse clinical factors, such as the time to relapse, age37,38 and WBC39 and cytogenetic risk factors,17,18,20 are most likely surrogate markers for underlying submicroscopic genetic abnormalities.40-42 With increased understanding of the biology of ALL, genetic factors are expected to be included in the future risk stratification and serve as targets for novel therapies.43-45 Despite the adjustments made to the NOPHO ALL-2000 protocol, OS did not differ significantly from the ALL-92 protocol: 5-year OS 89.1±1.1% and 87.6±0.8%, respectively.1 Although the relapse rate was lower after the ALL2000 treatment, it is expected that some of the late relapses from the ALL-2000 era are yet to occur. Although the pattern of relapse and outcome after relapse was very similar between the two NOPHO protocols, we observed a significant improvement in outcome for relapses occurring between 2002 and 2011 compared to 1992-2001, as well as a lower proportion of relapses generally associated with worse outcome (very early and early relapses) in the later period. In addition, we did not find a statistically significant difference in the CR2 rate or survival between the relapse protocols used during the study period, which supports the view that factors other than the protocol used explain the survival improvement and the changes in the relapse pattern between the two time periods. Minimal residual disease was measured in 73% of patients during the NOPHO ALL-2000 trial. However, although it was not used for risk stratification, it was optional to proceed to HSCT if MRD was 10-3 % or over after three months of treatment.1 The retrospective study design and the lack of detailed MRD-data in our cohort constitute a drawback, but to estimate the effect of MRD on survival in general we compared outcomes before and after the year 2002, roughly coinciding with the general introduction of MRD analysis in most Nordic childhood cancer centers. The use of MRD in the assessment of treatment response after reinduction and preceding allogeneic HSCT in CR2 was obligatory in the ALL-REZ BFM 2002 and RALLE protocols. However, although not obligatory in the NOPHO ALL-2000 HR protocol, it was still available in many centers, since evidence at that time supported the stratification by MRD over morphology.46-48 Therefore, after 2002 non-high-risk patients with high MRD levels after reinduction were recommended to undergo allogeneic HSCT in CR2 and a larger proportion of the high-risk patients was likely to be disease-free preceding HSCT.16,21 The introduction of MRD could, therefore, be one of the explanations for the observed overall reduction of first and second relapses over time. haematologica | 2016; 101(1)

Relapsed childhood ALL in the Nordic countries

Our results indicate that patients stratified as standardrisk at relapse have worse OS after HSCT in CR2 compared to chemotherapy only. Although we adjusted for base-line variables such as age, WBC at diagnosis and cytogenetics, this survival difference remained clearly significant. However, this may reflect the fact that those patients selected for HSCT had a higher risk based on the MRD response after re-induction. This would result in a selection of patients with a higher risk of death to the HSCT group and MRD negative patients to the chemotherapy group. Previous studies have shown superior outcome after HSCT in CR2 for MRD positive standard-risk patients,8,9 but in this study we did not find other risk factors or subgroups that seem to benefit from HSCT in CR2. Furthermore, the overall outcome of the SCTgroup improved in the second time-period, when MRD was presumably available speaking against this type of negative selection (data not shown). From 2015, the new international trial, IntReALL, will be the treatment of choice for relapsed childhood ALL in the Nordic countries. Our results indicate that the risk classification used in IntReALL is a reasonable approach, but we question whether early combined pre-B relapses should be classified as standard-risk instead of high-risk, since the 5-year overall survival for this subgroup was only 38.0% (Âą10.6%), despite the fact that 14 of the 21 patients underwent HSCT in CR2. In the ALL-REZ BFM 2002 protocol, intermediate-risk patients with high MRD levels after re-induction have been recommended to undergo allogeneic HSCT in CR2 if a donor is available.

References 7. 1. Schmiegelow K, Forestier E, Hellebostad M, et al. Long-term results of NOPHO ALL-92 and ALL-2000 studies of childhood acute lymphoblastic leukemia. Leukemia. 2010;24(2):345-354. 2. Hunger SP, Lu X, Devidas M, et al. Improved survival for children and adolescents with acute lymphoblastic leukemia between 1990 and 2005: a report from the children's oncology group. J Clin Oncol. 2012;30(14):16631669. 3.. Stary J, Zimmermann M, Campbell M, et al. Intensive Chemotherapy for Childhood Acute Lymphoblastic Leukemia: Results of the Randomized Intercontinental Trial ALL IC-BFM 2002. J Clin Oncol. 2014; 32(3):174184. 4. Schroeder H, Garwicz S, Kristinsson J, Siimes MA, Wesenberg F, Gustafsson G. Outcome after first relapse in children with acute lymphoblastic leukemia: a populationbased study of 315 patients from the Nordic Society of Pediatric Hematology and Oncology (NOPHO). Med Pediatr Oncol. 1995;25(5):372-378. 5. Moricke A, Zimmermann M, Reiter A, et al. Long-term results of five consecutive trials in childhood acute lymphoblastic leukemia performed by the ALL-BFM study group from 1981 to 2000. Leukemia. 2010;24 (2):265-284. 6. Escherich G, Horstmann MA, Zimmermann M, Janka-Schaub GE. Cooperative study group for childhood acute lymphoblastic leukaemia (COALL): long-term results of tri-

haematologica | 2016; 101(1)

10.

11.

12.

With these adjustments, the outcome for both good and poor responders has been similar (approximately 70% long-term OS), but the outcome for patients with early combined pre-B relapses has remained poor.8

Conclusion Over recent decades, improvements in the NOPHO ALL treatment have caused a reduction in the relapse rate. However, although improved survival over time was observed in this study, OS, especially for the high-risk patients, is still relatively poor. Most patients achieve second complete remission regardless of treatment protocol. But despite current treatment modalities, one-third of patients suffer second relapse. Therefore, better consolidation methods are needed without increasing the burden of treatment toxicities. Tailored risk-adapted treatment is the cornerstone of modern relapse therapy, but there is an urgent need for the development of new drugs and targeted therapies. There have been few reports on randomized controlled trials in patients with relapsed childhood ALL, and with the numbers of relapse patients decreasing, an international collaboration is very important to serve as a platform for progress to be made in the treatment of relapsed childhood ALL. Funding This project was supported with a grant from the Swedish Childhood Cancer Foundation, Barncancerfonden.

als 82,85,89,92 and 97. Leukemia. 2009;24 (2):298-308. Roy A, Cargill A, Love S, et al. Outcome after first relapse in childhood acute lymphoblastic leukaemia - lessons from the United Kingdom R2 trial. Br J Haematol. 2005;130(1):67-75. Eckert C, Henze G, Seeger K, et al. Use of allogeneic hematopoietic stem-cell transplantation based on minimal residual disease response improves outcomes for children with relapsed acute lymphoblastic leukemia in the intermediate-risk group. J Clin Oncol. 2013;31(21):2736-2742. Eckert C, von Stackelberg A, Seeger K, et al. Minimal residual disease after induction is the strongest predictor of prognosis in intermediate risk relapsed acute lymphoblastic leukaemia - long-term results of trial ALLREZ BFM P95/96. Eur J Cancer. 2013;49(6):1346-1355. Einsiedel HG, von Stackelberg A, Hartmann R, et al. Long-term outcome in children with relapsed ALL by risk-stratified salvage therapy: results of trial acute lymphoblastic leukemia-relapse study of the BerlinFrankfurt-Munster Group 87. J Clin Oncol. 2005;23(31):7942-7950. Parker C, Waters R, Leighton C, et al. Effect of mitoxantrone on outcome of children with first relapse of acute lymphoblastic leukaemia (ALL R3): an open-label randomised trial. Lancet. 2010;376(9757):20092017. Tallen G, Ratei R, Mann G, et al. Long-term outcome in children with relapsed acute lymphoblastic leukemia after time-point and site-of-relapse stratification and intensified

13.

14.

15.

16.

17.

18.

short-course multidrug chemotherapy: results of trial ALL-REZ BFM 90. J Clin Oncol. 2010;28(14):2339-2347. Saarinen-Pihkala UM, Parto K, Riikonen P, et al. RALLE pilot: response-guided therapy for marrow relapse in acute lymphoblastic leukemia in children. J Pediatr Hematol Oncol. 2012;34(4):263-270. Raetz EA, Borowitz MJ, Devidas M, et al. Reinduction platform for children with first marrow relapse of acute lymphoblastic Leukemia: A Children's Oncology Group Study[corrected]. J Clin Oncol. 2008;26(24):3971-3978. Paganin M, Zecca M, Fabbri G, et al. Minimal residual disease is an important predictive factor of outcome in children with relapsed 'high-risk' acute lymphoblastic leukemia. Leukemia. 2008;22(12):21932200. Coustan-Smith E, Gajjar A, Hijiya N, et al. Clinical significance of minimal residual disease in childhood acute lymphoblastic leukemia after first relapse. Leukemia. 2004;18(3):499-504. Forestier E, Johansson B, Gustafsson G, et al. Prognostic impact of karyotypic findings in childhood acute lymphoblastic leukaemia: a Nordic series comparing two treatment periods. For the Nordic Society of Paediatric Haematology and Oncology (NOPHO) Leukaemia Cytogenetic Study Group. Br J Haematol. 2000;110(1):147-153. Pui CH, Chessells JM, Camitta B, et al. Clinical heterogeneity in childhood acute lymphoblastic leukemia with 11q23 rearrangements. Leukemia. 2003;17(4):700706.

T. Oskarsson et al. 19. Raimondi SC, Zhou Y, Mathew S, et al. Reassessment of the prognostic significance of hypodiploidy in pediatric patients with acute lymphoblastic leukemia. Cancer. 2003;98(12):2715-2722. 20. Nachman JB, Heerema NA, Sather H, et al. Outcome of treatment in children with hypodiploid acute lymphoblastic leukemia. Blood. 2007;110(4):1112-1115. 21. Bader P, Kreyenberg H, Henze GH, et al. Prognostic value of minimal residual disease quantification before allogeneic stem-cell transplantation in relapsed childhood acute lymphoblastic leukemia: the ALL-REZ BFM Study Group. J Clin Oncol. 2009;27(3):377384. 22. Saarinen-Pihkala UM, Heilmann C, Winiarski J, et al. Pathways through relapses and deaths of children with acute lymphoblastic leukemia: role of allogeneic stemcell transplantation in Nordic data. J Clin Oncol. 2006;24(36):5750-5762. 23. Christensen MS, Heyman M, Mottonen M, Zeller B, Jonmundsson G, Hasle H. Treatment-related death in childhood acute lymphoblastic leukaemia in the Nordic countries: 1992-2001. Br J Haematol. 2005;131(1):50-58. 24. O'Connor D, Bate J, Wade R, et al. Infectionrelated mortality in children with acute lymphoblastic leukemia: an analysis of infectious deaths on UKALL2003. Blood. 2014;124(7):1056-1061. 25. Lund B, Asberg A, Heyman M, et al. Risk factors for treatment related mortality in childhood acute lymphoblastic leukaemia. Pediatr Blood Cancer. 2011;56(4):551-559. 26. Gray RJ. A Class of K-Sample Tests for Comparing the Cumulative Incidence of a Competing Risk. The Annals of Statistics. 1988;16(3):1141-1154. 27. Ward E, DeSantis C, Robbins A, Kohler B, Jemal A. Childhood and adolescent cancer statistics, 2014. CA: A Cancer Journal for Clinicians. 2014;64(2):83-103. 28. Malempati S, Gaynon PS, Sather H, La MK, Stork LC. Outcome after relapse among children with standard-risk acute lymphoblastic leukemia: Children's Oncology Group study CCG-1952. J Clin Oncol. 2007;25(36):58005807. 29. Nguyen K, Devidas M, Cheng SC, et al. Factors influencing survival after relapse from acute lymphoblastic leukemia: a Children's Oncology Group study. Leukemia. 2008;22(12):2142-2150. 30. Eckert C, Flohr T, Koehler R, et al. Very

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

early/early relapses of acute lymphoblastic leukemia show unexpected changes of clonal markers and high heterogeneity in response to initial and relapse treatment. Leukemia. 2011;25(8):1305-1313. Choi S, Henderson MJ, Kwan E, et al. Relapse in children with acute lymphoblastic leukemia involving selection of a preexisting drug-resistant subclone. Blood. 2007;110(2):632-639. Hasle H, Clemmensen IH, Mikkelsen M. Risks of leukaemia and solid tumours in individuals with Down's syndrome. Lancet. 2000;355(9199):165-169. Shah N, Al-Ahmari A, Al-Yamani A, Dupuis L, Stephens D, Hitzler J. Outcome and toxicity of chemotherapy for acute lymphoblastic leukemia in children with Down syndrome. Pediatr Blood Cancer. 2009;52(1):14-19. Buitenkamp TD, Izraeli S, Zimmermann M, et al. Acute lymphoblastic leukemia in children with Down syndrome: a retrospective analysis from the Ponte di Legno study group. Blood. 2014;123(1):70-77. Meyr F, Escherich G, Mann G, et al. Outcomes of treatment for relapsed acute lymphoblastic leukaemia in children with Down syndrome. Br J Haematol. 2013;162 (1):98-106. Bohnstedt C, Levinsen M, Rosthoj S, et al. Physicians compliance during maintenance therapy in children with Down syndrome and acute lymphoblastic leukemia. Leukemia. 2013;27(4):866-870. Mรถricke A, Zimmermann M, Reiter A, et al. Prognostic Impact of Age in Children and Adolescents with Acute Lymphoblastic Leukemia: Data from the Trials ALL-BFM 86, 90, and 95. Klin Padiatr. 2005;217(06): 310-320. Forestier E, Schmiegelow K, Nordic Society of Paediatric Haematology and Oncology NOPHO. The Incidence Peaks of the Childhood Acute Leukemias Reflect Specific Cytogenetic Aberrations. J Pediat Hematol Oncol. 2006;28(8):486-495. Vaitkeviciene G, Forestier E, Hellebostad M, et al. High white blood cell count at diagnosis of childhood acute lymphoblastic leukaemia: biological background and prognostic impact. Results from the NOPHO ALL-92 and ALL-2000 studies. Eur J Haematol. 2011;86(1):38-46. Yang JJ, Bhojwani D, Yang W, et al. Genome-wide copy number profiling reveals molecular evolution from diagnosis to relapse in childhood acute lymphoblastic

leukemia. Blood. 2008;112(10):4178-4183. 41. Staal FJ, de Ridder D, Szczepanski T, et al. Genome-wide expression analysis of paired diagnosis-relapse samples in ALL indicates involvement of pathways related to DNA replication, cell cycle and DNA repair, independent of immune phenotype. Leukemia. 2010;24(3):491-499. 42. Bhojwani D, Kang H, Moskowitz NP, et al. Biologic pathways associated with relapse in childhood acute lymphoblastic leukemia: a Children's Oncology Group study. Blood. 2006;108(2):711-717. 43. Hogan LE, Meyer JA, Yang J, et al. Integrated genomic analysis of relapsed childhood acute lymphoblastic leukemia reveals therapeutic strategies. Blood. 2011;118(19):5218-5226. 44. Kuiper RP, Waanders E, van der Velden VH, et al. IKZF1 deletions predict relapse in uniformly treated pediatric precursor B-ALL. Leukemia. 2010;24(7):1258-1264. 45. Chen IM, Harvey RC, Mullighan CG, et al. Outcome modeling with CRLF2, IKZF1, JAK, and minimal residual disease in pediatric acute lymphoblastic leukemia: a Children's Oncology Group study. Blood. 2012;119(15):3512-3522. 46. Panzer-Grumayer ER, Schneider M, Panzer S, Fasching K, Gadner H. Rapid molecular response during early induction chemotherapy predicts a good outcome in childhood acute lymphoblastic leukemia. Blood. 2000;95(3):790-794. 47. Biondi A, Valsecchi MG, Seriu T, et al. Molecular detection of minimal residual disease is a strong predictive factor of relapse in childhood B-lineage acute lymphoblastic leukemia with medium risk features. A case control study of the International BFM study group. Leukemia. 2000;14(11):1939-1943. 48. Donadieu J, Hill C. Early response to chemotherapy as a prognostic factor in childhood acute lymphoblastic leukaemia: a methodological review. Br J Haematol. 2001;115(1):34-45. 49. Lawson SE, Harrison G, Richards S, et al. The UK experience in treating relapsed childhood acute lymphoblastic leukaemia: a report on the medical research council UKALLR1 study. Br J Haematol. 2000;108 (3):531-543. 50. Domenech C, Mercier M, Plouvier E, et al. First isolated extramedullary relapse in children with B-cell precursor acute lymphoblastic leukaemia: results of the Cooprall-97 study. Eur J Cancer. 2008; 44(16):2461-2469.

haematologica | 2016; 101(1)

ARTICLE

Chronic Lymphocytic Leukemia

Clinical significance of bax/bcl-2 ratio in chronic lymphocytic leukemia

Maria Ilaria Del Principe,1 Michele Dal Bo, 2 Tamara Bittolo, 2 Francesco Buccisano, 1 Francesca Maria Rossi, 2 Antonella Zucchetto, 2 Davide Rossi, 3 Riccardo Bomben, 2 Luca Maurillo, 1 Mariagiovanna Cefalo, 1 Giovanna De Santis, 1 Adriano Venditti, 1 Gianluca Gaidano,3 Sergio Amadori,1 Paolo de Fabritiis,1 Valter Gattei,2 and Giovanni Del Poeta 1

Ematologia, Dipartimento di Biomedicina e Prevenzione, S. Eugenio Hospital and Università Tor Vergata, Roma; 2Unità Clinica e Sperimentale Onco-Ematologica, Centro di Riferimento Oncologico, I.R.C.C.S., Aviano; and 3Ematologia, Università del Piemonte Orientale, Novara, Italy 1

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Haematologica 2016 Volume 101(1):77-85

ABSTRACT

n chronic lymphocytic leukemia the balance between the pro-apoptotic and anti-apoptotic members of the bcl-2 family is involved in the pathogenesis, chemorefractoriness and clinical outcome. Moreover, the recently proposed anti-bcl-2 molecules, such as ABT-199, have emphasized the potential role of of bcl-2 family proteins in the context of target therapies. We investigated bax/bcl-2 ratio by flow cytometry in 502 patients and identified a cut off of 1.50 to correlate bax/bcl-2 ratio with well-established clinical and biological prognosticators. Bax/bcl-2 was 1.50 or over in 263 patients (52%) with chronic lymphocytic leukemia. Higher bax/bcl-2 was associated with low Rai stage, lymphocyte doubling time over 12 months, beta-2 microglobulin less than 2.2 mg/dL, soluble CD23 less than 70 U/mL and a low risk cytogenetic profile (P<0.0001). On the other hand, lower bax/bcl-2 was correlated with unmutated IGHV (P<0.0001), mutated NOTCH1 (P<0.0001) and mutated TP53 (P=0.00007). Significant shorter progression-free survival and overall survival were observed in patients with lower bax/bcl2 (P<0.0001). Moreover, within IGHV unmutated (168 patients) and TP53 mutated (37 patients) subgroups, higher bax/bcl-2 identified cases with significant longer PFS (P=0.00002 and P=0.039). In multivariate analysis of progression-free survival and overall survival, bax/bcl-2 was an independent prognostic factor (P=0.0002 and P=0.002). In conclusion, we defined the prognostic power of bax/bcl-2 ratio, as determined by a flow cytometric approach, and highlighted a correlation with chemoresistance and outcome in chronic lymphocytic leukemia. Finally, the recently proposed new therapies employing bcl-2 inhibitors prompted the potential use of bax/bcl-2 ratio to identify patients putatively resistant to these molecules.

Introduction Chronic lymphocytic leukemia (CLL) is a clinically heterogeneous disease characterized by the accumulation of CD5+CD23+ B-cell lymphocytes.1 The genetic and molecular mechanisms involved in the development of the disease are not well known, but recent observations suggest that the modulation of survival signals interfering with apoptosis may be a pivotal tool in the pathogenesis of CLL.2 Bcl-2 (B-cell lymphoma-2) family of antiapoptotic (bcl-2, bcl-xl, bcl-w and mcl-1) and proapoptotic (bax, bak and bok) proteins are critical regulators of apoptosis in CLL.3 In particular, bcl-2 has emerged as the most important protein in predicting haematologica | 2016; 101(1)

Correspondence: g.delpoeta@tin.it del.principe@med.uniroma2.it

Received: 8/6/2014. Accepted: 11/11/2015. Pre-published: 12/11/2015. doi:10.3324/haematol.2015.131854

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/1/77

M.I. Del Principe et al. A

Figure 1. Characteristics of bcl-2 and bax expressions. (A-E) Forward and side scatter of representative gated lymphocytes are shown. (B-F) Mouse IgG1 isotypic antibodies are used as controls for bcl-2 and bax positivities with their mean fluorescence intensities (MFI). (C-G) B-CLL (CD19+) cells (violet) were selected on the gated lymphocytes according to their phenotype. (D-H) Flow cytometric dot plots of bcl-2 and bax (violet) on CD19+ cells on peripheral blood sample of one representative CLL case with reported MFI values.

survival in CLL between 11 proteins that are implicated in the control of apoptosis, proliferation and differentiation.4 High levels of bcl-2 protein are found in CLL cells even in the absence of classical t(14;18) translocation that is detected frequently in follicular lymphomas.5 In this context, Cimmino et al.6 demonstrated that the loss of microRNAs (miR), miR-15a and miR-16-1, is involved in regulation of the bcl-2 family expression levels. Moreover, bcl-2 family members have also been linked to resistance of chemotherapy in CLL. In particular, Pepper et al.7 demonstrated that high levels of bcl-2 and low levels of bax protein expression are linked to chemoresistance in CLL cells treated with chlorambucil and a remarkable elevation of bax protein was found in CLL cells undergoing apoptosis. Several reports further confirmed that in CLL the balance between the proand anti-apoptotic members of the Bcl-2 family determines the chemotherapy sensitivity and survival.8,9 Finally, the recent introduction in the clinical use of novel potent oral pro-apoptotic BH3 peptidomimetics such as ABT-19910,11 and ABT-73712 has further emphasized the importance of bcl-2 family proteins in the context of CLL therapy. In the present study, we addressed the clinical impact of bax/bcl2 ratio, evaluated by a flow cytometric approach, as independent prognosticator of progression-free survival (PFS) and overall survival (OS) in CLL. In this context, we also correlated the bax/bcl-2 ratio with the other well-established clinical-biological prognosticators in CLL.

Methods Study design and patients This is a retrospective study that includes peripheral blood (PB) from 502 CLL patients recruited in a single center during the period 1989-2014. Approval of this study was obtained from the Institutional Review Board of the Centro di Riferimento 78

Oncologico (IRCSS) of Aviano (PN), Italy. Informed consent was provided in accordance with the Declaration of Helsinki. Patients were 279 males and 223 women with a median age of 66 years (range 33-89) at the time of diagnosis. Using the modified Rai staging system, 170 patients had a low stage, 318 an intermediate stage, and 14 a high stage. Median follow up was 7.9 years with 103 deaths and 399 censored patients. Two hundred and ninetyone patients out of 502 (58%) received chemotherapy for their disease. Eighty-seven patients were treated with a combination of chlorambucil plus prednisone and rituximab was added in 11 of them. One hundred and sixty patients received six courses of fludarabine monophosphate (Fludara; Genzyme, Modena, Italy) at 25 mg/m2 intravenously or 30-40 mg/m2 orally for five days every 28 days followed by four courses of rituximab (Mabthera; Roche, Basilea, Switzerland) at 375 mg/m2 for one day every week.13 Twenty-seven patients were treated with six courses of fludarabine, cyclophosphamide and rituximab.14 Finally, 17 patients aged over 60 years received bendamustine at 70 mg/m2 for two days and rituximab for one day at conventional dosages.15

Cellular immunophenotypic analysis All flow cytometric analyses were performed on a FACSCalibur flow cytometer (BD, Biosciences, CA, USA). The instrument was aligned and calibrated daily with the use of a 4-color mixture of CaliBRITE beads (BD, Biosciences) with FACSComp Sofware (BD, Biosciences), according to the manufacturerâ&#x20AC;&#x2122;s instructions. Triplecolor immunofluorescence analysis of surface antigens was performed by using combinations of phycoerythrin (PE), fluorescein isothiocyanate (FITC), and allophycocyanin (APC) monoclonal antibodies (MoAbs). The following conjugated MoAbs were used: anti-CD23 PE, anti-CD5 FITC, anti-CD5 APC, anti-CD38 PE, antiCD19 PE, anti-CD19 APC, anti-CD49d PE and anti-CD45 FITC (BD, Biosciences). The peripheral blood (PB) mononuclear cells were analyzed for surface expression of CD19/CD5/CD38, CD19/CD5/CD23 and CD19/CD5/CD49d in order to confirm CLL diagnosis. haematologica | 2016; 101(1)

bax/bcl-2 ratio in CLL

Bcl-2 and bax expression

All samples were tested at diagnosis or before any therapeutic approach. For multiparameter analysis of bcl-2 and bax combined with lymphoid antigens, mononuclear cells were initially incubated with CD19 PE and CD5 APC MoAbs for 30 min at 4°C. Subsequently, the cells, after washing twice in PBS, were fixed and permeabilized with Fix & Perm Cell Permeabilization kit® (Fix&Perm, Caltag Laboratories, Carlsbad, CA, USA). All samples were then incubated at 4°C for 30 min with either 10 mL anti-bcl-2 FITC or 20 mL unconjugated anti-bax. For bax, cells were further incubated at 4°C for 30 min with 100 mL FITC-

Bcl-2 and bax oncoproteins were evaluated by flow cytometry using the following two MoAbs: 1) Anti-human bcl-2 clone 124 FITC-conjugated MoAb (Dako, Glostrup, Denmark) is an immunoglobulin G1 (IgG1) that reacts specifically with bcl-2 oncoprotein associated with mitochondria, smooth endoplasmic reticulum, and perinuclear membrane;16,17 2) anti-bax clone Ab-2 (Oncogene, Cambridge, MA, USA) is an IgG1 that recognizes bax, normally localized to the cytoplasm, but translocating rapidly to the mitochondria after the induction of an apoptotic signal.18

Table 1. Distribution of prognostic factors in CLL according to bax/bcl-2 values.

Parameter Age <60 years >60 years B2M <2.2 g/L >2.2 g/L Mod-Rai I II-III sCD23 <70 UI/ml >70 UI/ml CD38 <30% >30% FISH Normal/del13q Int/Poor (+12, del11q, del17p) IGHV M UM TP53 M WT NOTCH-1 M WT

bax/bcl-2, no ≥1.50 <1.50

P¶

n§

10-year OS,%

10-year PFS, %

83 183

85 151

0.25

502

93 77

<0.0001

36 35

0.51

185 81

92 144

<0.0001

502

92 69

<0.0001

52 15

<0.0001

121 145

49 187

<0.0001

502

92 79

0.00001

59 24

<0.0001

193 44

101 109

<0.0001

447

92 60

<0.0001

52 6

<0.0001

222 44

151 85

<0.0001

502

87 70

<0.0001

44 12

<0.0001

221 38

120 110

<0.0001

489

90 64

<0.0001

46 9

<0.0001

221 38

106 130

<0.0001

495

92 61

<0.0001

47 13

<0.0001

8 258

29 207

0.00007

502

87 30

<0.0001

5 38

0.0001

9 257

49 187

<0.0001

502

85 60

<0.0001

0 40

<0.0001

Fisher exact tests were performed to evaluate the association between bax/bcl-2 values (below or above the established threshold set at ≥1.50 of positive CLL cells) and other prognostic factors. §Values refer to the number of cases analyzed for a given feature. *P values were calculated by the log rank test.

Table 2. Multivariate Cox regression analysis of progression-free survival.

Parameter Bax/bcl-2 NOTCH1 sCD23 B2M Mod-Rai CD38 IGHV FISH cytogenetics TP53

N=489 patients

N=435 patients

0.52 1.32 NT 1.91 2.08 1.79 0.73 1.37 2.32

0.000004 0.09 NT 0.000002 0.000008 0.00003 0.022 0.023 0.00002

0.57 1.26 1.57 1.65 1.59 1.65 0.77 1.34 2.24

0.0003 0.19 0.006 0.0011 0.0007 0.0009 0.11 0.05 0.0001

HR: hazard ratio; NT: not tested.

haematologica | 2016; 101(1)

M.I. Del Principe et al.

conjugated F(ab)2 fragment of goat antimouse immunoglobulin (dilution 1:50; Dako, Glostrup; Denmark). Negative controls were performed by incubating cells with non-specific isotype IgG1 antibodies (Figure 1B and F). The results were performed by gating lymphoid cells on a SSC versus FSC plot (Figure 1A and E) and examining bcl-2 and bax as mean fluorescence intensity (MFI) on CD19 positive/CD5 positive cells (Figure 1C, D, G and H).19 Bcl-2 and bax were then evaluated as relative mean fluorescence intensities (rMFIs), calculated as the ratio between bcl-2 or bax MoAbs MFI and non-specific MoAb MFI on B cells. Finally, the results were expressed as an index (bax/bcl-2) obtained by dividing rMFI bax and rMFI bcl-2. Estimation of bax/bcl-2 ratio yelding the best separation of 2 subgroups with different OS and/or PFS probabilities was made by applying various methods including Youden’s index, median value and receiver-operating characteristic (ROC) analysis. Finally, the threshold was set at the bax/bcl-2 median value equal or higher than 1.50 (range 0.27-6.10) confirmed also by ROC analysis (Online Supplementary Figure S1). Two small subsets of patients were studied using 20 ml of two directly conjugated anti–bax MoAbs (clone 2D2), one conjugated with FITC (Santa Cruz Biotechnology, Dallas, TX, USA) and the other with APC (Abcore, Ramona, CA, USA).

more than 2% from the corresponding germline gene were defined as mutated.23

NOTCH1 mutational status and TP53 mutational status The presence of NOTCH1 mutations was investigated with amplification refractory mutation system (ARMS) PCR for c.75447545delCT and by Sanger sequencing of NOTCH1 exon 34.24 Mutation analysis of TP53 exons 2 to 11 was carried out by DNA direct sequencing on an ABI Prism 3130 automated DNA sequence analyzer (Applied Biosystems, Foster City, CA, USA) according to the International Agency for Research on Cancer (IARC) guidelines (www.p53-iarc.fr) and analyzed with the Sequencing Analysis v.5.2 software (Applied Biosystems), as previously reported.25 Mutations were confirmed on both strands on independent amplimers and validated by the IARC TP53 Mutation Database R15, as previously reported.

Statistical analysis Correlations between bax/bcl-2 ratio and the other biological and clinical variables were based on the two-tailed Fisher exact test.

Soluble CD23 and B2-microglobulin detection Soluble CD23 (sCD23) immunoenzymometric assay and β2-microglobulin (β2M) determinations were performed as described elsewhere.20 The threshold of positivity was set at 70 U/mL for sCD23 and 2.2 g/L for β2M on the basis of normal laboratory references values.

Interphase FISH Separate hybridations were carried out for loci on chromosome 11, 12, 13 and 17. For chromosome 11, 13, 12 and 17 specific probes and procedures were used, as reported previously.21

IGHV mutation analysis Total RNAs were extracted and reverse-transcribed. The resulting cDNAs, checked for first strand synthesis,22 were amplified using a mixture of sense primers annealing to either VH1 through VH6 leader sequences or 5’ends of VH1-VH6 FR1. Purified amplifications were cloned and a minimum of 10-20 colonies for each case were sequenced; CLL specific IGHV transcripts, identified by sharing the same CDR3 in multiple clones were analyzed for percent mutation, as previously described. VH gene sequences deviating

Figure 2. Duration of response after fludarabine and rituximab in relation to bax/bcl-2 values. Progression-free survival after treatment was significantly longer within chronic lymphocytic leukemia subgroup showing bax/bcl-2 ≥1.50 (bax/bcl2+; 46% vs. 30% at 4 years; P=0.021).

Table 3. Multivariate Cox regression analyses of overall survival.

Parameter Bax/bcl-2 Age > or < 60 years B2-M TP53 CD38 NOTCH1 Mod-Rai IGHV FISH cytogenetics sCD23

N=489 patients

N=435 patients

0.41 3.91 2.50 2.74 1.13 1.83 1.81 0.37 1.85 NT

0.0022 0.000001 0.001 0.0001 0.58 0.02 0.03 0.0001 0.010 NT

0.46 3.51 2.11 2.71 1.03 1.76 1.64 0.41 1.88 1.37

0.011 0.00002 0.022 0.0004 0.91 0.045 0.057 0.001 0.016 0.27

HR: hazard ratio; NT,: not tested.

haematologica | 2016; 101(1)

bax/bcl-2 ratio in CLL

The clinical assessment of CLL patients was based both on the National Cancer Institute Working Group criteria26 and on the International Workshop on Chronic Lymphocytic Leukemia criteria after 2008.27 PFS and OS, measured from diagnosis, were estimated according to the Kaplan-Meier method and compared between groups by means of the log rank test. Cox proportional hazards regression models were used to assess the independent effect of covariables, treated as dichotomous, both on the PFS and OS.

FITC was 18.32 (range 4.0-38.9) and the same value of bax/bcl-2 ratio (≥1.50) was used as cut off. Interestingly, we confirmed the stability over time of bax/bcl-2 ratio in 46 of our untreated patients in which this index was repeated three times along the course of the disease (Online Supplementary Table S4). It is worthy of note that, in some patients treated with fludarabine, we observed that bax/bcl-2 ratio increased progressively from one day post treatment to one month post treatment and this increase was higher in responsive patients (data not shown).

Results Expression of bax/bcl-2 and association with other prognostic factors Bax and bcl-2 presented variable pattern of mean fluorescence intensity. Median rMFIs were 12.39 (range 3.85103.0) for bcl-2 and 17.44 (range 2.20-100.0) for bax, respectively. Median bax/bcl-2 ratio (bax/bcl-2) was 1.50 (range 0.26-6.10). The median bax/bcl-2 ratio was chosen as cut off to discriminate subgroups of cases, being very similar to the best cut off defined by ROC analysis (Online Supplementary Figure S1). Using this cut off, 263 patients (263 of 502, 52%) were bax/bcl-2 positive (≥1.50). An inverse correlation was observed between bax/bcl-2 levels and PB B-cell count (Spearman correlation, r=0.26, P<0.0001). Higher bax/bcl-2 was significantly associated with low Rai stage, lymphocyte doubling time (LDT) over 12 months, β2M less than 2.2 mg/dL and soluble CD23 less than 70 U/mL (P<0.0001) (Table 1). Moreover, lower bax/bcl-2 was significantly associated both with CD38 more than 30% (P<0.0001) (Table 1). and CD49d more than 30% (P=0.010) (Table 1). Four hundred and eightynine patients were analyzed by interphase FISH to evaluate deletions in chromosome bands 17p13, 11q23, 13q14, and trisomy of band 12q13. There was a significant correlation between higher bax/bcl-2 and low risk (normal karyotype or del13q) cytogenetics (P<0.0001) (Table 1). Conversely, lower bax/bcl-2 was significantly represented within the intermediate/high risk (trisomy 12, del11q and del17p) classes (P<0.0001) (Table 1). Significant associations were also found between lower bax/bcl-2 and IGHV unmutated status (P<0.0001) or NOTCH1 mutations (P<0.0001) or TP53 mutations (P=0.00007) (Table 1). Moreover, we analyzed a validation cohort of 233 CLL patients recruited in another Institution (Policlinico Tor Vergata) and analyzed in another laboratory located at University Tor Vergata obtaining similar results with regard to bax/bcl-2 ratio and all the above mentioned variables (Online Supplementary Table S1 and Figure S2). Interestingly, 54 of these patients were the same as our series with similar results regarding bax/bcl-2 ratio. We repeated these assays in 20 thawed samples using an antibax MoAb APC (Abcore, Ramona, CA, USA) and an antibcl-2 MoAb APC (Abcore, Ramona, CA, USA). Then we compared these bax/bcl-2 ratios with the bax/bcl-2 ratios previously obtained on the same patients using fresh cells and an anti-bax unconjugated MoAb. We obtained similar results in 20 of 20 cases (Online Supplementary Table S2) using the same cut off (≥1.50). Interestingly, significant correlations between bax/bcl-2 ratio and some other prognostic factors (Online Supplementary Table S3) were confirmed in 52 patients who were recently studied using an anti-bax FITC MoAb (clone 2D2, Santa Cruz Biotechnology, Dallas, TX, USA) and anti-bcl-2 FITC MoAb (clone 124). In this subset, median rMFI for bax haematologica | 2016; 101(1)

Relevance of bax/bcl-2 ratio as biological prognostic marker With regard to clinical behavior, a significant correlation was found between bax/bcl-2 and treatment. In particular, 197 of 236 (83%) with lower bax/bcl-2 received

Figure 3. Progression-free survival (PFS) and overall survival (OS) curves based on bax/bcl-2 values. Kaplan-Meier plot comparing PFS (A) and OS (B) based on the detection of bax/bcl-2 ratio ≥1.50 (bax/bcl2+) or <1.50 (bax/bcl2–). Bax/bcl2+ patients experienced both a longer PFS and OS (P<0.0001).

M.I. Del Principe et al.

chemotherapy, while only 90 (33%) of 263 patients with bax/bcl-2 1.50 or over were treated (P<0.0001). Moreover, 82 (35%) bax/bcl-2 negative patients received two or more treatment regimens versus 25 (9%) bax/bcl-2 positive patients (P<0.0001). Interestingly, patients with lower bax/bcl-2 went more rapidly to treatment (27 months vs. 36 months; P=0.019). It is worthy of note that, in a series of 160 patients treated in induction with fludarabine plus rituximab, all resistant cases (8 patients) were bax/bcl-2 negative, while among patients who achieved a response (152 patients), those bax/bcl-2 positive showed a significant longer response duration compared to bax/bcl-2 negative cases (46% vs. 30% at 4 years; P=0.021) (Figure 2). With regard to clinical outcome, significant shorter PFS and OS were observed in patients with lower bax/bcl-2 (10% vs. 52% at 16 years, P<0.0001 and 46% vs. 79% at 16 years, P<0.0001, respectively) (Figure 3A and B). Similar results were obtained when analyzing the impact on PFS and OS of IGHV gene mutations as well as of TP53 and NOTCH1 mutations (Table 1). A better refinement in the prognostic assessment of PFS and OS was obtained by combining bax/bcl-2 values with those of TP53 and NOTCH1 (Figure 4A and B, and D and E). For all these variables, bax/bcl-2 expression had true additive properties. In particular, positive or negative bax/bcl-2 combined both with wild-type or mutated TP53 or with wild-type or mutated NOTCH1 identified 2 subsets of patients, the former with the best prognosis and the latter with the worst prognosis with regard to both PFS (65% vs. 10% at 8 years, P<0.0001 and 66% vs. 4% at 8 years, P<0.0001) and OS (96% vs. 33% at 10 years, P<0.0001 and 95% vs. 52% at 10 years, P<0.0001). In all combinations, discordant

patients showed intermediate outcomes (Figure 4A and B, and D and E). Similar behavior was observed by combining bax/bcl-2 and IGHV mutational status in additional bivariate analyses, the shortest PFS and OS time intervals being found for bax/bcl-2 negative/UM IGHV patients (7% vs. 63% at 10 years and 52% vs. 97% at 10 years; P<0.0001) (Figure 4C and F). To further explore the prognostic impact of bax/bcl-2, we investigated its expression within TP53 mutated (n=37) and IGHV unmutated (n=168) subgroups that notoriously have the worst prognosis. As a matter of fact, higher bax/bcl-2 identified patients with a significant longer PFS (50% vs. 10% and 43% vs. 10% at 7 years; P=0.039 and P=0.00002, respectively) (Figure 5A and B), so confirming its very high prognostic impact.

Multivariate analysis The clinical impact of bax/bcl-2 as independent prognosticator both for PFS and OS was also checked by multivariate Cox proportional hazards analysis applied to models including the prognosticators proven to be significant in univariate analysis (Table 1). In a cohort of 489 patients, in which all the above mentioned prognostic factors were available at the same time, bax/bcl-2 was confirmed to be an independent prognosticator with regard to PFS along with mod-Rai, β2M, FISH, TP53 and IGHV (Table 2). Similarly, multivariate analysis selected age, β2M, bax/bcl2, FISH, NOTCH1, TP53 and IGHV as significant independent prognosticators for OS (Table 3). Equally, within a subset of 435 patients, in which sCD23 was included, bax/bcl2 was confirmed again as an independent prognosticator both with regard to PFS and OS (Tables 2 and 3).

Figure 4. Progression-free survival (PFS) and overall survival (OS) curves in relation to combined bax/bcl-2 ratio and TP53 or bax/bcl-2 and NOTCH1 or bax/bcl-2 and IGHV status. PFS and OS were significantly longer within the bax/bcl2+ (≥1.50) TP53 wild-type (wt) subgroup (A-D), or within bax/bcl2+ NOTCH1 wt patients (BE) or within bax/bcl2+IGHV mutated (m) cases (C-F). Discordant patients showed an intermediate outcome.

haematologica | 2016; 101(1)

bax/bcl-2 ratio in CLL

Discussion In CLL malignant clonal B cells accumulate because of dysregulated production and inappropriately prolonged survival due to impairment of apoptosis.2 In fact, the alteration of some apoptotic mechanisms such as the balance between the pro- (bax) and anti-apoptotic members (bcl2) of the bcl-2 family, is recognized to be an important factor not only in the pathogenesis but also in the development of chemorefractoriness in CLL.7 Thus, measures of bax/bcl-2 could provide information on the intrinsic chemosensitivity of CLL influencing response to treatment and clinical outcome. In the present study, we addressed the clinical impact of bax/bcl-2 ratio, evaluated with a flow cytometric approach, as independent prognostic factor in a large homogeneous series of 502 patients affected by CLL. We had previously demonstrated that the quantification of bcl-2 and bax, measuring fluorescence intensity by flow cytometry, is technically adequate and useful for a correct evaluation of clinical and prognostic correlations both in acute myeloid leukemia17 and in CLL.19 In the present study, we used a well-standardized and simple flow cytometric method for bcl-2 and bax determinations. In particular the methodological feasibility and reproducibility of this flow cytometric approach was confirmed by the finding that superimposable results of bax/bcl-2 ratios were obtained in a series of CLL cases for which bax/bcl-2 measures from fresh PB samples were repeated in two different laboratories (Online Supplementary Table S1), or when two directly conjugated anti-bax MoAbs were employed (Online Supplementary Tables S2 and S3). The robustness of this flow cytometry-based method was also highlighted by the fact that in a series of 46 untreated patients in which the bax/bcl-2 measures were performed three times along the course of the disease, but always before any treatment, the bax/bcl-2 ratio was seen to be stable over time (Online Supplementary Table S4). This stability of the ratio underlines a reliable use of bax/bcl-2 ratio as prognostic marker, although further analysis is required to clarify this important issue. However, to the best of our knowledge, this is the largest study in CLL patients on clinical significance of bax/bcl-2 index by flow cytometry. Clinically, by using a cut off of 1.50, representing the median value of bax/bcl-2 ratio of our case cohort, and further confirmed as best cut off by ROC analysis, we found significant correlations of bax/bcl-2 ratio both with tumor burden indicators (PB B lymphocytes, Rai stages, β2M, sCD23 and LDT) and with other biological prognosticators (CD38, CD49d, IGHV mutations, cytogenetic abnormalities, NOTCH1 and TP53). Specifically, we observed that a higher bax/bcl-2 ratio showed a strict correlation with some favorable clinical features such as low PB B cells and low/intermediate Rai stages. Moreover, low levels of sCD23 and β2M, known markers of disease activity in CLL, were correlated with higher apoptosis levels,28 although previously published studies did not find any significant association between some apoptosis regulation proteins such as bcl-2, mcl-1 and Rai stages, while their correlation with CD23 or β2M was not investigated.29-32 Most of these studies involved a small number of patients and methods other than flow cytometry, such as Western blotting, immunoblotting and PCR.33 However, Kitada et al.32 demonstrated that high bcl-2/bax ratio was haematologica | 2016; 101(1)

associated with elevated white blood cell count, similar to that found in our series. Moreover, we noted also that LDT was significantly longer in bax/bcl-2 positive patients, thus suggesting a significant correlation between this ratio and low proliferation. From a biological point of view, we found a significant relationship between this index and some well-known biological prognosticators. First, we observed a significant lower bax/bcl-2 ratio in CD38 and CD49d positive patients. It is well known that CD38 and CD49d overexpression are adverse prognostic markers in CLL19,34,35 and it has been reported that interactions involving CD38 and CD49d prevent spontaneous and drug-induced apoptosis of normal and neoplastic B cells.36 Interestingly, co-culture of CLL cells with human vascular endothelial cells determined a simultaneous overexpression of CD38, CD49d

Figure 5. Progression-free survival (PFS) curves based on bax/bcl-2 within IGHV unmutated subgroup and TP53 mutated subgroup. Bax/bcl2+ (≥1.50) patients showed a significant longer PFS both within TP53 mutated subgroup (A) and within IGHV unmutated subset (B).

M.I. Del Principe et al.

and bcl-2 with an enhanced tumor cell survival targeting NF-κB activation.37 Recently, it has been observed that CD49d was nearly universally expressed in trisomy 12 CLL patients.36 In our study, we demonstrated a significant correlation between lower bax/bcl-2 ratio and trisomy 12 (Table 1), so suggesting that this ratio may represent a further mechanism of chemoresistance in this subset. A strict association was also observed between bax/bcl-2 negative patients and other high risk (del11q and del17p) classes (Table 1), thus confirming the clinical impact of the apoptotic mechanisms in these poor risk cytogenetic subsets. Conversely, higher bax/bcl-2 was significantly represented in low risk (normal or del13q) cytogenetic groups (Table 1). In addition, we found that a lower bax/bcl-2 ratio was associated with NOTCH1 mutations (Table 1), unfavorable prognosticator in CLL,38-40 in keeping with previous studies showing that NOTCH1 mutations activate NFkB pathway resulting in upregulations of target genes such as bcl-2 in T-acute lymphoblastic leukemia.41,42 Significant associations were also found between lower bax/bcl-2 ratio and UM-IGHV or TP53 mutations (P<0.0001) (Table 1), thus confirming that biological mechanisms related to defective apoptosis play a pivotal role both in the chemoresistance and in the unfavorable clinical course of these patient subgroups. Results of this study also confirmed a correlation between bax/bcl-2 levels and chemoresistance in CLL, in keeping with previous studies.7,33,43,44 In particular, significant differences in treatment histories were found between bax/bcl-2 positive and negative patients. In fact, bax/bcl-2 negative patients underwent treatment more frequently and rapidly than bax/bcl-2 positive patients. Moreover, they presented a shorter response duration (Figure 2) and received, significantly, more chemotherapy regimens than patients with higher levels of bax/bcl-2. Clinically, we demonstrated that the bax/bcl-2 ratio had an impact for PFS and OS in our large series of CLL patients, together with CD38, IGHV mutations, cytogenetic abnormalities, NOTCH1 and TP53 (Table 1). Moreover, bax/bcl-2 ratio was confirmed to be an independent prognostic factor with regard both to PFS and OS in multivariate analysis, considering 489 patients in which all the most important recognized prognostic factors were included (Tables 2 and 3). Furthermore, bax/bcl-2 ratio demonstrated significant additive prognostic properties by the fact that positive or negative bax/bcl-2 combined both with wild-type or mutated TP53 or with wild-type or mutated NOTCH1 identified 2 subsets of patients, the former with the best prognosis and the latter with the worst prognosis with regard to both PFS and OS (Figure 4). The superior prognostic capacity of bax/bcl-2 with respect both to IGHV mutation status and TP53 mutations25,45,46 was corroborated by the observation that higher bax/bcl-2 ratio identified patients with a significant longer PFS within IGHV unmu-

References 1. Rozman C, Montserrat E. Chronic Lymphocytic leukemia. N Engl J Med. 1995;333(16):1052-1057. 2. Chiorazzi N, Rai KR, Ferrarini M. Chronic

tated and TP53 mutated subgroups, notoriously with the worst prognosis (Figure 5). The capacity of bax/bcl-2 index to identify subsets of patients with a different apoptotic profile that reflects prognosis could be of interest in the light of the recently introduced bcl-2 inhibitors such as ABT-199, a 2nd-generation BH3 mimetic, highly selective for bcl-2 with demonstrated high rates of activity in relapsed/refractory CLL patients.3,10 In this regard, the bax/bcl-2 ratio, determined with the flow cytometric approach, as proposed in the present study, could represent a proper predictive biomarker to identify CLL patients who could benefit from the use of ABT-199 or who, on the contrary, could be more effectively treated with conventional chemotherapy. Experiments of apoptosis testing ABT-199 on CLL cells in order to use bax/bcl-2 ratio for identifying the sensitivity degree of patients and thus monitoring the anti-bcl-2 therapy are in progress at our Institution (data not shown). In this context, antibodies able to detect bax conformational changes could be useful to characterize apoptosis levels.47 In conclusion, in the present study, we defined the prognostic power of bax/bcl-2 ratio, as determined by a flow cytometric approach, easily transferable to routine laboratory practice. Moreover, we highlighted the correlation of these two mitochondrial apoptotic proteins with chemoresistance and clinical outcome in CLL. Finally, the recently proposed new therapies employing bcl-2 inhibitors, such as ABT-199, prompted the potential use of bax/bcl-2 ratio to identify CLL cases putatively resistant to these molecules. Funding This study was supported in part by Ministero dell’Università e della Ricerca Scientifica e Tecnologica (MURST), Programmi di Ricerca di Interesse Nazionale, 2005; Ministero della Salute (Ricerca Finalizzata Istituto di Ricovero e Cura a Carattere Scientifico [IRCCS], Rome, Italy; the Associazione Italiana Ricerca Cancro (AIRC), Investigator Grant IG-13227; Progetto Ricerca Finalizzata I.R.C.C.S. n. RF-2009-1469205, n. RF2010-2307262; Progetto Giovani Ricercatori n.GR-20091475467, n.GR-2010-2317594, n.GR-2011-02347441, n.GR2011-02346826; Ministero della Salute, Rome, Italy; Fondazione Cariplo (grant 2012-0689); Associazione Italiana contro leLeucemie, Linfomi e Mielomi (AIL), Venezia Section, Pramaggiore Group, Italy; Fondazione per la Vita di Pordenone, Italy; Ricerca Scientifica Applicata, Regione Friuli Venezia Giulia (“Linfonet” Project), Trieste, Italy; “5x1000 Intramural Program”, Centro di Riferimento Oncologico, Aviano, Italy. F.A. is supported by a Beat-Leukemia fellowship. Acknowledgments Finally, all the authors thank the members of our Department of Haematology clinical staff for their invaluable support to our CLL clinical research program. Part of of this article was presented orally at the 56th Annual Meeting of the American Society of Hematology, San Francisco, CA, USA, December 6-9, 2014.

Lymphocytic leukemia. N Engl J Med. 2005;352(8):804-815. 3. Scarfò L, Ghia P. Reprogramming cell death: BCL2 family inhibition in hematological malignancies. Immunol Lett. 2013;155(1-2):36-39. 4. Faderl S, Keating MJ, Do KA, et al.

Expression profile of 11 proteins and their prognostic significance in patients with chronic lymphocytic leukemia (CLL.). Leukemia. 2002;16(6):1045-1052. 5. Hanada M, Delia D, Aiello A, Stadtmauer E, Reed JC. Bcl-2 gene hypomethylation and high levels espression in B-cell chronic

haematologica | 2016; 101(1)

bax/bcl-2 ratio in CLL

8. 9.

10.

11. 12.

13.

14.

15.

16. 17.

18.

19.

20.

lymphocytic leukemia. Blood. 1993;82(6):1820-1828. Cimmino A, Calin GA, Fabbri M, et al. miR-15 and miR-16 induce apoptosis by targeting BCL2. Proc Natl Acad Sci USA. 2005;102(39):13944-13949. Pepper C, Hoy T, Bentley P. Elevated bcl2/bax are a consistent feature of apoptosis resistance in B-cell chronic lymphocytic leukemia and are correlated with in vivo chemoresistance. Leuk Lymphoma. 1998;28(3-4):355-361. Fegan C, Pepper C. Apoptosis deregulation in CLL. Adv Exp Med Biol. 2013;792:151171. Williamson KE, Kelly JD, Hamilton PW, McManus D, Johnston SR. Bcl-2/Bax ratios in chronic lymphocytic leukaemia and their correlation with in vitro apoptosis and clinical resistance. Br J Cancer. 1998;78 (4):553554. Souers AJ, Leverson JD, Boghaert ER, et al. ABT-199, a potent and selective BCL-2 inhibitor, achieves antitumor activity while sparing platelets. Nat Med. 2013;19(2):202208. Davids MS, Letai A. ABT-199: a new hope for selective BCL-2 inhibition. Cancer Cell. 2013;23(2):139-141. Del Gaizo Moore V, Brown JR, Certo M, Love TM, Novina CD, Letai A. Chronic lymphocytic leukemia requires BCL2 to sequester prodeath BIM, explaining sensitivity to BCL2 antagonist ABT-737. J Clin Invest. 2007;117(1):112-121. Del Poeta G, Del Principe MI, Consalvo MA, et al. The addition of rituximab to fludarabine improves clinical outcome in untreated patients with ZAP-70 negative chronic lymphocytic leukemia. Cancer. 2005;104(12):2743-2752. Hallek M, Fisher K, Fingerle-Rowson G, et al. Addition of rituximab to fludarabine and cyclophosphamide in patients with chronic lymphocytic leukemia: a randomized, open-label, phase 3 trial. Lancet. 2010;376(9747):1164-1174. Cuneo A, Marchetti M, Marosi G, et al. Appropriate use of bendamustine in firstline therapy of chronic lymphocytic leukemia. Recommendations SIE; SIES, GITMO Group. Leuk Res. 2014;38(11):1269-1277. Korsmeyer SJ. Bcl-2 initiates a new category of oncogenes regulators of cell death. Blood. 1992;80(4):879-886. Del Poeta G, Venditti A, Del Principe MI, et al. Amount of spontaneous apoptosis detected by Bax/Bcl-2 ratio predicts outcome in acute myeloid leukemia (AML). Blood. 2003;101(6):2125-2131. Wolter KG, Hsu Y-T, Smith CL, Nechushtan A, Xi X-G, Youle RJ. Movement of bax from cytosol to mitochondria during apoptosis. J Cell Biol. 1997;139(5):1281-1292. Del Poeta G, Del Principe MI, Maurillo L, et al. Spontaneuos apoptosis and proliferation detected by BCL-2 and CD71 proteins are important progression indicators within ZAP-70 negative chronic lymphocytic leukemia. Leuk Lymphoma. 2010;51(1):95106. Del Poeta G, Maurillo L, Venditti A, et al. Clinical significance of CD38 expression in

haematologica | 2016; 101(1)

chronic lymphocytic leukemia. Blood. 2001;98(9):2633-2639. 21. Del Principe MI, Del Poeta G, Buccisano F, et al. Clinical significance of ZAP-70 protein expression in B-cell chronic lymphocytic leukemia. Blood. 2006;108(3):853-861. 22. Bomben D, Dal Bo M, Zucchetto A, et al. Mutational status of IgV(H) genes in B-cell chronic lymphocytic leukemia and prognosis: percent mutations or antigen-driven selection. Leukemia. 2005;19(8):1490-1492. 23. Degan M, Bomben R, Dal Bo M, et al. Analysis of IgV gene mutations in B cell chronic lymphocytic leukaemia according to antigen-driven selection identifies subgroups with different prognosis and usage of canonical somatic hypermutation machinery. Br J Haematol. 2004;126(1):2942. 24. Rossi D, Rasi S, Spina V, et al. Integrated mutational and cytogenetic analysis identifies new prognostic subgroups in chronic lymphocytic leukemia. Blood. 2013; 121(8):1403-1412. 25. Rossi D, Cerri M, Deambrogi C, et al. The prognostic value of TP53 mutations in chronic lymphocytic leukemia is independent of del17p13: implications of overall survival and chemorefractoriness. Clin Cancer Res. 2009;15(3):995-1004. 26. Cheson BD, Bennett JM, Grever M, et al. National Cancer Institute Sponsored Working Group Guidelines for chronic lymphocytic leukemia: revised guidelines for diagnosis and treatment. Blood. 1996;87(12):4990-4997. 27. Hallek M, Cheson BD, Catovsky D, et al. International Workshop on Chronic Lymphocytic Leukemia. 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):54465456. 28. Saka B, Atkan M, Sami U, Oner D, Sanem O, Dincol G. Prognostic importance of soluble CD23 in B-cell chronic lymphocytic leukemia. Clin Lab Haematol. 2006; 28(1):30-35. 29. Binet JL, Plunkett W, Robertson B, et al. What does apoptosis mean in CLL? Leuk Lymphoma. 1996;Suppl 2:47-52. 30. Robertson LE, Plunkett W, McConnel M, Keating MJ, McDonnel TJ. Bcl-2 expression in chronic lymphocytic leukemia is associated with a progressive pattern of disease. Leukemia. 1996;10(3):456-459. 31. Gottardi D, Alfarano A, De Leo AM, et al. In leukaemic CD5+ B cells the expression of BCL-2 gene family is shifted toward protection from apoptosis. Br J Haematol. 1996;94(4):612-618. 32. Kitada S, Andersen J, Akar S, et al. Expression of apoptosis-regulating proteins in chronic lymphocytic leukemia: correlations with in vitro and in vivo chemoresponses. Blood. 1998;91(9):3379-3389. 33. Saxena A, Viswanathan S, Moshynska O, Tandon P, Sankaran K, Sheridan DP.Mcl-1 and Bcl-2/Bax ratio are associated with treatment response but not with Rai stage in B-cell chronic lymphocytic leukemia. Am J Hematol. 2004;75(1):22-33.

34. Damle RN, Wasil T, Fais F, et al. Ig V gene mutation status and CD38 expression as novel prognostic indicators in chronic lymphocytic leukemia. Blood. 1999;94(6):18401847. 35. Gattei V, Bulian P, Del Principe MI, et al. Relevance of CD49d protein expression as overall survival and progressive disease prognosticator in chronic lymphocytic leukemia. Blood. 2008;111(2):865-873. 36. Zucchetto A, Benedetti D, Tripodo C, et al. CD38/CD31, the CCL3 and CCL4 chemokines, and CD49d vascular cell adhesion molecule-1 are interchained by sequential events sustaining chronic lymphocytic leukemia cell survival. Cancer Res. 2009;69(9):4001-4009. 37. Buggins AG, Pepper C, Patten PE, et al. Interaction with vascular endothelium enhances survival in primary chronic lymphocytic leukemia cells via NF-kB activation and de novo gene transcription. Cancer Res. 2010;70(19):7523-7533. 38. Fabbri G, Rasi S, Rossi D, et al. Analysis of the chronic lymphocytic leukemia coding genome: role of NOTCH1 mutational activation. J Exp Med. 2011;208(7):1389-1401. 39. Puente XS, Pinyol M, Quesada V, et al. Whole-genome sequencing identifies recurrent mutations in chronic lymphocytic leukaemia. Nature. 2011;475(7354):101105. 40. Del Giudice I, Rossi D, Chiaretti S, et al. NOTCH1 mutations in +12 chronic lymphocytic leukemia (CLL) confer an unfavorable prognosis, induce a distinctive transcriptional profiling and refine the intermediate prognosis of +12 CLL. Haematologica. 2012;97(3):437-441. 41. Koch U, Radtke F. Notch and cancer: a double-edged sword. Cell Mol Life Sci. 2007; 64(21):2746-2762. 42. Vilimas T, Mascarenhas J, Palomero T, et al. Targeting the NF-kappaB signaling pathway in Notch1-induced T-cell leukemia. Nat Med. 2007;13(1):70-77. 43. Faria JR, Yamamoto M, Faria RMD, Kerbauy J, Oliveira JSR. Fludarabine induces apoptosis in chronic lymphocytic leukemia- the role of P53,Bcl-2, Bax, Mcl-1, and Bag-1 proteins. Braz J Med Biol Res. 2006;39(3):327-333. 44. Molica S, Dattilo A, Giulino C, Levato D, Levato L. Increased bcl-2/bax ratio in B-cell chronic lymphocytic leukemia is associated with progressive pattern of disease. Haematologica. 1998;83(12):1122-1124. 45. Zenz T, Krรถber A, Scherer K, et al. Monoallelic TP53 inactivation is associated with poor prognosis in chronic lymphocytic leukemia: results from a detailed genetic characterization with long-term follow-up. Blood. 2008;112(8):3322-3329. 46. Dicker F, Herholz H, Schnittger S, et al. The detection of TP53 mutations in chronic lymphocytic leukemia independently predicts rapid disease progression and is highly correlated with a complex aberrant karyotype. Leukemia. 2009;23(1):117-124. 47. Bellosillo B, Villamor N, Lopez-Guillermo A, et al. Spontaneous and drug-induced apoptosis is mediated by conformational changes of bax and bak in B-cell chronic lymphocytic leukemia. Blood. 2002; 100(5):1810-1816.

ARTICLE EUROPEAN HEMATOLOGY ASSOCIATION

Plasma Cell Disorders

Ferrata Storti Foundation

Risk factors for venous thromboembolism in immunoglobulin light chain amyloidosis

Katherine M. Bever,1,5 Luke I. Masha,2 Fangui Sun,6 Lauren Stern,3,5 Andrea Havasi,3,5 John L. Berk,4,5 Vaishali Sanchorawala,1,5 David C. Seldin,1,5 and J. Mark Sloan1,5

Departments of Hematology/Oncology, 2Medicine, 3Nephrology, and 4Pulmonology, the 5Amyloidosis Center, Boston University School of Medicine; and 6Department of Biostatistics, Boston University School of Public Health, MA, USA 1

Haematologica 2016 Volume 101(1):86-90

ABSTRACT

Correspondence: mark.sloan@bmc.org

Received: 16/7/2015. Accepted: 7/10/2015. Pre-published: 9/10/2015. doi:10.3324/haematol.2015.133900

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/1/86

atients with immunoglobulin light chain amyloidosis are at risk for both thrombotic and bleeding complications. While the hemostatic defects have been extensively studied, less is known about thrombotic complications in this disease. This retrospective study examined the frequency of venous thromboembolism in 929 patients with immunoglobulin light chain amyloidosis presenting to a single referral center, correlated risk of venous thromboembolism with clinical and laboratory factors, and examined complications of anticoagulation in this population. Sixty-five patients (7%) were documented as having at least one venous thromboembolic event. Eighty percent of these patients had events within one year prior to or following diagnosis. Lower serum albumin was associated with increased risk of VTE, with a hazard ratio of 4.30 (CI 1.60-11.55; P=0.0038) for serum albumin less than 3 g/dL compared to serum albumin greater than 4 g/dL. Severe bleeding complications were observed in 5 out of 57 patients with venous thromboembolism undergoing treatment with anticoagulation. Prospective investigation should be undertaken to better risk stratify these patients and to determine the optimal strategies for prophylaxis against and management of venous thromboembolism.

Introduction Immunoglobulin light chain amyloidosis (AL) is a rare systemic disorder resulting from organ deposition of amyloid fibrils comprised of immunoglobulin light chains, typically associated with an underlying clonal plasma cell or lymphoproliferative disorder. AL is known to affect hemostasis with both severe thrombotic and bleeding complications reported.1-5 The predilection for bleeding complications has been attributed to acquired factor X deficiency, acquired von Willebrand syndrome, and vascular fragility due to amyloid fibril deposition in small capillaries.2,4,6-9 The last is primarily responsible for the periorbital ecchymoses considered pathognomonic of the disease. The etiology, risk, and optimal management of thrombotic complications in AL have not been established. The nephrotic syndrome, a common presentation of AL, is associated with hypercoagulability and may contribute to the risk of venous thromboembolism (VTE) in this population.10-15 Other potential contributors to thromboembolic risk include monoclonal gammopathy and plasma cell dyscrasia16 as well as AL treatments, specifically immunomodulatory agents (IMiDs)17-23 and those requiring central venous catheters. A means of risk stratification of VTE in this population is needed to guide prophylaxis and treatment strategies. In this study, we retrospectively examined the frequency of VTE events among patients with AL presenting to a single referral center, and determined specific factors that may influence the risk of VTE in this disease. haematologica | 2016; 101(1)

Risk factors for VTE in AL

Table 1. Base-line clinical characteristics.

Characteristic

Median age at diagnosis in years (range) N. female (%) N. non-white (%) Tobacco history in years, n. (%) None <10 10-30 30-50 >50 Unknown Primary organ involvement years, n. (%) Renal Cardiac Liver Gastrointestinal Nervous system Pulmonary Other/unknown Treatment history years, n. (%) Any IMiD Any ASCT Serum albumin (g/dL) >4 3-4 <3 Urine protein (g/day) ≤1 1-3.5 3.5-8 >8

Table 2. Description of events.

VTE (n=46)

All (n=824)

60 (26-80) 13 (28) 3 (6.5)

61 (26-90) 308 (37) 99 (12)

18 (39) 4 (8.7) 15 (33) 2 (4.3) 0 (0) 7 (15)

388 (47) 45 (5.5) 142 (17) 90 (11) 5 (0.6) 154 (19)

27 (59) 11 (24) 2 (4.3) 3 (6.5) 1 (2.2) 0 (0) 2 (4.3)

392 (48) 225 (27) 34 (4.1) 50 (6) 27 (3.3) 11 (1.3) 85 (10)

17 (37) 29 (63)

151 (18) 286 (35)

21 (35) 20 (33) 5 (8.3)

175 421 228

11 (18) 7 (12) 12 (20) 15 (25)

319 (39) 145 (56) 196 (24) 164 (20)

N: number at risk; VTE: venous thromboembolism; IMiD: immunomodulatory agent; ASCT: autologous stem cell transplantation.

Methods Study population Clinical and laboratory data were collected prospectively on patients presenting to the Amyloid Clinic at the Boston University Medical Center through an Institutional Review Board-approved protocol. All patients gave written informed consent in accordance with the Declaration of Helsinki. Patients typically returned at least annually for re-evaluation and repeat laboratory testing but generally received care at other institutions. Patients with AL who presented between January 2003 and September 2013 were included in this study. Cases of VTE were determined by searching the electronic medical record for the co-occurrence of “AL amyloidosis” along with terms related to VTE, including “thrombosis”, “DVT”, “PE”, and “embolism” in the above patients’ charts. Cases of VTE were verified by chart review. Only VTEs occurring within one year prior to diagnosis or at any time following the diagnosis of AL were included. Cases of arterial thrombosis and embolic cerebrovascular accidents were not included in this analysis. Bleeding events complicating anticoagulation treatments were identified by chart review in patients with confirmed VTE. Significant bleeding events were defined as bleeding requiring hospitalization, fatal bleeding, symptomatic bleeding into a critical area or organ, or bleeding requiring transfusion. haematologica | 2016; 101(1)

Characteristic VTE type DVT PE DVT/PE Portal vein thrombosis Unknown Treatment associated IMiD ASCT Other (bortezomib, melphalan, bendamustine, VAD) No or unknown CVC associated Yes No or unknown

Number 40 23 4 1 3 15 16 11 29 17 54

VTE: venous thromboembolism; DVT: deep venous thrombosis; PE: pulmonary embolism; UTP: urine total protein; IMiD: immunomodulatory agent; ASCT: autologous stem cell transplantation; VAD: vincristine, adriamycin, dexamethasone; CVC: central venous catheter.

Laboratory data from the initial evaluation at the Amyloid Clinic were used in this analysis. Urinary protein excretion was measured by 24-h collection when available. Urine protein to creatinine ratio was used to estimate proteinuria if 24-h collection was not available. Due to modification of the D-dimer assay during the period of study, the D-dimer result was converted to a binary positive or negative result based on the normal range established by the clinical laboratory at the Boston University Medical Center at the time. An intermediate result by the qualitative assay was considered to be positive. Data at the time of VTE were not available as the majority of events happened at other institutions.

Statistical analysis Patients’ characteristics were summarized by median and range for continuous variables, frequency and proportion for categorical variables (Tables 1 and 2). Time-dependent individual risk factors and the set of final models were assessed by Cox's regression model (Table 3). P<0.05 was considered statistically significant. The analysis was conducted by SAS 9.3.

Results Patients' characteristics A total of 929 patients with AL were identified on initial medical record search and included in the descriptive analysis. Of these, 105 patients were excluded from the statistical analysis because of incomplete data, including poor follow up, or because VTE occurred prior to diagnosis. Therefore, 824 patients were included in the final statistical analysis. Base-line characteristics of this group are described in Table 1.

Frequency of and description of events Out of 929 patients who underwent evaluation, 65 patients (7%) had at least one occurrence of VTE less than one year prior to or following diagnosis of AL. Among the 65 patients identified, there were a total of 71 VTEs diagnosed; 40 deep venous thromboses (DVTs), 23 pulmonary emboli (PEs), one portal vein thrombosis, and 4 occurrences of simultaneous DVT and PE. In 3 cases, the type of VTE could not be determined. Seventeen VTEs (23.9%) 87

K.M. Bever et al.

Figure 1. Number of venous thromboembolic events according to time from diagnosis.

occurred in the setting of indwelling central venous catheter (CVC) and 15 (21.1%) occurred within 100 days of IMiD therapy (Table 2). Patients undergoing treatment with IMiD therapy were taking full-dose aspirin prophylaxis or were on therapeutic anticoagulation unless contraindicated. Sixteen events (22.5%) occurred within 100 days of autologous stem cell transplantation (ASCT), of which 6 events were CVC associated. Of the twenty events (29%) that were not in the setting of indwelling CVC or treatment, 14 of these occurred in patients with nephrotic-range proteinuria. Fifty-two patients (80%) had events occurring within one year of diagnosis (Figure 1). VTE preceded the diagnosis of AL in 12 patients. These patients were included in the descriptive analysis on the presumption that the VTE occurred in the presence of undiagnosed disease, but excluded from statistical analysis. An additional 7 patients were excluded from further analysis because they did not return for follow up after the initial visit or because laboratory data were incomplete.

Risk of VTE in AL The results of univariate analysis of clinical and laboratory data and risk of VTE are shown in Table 3. An increasing risk of VTE with lower serum albumin was observed, with hazard ratio of 2.16 for serum albumin of 3-4g/dL (CI 0.80-5.81; P=0.13), and a hazard ratio of 4.30 in patients with serum albumin of less than 3 g/dL (CI 1.60-11.55; P=0.0038) (Figure 2). Increasing proteinuria was also associated with increased risk of VTE, with a hazard ratio of 1.84 for urine protein of 1-3.5 g per day (CI 0.74-4.57; P=0.19), 1.94 for urine protein of 3.5-8 g per day (CI 0.85-4.41, P=0.11), and 2.60 for urine protein of greater than 8 g per day (CI 1.19-5.72, P=0.017). Levels of serum albumin and proteinuria were correlated in this analysis, and thus urine protein was not included in final analysis. Age at diagnosis, sex, non-white race, fibrinogen, Ddimer, and partial thromboplastin time (PTT) were not significantly associated with an increased risk of VTE.

VTE among patients with nephrotic range proteinuria To determine the specific risk factors for VTE in the setting of nephrotic-range proteinuria, we examined the 382 patients with proteinuria of more than 3.5 g per day. In this subgroup, 37 patients (9.7%) had at least one documented episode of VTE within one year prior to or anytime following the diagnosis of AL. Univariate analysis of PTT, fibrinogen, D-dimer, urine protein excretion greater 88

Figure 2. Kaplan-Meier estimates of VTE-free survival for individuals with serum albumin less than 3g/dL (red), greater than or equal to 3g/dL and less than or equal to 4g/dL (green), and serum albumin greater than 4 g/dL (blue).

Table 3. Univariate analysis of risk factors associated with VTE. Age Sex Race White Non-white Partial thromboplastin time Fibrinogen D-dimer Normal Elevated Albumin (g/dL) >4 3-4 Less than 3 Urine protein (g/day) 1 1-3.5 3.5-8 >8

Hazard ratio (95% CI)

0.98 (0.95-1.01) 0.64 (0.34-1.21)

0.12 0.17

1.81 (0.56-5.83) 1.00 1.02 (0.98-1.02) 1.00 (1.00-1.003)

0.32 0.33 0.17

1.00 1.68 (0.93-3.04)

0.086

1.00 2.16 (0.80-5.81) 4.30 (1.60-11.55)

0.13 0.0038

1.00 1.84 (0.74-4.57) 1.94 (0.85-4.41) 2.60 (1.19-5.72)

0.19 0.11 0.017

N: number at risk; CI: confidence interval.

haematologica | 2016; 101(1)

Risk factors for VTE in AL

Table 4. Bleeding events in patients with VTE.

Event

Type of bleeding event

Anticoagulation type

ASCT-related

Transfusion required

Death from bleeding

1 2 3 4 5

Intraperitoneal bleed Intraorbital, GIB Hematuria Lower GIB Lower GIB

UFH drip, thrombolysis Warfarin Warfarin Warfarin Warfarin

Yes No No No No

Yes Unknown Unknown Yes Unknown

Yes No No No Yes

ASCT: autologous stem cell transplantation; GIB: gastrointestinal bleed; UFH: unfractionated heparin.

than 8 g per day, and albumin did not reveal any significant associations with risk of VTE in this subgroup.

Bleeding complications The incidence of bleeding complications among the cohort of patients with a documented VTE was assessed by chart review. Fifty-seven of the 65 patients with documented VTE had a treatment history which could be verified by chart review. Of those, 5 patients were documented as experiencing at least one episode of major bleeding related to anticoagulation (Table 4). In 2 of these cases, transfusion support was required. Of note, one patient was treated with thrombolysis followed by unfractionated heparin drip for high clot-burden pulmonary emboli on day 25 following ASCT; he developed intraperitoneal hemorrhage 21 days after thrombolysis leading to PEA arrest and death.

Discussion Venous thromboembolism and its treatment are associated with significant morbidity and mortality in AL. A better understanding of how often VTEs occur and in which patients is of extreme importance in guiding clinicians in the optimal management of these patients. Venous thromboembolism occurred in 7% of AL patients presenting to our center, and the risk of VTE was strongly associated with low serum albumin. The frequency of VTE observed in this study is consistent with what has previously been described in the nephrotic syndrome associated with membranous nephropathy11,12 and with what has been reported in AL and multiple myeloma.5,24,25 Interestingly, no renal vein thromboses were documented in this group; this may reflect the often clinically silent nature of these events, which have been observed in as many as 40-50% of patients with nephrotic syndrome,15 but were not screened for in our population. The majority of observed events occurred during treatment of the disease with chemotherapy, IMiDs, ASCT, or in the setting of an indwelling CVC. Among the events unrelated to treatment, 14 out of 20 occurred in patients with nephrotic-range proteinuria. The hypercoagulable state of nephrotic syndrome is thought to be multifactorial and, in part, a result of the imbalance of antithrombotic and prothrombotic factors involved in the coagulation cascade.26-29 The association between low serum albumin and risk of VTE has been previously described in nephrotic syndrome.11,12,14,30 Serum albumin is inversely correlated with the degree of urinary protein loss, and may be a surrogate marker for loss of endogenous antithrombotic proteins such as antithrombin III and plasminogen. Serum albumin levels may have direct effects on hemostasis and platelet aggregation26 and in addition levels may be decreased in critical illness and haematologica | 2016; 101(1)

thus correlate with more advanced disease. Low serum albumin should be a consideration in risk-stratifying patients during treatment of AL and further prospective studies are warranted to determine the benefits of prophylactic anticoagulation in this high-risk group. Given the high rate of bleeding complications in these patients, prophylactic anticoagulation cannot be recommended without a prospective randomized trial. In particular, the role of novel oral anticoagulants should be investigated given the potential better bleeding profile of these agents. Measures of the coagulation cascade or fibrinolysis such as PTT, fibrinogen and D-dimer were not associated with risk of VTE in this study. This may in part be explained by the fact that these data were collected at the time of initial presentation and not at the time of VTE diagnosis. Data on antithrombin III and factor VIII levels were not available at the time of this analysis, but have been demonstrated in some studies to correlate with risk of VTE.29,31 Prospective collection of these data may have value in further risk stratifying patients. The vast majority of documented VTEs in our cohort occurred within one year of diagnosis. Although this may be in part due to the median follow up of 2.3 years, it is consistent with reports of other studies of VTE in nephrotic syndrome patients11,14 and in MGUS/myeloma patients.16 The etiology of this increased incidence closer to the time of diagnosis is uncertain, but may reflect an increased risk of VTE associated with various treatments of the disease or a decrease in hypercoagulable state with treatment of the underlying disease. This may have therapeutic implications, particularly in decisions regarding duration of anticoagulation. This study is complimentary to others examining the incidence of VTE in AL patients, and, to our knowledge, is the first to correlate risk with serum albumin in this disease state. Our study was strengthened by the large number of patients with AL presenting to our center and the breadth of laboratory and clinical data available on these patients. Limitations of the study include its retrospective design and reliance on clinician documentation of venous thromboembolic events. The majority of patients presented to our center for consultation and received treatment elsewhere, and details of the circumstances surrounding the event, such as disease-specific treatment, immobility, surgery and infections, were therefore limited. The analysis was limited by the duration of patient follow up, which varied widely from less than one month to 13.2 years in the entire cohort. Median follow up was longer in the VTE cohort (2.36 years) than the controls (9.72 months); therefore, the observed incidence of VTE may be underestimated. Laboratory data were taken from the initial visit, and may not reflect the state of the patient at the time of VTE. In addition, complete data on treatment and 89

K.M. Bever et al.

response in these patients were not available to correlate with the occurrence of VTE. In summary, VTE is a frequent complication of AL and the majority of VTEs occurred within one year of diagnosis and most often in the setting of treatment of the disease. Decisions regarding treatment duration should take this into consideration. Serum albumin is strongly correlated with risk of VTE in patients with AL and may be a consideration in determining the optimal treatment for these patients. Further prospective trials are needed to determine the risk/benefit ratio of prophylactic anticoagulation, as major bleeding complications in AL patients on

References 1. Falk RH, Comenzo RL, Skinner M. The systemic amyloidoses. N Engl J Med. 1997;337(13):898-909. 2. Freeman B, Sloan JM, Seldin DC, Cowan AJ, Ruberg FL, Sanchorawala V. Multiple arterial and venous thromboembolic complications in AL amyloidosis and cardiac involvement: a case report and literature review. Amyloid. 2012;19(3):156-160. 3. Gamba G, Montani N, Anesi E, et al. Clotting alterations in primary systemic amyloidosis. Haematologica. 2000;85(3):289-292. 4. Yood RA, Skinner M, Rubinow A, Talarico L, Cohen AS. Bleeding manifestations in 100 patients with amyloidosis. Jama. 1983;249(10):1322-1324. 5. Halligan CS, Lacy MQ, Vincent Rajkumar S, Dispenzieri A, Witzig TE, Lust JA, et al. Natural history of thromboembolism in AL amyloidosis. Amyloid. 2006;13(1):31-36. 6. Skinner M, Talarico L, Cohen AS, St John A. The calcium-dependent interaction of blood clotting factors with primary amyloid fibrils. In: Glenner G, Costa P, de Freitas A, eds. Amyloid and amyloidosis. Amsterdam: Excerpta Medica, 1980:361365. 7. Kos CA, Ward JE, Malek K, et al. Association of acquired von Willebrand syndrome with AL amyloidosis. Am J Hematol. 2007;82(5):363-367. 8. Furie B, Greene E, Furie BC. Syndrome of acquired factor X deficiency and systemic amyloidosis in vivo studies of the metabolic fate of factor X. N Engl J Med. 1977;297(2):81-85. 9. Furie B, Voo L, McAdam KP, Furie BC. Mechanism of factor X deficiency in systemic amyloidosis. N Engl J Med. 1981;304(14):827-830. 10. Christiansen CF, Schmidt M, Lamberg AL, et al. Kidney disease and risk of venous thromboembolism: a nationwide population-based case-control study. J Thromb Haemost. 2014;12(9):1449-1454. 11. Barbour SJ, Greenwald A, Djurdjev O, et al. Disease-specific risk of venous throm-

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

anticoagulation were observed, including death in one patient receiving thrombolytics. Acknowledgments We gratefully acknowledge our colleagues in the Amyloidosis Center and Clinical Trials Office at Boston University School of Medicine and the Solomont Center for Hematology and Medical Oncology at Boston Medical Center who assisted with the evaluation and care of these patients. This research was supported by the Amyloid Research Fund and the Gruss Foundation. We particularly express our gratitude for the late Dr Seldinâ&#x20AC;&#x2122;s contribution to this paper and extraordinary work in the field of amyloidosis.

boembolic events is increased in idiopathic glomerulonephritis. Kidney Int. 2012;81(2): 190-195. Lionaki S, Derebail VK, Hogan SL, et al. Venous thromboembolism in patients with membranous nephropathy. Clin J Am Soc Nephrol. 2012;7(1):43-51. Kayali F, Najjar R, Aswad F, Matta F, Stein PD. Venous thromboembolism in patients hospitalized with nephrotic syndrome. Am J Med. 2008;121(3):226-230. Mahmoodi BK, ten Kate MK, Waanders F, et al. High absolute risks and predictors of venous and arterial thromboembolic events in patients with nephrotic syndrome: results from a large retrospective cohort study. Circulation. 2008;117(2):224-230. Kerlin BA, Ayoob R, Smoyer WE. Epidemiology and pathophysiology of nephrotic syndrome-associated thromboembolic disease. Clin J Am Soc Nephrol. 2012;7(3):513-520. Kristinsson SY, Fears TR, Gridley G, et al. Deep vein thrombosis after monoclonal gammopathy of undetermined significance and multiple myeloma. Blood. 2008;112(9):3582-3586. Bennett CL, Angelotta C, Yarnold PR, et al. Thalidomide- and lenalidomide-associated thromboembolism among patients with cancer. Jama. 2006;296(21):2558-2560. Cavo M, Zamagni E, Cellini C, et al. Deepvein thrombosis in patients with multiple myeloma receiving first-line thalidomidedexamethasone therapy. Blood. 2002;100(6):2272-2273. Knight R, DeLap RJ, Zeldis JB. Lenalidomide and venous thrombosis in multiple myeloma. N Engl J Med. 2006;354(19):2079-2080. Osman K, Comenzo R, Rajkumar SV. Deep venous thrombosis and thalidomide therapy for multiple myeloma. N Engl J Med. 2001;344(25):1951-1952. Zangari M, Anaissie E, Barlogie B, et al. Increased risk of deep-vein thrombosis in patients with multiple myeloma receiving thalidomide and chemotherapy. Blood. 2001;98(5):1614-1615. Menon SP, Rajkumar SV, Lacy M, Falco P,

23.

24.

25.

26.

27. 28.

29.

30.

31.

Palumbo A. Thromboembolic events with lenalidomide-based therapy for multiple myeloma. Cancer. 2008;112(7):1522-1528. Urbauer E, Kaufmann H, Nosslinger T, Raderer M, Drach J. Thromboembolic events during treatment with thalidomide. Blood. 2002;99(11):4247-4248. Srkalovic G, Cameron MG, Rybicki L, Deitcher SR, Kattke-Marchant K, Hussein MA. Monoclonal gammopathy of undetermined significance and multiple myeloma are associated with an increased incidence of venothromboembolic disease. Cancer. 2004;101(3):558-566. Srkalovic G, Cameron MG, Deitcher SR, Kattke-Marchant K, Hussein MA. Incidence and risk factors of venous thromboembolism (VTD) in patients with amyloidosis. Int Semin Surg Oncol. 2005;2:17. Yoshida N, Aoki N. Release of arachidonic acid from human platelets. A key role for the potentiation of platelet aggregability in normal subjects as well as in those with nephrotic syndrome. Blood. 1978;52(5): 969-977. Loscalzo J. Venous thrombosis in the nephrotic syndrome. N Engl J Med. 2013;368(10):956-958. Barbano B, Gigante A, Amoroso A, Cianci R. Thrombosis in nephrotic syndrome. Semin Thromb Hemost. 2013;39(5):469476. Kauffmann RH, de Graeff J, de la Riviere GB, van Es LA. Unilateral renal vein thrombosis and nephrotic syndrome. Report of a case with protein selectivity and antithrombin III clearance studies. Am J Med. 1976;60(7):1048-1054. Kuhlmann U, Steurer J, Bollinger A, Pouliadis G, Briner J, Siegenthaler W. [Incidence and clinical significance of thromboses and thrombo-embolic complications in nephrotic syndrome patients]. Schweiz Med Wochenschr. 1981;111(2728):1034-1040. Cherng SC, Huang WS, Wang YF, Yang SP, Lin YF. The role of lung scintigraphy in the diagnosis of nephrotic syndrome with pulmonary embolism. Clin Nucl Med. 2000;25(3):167-172.

haematologica | 2016; 101(1)

ARTICLE

Cell theraphy & Immunotherapy

Impact of immunosuppressive drugs on the therapeutic efficacy of ex vivo expanded human regulatory T cells

EUROPEAN HEMATOLOGY ASSOCIATION

Ferrata Storti Foundation

Cristiano Scottà,1 Giorgia Fanelli,1 Sec Julie Hoong,1 Marco Romano,1,2 Estefania Nova Lamperti,1 Mitalee Sukthankar,1 Giuliana Guggino,3 Henrieta Fazekasova,1 Kulachelvy Ratnasothy,1 Pablo D. Becker,1 Behdad Afzali,1,4 Robert I. Lechler,1 and Giovanna Lombardi1 Immunoregulation Laboratory, Division of Transplantation Immunology & Mucosal Biology, MRC Centre for Transplantation, King's College London, Guy's Hospital, UK; 2 Department of Experimental, Diagnostic and Specialty Medicine, Institute of Hematology “L. & A. Seràgnoli”, University of Bologna, Italy; 3Dipartimento di Biopatologia e Biotecnologie Mediche, University of Palermo, Italy; and 4Lymphocyte Cell Biology Section, Molecular Immunology and Inflammation Branch, National Institutes of Arthritis, and Musculoskeletal and Skin Diseases, National Institutes of Health, Bethesda, MD, USA 1

Haematologica 2016 Volume 101(1):91-100

ABSTRACT

mmunosuppressive drugs in clinical transplantation are necessary to inhibit the immune response to donor antigens. Although they are effective in controlling acute rejection, they do not prevent long-term transplant loss from chronic rejection. In addition, immunosuppressive drugs have adverse side effects, including increased rate of infections and malignancies. Adoptive cell therapy with human Tregs represents a promising strategy for the induction of transplantation tolerance. Phase I/II clinical trials in transplanted patients are already underway, involving the infusion of Tregs alongside concurrent immunosuppressive drugs. However, it remains to be determined whether the presence of immunosuppressive drugs negatively impacts Treg function and stability. We tested in vitro and in vivo the effects of tacrolimus, mycophenolate and methylprednisolone (major ISDs used in transplantation) on ex vivo expanded, rapamycin-treated human Tregs. The in vitro results showed that these drugs had no effect on phenotype, function and stability of Tregs, although tacrolimus affected the expression of chemokine receptors and IL-10 production. However, viability and proliferative capacity were reduced in a dose-dependent manner by all the three drugs. The in vivo experiments using a humanized mouse model confirmed the in vitro results. However, treatment of mice with only rapamycin maintained the viability, function and proliferative ability of adoptively transferred Tregs. Taken together, our results suggest that the key functions of ex vivo expanded Tregs are not affected by a concurrent immunosuppressive therapy. However, the choice of the drug combination and their timing and dosing should be considered as an essential component to induce and maintain tolerance by Treg.

Introduction Naturally occurring regulatory T cells (Tregs) play a critical role in maintaining peripheral tolerance to self-antigens and controlling autoimmune disease. They are also important to limit immune responses to foreign antigens such as alloantigens. As shown in experimental models, adoptive transfer of Tregs can ameliorate autoimmune diseases, graft-versus-host disease (GvHD) and also prevent solid haematologica | 2016; 101(1)

Correspondence: cristiano.scotta@kcl.ac.uk

Received: 13/4/2014. Accepted: 9/10/2015. Pre-published: 15/10/2015. doi:10.3324/haematol.2015.128934

Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/101/1/91

C. Scottà et al.

organ transplant rejection.1 More recently, we and others have demonstrated that adoptively transferred human Tregs can protect in vivo from human skin, vessel and islet transplant rejections.1-4 Adoptive cell therapy with human Tregs for the treatment of autoimmune diseases and for the induction of transplantation tolerance represents a promising strategy. Indeed, clinical trials using this approach have already demonstrated that Tregs are safe in the treatment of GvHD and Type 1 diabetes.5 We have recently identified a clinically applicable protocol for the expansion of human immunomagnetic beadseparated CD4+CD25+ Tregs.6 We have shown that the expansion of Tregs using rapamycin provides a new and refined approach for large-scale generation of functionally potent and phenotypically stable regulatory T cells, rendering them safe for clinical use in the settings of inflammatory diseases.6,7 Our group is currently involved in a multi-center phase I/II study to investigate the safety of infusing ex vivo expanded Tregs in solid organ transplantation (the ONE Study, funded by the European Union FP7 program). In this trial, Tregs expanded in vitro are being injected in kidney transplant patients receiving concurrent immunosuppressive drugs (ISDs). Immunosuppressive drugs in clinical transplantation are necessary to prevent acute graft rejection. Various compounds have been selected and applied according to their ability to control lymphocyte activation.8 However, despite the use of ISDs most of the transplants fail within ten years due to chronic allograft dysfunction.9 In addition, ISDs are linked to morbidity and mortality.8,10 Conventional immunosuppressive therapy employs drugs such as calcineurin inhibitors, e.g. cyclosporine (CsA) or tacrolimus (TAC), as well as mycophenolate mofetil (MMF), an ester prodrug of the immunosuppressant mycophenolic acid (MPA) and corticosteroids such as methyl-prednisolone (mPr). More recently, many transplant centers have started to investigate rapamycin (RAPA)-based immunosuppression. The precise mechanism of action of these drugs is well characterized and different reports have described their effect on different cell subsets. ISDs are part of clinical protocols that accompany the adoptive transfer of Tregs. For example, the combination of TAC, MMF and mPr is part of the clinical protocol in the One Study. There is, therefore, the need to investigate the influence of ISDs on the viability, function and stability of in vitro expanded and adoptively transferred human Tregs. Current evidence suggests that calcineurin inhibitors have markedly negative effects on circulating Tregs, mainly by interfering with IL-2 production and therefore affecting Treg function and survival.11,12 Conversely, corticosteroids have been reported to increase Treg frequency and FOXP3 expression in patients with autoimmune diseases.13,14 Finally, renal transplant recipients receiving MMF show significantly higher CD4+CD25hiFOXP3+ Tregs compared to patients on other treatments.12 Furthermore, studies on liver transplant patients with renal impairment on CsA and converted to MMF showed that the new treatment may reverse the negative effect of calcineurin inhibitors on Tregs.15 This study has focused on the consequences of using drugs commonly applied in the treatment of transplant patients, namely TAC, MPA and mPr, for the viability, 92

function and stability of ex vivo expanded Tregs. Their effects were compared to the in vivo treatment with RAPA.

Methods Cell isolation and separation Peripheral blood mononuclear cells (PBMC) from healthy donors were obtained from anonymized human leukocyte cones supplied by the National Blood Transfusion Service (NHS blood and transplantation, Tooting, London, UK). Human studies were conducted in accordance with the Declaration of Helsinki and approved by the Institutional Review Board of Guy’s Hospital (reference 09/H0707/86). Informed consent was obtained from all healthy donors prior to enrolment into the study. PBMC were isolated by lymphocyte (PAA) density gradient centrifugation. CD4+CD25+T cells were purified by negative selection of CD4+T cells followed by positive selection of CD25+T cells using miniMACS CD4+CD25+T Regulatory Cell Isolation Kit (MiltenyiBiotec) with a purity of 90-98%.

Human Treg-line generation Human Tregs were expanded ex vivo with RAPA, as previously described.6 In brief, human CD4+CD25+T cells were plated at 1x106/mL in X-Vivo15 (Lonza), containing rapamycin (100nM; LCLaboratories). Cells were activated with anti-CD3/CD28 beads (ratio bead:cell of 1:2; Invitrogen, UK). IL-2 (1000 IU/mL; Proleukin-Novartis) was added at day 4 post activation and replenished every two days. Cells were re-stimulated every 10-12 days and used after 36 days from the first activation (3 rounds of stimulation).

Antibodies for flow cytometry Stained single-cell suspension were analyzed with: CD3-FITC (OKT4), CD4-FITC/APC (OKT4) from Sigma-Aldrich; CD62LFITC (Greg-56; Invitrogen); CD25-PE (4E3), CD152-PE (14D3), CD127-PE-Cy7 (eBioRDR5), GITR-APC (eBioAITR), ICOS-PECy7 (ISA-3), CD39-FITC (eBioA1), IL-10-APC (JES3-9D7), IFN-γ-FITC (4S.B3), IL-17-PE (eBio64DEC17) and FOXP3-eFluor 660 (PCH101) from eBioscience; CD27-PE (M-T271) and Integrin β7-PE-Cy5 (FIB504) from BD-Bioscience, UK; HLA-DR-APC (L243), CLA-FITC (HECA-452) and CCR4-PerCP-Cy5.5 (TG6/CCR4) from Biolegend. Appropriate isotype controls from mouse or rat were used. Prior to use, all antibodies were titrated using normal resting or activated PBMC to establish optimal staining dilutions.

Treg culture with immunosuppressants TAC, MPA and mPr (Sigma-Aldrich) for the in vitro study were diluted in DMSO and used at different concentrations (see Results for details). To test their effect on Tregs, cells were activated with anti-CD3/CD28 beads (ratio 1:2), IL-2 (20 IU/mL) in X-Vivo15 in presence of drugs. Cell viability was measured by LIVE/DEAD®kit. Proliferation was evaluated by cell-count, Ki67 staining, CFDA-SE and CellTrace™ Violet (CTV) proliferation kit.

Xenogeneic GvHD and in vivo stability assay

Balb-c RAG–/–γc–/– mice were used between 8-14 weeks of age. Mice were maintained under specific pathogen-free conditions and handled in accordance with the Institutional Committees on Animal Welfare of the United Kingdom Home-Office (Home-Office Animals Scientific Procedures Act, 1986; reference PPL 70/7302). In order to establish whether injected Tregs could maintain a stable phenotype in the presence of immunosuppressive regimen, a XenoGvHD model was induced by intravenous transfer of 1x107 HLAhaematologica | 2016; 101(1)

Impact of immunosuppressive drugs on Treg function A2- PBMCs depleted of CD25+cells into Balb-c RAG–/–γc–/–mice. Animals were monitored 3-times weekly for body weight, other GvHD symptoms (hunched back, fur loss, skin inflammation), and assessment of human CD45+cell engraftment. Immunosuppressive treatment using drugs for infusion in transplantation therapy was administered by intra-peritoneal injection when animals were reconstituted with approximately 10% human CD45+ cells and daily during the next three days. ISDs were used as previously described: 16-18 TAC 2 mg/Kg (Prograf, Astellas), MPA 100 mg/Kg (CellCept, Roche), mPr 20 mg/Kg (Solu-Medrone, Pharmacia) and RAPA 300 mg/Kg (LC-Laboratories). HLA-A2+Tregs (1x107) were introduced into this environment after one day of drug administration. Spleen and lymph nodes were harvested three days after Treg transfer. Tregs were recovered by cell sorting on a 3-laser FACS-ARIA highspeed cell sorter (BD Biosciences). Additional information is available in the Online Supplementary Appendix.

Results Treg viability is affected by the presence of immunosuppressive drugs in vitro To determine the direct effect of immunosuppressive drugs (ISDs) on ex vivo expanded Tregs, CD4+CD25+ T cells were purified by magnetic bead separation and cultured in vitro with anti-CD3/CD28 beads, IL-2 (100 U/mL) and rapamycin (RAPA), as previously published.6 After 3 rounds of stimulation and 36 days in culture (GMP compliant protocol for the generation of Treg lines used in the

B B

Figure 1. Dose-response curves to derive the concentration of immunosuppressant. Ex vivo expanded and rapamycin-treated Tregs were activated with antiCD3/CD28 coated beads (bead:cell ratio = 1:2) and IL-2 (20 IU/mL) for three days in the presence of different concentration of drugs. (A-C) Concentration of TAC, MPA and methyl-prednisolone, respectively, leading to 50% of Treg viability. Data are combined from 4 independent experiments.

haematologica | 2016; 101(1)

Figure 2. Proliferative capacity of Tregs in the presence of immunosuppressants. (A) Total number of ex vivo expanded and rapamycin-treated Tregs activated with anti-CD3/CD28 coated beads (bead:cell ratio = 1:2) and IL-2 (20 IU/mL) and cultured for five days with medium (CTR), single immunosuppressive agent at LD50 (TAC, MPA and mPR) or their combination (IS Mix). Data are combined from 7 independent experiments. (B) Percentages of Tregs determined by CTV dilution profile with more than three cycles of cell division cultured in the same conditions described above. Data are from 3 independent experiments and presented as mean ± SD. *P<0.05, **P<0.01; ***P<0.001.

C. ScottĂ et al.

One Study), Tregs were activated with anti-CD3/CD28 beads and 20 IU/mL of IL-2 in the presence of tacrolimus (TAC), mycophenolate (MPA) and methylprednisolone (mPr). Variability in pharmacokinetics and pharmacodynamics makes the therapeutic concentrations of these drugs indicative only in clinical practice.19 Cross-sectional studies have revealed that trough levels may not correspond to the actual active concentrations on cells. Indeed, in blood, TAC is mainly associated with erythrocytes (>70%) and plasma proteins (>20%), and only a small fraction with lymphocytes (about 1%).19 MPA is mainly bound to serum albumin (>97%)20,21 and only the free fraction is thought to be responsible for the immunosuppressive effect.20,21 Other data showed that renal function, serum albumin concentrations, hemoglobin levels and immunosuppressive co-medication, such as TAC, are factors contributing to the variability and time-dependent pharmacokinetics of drug treatments.22,23 In this study, in which we aimed to translate in vitro the doses of ISDs used in clinical protocols, we tested a range of concentrations (TAC: 0.08-20 ng/mL; MPA: 0.08-2.5 Âľg/mL; mPr: 0.015-4 mg/mL) including both maximum and minimum therapeutic levels of ISDs, described in transplanted patients.24-26 The doses leading to 50% of Treg viability (LD50) (2 ng/mL, 1 mg/mL and 0.5 mg/mL for TAC, MPA and mPr, respectively) were selected and used for the in vitro analysis (Figure 1A-C). A significant reduction of cell proliferation was observed when Tregs were cultured for five days with either a single drug or the mixture of the three drugs (IS Mix), at the established LD50 (Figure 2A). Further analysis revealed that the number of cells going to division was affected mostly by TAC, MPA and by the IS Mix (Figure 2B).

The suppressive activities of in vitro expanded Tregs is maintained in the presence of immunosuppressive drugs The maintenance of the suppressive ability and stability of Tregs are relevant to the success of Treg therapy. Therefore, we investigated the suppressive ability of Tregs in the presence of ISDs. Tregs and CFSE-labeled responder T cells (Teff) were activated and co-cultured in the presence of LD50 dose of TAC, MPA and mPr in isolation or in combination for five days. Results in Figure 3 show that ISDs did not have any negative effect on the suppressive capacity of Tregs and MPA, mPr and IS Mix, but rather strengthened it (1:10 ratio). To further investigate the effect on Tregs expanded with drugs, Treg lines were cultured for five days with the LD50 dose of ISDs, then washed and co-cultured with activated CFSE-labeled Teff. We observed that the suppressive ability was maintained even in these conditions (Online Supplementary Figure S1).

Immunosuppressive treatments have no effect on the expression of Treg functional molecules but affect chemokine receptors Tregs have been described to express both surface and cytoplasmic molecules influencing their effector functions.27 Furthermore, the expression of chemokine receptors by Tregs can confer to these cells specific homing abilities. Treg lines (previously expanded in the presence of rapamycin) were cultured for five days in the presence of TAC, MPA, and mPr in isolation or in combination. A complete analysis of these markers on Tregs revealed that while the expression of most of the molecules was not modified by the drug treatments (Figure 4A and B), the level of expression of CD25 (MFI) on Tregs was reduced by the treatment with TAC (Figure 4B and C).

Figure 3. Effect of immunosuppressants on Treg suppressive ability. Proliferative responses (at 5 days) of Teff activated with anti-CD3/CD28 coated beads + IL-2 (20 IU/mL) and cultured alone or in co-culture with Tregs and different immunosuppressive agents. Proliferation of Teff, alone or in co-culture with Treg, was determined by CFSE dilution profile. Data are from 5 independent experiments. **P<0.01; ***P<0.001.

haematologica | 2016; 101(1)

Impact of immunosuppressive drugs on Treg function

Looking at the expression of chemokine receptors, we once again found that TAC reduced the expression of CCR4 and CLA (related to skin-homing) while the expression of integrin β7 (gut-homing) was up-regulated (Figure 4D).

ISDs maintain the stability of Tregs while IL-10 production is reduced by TAC We and others have previously shown that RAPA can selectively favor the expansion of Tregs, increase IL-10 production and reduce the frequency of cells releasing proinflammatory cytokines such as IL-17 and IFN-γ.6,28 To investigate the effect of drugs on the cytokine profile, Tregs were stimulated with anti-CD3/CD28 beads and IL2 in the presence of TAC, MPA and mPr in isolation or in combination. After five days, culture supernatants were harvested and tested for the presence of cytokines. The amount of cytokine was normalized to the total number of live cells at the end of culture. TAC and mPr decreased the production of IL-17 and IFN-γ (Figure 5A and B) while the presence of TAC reduced IL-10 production by Tregs (Figure 5C) suggesting that TAC may affect the regulatory function of Tregs.

ISDs reduce the proliferative ability of adoptively transferred Tregs in a humanized mouse model To extend the in vitro observations to an in vivo setting, we tested the effect of a concurrent immunosuppressive therapy on ex vivo expanded adoptively transferred RAPA-

treated Treg preparations in a humanized mouse model. Immunodeficient mice (RAG2–/–yc–/–mice) were injected with human HLA-A2- PBMC (1x107) to create an inflammatory environment in vivo.6 After 2-3 weeks, and long before the first signs of GvHD, one group of mice received daily intraperitoneal injection of IS Mix (TAC+mPr+MPA), for four days. Two other groups of mice received either the Treg permissive drug RAPA alone, 18 or the combination of RAPA and TAC, as many transplant centers have investigated sirolimus (RAPA)-based immunotherapies.29 Another group of mice received saline solution only. Twenty-four hours after the first injection of immunosuppressive drugs, human HLA-A2+ Tregs (1x107) were injected into the different groups of mice. After three additional days, mice were culled. Although in this setting mice did not show any sign of xeno-GvHD (e.g. ruffled fur, hunched posture), differences were observed in the size of the spleens between the four groups. Animals injected with drugs showed smaller spleens and a significant reduced number of total splenocytes compared to the control group (Online Supplementary Figure S3). The analysis of human/mouse chimerism in the spleens showed similar levels of human cell engraftment in all the groups of mice although the group receiving RAPA+TAC tended to be the lowest (Figure 6A). The average engraftment (±SD) in the control group was 51±22%: 54±16% in the IS Mix group, 45±8% in the RAPA group, and treated mice in the RAPA+TAC group 26±35% (Figure 6A). There was no difference in the percentage and the average number of adop-

Figure 4. Phenotype of Tregs cultured with immunosuppressive agents. (A) Expression of Treg markers on cells treated for five days with medium (CTR), single immunosuppressive agent (TAC, MPA and mPR) or their combination (IS Mix). Data are from 5 independent experiments. (B) FOXP3 and CD25 expression on Tregs from 2 healthy donors (HD1, HD2). Data are representative of 9 different Treg preparations cultured five days with either medium (CTR) or with the combination of 3 drugs (IS Mix). (C) Cumulative data showing CD25 expression (MFI) on untreated and drug-treated Tregs. (D). Expression of a panel of chemokine receptors on Tregs cultured in the same conditions described above. Data are from 5 independent experiments. **P<0.01; ***P<0.001.

haematologica | 2016; 101(1)

C. Scottà et al.

tively transferred Tregs, detected within the total human CD45+ cells, by gating on HLA-A2 expression, between the four groups of mice, although a very high variability was observed (Figure 6B). The average engraftment of Tregs was 25±32% in the control group, 13±23% in the IS Mix group, 27±25% in the RAPA group, and 19±17% in the RAPA+TAC group (Figure 6B). These data suggest that the concurrent in vivo treatments with ISDs had no effect on Treg engraftment (Figure 6B). Next, the number of dividing Tregs was investigated to understand whether any of the ISDs have affected the proliferation of Tregs in vivo. This was based on the possibility that ISDs, by altering the inflammatory environment and inhibiting, for example, IL-2 production, could have influenced the survival and proliferation of injected Tregs. As expected, we observed a significant difference in the number of proliferating non-Treg cells (HLA-A2-) between the control group (48±12%) and the three groups of mice treated with ISDs (IS Mix, 26±8%; RAPA, 22±11%; RAPA+TAC, 12±6%). However, the analysis of cycling Tregs (Ki67+ cells) showed that the IS Mix (4±1%) and RAPA+TAC (10±1%) treatments, but not RAPA alone, have decreased the number of proliferating cells in comparison to the control group (19±5%). Interestingly, the concurrent treatment with RAPA alone did not affect the division of Tregs (15±3%) (Figure 6C).

ISDs do not modify the phenotype, stability and function of adoptively transferred Tregs The ability of adoptively transferred Tregs to maintain the phenotype, cytokine profile and suppressive function in vivo in an inflammatory environment was analyzed in the humanized mouse model described above. The analysis of the phenotype of the cells recovered from the spleens of the three groups of mice treated with the ISDs showed that the expression of FOXP3, CD25 and CTLA4 molecules were maintained compared to the control group (Figure 7A). The analysis of chemokine receptors showed only a marginal increase in the number of Integrin-β7+ Tregs in the drug-treated groups but the differences did not reach statistical significance (Figure 7B). Next, in order to investigate whether the stability and function of Tregs were altered in vivo in the presence of immunosuppressive drugs and in a pro-inflammatory environment, we used cell sorting to purify human HLAA2+ Tregs from splenocytes and tested them in vitro for their function. The cytokine profiles of sorted Tregs was analyzed by intracellular staining.6 The results showed that IL-10, IL-17 and IFN-γ production were similar to that observed in vitro and the percentages of positive cells was identical between mice treated in vivo with all the different drugs or left untreated (Figure 7C). The suppressive capacity of ex vivo Tregs was tested in vitro. The results demonstrated that Tregs from either control or drug-treated mice maintained their suppressive ability (Figure 7D).

in vitro in the presence of IL-2 and RAPA to obtain therapeutic numbers of highly pure Tregs. In our clinical trial, Tregs are infused into patients receiving additional therapies in the form of a combination of TAC, mPr and MMF. However, in many transplant centers, immunosuppressive regimens based on the use of sirolimus (RAPA) are under investigation.5,29,30 In this study, we examianed whether the co-administration of immunosuppressive drugs could

Discussion Treg therapy is already in the clinic in the treatment of GvHD and Type I diabetes and has been shown to be safe.5 At KCL we have started a clinical trial in renal transplant patients (the One Study) in which Tregs are separated from the peripheral blood of the patients who are going forward to receive kidney transplants and expanded 96

Figure 5. Effect of immunosuppressants on the production of cytokines by Tregs. Amount of IL-17 (A), IFN-γ (B) and IL-10 (C) in the supernatants from 5-day culture of Tregs activated with anti-CD3/CD28 coated beads (bead:cell ratio = 1:2) and IL-2 (20 IU/mL) and in the presence of single immunosuppressive agent (TAC, MPA and mPR) or their combination (IS Mix). Data are from 3 independent experiments. *P<0.05.

haematologica | 2016; 101(1)

Impact of immunosuppressive drugs on Treg function

affect the phenotype, function and stability of in vitro expanded Tregs. We have demonstrated both in vitro and in vivo that the treatment with MMF, TAC and mPr, but not RAPA, decreased Treg viability and proliferation. In addition, none of the different ISDs altered Treg function and stability. However, we found that TAC had the biggest effect on Tregs, affecting not only their viability and proliferative capacity, but also the cytokine profile and the expression of some chemokine receptors. The negative effect of TAC is not surprising. Tregs cannot produce IL-2, but their survival and suppressive activity depend on the exogenous supply of this cytokine.31,32 TAC, by inhibiting calcineurin pathway, blocks IL-2 production from T cells making Tregs, and T effectors very susceptible to the inhibitory effect of this drug. This may explain, at least in part, the reduced size of the spleens and the lowered number of proliferating cells observed in vivo

in the animals treated with either the IS Mix or RAPA+TAC. Furthermore, our in vitro results, revealed a direct effect of TAC on IL-2RÎą (CD25) expression on Tregs. This inhibition, previously described with T lymphocytes,33 may aggravate even more the already critical Treg survival during immunosuppressive drug administration. Although previous data suggested that the treatment of transplant recipients with low doses of TAC may be beneficial to Tregs,34 there are many other studies providing evidence of the potentially harmful effect of calcineurin inhibitors on human Tregs.35-37 Although, we have not found any evidence of a negative effect of TAC on the FOXP3 expression, function and stability of our Treg preparations, some studies in animal models reported that CsA treatment reduced Foxp3 expression in natural diminished the frequencies of Tregs,32,38 CD4+CD25+Foxp3+ T cells,39 and failed to support the dif-

Figure 6. Effect of immunosuppressants on human cell engraftment in humanized mice. (A) Percentages of engraftment and absolute numbers of total human CD45+ cells in spleen of mice receiving immunosuppressive treatment (IS Mix, RAPA and RATA+TAC) and controls (CTR). (B) Percentages of engraftment and absolute numbers of human Tregs (HLAA2+ cells) in spleen of mice receiving immunosuppressive treatment (IS Mix, RAPA and RATA+TAC) and controls (CTR). (C) Percentages of proliferating Teff (left) and Tregs (right) in the four groups calculated as Ki-67+ cells. Data are representative of 4 independent experiments. *P<0.05; **P<0.01; ***P<0.001.

haematologica | 2016; 101(1)

C. Scottà et al.

ferentiation of the highly suppressive CD4+CD25+CD27+ subset upon alloantigen stimulation.40 In addition, calcineurin inhibitors have been shown to affect Treg stability by directly interfering with NFAT signaling, and, as a consequence, its interaction with FOXP3.37 In the experiment in which TAC was injected in mice together with RAPA, it was clear that TAC had a dominant effect on RAPA leading to a decrease in Treg viability and proliferation. The detrimental effect observed with ISDs on Tregs might not be limited to TAC, but MPA could have also played a role in vivo, as shown in vitro. In literature, the effect of this drug on Tregs has not been extensively studied and the few results reported are controversial. MPA has been shown to inhibit inosine 5'-monophosphate dehydrogenase, a rate-limiting enzyme in the de novo synthesis of guanine nucleotides; thereby it affects DNA synthesis and cell replication of conventional T cells, as well as Tregs.11 Our results confirmed this negative effect on cell replication in vitro and in vivo when administered as IS

Mix cocktail. Our findings are supported by data in literature describing the influence of this drug on the expansion of antigen specific Tregs and establishing long-term tolerance; Wu et al. showed, in a murine model, that MMF administration significantly inhibited the expansion of OVA-specific CD4+CD25+Foxp3+ Tregs after OVA immunization.41 Similarly, Lim et al. showed that MPA adversely affected the therapeutic effectiveness of adoptively transferred Tregs, reducing the median survival time of allogenic skin transplants in recipient mice from 21 to 16 days.17 On the other hand, some authors suggested that the administration of MPA have a positive role in Tregs. Zeiser et al. demonstrated that the concurrent administration of freshly isolated Tregs and MPA in a murine model of GvHD did not significantly inhibit either Treg expansion or suppressive ability and showed that the overall survival of the mice was extended.32 Other reports suggested a role for this drug in the induction of a more tolerogenic environment. Indeed, Shibutani et al. showed that generation of Tregs in mice by intra-tracheal delivery

Figure 7. Effect of immunosuppressants on Treg phenotype and function in a humanized mouse model. (A) Percentages of FOXP3+, CTLA+ and CD25+ Tregs recovered from mice receiving immunosuppressive treatment (IS Mix, RAPA and RATA+TAC) and controls (CTR). (B) Percentages of CLA+ and Integrin-β7+ Tregs in the 4 groups of mice. (C) Percentages of Tregs producing IL-17, IFN-γ and IL-10, respectively, in the 4 groups of animals. (D) Representative histograms showing the suppressive ability of Tregs recovered from mice under immunosuppressive treatments and controls. Treg:Teff ratio 1:2 and 1:4. Data are representative of 4 independent experiments.

haematologica | 2016; 101(1)

Impact of immunosuppressive drugs on Treg function

of alloantigen was facilitated by the presence of MPA.42 In vitro studies further confirmed the pro-tolerogenic effect of MPA on the generation of Tregs43 and on the enhancement of FOXP3 expression and co-inhibitory molecules, such as PD-1 and CTLA-4 in total CD4+ T cells.44 The treatment with mPr alone in vitro or in vivo was not harmful for Tregs. Steroids such as dexamethasone and prednisolone have been used for decades as a basis for the treatment of inflammatory diseases and many authors have described positive effects on the maturation and expansion of Tregs in autoimmune diseases. Glucocorticoids have been suggested to amplify the IL-2-dependent expansion of FOXP3+CD4+CD25+ T cells in vivo,45 increase FOXP3 expression by Tregs in patients affected by asthma,14 and restore the impaired suppressive function of Tregs in patients with relapsing multiple sclerosis.46 Altogether, steroids may have a beneficial effect on Tregs by influencing other cells and altering the inflammatory environment. This idea was suggested by some studies reporting that glucocorticoids negatively control both Th1- and Th17-polarization in mice and humans. Data showed that this treatment can affect the production of Th1 or Th17 polarizing cytokines produced by innate immune cells, as well as T-cell capability to respond to these cytokines.47,48 Furthermore, recent in vitro and in vivo studies showed that glucocorticoids can favor the establishment of pro-tolerogenic conditions. Human DCs treated with glucocorticoids up-regulate the GC-Induced Leucine Zipper protein (GILZ) and differentiate into tolerogenic cells capable of inducing IL-10–producing antigen-specific Tregs.49 Finally, our data have shown that the in vivo administration of RAPA maintained the survival and proliferation of adoptively transferred Tregs. In contrast, the proliferation of non-Treg cells was inhibited. Similar results were obtained in a murine model of heart transplant in which the adoptive transfer of a small number of alloantigen-specific Treg in the presence of a low dose of RAPA induced long-term survival of cardiac allografts.50 The beneficial effect of RAPA on Tregs was also observed in a humanized-mouse model where segments of human arterial were transplanted into immunodeficient mice reconstituted with allogeneic human peripheral blood cells. In this model, the authors showed that the presence of RAPA enhanced the ability of sub-therapeutic numbers of human Tregs to prevent transplant rejection.18 The differential in vivo effect of RAPA and TAC is of interest. Although they are structurally similar and bind the same target (FK-binding protein), RAPA leaves the calcineurin pathway untouched. Our results confirmed data in literature showing that RAPA rather than CNI, maintains Treg function without decreasing the efficacy of the

References 1. Issa F, Hester J, Goto R, Nadig SN, Goodacre TE, Wood K. Ex vivo-expanded human regulatory T cells prevent the rejection of skin allografts in a humanized mouse model. Transplantation. 2010;90(12):1321-1327. 2. Putnam AL, Safinia N, Medvec A, et al. Clinical grade manufacturing of human alloantigen-reactive regulatory T cells for use

haematologica | 2016; 101(1)

immunotherapy and leads to a prolonged allograft survival.32 Altogether the data obtained from our study suggest that ISDs do not affect the key functions of ex vivo expanded Tregs. However, the treatment can have detrimental effects on their viability and proliferative capacity. It has been reported that the percentage of circulating Tregs decreases in the months after transplantation in patients on immunosuppression with calcineurin inhibitors.35,36 Similarly, a clinical study showed that the conversion of kidney transplant recipients from TAC+MPA treatment to RAPA mono-therapy, significantly increased the percentage and absolute number of circulating Tregs.30 Therefore, our results confirm that TAC and MPA may have an unhelpful effect, and suggest that more tolerance-permissive agents such as RAPA should be used as adjunctive therapy with Tregs. Finally, our data indicate that the use of a concurrent immunosuppression treatment can be compatible with adoptive Treg therapy. However, the choice of specific drugs, as well as their timing and dosing, could be an essential component of strategies to induce and maintain tolerance mediated by Treg therapy. Acknowledgments The authors would like to thank Alberto Sanchez Fueyo, Marc Martinez-Llordella and Veronica Coelho for helpful suggestions, Kate John (Viapath) for technical support. The authors would like to acknowledge financial support from the Department of Health through the National Institute for Health Research (NIHR) comprehensive Biomedical Research Centre award to Guy’s & St Thomas’ NHS Foundation Trust in partnership with King’s College London and King’s College Hospital NHS Foundation Trust. The authors also acknowledge the support of the MRC Centre for Transplantation. Funding This work was funded by grants from the National Institute for Health Research, Medical Research Council (BA, GL and RIL), British Heart Foundation (PB, KR, GL and RIL), the Academy of Medical Sciences/Wellcome Trust (BA), Guy’s and St Thomas’ Charity (RIL and GL), the European Union 7th Framework Programme (EU FP7) ONE Study (GL, RIL and CS) and King's Health Partners Research and Development Challenge Fund (CS; grant number R1405170), Guy's and St Thomas’ Charity (CS, BA, RIL and GL). Research was also supported by the National Institute for Health Research (NIHR) Biomedical Research Centre based at Guy's and St Thomas’ National Health Service (NHS) Foundation Trust and King's College London. The views expressed are those of the author(s) and not necessarily those of the NHS, the NIHR, or the Department of Health.

in transplantation. Am J Transplant. 2013;13(11):3010-3020. 3. Xiao F, Ma L, Zhao M, et al. Ex vivo expanded human regulatory T cells delay islet allograft rejection via inhibiting islet-derived monocyte chemoattractant protein-1 production in CD34+ stem cells-reconstituted NOD-scid IL2rgammanull mice. PLoS One. 2014;9(3):e90387. 4. Sagoo P, Ali N, Garg G, Nestle FO, Lechler RI, Lombardi G. Human regulatory T cells

with alloantigen specificity are more potent inhibitors of alloimmune skin graft damage than polyclonal regulatory T cells. Sci Transl Med. 2011;3(83):83ra42. 5. Edozie FC, Nova-Lamperti EA, Povoleri GA, et al. Regulatory T-cell therapy in the induction of transplant tolerance: the issue of subpopulations. Transplantation. 2014;98(4): 370-379. 6. Scotta C, Esposito M, Fazekasova H, et al. Differential effects of rapamycin and

C. ScottĂ et al.

8. 9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

100

retinoic acid on expansion, stability and suppressive qualities of human CD4(+)CD25(+)FOXP3(+) T regulatory cell subpopulations. Haematologica. 2013;98(8): 1291-1299. Afzali B, Edozie FC, Fazekasova H, et al. Comparison of regulatory T cells in hemodialysis patients and healthy controls: implications for cell therapy in transplantation. Clin J Am Soc Nephrol. 2013;8(8):13961405. Halloran PF. Immunosuppressive drugs for kidney transplantation. N Engl J Med. 2004;351(26):2715-2729. Briganti EM, Russ GR, McNeil JJ, Atkins RC, Chadban SJ. Risk of renal allograft loss from recurrent glomerulonephritis. N Engl J Med. 2002;347(2):103-109. Johnston SD, Morris JK, Cramb R, Gunson BK, Neuberger J. Cardiovascular morbidity and mortality after orthotopic liver transplantation. Transplantation. 2002;73(6):901906. Miroux C, Morales O, Ghazal K, et al. In vitro effects of cyclosporine A and tacrolimus on regulatory T-cell proliferation and function. Transplantation. 2012;94(2): 123-131. Fourtounas C, Dousdampanis P, Sakellaraki P, et al. Different immunosuppressive combinations on T-cell regulation in renal transplant recipients. Am J Nephrol. 2010;32(1):19. Braitch M, Harikrishnan S, Robins RA, et al. Glucocorticoids increase CD4CD25 cell percentage and Foxp3 expression in patients with multiple sclerosis. Acta Neurol Scand. 2009;119(4):239-245. Karagiannidis C, Akdis M, Holopainen P, et al. Glucocorticoids upregulate FOXP3 expression and regulatory T cells in asthma. J Allergy Clin Immunol. 2004;114(6):14251433. Demirkiran A, Sewgobind VD, van der Weijde J, et al. Conversion from calcineurin inhibitor to mycophenolate mofetil-based immunosuppression changes the frequency and phenotype of CD4+FOXP3+ regulatory T cells. Transplantation. 2009;87(7):10621068. Liang H, Liao C, Qi Z, et al. Rapamycin or tacrolimus alone fails to resist cardiac allograft accelerated rejection mediated by alloreactive CD4(+) memory T cells in mice. Transpl Immunol. 2010;22(3-4):128-136. Lim DG, Koo SK, Park YH, et al. Impact of immunosuppressants on the therapeutic efficacy of in vitro-expanded CD4+CD25+Foxp3+ regulatory T cells in allotransplantation. Transplantation. 2010;89(8):928-936. Hester J, Schiopu A, Nadig SN, Wood KJ. Low-dose rapamycin treatment increases the ability of human regulatory T cells to inhibit transplant arteriosclerosis in vivo. Am J Transplant. 2012;12(8):2008-2016. Zahir H, McCaughan G, Gleeson M, Nand RA, McLachlan AJ. Changes in tacrolimus distribution in blood and plasma protein binding following liver transplantation. Ther Drug Monit. 2004;26(5):506-515. Bullingham RE, Nicholls AJ, Kamm BR. Clinical pharmacokinetics of mycophenolate mofetil. Clin Pharmacokinet. 1998;34(6):429-455. de Winter BC, van Gelder T, Sombogaard F,

22.

23.

24.

25.

26.

27. 28.

29.

30.

31.

32.

33.

34.

35.

36.

Shaw LM, van Hest RM, Mathot RA. Pharmacokinetic role of protein binding of mycophenolic acid and its glucuronide metabolite in renal transplant recipients. J Pharmacokinet Pharmacodyn. 2009;36(6): 541-564. Colom H, Lloberas N, Andreu F, et al. Pharmacokinetic modeling of enterohepatic circulation of mycophenolic acid in renal transplant recipients. Kidney Int. 2014;85(6): 1434-1443. van Hest RM, Mathot RA, Pescovitz MD, Gordon R, Mamelok RD, van Gelder T. Explaining variability in mycophenolic acid exposure to optimize mycophenolate mofetil dosing: a population pharmacokinetic meta-analysis of mycophenolic acid in renal transplant recipients. J Am Soc Nephrol. 2006;17(3):871-880. Wallemacq P, Armstrong VW, Brunet M, et al. Opportunities to optimize tacrolimus therapy in solid organ transplantation: report of the European consensus conference. Ther Drug Monit. 2009;31(2):139-152. Rostin M, Barthe P, Houin G, Alvinerie M, Bouissou F. Pharmacokinetics of prednisolone in children with the nephrotic syndrome. Pediatr Nephrol. 1990;4(5):470-473. Borrows R, Chusney G, Loucaidou M, et al. Mycophenolic acid 12-h trough level monitoring in renal transplantation: association with acute rejection and toxicity. Am J Transplant. 2006;6(1):121-128. Shevach EM. Mechanisms of foxp3+ T regulatory cell-mediated suppression. Immunity. 2009;30(5):636-645. Tresoldi E, Dell' Albani I, Stabilini A, et al. Stability of human rapamycin-expanded CD4+CD25+ T regulatory cells. Haematologica. 2011;96(9):1357-1365. Formica RN Jr, Lorber KM, Friedman AL, et al. Sirolimus-based immunosuppression with reduce dose cyclosporine or tacrolimus after renal transplantation. Transplant Proc. 2003;35(3 Suppl):95S-98S. Hendrikx TK, Velthuis JH, Klepper M, et al. Monotherapy rapamycin allows an increase of CD4 CD25 FoxP3 T cells in renal recipients. Transpl Int. 2009;22(9):884-891. Furtado GC, Curotto de Lafaille MA, Kutchukhidze N, Lafaille JJ. Interleukin 2 signaling is required for CD4(+) regulatory T cell function. J Exp Med. 2002;196(6):851857. Zeiser R, Nguyen VH, Beilhack A, et al. Inhibition of CD4+CD25+ regulatory T-cell function by calcineurin-dependent interleukin-2 production. Blood. 2006;108(1): 390-399. Mascarell L, Truffa-Bachi P. Cyclosporin A therapy differently affects immunologicalrelevant gene expression following immunization. Immunol Lett. 2002;84(2):137-143. Wang Z, Shi B, Jin H, Xiao L, Chen Y, Qian Y. Low-dose of tacrolimus favors the induction of functional CD4(+)CD25(+)FoxP3(+) regulatory T cells in solid-organ transplantation. Int Immunopharmacol. 2009;9(5):564-569. Codarri L, Vallotton L, Ciuffreda D, et al. Expansion and tissue infiltration of an allospecific CD4+CD25+CD45RO+IL7Ralphahigh cell population in solid organ transplant recipients. J Exp Med. 2007;204 (7):1533-1541. Demirkiran A, Kok A, Kwekkeboom J, et al. Low circulating regulatory T-cell levels after

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

acute rejection in liver transplantation. Liver Transpl. 2006;12(2):277-284. Akimova T, Kamath BM, Goebel JW, et al. Differing effects of rapamycin or calcineurin inhibitor on T-regulatory cells in pediatric liver and kidney transplant recipients. Am J Transplant. 2012;12(12):3449-3461. Baan CC, van der Mast BJ, Klepper M, et al. Differential effect of calcineurin inhibitors, anti-CD25 antibodies and rapamycin on the induction of FOXP3 in human T cells. Transplantation. 2005;80(1):110-117. Segundo DS, Ruiz JC, Izquierdo M, et al. Calcineurin inhibitors, but not rapamycin, reduce percentages of CD4+CD25+FOXP3+ regulatory T cells in renal transplant recipients. Transplantation. 2006;82(4):550-557. Coenen JJ, Koenen HJ, van Rijssen E, Hilbrands LB, Joosten I. Rapamycin, and not cyclosporin A, preserves the highly suppressive CD27+ subset of human CD4+CD25+ regulatory T cells. Blood. 2006;107(3):10181023. Wu T, Zhang L, Xu K, et al. Immunosuppressive drugs on inducing Agspecific CD4(+)CD25(+)Foxp3(+) Treg cells during immune response in vivo. Transpl Immunol. 2012;27(1):30-38. Shibutani S, Inoue F, Aramaki O, et al. Effects of immunosuppressants on induction of regulatory cells after intratracheal delivery of alloantigen. Transplantation. 2005;79(8): 904-913. Lagaraine C, Lemoine R, Baron C, Nivet H, Velge-Roussel F, Lebranchu Y. Induction of human CD4+ regulatory T cells by mycophenolic acid-treated dendritic cells. J Leukoc Biol. 2008;84(4):1057-1064. He X, Smeets RL, Koenen HJ, et al. Mycophenolic acid-mediated suppression of human CD4+ T cells: more than mere guanine nucleotide deprivation. Am J Transplant. 2011;11(3):439-449. Chen X, Oppenheim JJ, Winkler-Pickett RT, Ortaldo JR, Howard OM. Glucocorticoid amplifies IL-2-dependent expansion of functional FoxP3(+)CD4(+)CD25(+) T regulatory cells in vivo and enhances their capacity to suppress EAE. Eur J Immunol. 2006;36(8): 2139-2149. Xu L, Xu Z, Xu M. Glucocorticoid treatment restores the impaired suppressive function of regulatory T cells in patients with relapsing-remitting multiple sclerosis. Clin Exp Immunol. 2009;158(1):26-30. Blotta MH, DeKruyff RH, Umetsu DT. Corticosteroids inhibit IL-12 production in human monocytes and enhance their capacity to induce IL-4 synthesis in CD4+ lymphocytes. J Immunol. 1997;158(12):5589-5595. Luther C, Adamopoulou E, Stoeckle C, et al. Prednisolone treatment induces tolerogenic dendritic cells and a regulatory milieu in myasthenia gravis patients. J Immunol. 2009;183(2):841-848. Calmette J, Ellouze M, Tran T, et al. Glucocorticoid-induced leucine zipper enhanced expression in dendritic cells is sufficient to drive regulatory T cells expansion in vivo. J Immunol. 2014;193(12):5863-5872. Raimondi G, Sumpter TL, Matta BM, et al. Mammalian target of rapamycin inhibition and alloantigen-specific regulatory T cells synergize to promote long-term graft survival in immunocompetent recipients. J Immunol. 2010;184(2):624-636.

haematologica | 2016; 101(1)