Haematologica, Volume 107, Issue 7 by Haematologica

haematologica Editor-in-Chief Jacob M. Rowe (Jerusalem)

Deputy Editors Carlo Balduini (Pavia), Jerry Radich (Seattle)

Associate Editors Hélène Cavé (Paris), Monika Engelhardt (Freiburg), Steve Lane (Brisbane), Pier Mannuccio Mannucci (Milan), Pavan Reddy (Ann Arbor), David C. Rees (London), Francesco Rodeghiero (Vicenza), Gilles Salles (New York), Kerry Savage (Vancouver), Aaron Schimmer (Toronto), Richard F. Schlenk (Heidelberg), Sonali Smith (Chicago)

Statistical Consultant Catherine Klersy (Pavia)

Editorial Board Walter Ageno (Varese), Sarit Assouline (Montreal), Andrea Bacigalupo (Roma), Taman Bakchoul (Tübingen), Pablo Bartolucci (Créteil), Katherine Borden (Montreal), Marco Cattaneo (Milan), Corey Cutler (Boston), Kate Cwynarski (London), Mary Eapen (Milwaukee), Francesca Gay (Torino), Ajay Gopal (Seattle), Alex Herrera (Duarte), Shai Izraeli (Ramat Gan), Martin Kaiser (London), Marina Konopleva (Houston), Johanna A. Kremer Hovinga (Bern), Nicolaus Kröger (Hamburg), Austin Kulasekararaj (London), Shaji Kumar (Rochester), Ann LaCasce (Boston), Anthony R. Mato (New York), Matthew J. Maurer (Rochester), Neha Mehta-Shah (St. Louis), Alison Moskowitz (New York), Yishai Ofran (Haifa), Farhad Ravandi (Houston), John W. Semple (Lund), Liran Shlush (Toronto), Sara Tasian (Philadelphia), Pieter van Vlieberghe (Ghent), Ofir Wolach (Haifa), Loic Ysebaert (Toulouse)

Managing Director Antonio Majocchi (Pavia)

Editorial Office Lorella Ripari (Office & Peer Review Manager), Simona Giri (Production & Marketing Manager), Paola Cariati (Graphic Designer), Giulia Carlini (Graphic Designer), Igor Poletti (Graphic Designer), Marta Fossati (Peer Review), Diana Serena Ravera (Peer Review), Laura Sterza (Account Administrator)

Assistant Editors Britta Dost (English Editor), Rachel Stenner (English Editor), Bertie Vitry (English Editor), Massimo Senna (Information technology), Idoya Lahortiga (Graphic artist)

Haematologica | 107 - June 2022

Brief information on Haematologica 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, Original articles, Review articles, Perspective articles, Editorials, Guideline articles, Letters to the Editor, Case reports & Case series and Comments. 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 at 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. Subscription. Detailed information about subscriptions is available at www.haematologica.org. Haematologica is an open access journal and access to the online journal 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 printed edition for the year 2022 are as following: Institutional: Euro 700 Personal: Euro 170 Advertisements. Contact the Advertising Manager, Haematologica Office, via Giuseppe Belli 4, 27100 Pavia, Italy (phone +39.0382.27129, fax +39.0382.394705, e-mail: marketing@haematologica.org). Disclaimer. Whilst every effort is made by the publishers and the editorial board to see that no inaccurate or misleading data, opinion or statement appears in this journal, they wish to make it clear that the data and opinions appearing in the articles or advertisements herein are the responsibility of the contributor or advisor concerned. Accordingly, the publisher, the editorial board and their respective employees, officers and agents accept no liability whatsoever for the consequences of any inaccurate or misleading data, opinion or statement. Whilst all due care is taken to ensure that drug doses and other quantities are presented accurately, readers are advised that new methods and techniques involving drug usage, and described within this journal, should only be followed in conjunction with the drug manufacturer’s own published literature.

Direttore responsabile: Prof. Carlo Balduini; Autorizzazione del Tribunale di Pavia n. 63 del 5 marzo 1955. Printing: Press Up, zona Via Cassia Km 36, 300 Zona Ind.le Settevene - 01036 Nepi (VT)

Associated with USPI, Unione Stampa Periodica Italiana. Premiato per l’alto valore culturale dal Ministero dei Beni Culturali ed Ambientali Haematologica | 107 - June 2022

Table of Contents Volume 107, Issue 7: July 2022 About the Cover 1493

Images from the Haematologica Atlas of Hematologic Cytology: parvovirus-induced pure red cell aplasia Rosangela Invernizzi https://doi.org/10.3324/haematol.2022.281331

Landmark Papers in Hematology 1494

Graft-versus-host disease in allogeneic transplantation: the good and the bad Nicolaus-Martin Kröger https://doi.org/10.3324/haematol.2022.281425

1496

Is allogeneic transplantation the preferred therapy for older patients with acute myeloid leukemia? Arnold Ganser https://doi.org/10.3324/haematol.2021.279820

1498

Early death in acute promyelocytic leukemia: time to redefine risk groups Meira Yisraeli Salman and Yishai Ofran https://doi.org/10.3324/haematol.2021.280446

1501

Did brentuximab vedotin’s rise to the top ECHELON of Hodgkin therapeutics invalidate AETHERA results? Mehdi Hamadani https://doi.org/10.3324/haematol.2021.280284

Editorials

Review Articles 1503

How I diagnose and treat chronic myelomonocytic leukemia Mrinal M. Patnaik https://doi.org/10.3324/haematol.2021.279500

1518

Acute Myeloid Leukemia Outcomes of older patients aged 60 to 70 years undergoing reduced intensity transplant for acute myeloblastic leukemia: results of the NCRI acute myeloid leukemia 16 trial Nigel H. Russell, et al. https://doi.org/10.3324/haematol.2021.279010

1528

Acute Myeloid Leukemia A risk score based on real-world data to predict early death in acute promyelocytic leukemia Albin Österroos et al. https://doi.org/10.3324/haematol.2021.280093

Articles

Haematologica | 107 - July 2022

III

1538

Cell Therapy and Immunotherapy Therapeutic targeting of endoplasmic reticulum stress in acute graft-versus-host disease Eileen Haring et al. https://doi.org/10.3324/haematol.2021.278387

1555

Cell Therapy and Immunotherapy Cardiovascular events in patients treated with chimeric antigen receptor T-cell therapy for aggressive B-cell lymphoma Raphael E. Steiner et al. https://doi.org/10.3324/haematol.2021.280009

1567

Hemostasis Renal function and clinical outcome of patients with cancer-associated venous thromboembolism randomized to receive apixaban or dalteparin. Results from the Caravaggio trial Cecilia Becattini et al. https://doi.org/10.3324/haematol.2021.279072

1577

Hemoglobinopathies Genetic modifiers of fetal hemoglobin affect the course of sickle cell disease in patients treated with hydroxyurea Pierre Allard et al. https://doi.org/10.3324/haematol.2021.278952

1589

Iron Metabolism & its Disorders Hepcidin regulation in Kenyan children with severe malaria and non-typhoidal Salmonella bacteremia Kelvin M. Abuga et al. https://doi.org/10.3324/haematol.2021.279316

1599

Myeloproliferative Disorders Retrospective analysis of pacritinib in patients with myelofibrosis and severe thrombocytopenia Srdan Verstovsek et al. https://doi.org/10.3324/haematol.2021.279415

1608

Non-Hodgkin Lymphoma Phase I/II clinical trial of temsirolimus and lenalidomide in patients with relapsed and refractory lymphomas Ajay Major et al. https://doi.org/10.3324/haematol.2021.278853

1619

Non-Hodgkin Lymphoma Whole-genome profiling of primary cutaneous anaplastic large cell lymphoma Armando N. Bastidas Torres et al. https://doi.org/10.3324/haematol.2020.263251

1633

Non-Hodgkin Lymphoma Total metabolic tumor volume as a survival predictor for patients with diffuse large B-cell lymphoma in the GOYA study Lale Kostakoglu et al. https://doi.org/10.3324/haematol.2021.278663

1643

Platelet Biology and its Disorders Platelet dysfunction in platelet-type von Willebrand disease due to the constitutive triggering of the Lyn-PECAM1 inhibitory pathway Loredana Bury et al. https://doi.org/10.3324/haematol.2021.278776

1655

Platelet Biology and its Disorders Establishment of a Bernard-Soulier syndrome model in zebrafish Qing Lin et al. https://doi.org/10.3324/haematol.2021.278893

Haematologica | 107 - July 2022

III

1669

Platelet Biology and its Disorders Single platelet and megakaryocyte morpho-dynamics uncovered by multicolor reporter mouse strains in vitro and in vivo Leo Nicolai et al. https://doi.org/10.3324/haematol.2021.278896

Letters to the Editor 1681

1687

Vaccine-induced immune thrombotic thrombocytopenia: a possible pathogenic role of ChAdOx1 nCoV-19 vaccine-encoded soluble SARS-CoV-2 spike protein Manuela De Michele et al. https://doi.org/10.3324/haematol.2021.280180

1693

All-oral triplet combination of ixazomib, lenalidomide, and dexamethasone in newly diagnosed transplant-eligible multiple myeloma patients: final results of the phase II IFM 2013-06 study Cyrille Touzeau et al. https://doi.org/10.3324/haematol.2021.280394

1698

Complement C1s inhibition with sutimlimab results in durable response in cold agglutinin disease: CARDINAL study 1-year interim follow-up results Alexander Röth et al. https://doi.org/10.3324/haematol.2021.279812

1703

1709

Decitabine salvage for TP53-mutated, relapsed/refractory acute myeloid leukemia after cytotoxic induction therapy Francesca Ferraro et al. https://doi.org/10.3324/haematol.2021.280153

1714

Tricuspid-valve regurgitant jet velocity as a risk factor for death in β-thalassemia Giorgio Derchi et al. https://doi.org/10.3324/haematol.2021.280389

1719

BRAF V600E-positive cells as molecular markers of bone marrow disease in pediatric Langerhans cell histiocytosis Ko Kudo, et al. https://doi.org/10.3324/haematol.2021.279857

1726

Treatment emergent peripheral neuropathy in the CASSIOPEIA trial Cathelijne Fokkema et al. https://doi.org/10.3324/haematol.2021.280567

1731

Prolonged viral replication in patients with hematologic malignancies hospitalized with COVID-19 Carolina Garcia-Vidal et al. https://doi.org/10.3324/haematol.2021.280407

Haematologica | 107 - July 2022

III

ABOUT THE COVER

R. Invernizzi

Images from the Haematologica Atlas of Hematologic Cytology: parvovirus-induced pure red cell aplasia Rosangela Invernizzi University of Pavia, Pavia, Italy E-mail: rosangela.invernizzi@unipv.it https://doi.org/10.3324/haematol.2022.281331

Pure red cell aplasia (PRCA) is a rare disorder characterized by normochromic-normocytic or normochromic-macrocytic anemia, marked reticulocytopenia and almost complete absence of erythroid precursors in the bone marrow. In most cases the granulocytic and megakaryocytic lineages are normal. This disorder includes congenital and acquired conditions. In older children and adults the most common form of acute PRCA is due to parvovirus B19 infection. The virus invades and destroys red cell progenitors, and the aplasia is terminated when neutralizing antibodies develop. In immunocompromised patients, parvovirus B19 infection can cause a chronic form of PRCA. Distinctive morphological features of parvovirus-induced PRCA are illustrated in the Figure showing bone marrow smears from a heart transplant recipient presenting with severe normochromic normocytic anemia, a low reticulocyte count, and normal leukocyte, differential and platelet counts, as well as high levels of erythropoietin. (A and B) There is marked hypoplasia of the erythroblastic lineage with almost total absence of maturing erythroblasts. Note the atypical hyperbasophilic giant cells with very large, prominent nucleoli, cytoplasmic vacuolization and blebs. These giant cells are proerythroblasts with morphological features pathognomonic for parvovirus B19 infection. Panel C shows an enormous proerythroblast with numerous nuclear inclusions. (D) The atypical erythroblasts show an abnormal, very strongly positive periodic acid Schiff (PAS) reaction. The diagnosis of parvovirus B19-induced PRCA was confirmed by immunocytochemical and molecular tests; however, serology was negative, demonstrating its limited diagnostic usefulness in cases with immunodeficiency. Disclosures No conflicts of interest to disclose.

References 1. Invernizzi R. Pure red cell aplasia. Haematologica. 2020;105(Suppl 1):184-187. Haematologica | 107 July 2022

1493

LANDMARK PAPER IN HEMATOLOGY

N-M. Kröger

Graft-versus-host disease in allogeneic transplantation: the good and the bad Nicolaus-Martin Kröger Department of Stem Cell Transplantation, University Medical Center Hamburg-Eppendorf, Hamburg, Germany E-mail: n.kroeger@uke.de doi:10.3324/haematol.2022.281425

TITLE

Antileukemic effect of graft-versus-host disease in human recipients of allogeneic-marrow grafts.

AUTHORS Weiden PL, Flournoy N, Thomas ED, Prentice R, Fefer A, Buckner CD, Storb R. JOURNAL New England Journal of Medicine. 1979;300(19):1068-1073. PMID: 34792.

Now after more than 60 years of clinical research into allogeneic hematopoietic stem cell transplantation this treatment approach is well established as a curative option for many hematologic malignancies. One of the major complications of this treatment is an immunologicalbased complication which was initially named ‘secondary disease’ and is now well known as graft-versus-host disease (GvHD). The preclinical observation that GvHD in mice also resulted in eradication of leukemic cells was described by Mathe et al. as the graft-versus-leukemia effect.1 Clinically, the powerful correlation between GvHD and an antileukemic effect in humans was convincingly demonstrated by Weiden and colleagues from Seattle in

two seminal studies published in 1979 for acute GvHD and in 1981 for chronic GvHD.2,3 Patients who experienced acute GvHD had a 2.5 times lower relative relapse rate than those without GvHD. Notably, during the first 130 days after allogeneic stem cell transplantation the relapse rate among patients with acute GvHD grade II to IV was ten times lower than that in patients without (grade 0 to I) GvHD and 13 times lower than that among syngeneic graft recipients (Figure 1).2 Despite this lower incidence of relapse the benefit for overall survival was offset by a higher non-relapse mortality in patients with GvHD. This described clinical observation highlighted the double-edged sword of GvHD

Figure 1. Kaplan-Meier probability of remaining in remission after allogeneic bone marrow transplantation for acute leukemia according to the occurrence of acute graft-versus-host disease. Figure drawn from Weiden et al. NEJM 1979.2 GVHD: graftversus-host disease. Haematologica | 107 July 2022

1494

LANDMARK PAPER IN HEMATOLOGY

N-M. Kröger

regarding toxicity and cure and has boosted preclinical and clinical research in order to separate GvHD from the graftversus-leukemia effect and to develop clinical strategies to reduce GvHD without inducing a higher rate of relapse. This work, which demonstrated the crucial effect of donor T cells in eliminating leukemia cells, was also the rationale for the use of donor T cells after transplantation (donor leukocyte infusion) to treat relapse in chronic myeloid leukemia and other diseases, resulting in impressive remission rates.4 Later, multiple approaches to reduce the risk of GvHD without increasing the risk of relapse were investigated and were used in clinical practice. These approaches included antithymocyte globulin and selected

depletion of α/β T cells, as well as an increased antileukemic effect with adoptive immunotherapy using leukemicspecific T cells, chimeric antigen receptor T cells or natural killer cells among others. However, despite clinical improvement in reducing GvHD and harnessing the graft-versus-leukemia effect, initiated by Weiden’s seminal paper, a clear separation between GvHD and the graft-versus-leukemia effect still remains, in 2022, the holy grail of allogeneic stem cell transplantation.1-4 Disclosures No conflicts of interest to disclose.

References 1. Mathe G, Amiel JL, Schwarzenberg L, Cattan A, Schneider M. Adoptive immunotherapy of acute leukemia: experimental and clinical results. Cancer Res. 1965;25(9):1525-1531. 2. Weiden PL, Flournoy N, Thomas ED, Prentice R, Fefer A, Buckner CD, Storb R. Antileukemic effect of graft-versus-host disease in human recipients of allogeneic-marrow grafts. N Engl J Med. 1979;300(19):1068-1073. 3. Weiden PL, Sullivan KM, Flournoy N, Storb R, Thomas ED;

Seattle Marrow Transplant Team. Antileukemic effect of chronic graft-versus-host disease: contribution to improved survival after allogeneic marrow transplantation. N Engl J Med. 1981;304(25):1529-1533. 4. Kolb HJ, Schattenberg A, Goldman JM, et al. Graft-versusleukemia effect of donor lymphocyte transfusions in marrow grafted patients. Blood. 1995;86(5):2041-2050.

Haematologica | 107 July 2022

1495

EDITORIAL

A. Ganser

Is allogeneic transplantation the preferred therapy for older patients with acute myeloid leukemia? Arnold Ganser Department of Hematology, Hemostasis, Oncology, and Stem Cell Transplantation, Hannover Medical School, Hannover, Germany E-mail: ganser.arnold@mh-hannover.de https://doi.org/10.3324/haematol.2021.279820

The treatment of acute myeloid leukemia (AML) has seen tremendous developments over the last few years, but treatment approaches and results are predominantly determined by selection based on age, karyotype, and molecular genotype. Treatment recommendations mainly follow the European LeukemiaNet guidelines,1 especially regarding the use of allogeneic hematopoietic stem cell transplantation (HSCT) as post-remission consolidation therapy for adverse-risk disease. However, these recommendations are category 2A only with low-level evidence but uniform panel consensus, since results from randomized clinical trials are lacking.2 The deficits with the system, in addition, include a bias in that the dataset for its establishment almost exclusively consists of treatment results in younger fit patients with de novo AML who were treated with intensive induction chemotherapy followed by high-dose cytosine arabinoside or HSCT. Recommendations for post-remission therapy in elderly AML patients have not explicitly addressed HSCT but point to the high degree of selection even with reduced intensity conditioning (RIC).3 In this issue of Haematologica, Russell et al. now present a considerable set of data on the outcome of older AML patients, aged 60 to 70 years, treated intensively within the National Cancer Research Institute (NCRI) AML 16 study with various induction regimens followed by RIC transplantation in the case of non-favorable cytogenetics and the presence of a fully matched related or unrelated (at least 9/10 HLA-match) donor.4 Out of 932 patients treated between 2006 and 2012, 788 continued on some sort of chemotherapy, while 144 underwent HSCT (sibling n=52; matched unrelated donor n=92). The survival rate at 5 years was 37% among the transplanted patients compared to 20% in the chemotherapy arm (P<0.001). There was no significant difference in survival between patients transplanted with grafts from siblings or matched unrelated donors. Dividing patients into groups according to Wheatley risk,5 all three risk groups benefited from HSCT. Although mutation status was not known in the majority

of patients, benefit was also seen in patients with an FLT3-ITD and/or NPM1 mutation with no difference among genotypic subgroups. Thus, Russell et al. conclude that “RIC transplantation is an attractive option for older AML patients lacking favorable risk cytogenetics”.4 Although the data are encouraging they do not fully solve the current problems in the elderly AML patient population with regard to appropriate post-remission therapy. In the NRCI AML 16 trial the patients had to be fit for intensive induction chemotherapy which only applies to the minority of the patients. Recent developments with new effective drugs and combination therapies have not yet been addressed. These include, for example, the use of additional FLT3 inhibitors6 or CPX-3517 in fit patients who can be treated intensively, or combinations of hypomethylating agents plus venetoclax which have now become standard of care in less fit AML patients. The latter combination is especially effective in AML with NPM1 and IDH mutations.8 These developments, especially in the unfit population, might even lead to a more dynamic approach, since patients unfit at the time of AML diagnosis might become fit for RIC HSCT once they have entered complete remission with restoration of normal hematopoietic function.9 The NCRI AML 16 data support the now common practice of offering RIC HSCT as post-remission therapy to these patients. However, it would be preferable to design randomized controlled trials to demonstrate the advantage of this approach compared to other post-remission therapies including maintenance therapy, e.g., with oral azacitidine.10 The development of novel drugs will certainly lead to new risk stratification in AML and treatment recommendations which include HSCT also in older AML patients.11 Not only has HSCT undergone rapid progress, but the whole therapeutic landscape is in flux and it would be desirable to have data from randomized controlled trials in order to inform decisions in patients’ care. Disclosures No conflicts of interest to disclose.

References 1. Döhner H, Estey E, Grimwade D, et al. Diagnosis and management of AML in adults: 2017 ELN recommendations from an international expert panel. Blood. 2017;129(4):424-447.

2. Kim MS, Cai J, Maniar A, et al. Comparison of classification of indications for allogeneic and autologous transplant for adults in ASTCT guidelines and evidence available in published

Haematologica | 107 July 2022

1496

EDITORIAL

A. Ganser

literature. JAMA Intern Med. 2021;e214826. 3. Sekeres MA, Guyatt G, Abel G, et al. American Society of Hematology 2020 guidelines for treating newly diagnosed acute myeloid leukemia in older adults. Blood Adv. 2020;4(15):3528-3549. 4. Russell NH, Hills RK, Thoma A, et al. Outcomes of older patients aged 60 to 70 years undergoing reduced intensity transplant for acute myeloblastic leukemia: results of the NCRI acute myeloid leukemia 16 trial. Haematologica. 2022;107(7):1518-1527. 5. Wheatley K, Brookes CL, Howman AJ, et al. Prognostic factor analysis of the survival of elderly patients with AML in the MRC AML 11 and LRF AML 14 trials. Br J Haematol. 2009;145(5):598-605. 6. Stone RM, Mandrekar SJ, Sanford BL, et al. Midostaurin plus chemotherapy for acute myeloid leukemia with a FLT3 mutation. N Engl J Med. 2017;377(5):454-464. 7. Lancet JE, Uy GL, Cortes JE, et al. CPX-351 (cytarabine and daunorubicin) liposome for injection versus conventional

cytarabine plus daunorubicin in older patients with newly diagnosed secondary acute myeloid leukemia. J Clin Oncol. 2018;36(26):2684-2692. 8. DiNardo CD, Jonas BA, Pullarkat V, et al. Azacitidine and venetoclax in previously untreated acute myeloid leukemia. N Engl J Med. 2020;383(7):617-629. 9. Short NJ, Kantarjian H. When less is more: reevaluating the role of intensive chemotherapy for older adults with acute myeloid leukemia in the modern era. J Clin Oncol. 2021;39(28):3104-3108. 10. Wei AH, Döhner H, Pocock C, et al. QUAZAR AML-001 Trial Investigators. Oral azacitidine maintenance therapy for acute myeloid leukemia in first remission. N Engl J Med. 2020;383(26):2526-2537. 11. Short NJ, Tallman MS, Pollyea DA, Ravandi F, Kantarjian H. Optimizing risk stratification in acute myeloid leukemia: dynamic models for a dynamic therapeutic landscape. J Clin Oncol. 2021;39(23):2535-2538.

Haematologica | 107 July 2022

1497

EDITORIAL

M.Y. Salman and Y. Ofran

Early death in acute promyelocytic leukemia: time to redefine risk groups Meira Yisraeli Salman and Yishai Ofran Department of Hematology, Shaare Zedek Medical Center, Jerusalem, Israel E-mail: yofran@szmc.org.il https://doi.org/10.3324/haematol.2021.280446

Acute promyelocytic leukemia (APL) stands out among its myeloid counterparts with unique clinical, morphological and cytogenetic elements. Sixty-five years ago, it was coined the “most malignant form of acute leukemia”,1 with single digit remission rates and a median survival of 3.5 weeks. In 1973, a hallmark study by Bernard et al.2 described the exquisite sensitivity of APL to anthracyclines, rendering this fatal disease curable, although mortality remained high. The incorporation of all-trans retinoic acid (ATRA) and the later introduction of arsenic treatment significantly improved survival and gradually reduced the anxiety that a diagnosis of APL used to evoke among physicians. However, still to this day, early death rates, primarily due to severe coagulopathy, may reach 20-30%. Bernard and colleagues were also among the first to delineate prognostic factors important for survival, including elevated white blood cell (WBC) count and low levels of fibrinogen. Fast-forward to the year 2000, a risk-based model was being established, stratifying patients into three risk categories.3 The model, which became known as the Sanz risk model, defined high-risk patients as those with a WBC count above 10x109/L at presentation, medium-risk patients as those with a WBC count less than 10 x109/L and a platelet count less than 40x109/L and low-risk patients as those with a WBC count less than 10x109/L and a platelet count of more than 40x109/L. However, considering the very low relapse rate in the current arsenic+ATRA era, survival risk scores should now be modified to concentrate on the risk of early death rather than relapse. In this issue of Haematologica, Österroos et al.4 set out to do just that, and present an elegant and simple scoring system for identifying APL patients most in danger of early death. Using real-world data from the Swedish AML registry, they created a model that is externally validated. The training cohort for the development of the prediction model included 301 consecutive adult patients diagnosed with APL in Sweden during the two decades leading up to 2020. The validation cohort encompassed 129 consecutive APL patients diagnosed at a university hospital in Portugal. The treatment protocols were slightly different between the two groups, but both were based on idarubicin induction plus oral ATRA until complete remission in the prearsenic era. From 2016 in the test cohort and 2018 in the validation one, arsenic replaced idarubicin for low- and intermediate-risk patients, and was added on for the high-

risk ones. The score that was developed appears to predict early death more accurately than the Sanz criteria. The three most important prognostic parameters that emerged from the analysis by Österroos and colleagues are WBC and platelet counts at diagnosis, as in the Sanz model, as well as age, which was not included in previous models. An additional and frankly, surprising, finding is that the point at which the risk of early death begins to rise may be from a WBC count as low as 2.2x109/L. The model stratifies patients into three WBC categories that correlate with risk; less than 3x109/L, between 3-5x109/L and more than 5x109/L. Platelet counts are dichotomized with a cutoff of 30x109/L, and age is divided into four subgroups. The identification of the age contribution and the steep increase in risk with WBC counts between 2x109/L and 10x109/L are probably the main contributing factors to the score’s sensitivity. The score has clear, potentially practice-changing implications. Preventing early death by applying immediate aggressive supportive measures to all patients at high risk is imperative. Thus, identification of high-risk patients “hiding” among patients who were previously incorrectly considered to be of low risk, can improve patient care and save lives. Other suggested risk models focus primarily on bleeding risk,5 while this model encompasses all early death risks. To best apply this new risk model in clinical practice, one should focus on the risk calculator presented in Figure 3 in the paper by Österroos et al.4 We recommend that all patients who present with a risk of early death >10% be treated with aggressive supportive measures. According to the proposed model, these high and very-high risk groups include all patients with a WBC count greater than 5x109/L as well as those with lower WBC counts but with advanced age. The presented scoring system has several compelling features. Based on real-world data, the cohorts had no exclusion criteria and included all patients, even those who were too ill or too old to be included in prospective clinical trials, thus greatly improving the reliability of the data. The authors also developed an easy-to-use online tool, readily available for any clinician to use at the bedside in the Emergency Department. The fact that age emerged as an important factor predicting early death is worth dwelling on. It is well known that age is a major factor both in acute leukemia pathogenesis

Haematologica | 107 July 2022

1498

EDITORIAL

M.Y. Salman and Y. Ofran

and prognosis, and in pathologies of the coagulation system. It would be interesting to study whether it is the effect of aging on leukemia or on the coagulation system itself which contributes to the increase in early death risk. As the authors stated, only a small percentage of patients in the cohort received the combination of arsenic with ATRA, as this treatment became prevalent only late during the study period. In addition, future modifications of this score beg inclusion of molecular data if these could be acquired early after diagnosis. One more point to consider may be the inclusion of general coagulopathy risk factors, such as those known to be involved in veno-occlusive diseases, in the scoring system.6 Clearly more validation is necessary in order to create clinical guidelines based on this score, primarily relating to increased surveillance and to recommending risk-adapted thresholds for platelet,

cryoprecipitate and/or fresh-frozen plasma transfusions. To conclude, Österoos and colleagues developed a practical and effective score that may allow identification of highrisk patients hidden among those previously classified as low risk. Aggressive supportive care, if applied, can save the lives of these patients. Perhaps the entire risk-stratification system in APL that is currently guiding treatment should be revamped, and focus on risk of early death. Disregarding age as an important prognostic factor should be remedied, and the WBC cutoff, if used, should be age dependent. Disclosures No conflicts of interest to disclose. Contributions MYS and YO co-wrote the manuscript

References 1. Hillestad LK. Acute promyelocytc leukemia . Acta Med Scand.1957;159(3):189-194. 2. Bernard J, Weil M, Boiron M, et al. Acute promyelocytic leukemia: results of treatment by daunorubicin. Blood. 1973;41(4):489-496. 3. Sanz MA, Lo Coco F, Martin G, et al. Definition of relapse risk and role of nonanthracycline drags for consolidation in patients with acute promyelocytic leukemia: a joint study of the PETHEMA and GIMEMA cooperative groups. Blood. 2000;96(4):1247–1253.

4. Österroos A, Maia T, Eriksson A, et al. A risk score based on realworld data to predict early death in acute promyelocytic leukemia. Haematologica. 2022;107(7):1528-1537. 5. Mitrovic M, Suvajdzic N, Bogdanovic A, et al. International Society of Thrombosis and Hemostasis Scoring System for disseminated intravascular coagulation ≥6: a new predictor of hemorrhagic early death in acute promyelocytic leukemia. Med Oncol. 2013;30(1):478. 6. Mitrovic M, Suvajdzic N, Elezovic I, et al. Thrombotic events in acute promyelocytic leukemia. Thromb Res. 2015;135(4):588-593.

Haematologica | 107 July 2022

1499

EDITORIAL

M. Hamadani

Did brentuximab vedotin’s rise to the top ECHELON of Hodgkin therapeutics invalidate AETHERA results? Mehdi Hamadani BMT & Cellular Therapy Program, Department of Medicine, Medical College of Wisconsin, Milwaukee, WI, USA E-mail: mhamadani@mcw.edu https://doi.org/10.3324/haematol.2021.280284

High-dose therapy (HDT) with autologous hematopoietic cell transplantation (autoHCT) has been considered the standard-of-care for relapsed chemosensitive classical Hodgkin lymphoma (cHL), since the end of the last century.1,2 While this simple and time-tested modality can eradicate disease in the majority of patients undergoing this procedure, recurrent cHL remains the most common cause of treatment failure and death following autoHCT. Several prognostic factors (primary refractory disease, early relapse after frontline treatment, poor patient performance status, extranodal involvement, positron emission tomography-avid residual disease prior to autoHCT, etc.) have been known to predict the risk of relapse after HDT and autoHCT.3 However, pharmacological interventions to mitigate the risk of lymphoma relapse after autoHCT had remained elusive, until the clinical availability of brentuximab vedotin. Brentuximab vedotin is an anti-CD30 antibody conjugated by a protease-cleavable linker to a microtubule-disrupting agent, monomethyl auristatin E (MMAE). Brentuximab ve-

dotin derives its antitumor activity by binding the antibody-drug conjugate (ADC) to CD30-expressing cells, leading to internalization of the ADC–CD30 complex followed by proteolytic cleavage and release of MMAE. MMAE disrupts the microtubule network by binding to tubulin within the cell, thereby causing cell cycle arrest and apoptotic death of the cell.4 The key phase II registration study of brentuximab vedotin (1.8 mg/kg every 3 weeks) in relapsed/refractory cHL (n=102; overall response rate = 75%)5 eventually paved the path to the prospective, multicenter, phase III randomized AETHERA trial, in which 327 patients with cHL at increased risk of progression after autoHCT, who had been treated with a minimum of two prior systemic therapies and had achieved a complete or partial remission or had stable disease at the time of autoHCT, were randomized to receive brentuximab vedotin consolidation or placebo (every 3 weeks for up to 16 cycles) after HDT.6 Patients were eligible for the study if they met one of the following criteria: primary refractory disease following first-line therapy, first remission dur-

Table 1. Indications for brentuximab vedotin treatment of classical Hodgkin lymphoma approved by the United States Food and Drug Administration and the European Medicines Agency.

Indications for brentuximab vedotin treatment of cHL according to the FDA Previously untreated stage III/IV cHL Adult patients with previously untreated stage III/IV cHL in combination with doxorubicin, vinblastine, and dacarbazine. cHL post-autoHCT consolidation1 Adult patients with cHL at high risk of relapse or progression as post autoHCT consolidation. Relapsed/refractory cHL Adult patients with cHL after failure of autoHSCT or after failure of at least two prior multi-agent chemotherapy regimens in patients who are not auto-HSCT candidates. Indications for brentuximab vedotin treatment of cHL according to the EMA Previously untreated stage III/IV cHL Adult patients with previously untreated CD30+ stage IV cHL in combination with doxorubicin, vinblastine and dacarbazine. cHL post-autoHCT consolidation* Treatment of adult cHL patients at increased risk of relapse or progression following autoHCT. Relapsed/refractory cHL 1. following autoHCT, or 2. following at least two prior therapies when autoHCT or multi-agent chemotherapy is not a treatment option. *Increased risk defined for the AETHERA trial as either primary refractory cHL, relapsed cHL with an initial remission duration of <12 months, or extranodal involvement at the start of pre-transplantation salvage chemotherapy. cHL: classical Hodgkin lymphoma; FDA: Food and Drug Administration; EMA: European Medicines Agency; autoHCT: autologous hematopoietic stem cell transplantation. Haematologica | 107 July 2022

1500

EDITORIAL

M. Hamadani

ation of less than 12 months, or extranodal involvement at the start of salvage chemotherapy. After a median followup of 30 months, the estimated 2-year rate of progression-free survival by independent review was 63% (95% confidence interval [95% CI]: 55–70) in the brentuximab vedotin group and 51% (95% CI: 43–59) in the placebo group (hazard ratio = 0.57; 95% CI: 0·40–0·81). Overall survival was 88% at 2 years in both arms. Of note, the AETHERA trial only enrolled cHL patients who had not previously received treatment with brentuximab vedotin. However, since publication of the results of AETHERA, brentuximab vedotin (at a dose of 1.2 mg/kg up to a maximum of 120 mg every 2 weeks in combination with doxorubicin, vinblastine, and dacarbazine) has become approved in the frontline setting based on the results of ECHELON17 (Table 1) and is being increasingly used in preautoHCT salvage regimens.8 Accordingly, in current clinical practice, the proportion of brentuximab vedotin-naïve highrisk cHL patients undergoing autoHCT is declining. While international consensus guidelines suggest the use of postautoHCT brentuximab vedotin maintenance in high-risk cHL patients with limited prior exposure to brentuximab vedotin (defined as approximately ≤4-6 cycles),9 these recommendations lack supportive evidence. In a Letter to the Editor published in this issue of Haematologica, Marouf et al.10 report a retrospective nationwide French cohort (AMAHRELIS) study, examining the real-life outcome of cHL patients who received posttransplant brentuximab vedotin maintenance (n=115) during 2012-2017. Since brentuximab vedotin received European Commission approval in 2016,11 it is not clear whether these patients were treated on local clinical trials or had access through compassionate use protocols. Compared to AETHERA, more patients in the AMAHRELIS cohort received escalated BEACOPP in firstline (16% vs. 37%), had a negative positron emission tomography scan prior to autoHCT (47% vs. 82%) and underwent more than one salvage therapies (43% vs. 51%). Ninety-five percent of patients in the AMAHRELIS cohort met the AETHERA definition of high-risk disease.

More importantly, 70% (n=81) of patients in the AMAHRELIS cohort had been exposure to brentuximab vedotin prior to autoHCT. The mean number of brentuximab vedotin doses administered after autoHCT was 11 (range, 318), without difference between patients who had and had not been previously exposed to brentuximab vedotin. Treatment-related events led to maintenance discontinuation in 10% of patients, which is surprising, since nearly a third of brentuximab vedotin-naïve AETHERA patients discontinued treatment due to adverse events.8 The 2year progression-free survival of patients in AMAHRELIS was 75% and was not affected by pre-transplant exposure to brentuximab vedotin. These rates seem to be numerically higher than rates reported in the AETHERA study; a potential explanation for these improved outcomes is a higher proportion of patients with complete metabolic remission in the retrospective cohort prior to transplantation. With limitations of a retrospective cohort in mind, these results are noteworthy and provide much needed evidence supporting the use of brentuximab vedotin maintenance after autoHCT in cHL patients with prior exposure to this agent. In the recent past, nonrandomized data on post-transplant maintenance with checkpoint inhibitors have shown a high degree of disease control.12 Whether checkpoint inhibitors are superior to brentuximab vedotin in the post-transplant maintenance space merits examination, especially when considering the recent results of the KEYNOTE-204 trial showing a progression-free survival benefit of pembrolizumab compared with brentuximab vedotin, in patients with cHL who have relapsed after autoHCT or are ineligible for autoHCT.13 Disclosures MH has received research support or funding from Takeda Pharmaceutical Company and Spectrum Pharmaceuticals; has provided consultancy services for Incyte Corporation, ADC Therapeutics, Omeros, MorphoSys, Kite, Genmab, SeaGen, and Gamida Cell; and has participated in speakers’ bureaus for Sanofi Genzyme, AstraZeneca, and BeiGene.

References 1. Linch DC, Winfield D, Goldstone AH, et al. Dose intensification with autologous bone-marrow transplantation in relapsed and resistant Hodgkin's disease: results of a BNLI randomised trial. Lancet. 1993;341(8852):1051-1054. 2. Schmitz N, Pfistner B, Sextro M, et al. Aggressive conventional chemotherapy compared with high-dose chemotherapy with autologous haemopoietic stem-cell transplantation for relapsed chemosensitive Hodgkin's disease: a randomised trial. Lancet. 2002;359(9323):2065-2071. 3. Satwani P, Ahn KW, Carreras J, et al. A prognostic model predicting autologous transplantation outcomes in children, adolescents and young adults with Hodgkin lymphoma. Bone Marrow Transplant. 2015;50(11):1416-1423.

4. Doronina SO, Toki BE, Torgov MY, et al. Development of potent monoclonal antibody auristatin conjugates for cancer therapy. Nat Biotechnol. 2003;21(7):778-784. 5. Younes A, Gopal AK, Smith SE, et al. Results of a pivotal phase II study of brentuximab vedotin for patients with relapsed or refractory Hodgkin's lymphoma. J Clin Oncol. 2012;30(18):2183-2189. 6. Moskowitz CH, Nademanee A, Masszi T, et al. Brentuximab vedotin as consolidation therapy after autologous stem-cell transplantation in patients with Hodgkin’s lymphoma at risk of relapse or progression (AETHERA): a randomised, double-blind, placebo-controlled, phase 3 trial. Lancet. 385(9980):1853-1862. 7. Connors JM, Jurczak W, Straus DJ, et al. Brentuximab vedotin

Haematologica | 107 July 2022

1501

EDITORIAL

M. Hamadani

with chemotherapy for stage III or IV Hodgkin’s lymphoma. N Engl J Med. 2017;378(4):331-344. 8. Moskowitz AJ, Schöder H, Yahalom J, et al. PET-adapted sequential salvage therapy with brentuximab vedotin followed by augmented ifosamide, carboplatin, and etoposide for patients with relapsed and refractory Hodgkin’s lymphoma: a non-randomised, open-label, single-centre, phase 2 study. Lancet Oncol. 2015;16(3):284-292. 9. Kanate AS, Kumar A, Dreger P, et al. Maintenance therapies for Hodgkin and non-Hodgkin lymphomas after autologous transplantation: a Consensus Project of ASBMT, CIBMTR, and the Lymphoma Working Party of EBMT. JAMA Oncol. 2019;5(5):715-722. 10. Marouf A, Cottereau AS, Kanoun S, et al. Outcomes of refractory or relapsed Hodgkin lymphoma patients with post autologous

stem cell transplantation brentuximab vedotin maintenance : a French multicenter observational cohort study. Haematologica. 2022;107(7):1685-1690. 11. https://www.takedaoncology.com/en/news/newsreleases/takeda-receives-european-commission-approval-of-ad cetris-brentuximab-vedotin-for-consolidation-treatment-inpost-transplant-hodgkin-lymphoma/Last accessed Nov 18, 2021. 12. Armand P, Chen Y-B, Redd RA, et al. PD-1 blockade with pembrolizumab for classical Hodgkin lymphoma after autologous stem cell transplantation. Blood. 2019;134(1):22-29. 13. Kuruvilla J, Ramchandren R, Santoro A, et al. Pembrolizumab versus brentuximab vedotin in relapsed or refractory classical Hodgkin lymphoma (KEYNOTE-204): an interim analysis of a multicentre, randomised, open-label, phase 3 study. Lancet Oncol. 2021;22(4):512-524.

Haematologica | 107 July 2022

1502

REVIEW ARTICLE

How I diagnose and treat chronic myelomonocytic leukemia Mrinal M. Patnaik

Correspondence:

Division of Hematology, Department of Internal Medicine, Mayo Clinic, Rochester, MN, USA

Mrinal M. Patnaik patnaik.mrinal@mayo.edu Received: December 20, 2021. Accepted: February 24, 2022. Prepublished: March 3, 2022. https://doi.org/10.3324/haematol.2021.279500 ©2022 Ferrata Storti Foundation Haematologica material is published under a CC-BY-NC license

Abstract Chronic myelomonocytic leukemia (CMML) is a myelodysplastic syndrome/myeloproliferative overlap neoplasm characterized by sustained peripheral blood monocytosis and an inherent risk for transformation to acute myeloid leukemia (15-30% over 3-5 years). While CMML is morphologically classified into CMML-0, 1 and 2 based on peripheral blood and bone marrow promonocyte/blast counts, a more clinically relevant classification into dysplastic and proliferative subtypes, based on the presenting white blood cell count, is helpful in prognostication and therapeutics. CMML is a neoplasm associated with aging, occurring on the background of clonal hematopoiesis, with TET2 and SRSF2 mutations being early initiating events. The subsequent acquisitions of ASXL1, RUNX1, SF3B1 and DNMT3A mutations usually give rise to dysplastic CMML, while ASXL1, JAK2V617F and RAS pathway mutations give rise to proliferative CMML. Patients with proliferative CMML have a more aggressive course with higher rates of transformation to acute myeloid leukemia. Allogeneic stem cell transplant remains the only potential cure for CMML; however, given the advanced median age at presentation (73 years) and comorbidities, it is an option for only a few affected patients (10%). While DNA methyltransferase inhibitors are approved for the management of CMML, the overall response rates are 40-50%, with true complete remission rates of <20%. These agents seem to be particularly ineffective in proliferative CMML subtypes with RAS mutations, while the TET2mutant/ASXL1wildtype genotype seems to be the best predictor for responses. These agents epigenetically restore hematopoiesis in responding patients without altering mutational allele burdens and progression remains inevitable. Rationally derived personalized/targeted therapies with disease-modifying capabilities are much needed.

Introduction Chronic myelomonocytic leukemia (CMML) is a myeloid neoplasm characterized by sustained peripheral blood monocytosis (absolute monocyte count ≥1x109/L, with monocytes accounting for ≥10% of the white blood cells), predominantly arising in the context of age-related clonal hematopoiesis, with overlapping features of myelodysplastic syndromes (MDS) and myeloproliferative neoplasms (MPN).1-3 The exact incidence and prevalence rates for CMML are hard to define, with Surveillance, Epidemiology, and End Results (SEER) registry data demonstrating an incidence of 0.4 cases per 100,000, with most studies showing a clear male preponderance.4-7 The median age of presentation for CMML patients is between 70-75 years, with “young CMML” patients being operationally defined as those who present at <65 years of age.4-8. At the genome level CMML is relatively homogeneous,

demonstrating approximately 10-12 somatic variants per kilobase of coding region, with most pathogenic variants involving TET2 (60%), ASXL1 (40%), SRSF2 (50%) and RAS pathway (30%) genes. However, clinically the disease is very heterogenous in presentation and outcomes, making diagnostic, prognostic and therapeutic decision-making challenging.2,3,9-11 Broadly, CMML can be classified into dysplastic CMML (dCMML), presenting with cytopenias and clinical signs and symptoms related to the same (fatigue, bruising and transfusion dependence) and proliferative CMML (pCMML), presenting with significant myeloproliferation, extramedullary hematopoiesis and associated constitutional symptoms (fever, weight loss, night sweats, anorexia, pruritus, bone pain and cachexia).10 From a classification perspective, for several years CMML was classified as a subtype of MDS, with the World Health Organization (WHO) rightfully and formally classifying CMML as an MDS/MPN overlap neoplasm, from 2002 on-

Haematologica | 107 July 2022

1503

REVIEW ARTICLE - How I diagnose and treat CMML

M.M. Patnaik

wards.1 In 2015, the International Working Group for MDS/MPN overlap neoplasms proposed CMML-specific disease response criteria, providing support for the recognition of CMML as a specific disease entity and providing impetus for CMML-specific clinical trials.12 These changes have clearly incentivized the development of disease-specific diagnostic, prognostic, and therapeutic strategies for patients with CMML. In this review, I discuss my approach to the diagnosis, prognosis, and management of patients with CMML.

JAK2V617F), epigenetic regulation (SETBP1, DNMT3A and EZH2), transcription factors (RUNX1) and pre mRNA splicing (SF3B1 and U2AF1), shaping clinical phenotypes (Figure 1).2,3,13 There are pCMML subtypes in which oncogenic RAS pathway mutations (NRAS, CBL, KRAS and PTPN11) are clear initiating driver mutations, occurring early in the course of disease, associated with poor outcomes (decreased survival and higher rates of transformation to acute myeloid leukemia [AML]) (Figure 1).10,14 Unlike in MDS, MPN or AML, TP53 mutations are extremely infrequent in CMML (<1%), and are only really encountered at the time of CMML to AML transformation, or in the context of ther11,15,16 Diagnosis and differential diagnosis apy-related CMML. The 2016 iteration of the WHO classification of myeloid of chronic myelomonocytic leukemia: neoplasms has outlined diagnostic criteria for CMML which CMML is a neoplasm associated with aging, often arising include the presence of sustained peripheral blood monoin the background of clonal hematopoiesis (bi-allelic TET2, cytosis, absence of reactive causes, the presence of <20% or TET2/SRSF2 mutations), with the subsequent acquisi- blasts and promonocytes (blast equivalents) in the periphtion of mutations involving signaling (RAS pathway or eral blood and bone marrow, and exclusion of molecularly

Figure 1. Clonal evolutionary dynamics in patients with chronic myelomonocytic leukemia. The dynamics of clonal evolution in patients with chronic myelomonocytic leukemia (CMML) demonstrating the early acquisition of TET2 and SRSF2 mutations in hematopoietic stem cells and common myeloid progenitor cells, followed by acquisition of signaling mutations (NRAS, KRAS, CBL, PTPN11, JAK2V617F), mutations in additional epigenetic regulator genes (ASXL1, EZH2, DNMT3A) and splicing components (SF3B1, U2AF1) resulting in dysplastic and proliferative subtypes of CMML. HSC: hematopoietic stem cell; CMP: common myeloid progenitor cell; CHIP: clonal hematopoiesis of indeterminate potential; dCMML: dysplastic CMML; pCMML:proliferative CMML; AML: acute myeloid leukemia; SCNA: somatic copy number alterations. Haematologica | 107 July 2022

1504

REVIEW ARTICLE - How I diagnose and treat CMML

M.M. Patnaik

Table 1. 2017, World Health Organization criteria for the diagnosis of chronic myelomonocytic leukemia. 1

Persistent peripheral blood monocytosis ≥1 x109/L, with monocytes accounting for ≥10% of the white blood cell count

Not meeting criteria for BCR::ABL1 rearranged chronic myeloid leukemia, primary myelofibrosis, polycythemia vera and essential thrombocythemia *

No evidence for PDGFRA, PDGFRB, or FGFR1 rearrangements or the PCM1-JAK2 fusion

<20% blasts in the peripheral blood and bone marrow**

Dysplasia in one or more myeloid lineages. If myelodysplasia is absent or minimal, the diagnosis of CMML can still be made if the other requirements are met and

An acquired clonal cytogenetic or molecular genetic abnormality is present in hematopoietic cells***, or

The monocytosis has persisted for >3 months and all other causes of monocytosis have been excluded

*Cases of myeloproliferative neoplasms can be associated with monocytosis or monocytosis can develop during the disease. These cases may simulate chronic myelomonocytic leukemia (CMML). In these rare instances, a previous documented history of myeloproliferative neoplasm (MPN) excludes CMML, whereas the presence of MPN features in the bone marrow and/or of MPN-associated mutations (JAK2, CALR, or MPL) tend to support MPN with monocytosis rather than CMML. **Blasts and blast equivalents include myeloblasts, monoblasts, and promonocytes. Promonocytes are monocytic precursors with abundant light gray or slightly basophilic cytoplasm with a few scattered, fine lilaccolored granules, finely distributed, stippled nuclear chromatin, variably prominent nucleoli, and delicate nuclear folding or creasing. Abnormal monocytes, which can be present both in the peripheral blood and bone marrow, are excluded from the blast count. ***Presence of mutations in genes often associated with CMML (e.g., TET2, SRSF2, ASXL1, and SETBP1) in the proper clinical context can be used to support a diagnosis. It should be noted however, that many of these mutations can be age-related or be present in subclones. Therefore, caution would have to be used in the interpretation of these genetic results.

defined myeloid neoplasms that can present with monocytosis (BCR-ABL1, PDGFRA, PDGFRB, FGFR1 and PCM1-JAK2 rearrangements), with or without dysplasia (Table 1).1 In the absence of dysplasia, a diagnosis of CMML can be made if the monocytosis has persisted for ≥3 months, reactive causes have been excluded, or somatic cytogenetic or molecular markers frequent in CMML (e.g., ASXL1, TET2, SRSF2 and SETBP1) can be documented. While this approach is very reasonable, there are limitations and important nuances associated with these criteria. While absolute monocytosis is uncommon in chronic myeloid leukemia, it can occur in BCR-ABL1 p190 isoformdriven disease and, regardless of the presence or absence of a “myelocyte bulge”, assessment for BCR-ABL1 fusions by fluorescence in situ hybridization, cytogenetics, and/or molecular techniques should be pursued.17 Chromosomal translocations/rearrangements involving PDGFRA and PDGFRB can give rise to myeloid neoplasms, often characterized by prominent eosinophilia and responsiveness to imatinib.18-20 Among these, PDGFRB-rearranged myeloid neoplasms can be associated with absolute monocytosis (<1% of all cases morphologically diagnosed as CMML), usually with concomitant eosinophilia, and given their unique responsiveness to imatinib are best classified as molecularly defined neoplasms and not as CMML.18-20 More than 20 fusion partner genes have been described with PDGFRB, with t(5;12)(q31-q32;p13), giving rise to the ETV6(TEL)-PDGFRB fusion, being the most common.21 The FIP1L1-PDGFRA fusion arising due to the CHIC2 deletion is the most common PDGFRA aberration and is uncommonly associated with monocytosis.19 While most PDGFRB re-

arrangements can be identified by conventional karyotyping, the FIP1L1-PDGFRA fusion is karyotypically occult and can only be detected by fluorescence in situ hybridization or molecular analyses. Similarly, FGFR1 and PCM1-JAK2 rearrangements are very uncommon causes of monocytosis and are more commonly associated with eosinophilia. Monocytosis can occur in the context of other myeloid neoplasms such as MDS and MPN and is associated with poor outcomes in MPN.22,23 While monocytosis in MDS can be a reflection of an ongoing evolutionary trajectory to CMML (oligo-monocytic CMML), in MPN, the utilization of monocyte repartitioning flow cytometry (discussed below) and driver mutation status can help differentiate CMML from MPN with monocytosis.23,24 Among classical MPNdriver mutations, while JAK2V617F can occur in 10% of CMML patients, mutations involving MPL and CALR are extremely infrequent and their detection should raise questions with regards to a bona fide CMML diagnosis.3,25 Occasionally NPM1 and FLT3 driver mutations are identified in CMML patients with excess blasts (5-19%).26,27 For all practical purposes I consider these cases as acute myelomonocytic leukemia in evolution and treat them as such. Reactive monocytosis is very common in practice and while viral infections and recovering bone marrow (from injury, drugs or chemotherapy) are frequent causes, sustained reactive monocytosis is more common in chronic infections such as subacute bacterial endocarditis, tuberculosis, brucellosis, leishmaniasis and leprosy and in autoimmune/inflammatory disorders such as systemic lupus erythematosus, sarcoidosis and mixed connective tissue

Haematologica | 107 July 2022

1505

REVIEW ARTICLE - How I diagnose and treat CMML

M.M. Patnaik

disorder.28 Reactive monocytosis can also be seen in the context of metastatic visceral neoplasms, either due to enhanced mobilization of monocytes from the bone marrow, or due to increased monopoiesis mediated by CCL2 (C-C motif chemokine ligand 2).29

Role of flow cytometry, next-generation sequencing and bone marrow biopsies in the diagnosis of chronic myelomonocytic leukemia Conventional flow cytometry has a limited diagnostic role in CMML, given that immature/neoplastic monocytes do not express unique surface markers and that promonocytes/monoblasts are frequently CD34-negative (blast marker).30 Flow abnormalities that can be detected in CMML include abnormal myeloid maturation patterns involving CD11b, CD13 and CD16, along with aberrant expression of CD56 on monocytes.30 Monocyte repartitioning by flow cytometry has gained popularity, especially given its ability to differentiate CMML from other reactive and clonal causes of monocytosis.31-33 Based on the expression of CD14 and CD16, monocytes can be divided into three categories: CD14+/CD16- classical (M01), CD14low/CD16+ intermediate (M02), and CD14-/CD16+ non-classical monocytes (M03) (Figure 2).31,32 These subsets differ in their chemokine receptor expression, phagocytic activity, epigenetic profiles and have unique metabolic pathway dependencies.32,34 In CMML, a pivotal study demonstrated an increase in the M01 subset, with an established cutoff >94% being associated with sensitivity and specificity values of 90.6% and 95.1%, respectively.32 These findings have been validated independently and this method importantly is effective in identifying MDS patients whose disease eventually evolves into CMML and in distinguishing CMML from MPN with monocytosis.31,33,35,36 False negative findings secondary to autoimmunity/inflammation (expansion of the M02 fraction) and false positive findings in myeloid neoplasms such as MDS, atypical chronic myeloid leukemia and classical chronic myeloid leukemia have been documented.33,37,38 I do use this flow cytometry assay for screening patients who present with sustained monocytosis, especially when there is suspicion of an underlying clonal process and to differentiate CMML-associated monocytosis from other myeloid neoplasms with monocytosis.39 Next-generation sequencing assays are an important part of the diagnosis and prognostication of CMML. Virtually all CMML patients will have detectable somatic mutations involving genes regulating the epigenome, splicing, signaling and transcription.4,5,11,40 Clonal compositions help to define pCMML and dCMML subtypes and contribute towards AML

transformation.10 Single-cell sequencing data have shown that TET2 mutations are usually the founder mutations occurring at the hematopoietic stem cell level.41 These mutations impact multipotent progenitor and common myeloid progenitor cells, skewing differentiation towards granulomonocytic progenitors and mature monocytes, respectively (Figure 1).41 Second-order mutations tend to accumulate in multipotent progenitor and common myeloid progenitor cells, often involving additional sites/alleles on TET2, spliceosome component genes (SRSF2, rarely SF3B1 and U2AF1) and additional epigenetic regulators (ASXL1, rarely EZH2).11,41-43 Signaling mutations, such as NRAS, CBL, PTPN11, KRAS, NF1 and JAK2V671F, can sometimes be early/founder mutations, with inherent hypersensitivity of hematopoietic stem/progenitor cells to granulocyte-macrophage colony-stimulating factor (like the pCMML pediatric counterpart, juvenile myelomonocytic leukemia), but can also be later/subclonal events, giving rise to pCMML.10,41,44 Similarly, acquisition of additional epigenetic (DNMT3A, rarely IDH1/2) and splicing mutations (SF3B1) and mutations involving transcription factors (RUNX1) often gives rise to a dCMML phenotype.40,45 Somatic copy number alterations, especially copy neutral loss of heterozygosity, is common in CMML and frequently involves TET2 (4q24) and CBL (11q23), playing a role in clonal evolution.10 Genetic alterations involving protein coding regions, including copy number gains and losses in driver mutations were only able to explain 44% of CMML cases that transformed to AML, indicating that mechanisms of

Figure 2. Flow cytometric analysis of monocyte repartitioning. Flow cytometry demonstrating a markedly expanded M01 (CD14+/CD16–) monocyte fraction, approximating 98.79%, in a patient with chronic myelomonocytic leukemia. M01: classical monocytes; M02: intermediate monocytes; M03: non-classical monocytes.

Haematologica | 107 July 2022

1506

REVIEW ARTICLE - How I diagnose and treat CMML

M.M. Patnaik

AML transformation remain to be elucidated.10 It is important to note that germline variants have been implicated in the pathogenesis of CMML, with cases having been documented in the context of germline mutations involving RUNX1, ANKRD26, ETV6, CHEK2, CDK2NA and GATA2.46,47 Recently, a 700 kb germline duplication localized to 14.q32.2, resulting in overexpression of ATG2B and GSKIP genes, was described and has been associated with familial MPN and CMML.48 Based on family histories, age

of onset and the heterozygous nature of variant allele fractions, I routinely assess germline DNA from extracted hair follicles/skin fibroblasts to assess for germline predisposition syndromes.49 In CMML, bone marrow biopsies are often hypercellular with granulocytic hyperplasia and mild to modest dysplasia (Figure 3D, E). Bone marrow monocytosis can be present, but is often difficult to appreciate and immunohistochemical studies that aid in the identification of monocytes and

Figure 3. Peripheral blood and bone marrow findings in patients with chronic myelomonocytic leukemia. (A) Peripheral blood smear with hypogranular neutrophil (black arrow) and promonocytes (black arrowhead). (B) Bone marrow aspirate with promonocytes (black arrowhead). (C) Butyrate esterase/chloroacetate esterase stain demonstrates numerous butyrate esterase-positive monocytes, as well as dual esterase-positive cells (inset). (D) Bone marrow aspirate with increased blasts. (E) Bone marrow biopsy demonstrating a hypercellular bone marrow. (F) Small plasmacytoid dendritic cell nodules CD123+ (left) and CD303+ (right). Haematologica | 107 July 2022

1507

REVIEW ARTICLE - How I diagnose and treat CMML

M.M. Patnaik

their precursors are recommended (Figure 3C).50 Almost 80% of patients will demonstrate micro-megakaryocytes with abnormal nuclear contours and lobations, and 20-30% of patients can have an increase in reticulin fibrosis.50 Approximately 30% of patients demonstrate nodules composed of mature plasmacytoid dendritic cells that are clonal (CD123+,lineage-negative, CD45+, CD11c–, CD33–, HLADR+, BDCA-2+ and BDCA-4+), often have RAS pathway mutations and predict for an inferior AML-free survival (Figure 3F).51 The identification of promonocytes requires expertise and these cells should be summated with blasts when estimating the blast count (Figure 3A, B).52 Promonocytes are described as monocytic precursors that have a delicately convoluted, folded or grooved nucleus with finely dispersed chromatin, a small indistinct or absent nucleolus, and finely granulated cytoplasm.52,53 On immunophenotyping the abnormal bone marrow cells often express myelomonocytic antigens such as CD13 and CD33, with variable expression of CD14, CD68 and CD64. Markers of aberrant expression include CD2, CD15, and CD56 or decreased expression of CD13, CD14, HLA-DR, CD64 or CD36. The presence of myeloblasts can often be detected by expression of CD34. The most reliable markers on immunohistochemistry include CD68R and CD163. On cytochemical analysis, monocytes are often positive for non-specific esterases and lysozyme, while the granulocytic precursors are often positive for lysozyme and chloroacetate esterase (Figure 3C). This technique can help to differentiate CMML from other overlap neoplasms in which bone marrow monocytosis is uncommon. Conventional karyotyping on the bone marrow is important for cytogenetic risk stratification, with approximately 70-80% of patients demonstrating a normal karyotype. Common abnormalities include +8 and –Y, with isolated del5q, complex and monosomal karyotypes being very uncommon (complex and monosomal karyotypes can be seen in patients with therapy-related CMML).16,54,55 The CMML-specific prognostic scoring system (CPSS) and the Mayo-French cytogenetic risk stratification system are two commonly used karyotype-based prognostic models for patients with CMML (Table 2).54,55

entities with monocytosis. Oligomonocytic CMML. This category encompasses patients who present with sustained relative monocytosis (≥10% of white blood cells) and absolute monocytosis not meeting current diagnostic criteria for CMML (absolute monocyte count 0.5-<1.0x109/L).24 Based on the 2016 WHO classification, these patients would be classified as having either MDS or MDS/MPN-Unclassifiable.1 Except for an absolute monocyte count of ≥1x109/L, if these patients meet other CMML diagnostic criteria, along with a M01 fraction >94% on monocyte repartitioning flow cytometry, and a molecular signature consistent with CMML (TET2, SRSF2, and ASXL1), I consider them as having oligomonocytic CMML and follow and manage them as such. Over time, several of these patients will have clonal evolution to either CMML or secondary AML. CMML associated with a concomitant myeloid neoplasm. This category includes several variants, with the two most important being systemic mastocytosis (SM) with CMML (SM-CMML) and JAK2V617F-mutant CMML24 CMML is the most frequent hematologic neoplasm associated with SM. I diagnose SM-CMML in patients who meet WHO criteria for both entities. In these patients, the SM component can present as either indolent or aggressive SM, with mast cell leukemia being extremely infrequent.56 In a Mayo Clinic study of 50 patients with SM-CMML, survival outcomes were similar to those of patients with CMML (24 months for SM-CMML vs. 18 months for CMML; P=0.08), There was a higher frequency of KIT and CBL mutations in SM-CMML and CMML-based prognostic models were not effective in risk stratification in this condition due to the confounding impact of SM.56 In CMML patients who have a detectable KITD816V mutation, a known driver oncogene in SM (>90% of cases), I universally assess for concomitant SM clinically and by using laboratory techniques such as serum tryptase levels (>20 ng/mL), bone marrow morphology, flow cytometry (aberrant expression of CD2 and/or CD25 on mast cells) and immunohistochemistry (CD117 and tryptase). For SM-CMML patients there are several exciting KIT-directed targeted therapies for the SM component (e.g., midostaurin and avapritinib) and given that in SMCMML, neoplastic monocytes are also often KIT mutated,57 these drugs can sometimes effectively decrease monocyte How I approach chronic counts and infiltrative disease burden.58 JAK2V617F-mumyelomonocytic leukemia variants tant CMML usually gives rise to a pCMML subtype with and molecularly defined entities higher hemoglobin levels and absolute monocyte counts, with monocyte repartitioning by flow being a useful tool with monocytosis to distinguish this entity from MPN with monocytosis.59 On The diagnosis and management of CMML variants and rare occasions, it becomes very difficult to distinguish the molecularly defined entities presenting with monocytosis, concomitant presence of a JAK2V617F MPN with bona fide mimicking CMML, require special attention. These variants CMML and in these instances, I manage each symptomatic can broadly be divided into three categories: (i) oligo- component individually. monocytic CMML; (ii) CMML associated with a concomi- Molecularly defined entities with monocytosis. This cattant myeloid neoplasm; and (iii) molecularly defined egory includes myeloid and lymphoid neoplasms presHaematologica | 107 July 2022

1508

REVIEW ARTICLE - How I diagnose and treat CMML

M.M. Patnaik

enting with monocytosis in the context of rearrangements involving PDGFRA, PDGFRB, FGFR1 and PCM1-JAK2. These entities are best designated by their molecular drivers and are currently not referred to as CMML. The

recognition of these entities is very important given that they can present with a lymphoproliferative component and eosinophilia and can be targeted with tyrosine kinase inhibitors.19,60

Table 2. Chronic myelomonocytic leukemia risk stratification models.

OS: overall survival; yr: year; mo: months; MS: median survival; AMC: absolute monocyte count, IMC: immature myeloid cells; WBC: white blood cell count; WHO_ World Health Organization; FAB: French American and British classification; WT: wild type, MT: mutant, AML-TR: acute myeloid leukemia transformation rate, RBC transf: red blood cell transfusion; CMML: chronic myelomonocytic leukemia; MDS: myelodysplastic syndrome; MPN: myeloproliferative neoplasm. Haematologica | 107 July 2022

1509

REVIEW ARTICLE - How I diagnose and treat CMML

How I prognosticate outcomes for patients with chronic myelomonocytic leukemia

M.M. Patnaik

While there are numerous prognostic models for patients with CMML, models such as the Bournemouth, Lille, International Prognostic Scoring System (IPSS) and revisedIPSS scores are primarily designed for patients with MDS and excluded patients with pCMML.61,62 The MD Anderson prognostic system (MDAPS) is CMML-specific and identified a hemoglobin level <12 g/dL, presence of circulating immature myeloid cells (myelocytes, promyelocytes and metamyelocytes), absolute lymphocyte count >2.5x109/L and ≥10% bone marrow blasts as independent predictors for inferior survival.63 The Global MDAPS was then developed for patients with MDS, secondary MDS and CMML, with prognostic factors including; older age, poor performance status, thrombocytopenia, anemia, increased bone marrow blasts, leukocytosis (>20x109/L), chromosome 7 or complex cytogenetic abnormalities and a prior history of red blood cell transfusions.64 This model identified four prognostic groups with median survivals of 54 (low), 25 (intermediate-1), 14 (intermediate-2) and 6 months (high), respectively.64 The CPSS model was developed in Europe and identified pCMML versus dCMML subtypes, WHO CMML-subtypes, red blood cell transfusion dependency and the CPSS cytogenetic risk stratification system as being prognostic for survival.54,65 The Mayo prognostic model identified hemoglobin <10 g/dL, platelet count <100×109/L, absolute monocyte count >10×109/L and circulating immature myeloid cells as being independently prognostic.6 The discovery of somatic mutations in CMML resulted in the development of contemporary molecular prognostic models. The Group Francophone des Myelodysplasies (GFM) demonstrated an adverse prognostic effect for truncating ASXL1 mutations in 312 patients with CMML; additional risk factors included age >65 years, white blood cell count >15×109/L, platelet count <100×109/L and hemoglobin <10 g/dL in females and <11 g/dL in males.5 The GFM model assigns three adverse points for white blood cell count >15×109/L and two adverse points for each one of the remaining risk factors, resulting in a three-tiered risk stratification; low (0–4 points), intermediate (5–7) and high (8–12), with respective median survivals of 56, 27.4 and 9.2 months.5 To further clarify the prognostic relevance of ASXL1 mutations, the Mayo Molecular Model (MMM) was developed as a collaborative effort between the GFM and Mayo Clinic (n=466).66 Adverse prognostic factors included truncating ASXL1 mutations, absolute monocyte count >10×109/L, hemoglobin <10 g/dL, platelets <100×109/L and circulating immature myeloid cells. Based on these variables a regression coefficient-based prognostic model was developed with the following risk cat-

egories; high (≥3 risk factors), intermediate-2 (2 risk factors), intermediate-1 (1 risk factor), and low (no risk factors) risk, with median survivals of 16, 31, 59 and 97 months, respectively.67 The CPSS model was also updated to include gene mutations involving ASXL1, RUNX1, NRAS and SETBP1 (CPSS-Mol).4 Gene mutations along with karyotypic abnormalities are used to calculate the CPSS genetic score. One point each is assigned for an intermediate-1 genetic score, white cell count ≥13x109/L, bone marrow blasts ≥5% and red blood cell transfusion dependency, two points for intermediate-2 genetic score and three points for a high risk genetic score.4 The CPSS-Mol stratifies patients into four categories, low (0 risk factors), intermediate-1 (1 risk factor), intermediate-2 (2-3 risk factors) and high (≥4 risk factors) risk, with median overall survival not reached, 64, 37 and 18 months; with 4-year leukemic transformation rates of 0%,3%, 21% and 48%, respectively.4 Table 2 highlights relevant CMML-specific prognostic models along with their component variables. Recently a CMML transplant model was developed (CMML transplant score) which assigned four points for the presence of ASXL1 and/or NRAS mutations, four points for bone marrow blasts >2% and one point for each hematopoietic stem cell transplant (HSCT) comorbidity index, effectively stratifying for both overall survival and non-relapse mortality.68 In practice, any of the three molecularly integrated CMML prognostic models can be used for risk stratification. While these models have not been formally compared against each other, they share several overlapping prognostic features, especially anemia, elevated white blood cell counts and truncating ASXL1 mutations.4,5,11 I use both the MMM and the CPSS-Mol model for risk stratification at my institution.

How I manage patients with chronic myelomonocytic leukemia The first step in the management of CMML patients is establishing an accurate diagnosis, followed by personalized risk stratification. Using any one of the molecularly integrated prognostic models, CMML patients can be stratified into lower risk and higher risk groups (Figure 4). Management strategies for these two risk groups are described below. Lower-risk chronic myelomonocytic leukemia I define lower-risk CMML patients as those who fall into low and intermediate-1 risk categories based on the MMM and the CPSS-Mol, or the low-risk category of the GFM model. On average these patients have a median overall survival of 60-100 months4,5,11 and the following treatment strategies can be adopted for their care:

Haematologica | 107 July 2022

1510

REVIEW ARTICLE - How I diagnose and treat CMML

M.M. Patnaik

Figure 4. Management algorithm for chronic myelomonocytic leukemia based on risk stratification using the Mayo Molecular Model (Patnaik MM et al. Leukemia 2014). CMML: chronic myelomonocytic leukemia, DNMTi: DNA methyltransferase inhibitors, X: new clinical trial investigational agent.

Observation and supportive care. Several low-risk patients can be observed without any CMML-directed therapies, with serial blood count measurements and management of symptoms as needed. There are some data to suggest that in pCMML, permissive leukocytosis/monocytosis can be associated with increased lysozyme levels and a higher prevalence of chronic kidney disease (lysozyme nephropathy), and in select asymptomatic pCMML patients with white blood cell counts >30x109/L, I do recommend hydroxyurea to control blood counts and proliferative features.69 Side effects associated with hydroxyurea include nausea, vomiting, diarrhea, fever, mouth sores, photosensitivity, myelosuppression, and chronic non-healing leg/ankle ulcers. Anemia. Ineffective erythropoiesis contributes to significant morbidity and mortality in CMML. While there are limited CMML-specific prospective data for anemia management, akin to MDS, the management of anemia largely centers around the use of red blood cell transfusions and erythropoiesis-stimulating agents. Extrapolating from MDS-based trials, in CMML, erythropoiesis-stimulating agents are also more likely to be effective in lower-risk patients, especially those with endogenous erythropoietin levels <200 U/L and those with low or no dependency on red blood cell transfusions (40-70% response rates).70-72 The median duration of response to erythropoiesis-stimulating agents is 12-18 months, with limited options after progression. I usually use fixed doses of recombinant human erythropoietin or darbepoetin and strictly avoid the use of granulocyte – colony-stimulating factor) given the higher baseline risk of splenic rupture in CMML patients.73 I closely monitor for

adverse vascular side effects associated with erythropoiesis-stimulating agents, such as treatment-emergent hypertension and thromboembolism and do not administer these agents when hemoglobin levels are >11 g/dL. Luspatercept is a recombinant fusion protein that traps GDF 11 and activin ligands belonging to the TGF-β superfamily, decreasing SMAD2 and SMAD3 signaling, enabling latestage erythroid maturation and has been approved by the Food and Drug Administration for patients with β-thalassemia and MDS-ring sideroblasts.74 While there are no clear safety or efficacy data on the use of luspatercept in CMML, I do consider off-label use in a select group of SF3B1-mutant CMML patients with bone marrow-ring sideroblasts, who are ineligible for erythropoiesis-stimulating agents or in whom these agents have failed.43 Luspatercept is in general well tolerated with side effects including headaches, bone pain, arthralgia and fatigue. We have recently defined SF3B1mutant CMML as a CMML subtype with predominant dysplastic features, with a low frequency of ASXL1 mutations, higher frequency of JAK2V61F mutations, concurrent splicing mutations, and a superior AML-free survival.43 Other options for anemia management include danazol (an anabolic steroid), lenalidomide (an immunomodulatory agent; note - isolated del5q is seen in <1% of CMML cases) and DNA methyltransferase (DNMT) inhibitors, such as 5-azacitidine, decitabine and oral decitabine combined with cedazuridine (cytidine deaminase inhibitor), given in either conventional doses, or in attenuated dose schedules.75,76 Given that these strategies for managing anemia are either suboptimal or not durable, I strongly encourage participation in clinical trials.

Haematologica | 107 July 2022

1511

REVIEW ARTICLE - How I diagnose and treat CMML

M.M. Patnaik

Splenomegaly. Symptomatic splenomegaly can be a significant issue in patients with pCMML. Clinical issues related to splenomegaly include early satiety, abdominal pain and tenderness, constitutional symptoms, referred shoulder pain, hiccoughs and mechanical obstruction of abdominal organs.77 Splenic infarction and spontaneous splenic rupture can result in abdominal catastrophes.73 I usually manage symptomatic splenomegaly, or massive splenomegaly, with cytoreductive therapy, with hydroxyurea being my first choice. There are recent encouraging data on the use of ruxolitinib, a JAK1/2 inhibitor, in patients with CMML, with 43% of CMML patients with baseline splenomegaly demonstrating a spleen response.78 Ruxolitinib was well tolerated with the two most common grade 3 and 4 treatment-related toxicities being anemia (10%) and thrombocytopenia (6%). I have used ruxolitinib off-label in select patients with good effect. DNMT inhibitors, splenic radiation and splenectomy are generally avoided, given inherent complications such as worsening cytopenias with DNMT inhibition, lack of durable responses with radiation and surgical morbidity/mortality associated with splenectomy. Thrombocytopenia. Thrombocytopenia in CMML has diverse etiologies including splenic sequestration from hypersplenism, immune-mediated thrombocytopenia and bone marrow failure from progressive disease. Autoimmune phenomena, including immune-mediated thrombocytopenia, can be seen in 20-30% of CMML patients.79,80 While corticosteroids and rituximab have been used for patients with suspected immune-mediated thrombocytopenia, the use of thrombopoietin analogs, especially eltrombopag in CMML needs caution. There are reports of pCMML patients demonstrating worsening proliferative features, circulating blasts and bone marrow fibrosis on exposure to eltrombopag.81 The GFM however has completed a yet to be published phase II trial (NCT02323178) assessing the safety of eltrombopag in CMML patients with severe thrombocytopenia (platelet count <50x109/L). In this study eltrombopag was relatively well tolerated (median dose 150 mg; range, 100-300 mg), with 46.7% of patients achieving a platelet response (10 with dCMML and 4 with pCMML) that in general was not durable (median duration 3.4 months; range, 1.7-11.6 months). I use extreme caution when prescribing eltrombopag for pCMML patients with proliferative features. Other options for thrombocytopenia include splenectomy when immunemediated thrombocytopenia or splenic sequestration is suspected and DNMT inhibitors if the etiology is diseaserelated bone marrow dysfunction/failure. Autoimmune manifestations. Autoimmune and systemic inflammatory manifestations such as erythema nodosum, leukocytoclastic vasculitis, Sweet syndrome, polymyalgia rheumatica, seronegative arthritis, and mixed connective tissue disorder-like syndromes can be seen in 20-30% of

patients, with manifestations often preceding the diagnosis of CMML.79,80 With growing evidence on the role played by clonal hematopoiesis-clones in amplifying inflammation and endothelial dysfunction, there is more understanding on the pathobiology of inflammation and autoimmunity in CMML.82-84 Cytokines whose levels are elevated in CMML patients include, IL-8, IP-10, IL-1RA, TNFα, IL-6, MCP-1/CCL2, HGF, M-CSF, VEGF, IL-4, and IL-2RA, with decreased levels of IL-10 being associated with adverse prognosis.85 The transcriptional signature of CMML monocytes is also highly inflammatory, with upregulation of multiple inflammatory pathways, including TNF-α, IL-6 and IL-17.86 While corticosteroids and steroid-sparing/disease-modifying agents are often used in the management of these symptoms, I use DNMT inhibitors in conventional or low doses, for more durable responses.80 Azathioprine is a steroid-sparing immunosuppressive agent that I strictly avoid, given its strong association with therapyrelated myeloid neoplasms. Higher-risk chronic myelomonocytic leukemia I define higher-risk CMML patients as those who fall into intermediate-2 and high-risk categories based on the MMM and the CPSS-Mol, or the high-risk category of the GFM model. On average these patients have a median overall survival <2 years4,5,11 and the following treatment strategies can be adopted for their care. Allogeneic HSCT. Allogeneic HSCT remains the only potentially curative option for patients with CMML. However, given the older age at presentation and associated morbidities, most patients are not eligible.87 At our institution, CMML patients with higher-risk disease diagnosed <75 years of age and with an acceptable HSCT-comorbidity index (deemed by an expert committee) are usually referred for allogeneic HSCT.88 A recent consensus document from an expert panel does recommend upfront allogeneic HSCT for intermediate-2 and high-risk CMML patients (CPSS risk stratification), with <10% bone marrow blasts.89 For patients with >10% bone marrow blasts, I usually cytoreduce with DNMT inhibition or AML-like induction therapy, especially in “younger and fit” patients.8,88 In the largest study to date, the European Group for Blood and Bone Marrow Transplant reported outcomes of 513 patients (median age 53 years) of whom 249 received myeloablative conditioning and 228 received reduced intensity conditioning.90 The 4-year nonrelapse mortality was 41% and relapse rate 32%, accounting for a 4-year relapse-free survival rate of 27% and an overall survival rate of 33%.90 In this study the only factor prognostic for favorable outcomes was the achievement of a complete remission prior to HSCT. Data with regard to the use of DNMT inhibitors prior to HSCT in CMML are largely retrospective and somewhat controversial, given their lack of disease-modifying efficacy and propensity to worsen existing cytopenias.91 In general, I try to avoid DNMT in-

Haematologica | 107 July 2022

1512

REVIEW ARTICLE - How I diagnose and treat CMML

M.M. Patnaik

Figure 5. Impact of mutations on survival of patients stratified according to treatment with DNA methyltransferase inhibitors. (A, B) Kaplan-Meier curves demonstrating the impact of TET2 mutations (A) and the TET2mutant/ASXL1wildtype genotype (B) on survival outcomes, stratified by treatment or no treatment with DNA methyltransferase inhibitors (HMA: hypomethylating agents) in the Mayo Clinic CMML cohort. OS: overall survival.

hibition prior to HSCT, unless it is needed to cytoreduce a patient prior to conditioning therapy. A recent Mayo Clinic study confirmed the survival benefit offered by allogeneic HSCT in higher-risk CMML patients, with 5-year overall survival rates of 51% for patients with chronic phase CMML and 19% for those with blast transformed CMML.88 Soberingly the graft-versus-host disease relapse-free survival in the chronic phase cohort was only 7 months. With a greater availability of donor sources, alternative donor transplant strategies and an improving arsenal for management of graft-versus-host disease, we hope that allogeneic HSCT becomes a viable option for a greater number of patients. DNMT inhibitors. DNMT inhibitors such as 5-azacitidine, decitabine and the oral combination of decitabine with cedazuridine are the only drugs approved for the management of CMML by the U.S Food and Drug Administration. In Europe, 5-azacitidine remains the only drug approved for the management of CMML (>10% blasts). The approval of these agents was largely based on MDS predominant trials that included a small number of CMML patients, all of whom had a white blood cell count <12x109/L (dCMML).92,93 The overall response rates to DNMT inhibitors are approximately 40-50%, with true complete remission rates of <20%.94,95 In a seminal study, elaborate sequencing data demonstrated that DNMT inhibitors induce responses in CMML patients by epigenetically restoring normal

hematopoiesis, without impacting mutational allele burdens, with disease progression to AML remaining inevitable.96 The struggle with using DNMT inhibitors in CMML is the lack of predictors or biomarkers of response. In a large multi-institutional study, we identified the presence of TET2 mutations in the absence of ASXL1 mutations being the best markers of response to DNMT inhibitors (Figure 5), similarly to the situation in MDS.97-99 In a recent prospective randomized trial assessing the efficacy of decitabine versus hydroxyurea in higher-risk pCMML patients (n=170), there was no difference in overall survival or event-free survival between the hydroxyurea and decitabine arms (NCT02214407; Itzykson et al. ASH 2020). Adverse effects of DNMT inhibitors include nausea, vomiting, diarrhea, fatigue, and myelosuppression. I proactively educate my patients about these adverse effects and optimize supportive care for the preemptive management of the same. I largely use DNTM inhibitors in dCMML subtypes with higher risk disease or clinically significant cytopenias, especially if they are TET2 mutant and ASXL1 wildtype (Figure 5). I avoid using DNMT inhibtors in pCMML patients with dominant RAS pathway mutations or with highly proliferative features. For these patients I proactively seek clinical trials with novel therapeutic agents. For patients who respond to DNMT inhibition, the median duration of re-

Haematologica | 107 July 2022

1513

REVIEW ARTICLE - How I diagnose and treat CMML

M.M. Patnaik

sponse is 12-18 months and survival after progression is dismal, ranging between 6-9 months.94 Given recent data suggesting improved survival rates in AML patients treated in combination with azacitidine and the bcl-2 inhibitor venetoclax, there has been interest in assessing this combination in CMML.100 Preclinical BH3 mimetic profiling data in CMML suggest that the malignant monocytes are addicted to mcl1 instead of bcl2 and that combination therapy with mcl1 and MAPK inhibitors might be a successful treatment strategy.101 While prospective data are awaited, a small retrospective series including CMML patients with blast transformation demonstrated suboptimal response rates (overall response rate - 50% for CMML), with no significant difference from response rates seen with DNMT inhibitors alone.102 Clinical trials. I actively seek out clinical trials for all our CMML patients, given the suboptimal response rates to conventional strategies. We have successfully carried out CMML-specific clinical trials and urge the scientific community to stop grouping CMML with MDS or MPN, given the unique biology of CMML. CMML-specific trials that have been completed or are currently accruing include trials assessing the safety and efficacy of lenzilumab (an anti-GM-CSF monoclonal antibody: NCT02546284), tipifarnib (a farnesyl transferase inhibitor: NCT02807272), ruxolitinib (a JAK1/2 inhibitor: NCT03722407), cobimetinib (a MEK inhibition: NCT04409639) and tagraxofusp (a protein conjugate involving IL-3 and truncated components of the diphtheria toxin: NCT02268253) in patients with CMML.103105 Preclinical data from our laboratory have demonstrated a unique RAS-KMT2A-PLK1 axis defining the pCMML phenotype, with PLK1 inhibition with the oral, selective PLK1 inhibitor onvansertib demonstrating excellent preclinical activity.10,106

Conclusions CMML is a unique MDS/MPN overlap neoplasm with relative genetic homogeneity, but with marked clinical heterogeneity. The disease is seen in the elderly and frequently develops on the background of clonal hematopoiesis, with recurrent somatic mutations involving TET2, ASXL1, SRSF2 and the RAS pathway defining dysplastic and proliferative subtypes of CMML. For several years CMML was considered as a subtype of MDS, but from 2002 onwards, CMML has been rightfully recognized as a unique neoplasm with the development of CMML-specific prognostic models, response criteria, preclinical models and, most importantly, clinical trials; heralding a new future for this disease and affected patients. Several challenges remain, including the lack of uniform consensus on personalized prognostication and more importantly, the identification of disease-modifying targets and therapies that might ameliorate disease progression, improve quality of life, and potentially offer a cure. Disclosures My institution has received research funding from Stem Line therapeutics and Kura Oncology. Acknowledgments I would like to acknowledge all the CMML patients who have entrusted their lives and care in my hands; my institution and mentors for supporting me and my national and international collaborators and sponsors. I would also like to acknowledge DRS, and Matthew Howard and Michael Timm from the Department of Laboratory Medicine and Pathology for helping with blood and bone marrow images and flow cytometry data, respectively.

References 1. Arber DA, Orazi A, Hasserjian R, et al. The 2016 revision to the World Health Organization classification of myeloid neoplasms and acute leukemia. Blood. 2016;127(20):2391-2405. 2. Patnaik MM, Lasho T. Myelodysplastic syndrome/myeloproliferative neoplasm overlap syndromes: a focused review. Hematology Am Soc Hematol Educ Program. 2020;2020(1):460-464. 3. Patnaik MM, Lasho TL. Genomics of myelodysplastic syndrome/myeloproliferative neoplasm overlap syndromes. Hematology Am Soc Hematol Educ Program. 2020;2020(1):450-459. 4. Elena C, Galli A, Such E, et al. Integrating clinical features and genetic lesions in the risk assessment of patients with chronic myelomonocytic leukemia. Blood. 2016;128(10):1408-1417. 5. Itzykson R, Kosmider O, Renneville A, et al. Prognostic score including gene mutations in chronic myelomonocytic leukemia. J Clin Oncol. 2013;31(19):2428-2436. 6. Patnaik MM, Padron E, LaBorde RR, et al. Mayo prognostic model for WHO-defined chronic myelomonocytic leukemia:

ASXL1 and spliceosome component mutations and outcomes. Leukemia. 2013;27(7):1504-1510. 7. Rollison DE, Howlader N, Smith MT, et al. Epidemiology of myelodysplastic syndromes and chronic myeloproliferative disorders in the United States, 2001-2004, using data from the NAACCR and SEER programs. Blood. 2008;112(1):45-52. 8. Patnaik MM, Wassie EA, Padron E, et al. Chronic myelomonocytic leukemia in younger patients: molecular and cytogenetic predictors of survival and treatment outcome. Blood Cancer J. 2015;5(2):e280. 9. Ball M, List AF, Padron E. When clinical heterogeneity exceeds genetic heterogeneity: thinking outside the genomic box in chronic myelomonocytic leukemia. Blood. 2016;128(20):2381-2387. 10. Carr RM, Vorobyev D, Lasho T, et al. RAS mutations drive proliferative chronic myelomonocytic leukemia via a KMT2APLK1 axis. Nat Commun. 2021;12(1):2901. 11. Patnaik MM, Itzykson R, Lasho TL, et al. ASXL1 and SETBP1 mutations and their prognostic contribution in chronic

Haematologica | 107 July 2022

1514

REVIEW ARTICLE - How I diagnose and treat CMML

M.M. Patnaik

myelomonocytic leukemia: a two-center study of 466 patients. Leukemia. 2014;28(11):2206-2212. 12. Savona MR, Malcovati L, Komrokji R, et al. An international consortium proposal of uniform response criteria for myelodysplastic/myeloproliferative neoplasms (MDS/MPN) in adults. Blood. 2015;125(12):1857-1865. 13. Mason CC, Khorashad JS, Tantravahi SK, et al. Age-related mutations and chronic myelomonocytic leukemia. Leukemia. 2016;30(4):906-913. 14. You X, Liu F, Binder M, et al. Asxl1 loss cooperates with oncogenic Nras in mice to reprogram immune microenvironment and drive leukemic transformation. Blood. 2022;139(7):1066-1079. 15. Bernard E, Nannya Y, Hasserjian RP, et al. Implications of TP53 allelic state for genome stability, clinical presentation and outcomes in myelodysplastic syndromes. Nat Med. 2020;26(10):1549-1556. 16. Patnaik MM, Vallapureddy R, Yalniz FF, et al. Therapy relatedchronic myelomonocytic leukemia (CMML): molecular, cytogenetic, and clinical distinctions from de novo CMML. Am J Hematol. 2018;93(1):65-73. 17. Ohtake S. Chronic myelogenous leukemia with p190BCR-ABL expression: the missing link with monocytosis. Intern Med. 2002;41(12):1092-1093. 18. Patnaik MM, Lasho TL, Finke CM, Pardanani A, Tefferi A. Targeted next generation sequencing of PDGFRB rearranged myeloid neoplasms with monocytosis. Am J Hematol. 2016;91(3):E12-14. 19. Pardanani A, Ketterling RP, Li CY, et al. FIP1L1-PDGFRA in eosinophilic disorders: prevalence in routine clinical practice, long-term experience with imatinib therapy, and a critical review of the literature. Leuk Res. 2006;30(8):965-970. 20. Apperley JF, Gardembas M, Melo JV, et al. Response to imatinib mesylate in patients with chronic myeloproliferative diseases with rearrangements of the platelet-derived growth factor receptor beta. N Engl J Med. 2002;347(7):481-487. 21. Golub TR, Barker GF, Lovett M, Gilliland DG. Fusion of PDGF receptor beta to a novel ets-like gene, tel, in chronic myelomonocytic leukemia with t(5;12) chromosomal translocation. Cell. 1994;77(2):307-316. 22. Tefferi A, Shah S, Mudireddy M, et al. Monocytosis is a powerful and independent predictor of inferior survival in primary myelofibrosis. Br J Haematol. 2018;183(5):835-838. 23. Patnaik MM, Timm MM, Vallapureddy R, et al. Flow cytometry based monocyte subset analysis accurately distinguishes chronic myelomonocytic leukemia from myeloproliferative neoplasms with associated monocytosis. Blood Cancer J. 2017;7(7):e584. 24. Valent P, Orazi A, Savona MR, et al. Proposed diagnostic criteria for classical chronic myelomonocytic leukemia (CMML), CMML variants and pre-CMML conditions. Haematologica. 2019;104(10):1935-1949. 25. Patnaik MM, Pophali PA, Lasho TL, et al. Clinical correlates, prognostic impact and survival outcomes in chronic myelomonocytic leukemia patients with the JAK2V617F mutation. Haematologica. 2019;104(6):e236-e239. 26. Daver N, Strati P, Jabbour E, et al. FLT3 mutations in myelodysplastic syndrome and chronic myelomonocytic leukemia. Am J Hematol. 2013;88(1):56-59. 27. Vallapureddy R, Lasho TL, Hoversten K, et al. Nucleophosmin 1 (NPM1) mutations in chronic myelomonocytic leukemia and their prognostic relevance. Am J Hematol. 2017;92(10):E614-E618. 28. Patnaik MM, Tefferi A. Chronic myelomonocytic leukemia: 2020 update on diagnosis, risk stratification and management. Am J

Hematol. 2020;95(1):97-115. 29. Kiss M, Caro AA, Raes G, Laoui D. Systemic reprogramming of monocytes in cancer. Front Oncol. 2020;10:1399. 30. Shen Q, Ouyang J, Tang G, et al. Flow cytometry immunophenotypic findings in chronic myelomonocytic leukemia and its utility in monitoring treatment response. Eur J Haematol. 2015;95(2):168-176. 31. Patnaik MM, Timm MM, Vallapureddy R, et al. Flow cytometry based monocyte subset analysis accurately distinguishes chronic myelomonocytic leukemia from myeloproliferative neoplasms with associated monocytosis. Blood Cancer J. 2017;7(7):e584. 32. Selimoglu-Buet D, Wagner-Ballon O, Saada V, et al. Characteristic repartition of monocyte subsets as a diagnostic signature of chronic myelomonocytic leukemia. Blood. 2015;125(23):3618-3626. 33. Selimoglu-Buet D, Badaoui B, Benayoun E, et al. Accumulation of classical monocytes defines a subgroup of MDS that frequently evolves into CMML. Blood. 2017;130(6):832-835. 34. Schmidl C, Renner K, Peter K, et al. Transcription and enhancer profiling in human monocyte subsets. Blood. 2014;123(17):e90-99. 35. Talati C, Zhang L, Shaheen G, et al. Monocyte subset analysis accurately distinguishes CMML from MDS and is associated with a favorable MDS prognosis. Blood. 2017;129(13):1881-1883. 36. Wagner-Ballon O, Bettelheim P, Lauf J, et al. ELN iMDS flow working group validation of the monocyte assay for chronic myelomonocytic leukemia diagnosis by flow cytometry. Cytometry B Clin Cytom. 2021 Dec 30. [Epub ahead of print] 37. Hudson CA, Burack WR, Leary PC, Bennett JM. Clinical utility of classical and nonclassical monocyte percentage in the diagnosis of chronic myelomonocytic leukemia. Am J Clin Pathol. 2018;150(4):293-302. 38. Pophali PA, Timm MM, Mangaonkar AA, et al. Practical limitations of monocyte subset repartitioning by multiparametric flow cytometry in chronic myelomonocytic leukemia. Blood Cancer J. 2019;9(9):65. 39. Cargo C, Cullen M, Taylor J, et al. The use of targeted sequencing and flow cytometry to identify patients with a clinically significant monocytosis. Blood. 2019;133(12):1325-1334. 40. Patnaik MM, Barraco D, Lasho TL, et al. DNMT3A mutations are associated with inferior overall and leukemia-free survival in chronic myelomonocytic leukemia. Am J Hematol. 2017;92(1):56-61. 41. Itzykson R, Kosmider O, Renneville A, et al. Clonal architecture of chronic myelomonocytic leukemias. Blood. 2013;121(12):2186-2198. 42. Patnaik MM, Lasho TL, Finke CM, et al. Spliceosome mutations involving SRSF2, SF3B1, and U2AF35 in chronic myelomonocytic leukemia: prevalence, clinical correlates, and prognostic relevance. Am J Hematol. 2013;88(3):201-206. 43. Wudhikarn K, Loghavi S, Mangaonkar AA, et al. SF3B1-mutant CMML defines a predominantly dysplastic CMML subtype with a superior acute leukemia-free survival. Blood Adv. 2020;4(22):5716-5721. 44. Caye A, Strullu M, Guidez F, et al. Juvenile myelomonocytic leukemia displays mutations in components of the RAS pathway and the PRC2 network. Nat Genet. 2015;47(11):1334-1340. 45. DiFilippo EC, Coltro G, Carr RM, et al. Spectrum of abnormalities and clonal transformation in germline RUNX1 familial platelet disorder and a genomic comparative analysis with somatic RUNX1 mutations in MDS/MPN overlap neoplasms. Leukemia. 2020;34(9):2519-2524.

Haematologica | 107 July 2022

1515

REVIEW ARTICLE - How I diagnose and treat CMML

M.M. Patnaik

46. St Martin EC, Ferrer A, Wudhikarn K, et al. Clinical features and survival outcomes in patients with chronic myelomonocytic leukemia arising in the context of germline predisposition syndromes. Am J Hematol. 2021;96(9):E327-E330. 47. Perez Botero J, Oliveira JL, Chen D, et al. ASXL1 mutated chronic myelomonocytic leukemia in a patient with familial thrombocytopenia secondary to germline mutation in ANKRD26. Blood Cancer J. 2015;5(5):e315. 48. Saliba J, Saint-Martin C, Di Stefano A, et al. Germline duplication of ATG2B and GSKIP predisposes to familial myeloid malignancies. Nat Genet. 2015;47(10):1131-1140. 49. Martin ES, Ferrer A, Mangaonkar AA, et al. Spectrum of hematological malignancies, clonal evolution and outcomes in 144 Mayo Clinic patients with germline predisposition syndromes. Am J Hematol. 2021;96(11):1450-1460. 50. Swerdlow SH, Campo E, Harris NL, et al., editors. WHO Classification of Tumors of Haematopoietic and Lymphoid Tissues. Lyon: International Agency for Research on Cancer. 2008. 51. Lucas N, Duchmann M, Rameau P, et al. Biology and prognostic impact of clonal plasmacytoid dendritic cells in chronic myelomonocytic leukemia. Leukemia. 2019;33(10):2466-2480. 52. Swederlow SH, Campo E, Harris NL, et al, editors. WHO Classification of Tumors of the Haematopoietic and Lymphoid Tissue. 4 th ed. Lyon: International Agency for Research on Cancer (IARC), 2008. 53. Bain BJ. What is a promonocyte? Am J Hematol. 2013;88(10):919. 54. Such E, Cervera J, Costa D, et al. Cytogenetic risk stratification in chronic myelomonocytic leukemia. Haematologica. 2011;96(3):375-383. 55. Wassie EA, Itzykson R, Lasho TL, et al. Molecular and prognostic correlates of cytogenetic abnormalities in chronic myelomonocytic leukemia: a Mayo Clinic-French Consortium Study. Am J Hematol. 2014;89(12):1111-1115. 56. Patnaik MM, Rangit V, Lasho TL, et al. A comparison of clinical and molecular characteristics of patients with systemic mastocytosis with chronic myelomonocytic leukemia to CMML alone. Leukemia. 2018;32(8):1850-1856. 57. Sotlar K, Fridrich C, Mall A, et al. Detection of c-kit point mutation Asp-816 --> Val in microdissected pooled single mast cells and leukemic cells in a patient with systemic mastocytosis and concomitant chronic myelomonocytic leukemia. Leuk Res. 2002;26(11):979-984. 58. DeAngelo DJ, Radia DH, George TI, et al. Safety and efficacy of avapritinib in advanced systemic mastocytosis: the phase 1 EXPLORER trial. Nat Med. 2021;27(12):2183-2191. 59. Patnaik MM, Pophali PA, Lasho TL, et al. Clinical correlates, prognostic impact and survival outcomes in chronic myelomonocytic leukemia patients with the JAK2V617F mutation. Haematologica. 2019;104(6):e236-e239. 60. Patnaik MM, Ketterling RP, Tefferi A. FGFR1 rearranged hematological neoplasms - molecularly defined and clinically heterogeneous. Leuk Lymphoma. 2018;59(7):1520-1522. 61. Padron E, Garcia-Manero G, Patnaik MM, et al. An international data set for CMML validates prognostic scoring systems and demonstrates a need for novel prognostication strategies. Blood Cancer J. 2015;5(7):e333. 62. Patnaik MM, Tefferi A. Chronic myelomonocytic leukemia: focus on clinical practice. Mayo Clin Proc. 2016;91(2):259-272. 63. Onida F, Kantarjian HM, Smith TL, et al. Prognostic factors and scoring systems in chronic myelomonocytic leukemia: a retrospective analysis of 213 patients. Blood. 2002;99(3):840-849.

64. Kantarjian H, O'Brien S, Ravandi F, et al. Proposal for a new risk model in myelodysplastic syndrome that accounts for events not considered in the original International Prognostic Scoring System. Cancer. 2008;113(6):1351-1361. 65. Such E, Germing U, Malcovati L, et al. Development and validation of a prognostic scoring system for patients with chronic myelomonocytic leukemia. Blood. 2013;121(15):3005-3015. 66. Wassie EA, Itzykson R, Lasho TL, et al. Molecular and prognostic correlates of cytogenetic abnormalities in chronic myelomonocytic leukemia: a Mayo Clinic-French Consortium Study. Am J Hematol. 2014;89(12):1111-1115. 67. Patnaik MM, Itzykson R, Lasho TL, et al. ASXL1 and SETBP1 mutations and their prognostic contribution in chronic myelomonocytic leukemia: a two-center study of 466 patients. Leukemia. 2014;28(11):2206-2212. 68. Gagelmann N, Badbaran A, Beelen DW, et al. A prognostic score including mutation profile and clinical features for patients with CMML undergoing stem cell transplantation. Blood Adv. 2021;5(6):1760-1769. 69. Santoriello D, Andal LM, Cox R, D'Agati VD, Markowitz GS. Lysozyme-induced nephropathy. Kidney Int Rep. 2017;2(1):84-88. 70. Fenaux P, Santini V, Spiriti MAA, et al. A phase 3 randomized, placebo-controlled study assessing the efficacy and safety of epoetin-alpha in anemic patients with low-risk MDS. Leukemia. 2018;32(12):2648-2658. 71. Santini V. Anemia as the main manifestation of myelodysplastic syndromes. Semin Hematol. 2015;52(4):348-356. 72. Park S, Greenberg P, Yucel A, et al. Clinical effectiveness and safety of erythropoietin-stimulating agents for the treatment of low- and intermediate-1-risk myelodysplastic syndrome: a systematic literature review. Br J Haematol. 2019;184(2):134-160. 73. Pophali P, Horna P, Lasho TL, et al. Splenectomy in patients with chronic myelomonocytic leukemia: indications, histopathological findings and clinical outcomes in a single institutional series of thirty-nine patients. Am J Hematol. 2018;93(11):1347-1357. 74. Fenaux P, Platzbecker U, Mufti GJ, et al. Luspatercept in patients with lower-risk myelodysplastic syndromes. N Engl J Med. 2020;382(2):140-151. 75. Jabbour E, Short NJ, Montalban-Bravo G, et al. Randomized phase 2 study of low-dose decitabine vs low-dose azacitidine in lower-risk MDS and MDS/MPN. Blood. 2017;130(13):1514-1522. 76. Burgstaller S, Stauder R, Kuehr T, et al. A phase I study of lenalidomide in patients with chronic myelomonocytic leukemia (CMML) - AGMT_CMML-1. Leuk Lymphoma. 2018;59(5):1121-1126. 77. Hoversten K, Vallapureddy R, Lasho T, et al. Nonhepatosplenic extramedullary manifestations of chronic myelomonocytic leukemia: clinical, molecular and prognostic correlates. Leuk Lymphoma. 2018;59(12):2998-3001. 78. Hunter AM, Newman H, Dezern AE, et al. Integrated human and murine clinical study establishes clinical efficacy of ruxolitinib in chronic myelomonocytic leukemia. Clin Cancer Res. 2021;27(22):6095-6105. 79. Peker D, Padron E, Bennett JM, et al. A close association of autoimmune-mediated processes and autoimmune disorders with chronic myelomonocytic leukemia: observation from a single institution. Acta Haematol. 2015;133(2):249-256. 80. Zahid MF, Barraco D, Lasho TL, et al. Spectrum of autoimmune diseases and systemic inflammatory syndromes in patients with chronic myelomonocytic leukemia. Leuk Lymphoma. 2017;58(6):1488-1493. 81. Ramadan H, Duong VH, Al Ali N, et al. Eltrombopag use in

Haematologica | 107 July 2022

1516

REVIEW ARTICLE - How I diagnose and treat CMML

M.M. Patnaik

patients with chronic myelomonocytic leukemia (CMML): a cautionary tale. Clin Lymphoma Myeloma Leuk. 2016;16Suppl:S64-66. 82. Jaiswal S, Ebert BL. Clonal hematopoiesis in human aging and disease. Science. 2019;366(6465):eaan4673. 83. Jaiswal S, Fontanillas P, Flannick J, et al. Age-related clonal hematopoiesis associated with adverse outcomes. N Engl J Med. 2014;371(26):2488-2498. 84. Jaiswal S, Natarajan P, Silver AJ, et al. Clonal hematopoiesis and risk of atherosclerotic cardiovascular disease. N Engl J Med. 2017;377(2):111-121. 85. Niyongere S, Lucas N, Zhou JM, et al. Heterogeneous expression of cytokines accounts for clinical diversity and refines prognostication in CMML. Leukemia. 2019;33(1):205-216. 86. Franzini A, Pomicter AD, Yan D, et al. The transcriptome of CMML monocytes is highly inflammatory and reflects leukemiaspecific and age-related alterations. Blood Adv. 2019;3(20):2949-2961. 87. Padron E, Garcia-Manero G, Patnaik MM, et al. An international data set for CMML validates prognostic scoring systems and demonstrates a need for novel prognostication strategies. Blood Cancer J. 2015;5(7):e333. 88. Pophali P, Matin A, Mangaonkar AA, et al. Prognostic impact and timing considerations for allogeneic hematopoietic stem cell transplantation in chronic myelomonocytic leukemia. Blood Cancer J. 2020;10(11):121. 89. de Witte T, Bowen D, Robin M, et al. Allogeneic hematopoietic stem cell transplantation for MDS and CMML: recommendations from an international expert panel. Blood. 2017;129(13):1753-1762. 90. Symeonidis A, van Biezen A, de Wreede L, et al. Achievement of complete remission predicts outcome of allogeneic haematopoietic stem cell transplantation in patients with chronic myelomonocytic leukaemia. A study of the Chronic Malignancies Working Party of the European Group for Blood and Marrow Transplantation. Br J Haematol. 2015;171(2):239-246. 91. Kongtim P, Popat U, Jimenez A, et al. Treatment with hypomethylating agents before allogeneic stem cell transplant improves progression-free survival for patients with chronic myelomonocytic leukemia. Biol Blood Marrow Transplant. 2016;22(1):47-53. 92. Fenaux P, Mufti GJ, Hellstrom-Lindberg E, et al. Efficacy of azacitidine compared with that of conventional care regimens in the treatment of higher-risk myelodysplastic syndromes: a randomised, open-label, phase III study. Lancet Oncol. 2009;10(3):223-232. 93. Silverman LR, Demakos EP, Peterson BL, et al. Randomized controlled trial of azacitidine in patients with the myelodysplastic syndrome: a study of the cancer and leukemia

group B. J Clin Oncol. 2002;20(10):2429-2440. 94. Coston T, Pophali P, Vallapureddy R, et al. Suboptimal response rates to hypomethylating agent therapy in chronic myelomonocytic leukemia; a single institutional study of 121 patients. Am J Hematol. 2019;94(7):767-779. 95. Santini V, Allione B, Zini G, et al. A phase II, multicentre trial of decitabine in higher-risk chronic myelomonocytic leukemia. Leukemia. 2018;32(2):413-418. 96. Merlevede J, Droin N, Qin T, et al. Mutation allele burden remains unchanged in chronic myelomonocytic leukaemia responding to hypomethylating agents. Nature Commun. 2016;7:10767. 97. Bejar R, Lord A, Stevenson K, et al. TET2 mutations predict response to hypomethylating agents in myelodysplastic syndrome patients. Blood. 2014;124(17):2705-2712. 98. Coltro G, Mangaonkar AA, Lasho TL, et al. Clinical, molecular, and prognostic correlates of number, type, and functional localization of TET2 mutations in chronic myelomonocytic leukemia (CMML)-a study of 1084 patients. Leukemia. 2019;34(5):1407-1421. 99. Duchmann M, Yalniz FF, Sanna A, et al. Prognostic role of gene mutations in chronic myelomonocytic leukemia patients treated with hypomethylating agents. EBioMedicine. 2018;31:174-181. 100. DiNardo CD, Jonas BA, Pullarkat V, et al. Azacitidine and venetoclax in previously untreated acute myeloid leukemia. N Engl J Med. 2020;383(7):617-629. 101. Sevin M, Debeurme F, Laplane L, et al. Cytokine-like protein 1induced survival of monocytes suggests a combined strategy targeting MCL1 and MAPK in CMML. Blood. 2021;137(24):3390-3402. 102. Saliba AN, Litzow MR, Gangat N, et al. Outcomes of venetoclaxbased therapy in chronic phase and blast transformed chronic myelomonocytic leukemia. Am J Hematol. 2021;96(11):E433-E436. 103. Padron E, Dezern A, Andrade-Campos M, et al. A multiinstitution phase 1 trial of ruxolitinib in patients with chronic myelomonocytic leukemia (CMML). Clin Cancer Res. 2016;22(15):3746-3754. 104. Padron E, Painter JS, Kunigal S, et al. GM-CSF-dependent pSTAT5 sensitivity is a feature with therapeutic potential in chronic myelomonocytic leukemia. Blood. 2013;121(25):5068-5077. 105. Patnaik MM, Sallman DA, Mangaonkar A, et al. Phase 1 study of lenzilumab, a recombinant anti-human GM-CSF antibody, for chronic myelomonocytic leukemia (CMML). Blood. 2020;136(7):909-913. 106. Casolaro A, Golay J, Albanese C, et al. The Polo-like kinase 1 (PLK1) inhibitor NMS-P937 is effective in a new model of disseminated primary CD56+ acute monoblastic leukaemia. PLoS One. 2013;8(3):e58424.

Haematologica | 107 July 2022

1517

ARTICLE - Acute Myeloid Leukemia

Outcomes of older patients aged 60 to 70 years undergoing reduced intensity transplant for acute myeloblastic leukemia: results of the NCRI acute myeloid leukemia 16 trial Nigel H. Russell,1 Robert K. Hills,2 Abin Thomas,3 Ian Thomas,3 Lars Kjeldsen,4 Mike Dennis,5 Charles Craddock,6 Sylvie Freeman,7 Richard E. Clark8 and Alan K. Burnett9 1

Department of Haematology, Nottingham University Hospitals, Nottingham, UK; Nuffield Department of Population Health, University of Oxford, Oxford, UK; 3Centre for Trials Research, College of Biomedical & Life Sciences, Cardiff University, Cardiff, UK; 4Department of Haematology, Rigshospitalet, Copenhagen, Denmark; 5Department of Haematology, Christie Hospital, Manchester, UK; 6Department of Haematology, Queen Elizabeth Hospital, Birmingham, UK; 7Department of Immunology and Immunotherapy, University of Birmingham, Birmingham, UK; 8Department of Molecular and Clinical Cancer Medicine, University of Liverpool, UK and 9Blackwaterfoot, Isle of Arran, UK

Correspondence: Nigel Russell nigel.russell@nottingham.ac.uk Received: June 22, 2021. Accepted: August 19, 2021. Prepublished: October 14, 2021. https://doi.org/10.3324/haematol.2021.279010 ©2022 Ferrata Storti Foundation Haematologica material is published under a CC BY-NC license

Abstract Reduced intensity conditioning (RIC) transplantation is increasingly offered to older patients with acute myeloblastic leukemia. We have previously shown that a RIC allograft, particularly from a sibling donor, is beneficial in intermediate-risk patients aged 35-65 years. We here present analyses from the NCRI AML16 trial extending this experience to older patients aged 60-70 inclusive lacking favorable-risk cytogenetics. Nine hundred thirty-two patients were studied, with RIC transplant in first remission given to 144 (sibling n=52, matched unrelated donor n=92) with a median follow-up for survival from complete remission of 60 months. Comparisons of outcomes of patients transplanted versus those not were carried out using Mantel-Byar analysis. Among the 144 allografted patients, 93 had intermediate-risk cytogenetics, 18 had adverse risk and cytogenetic risk group was unknown for 33. In transplanted patients survival was 37% at 5 years, and while the survival for recipients of grafts from siblings (44%) was better than that for recipients of grafts from matched unrelated donors (34%), this difference was not statistically significant (P=0.2). When comparing RIC versus chemotherapy, survival of patients treated with the former was significantly improved (37% versus 20%, hazard ratio = 0.67 [0.53-0.84]; P<0.001). When stratified by Wheatley risk group into good, standard and poor risk there was consistent benefit for RIC across risk groups. When stratified by minimal residual disease status after course 1, there was consistent benefit for allografting. The benefit for RIC was seen in patients with a FLT3 ITD or NPM1 mutation with no evidence of a differential effect by genotype. We conclude that RIC transplantation is an attractive option for older patients with acute myeloblastic leukemia lacking favorable-risk cytogenetics and, in this study, we could not find a group that did not benefit.

Introduction The outcome for older patients with acute myeloblastic leukemia (AML) has not shown the survival improvements that have been achieved in younger patients.1 The reasons for this include patient-related factors that can affect tolerance to the intensive chemotherapy regimens employed in younger patients but also the biological features of the disease. Patient-related factors include concurrent medical conditions, such as performance status and co-morbidities, which can adversely affect outcome.1,2 Disease-related characteristics include an increased incidence of poor-risk cytogenetics,3,4 an increased incidence

of patients with secondary AML resulting from progression of an antecedent myelodysplastic syndrome1,4 and an increased expression of multidrug resistance mechanisms.5 As a consequence even when patients achieve a complete remission (CR) the risk of disease relapse remains high1 and although over 65% of patients >60 years can achieve a CR with intensive chemotherapy approximately only 20% will survive to 5 years.6 Thus, interventions based on preventing relapse are a priority for development. One therapeutic approach to reduce relapse is allogeneic stem cell transplantation. With the advent of RIC regimens, this therapy has been increasingly applied to older AML patients. Here the intention is to apply a cura-

Haematologica | 107 July 2022

1518

ARTICLE - RIC transplant in the NCRI AML16 trial

N.H. Russell et al.

tive therapy but one that is known to have a significant risk of procedural mortality. In patients aged 40-59 years, a benefit for RIC over chemotherapy has been reported but in our experience and those of others, this was confined to sibling donors.7,8 A number of retrospective studies of registry data in patients over 60 years have been reported. The European Group for Blood and Marrow Transplantation reported a 4-year survival of 27% in patients over 60 years with worse outcome in patients with poor-risk cytogenetics.9 A study from the Center for International Blood and Marrow Transplant Research reported similar results in older patients undergoing RIC in first CR (CR1).10 In that study multivariate analysis showed no impact of age on relapse, non-relapse mortality or survival. The Seattle consortium also confirmed that increasing age over 60 years was not associated with adverse outcomes whereas more co-morbidities and adverse cytogenetics were.11 A multicenter prospective phase II study also found a benefit with an overall survival at 2 years of 40%.12 Studies comparing RIC transplantation with standard chemotherapy in this age group are however lacking although recently the ALFA-1200 trial reported benefit in high-risk older patients only.13 The NCRI AML16 trial for patients with AML and high-grade myelodysplastic syndrome (>10% blasts) over the age of 60 years permitted RIC transplant in CR1 for patients with a suitably matched sibling or unrelated donor if they lacked favorable-risk cytogenetics. Here we report on a comparison of RIC transplantation with chemotherapy in older patients fit for intensive chemotherapy.

Methods Between December 2006 and August 2012, 1,880 patients with AML or high-grade myelodysplastic syndrome (defined as >10% blasts) from the UK and Denmark entered the NCRI AML16 trial (EUDRACT 2005-002847-14) for patients >60 years who were fit for intensive therapy. The trial schema and major outcomes have been published6 but, briefly, induction chemotherapy was daunorubicin/cytarabine or daunorubicin/clofarabine with or without a single dose of gemtuzumab ozogamicin (3 mg/m2) or daunorubicin/cytarabine with or without etoposide and with or without all-trans retinoic acid. Patients could be randomized between two or three courses of therapy and maintenance or not with azacytidine. Patients not in partial remission after course 1 were excluded from the randomization to two or three courses of therapy. The overall CR or incomplete CR (CRi) rate was 67% and did not differ between the arms. Patients in remission (CR or CRi) and who did not have favorable-risk disease could receive a RIC allograft from a matched sibling or matched unrelated donor (MUD) if considered fit for transplantation by the in-

vestigator. Unrelated donor transplants were permissible if they were matched at HLA A, B, C, DRB1 and DQ at the allele level (10/10 HLA-match) or mismatched at a single locus (9/10 HLA-match). Details of the RIC conditioning regimens, which were not protocol-specified, are given in the Online Supplementary Appendix. Graft-versus-host disease prophylaxis and supportive care were in accordance with the local policy of the individual institutions. In this study a total of 148 transplants were performed in CR1 but as only four of these took place in patients >70 years this analysis is confined to the 144 patients aged 70 and less who did not have favorable-risk leukemia. Nine hundred thirty-two patients who had these characteristics were studied with a median follow-up from CR of 60 months. Of the 144 RIC transplants, 52 were from matched sibling donors and 92 from MUD. The characteristics of all the patients in this study are shown in Table 1. The AML16 trial was approved by Council for Research Ethics Committees (COREC) and the local research ethics committee and conforms to local research governance procedures at each center. A copy of a center's local research ethics committee approval and site-specific assessment was lodged with the trial office at the University of Birmingham Clinical Trials Unit (BCTU) before entering patients. Statistical considerations Endpoint definitions follow the recommendations of the International Working Group.14 Only transplantation in first remission is considered here; outcomes following transplantation or not are summarized using the Kaplan-Meier method and forest plots. The outcomes of the MUD and sibling transplant groups were compared using log-rank tests or Cox regression. When comparing the relapse risk, we considered death as a competing risk and a competing risk regression method was used for the analysis. For the comparison of transplant versus no transplant, to counteract the immortal time bias introduced by patients needing to have survived long enough to receive a transplant, Mantel-Byar methodology15 was used. Here all patients entering CR start in the “no transplant” group, with no patients in the transplant group. When a patient is transplanted, they change to the allograft group at the time of transplantation. This avoids early deaths counting against non-transplanted patients as transplanted patients must have survived long enough to receive a transplant and allows for the fact that MUD transplants tend to take place later than those from sibling donors. For the subgroup analysis using the forest plots, we tested for heterogeneity across all the subgroups with a test for trend wherever applicable. We summarize the characteristics of the patients across the group using frequencies and percentages for categorical data, and medians and quartile ranges for quantitative data. Comparisons of pa-

Haematologica | 107 July 2022

1519

ARTICLE - RIC transplant in the NCRI AML16 trial

N.H. Russell et al.

tients’ characteristics are based on c2, Mantel-Haenszel tests for trend, or Wilcoxon rank sum tests as appropriate. All outcomes are summarized at 5 years of follow-up and statistical significance is set at P<0.05.

Results Patient characteristics Nine hundred thirty-two patients aged 60-70 inclusive who entered remission (CR or CRi) and did not have favorable-risk AML were studied with a median follow-up from remission of 60 months. A CONSORT diagram is shown in Online Supplementary Figure S1. Of the 144 RIC transplants performed, 52 were from matched sibling donors and 92 from MUD. The characteristics of all the patients in this study as well as those treated with chemotherapy and those treated with RIC are shown in Table 1. Patients selected for transplantation were more likely to be 65 years or younger and to have a better performance score. The transplant and non-transplant groups were balanced in terms of Wheatley risk groups which subdivides patients into three groups using a prospectively validated risk score comprising cytogenetics, white blood cell count, de novo or secondary AML, age and performance score.16 Of the transplant recipients, 83 were Wheatley good risk, 39 standard risk and 22 poor risk. Of the 144 RIC transplants, 52 were from matched sibling donors and 92 were from MUD. Ninety-three patients had intermediate-risk cytogenetics, 18 had adverse cytogenetics and cytogenetic risk status was unknown for 33. Of the 18 patients with adverse-risk cytogenetics, 15 (83.3%) received transplants from an unrelated donor compared to 57/93 (61.2%) with intermediate-risk cytogenetics. Outcomes after reduced intensity conditioning transplants Detailed outcome data are provided in Table 2. The overall survival at 5 years after remission of the 144 recipients of a RIC transplant was 37% and was not significantly different between those with sibling (44%) or unrelated (34%) donors (unadjusted hazard ratio [HR]=1.35 [0.85–2.13] P=0.2; adjusted for risk group: HR=1.28 [0.81–2.02] P=0.3) (Figure 1). Likewise, there was no significant difference in non-relapse mortality at 5 years between recipients of sibling (28%) and MUD (36%) transplants (unadjusted HR=1.45 [0.78–2.73] P=0.2; adjusted for risk group: HR=1.38 [0.73–2.60] P=0.3; with relapse analyzed as a competing risk) (Online Supplementary Figure S2). We also compared the outcome of sibling and MUD transplants by Wheatley risk group. This analysis showed no survival difference between sibling and MUD transplants for good (51% vs. 42%), standard (49% vs. 31%) or poor risk (0% vs. 13%) patients (HR=1.26 [0.8-2.00], test for trend

P=0.7) although relatively few sibling transplants (7 sibling, 15 MUD) were performed for poor-risk patients. The lack of evidence for a difference between outcomes of sibling and MUD RIC transplants led us to combine the two transplant types together for analyses comparing the outcome of transplantation or not. Apart from survival, chronic graft-versus-host disease can have a major impact on post-transplant quality of life. The majority of patients in this study received in vivo T-lymphocyte depletion (Online Supplementary File) and although quality of life outcomes were not collected, the incidence of severe chronic graft-versus-host disease at 12 months was only 12%, was absent in 38%, mild in 26% and moderate in 24%. Comparison of reduced intensity conditioning transplantation with chemotherapyy When comparing RIC transplantation with no transplantation, survival was significantly improved by transplantation (37% vs. 20%, HR=0.67 [0.53-0.84], P<0.001) (Figure 2). Relapse-free survival was also improved by transplantation (32% vs. 13%, HR=0.56 [0.45-0.70], P<0.001). The cumulative incidence of death in remission at 5 years among the transplant group was 34% which compares to 10% with chemotherapy alone (unadjusted HR=5.02 [3.46–7.29] P<0.001; adjusted for risk group: HR=5.02 [3.46–7.29] P<0.001; with relapse analyzed as a competing risk) (Table 2). To compensate for selection factors for transplant the comparison was repeated for the Wheatley risk groups. Consistent benefit for RIC allografting was seen in all risk groups with no evidence of any interaction (P value for trend 0.86) and the adjusted HR was 0.68 (0.54-0.85), P<0.001. (Figure 3) Overall survival at 5 years for patients in the Wheatley good-risk group was 45% vs. 26%, standard risk 36% vs. 21% and poor risk 12% vs. 7% (Online Supplementary Figure S3A-C). Looking at age as a factor the survival of patients aged 60-65 (n=99) and 66-70 (n=45) at 5 years undergoing RIC was 39% and 33% compared to 23% and 18% for those not transplanted (P value for heterogeneity 0.95) (Online Supplementary Figure S4). The benefit of RIC was seen in patients with intermediate risk, adverse risk or unknown risk cytogenetics (Online Supplementary Figure S5) and in those with a FLT3-ITD or NPM1 mutation with no evidence of a differential effect by genotype (P value for heterogeneity by subgroups 0.2 and 0.7, respectively) although mutation results were not available for all patients (Figure 4). Relapse risk at 5 years was significantly reduced by transplantation (RIC 34%, chemotherapy alone 77%, HR=0.30 [0.22-0.40] P<0.001; with death analyzed as a competing risk) (Table 2 and Online Supplementary Figure S6) and was not significantly different between recipients of sibling or MUD transplants when adjusted for risk group (30% and 33% at 5 years, respectively, P=0.9).

Haematologica | 107 July 2022

1520

ARTICLE - RIC transplant in the NCRI AML16 trial

N.H. Russell et al.

Table 1. Demographics of 932 eligible patients in the AML16 reduced intensity conditioning study.

Characteristics N. of patients Age in years, N (%) 60-65 66-70 Median (Q1-Q3) Sex, N (%) Male Female WHO PS(at diagnosis), N (%) 0 1 2 3 4 Diagnosis, N (%) De novo Secondary High Risk MDS Cytogenetics, N (%) Intermediate Adverse Unknown White cell count x109/L, N (%) 0-9.9 10-49.9 50-99.9 100+ Median (Q1 – Q3) Risk group, N (%) Good Standard Poor HCTCI score, N (%) 0 1 or 2 3 or more Unknown ITD, N (%) WT Mutant Fail Unknown NPM1, N (%) WT Mutant Fail Unknown ITD/NPM1, N (%) WT/WT WT/Mut Mut/WT Mut/Mut Unknown Donor type, N (%) Sibling MUD

Overall 932

No-allograft 788

Allograft 144

P-value

480 (51%) 452 (49%) 65 (63-68)

381 (48%) 407 (52%) 66 (63-68)

99 (69%) 45 (31%) 64 (62-66)

<0.001 <0.001**

543 (58%) 389 (42%)

449 (57%) 339 (43%)

94 (65%) 50 (35%)

0.062

578 (62%) 306 (33%) 28 (3%) 18 (2%) 2 (<0.5%)

469 (60%) 273 (35%) 27 (3%) 17 (2%) 2 (<0.5%)

109 (76%) 33 (23%) 1 (1%) 1 (1%) 0

<0.001*

691 (74%) 134 (14%) 107 (12%)

597 (76%) 110 (14%) 81 (10%)

94 (65%) 24 (17%) 26 (18%)

0.012

606 (83%) 122 (17%) 204

513 (83%) 104 (17%) 171

93 (84%) 18 (16%) 33

0.868

594 (64%) 216 (23%) 75 (8%) 47 (5%) 4.3 (1.9-23.3)

485 (62%) 187 (24%) 72 (9%) 44 (5%) 4.7 (1.9-27)

109 (76%) 29 (20%) 3 (2%) 3 (2%) 3.2 (1.4-9.3)

0.001* <0.001**

547 (49%) 271 (29%) 204 (22%)

374 (48%) 232 (29%) 182 (23%)

83 (58%) 39 (27%) 22 (15%)

0.013*

474 (51%) 313 (34%) 139 (15%) 6

390 (50%) 267 (34%) 126 (16%) 5

84 (50%) 46 (34%) 13 (9%) 1

0.016*

341 (82%) 76 (18%) 9 506

290 (81%) 70 (19%) 6 422

51 (90%) 6 (10%) 3 84

0.105

281 (73%) 102 (27%) 14 535

237 (72%) 93 (28%) 11 447

44 (83%) 9 (17%) 3 88

0.087

246 (65%) 63 (17%) 30 (8%) 37 (10%) 556

207 (64%) 57 (18%) 26 (8%) 35 (11%) 463

39 (76%) 6 (12%) 4 (8%) 2 (4%) 93

0.251

52 (36%) 92 (64%)

All tests are the c2-test, except: *Mantel-Haenszel test for trend and **Wilcoxon rank-sum test. Q1: first quartile; Q3: third quartile; WHO PS, World Health Organization performance status; MDS: myelodysplastic syndrome; WBC: white blood cell count; HCTCI: Hematopoietic Cell Transplantation Comorbidity Index; ITD: internal tandem duplication; WT: wild-type; Mut: mutated; MUD: matched unrelated donor Haematologica | 107 July 2022

1521

ARTICLE - RIC transplant in the NCRI AML16 trial

N.H. Russell et al.

On entry to AML16, all patients had their Hematopoietic Cell Transplant Specific Co-Morbidity Index (HCT-CI) measured.17 This measurement was available for 926/932 patients who entered CR1 and was 0 in 474 (51.9%), 1-2 in

313 (33.6%) and 3+ in 139 (14.9%). For patients proceeding to RIC transplantation the HCT-CI was available for 143/144 patients and the relative frequencies of scores 0, 1-2 and 3+ were 59%, 32% and 9%, respectively. When sur-

Table 2. Survival estimates.

Category

*Death as a competing risk; **Relapse as a competing risk. HR: hazard ratio; 95% CI: 95% confidence interval; CR1: first complete remission; CI: cumulative incidence.

Figure 1. Survival after first complete remission comparing sibling and unrelated donors. CR1: first complete remission; Obs: observed; Exp: expected; MUD: matched unrelated donor; Allo_Type: type of allogeneic transplant. Haematologica | 107 July 2022

1522

ARTICLE - RIC transplant in the NCRI AML16 trial

N.H. Russell et al.

Figure 2. Overall survival comparing reduced intensity conditioning transplantation or not for acute myeloid leukemia in first complete remission. RIC: reduced intensity conditioning; CR1: first complete remission; Obs: observed; Exp: expected.

vival was analyzed by HCT-CI risk group we observed that all HCT-CI risk groups had benefit (P value for heterogeneity by subgroups 0.08). Paradoxically there was a trend for increased benefit with higher HCT-CI (P value for trend 0.03) although numbers transplanted with an HCT-CI >3 were small (Online Supplementary Figure S7). Impact of minimal residual disease after course 1 on transplant outcome Flow minimal residual disease (MRD) status after course 1 was available for 323 patients, as previously reported.18 These included 36 patients who went on to have a RIC transplant. When stratified by post-course 1 flow MRD status (MRD-negative, MRD unknown, MRD-positive and no CR/CRi after course 1) there was again consistent benefit for allografting. (HR=0.68 [0.54-0.85] P<0.001) with no heterogeneity (P=0.20) (Figure 5).

Discussion Here we report on a large prospective evaluation of RIC transplantation compared to chemotherapy in AML patients aged 60-70 years inclusive in first remission who were considered fit for intensive treatment. Our experience shows that when comparing an allograft with no transplant, survival was significantly improved for the former with overall survival at 5 years of 37% compared to 20% for patients treated with chemotherapy. This bene-

fit was present in patients with a good, intermediate or poor prognosis as stratified by the Wheatley index, which includes both disease- and patient-specific factors that can affect outcome. Therefore, this benefit was seen in those with adverse-risk features including secondary AML, adverse-risk cytogenetics and older age and was seen in those with a FLT3 ITD. The benefit from RIC transplantation was due to a significant reduction in the risk of relapse, which outweighed the negative impact of excess transplant-related mortality. These results are consistent with the results from our AML15 trial in which we showed a benefit for RIC transplantation for the age range 35-65 years although in that study only when sibling donors were used.7 In AML16 the benefit from transplantation was seen with both sibling and unrelated donors and although sibling transplants generally performed better, survival was not significantly different particularly as more patients with adverse-risk features received transplants from unrelated donors. Although in this study we adjusted for known factors only 15% of patients who entered remission underwent RIC so there might be other selection factors operative that could have biased the results. In our NCRI AML16 trial, we previously reported that MRD status after course 1 was the most important independent prognostic factor for relapse and survival, being independent of age, cytogenetics, white blood cell count, secondary disease and performance status.18 In this analysis we could not find any evidence of an interaction with treatment outcome following RIC and all groups appeared to

Haematologica | 107 July 2022

1523

ARTICLE - RIC transplant in the NCRI AML16 trial

N.H. Russell et al.

Figure 3. Mantel-Byar analysis of overall survival by Wheatley risk group. SCT: stem cell transplant; O-E: observed – expected; HR: hazard ratio; 95% CI: 95% confidence interval.

Figure 4. Mantel-Byar analysis of overall survival by FLT3/NPM1 status. SCT: stem cell transplant; O-E: observed – expected; HR: hazard ratio; 95% CI: 95% confidence interval; ITD: internal tandem duplication; WT: wild-type.

Figure 5. Mantel-Byar analysis of overall survival by minimal residual disease status after course 1. MRD: minimal residual disease; SCT: stem cell transplant; O-E: observed – expected; HR: hazard ratio; 95% CI: 95% confidence interval; CR: complete remission; C1: course 1. Haematologica | 107 July 2022

1524

ARTICLE - RIC transplant in the NCRI AML16 trial

N.H. Russell et al.

benefit independently of their post-course 1 MRD status including those patients who were MRD-positive or had failed to achieve a CR with their first course of chemotherapy. MRD status was not re-assessed immediately pre-transplant and could have changed as a consequence of subsequent chemotherapy, so whether the benefit of transplantation in post-course 1 MRD-positive patients is confined to those becoming MRD-negative after course 2, is being studied in the NCRI AML18 trial. . One factor that can affect outcome following an allogeneic transplant is the presence of co-morbidities with an HCTCI score of 3+ being associated with a significantly higher risk of non-relapse mortality.19 The majority of patients entering AML16 had an HCT-CI of 0 or 1-2 and we could not detect an adverse impact of higher HCT-CI on outcome although the numbers of patients with a high score were small. Furthermore, our assessment of HCT-CI was made at trial entry not immediately pre-transplant so there could have been changes in the score of some patients as a consequence of receiving induction treatment. However, our findings do suggest that the majority of patients in this age group being treated with intensive chemotherapy are likely to have a health status reflected by a co-morbidity score of 3 or less which would not preclude them from being considered for RIC transplantation. Despite these findings, only a minority of patients underwent transplantation suggesting that this modality of therapy has been underused although transplant rates have increased to 31% in our current NCRI AML18 trial.20 In summary, our results show that while, as patients get older, there will inevitably be selection for treatments based on fitness, our analysis adjusted for the clinical features of the disease shows that RIC transplantation results in improved survival for older patients with AML. Despite these findings and considering the significant increase in transplant-related mortality associated with transplantation more precise risk stratification is required in the selection of older patients for transplants. This may be provided by more detailed genomic analysis particularly in intermediate-risk patients for whom it has been reported that the benefit of transplantation is limited to those with gene mutations typical of secondary AML.13 Disclosures No conflicts of interest to disclose. Contributions AKB. NHR, and RKH conceived and designed the study; AKB, NHR, CC, LK, REC, and SF provided study materials or patients; AKB, RKH, NHR, CC, MD, LH, REC, SF, IT, and AT collected and assembled data; AKB, RKH, NHR, and AT analyzed and interpreted the data; NHR drafted the paper which was revised and approved by all authors.

Acknowledgments The authors would like to thank the Cardiff University Haematology Trials Unit staff for supervision of the trial and the following investigators: Aalborg University Hospital: Kallenbach; Aarhus University Hospital: Starklint, Norgaard, Friis, Steffensen, Moelle, Skov Holm; Aberdeen Royal Infirmary: Culligan, Tighe; Addenbrooke's Hospital: Craig, Crawley, Marcus; Arrowe Park Hospital: Butt, Dasgupta; Barnet General Hospital: Virchis; Basingstoke and North Hampshire Hospital: Simpson, Milne; Belfast City Hospital: Cuthbert, McMullin, Jones; Birmingham Heartlands Hospital: Milligan, Lovell, Paneesha, Smith, Pratt; Blackpool Victoria Hospital: Cahalin; Borders General Hospital: Okhandiar, Tucker; Bradford Royal Infirmary: Ackroyd, Williams, Newton; Bristol Haematology & Oncology Centre: Standen, Evely, Mehta, Robinson, Bird, Robinson, Marks; Cheltenham General Hospital: Blundell, Rye, Lush, Robinson; Chesterfield & North Derbyshire Royal Hospital: Wodzinski, Cutting; Christie Hospital: Dennis, Liakopoulou; Churchill Hospital: Vyas, Peniket; City Hospitals Sunderland: Pemberton, Marshall, Hervey, Lyons; Colchester General Hospital: Maboreke, Hamblin; Countess of Chester Hospital: Lee, Tueger; Crosshouse Hospital: Mccoll, Maclean; Croydon University Hospital: Osuji, Pollard; Derbyshire Royal Infirmary: McKernan, Smith; Derriford Hospital: Rule, Hunter, Nokes, Copplestone, Hamon; Doncaster Royal Infirmary: Kaul, Paul; Ealing Hospital: Abrahamson; Eastbourne District General Hospital: Grace; Falkirk Community Hospital: Neilson; Gartnavel General Hospital: Fitzsimmonds, Soutar, Drummond, McKay; Gloucestershire Royal Hospital: Chown, Frewin, MacPherson, Rye; Guy's Hospital: Carr, Smith; Hammersmith Hospital: Apperley; Hemel Hempstead General Hospital: Harrison, Wood; Hereford County Hospital: Robinson; Herlev University Hospital: Hoegh Dufva, Helleberg, Jensen; Hillingdon Hospital: Jan-Mohamed, Kaczmarski; Hull Royal Infirmary: Ali, Carter; Ipswich Hospital: Dodd, Ademokun; James Paget Hospital: Gomez; Kent & Canterbury Hospital: Pocock, Saieed, Ratnayake, Zwaan, Lindsay; Kettering General Hospital: Kwan, Lyttleton; King's College Hospital (Denmark Hill): Lucas; Kingston Hospital: Sykes; Leicester Royal Infirmary: Hunter, Garg; Lincoln County Hospital: Saravanamuttu; Maidstone District General Hospital: Gale, Rassam, Gillett; Manchester Royal Infirmary: Lucas, Yin, Burthem, Tholouli; Medway Maritime Hospital: Aldouri, Andrews, Eden; Monklands Hospital: Mitchell, Patterson; Musgrove Park Hospital: Bolam, Mannari; New Cross Hospital: Basu, Jacob, Handa; Ninewells Hospital: Gelly, Tauro, Meiklejohn, Gowans, Marron; Norfolk & Norwich University Hospital: Lawes; North Staffordshire Hospital: Chasty, Stewart; Northampton General Hospital (Acute): Mittal; Northwick Park Hospital: Panoskaltsis; Nottingham University Hospitals NHS Trust - City Hospital Campus: Das-Gupta, Russell, Byrne; Odense University Hospital: Friis; Pilgrim Hospital: Tringham, Saravanamuttu; Pinderfields General Hospital:

Haematologica | 107 July 2022

1525

ARTICLE - RIC transplant in the NCRI AML16 trial

N.H. Russell et al.

Ashcroft, Moreton, Wright, Patil, Chapple; Poole General Hospital: Bell, Jack; Princess Royal University Hospital: Vadher, De Lord, Lakhani; Queen Alexandra Hospital: Dignum, Corser, Cranfield, Ganczakowski; Queen Elizabeth Hospital (Kings Lynn): Keidan, Coates; Queen Elizabeth Hospital Birmingham: Murray, Craddock, Mahendra, Cook; Queen Elizabeth Hospital, Woolwich: Cheung; Queen's Hospital, Romford: Grant, Brownell, Stevens, Hemmaway; Raigmore Hospital: Forsyth, Lush; Rigshospitalet University Hospital: Kjeldsen, Jensen, Nielsen, Dimitrijevic, Bjerrum, Niemann, Schoelkopf, Friis, Kampmann, Groenbaek; Rotherham District General Hospital: Barker, Taylor; Royal Berkshire Hospital: Mucklow, Ramasamy, Simpson; Royal Bournemouth General Hospital: Killick, Hall, Walewska; Royal Cornwall Hospital (Treliske): Pottinger, Parkins, Creagh, Royal Devon & Exeter Hospital (Wonford): Hamilton, Joiner, Lee, Ruell, Todd, Rudin, Coppell, Kerr, Royal Free Hospital: Kottaridis, Mehta, McNamara, Royal Gwent Hospital: Jenkins, Jackson, Osman, Royal Hallamshire Hospital: Dalley, Reilly, Snowden; Royal Stoke University Hospital (University Hospital of North Midlands NHS Trust): Chasty, Stewart, Chanda, Karunanithi; Royal Sussex County Hospital: Duncan, Corbett, Hill; Royal United Hospital Bath: Crowe, Wexler, Knechtli; Russells Hall Hospital: Fernandes, Neilson, Bareford, Taylor; Salford Royal Hospital: Jowitt, Houghton; Salisbury District Hospital: Cullis, Grand, Everington; Sandwell General Hospital: Hasan, Singleton Hospital: Sati, Al-Ismail, Benton; Southampton General Hospital: Richardson, Orchard, Jenner; Southern General Hospital: Morrison, Macdonald; St Bartholomew's Hospital: Oakervee; St George's Hospital: Willis, St Helier Hospital: Mercieca, Rice, Zuha, Kolade; St James's University Hospital: Bowen, Johnson, McVerry; St Richard's Hospital: Janes, Bevan; Staffordshire General Hospital: Revell; Stirling Royal Infirmary: Neilson; Stoke Mandeville Hospi-

tal: Eagleton; The Alexandra Hospital: Clark; The Ayr Hospital: Eynaud; The Great Western Hospital: Gray, Blesing, Sternberg, Green; The James Cook University Hospital: Wood, Plews; The Royal Bolton Hospital: Grey; The Royal Liverpool University Hospital: Clark; The Royal Oldham Hospital: Allamedine, Elhanash, Osborne, Greenfield, Sen; The Royal Victoria Infirmary: Jackson, Jones, Lennard, O'Brien; Torbay District General Hospital: Turner, Rymes, Roberts, Smith; University College Hospital: Khwaja, Ardeshna, Yong; University Hospital Aintree: Woodcock, Sadik, Salim; University Hospital Coventry (Walsgrave): Jackson, Harrison, Bokhari, Arbuthnot, Jobanputra; University Hospital Lewisham: Mir, Yeghen; University Hospital of Wales: Poynton, Rowntree, Burnett, Kell, Knapper; Victoria Hospital: Williamson; Victoria Infirmary: Tansey, Sharp; Warwick Hospital: Borg; Western General Hospital: Roddie, Johnson; Wexham Park Hospital: Blienz, Moule, Philpott; Whiston Hospital: Nicholson, Satchi; Wishaw General Hospital: Helenglass; Worcestershire Royal Hospital: Shafeek, Pemberton; Worthing Hospital: O'Driscoll; Wycombe General Hospital: Pattinson, Kelly; York Hospital: Bond, Munro; Ysbyty Glan Clwyd: Hoyle, Heartin; Ysbyty Gwynedd District General Hospital: Seale. Funding Research support for the NCRI AML16 trial was provided by Cancer Research UK (grant number A6031). Data-sharing statement Anonymised data from the Sponsor’s dataset are available to independent researchers through a standard process, which includes an internal feasibility assessment and scientific review process. Any data release is subject to participant consent and any existing contractual obligations with funders and collaborators

References 1. Burnett A, Wetzler M, Löwenberg B. Therapeutic advances in acute myeloid leukemia. J Clin Oncol. 2010;28(5):4333-4338. 2. Walter RB, Kantarjian HM, Huang X, et al. Effect of complete remission and responses less than complete remission on survival in acute myeloid leukemia: a combined Eastern Cooperative Oncology Group, Southwest Oncology Group, and M.D. Anderson Cancer Center study. J Clin Oncol. 2010;28(10):1766-1771. 3. van der Holt B, Breems DA, Berna Beverloo H, et al. Various distinctive cytogenetic abnormalities in patients with acute myeloid leukaemia aged 60 years and older express adverse prognostic value: results from a prospective clinical trial. Br J Haematol. 2007;136(1):96-105. 4. Appelbaum FR, Gundacker H, Head DR, et al. Age and acute myeloid leukemia. Blood. 2006;107:3481-3485. 5. Leith CP, Chir B, Kopecky KJ, et al. Acute myeloid leukemia in the elderly: assessment of multidrug resistance (MDR1) and cytogenetics distinguishes biologic subgroups with remarkably

distinct responses to standard chemotherapy. A Southwest Oncology Group study. Blood. 1997;89(9):3323-3329. 6. Burnett AK, Russell NH, Hills RK et al. Addition of gemtuzumab ozogamicin to induction chemotherapy improves survival in older patients with acute myeloid leukemia. J Clin Oncol. 2012;30(32):3924-3931. 7. Russell NH, Kjeldsen L, Craddock C, et al. A comparative assessment of the curative potential of reduced intensity allografts in acute myeloid leukaemia. Leukemia. 2015;29(7):1478-1484. 8. Cornelissen JJ, Versluis J, Passweg JR, et al. Comparative therapeutic value of post-remission approaches in patients with acute myeloid leukemia aged 40-60 years. Leukemia. 2015;29(5):1041-1050. 9. Lim Z, Brand R, Martino R, et al. Allogeneic hematopoietic stemcell transplantation for patients 50 years or older with myelodysplastic syndromes or secondary acute myeloid leukemia. J Clin Oncol. 2010;28 (3):405-411.

Haematologica | 107 July 2022

1526

ARTICLE - RIC transplant in the NCRI AML16 trial

N.H. Russell et al.

10. McClune BL, Weisdorf DJ, Pedersen TL, et al. Effect of age on outcome of reduced-intensity hematopoietic cell transplantation for older patients with acute myeloid leukemia in first complete remission or with myelodysplastic syndrome. J Clin Oncol. 2010;28(11):1878-1887. 11. Sorror ML, Sandmaier BM, Storer BE, et al. Long-term outcomes among older patients following nonmyeloablative conditioning and allogeneic hematopoietic cell transplantation for advanced hematologic malignancies. JAMA. 2011;306(17):1874-1883. 12. Devine SM, Owzar K, Blum W, et al. Phase II study of allogeneic transplantation for older patients with acute myeloid leukemia in first complete remission using a reduced-intensity conditioning regimen: results from Cancer and Leukemia Group B 100103 (Alliance for Clinical Trials in Oncology)/Blood and Marrow Transplant Clinical Trial Network 0502. J Clin Oncol. 2015;33(35):4167-4175. 13 Gardin C, Pautas C, Fournier E, et al Added prognostic value of secondary AML-like gene mutations in ELN intermediate-risk older AML: ALFA-1200 study results. Blood Adv. 2020;4(9):1942-1949. 14. Cheson BD, Bennett JM, Kopecky KJ, et al. Revised recommendations of the International Working Group for diagnosis, standardization of response criteria treatment

outcomes, and reporting standards for therapeutic trials in acute myeloid leukemia. J Clin Oncol. 2003;21(24):4642-4649. 15. Mantel N, Byar DP. Evaluation of response-time data involving transient states: an illustration using heart-transplant data. J Am Stat Assoc. 1974;69(8):81-86. 16. Wheatley K, Brookes CL, Howman AJ, et al. Prognostic factor analysis of the survival of elderly patients with AML in the MRC AML11 and LRF AML14 trials. Br J Haematol. 2009;145(5):598-605. 17. Sorror ML, Sandmaier BM, Storer BE, et al. Comorbidity and disease status-based risk stratification of outcomes among patients with acute myeloid leukemia or myelodysplasia receiving allogeneic hematopoietic cell transplantation. J Clin Oncol. 2007;25 (27):4246-4254 18. Freeman SD, Virgo P, Couzens S, et al. Prognostic relevance of treatment response measured by flow cytometric residual disease detection in older patients with acute myeloid leukemia. J Clin Oncol. 2013;31(32):4123-4131. 19. Sorror ML, Logan BR, Zhu X, et al. Prospective validation of the predictive power of the Hematopoietic Cell Transplantation Comorbidity Index: a Center for International Blood and Marrow Transplant Research study. Biol Blood Marrow Transplant. 2015;21(8):1479-1487.

Haematologica | 107 July 2022

1527

A. Österroos et al.

ARTICLE - Acute Myeloid Leukemia

A risk score based on real-world data to predict early death in acute promyelocytic leukemia Albin Österroos,1 Tânia Maia,2 Anna Eriksson,1 Martin Jädersten,3 Vladimir Lazarevic,4,5 Lovisa Wennström,6 Petar Antunovic,7 Jörg Cammenga,7 Stefan Deneberg,3 Fryderyk Lorenz,8 Lars Möllgård,6 Bertil Uggla,9 Emma Ölander,10 Eliana Aguiar,2 Fernanda Trigo,2 Martin Höglund,1 Gunnar Juliusson4,5 and Sören Lehmann1,3 Department of Medical Sciences, Uppsala University, Uppsala, Sweden; 2Department of Clinical Hematology, University Hospital Center of São João, Porto, Portugal; 3Department of Hematology, Karolinska University Hospital, Stockholm, Sweden; 4Department of Hematology, Skåne University Hospital, Lund, Sweden; 5Stem Cell Center, Department of Hematology, Department of Laboratory Medicine, Lund University, Lund, Sweden: 6 Department of Hematology, Sahlgrenska University Hospital, Gothenburg, Sweden; 7 Department of Hematology, Linköping University Hospital, Linköping, Sweden; 8Department of Hematology, Norrland University Hospital, Umeå, Sweden; 9Department of Medicine, Division of Hematology, Örebro University Hospital, Örebro, Sweden and 10Department of Hematology, Sundsvall Hospital, Sundsvall, Sweden 1

Correspondence: Sören Lehmann soren.lehmann@medsci.uu.se Received: September 29, 2021. Accepted: December 21, 2021. Prepublished: January 27, 2022. https://doi.org/10.3324/haematol.2021.280093 ©2022 Ferrata Storti Foundation Haematologica material is published under a CC BY-NC license

Abstract With increasingly effective treatments, early death (ED) has become the predominant reason for therapeutic failure in patients with acute promyelocytic leukemia (APL). To better prevent ED, patients with high-risk of ED must be identified. Our aim was to develop a score that predicts the risk of ED in a real-life setting. We used APL patients in the populationbased Swedish AML Registry (n=301) and a Portuguese hospital-based registry (n=129) as training and validation cohorts, respectively. The cohorts were comparable with respect to age (median, 54 and 53 years) and ED rate (19.6% and 18.6%). The score was developed by logistic regression analyses, risk-per-quantile assessment and scoring based on ridge regression coefficients from multivariable penalized logistic regression analysis. White blood cell count, platelet count and age were selected by this approach as the most significant variables for predicting ED. The score identified low-, high- and very high-risk patients with ED risks of 4.8%, 20.2% and 50.9% respectively in the training cohort and with 6.7%, 25.0% and 36.0% as corresponding values for the validation cohort. The score identified an increased risk of ED already at sub-normal and normal white blood cell counts and, consequently, it was better at predicting ED risk than the Sanz score (AUROC 0.77 vs. 0.64). In summary, we here present an externally validated and population-based risk score to predict ED risk in a real-world setting, identifying patients with the most urgent need of aggressive ED prevention. The results also suggest that increased vigilance for ED is already necessary at sub-normal/normal white blood cell counts.

Introduction Acute promyelocytic leukemia (APL) accounts for approximately 5-8% of all cases of acute myeloid leukemia (AML) and is a distinct AML subtype characterized by the balanced reciprocal translocation t(15;17) that results in the fusion oncoprotein PML-RARA (promyelocytic leukemia gene-retinoic acid receptor alpha). APL patients have cure rates of approximately 90% with current treatment regimens.1 Nonetheless, patients with APL are at a high risk of early death (ED) due to a coagulopathy that is characterized especially by a hyperfibrinolytic state.2 This leads to a risk of fatal hemorrhages and thromboses with a reported ED rate within 30 days after diagnosis of up to 30%.3 While clinical trials usually report ED rates of 5-8%,4-8 higher ED

rates are found in population-based studies9-13 that do not exclude patients due to age, comorbidities, poor performance status or ongoing hemorrhages and thromboses.1 Due to the effectiveness of the current treatment, with almost no primary treatment resistance and very few relapses, ED has become the main obstacle to long-term survival in APL.13 Approximately two-thirds of all ED occur within 1 week after diagnosis and the median time to onset of bleeding is 5 days.14,15 An early clinical suspicion of APL is necessary for prompt and adequate surveillance as well as prompt pre-emptive measures, even before the diagnosis is established, to decrease the risk of ED. Although risk factors for ED are well known, clinically useful and validated tools for predicting ED are lacking. Such tools could help clinicians to identify APL patients in need of more intensive surveillance and more aggressive ED-preventing interven-

Haematologica | 107 July 2022

1528

ARTICLE - Predicting early death in APL

A. Österroos et al.

tions. Currently, the Sanz risk model is often used to predict ED although the model was developed to predict relapse after treatment with all-trans retinoic acid (ATRA) and idarubicin.16 Other scores have been proposed17,18 but an externally validated system based on population-based data is still needed. Importantly, a clinically useful score system should be based on studies using unselected population-based APL cohorts that are not restricted to patients fitting inclusion criteria in clinical trials. In this study, we aimed to develop a prediction model for ED in newly diagnosed APL patients. We used real-world data and an external validation cohort to propose a clinically useful risk score based on age, white blood cell (WBC) count and platelet count at the time of diagnosis.

Methods Cohorts All adult patients diagnosed with APL in Sweden from January 1997 to December 2020 (n=301) were included in a training cohort for development of the prediction model. This model was validated against all APL patients (n=129) diagnosed and treated at the University Hospital Center of São João, Porto, Portugal, between January 2005 and April 2019. Clinical data for the training cohort were retrieved from the Swedish AML Registry (coverage >98%19) with additional data collected from the patients’ records. Data for the validation cohort were retrieved from the patients’ records. The diagnosis of APL was confirmed by cytogenetics, fluorescence in situ hybridization and/or reverse transcriptase polymerase chain reaction analysis for the PML-RARA fusion. Induction therapies Patients in the test cohort <70 years received oral ATRA 45 mg/m2 daily until complete remission plus idarubicin 12 mg/m2 on days 1, 3 and 5 plus cytarabine 200 mg/m2 days 1-7 until October 2016. Patients >70 years were given ATRA 45 mg/m2 daily until complete remission plus idarubicin 12 mg/m2 on days 2, 4 and 6. From October 2016, low- and intermediate-risk patients received oral ATRA 45 mg/m2 daily plus arsenic trioxide (ATO) at a dose of 0.3 mg/kg intravenously, given according to the Medical Research Council protocol.8 From 2019, also high-risk patients were treated with ATO and ATRA at the same doses plus idarubicin at a dose of 6-12 mg/m2 (adjusted for age) on days 2, 4, 6 and 8.20 Patients in the validation cohort were given treatment according to the AIDA protocol21 with idarubicin 12 mg/m2 for 4 days plus oral ATRA 45 mg/m2 daily until complete remission until April 2018. From April 2018, low- and inter-

mediate-risk patients were treated according to the PETHEMA LPA 2017 protocol with intravenous ATO at a dose of 0.15 mg/kg and oral ATRA 45 mg/m2 daily until complete remission whereas high-risk patients <70 years were still treated according to the AIDA protocol. Online Supplementary Figure S1 shows an overview of the induction therapies administered and the ED rate observed per group. Statistical analyses and risk score development ED was defined as death from any cause within 30 days of diagnosis. Descriptive statistical comparisons were performed using the Fisher exact test for categorical variables and Mann-Whitney U test for continuous variables. Online Supplementary Figure S2 shows how the prediction model was developed and more details on the statistics are presented in the Online Supplementary Material. In short, univariate and multivariate logistic regression analyses were performed with regard to ED using clinical and laboratory variables. Clinically useful cut-offs were determined for the statistically significant variables and multivariable penalized logistic regression analysis was performed to obtain ridge regression coefficients per variable and per level. Points were assigned by dividing each ridge regression coefficient by the lowest coefficient followed by rounding to the nearest integer. The risk score was developed in line with the Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis (TRIPOD) statement.22 All statistical analyses were performed using the computing environment R version 4.0.223 with packages described in the Online Supplementary Methods. The regional ethics review board in Stockholm, Sweden, approved the study, which was performed in accordance with the Declaration of Helsinki.

Results Patients’ characteristics The clinical characteristics of the patients in the two cohorts are described in Table 1. The training cohort included 301 patients of whom 19.6% (n=59) died within 30 days (Figure 1A). The validation cohort included 129 patients of whom 18.6% (n=24) died within 30 days (P=0.92) (Figure 1B). Similarly, 11.6% (n=35) and 10.1% (n=13) died within 7 days of diagnosis, accounting for 59.3% and 54.2% of all ED in the training and validation cohorts, respectively (P=0.76). The training and validation cohorts were similar with regard to age (median 54 vs. 53 years, P=0.16), gender distribution (50.2% vs. 56.6% women, P=0.26) and median time to ED (6.0 vs. 4.5 days, P=0.66). Approximately one third of the patients in both cohorts were classified as having a high Sanz risk score (28.0% vs. 28.7%). Hemoglo-

Haematologica | 107 July 2022

1529

ARTICLE - Predicting early death in APL

A. Österroos et al.

Table 1. Characteristics of the training and validation cohorts.

Training cohort (N=301)

Validation cohort (N=129)

Jan 1997 - Dec 2020

Jan 2005 - Apr 2019

Setting

Nationwide

University hospital

Age (median, range)

54 (17-89)

53 (18-82)

0.16

Women (n, %) WHO status (n, %) 0 1 2 3 4 Sanz score (n, %) Low Intermediate High Hemoglobin, g/L (median, range) White blood cells, x109/L (median, range) Platelets, x109/L (median, range) Lactate dehydrogenase, mkat/L (median, range) C-reactive protein, mg/L (median, range)

151 (50.2)

73 (56.6)

0.26 0.84

70 (27.5) 119 (46.7) 45 (17.6) 10 (3.9) 11 (4.3)

30 (28.6) 45 (42.9) 20 (19.0) 3 (2.9) 7 (6.7)

Data collection period

Fibrinogen, g/L (median, range) Creatinine, mmol/L (median, range) Albumin, g/L (median, range) Time to early death, days (median, range) Death within 7 days (n, %) Death within 30 days (n, %)

P-value

0.001 95 (34.5) 103 (37.5) 77 (28.0)

23 (17.8) 69 (53.5) 37 (28.7)

98 (34-151)

89 (55-137)

0.001

2.1 (0-254)

2.5 (0.3-256)

0.39

30 (4-366)

23 (5-137)

0.02

5.2 (0-61.9)

5.4 (0.5-382)

0.29

17 (0-450)

19.1 (0.9-442)

0.71

1.5 (0.05-4.8)

1.4 (0-5.2)

0.53

78 (42-407)

70 (27-346)

< 0.001

38 (22-54)

40.7 (16-50.8)

0.001

6 (0-29)

4.5 (0-24)

0.66

35 (11.6) 59 (19.6)

13 (10.1) 24 (18.6)

0.76 0.92

P-values for the Fisher exact test for categorical variables and the Mann-Whitney U test for continuous variables. WHO: World Health Organization.

bin concentration was lower at the time of diagnosis in the validation cohort (98 g/L vs. 89 g/L, P=0.001) while there were no significant differences between the cohorts with regard to World Health Organization (WHO) performance status, WBC count, platelet count, and lactate dehydrogenase (LDH), C-reactive protein (CRP) and fibrinogen levels. The cause of death was available for 46 and 23 ED patients in the training and validation cohorts, respectively. Of all deaths, 59.4% were due to bleeding or thrombosis with a median time to ED of only 2 days (range, 0-27 days). Other major causes of death were multiorgan failure (including ED due to differentiation syndrome) (18.8%) and infections/sepsis (17.4%). The median time to ED for nonhemorrhagic and non-thrombotic patients was 11 days (range, 1-24 days).

Comparison of early death versus non-early death patients in the training cohort Table 2 shows a comparison between ED and non-ED patients in the training cohort. ED patients were older than non-ED patients (median age: 65 vs. 52 years, P<0.001). Furthermore, 54.2% of ED patients had a WHO performance status of 2-4 at the time of diagnosis compared to 14.0% of non-ED patients (P<0.001). ED patients also had higher levels of WBC, LDH, CRP and creatinine compared to non-ED patients. Platelet counts and albumin levels were significantly lower in ED patients. There were no differences in gender distribution or hemoglobin levels between ED and non-ED patients. Multivariate logistic regression analyses were performed with regard to ED with the initial inclusion of age, hemoglobin, WBC and platelets in the training cohort. These

Haematologica | 107 July 2022

1530

ARTICLE - Predicting early death in APL

A. Österroos et al.

Figure 1. Distribution of early deaths. (A, B) The number and deaths per day from time of diagnosis in the training cohort (A) and the validation cohort (B).

variables were included as they are readily available at the time of APL diagnosis, had no need for conversion factors (e.g. creatinine and LDH) and to keep the events-per-variable >10 as recommended.22 LDH, which was associated with ED in univariate analysis, was excluded due to a strong correlation with WBC count with a Pearson correlation coefficient of 0.62 (95% confidence interval (95% CI): 0.55-0.69, P<0.001). Values for hemoglobin, WBC and/or platelets had to be imputed for 26 (8.6%) patients who lacked these data. There were no differences between patients with complete and incomplete data, suggesting that data were missing at random (Online Supplementary Table S1). Hemoglobin showed no prognostic power and was removed after backward selection (Online Supplementary Table S2). The odds ratio (OR) for age was 1.50 (95% CI: 1.24-1.83) per 10-year increase, 1.10 (95% CI: 1.05-1.16) per 5x109/L increment for WBC and 0.86 (95% CI: 0.77-0.95) per 10x109/L increment for platelets.

vided the training cohort into ten quantiles by age, WBC or platelet levels. There was a continuous linear increase in the risk of ED with either age ≥50 years or WBC ≥2.2x109/L (Figure 2). Importantly, the risk of ED increased already at sub-normal WBC levels and with a steep increase in ED risk within the normal WBC range. Contrariwise, the risk of ED steadily declined when platelet levels reached ≥30x109/L. Based on the observed risks, we divided age, WBC and platelets into subgroups (Table 3). Age was subdivided into four categories: <50 years, 50-59 years, 60-69 years and ≥70 years; WBC into three categories: <3.0x109/L, 3.05.0x109/L and >5.0x109/L; and platelets into two categories: ≥30x109/L and <30x109/L. Interestingly and as noted above, already sub-normal WBC levels had a clear association with ED, as shown in Online Supplementary Figure S3. The points per risk group variables were obtained by performing multivariable penalized logistic regression analyses to obtain ridge regression coefficients. We assigned Increasing early death is seen with rising white blood points to each category by dividing each regression coefcell count already at sub-normal and normal white cell ficient by the lowest regression coefficient and rounding counts to the nearest integer (Table 3). A total score was assigned Based on the multivariate logistic regression analysis, we to each patient by adding the points for each variable. investigated the risk of ED per variable level to obtain Zero, 1, 2 or 3 points were assigned based on age, 0, 1 or clinically useful cut-offs for a predictive risk score. We di- 3 points based on WBC count and 0 or 1 point based on Haematologica | 107 July 2022

1531

ARTICLE - Predicting early death in APL

A. Österroos et al.

Table 2. Characteristics and logistic regression analyses for early death or non-early death patients in the training cohort (n=301).

Age (median, range) Women (n, %)

ED (N=59)

Non-ED (N=242)

P-value*

Odds ratio

95% CI

65 (17-86)

52 (18-89)

< 0.001

1.04

1.02-1.06

23 (39.0)

128 (52.9)

0.08

0.57

0.32-1.02

< 0.001

4.63

2.93-7.30

WHO status (n, %) 0

2 (4.4)

68 (32.5)

12 (26.1)

107 (51.2)

14 (30.4)

31 (14.8)

8 (17.4)

2 (1.0)

10 (21.7)

1 (0.5)

97 (46-140)

98 (34-151)

0.72

1.00

0.98-1.01

10.6 (0.4-254)

1.7 (0-166)

< 0.001

1.02

1.01-1.03

Platelets, x109/L (median, range)

21 (5-143)

34 (4-366)

0.001

0.98

0.97-1.00

Lactate dehydrogenase, mkat/L (median, range)

9.8 (3.3-48)

4.8 (0-61.9)

< 0.001

1.12

1.06-1.18

C-reactive protein, mg/L (median, range)

30 (0-450)

14 (0-270)

0.02

1.01

1.00-1.01

1.5 (0.05-3.1)

1.5 (0.5-4.8)

0.26

0.63

0.35-1.11

Creatinine, μmol/L (median, range)

93 (46-407)

75 (42-219)

< 0.001

1.03

1.01-1.04

Albumin, g/L (median, range)

35 (22-47)

38.5 (22-54)

0.06

0.94

0.89-0.99

Hemoglobin, g/L (median, range) White blood cells, x109/L (median, range)

Fibrinogen, g/L (median, range)

P-values for the Fisher exact test for categorical variables and the Mann-Whitney U test for continuous variables. ED: early death, i.e. death within 30 days of diagnosis; 95% CI: 95% confidence interval; WHO: World Health Organization.

platelet levels. The risk score sum for each patient ranged from 0 to 7 points with a higher predicted risk of ED as the score sum increased and with ED risks ranging from 2.3% in patients with 0 points to 72.0% in patients with 7 points (Online Supplementary Figure S4). We further stratified the patients into three risk groups based on assigned points and the predicted risk of ED. Patients with 0-2 points had a predicted risk of ED between 0-10% and constituted a low-risk group whereas patients assigned 3-4 points were classified as a highrisk group with a 10-30% predicted risk of ED. Patients with 5-7 points had >30% predicted risk of ED and were defined as a very high-risk group (Figure 3). An online calculator for the score is provided at: apl-earlydeath.shinyapps.io/risk-score. Risk score performance in the training and validation cohorts The risk score distribution in the training cohort is shown in Figure 4A and Online Supplementary Table S3. The 125 (41.5%) patients who were categorized as lowrisk patients (0-2 score points) had an observed ED rate of 4.8%. The 119 (39.5%) patients who were classified as high-risk (3-4 score points) had an observed ED rate of 20.2% while the remaining 57 (18.9%) patients were categorized as very high-risk (5-7 score points) and had an observed ED rate of 50.9%. The median time to ED in the

low-, high- and very high-risk groups was 14 days (range, 1-22 days), 4 days (range, 0-29 days) and 6 days (range, 0-26 days), respectively. The area under the receiver operating characteristic (AUROC) curve for ED in the training cohort was 0.81 (95% CI: 0.75-0.87) indicating good discriminative potential. We applied the predictive score to all patients in the validation cohort with the same definitions of the score variables and of ED as for the training cohort. The risk score distribution in the external validation cohort is shown in Figure 4B and Online Supplementary Table S3. The proportions of patients per risk group were similar to those of the training cohort with 46.5%, 34.1% and 19.4% in the low-, high- and very high-risk groups, respectively. Importantly, the observed ED rates per risk group were in line with those of the training cohort with 6.7%, 25.0% and 36.0% of patients in the low-, high- and very high-risk groups, respectively (Figure 4C). The AUROC for ED was 0.70 (95% CI: 0.58-0.82) in the validation cohort. ATRA/ATO induction therapy was introduced late during the study period and was used predominantly in Sanz low- and intermediate-risk patients. In total, 85 patients were treated with ATRA/ATO as induction therapy with only five ED (5.9%) (Online Supplementary Figure S1). These numbers were too low to further validate the score in ATRA/ATO-treated patients specifically.

Haematologica | 107 July 2022

1532

ARTICLE - Predicting early death in APL

A. Österroos et al.

Figure 2. Risk of early death per variable and quantile. The training cohort (n=301) was divided into ten quantiles for age, platelets and white blood cell counts, with approximately 30 patients per quantile. Locally estimated scatterplot smoothing curves are shown for each variable with regard to rate of early death on the y-axis. ED: early death; WBC: white blood cell count.

Table 3. Table 3. Multivariable penalized logistic regression for Comparison to other predictive scores While the Sanz risk score was originally developed as a tool early death in the training cohort. for predicting APL relapse,16 the score has also shown progRidge Assigned nostic value for ED.11,17,24 Seventy-seven (28.0%) patients in regression points* the training cohort and 37 (28.7%) in the validation cohort coefficients were classified as high Sanz risk (Table 1). The observed ED < 50 ref. 0 rates in the training cohort were 9.4%, 17.7% and 35.4% for low, intermediate and high Sanz risk groups, respectively. 50-59 0.63 1 However, in the validation cohort, respective ED rates were Age (years) 60-69 1.30 2 26.1%, 8.7% and 32.4% indicating less accurate ED prediction with a high ED rate in the low Sanz risk group. ≥ 70 1.63 3 Merging patients in both our cohorts, the net reclassifica<3 ref. 0 tion improvement compared to the Sanz risk was 23.3% with regard to the proposed risk groups (Online SupplemenWBC (x109/L) 3-5 0.70 1 tary Table S4, Online Supplementary Figure S5). Sanz risk >5 1.71 3 overestimated the risk of ED in non-ED patients of whom 41.5% could be accurately stratified to a lower risk with our ≥ 30 ref. 0 Platelets score. The AUROC of our new risk score was 0.77 (95% CI: (x109/L) < 30 0.61 1 0.72-0.83) compared to 0.64 (95% CI: 0.58-0.71) for the Sanz Intercept -3.42 risk score (Figure 5) when merging our cohorts. We also compared our risk score to two previously pub- *Points were assigned by dividing each regression coefficient by the lished risk scores designed to predict ED in APL but to our lowest regression coefficient and rounding to the nearest integer. ED: early death, i.e. death within 30 days of diagnosis; WBC: white blood knowledge not yet externally validated. The suggested re- cells, ref: Reference. Haematologica | 107 July 2022

1533

ARTICLE - Predicting early death in APL

A. Österroos et al.

Figure 3. Risk score algorithm. The total sum of assigned points per variable level can rapidly be found in the tabulated chart above. Higher points indicate a higher risk for early death (ED). Colors indicate the predicted risk of ED as shown in the boxes below.

vised Sanz risk score by Lou and colleagues risk stratifies The Swedish cohort included all patients diagnosed with patients into four categories based on age, WBC and pla- APL nationwide between 1997 and 2020. To the best of our telets.17 The risk score developed by Cai and colleagues knowledge, this is the first study to utilize a truly popuassigns points based on age, WBC, platelets and LDH and was internally validated at the time of publication.18 In A B comparison to an AUROC of 0.77 (95% CI: 0.72-0.83) for our risk score, these models obtained lower AUROC of 0.75 (95% CI: 0.69-0.80) for the score by Cai et al. and 0.72 (95% CI: 0.66-0.77) for the score by Lou et al. (Figure 5).

Discussion ED emerges as the most important remaining challenge in APL treatment. The assessment of an individual’s risk of ED can have an impact on treatment strategy, medical intervention, level of surveillance and frequency of monitoring. Despite these facts, there are no clinically used and/or validated risk stratification tools to date concerning the risk of ED. Clinical trials often underestimate the true rate of ED, mainly due to the exclusion of patients presenting with high age, poor performance status, major hemorrhage or lifethreatening coagulopathy. This results in differences in ED rates between clinical trials and population-based reports.413 A tool to assess the risk of ED, and that is relevant for clinical routine, therefore needs to be based on populationbased data without the application of exclusion criteria. In this study, we used real-world data from the nationwide population-based Swedish AML Registry, one of the few AML registries in the world with close to 100% coverage.19

Figure 4. Risk score distribution and calibration plot. (A, B) Distribution of risk score groups in the training (A) and validation (B) cohorts, with green indicating low-risk, yellow indicating high-risk and red indicating very-high risk groups. (C) Calibration plot with the observed proportion of early deaths (ED) per risk group in the training cohort (blue) and validation cohort (green).

Haematologica | 107 July 2022

1534

ARTICLE - Predicting early death in APL

A. Österroos et al.

Figure 5. Receiver operating characteristic curves with areas under the curve (AUC) for the risk scores assessed: the score suggested here, the Sanz score, and the scores of. Lou et al.17 and Cai et al.18

lation-based APL cohort to develop a risk score for ED. Based on uni- and multivariable logistic regression analyses, a risk-per-quantile assessment and multivariable penalized logistic regression analysis to obtain ridge regression coefficients, a model was developed that identified the most significant risk factors and their optimal cut-off values. Patients were divided into three risk groups based on their total score points, identifying low-risk, high-risk and very high-risk patients with ED risks of <10%, 10-30% and >30%, respectively. The model identified WBC, platelets and age as the most significant risk factors and was, therefore, based on these parameters. Although studies differ somewhat with respect to risk factors for ED, WBC, platelets and age are recurrent in numerous studies.3,11,13,25 However, in contrast to the commonly used 10x109/L cut-off value for WBC, we found the risk of ED increased already at sub-normal WBC levels from approximately 2x109/L, then increased steeply within the normal WBC range. This suggests the need for increased vigilance for ED already at WBC values below or within the normal range. The increase of ED at WBC levels below 10x109/L in part explains why the score showed better discriminative power for ED compared to the Sanz risk score. Moreover, the risk score also performed better than the previously published ED risk scores17,18 which, to our knowledge, also remain externally unvalidated. Nonetheless, the model by Cai and col-

leagues could be validated in our study and may act as an alternative for assessing ED risk. For studies that include measurement of the WHO performance status, this is also associated with a higher risk of ED,26 as also in our study. However, the performance status at presentation is a subjective factor and it can also be strongly affected by already manifest intracerebral bleeding or thrombosis which makes it less suitable as a predictive factor for ED prevention (unpublished data). The presence of FLT3-ITD, CD2 expression and bcr3 PML-RARA transcript have also been associated with a higher risk of thrombosis although not uniformly.27-29 Such molecular data were not available for our cohorts but could potentially further improve the risk stratification. Limitations to our study include the use of retrospective data where the prediction of ED was made with the knowledge of outcome data that could introduce bias. Moreover, both the training and the validation cohorts only span a short time period after the introduction of ATO as a standard induction therapy. Induction treatment with ATO in combination with ATRA was introduced late during the study period and while ATO/ATRA-treated patients are included in this study, the low numbers of patients were not sufficient to accurately validate the score solely for ATO/ATRA-treated patients. This warrants further validation of the proposed risk score, preferably in prospective cohorts, in which patients are treated uniformly with ATRA

Haematologica | 107 July 2022

1535

ARTICLE - Predicting early death in APL

A. Österroos et al.

and ATO. Another factor that is not accounted for in our score, as well as in the other APL scores, is the presence of comorbidities and the risk of ED. Comorbidities likely play a role in the risk of ED but how to include them accurately in a score system remains a challenge. An obvious question is how a predictive score for ED in APL should be used in clinical routine and what type of measures should be taken in APL patients with a high or very high risk of ED. Current European LeukemiaNet (ELN) guidelines recommend treatment with fibrinogen (and/or cryoprecipitate or fresh-frozen plasma) to keep fibrinogen levels >100-150 mg/dL and platelet transfusions to maintain levels >30-50x109/L in all newly diagnosed APL patients. The score presented here can potentially be used to identify patients who need a more aggressive approach regarding monitoring frequency, level of surveillance and/or more rapid action when levels are below the thresholds. Even higher thresholds for platelet transfusions could potentially be used for patients at the highest risk, but such recommendations need to be further evaluated. Moreover, the need for frequent monitoring and quick action may require surveillance in intensive care units, which could be contacted already when a patient with an especially high ED risk is identified. We conclude that WBC values already below or within the normal range in APL patients are associated with an increased risk of ED which should be considered when assessing the risk of ED in APL patients. We here propose a risk score for ED in APL designed for the real-life situation and that can be used to risk-stratify patients when deciding upon level of surveillance and transfusion thresholds in order to overcome ED as one of the last standing barriers to successful APL treatment.

An online calculator for the score is provided at: aplearly-death.shinyapps.io/risk-score while Figure 3 could also be used directly to assess the risk for an individual patient. Disclosures No conflicts of interest to disclose. Contributions AÖ and SL designed the study. AÖ and SL conceived, designed and performed the analyses. AÖ designed the figures. AÖ and SL wrote the paper. AÖ, AE, MJ, VL, LW, PA, JC, SD, FL, LM, BU, EÖ, MH, GJ and SL contributed data for the training cohort and the Swedish AML Registry. TM, EA and FT collected data for the validation cohort. All authors critically reviewed the manuscript and approved the final draft. Acknowledgments The authors would like to thank everyone reporting to the Swedish AML Registry. The authors thank Mattias Mattsson, MD, PhD, at Uppsala University Hospital for valuable input during manuscript writing. Funding The Swedish AML Registry is supported by the Swedish Association of Local Authorities and Regions, Region Skåne, Regionalt Cancercentrum Syd, and the Swedish Cancer Society. Data-sharing statement Questions regarding data sharing should be addressed to the corresponding author.

References 1. Stahl M, Tallman MS. Acute promyelocytic leukemia (APL): remaining challenges towards a cure for all. Leuk Lymphoma. 2019;60(13):3107-3115. 2. Mantha S, Tallman MS, Soff GA. Whatʼs new in the pathogenesis of the coagulopathy in acute promyelocytic leukemia? Curr Opin Hematol. 2016;23(2):121-126. 3. Lehmann S. Early death in APL. In: Abla O, Lo Coco F, Sanz MA, editors. Acute Promyelocytic Leukemia: A Clinical Guide. Cham: Springer International Publishing; p71-86. 4. Sanz MA, Montesinos P, Vellenga E, et al. Risk-adapted treatment of acute promyelocytic leukemia with all-trans retinoic acid and anthracycline monochemotherapy: long-term outcome of the LPA 99 multicenter study by the PETHEMA group. Blood. 2008;112(8):3130-3134. 5. Asou N, Kishimoto Y, Kiyoi H, et al. A randomized study with or without intensified maintenance chemotherapy in patients with acute promyelocytic leukemia who have become negative for PML-RARα transcript after consolidation therapy: the Japan Adult Leukemia Study Group (JALSG) APL97 study. Blood. 2007;110(1):59-66.

6. Lengfelder E, Haferlach C, et al. for the German AML Cooperative Group (AMLCG). High dose ara-C in the treatment of newly diagnosed acute promyelocytic leukemia: long-term results of the German AMLCG. Leukemia. 2009;23(12):2248-2258. 7. Avvisati G, Lo-Coco F, Paoloni FP, et al. AIDA 0493 protocol for newly diagnosed acute promyelocytic leukemia: very long-term results and role of maintenance. Blood. 2011;117(18):4716-4725. 8. Burnett AK, Russell NH, Hills RK, et al. Arsenic trioxide and alltrans retinoic acid treatment for acute promyelocytic leukaemia in all risk groups (AML17): results of a randomised, controlled, phase 3 trial. Lancet Oncol. 2015;16(13):1295-1305. 9. Jeddi R, Kacem K, Ben Neji H, et al. Predictive factors of alltrans-retinoic acid related complications during induction therapy for acute promyelocytic leukemia. Hematology. 2008;13(3):142-146. 10. McClellan JS, Kohrt HE, Coutre S, et al. Treatment advances have not improved the early death rate in acute promyelocytic leukemia. Haematologica. 2012;97(1):133-136. 11. Park JH, Qiao B, Panageas KS, et al. Early death rate in acute

Haematologica | 107 July 2022

1536

ARTICLE - Predicting early death in APL

A. Österroos et al.

promyelocytic leukemia remains high despite all-trans retinoic acid. Blood. 2011;118(5):1248-1254. 12. Altman JK, Rademaker A, Cull E, et al. Administration of ATRA to newly diagnosed patients with acute promyelocytic leukemia is delayed contributing to early hemorrhagic death. Leuk Res. 2013;37(9):1004-1009. 13. Lehmann S, Deneberg S, Antunovic P, et al. Early death rates remain high in high-risk APL: update from the Swedish Acute Leukemia Registry 1997–2013. Leukemia. 2017;31(6):1457-1459. 14. Yanada M, Matsushita T, Asou N, et al. Severe hemorrhagic complications during remission induction therapy for acute promyelocytic leukemia: incidence, risk factors, and influence on outcome. Eur J Haematol. 2007;78(3):213-219. 15. Breen KA, Grimwade D, Hunt BJ. The pathogenesis and management of the coagulopathy of acute promyelocytic leukaemia. Br J Haematol. 2012;156(1):24-36. 16. Sanz MA, Lo Coco F, Martín G, et al. Definition of relapse risk and role of nonanthracycline drugs for consolidation in patients with acute promyelocytic leukemia: a joint study of the PETHEMA and GIMEMA cooperative groups. Blood. 2000;96(4):1247-1253. 17. Lou Y, Ma Y, Sun J, et al. Effectivity of a modified Sanz risk model for early death prediction in patients with newly diagnosed acute promyelocytic leukemia. Ann Hematol. 2017;96(11):1793-1800. 18. Cai P, Wu Q, Wang Y, Yang X, Zhang X, Chen S. An effective early death scoring system for predicting early death risk in de novo acute promyelocytic leukemia. Leuk Lymphoma. 2020;61(8):1989-1995. 19. Juliusson G, Antunovic P, Derolf Å, et al. Age and acute myeloid leukemia: real world data on decision to treat and outcomes from the Swedish Acute Leukemia Registry. Blood. 2009;113(18):4179-4187. 20. Iland HJ, Collins M, Bradstock K, et al. Use of arsenic trioxide in remission induction and consolidation therapy for acute promyelocytic leukaemia in the Australasian Leukaemia and Lymphoma Group (ALLG) APML4 study: a non-randomised phase 2 trial. Lancet Haemat. 2015;2(9):e357-e366.

21. Pavlou M, Ambler G, Seaman SR, et al. How to develop a more accurate risk prediction model when there are few events. BMJ. 2015;351:h3868. 22. Moons KGM, Altman DG, Reitsma JB, et al. Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis (TRIPOD): explanation and elaboration. Ann Intern Med. 2015;162(1):W1-73. 23. R Core Team (2017). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org. 24. Rashidi A, Goudar RK, Sayedian F, et al. All-trans retinoic acid and early mortality in acute promyelocytic leukemia. Leuk Res. 2013;37(10):1391-1392. 25. Song Y-H, Peng P, Qiao C, Zhang R, Li J-Y, Lu H. Low platelet count is potentially the most important contributor to severe bleeding in patients newly diagnosed with acute promyelocytic leukemia. Onco Targets Ther. 2017;10:4917-4924. 26. Mantha S, Goldman DA, Devlin SM, et al. Determinants of fatal bleeding during induction therapy for acute promyelocytic leukemia in the ATRA era. Blood. 2017;129(13):1763-1767. 27. Breccia M, Avvisati G, Latagliata R, et al. Occurrence of thrombotic events in acute promyelocytic leukemia correlates with consistent immunophenotypic and molecular features. Leukemia. 2007;21(1):79-83. 28. Chen C, Huang X, Wang K, Chen K, Gao D, Qian S. Early mortality in acute promyelocytic leukemia: potential predictors. Oncol Lett. 2018;15(4):4061-4069. 29. Montesinos P, de la Serna J, Vellenga E, et al. Incidence and risk factors for thrombosis in patients with acute promyelocytic leukemia. Experience of the PETHEMA LPA96 and LPA99 protocols. Blood. 2006;108(11):1503.

Haematologica | 107 July 2022

1537

ARTICLE - Cell Therapy and Immunotherapy

Therapeutic targeting of endoplasmic reticulum stress in acute graft-versus-host disease Eileen Haring,1,2,3 Geoffroy Andrieux,3,4 Franziska M. Uhl,1,2 Máté Krausz,5 Michele Proietti,5 Barbara Sauer,1 Philipp R. Esser,6 Stefan F. Martin,6 Dietmar Pfeifer,1 Annette Schmitt-Graeff,7 Justus Duyster,1 Natalie Köhler,1,8 Bodo Grimbacher,5,8,9,10 Melanie Boerries,3,4,11 Konrad Aumann,12 Robert Zeiser1,4,8 and Petya Apostolova1,3 Department of Medicine I, Medical Center - Faculty of Medicine, University of Freiburg, Freiburg; 2Faculty of Biology, Albert-Ludwigs-University, Freiburg; 3German Cancer Consortium (DKTK), Partner site Freiburg and German Cancer Research Center (DKFZ), Heidelberg; 4 Institute of Medical Bioinformatics and Systems Medicine, Medical Center - Faculty of Medicine, University of Freiburg, Freiburg; 5Institute for Immunodeficiency, Center for Chronic Immunodeficiency (CCI), Medical Center, Faculty of Medicine, Albert-Ludwigs-University, Freiburg; 6Allergy Research Group, Department of Dermatology, Medical Center - Faculty of Medicine, University of Freiburg, Freiburg; 7University of Freiburg, Freiburg; 8CIBSS – Center for Integrative Biological Signalling Studies, University of Freiburg, Freiburg; 9DZIF – German Center for Infection Research, Satellite Center Freiburg, Freiburg; 10RESIST – Cluster of Excellence 2155 to Hannover Medical School, Satellite Center Freiburg, Freiburg; 11 Comprehensive Cancer Center Freiburg (CCCF), Medical Center - Faculty of Medicine, University of Freiburg, Freiburg, and 12Department of Pathology, Medical Center – University of Freiburg, Faculty of Medicine, University of Freiburg, Freiburg, Germany 1

Correspondence: Petya Apostolova petya.apostolova@uniklinik-freiburg.de Received: January 17, 2021. Accepted: August 6, 2021. Prepublished: August 19, 2021. https://doi.org/10.3324/haematol.2021.278387 ©2022 Ferrata Storti Foundation Haematologica material is published under a CC BY-NC license

Abstract Acute graft-versus-host disease (GvHD) is a life-threatening complication of allogeneic hematopoietic cell transplantation (allo-HCT), a potentially curative treatment for leukemia. Endoplasmic reticulum (ER) stress occurs when the protein folding capacity of the ER is oversaturated. How ER stress modulates tissue homeostasis in the context of alloimmunity is not well understood. We show that ER stress contributes to intestinal tissue injury during GvHD and can be targeted pharmacologically. We observed high levels of ER stress upon GvHD onset in a murine alloHCT model and in human biopsies. These levels correlated with GvHD severity, underscoring a novel therapeutic potential. Elevated ER stress resulted in increased cell death of intestinal organoids. In a conditional knockout model, deletion of the ER stress regulator transcription factor Xbp1 in intestinal epithelial cells induced a general ER stress signaling disruption and aggravated GvHD lethality. This phenotype was mediated by changes in the production of antimicrobial peptides and the microbiome composition as well as activation of pro-apoptotic signaling. Inhibition of inositol-requiring enzyme 1α (IRE1α), the most conserved signaling branch in ER stress, reduced GvHD development in mice. IRE1α blockade by the small molecule inhibitor 4m8c improved intestinal cell viability, without impairing hematopoietic regeneration and T-cell activity against tumor cells. Our findings in patient samples and mice indicate that excessive ER stress propagates tissue injury during GvHD. Reducing ER stress could improve the outcome of patients suffering from GvHD.

Introduction Acute graft-versus-host disease (GvHD) is a life-threatening complication of allogeneic hematopoietic cell transplantation (allo-HCT). In particular, GvHD of the gastrointestinal tract (GI-GvHD) remains one of the most frequent causes of allo-HCT-related morbidity and mortality.1 GI-GvHD results from a complex multi-step crosstalk between extensive epithelial tissue damage in the patient and activation of the allo-reactive immune system

transferred with the donor graft.2 Enterocytes are subjected to cellular stress and undergo apoptosis. As a result, patients can develop diarrhea, dehydration, intestinal bleeding, hypalbuminemia, and generalized infections.2 Recently, it has been shown that GvHD is also associated with changes of the microbiome diversity and composition.3 Standard treatment for GI-GvHD consists of corticosteroids or other compounds that target immune cell activation.4 Several factors supporting intestinal repair mechanisms have been identified in pre-clinical studies,

Haematologica | 107 July 2022

1538

ARTICLE - Targeting ER stress in graft-versus-host disease including IL-22,5,6 R-spondin7,8 and glucagon-like peptide 2.9 Recently, we have demonstrated a protective role for bile acids, which reduce cytokine-mediated intestinal injury and decrease intestinal antigen presentation.10 These novel developments show as a proof-of-principle that local regulation of repair mechanisms and inflammation is a successful approach to treat GI-GvHD and possibly other inflammatory diseases of the intestine. In this study, we set out to decipher the role of endoplasmic reticulum (ER) stress, a cellular stress reaction, in the context of GI-GvHD. ER stress occurs upon accumulation of unfolded proteins in the ER lumen which can result from hypoxia, tissue damage, or pathogen exposure. These unfolded proteins are sensed by three molecules in the ER membrane which activate signaling cascades, commonly known as the unfolded protein response (UPR).11 The most conserved UPR transducer is the kinase inositol-requiring enzyme 1 α (IRE1α). This kinase has a ribonuclease activity and splices the mRNA of the transcription factor X-box binding protein 1 (XBP1), which then translocates to the nucleus and regulates the expression of chaperones, foldases and enzymes for lipid metabolism.12,13 The two other UPR branches are mediated by the protein kinase RNA-like ER kinase (PERK) and the activating transcription factor 6 (ATF6).11 Their downstream signaling cascades lead to a reduction of the intracellular protein overload. The fundamental task of the UPR is to assist the cell in coping with the large amount of unfolded proteins. However, prolonged activation of the UPR also activates pro-apoptotic kinases and transcription factors.11,14,15 Another mechanism contributing to ER stress-induced cell death is mediated via regulated inositol-requiring enzyme 1-dependent decay of mRNA (RIDD).16 During this process, IRE1α splices and thus inactivates multiple mRNA which results in cell death. These opposite effects of ER stress on cell fate underline the fact that a tight regulation of the UPR is necessary to maintain physiological balance and homeostasis in stressed cells (Figure 1A). Our study aimed to close the gap in understanding the function of intestinal UPR and its therapeutic potential in GI-GvHD. We observed that ER stress occurred during GvHD development in mice and humans and correlated with GI-GvHD onset and severity. Using intestinal organoids and a genetic conditional knockout mouse model, we found that dysregulated chronic ER stress reduced intestinal cell viability and aggravated GvHD severity by altercating anti-microbial peptide production, the microbiome composition and pro-apoptotic pathway activity. Pharmacological inhibition of IRE1α, one of three ER stress signaling pathways, improved GvHD outcome by directly protecting the intestinal epithelium. Importantly, this intervention did not impair hematopoietic regeneration, T-cell proliferation or anti-tumor killing capacity.

E. Haring et al. Our study proposes the inhibition of this ER stress pathway as a novel approach for the treatment of GI-GvHD.

Methods Mice BALB/c (H-2Kd) and C57BL/6 (H-2Kb) mice were purchased from Janvier Labs (Le Genest-Saint-Isle, France). Luciferase-transgenic C57BL/6 mice (H-2Kb) were bred in the animal facility of the Center for Clinical Research at the Medical Center – University of Freiburg, Germany. Mice carrying a floxed Xbp1 allele (Xbp1fl/fl) have been previously described17 and were a kind gift from Dr. Laurie Glimcher, Cornell University, USA. B6. Cg-Tg(Vil1-cre)997Gum/J mice expressing the Cre recombinase in villus and crypt intestinal epithelial cells (VilCre) were purchased from the Jackson Laboratory. VilCreXbp1fl/fl mice and Xbp1fl/fl littermates were generated by crossing these two strains. All animals were housed under specific pathogen-free conditions at the animal facility of the Center for Clinical Research (ZKF, Freiburg, Germany). Genotypes were confirmed via polymerase chain reaction (PCR). All animal protocols (G16/018, G-17/063, G-18/063; X-13/07J; X-15/10A; X20/06K) were approved by the Federal Ministry for Nature, Environment and Consumer Protection of the state of BadenWürttemberg, Germany. All other methods are provided in the Online Supplementary Appendix.

Results Unfolded protein response is activated during early graft-versus-host disease development in mice We first characterized the UPR activity in the intestine of healthy mice and after GvHD induction by quantifying the splicing of Xbp1, which is one of the initiating events of UPR signaling. We observed increased Xbp1 splicing in the colon of mice 7 days after allo-HCT when compared to untreated mice (Online Supplementary Figure S1A-C). In order to study the exact time course of UPR activation, we performed quantification of the established ER stress markers Xbp1s/Xbp1u (spliced Xbp1 in relation to unspliced Xbp1) and Hspa5 (encoding for the chaperone GRP78, also known as BIP) at multiple time points after allo-HCT (Figure 1B). We further set out to distinguish whether ER stress was induced as a consequence of the conditioning treatment alone (total body irradiation [TBI]) or by the allo-immune reaction that marks GvHD development. We observed an upregulation of the UPR-related markers in the small intestine (Online Supplementary Figure S1D) and in the colon (Figure 1C) of both TBI and GvHD mice when compared to healthy mice. While ER stress markers were

Haematologica | 107 July 2022

1539

ARTICLE - Targeting ER stress in graft-versus-host disease

E. Haring et al.

Figure 1. Graft-versus-host disease induction leads to endoplasmic reticulum stress in the murine intestine. (A) Schematic overview of the three different unfolded protein response (UPR) branches induced by endoplasmic reticulum (ER) stress, created with BioRender.com. (B, C) BALB/c mice received either only total body irradiation (TBI) with 10 Gy or underwent complete allogeneic hematopoietic cell transplantation (allo-HCT) as described in Methods with 5x106 bone marrow (BM) cells and 3x105 CD4+ and CD8+ T cells isolated from the spleen of a C57BL/6 donor. Untreated BALB/c mice were used as a control. (B) Overview of the experimental setup, created with BioRender.com. (C) Quantitative real-time polymerase chain reaction (PCR) analysis of the mRNA expression of selected UPR marker genes in the colon with Actb as a reference gene. Samples were isolated on different time points after TBI or allo-HCT as indicated. The dashed line represents the expression level in untreated mice, which Continued on following page. Haematologica | 107 July 2022

1540

ARTICLE - Targeting ER stress in graft-versus-host disease

E. Haring et al.

was set to 1. Data were pooled from n=11 mice for day 0 (d0) and n=5-7 mice/group for the other time points. The P-values were calculated using the ordinary one-way ANOVA with correction for multiple comparisons. Statistical comparisons between “GvHD” and “untreated” are highlighted in red color, statistical comparisons between “TBI” and “untreated” are highlighted in grey color, *P<0.05, ***P<0.0001. (D) Representative immunohistochemistry staining images for GRP78 in murine colon tissue sections. Sections from an untreated mouse (“no inflammation”) and from an allo-transplanted BALB/c mouse (“GvHD”) on d7 after allo-HCT are shown. Arrows point to GRP78-expressing cells. Scale bars 100 mM. (E) Quantification of the GRP78 expression score. Each dot represents a single mouse, n=4-5 mice/group. The P-value was calculated using a two-tailed unpaired Mann Whitney U test. (F) Representative immunohistochemistry staining images for CHOP in murine colon tissue sections. Sections from an untreated mouse (“no inflammation”) and from an allo-transplanted BALB/c mouse (“GvHD”) on d7 after allo-HCT are shown. Arrows point to GRP78-expressing cells. Scale bars 100 mM. (G) Quantification of the CHOP expression score. Each dot represents a single mouse, n=4-5 mice/group. The P-value was calculated using a two-tailed unpaired Mann Whitney U test.

induced by TBI as early as day 4, they dropped back to normal more rapidly than in mice with GvHD. Especially 6-8 days after GvHD induction in our model, Xbp1s/Xbp1u and Hspa5 expression were significantly higher in mice developing GvHD compared to TBI mice and healthy controls (Figure 1C). After that time point, ER stress levels in the colon declined and were similar to baseline (Figure 1C) whereas in the small intestine a second increase was observed on day 14 after allo-HCT (Online Supplementary Figure S1D). These data indicate an active UPR during early GI-GvHD development. While we observed that TBI alone can induce the UPR, transfer of T cells aggravated ER stress additionally. In line with this, immunohistochemistry for GRP78 showed increased expression in the colon of GvHD mice when compared to healthy colon tissue (Figure 1D to E). Although the aim of the UPR is to support the cellular survival in stress conditions by reducing the load of unfolded proteins, pro-apoptotic cascades are engaged during prolonged ER stress. C/EBP homologous protein (CHOP, encoded by the gene Ddit3) is one of the transcription factors mediating apoptosis in the context of the UPR. We found that CHOP is expressed by the healthy colonic epithelium, especially at the top of the crypts. Interestingly, the expression in GvHD-developing mice extended to the whole crypt (Figure 1F, G). As epithelial regeneration has its origin at the bottom of the crypts where intestinal stem cells reside, CHOP-induced apoptosis in this area might reduce the capacity of the intestine to regenerate itself. Together, we observed that GI-GvHD development in mice is marked by increased levels of ER stress. Unfolded protein response is activated during graft-versus-host disease (GvHD) onset in humans and correlates with GvHD severity GI-GvHD is a major factor contributing to morbidity and mortality in allo-HCT recipients. In order to assess whether the UPR was activated in humans, we examined biopsies of GvHD patients for the expression of GRP78 and CHOP. Healthy tissue and biopsies from colitis patients were used as negative and positive controls, respectively. In the uninflamed intestine, expression of GRP78 was confined to immune cells that were interspersed between

crypts, and staining in epithelial cells was low (Figure 2A). Confirming a previously shown connection between ER stress and colitis,17 GRP78 expression in the epithelium was strong in colitis samples. When analyzing biopsies from GvHD patients, we observed increased expression in higher-grade GvHD compared to lower-grade GvHD (Figure 2A and B). Similarly, healthy colon tissue showed only low expression of CHOP. We observed a more intensive CHOP signal in biopsies from patients with GvHD grade 2-3, especially from injured crypts that were undergoing cell death (Figure 2C). Histology data were supported by the analysis of a publicly available RNA sequencing data set (GSE134662), generated by Holtan et al.18 We found an upregulation of various ER stress-related genes when comparing gut biopsies from newly diagnosed or steroid-refractory GvHD to healthy tissue (Online Supplementary Figure S2A, B). Gene set enrichment analysis confirmed a differential expression of genes matched to the GO term "Response to endoplasmic reticulum stress" (Online Supplementary Figure S2C). Collectively, our data show that ER stress occurs during GvHD development in mice and humans, and that its levels correlate with GvHD severity. Chronic endoplasmic reticulum stress aggravates intestinal cell death and graft-versus-host disease Based on our observation that the UPR is activated during GvHD, we investigated the cellular impact of ER stress induction on the intestine using murine small intestinal organoids that were established and cultured as previously described.19 Exposure of organoids to the ER stress inducer tunicamycin resulted in cell death as observed by microscopy and flow cytometry (Figure 3A-D). Release of DAMP from dying cells contributes to the activation of antigenpresenting cells in the context of GvHD.20,21 We detected increased concentrations of the DAMP uric acid in the supernatants of tunicamycin-treated organoids (Figure 3E). Next, we investigated if UPR pathway disruption affects GvHD outcome in a murine allo-HCT model. Since the IRE1α pathway is the most conserved branch among different species, we focused on the key transcription factor XBP1 which is downstream of IRE1α and gets activated by the RNase activity of this enzyme. In order to test whether

Haematologica | 107 July 2022

1541

ARTICLE - Targeting ER stress in graft-versus-host disease intestinal Xbp1 plays a role for the development of GvHD, we crossed VilCre mice bearing the Cre-recombinase under the control of the Villin-promoter with Xbp1fl/fl mice carrying a floxed exon 2 in the Xbp1 gene. The specific deletion of the exon 2 of Xbp1 in the intestine of VilCre Xbp1fl/fl mice (here referred to as Xbp1DIEC) was confirmed by quantitative PCR analysis of the intestine in comparison to Xbp1fl/fl (here referred to as Xbp1+/+) littermate controls (Online Supplementary Figure S3A). Subsequently, we asked if intestinal Xbp1 deletion had an impact on the development of GvHD. Survival of Xbp1DIEC mice after induction of GvHD was significantly reduced when compared to Xbp1+/+ littermate controls (Figure 3F, G). Furthermore, his-

E. Haring et al. topathological GvHD scores were increased in Xbp1DIEC mice in two different allo-HCT models with TBI or chemotherapy as a conditioning treatment (Figure 3H; Online Supplementary Figure S3B, C). Altogether, these data show that chronic ER stress induces intestinal cell death and results in deteriorated GvHD outcome in different murine models. Intestinal deletion of Xbp1 alters anti-microbial peptides and microbiome composition of the intestine Next, we explored the mechanism by which chronic ER stress increased GvHD. Gene expression analysis revealed that multiple genes belonging to the defensin (Def) family

Figure 2. Levels of endoplasmic reticulum stress markers correlate with graft-versus-host disease severity in patients. (A, B) Immunohistochemistry staining for GRP78 in human colon tissue samples. (A) Representative images from healthy and diseased colon tissues as indicated. Arrows point to GRP78-expressing cells. The colitis sample served as a positive control. Scale bars 100 mm. (B) Quantification of the GRP78 expression score in graft-versus-host disease (GvHD) samples. Statistical analysis of n=7 (GvHD grade 1-2) and n=5 (GvHD grade 3) patient biopsies. The P-value was calculated using the two-tailed unpaired Mann Whitney U test. (C) Representative images of an immunohistochemistry staining for CHOP in human healthy and diseased colon tissues as indicated. Arrows point to CHOP-expressing cells. Scale bars 100 mm. Haematologica | 107 July 2022

1542

ARTICLE - Targeting ER stress in graft-versus-host disease

E. Haring et al.

Figure 3. Chronic endoplasmic reticulum stress induces intestinal cell death and a more severe graft-versus-host disease phenotype. (A) Representative images of BALB/c intestinal organoids treated for 24 hours (h) with 1 mg/mL tunicamycin or vehicle control. (B) Quantification of living organoids after vehicle or tunicamycin treatment as in (A) performed by manual microscope counting. Data were normalized to the vehicle group. Statistical analysis of n=3 biologically independent experiments. The Pvalue was calculated using the two-tailed unpaired Student’s t-test. (C, D) Intestinal organoids were cultured as described in (A). Organoids were digested and the proportion of dead cells was determined by flow cytometry. (C) Quantification of the percentage of dead cells. The P-value was calculated using the two-tailed unpaired Student’s t-test. (D) Representative flow cytometry dot plots. (E) Organoids were treated with 0.15 mg/mL tunicamycin for 12 h and afterwards allowed to rest for further 24 h in normal medium. Levels of uric acid were measured in the supernatant medium, in which the organoids were rested. Concentrations were measured in total ng/mL, n=2-3 individual biological replicates/group. The P-value was calculated using the two-tailed unpaired Student’s t-test. (F) Transplantation model with BALB/c (H-2Kd) as donor and Xbp1DIEC or Xbp1+/+ mice as recipients, created with Biorender.com. (G) Survival of Xbp1DIEC and Xbp1+/+ mice after allogeneic hematopoietic cell transplantation (allo-HCT). Data were pooled from three independent experiments with n=14 mice/group. The P-value was calculated using the two-sided Mantel-Cox test. (H) Histopathology scores of the small intestine and colon from Xbp1DIEC and Xbp1+/+ recipients on day 14 after allo-HCT. Data are pooled from n=5 mice/group. Each dot represents a single mouse. The Pvalues were calculated using the two-tailed unpaired Mann Whitney U test. Haematologica | 107 July 2022

1543

ARTICLE - Targeting ER stress in graft-versus-host disease were downregulated in Xbp1DIEC mice with GvHD when compared to littermate controls (Figure 4A). Amongst these were Defa1, Defa5, Defa20, Defa21, Defa22, Defa30, Defa34, Defa39, and Defa41. Defensins are anti-microbial peptides, secreted by Paneth cells. A decrease in defensin gene expression is in line with previously reported loss of Paneth cells in Xbp1DIEC mice.17 We next explored how the expression of anti-microbial peptides varied during GvHD development. We observed reduced mRNA expression of Defa1, Defa4, lysozyme (Lyz) and regenerating islet-derived protein 3γ (Reg3g) in untreated Xbp1DIEC mice compared to Xbp1+/+ littermates. GvHD induction reduced initially the expression of all four genes (Figure 4B). On day 14 after allo-HCT, anti-microbial peptide expression in Xbp1+/+ mice had returned to normal levels whereas it stayed significantly lower in Xbp1DIEC mice (Figure 4B). Previous studies have suggested that the intestinal dysbiosis, which is frequently observed in GvHD, results at least in part from the reduction of Paneth cells and their products in intestinal crypts.22 We hypothesized that reduced anti-microbial peptide expression in Xbp1DIEC animals might be accompanied by changes in the microbiome. We analyzed the microbiome of Xbp1DIEC and Xbp1+/+ control mice with or without GvHD induction. We observed changes on the phylum level in the untreated condition with a shift towards more Bacteroidetes and less Firmicutes in Xbp1DIEC mice (Figure 4C, D). Loss of Firmicutes, and specifically some Clostridia species has previously been linked to GvHD.23 We found that within the Firmicutes phylum, the Clostridia class was particularly reduced with decreased abundances of the families Ruminococcaceae and Lachnospiraceae in Xbp1DIEC mice (Figure 4E), the latter of which has been associated with a protective effect against lethal GvHD.24,25 In the early phase after allo-HCT, there were no differences in the microbiome composition between Xbp1+/+ and Xbp1DIEC mice (Figure 4C, D), mirroring the comparable levels of anti-microbial peptides that we observed at this time point (Figure 4B). These data indicate that ER stress signaling dysregulation caused by loss of Xbp1 induces a loss of anti-microbial peptides, coinciding with a shift in the microbiome composition in the steady-state that is equivalent to the microbiome disruption during lethal GvHD development. Loss of Xbp1 promotes a shift towards pro-apoptotic unfolded protein response signaling which can be reversed by a pharmacological intervention We further hypothesized that the observed aggravated GvHD phenotype might be caused by upregulation of proapoptotic UPR signaling in Xbp1DIEC mice. RIDD is an IRE1αdependent process that gets activated upon prolonged ER stress. IRE1α RNase activity cleaves several mRNA with subsequent apoptosis induction. We observed signifi-

E. Haring et al. cantly decreased levels of the RIDD targets bone marrow stromal cell antigen 2 (Bst2), lysosomal-associated membrane protein 1 (Lamp1), carboxylesterase 1F (Ces1f), solute carrier family 35, member B1 (Slc35b1), ribophorin II (Rpn2) and heparan-α-glucosamide N-acetyltransferase (Hgsnat) in the intestine of Xbp1DIEC recipient mice (Figure 4A; Figure 5A, B). Consistent with the hypothesis that Xbp1 deletion favors other, pro-apoptotic ER stress branches, Ddit3, encoding for CHOP, was upregulated (Figure 5C). We then explored the question of whether compensatory over-activation of IRE1α-related RIDD mediates the aggravated GvHD phenotype in mice with intestinal Xbp1 deletion. We employed a small molecule inhibitor of IRE1α, 4m8c (7-hydroxy-4-methyl-2-oxo-2H-1benzopyran-8-carboxaldehyde). This compound blocks the RNase activity of IRE1α, including RIDD, but spares its kinase activity. We observed that treatment of Xbp1DIEC mice with 4m8c restored the survival upon GvHD induction to the level observed in Xbp1+/+ littermates (Figure 5D). Taken together, we show that selective inactivation of intestinal Xbp1 results in a major dysregulation of the UPR with a pro-apoptotic signature. Inhibition of IRE1α improves graft-versus-host disease outcome without impairing immune reconstitution or the graft-versus-leukemia effect We observed that inhibition of IRE1α reverses the toxicity of chronic ER stress in Xbp1DIEC mice and therefore studied the potential of the IRE1α inhibitor as a therapeutic compound for general GvHD development and outcome. In vivo treatment of mice who had developed GvHD with the IRE1α inhibitor 4m8c ameliorated GvHD severity, as observed by prolonged survival, reduced clinical GvHD signs and decreased histopathology score (Figure 6A, D). A beneficial effect, albeit not as strong as with 4m8c, was observed with a second IRE1α inhibitor, B-I09 (Online Supplementary Figure S4A). On the contrary, inhibition of PERK signaling did not improve the GvHD outcome (Online Supplementary Figure S4B). Given the beneficial effect of IRE1α inhibition over PERK inhibition, we focused on the effects of the former pathway in our further analysis. In line with the hypothesis that inhibition of IRE1α reduces GvHD-related intestinal apoptosis, 4m8c decreased intestinal cell death in response to treatment with the pro-inflammatory cytokine TNF or the ER stress inducer tunicamycin in vitro in the small intestinal cell line MODEK (Figure 6E, G). Long-term success of allo-HCT strongly depends on an intact immune reconstitution that guarantees protection against pathogens and malignancy relapse. We therefore next analyzed whether IRE1α inhibition had an impact on peripheral blood leukocyte and thymus regeneration (Figure 7A). We found that, as expected, hemoglobin, platelets and white blood cells were significantly decreased in the

Haematologica | 107 July 2022

1544

ARTICLE - Targeting ER stress in graft-versus-host disease

E. Haring et al.

A B

Continued on following page. Haematologica | 107 July 2022

1545

ARTICLE - Targeting ER stress in graft-versus-host disease

E. Haring et al.

Figure 4. The deletion of intestinal Xbp1 disrupts antimicrobial peptide production and induces changes in the microbiome. (D) Heatmap showing the top 50 differentially expressed genes in the small intestine of Xbp1DIEC and Xbp1+/+ recipient mice on day 14 after allogeneic hematopoietic cell transplantation (allo-HCT) performed as shown in Figure 3F. Color legend “Z-score” indicates the row-wise scaling of the normalized intensity, whereas “fc” indicates the log2 fold change between Xbp1+/+ and Xbp1DIEC recipient mice. (B) Quantitative real-time polymerase chain reaction (PCR) analysis of the mRNA expression of the antimicrobial peptides Defa1, Defa4, Lyz and Reg3g in the indicated groups with Actb as a reference gene. Data were pooled from n=5-14 mice/group. Each dot represents a single mouse. The P-values were calculated using the ordinary one-way ANOVA with correction for multiple comparisons. (C) Simplified overview of bacteria taxonomy with a focus on the phylum Firmicutes. (D, E) Microbiome analysis of untreated (CTRL) and graft-versus-host disease (GVHD)-developing Xbp1DIEC and Xbp1+/+ mice on day 5 after allo-HCT. Data were pooled from n=5 independent biological replicates. The P-values were calculated using an ordinary one-way ANOVA with correction for multiple comparisons. (D) Phylum composition. (E) Relative abundance of the phyla Bacteroidetes and Firmicutes, the class Clostridia and the families Ruminococcaceae and Lachnospiraceae.

Figure 5. Loss of Xbp1 favors pro-apoptotic signaling. (A-D) Xbp1DIEC and Xbp1+/+ mice underwent bone marrow transplantation (BMT) as in Figure 3F. (A) Volcano plot showing gene expression in Xbp1DIEC vs. Xbp1+/+ recipient mice on day 14 after allogeneic hematopoietic cell transplantation (allo-HCT). Red circles highlight mRNA that are targets of RIDD. (B) Quantitative real-time polymerase chain reaction (PCR) analysis of the mRNA expression of the RIDD target genes Lamp1, Rpn1 and Bst2 in the indicated groups with Actb as a reference gene. Data were pooled from n=10 mice in the Xbp1+/+ group and n=7 mice in the Xbp1DIEC group. Each dot represents a single mouse. P-values were calculated using the two-tailed unpaired Student’s t-test. (C) Quantitative real-time PCR analysis of the mRNA expression of the pro-apoptotic endoplasmic reticulum (ER) stress gene Ddit3 (encodes for CHOP) with Actb as a reference gene. Data were pooled from n=6 mice/group. Each dot represents a single mouse. P-values were calculated using the two-tailed unpaired Student’s t-test. (D) Survival of Xbp1DIEC and Xbp1+/+ mice after allo-HCT and treatment with the IRE1α inhibitor 4m8c. Data were pooled from three independent experiments with n=13-14 mice/group. Each dot represents a single mouse. P-values were calculated using the two-sided Mantel-Cox test. Haematologica | 107 July 2022

1546

ARTICLE - Targeting ER stress in graft-versus-host disease

E. Haring et al.

Continued on following page. Haematologica | 107 July 2022

1547

ARTICLE - Targeting ER stress in graft-versus-host disease

E. Haring et al.

Figure 6. Pharmacological IRE1α inhibition improves graft-versus-host disease outcome in mice. (A) Transplantation model with C57BL/6 (H-2Kb) as donor and BALB/c (H-2Kd) as recipient. Recipient animals were treated with 10 mg/kg body weight IRE1α inhibitor 4m8c (from day 0 until day 10 after bone marrow transplantation [BMT]) or an equal volume of vehicle by a daily intraperitoneal (i.p.) injection. Schematic overview created with BioRender.com. (B) Survival of BALB/c mice transplanted and treated with 4m8c as shown in (A). Data were pooled from two independent experiments, n=14 mice/group. The P-value was calculated using the two-sided Mantel-Cox test. (C) Clinical graft-versus-host disease (GvHD) scores combined from weight loss, skin lesions, hunching posture, dull fur and diarrhea in BALB/c mice transplanted as shown in (A). Data were pooled from n=8 mice/group. The P-values were calculated using the two-tailed unpaired Mann-Whitney U test. (D) Histopathology scores of the small intestine and colon from BALB/c mice transplanted and treated as in (A) on day 14 after allogeneic hematopoietic cell transplantation (allo-HCT). Data are pooled from n=7 (BM+vehicle group) and n=10 mice (BM+Tcell+vehicle and BM+Tcell+4m8c groups). P-values were calculated using the two-tailed unpaired Mann-Whitney U test. (E, F) Analysis of MODE-K cell viability after treatment with TNF (20 ng/mL) ± 4m8c (concentrations as indicated) for 48 hours (h) performed by flow cytometry. (E) Representative flow cytometry dot plots. (F) Quantification of the percentages of dead cells. Data were normalized to the TNFtreated group. Statistical analysis of n=3 biologically independent experiments performed in technical duplicates or triplicates. The P-values were calculated using the ordinary one-way ANOVA test. (G) Analysis of MODE-K cell viability after treatment with 0.15 mg/mL tunicamycin ± 4m8c (concentrations as indicated) for 48 hours. Quantification of the percentages of dead cells. Statistical analysis of n=3 independent experiments each performed in technical triplicates. Data were normalized to the tunicamycin treatment group. The P-values were calculated using the ordinary one-way ANOVA test.

blood of mice on day 14 after allo-HCT when compared to untreated mice. By day 29, hemoglobin and platelets had recovered to normal levels, whereas the white blood cells, and specifically B and T lymphocytes remained reduced by about 80% (Figure 7B, C). Importantly, there was no difference between vehicle- and 4m8c-treated animals (Figure 7B, C). Peripheral blood T-cell differentiation into naïve, effector and central memory T cells was also not affected by IRE1α inhibition (Online Supplementary Figure S5A). Thymus pathology following TBI and GvHD development can impair the T-cell selection process leading to the emergence of an autoreactive population and a failure to generate functional T cells that fight cancer cells and pathogens. Analysis of the thymic T cell populations showed that, independently of 4m8c treatment, CD4+ CD8+ double positive (DP), that are the most frequent population in the healthy thymus, represent only 8% of the cells in allo-HCT recipients. In the same time, the single positive CD4+ population had expanded to 72% (Figure 7D) after alloHCT. In the non-hematopoietic compartment, we observed no changes in the percentages of thymus epithelial cells or fibroblasts in 4µ8c- compared to vehicle-treated animals (Figure 7E). Overall, these data provide evidence that hematopoietic and thymus regeneration is not affected by administration of 4m8c to d29, although longterm studies of lymphocyte regeneration and thymic populations would be required to provide definitive confirmation. Long-term malignancy control by allo-reactive T cells is essential for the success of allo-HCT. We assessed the expansion of allo-reactive T cells by bioluminescence imaging and flow cytometry and found that 4α8c did not impact T-cell expansion and differentiation in vivo (Figure 8A-D). We next tested whether IRE1α inhibition interfered with the graft-versus-tumor effect. We activated T cells via co-culture with allogeneic bone marrow-derived dendritic cells in the presence or absence of 4m8c. After 72 hours of activation, the T cells were incubated

with A20 lymphoma cells and the killing efficacy was evaluated by flow cytometry (Figure 8E). We observed a similar capacity of vehicle- and 4m8c-pretreated T cells to eliminate A20 cells (Figure 8F). Collectively, these data show that pharmacological inhibition of IRE1α does not impair the killing capacity of allo-reactive T cells.

Discussion Acute GvHD affects about 50% of allo-HCT recipients with more than 10% of all patients suffering from a severe course of the disease.1 Recent efforts to improve the efficacy of treatment and prophylaxis range from blockade of T-cell activation,26,27 to inhibition of cytokine signaling28,29 and epigenetic therapies.30 In this study, we hypothesized that intestinal ER stress could be a target for novel GvHD therapies. We observed an upregulation of multiple ER stress markers following irradiation and GvHD induction. This is in line with a previous report showing upregulation of chaperones after allogeneic transplantation.31 Upregulation of ER stress markers has also been observed in the lacrimal gland, small intestine, skin and liver of mice developing chronic GvHD.32 In an extended analysis of a publicly available RNA sequencing data set (GSE134662),18 we found increased expression of several UPR genes in human GvHD samples compared to healthy tissue. Furthermore, we found that the expression of two ER stress markers, GRP78 and CHOP, directly correlated with GvHD severity in a cohort of patients at our center. The association of ER stress marker expression with higher-grade GvHD in patients strengthens the implication that modulation of intestinal UPR might be a successful treatment approach. We next hypothesized that chronic unresolved ER stress might aggravate GvHD. We performed experiments using murine intestinal organoids as an in vitro model, and with a conditional knockout model of the transcription factor Xbp1 in vivo. Induction of ER stress by a chemical compound in-

Haematologica | 107 July 2022

1548

ARTICLE - Targeting ER stress in graft-versus-host disease

E. Haring et al.

Figure 7. IRE1α inhibition does not impair thymic and peripheral blood regeneration. (A-E) BALB/c mice underwent transplantation and were treated with 4m8c as described in Figure 6A. On day 14 and day 29 after allogeneic hematopoietic cell transplantation (allo-HCT), leukocyte subpopulations and phenotype in the peripheral blood was analyzed. In addition, thymus reconstitution was assessed on day 29. Data were pooled from n=4 mice in the untreated group and n=9 mice each for the other groups. Each dot represents a single mouse. P-values were calculated using the two-tailed unpaired Student’s t-test. ns: not significant. (A) Schematic overview of the experiment, created with BioRender.com. (B) Hemoglobin, platelet and white blood cell count including differentiation into granulocytes, monocytes and lymphocytes from mice treated as shown in (A). (C) Percentages of leukocyte subpopulations assessed by flow cytometry. Pre-gating on single cells – living cells – CD45+ was performed. Neutrophils were defined as CD11b+ Ly6G+. Monocytes were defined as CD11b+ Ly6G–. (D) Percentages of double positive (DP, CD4+CD8+), single CD4+, single CD8+ and double negative (DN, CD4–CD8–) T cells in the thymus of recipient mice. Pre-gating on single cells- living cells – CD45+ – CD3+ was performed. (E) Percentages of medullary thymic epithelial cells (mTEC, CD45– EpCAM+ Ly51+), cortical thymic epithelial cells (cTEC, CD45- EpCAM+ Ly51–) and fibroblasts (CD45–CD140+) in the thymus of recipient cells. Pre-gating on single cells and living cells was performed. Haematologica | 107 July 2022

1549

ARTICLE - Targeting ER stress in graft-versus-host disease

E. Haring et al.

Continued on following page. Haematologica | 107 July 2022

1550

ARTICLE - Targeting ER stress in graft-versus-host disease

E. Haring et al.

Figure 8. IRE1α inhibitor treatment does not affect alloreactive T-cell expansion. (A, B) BALB/c mice underwent transplantation as described in Figure 6A using luciferase-transgenic T cells. (A) Representative bioluminescence (BLI) images from several time points. (B) Quantification of the BLI measurement is shown. Signal was quantified from the whole body in photons/sec/mouse. Data were pooled from n=6 mice/group. Each dot represents a single mouse. One representative result from three independent experiments is shown. (C, D) Flow cytometry analysis of T cells isolated from recipient spleens and colons on day 14 after allogeneic hematopoietic cell transplantation (allo-HCT) (C57BL/6 in BALB/c model). (C) Representative flow cytometry dot plots from the spleen are shown. (D) Combined data pooled from n=5-6 mice/group for the percentage of donor CD4+ and CD8+ T cells in the spleen (upper panel) and colon lamina propria (lower panel). Each dot represents a single mouse. One representative result from two independent experiments is shown. P-values were calculated using the two-tailed unpaired Student’s t-test. ns: not significant. (E) C57BL/6 CD8+ T cells were activated in a co-culture with allogeneic BALB/c bone marrow-derived dendritic cells in the presence or absence of 5 mM 4m8c for 72 hours (h). CD8+ T cells were then incubated with GFP+ A20 lymphoma cells and the percentage of dead cells after 24 h of culture was assessed by flow cytometry. Schematic overview created with BioRender.com. (F) Percentage of dead A20 cells alone or at three different co-culture ratios with CD8+ T cells activated as described in (E). The experiment was repeated three times with four or five replicates each. One representative result is shown.

duced cell death in intestinal organoids with an increased release of the DAMP uric acid. Production of uric acid as a consequence of tissue damage during the conditioning treatment has been shown to promote GvHD by activating antigen-presenting cells.20,21 Short-term ER stress induces a stress-associated transcriptomic reprogramming with genes related to cellular signaling, expression regulation, metabolism, the cytoskeleton and others being upregulated.33 Among these are multiple inflammation-associated genes, in particular chemokines, cytokines and acute phase proteins.33 Higher expression of these mediators due to chronic activation of the UPR might perpetuate the tissue damage during GvHD by supporting the characteristic inflammatory microenvironment of the disease. Deletion of Xbp1 led to a global dysregulation of the UPR in mice developing GvHD and was associated with a significantly decreased survival rate and increased GvHD histopathology. In a preclinical model of liver injury, mice with liver-specific deletion of Xbp1 showed initially a similar UPR activation after chemical ER stress induction with tunicamycin when compared to wild-type animals. However, this activation was significantly prolonged indicating the development of chronic ER stress in these animals.34 UPR dysregulation by Xbp1 deletion changes the composition of the intestinal epithelium. Previous studies by Kaser et al. show that mice with a Xbp1 deletion in the intestinal epithelium had a loss of Paneth and goblet cells with rising age while at the same time intestinal stem cells and transitamplifying cells appeared increased.17,35 Upon GvHD induction, a loss of Paneth cell occurs and its severity is a predictive marker of GvHD outcome.22,36,37 In our study, this loss was more pronounced in animals with a selective deletion of Xbp1 with multiple defensin subtypes, which are produced by Paneth cells, being amongst the most differentially regulated genes in Xbp1DIEC mice. One additional mechanism by which ER stress might modulate intestinal inflammation is by altering the microbiome. The microbiome has become increasingly important for the understanding of GvHD pathology. Several studies in mice and humans show that GvHD severity and mortality are correlated with reduced microbial diversion and a shift from protective towards det-

rimental bacterial species.38,39 A protective role has been shown for the phylum Firmicutes and specifically for members of the Clostridia class.24,25 The beneficial effects of these bacteria could be attributed to the generation of antiinflammatory metabolites, such as short-chain fatty acids. To the best of our knowledge, this study is the first one to describe the intestinal microbiome composition in Xbp1DIEC mice. We observed that these mice have reduced abundancy of Firmicutes compared to their wild-type littermates. In particular, the Clostridia class including the Lachnospiraceae family were decreased. These changes closely resemble the changes observed during GvHD induction. One possible explanation for this dysbiosis is, that the intestinal environment is changed due to the reduced numbers of Paneth cells, and the lower concentrations of their anti-microbial products, defensins, lysozyme and Reg3γ. In our view, these data support the hypothesis that there are pre-existing changes in the microbiome in Xbp1DIEC mice that predispose to GvHD development. The UPR provides multiple targets for signaling modulation by inhibition of specific branches. We proposed that specific UPR signaling inhibition might aid in modulating ER stress signaling from pro-apoptotic to cell-protective pathways. Here, we used 4m8c and B-I09, inhibitors of IRE1α RNase activity, and GSK2606414 as a PERK inhibitor. Administration of IRE1α inhibitors, but not the PERK inhibitor, showed a beneficial effect in GvHD-developing mice. Mechanistically, 4m8c decreased TNF-associated intestinal cell death. A crosstalk between IRE1α and TNF signaling has been documented previously.40 ER stress can induce the expression of various pro-inflammatory cytokines.41 Indeed, administration of 4m8c in a murine model of rheumatoid arthritis decreased disease severity by blocking pro-inflammatory cytokine secretion.42 At the same time, TNF is known to induce ER stress.43 It is conceivable that pro-apoptotic UPR employs at least in part the same pathways as TNF signaling so that inhibition of IRE1α can ameliorate the cytotoxic effects of both. Interestingly, 4m8c showed a higher protective effect in GvHD animals, compared to B-I09. These differences might be due to variations in bioavailability after intraperitoneal injection. Another potential consideration would be

Haematologica | 107 July 2022

1551

ARTICLE - Targeting ER stress in graft-versus-host disease their specificity. Both compounds have been developed as potent and specific inhibitors of IRE1α RNase activity, and little is known about potential off-target effects. In a previous study, 4m8c was reported to have anti-oxidant capacity in endothelial and pancreatic cells by inhibiting xanthine/xanthine oxidase-catalyzed superoxide production and angiotensin II-induced ROS production.44 In light of this report, it is conceivable that 4µ8c has combined beneficial effects in GvHD due to IRE1α inhibition, and an anti-oxidative effect, which makes it a particularly interesting compound for further testing. Besides having influence on signaling in intestinal cells, systemic IRE1α inhibition might potentially have effects on immune cell reconstitution and immune responses against pathogens and malignant cells. Our data show that 4m8c did neither cause alterations in platelet and hemoglobin recovery, nor on lymphocyte differentiation and effector memory development. Furthermore, we did not observe a change in T-cell expansion and cytokine production upon administration of 4m8c. The capacity to kill A20 lymphoma cells was also not affected. These data are in line with previous reports, showing that dendritic cells, which were activated in the presence of the IRE1α inhibitor B-I09, were capable of eliciting strong anti-tumor responses in T cells.45 In addition, a direct anti-tumor effect of B-I09 against chronic lymphocytic leukemia cells has been reported.46 Interestingly, specific inhibition of IRE1α significantly inhibited influenza A viral replication in cell lines.47 In light of these publications, our results provide encouraging evidence that 4m8c treatment might spare anti-viral and anti-tumor responses while diminishing GvHD severity, even though further in vivo studies are necessary to provide additional confirmation. Contrary to acute GvHD pathogenesis, donor B cells play an essential role in the context of chronic GvHD. These cells secrete antibodies in high quantities and are therefore highly dependent on a properly functioning UPR and are sensitive to ER stress; additionally it was found that XBP1 is essential for plasma cell differentiation.48 It is therefore conclusive that several attempts were made to target ER stress and the UPR in the setting of chronic GvHD. The administration of the IRE1 inhibitor B-I09 was found to reduce clinical features in a cutaneous model of chronic GvHD, with infiltrations of the skin by donor T cells and dendritic cells being reduced.49 In another study, the use of the chemical chaperone 4-phenylbutyric acid (4-PBA) led to the amelioration of chronic GvHD-induced fibrosis.32 Based on our data, we propose that excessive ER stress and the activation of the UPR are mechanisms which mediate tissue injury during intestinal GvHD. Our study provides the first evidence that administration of an IRE1α inhibitor is a pharmacologi-

E. Haring et al. cal intervention that reduces intestinal GvHD in mice and should be considered for testing in GvHD patients. Disclosures RZ received speakers fees from Novartis, Incyte and Mallinckrodt. Contributions EH, RZ and PA developed the overall concept and designed research; EH, FMU, BS, DP, and PA conducted the experiments and analyzed data; GA and MB performed bioinformatics analysis; MK, MP and BG performed and analyzed 16S rRNA sequencing; ASG and KA developed and analyzed histological stains; PE, SM, NK and JD provided reagents and/or conceptual input; PA and RZ provided funding and supervision; EH and PA wrote the manuscript. All authors discussed the results and contributed to the final manuscript. Acknowledgments The authors would like to acknowledge Dr. L. Glimcher for providing the XBP1 flox/flox mice as a gift; Dr. D. Kaiserlian for providing the MODE-K cell line; Dr. H. Andrlova for providing protocols for experimental procedures; and K. Gräwe for performing immunohistochemical stainings. Funding PA was supported by the Else Kröner-Fresenius-Stiftung (EKFS 2015_A147 to PA), the German Cancer Consortium (DKTK, FR 01-375) and a scholarship from the Berta Ottenstein Program for Clinician Scientists, Faculty of Medicine, Medical Center – University of Freiburg, Germany. RZ is supported by the Deutsche Forschungsgemeinschaft (DFG): SFB1479 (Project ID: 441891347), SFB1160, TRR167 and SFB850, the INTERREG V European regional development fund (European Union) program (Project 3.2 TRIDIAG), the European Union: GVHDCure Proposal n° 681012 ERC consolidator grant, the Deutsche Krebshilfe (grant number 70113473), the Jose-Carreras Leukemia foundation (grant number DJCLS 01R/2019) and the Wilhelm Sander Stiftung (grant 2008.046.4). NK was supported by the German Research Foundation (DFG) under German’s Excellence Strategy (CIBSS - EXC 2189 – Project ID 390939984). MB is supported by the Deutsche Forschungsgemeinschaft (DFG) – SFB 850 subprojects C9 and Z1, SFB1479 (Project ID: 441891347- S1), SFB1160 (Project Z02), SFB1453 (Project S1) and TRR167 (Project Z01), the German Federal Ministry of Education and Research by MIRACUM within the Medical Informatics Funding Scheme (FKZ 01ZZ1801B). Data-sharing statement Microarray data are available on GEO under accession number GSE156469 (with the token uhgrqykubhmppul).

Haematologica | 107 July 2022

1552

ARTICLE - Targeting ER stress in graft-versus-host disease

E. Haring et al.

References 1. Zeiser R, Blazar BR. Acute graft-versus-host disease - biologic process, prevention, and therapy. N Engl J Med. 2017;377(22):2167-2179. 2. Zeiser R. Advances in understanding the pathogenesis of graftversus-host disease. Br J Haematol. 2019;187(5):563-572. 3. Köhler N, Zeiser R. Intestinal microbiota influence immune tolerance post allogeneic hematopoietic cell transplantation and Intestinal GVHD. Front Immunol. 2018;9:3179. 4. Malard F, Huang XJ, Sim JPY. Treatment and unmet needs in steroid-refractory acute graft-versus-host disease. Leukemia. 2020;34(5):1229-1240. 5. Lindemans CA, Calafiore M, Mertelsmann AM, et al. Interleukin22 promotes intestinal-stem-cell-mediated epithelial regeneration. Nature. 2015;528(7583):560-564. 6. Zhang X, Liu S, Wang Y, et al. Interleukin22 regulates the homeostasis of the intestinal epithelium during inflammation. Int J Mol Med. 2019;43(4):1657-1668. 7. Takashima S, Kadowaki M, Aoyama K, et al. The Wnt agonist Rspondin1 regulates systemic graft-versus-host disease by protecting intestinal stem cells. J Exp Med. 2011;208(2):285-294. 8. Hayase E, Hashimoto D, Nakamura K, et al. R-Spondin1 expands Paneth cells and prevents dysbiosis induced by graft-versushost disease. J Exp Med. 2017;214(12):3507-3518. 9. Norona J, Apostolova P, Schmidt D, et al. Glucagon-like peptide 2 for intestinal stem cell and Paneth cell repair during graftversus-host disease in mice and humans. Blood. 2020;136(12):1442-1455. 10. Haring E, Uhl FM, Andrieux G, et al. Bile acids regulate intestinal antigen presentation and reduce graft-versus-host disease without impairing the graft-versus-leukemia effect. Haematologica. 2021;106(8):2131-2146. 11. Hetz C, Chevet E, Harding HP. Targeting the unfolded protein response in disease. Nat Rev Drug Discov. 2013;12(9):703-719. 12. Calfon M, Zeng H, Urano F, et al. IRE1 couples endoplasmic reticulum load to secretory capacity by processing the XBP-1 mRNA. Nature. 2002;415(6867):92-96. 13. Yoshida H, Matsui T, Yamamoto A, Okada T, Mori K. XBP1 mRNA is induced by ATF6 and spliced by IRE1 in response to ER stress to produce a highly active transcription factor. Cell. 2001;107(7):881-891. 14. Nishitoh H, Matsuzawa A, Tobiume K, et al. ASK1 is essential for endoplasmic reticulum stress-induced neuronal cell death triggered by expanded polyglutamine repeats. Genes Dev. 2002;16(11):1345-1355. 15. Urano F, Wang X, Bertolotti A, et al. Coupling of stress in the ER to activation of JNK protein kinases by transmembrane protein kinase IRE1. Science. 2000;287(5453):664-666. 16. Han D, Lerner AG, Vande Walle L, et al. IRE1alpha kinase activation modes control alternate endoribonuclease outputs to determine divergent cell fates. Cell. 2009;138(3):562-575. 17. Kaser A, Lee AH, Franke A, et al. XBP1 links ER stress to intestinal inflammation and confers genetic risk for human inflammatory bowel disease. Cell. 2008;134(5):743-756. 18. Holtan SG, Shabaneh A, Betts BC, et al. Stress responses, M2 macrophages, and a distinct microbial signature in fatal intestinal acute graft-versus-host disease. JCI Insight. 2019;5(17):e129762. 19. Sato T, Clevers H. Primary mouse small intestinal epithelial cell cultures. Methods Mol Biol. 2013;945:319-328. 20. Wilhelm K, Ganesan J, Muller T, et al. Graft-versus-host disease is enhanced by extracellular ATP activating P2X7R. Nat Med.

2010;16(12):1434-1438. 21. Jankovic D, Ganesan J, Bscheider M, et al. The Nlrp3 inflammasome regulates acute graft-versus-host disease. J Exp Med. 2013;210(10):1899-1910. 22. Eriguchi Y, Takashima S, Oka H, et al. Graft-versus-host disease disrupts intestinal microbial ecology by inhibiting Paneth cell production of alpha-defensins. Blood. 2012;120(1):223-231. 23. Simms-Waldrip TR, Sunkersett G, Coughlin LA, et al. Antibioticinduced depletion of anti-inflammatory Clostridia is associated with the development of graft-versus-host disease in pediatric stem cell transplantation patients. Biol Blood Marrow Transplant. 2017;23(5):820-829. 24. Jenq RR, Taur Y, Devlin SM, et al. Intestinal Blautia is associated with reduced death from graft-versus-host disease. Biol Blood Marrow Transplant. 2015;21(8):1373-1383 25. Mathewson ND, Jenq R, Mathew AV, et al. Gut microbiomederived metabolites modulate intestinal epithelial cell damage and mitigate graft-versus-host disease. Nat Immunol. 2016;17(5):505-513. 26. Wysocki CA, Burkett SB, Panoskaltsis-Mortari A, et al. Differential roles for CCR5 expression on donor T cells during graft-versus-host disease based on pretransplant conditioning. J Immunol. 2004;173(2):845-854. 27. Taylor PA, Ehrhardt MJ, Lees CJ, et al. Insights into the mechanism of FTY720 and compatibility with regulatory T cells for the inhibition of graft-versus-host disease (GVHD). Blood. 2007;110(9):3480-3488. 28. Kennedy GA, Varelias A, Vuckovic S, et al. Addition of interleukin-6 inhibition with tocilizumab to standard graftversus-host disease prophylaxis after allogeneic stem-cell transplantation: a phase 1/2 trial. Lancet Oncol. 2014;15(13):1451-1459. 29. Zeiser R, von Bubnoff N, Butler J, et al. Ruxolitinib for glucocorticoid-refractory acute graft-versus-host disease. N Engl J Med. 2020;382(19):1800-1810. 30. Choi SW, Braun T, Chang L, et al. Vorinostat plus tacrolimus and mycophenolate to prevent graft-versus-host disease after related-donor reduced-intensity conditioning allogeneic haemopoietic stem-cell transplantation: a phase 1/2 trial. Lancet Oncol. 2014;15(1):87-95. 31. Joly AL, Deepti A, Seignez A, et al. The HSP90 inhibitor, 17AAG, protects the intestinal stem cell niche and inhibits graft versus host disease development. Oncogene. 2016;35(22):2842-2851. 32. Mukai S, Ogawa Y, Urano F, Kudo-Saito C, Kawakami Y, Tsubota K. Novel treatment of chronic graft-versus-host disease in mice using the ER stress reducer 4-phenylbutyric acid. Sci Rep. 2017;7:41939. 33. Tsalikis J, Pan Q, Tattoli I, et al. The transcriptional and splicing landscape of intestinal organoids undergoing nutrient starvation or endoplasmic reticulum stress. BMC Genomics. 2016;17(1):680. 34. Olivares S, Henkel AS. Hepatic Xbp1 gene deletion promotes endoplasmic reticulum stress-induced liver injury and apoptosis. J Biol Chem. 2015;290(50):30142-30151. 35. Niederreiter L, Fritz TM, Adolph TE, et al. ER stress transcription factor Xbp1 suppresses intestinal tumorigenesis and directs intestinal stem cells. J Exp Med. 2013;210(10):2041-2056. 36. Levine JE, Huber E, Hammer ST, et al. Low Paneth cell numbers at onset of gastrointestinal graft-versus-host disease identify patients at high risk for nonrelapse mortality. Blood. 2013;122(8):1505-1509. 37. Eriguchi Y, Nakamura K, Hashimoto D, et al. Decreased secretion

Haematologica | 107 July 2022

1553

ARTICLE - Targeting ER stress in graft-versus-host disease of Paneth cell alpha-defensins in graft-versus-host disease. Transpl Infect Dis. 2015;17(5):702-706. 38. Holler E, Butzhammer P, Schmid K, et al. Metagenomic analysis of the stool microbiome in patients receiving allogeneic stem cell transplantation: loss of diversity is associated with use of systemic antibiotics and more pronounced in gastrointestinal graft-versus-host disease. Biol Blood Marrow Transplant. 2014;20(5):640-645. 39. Taur Y, Jenq RR, Perales MA, et al. The effects of intestinal tract bacterial diversity on mortality following allogeneic hematopoietic stem cell transplantation. Blood. 2014;124(7):1174-1182. 40. Hu P, Han Z, Couvillon AD, Kaufman RJ, Exton JH. Autocrine tumor necrosis factor alpha links endoplasmic reticulum stress to the membrane death receptor pathway through IRE1alphamediated NF-kappaB activation and down-regulation of TRAF2 expression. Mol Cell Biol. 2006;26(8):3071-3084. 41. Bettigole SE, Glimcher LH. Endoplasmic reticulum stress in immunity. Annu Rev Immunol. 2015;33;107-138. 42. Qiu Q, Zheng Z, Chang L, et al. Toll-like receptor-mediated IRE1alpha activation as a therapeutic target for inflammatory arthritis. EMBO J. 2013;32(18):2477-2490. 43. Denis RG, Arruda AP, Romanatto T, et al. TNF-alpha transiently

E. Haring et al. induces endoplasmic reticulum stress and an incomplete unfolded protein response in the hypothalamus. Neuroscience. 2010;170(4):1035-1044. 44. Chan SMH, Lowe MP, Bernard A, Miller AA, Herbert TP. The inositol-requiring enzyme 1 (IRE1alpha) RNAse inhibitor, 4micro8C, is also a potent cellular antioxidant. Biochem J. 2018;475(5):923-929. 45. Betts BC, Locke FL, Sagatys EM, et al. Inhibition of human dendritic cell ER stress response reduces T cell alloreactivity yet spares donor anti-tumor immunity. Front Immunol. 2018;9:2887. 46. Tang CH, Ranatunga S, Kriss CL, et al. Inhibition of ER stressassociated IRE-1/XBP-1 pathway reduces leukemic cell survival. J Clin Invest. 2014;124(6):2585-2598. 47. Hassan IH, Zhang MS, Powers LS, et al. Influenza A viral replication is blocked by inhibition of the inositol-requiring enzyme 1 (IRE1) stress pathway. J Biol Chem. 2012;287(7):4679-4689. 48. Reimold AM, Iwakoshi NN, Manis J, et al. Plasma cell differentiation requires the transcription factor XBP-1. Nature. 2001;412(6844):300-307. 49. Schutt SD, Wu Y, Tang CH, et al. Inhibition of the IRE-1alpha/XBP1 pathway prevents chronic GVHD and preserves the GVL effect in mice. Blood Adv. 2018;2(4):414-427.

Haematologica | 107 July 2022

1554

ARTICLE - Cell Therapy and Immunotherapy

Cardiovascular events in patients treated with chimeric antigen receptor T-cell therapy for aggressive B-cell lymphoma Raphael E. Steiner,1* Jose Banchs,2* Efstratios Koutroumpakis,2** Melody Becnel,1** Cristina Gutierrez,3 Paolo Strati,1 Chelsea C. Pinnix,4 Lei Feng,5 Gabriela Rondon,6 Catherine Claussen,1 Nicolas Palaskas,2 Kaveh Karimzad,2 Sairah Ahmed,1 Sattva S. Neelapu,1 Elizabeth Shpall,6 Michael Wang,1 Francisco Vega,7 Jason Westin,1# Loretta J. Nastoupil1# and Anita Deswal2# Lymphoma and Myeloma, MD Anderson Cancer Center; Cardiology, MD Anderson Cancer Center; 3Critical Care& Respiratory Care, MD Anderson Cancer Center; 4Radiation Oncology, MD Anderson Cancer Center; 5Biostatistics, MD Anderson Cancer Center; 6Stem Cell Transplantation, MD Anderson Cancer Center and 7Hematophathology, MD Anderson Cancer Center, Houston, TX, USA 1

Correspondence: Raphael Steiner RESteiner1@mdanderson.org. Received: September 12, 2021. Accepted: November 3, 2021. Prepublished: November 11, 2021. https://doi.org/10.3324/haematol.2021.280009 ©2022 Ferrata Storti Foundation Haematologica material is published under a CC BY-NC license

RES and JBcontributed equally as co-first authors. EK and MB contributed equally as co-second authors. # JW, LJN and AD contributed equally as co-senior authors. *

Abstract Standard of care (SOC) chimeric antigen receptor (CAR) T-cell therapies such as axicabtagene ciloleucel (axi-cel) and tisagenlecleucel (tisa-cel) are associated with multisystem toxicities. There is limited information available about cardiovascular (CV) events associated with SOC axi-cel or tisa-cel. Patients with CV comorbidities, organ dysfunction, or lower performance status were often excluded in the clinical trials leading to their Food and Drug Adminsitration approval. An improved understanding of CV toxicities in the real-world setting will better inform therapy selection and management of patients receiving these cellular therapies. Here, we retrospectively reviewed the characteristics and outcomes of adult patients with relapsed/refractory large B-cell lymphoma treated with SOC axi-cel or tisa-cel. Among the 165 patients evaluated, 27 (16%) developed at least one 30-day (30-d) major adverse CV event (MACE). Cumulatively, these patients experienced 21 arrhythmias, four exacerbations of heart failure/cardiomyopathy, four cerebrovascular accidents, three myocardial infarctions, and one patient died due to myocardial infaction. Factors significantly associated with an increased risk of 30-d MACE included age ≥60 years, an earlier start of cytokine release syndrome (CRS), CRS ≥ grade 3, long duration of CRS, and use of tocilizumab. After a median follow-up time of 16.2 months (range, 14.3-19.1), the occurrence of 30-d MACE was not significantly associated with progression-free survival or with overall survival. Our results suggest that the occurrence of 30-d MACE is more frequent among patients who are elderly, with early, severe, and prolonged CRS. However, with limited follow-up, larger prospective studies are needed, and multidisciplinary management of these patients is recommended.

Introduction Patients with relapsed/refractory aggressive large B-cell lymphoma (LBCL) can achieve durable remissions with anti-CD19 chimeric antigen receptor T (CAR-T) cell therapies.1-5 Multicenter clinical trials led to US Food and Drug Administration (FDA) approval of the CAR-T cell products axicabtagene ciloleucel (axi-cel) in October 2017, tisagenlecleucel (tisa-cel) in May 2018, and Lisocabtagene maraleucel in February 2021 for relapsed or refractory LBCL after two or more lines of systemic therapy.6-8 In the trials

leading to their FDA approval, patients with cardiovascular (CV) comorbidities, organ dysfunction, or lower performance status were often excluded. These CAR-T cell therapies are associated with significant toxicities. Notably, cytokine release syndrome (CRS) and immune cell-associated neurotoxicity syndrome (ICANS) are commonly observed.9 In addition, CAR-T cell therapy has been associated with cardiotoxicity, which may result in prolonged length of hospital stay, admission to the intensive care unit for vasopressor support, or cardiac death.10 Early diagnosis and management of CV complica-

Haematologica | 107 July 2022

1555

ARTICLE - Cardiovascular events post T-cell therapy

R.E. Steiner et al.

tions may occur in patients with CRS and require awareness and multidisciplinary collaboration.11,12 The exact mechanisms of CAR-T cell-associated cardiac dysfunction are unclear. Potential hypotheses include interleukin-6–mediated myocardial stunning13 during CRS.14 With the increased utilization of CAR-T cell therapy outside of clinical trials, more data characterizing the CV toxicities of standard of care (SOC) CAR-T cell therapy among LBCL adults are needed to assist clinical decisions. An adequate cardiac reserve is required to withstand the clinical manifestations of CRS. Second, many adults, including elderly patients treated with CAR-T cell therapy, especially outside of clinical trials, have pre-existing CV disease in addition to prior cardiotoxic chemo- or radiation therapy. These different factors can predispose patients to a higher risk for CV events after CAR-T therapy. Hence, we report the results from a contemporary cohort of patients with relapsed or refractory LBCL treated at our institution with SOC axi-cel and tisa-cel and describe the clinical, CV characteristics and outcomes and their relation to CRS.

Methods Patient selection The study cohort consists of 165 consecutive adult patients (≥18 years old) with relapsed or refractory LBCL after ≥2 lines of systemic therapy, treated with SOC axicel and tisa-cel at the University of Texas MD Anderson Cancer Center (MDACC), between January 2018 and April 2020. The study excluded patients with unapproved indications such as Richter syndrome and post-transplant lymphoproliferative disorders because of their different biology and outcomes. SOC was defined as the administration of commercial products outside of any clinical trial.9 Follow-up data were collected through February 19, 2021. The study was approved by the Institutional Review Board (IRB) of MDACC and conducted in accordance with our institutional guidelines and the principles of the Declaration of Helsinki.9 The IRB approved the request of waiver of informed consent and a waiver of authorizations as the study does not involve diagnostic or therapeutic intervention or any type of direct patient contact. Covariates, definitions, and response assessment Data on the covariates were extracted retrospectively from the electronic medical records.15 A cardiologist reviewed all electrocardiogram (EKG) (EK) and echocardiogram studies (JB). Cardiac testing was performed at the discretion of the treating team at any time and was not pre-specified. Myocardial infarction (MI) was defined as angina or anginal

equivalent symptoms with cardiac enzyme elevation, with or without EKG/echocardiographic changes. New or worsening cardiomyopathy (CMP) was defined as a reduction in left ventricular ejection fraction (LVEF) >10% from baseline to <50% during the hospitalization for CAR-T cell therapy.14 Major adverse cardiovascular events (MACE) were defined as a composite of arrhythmias requiring an intervention, new or worsening CMP, exacerbation of heart failure (HF), cerebrovascular accident (CVA), MI, or CV death. CV death was defined as death due to HF, cardiogenic shock, arrhythmia, MI, or cardiac arrest.16 CRS and ICANS were prospectively graded for up to 30 days (30-d) after CAR-T cell infusion according to the CAR toxicity (CARTOX) grading system from January 2018 to April 2019 and ASTCT criteria from May 2019 onward.17,18 CRS and ICANS were managed according to our institutional CARTOX guidelines. Notably, patients requiring tocilizumab received 8 mg/kg dosing for a maximum of four doses. All MACE were treated by cardiologists per disease guidelines and clinical expertise. Of note, all patients receiving CAR-T cell therapy are monitored by telemetry at our institution. Lymphoma response status was determined by the Lugano 2014 classification.19 Statistical methods Fisher’s exact test or Chi-square test was used to evaluate associations between categorical variables. The Wilcoxon rank sum test was used to evaluate differences in continuous variables between patient groups. Multivariable logistic regression models were fitted to estimate the effects of important covariates on 30-day (30-d) MACE. The variables with a P-value <0.2 from the univariable analysis were included in the full multivariable model, and a backward selection method was used for the final model. Progression-free survival (PFS) time was calculated from the start of CAR-T cell infusion to the progression of disease or death, whichever occurred first. Patients without event were censored at the last follow-up. Overall survival (OS) was calculated from the start of CAR-T cell infusion to death or last follow-up. A multidisciplinary committee determined the cause of death, including a cardiologist, intensivist, and medical oncologist. The Kaplan-Meier method was used to estimate PFS and OS. The Cox proportional hazards models were used for multivariable analysis. The Schoenfeld residual was used to check the proportional hazards assumption. The variables with a P-value <0.2 from the univariable analysis were included in the initial full model. A backward selection method was used, and a significant level of 0.1 was set as the criterion for a covariate to stay in the model. Landmark analysis for PFS/OS by 30d MACE post-infusion was performed using the landmark time of 30 d. The patients with follow-up time less than 30d were excluded from this analysis. 30-d MACE was forced

Haematologica | 107 July 2022

1556

ARTICLE - Cardiovascular events post T-cell therapy

R.E. Steiner et al.

in the final model to assess its effect on PFS/OS with the adjustment of other covariates in the model.

Results Baseline patient characteristics The baseline characteristics of the study cohort of 165 patients with relapsed or refractory LBCL are shown in Table 1. Patients had a median age of 60 years (range, 18-88 years), and the majority were male (72%) and white (74%). They had received a median number of three previous therapy lines, a median cumulative dose of anthracycline of 300 mg/m2, and 12% had a history of radiotherapy of the chest. Overall, 9% of patients had a history of coronary artery disease (CAD), and 8% had a history of HF. Most patients received axi-cel (94%; median age 60 years; range, 18-85 years), and ten patients received tisa-cel (6%; median age 66 years; range, 28-88 years). Description of 30-day major adverse cardiovascular events Overall, 27 patients (16%) developed at least one 30-d MACE, with a total of 33 30-d MACE (Figure 1A). Of all MACE, 64% were arrhythmias that triggered a medical intervention even if asymptomatic, 12% were CVA, 12% were new or exacerbated prior HF or worsening of CMP, 6% were non-fatal MI, and one was a CV death due to possible MI. Figure 1B shows the distribution of the first and subsequent MACE by patients. The timing of the 30-day MACE is depicted in Figure 1C. The first MACE occurred at a median of 7 days (range, 0-29 days) after CAR-T cell infusion (day 0), and 91% of all events occurred within the first 16 days after the infusion. Six patients had one recurrence of 30-d MACE, with five of these six events occurring in patients who had an arrhythmia as the initial event (Figure 1B). The baseline characteristics between patients who did and did not develop a MACE are shown in Table 1. Patients that developed MACE were older (median age 69 years)) compared with patients that did not develop MACE (median age 59 years). However, younger patients also presented with severe non-arrhythmic CV events. Nonarrhythmic CV events include MI, cerebrovascular accident, heart failure, cardiomyopathy, and CV-related death. Although the prevalence of CV risk factors and pre-existing cardiac disease was numerically higher in patients with MACE compared to those without MACE, these differences did not reach statistical significance (trend towards significance for coronary artery disease). Among the patients who developed arrhythmias triggering an intervention, the median age was 68 years (range, 4282 years). Almost half these arrhythmias (48%) were short runs of non-sustained ventricular tachycardia (NSVT), with

a median number of eight beats (range, 3-11), and only one event was associated with symptoms of palpitations. Episodes of NSVT were detected, given patients were monitored on telemetry as standard practice. In the ten (48%) patients with at least one episode of atrial fibrillation (AF), the median age was 70 years (range, 54-83 years), and among them, only two had a prior history of AF before CAR-T cell therapy. Of the ten patients treated with tisacel, two patients aged 71 and 81 years old developed a 30d MACE, an episode of AF, and NSVT, respectively. All the patients who presented with at least one episode of arrhythmia had a rapid resolution of the event, and 55% of them were discharged with a new cardiac medication such as metoprolol and/or amiodarone. Of the three patients with MI, the median age was 77 years (range, 60-83 range), two had a prior history of CAD, and all had a least one CV risk factor. One patient required cardiac catheterization without revascularization; the two other patients could not get aspirin because of grade 4 thrombocytopenia and did not undergo invasive cardiac evaluation. The CV death also occurred in an elderly patient, 76 years old, with a history of diabetes mellitus and hypertension, in the setting of transient CRS maximal grade 1, requiring tocilizumab. He developed a cardiac arrest due to suspected inferior MI on day 15 post-CAR-T cell infusion. Three patients developed clinical HF, and one patient developed a worsening CMP with further reduction in LVEF without symptomatic HF (median age of patients with HF or cardiomyopathy was 62 years). Of the three patients with symptomatic HF post-CAR-T cell therapy, at baseline, one patient had a history of mild asymptomatic ischemic CMP, one patient had a history of HF with preserved LVEF, and one patient had a history of AF with preserved LVEF. All patients with symptomatic HF developed a significant reduction in LVEF. These patients had a median baseline LVEF of 50% (range, 47-63%). None of the HF exacerbations or CMP were related to CAD or MI. Among the four patients with clinical HF/worsening CMP, three presented with progression of LBCL at day 30, and one died of multiple etiologies at day 50. Among the four patients who presented a CVA (median age 61 years), three had at least one CV risk factor, and two of them had AF. A young 24-year-old patient without risk factors was postulated to have CVA-like brain injury possibly related to ICANS. Of note, among all the patients who presented a 30-d MACE, only one was readmitted in the following 6 months for a CV reason: atrial fibrillation with rapid ventricular rate. Echocardiographic characteristics Overall, 129 patients (78%) had a baseline echocardiogram before CAR-T cell therapy infusion, with a median baseline LVEF of 54% (range, 38-75%). Although numerically lower

Haematologica | 107 July 2022

1557

ARTICLE - Cardiovascular events post T-cell therapy

R.E. Steiner et al.

in patients with 30-d MACE (median LVEF 53%), the difference in LVEF compared to patients without MACE (58%) was not statistically significant, P=0.131 (Table 1). Moreover, 67% of non-arrhythmic MACE occurred in pa-

tients with a baseline LVEF of at least 50%, as shown in Figure 2. Baseline diastolic function could be reported in 123 patients: normal in 51%, grade 1 (mild/impaired relaxation) in 57 (46%), and grade 2 (moderate/pseudonormal)

Figure 1. Nature, recurrences, and timing of 30-day major adverse cardiovascular events (A) Cumulative occurrences of 30-day major adverse cardiovascular event (30-d MACE). (B) Nature and recurrences of 30-d MACE. *Events happened on the same day, counted as 1 atrial fibrillation (AF) event. The patients who presented clinical heart failure had a decrease in left ventricular ejection fraction. (C). Timing of 30-d MACE. Day 0 represents the day of chimeric antigen receptor T-cell therapy infusion. AF: atrial fibrillation; CMP: cardiomyopathy; HF: heart failure; CV: cardiovascular; CVA: cerebrovascular accident; MI: myocardial infarction; NSVT: non-sustained ventricular tachycardia; SVT: supraventricular tachycardia. Haematologica | 107 July 2022

1558

ARTICLE - Cardiovascular events post T-cell therapy

R.E. Steiner et al.

Table 1. Baseline patient characteristics and association with 30-day major adverse cardiovascular events Patients who presented at least one 30-d MACE (N = 27) Characteristics/ Outcomes

Patients who Patients who presented All patients did not present at least one 30(N= 165) 30-d MACE d MACE (N = 138) (N = 27)

Arrhythmic event(s) only (N = 15)

At least one nonarrhythmic event (N = 12)

0.001

68 [42-82]

70 [24-83]

13 (48%)

8 (30%)

2 (7%)

4 (15%)

11 (41%)

10 (37%)

1 (4%) 0 (0%) 13 (48%) 1 (4%)

1 (4%) 0 (0%) 10 (37%) 1 (4%)

2 (7%)

3 (11%)

1 (4%) 5 (19%) 3 (11%) 6 (22%)

6 (22%) 1 (4%) 8 (30%) 5 (19%)

Cohort Age, median [range], y

60 [18-88]

59 [18-88]

69 [24-83]

Age >60 years

87 (53%)

66 (48%)

21 (78%)

Age <60 years

78 (47%)

72 (52%)

6 (22%)

Male

118 (72%)

97 (70%)

21 (78%)

Race Asian African American White Hispanic

8 (5%) 13 (8%) 123 (74%) 21 (13%)

6 (4%) 13 (9%) 100 (73%) 19 (14%)

2 (7%) 0 (0%) 23 (86%) 2 (7%)

ECOG performance status >1

129 (78%)

21 (15%)

5 (19%)

HCT-CI NA 0 1-2 ≥3

26 (16%) 37 (22%) 55 (33%) 47 (28%)

25 (18%) 31 (22%) 44 (32% 38 (28%)

7 (26%) 6 (22%) 11 (41%) 9 (33%)

3 [2-11]

3 [2-6]

0.915

2 [2-6]

3 [2-5]

42 (26%)

37 (27%)

5 (19%)

0.472

3 (11%)

2 (7%)

Cum dose anthracycline, mg/m2, median [range], N

300 [0-632]

300 [50-370]

0.660

300 [50-300]

225 [80-370]

History of radiotherapy of chest

20 (12%)

17 (12%)

3 (11%)

1.000

1 (4%)

2 (7%)

Hypercholesterolemia

39 (24%)

34 (25%)

5 (19%)

0.623

2 (7%)

3 (11%)

Smoking (history or active)

58 (35%)

46 (34%)

12 (44%)

0.280

9 (33%)

3 (11%)

Diabetes Mellitus

37 (23%)

29 (21%)

8 (30%)

0.336

3 (11%)

5 (19%)

Hypertension

70 (43%)

55 (40%)

15 (56%)

0.139

7 (26%)

8 (30%)

History of coronary artery disease

14 (9%)

9 (7%)

5 (19%)

0.057

2 (7%)

3 (11%)

Heart failure

13 (8%)

9 (7%)

4 (15%)

0.231

2 (7%)

History of coronary artery disease or heart failure

22 (13%)

16 (12%)

6 (22%)

0.141

3 (33%)

3 (11%)

0.004 0.431

0.212

0.772

0.897

Previous treatment Previous therapies, median [range], N Previous ASCT

CV risk factors and baseline CVD

Baseline echocardiographic features# Left ventricular ejection fraction, median [range]

58% [38-75]

53% [39-68]

0.131

58% [39-66]

50% [44-68]

Presence of diastolic dysfunction

49%

43%

82%

0.004

89%

75%

Axi-cel

155 (94%)

130 (94%)

25 (93%)

13 (48%)

12 (44%)

Tisa-cel

10 (6%)

8 (6%)

2 (7%)

0 (0%)

CAR-T cell therapy 0.669

Values are n (%) except as noted. % values reflect the proportion of patients who had no 30-d MACE, at least one 30-d MACE, and all the patients. P-value reflects the comparison of patients who did not present a 30-d MACE with patients who had at least one MACE. Bold numbers indicate a significant P-value. CAR-T cell: chimeric antigen receptor T cell; ASC: autologous stem cell transplant; CAR: chimeric antigen receptor; cum: cumulative; CV: cardiovascular; CVD: cardiovascular disease; d: day; ECOG: Eastern Cooperative Oncology Group; HCT-CI: Hematopoietic Cell Transplantation–Specific Comorbidity Index15; MACE: major adverse cardiovascular event; n: number of patients; NA: not available; Axi-cel: axicabtagene ciloleucel; Tisa-cel: tisagenlecleucel; y: years. Non-arrhythmic events include myocardial infarction, cerebrovascular accident, heart failure, cardiomyopathy, and cardiovascular-related death. #not done in all patients. Haematologica | 107 July 2022

1559

ARTICLE - Cardiovascular events post T-cell therapy

R.E. Steiner et al.

in three patients (3%). Patients who presented with any baseline diastolic dysfunction were more at risk for 30-d MACE (82% in the 30-d MACE group vs 43% in the nonMACE group, P=0.004). Among the 27 patients who had at least one repeat echocardiogram during the 30-d following CAR-T cell therapy, only six patients had a decrease of LVEF of at least 10% (range, 10-21%); however, only four of them experienced a decrease of LVEF under 50%. Among the latter, one patient had documented rapid resolution of the LVEF decrease but died of lymphoma progression on day 53 (Figure 3). Another patient had no documented recovery of LVEF and died of lymphoma progression on day 49. The two latter patients did not have a documented repeat echocardiogram to date but had clinical recovery of HF in less than a month. Inflammation and 30-day major adverse cardiovascular events Overall, 151 (92%) patients had CRS, and 100 (61%) had ICANS of any grade. As shown in Table 2, patients with 30d MACE were more likely to have an earlier manifestation of CRS and longer and more severe episodes of CRS requiring tocilizumab (especially patients with non-arrhythmic events) than patients who did not have 30-d MACE (P=0.016). There was no significant association between 30-d MACE and ICANS, the requirement for ste-

roids, stay in the ICU, use of vasopressive and inotropic support, and intubation/mechanical ventilation. Moreover, among the three patients who presented a 30-d MACE requiring intubation, only one patient was intubated for CV reasons (cardiac arrest). Compared with patients without 30-d MACE (n=138), 30-d MACE was not associated with higher median peak ferritin, CRP, or time to the first peak of CRP or ferritin. Independent predictors of major adverse cardiovascular events On multivariable analyses, the only two independent predictors of 30-MACE were age (odds ratio 3.98, 95% confidence interval [CI]: 1.47-10.78 for MACE in patients > 60 years vs. those <60 years) and development of grade 3 or higher CRS (odds ratio 4.66, 95% CI: 1.68-12.95; Online Supplementary Tables S1 and S2, Online Supplementary Appendix). Outcome of patients with major adverse cardiovascular events After a median follow-up time of 16.2 months (range, 14.319.1 months) for censored observations, the occurrence of 30-d MACE was not significantly associated with hospital survival, PFS (12-month PFS 38% if 30-d MACE vs. 42% without 30-d MACE, P=0.552, log-rank test, Figure 4), or with OS (12-month OS 58% if 30-d MACE vs. 62% without

Figure 2. Baseline left ventricular ejection fraction and occurrence of 30-day major adverse cardiovascular events. *The numbers between parentheses represent the proportion of patients who had at least a 30-day major adverse cardiovascular event (30d MACE) within the represented range of baseline left ventricular ejection fraction (LVEF). CMP: cardiomyopathy; CV: cardiovascular; CVA: cerebrovascular accident; d: day; HF: heart failure; MI: myocardial infarction. Haematologica | 107 July 2022

1560

ARTICLE - Cardiovascular events post T-cell therapy

R.E. Steiner et al.

Figure 3. Evolution of left ventricular ejection fraction of patients who presented a drop of ejection fraction of at least 10% during day 0-30. LVEF: left ventricular ejection fraction. Day 0 represents the day of chimeric antigen receptor T-cell therapy infusion. The colored dashed lines indicate the day of death for the patient of the corresponding color.

30-d MACE P=0.519, log-rank test) in a landmark analysis using the landmark time of 30 days. The multivariable analyses also demonstrated no significant association of: 30-d MACE with OS or PFS. Moreover, the association between arrhythmias, non-arrhythmic MACE, and the outcome was not significant for either PFS or OS. Of note, among the 15 patients with 30-d MACE who died, eight died of lymphoma progression, two of sepsis, two of multifactorial etiologies, two of unknown causes, and only one died of a CV cause (MI death).

Discussion This large, single-institution cohort study describes the incidence and baseline clinical and echocardiographic correlates of MACE in a population of contemporary patients with relapsed or refractory LBCL treated with SOC CAR-T cell therapy axi-cel or tisa-cel. Our key findings in real-world patients treated outside of clinical trials include i) 16% patients developed at least one 30-d MACE; ii) almost two-thirds of the 30-d MACE were cardiac arrhythmias which were managed easily with standard therapy; iii) of all patients, only 7% developed non-arrhythmic CV events, with one CV death; iv) development of 30-d MACE did not impact intermediate or longer-term global outcomes of OS or PFS. Notably, amongst the 9% of patients that developed arrhythmias triggering treatment changes, almost half were short runs of asymptomatic NSVT, which were only noted because of the routine use of telemetry. Whether any treatment, usually in the form of beta-blockers, is even needed in these patients is unclear at this time. As new CAR products gain FDA approval with more manageable

toxicity profiles and institutions gain experience managing these toxicities, increased outpatient CAR administration and monitoring can be expected.20 Given the strong association of age with MACE, especially with arrhythmias, further multicenter studies are needed to evaluate the clinical usefulness of routine telemetry, particularly in patients younger than 60 years without cardiac comorbidities. Also, given the strong association of MACE with higher grades of CRS, closer monitoring could be triggered by the development of higher grades of CRS. In a prior retrospective study by Alvi et al. of 137 lymphoma and myeloma patients treated with CAR-T cell therapy (almost half as part of clinical trials), 17 CV events occurred (6 CV deaths, 6 decompensated HF, and 5 arrhythmias).16 They did not include asymptomatic drops in LVEF, asymptomatic NSVT, or CVA in the count of CV events. If we evaluate comparable events in our cohort, it is reassuring to note that only 18 such CV events were observed in a larger SOC cohort (165 patients) with only one CV death. This is further notable given that patients in our cohort had more CV risk factors and CV comorbidities at baseline compared to those of Alvi et al. and were of comparable age. Although not definitive, given the overall lower number of events, the lower number of CV deaths in our cohort could be associated with increased recognition and improved management of CRS, including increased use of tocilizumab (65% in our cohort vs. 41% in the prior cohort).16 In our study, age of 60 years or older and ≥ grade 3 CRS were independent predictors of 30-d MACE. Although baseline CV risk factors and disease were numerically higher in patients with CV events described in the results, these may not have been statistically significant given an overall lower number of events, especially non-arrhythmic

Haematologica | 107 July 2022

1561

ARTICLE - Cardiovascular events post T-cell therapy

R.E. Steiner et al.

Table 2. Inflammation and 30-day major adverse cardiovascular events Patients who presented at least one 30-d MACE (n = 27)

All patients (n= 165)

Patients who did not present 30-d MACE (n = 138)

Patients who presented at least one 30-d MACE (n = 27)

3 [0-15]

2 [0-9]

0.025

2 [0-9]

2 [0-7]

23 (14%)

14 (10%)

9 (33%)

0.002

3 (11%)

6 (22%)

6 [0-48]

5 [0-48]

11 [0-48]

0.036

7 [1-20]

17 [0-48]

107 (65%) 6 [0-40] 1 [1-4]

84 (61%) 6 [0-40] 1 [1-4]

23 (85%) 5 [3-10] 1 [1-4]

0.016 0.403 0.789

12 (44%) 5 [3-10] 1 [1-3]

11 (41%) 5 [3-10] 1 [1-4]

100 (61%)

80 (58%)

20 (74%)

0.105

10 (37%)

6 [0-18]

5 [1-14]

0.443

5 [3-10]

5 [1-14]

51 (31%)

41 (30%)

10 (37%)

0.451

3 (11%)

7 (26%)

Duration between first and last day of ICANS any grade, d [range]

5 [0-82]

4.5 [0-82]

7 [0-49]

0.220

5 [0-24]

7 [3-49]

Corticosteroid use* Time to first dose, median [range], d Median cumulative dose, mg

91 (55%) 7 [1-16] 170 [8-13,043]

72 (52%) 6 [1-16] 180 [8-13,043]

19 (70%) 7 [4-15] 140 [10-11,750]

0.082 0.914 0.653

8 (30%) 7 [4-8] 106 [10-494]

11 (40%) 7 [5-15] 200 [20-11,750]

ICU stay

60 (36%)

47 (34%)

13 (48%)

0.164

4 (15%)

9 (33%)

Required vasopressive and/or inotropic support

13 (8%)

12 (9%)

1 (4%)

0.696

1 (4%)

Mechanical ventilation

12 (7%)

9 (7%)

3 (11%)

0.4485

3 (11%)

17 [7-99]

16 [7-99]

23 [10-54]

0.073

16 [10-54]

30 [16-54]

117 [5-490]

109 [5-490]

121 [6-296]

0.098

109 [6-257]

164 [97-296]

5 [1-371]

5 [3-10]

0.769

5 [3-10]

First peak of Ferritin, median [range], mg/L,

1902 [83-100,001]

2005 [102-100,001]

1752 [83-34,513]

0.939

1567 [83-12,287]

2053 [266-34,513]

Time to first peak of Ferritin, median [range], d

9 [1-40]

9 [2-19]

0.697

9 [2-19]

11 [6-18]

Characteristics/Outcomes

Arrhythmic At least one nonevent(s) only arrhythmic event (n = 15) (n = 12)

CRS Time to CRS onset, d ≥grade 3 CRS Duration between first and last day of CRS any grade, median [range], d Tocilizumab use Time to first dose, median [range], d Cumulative number of doses median [range] ICANS ICANS any grade time to ICANS onset, median [range], d ≥grade 3 ICANS

Intensive care management

Inpatient stay Length of inpatient stay, median [range], d Inflammatory markers First peak of CRP, median [range], mg/L Time to first peak of CRP, median [range], d

Values are n (%) except as noted. % values reflect the proportion of patients who did not experience a 30-day major adverse cardiovascular events (30-d MACE), at least one 30-d MACE, and all the patients. P-value reflects the comparison of patients who did not experience a 30d MACE with patients who experienced at least one MACE. Bold numbers indicate a significant P-value. CRS: cytokine release syndrome; Cum.: cumulative; d: day; ICANS: immune cell-associated neurotoxicity syndrome; ICU: intensive care unit; n: number; CRP: c-reactive protein. *Dexamethasone or dexamethasone equivalent. Non-arrhythmic events include myocardial infarction, cerebrovascular accident, heart failure, cardiomyopathy, and cardiovascular-related death.

Haematologica | 107 July 2022

1562

ARTICLE - Cardiovascular events post T-cell therapy

R.E. Steiner et al.

events. In another retrospective study by Ganatra et al. with 187 patients with relapsed refractory follicular lymphoma and LBCL, the median age was 63 years. Patients who developed cardiomyopathy were older and had a greater prevalence of hyperlipidemia and CAD.14 In the retrospective study of Lefebvre et al. with 145 adult patients undergoing anti-CD19 CAR-T cell therapy with a median age of 60 years, age was not associated with MACE (P=0.431). Overall, 31 patients had MACE (41 events) at a median time of 11 days (interquartile range, 6-151 days) after CAR-T cell infusion (median follow-up period was 456 days [Interquartile range, 128-1,214 days]). However, the population was not limited to LBCL treated SOC CAR-T cell therapy and included ALL and CLL patients.21 Advanced age is a risk factor for increased comorbidities and complications. However, in a subgroup analysis of the trial ZUMA-12,4 of Neelapu et al., axi-cel induced a high rate of durable responses with a manageable safety profile, regardless of age, suggesting that age alone should not limit

CAR-T cell therapy, although it may prompt pre-therapy optimization and closer monitoring after infusion.22 The occurrence of 30-d MACE was associated with the earlier start of CRS in our study and with a more extended duration of CRS, and expectedly with higher use of tocilizumab. Our population consisted exclusively of LBCL patients treated with SOC axi-cel and tisa-cel, and comparisons with other studies may be challenging given the use of different noncommercial products and heterogeneous populations. In the study of Nastoupil et al., the overall incidence of CRS was comparable to the trial ZUMA-1. Still, grade ≥3 CRS was slightly lower at 7% versus 11% in ZUMA-1.1 This difference might be accounted for by greater use of tocilizumab and corticosteroids23 (62% and 55%, respectively) in the study of Nastoupil et al. compared with the trial ZUMA-1 (43% and 27%, respectively) in line with evolving practice patterns for toxicity management. CRS may result in depressed myocardial function, even if transient, explaining the association with MACE.

Figure 4. Association between 30-day major adverse cardiovascular and arrhythmic events and survival. (A) Progression-free survival (PFS) of patients with and without 30-day arrhythmia requiring intervention. (B) PFS of patients with and without 30day major adverse cardiovascular event (30-d MACE). (C) Overall survival (OS) of patients with and without 30-d arrhythmia requiring intervention. (D) OS of patients with and without 30-d MACE. Haematologica | 107 July 2022

1563

ARTICLE - Cardiovascular events post T-cell therapy

R.E. Steiner et al.

Further, it is possible that patients with CRS, especially with hypotension and tachycardia, which are cardiac stressors and in which patients receive large quantities of intravenous fluids, can worsen the volume overload state and precipitate myocardial ischemia.21 Our data did not show a statistically significant association between baseline LVEF and 30-d MACE. The median baseline LVEF of our study population was 54% (range, 38.1-75%), compared to 60% (range, 35-75%) in the study of Ganatra et al.14 In the latter study, baseline LVEF was not associated either with new or worsening cardiomyopathy (P=0.23). In addition, a lower baseline LVEF was not associated with poorer outcomes in our study. Nevertheless, the number of patients with low baseline LVEF was small, and further studies are needed to evaluate such a population's safety to undergo CAR-T cell therapy. We noted that patients who presented with any baseline diastolic dysfunction were at increased risk for 30-d MACE (P=0.004). However, diastolic dysfunction is prevalent with increasing age.24 In a subset analysis in patients for whom baseline echocardiogram data was available, baseline diastolic dysfunction did not prove to be an independent risk factor for 30-d MACE (P=0.109). Therefore, diastolic dysfunction may be a marker of older age and comorbidities, such as hypertension. In contrast, in the study by Lefebvre et al., there was no association between diastolic dysfunction and MACE (P=0.866).21 Of note, the evaluation of left ventricular diastolic function with echocardiographic methods may not always be standardized, and there have been two significantly different guideline statements in terms of criteria published from the same cardiology society since 2009.25,26 In our study, diastolic function on echocardiography was examined by uniformly standardized criteria for this analysis. Future studies examining larger cohorts may be needed to further study the association of diastolic dysfunction with MACE. Among the 27 patients of our study who had at least one repeat echocardiogram during the 30 days following CART cell therapy, only four patients had a decrease in LVEF of at least 10% to LVEF under 50%. None of these patients presented grade ≥3 CRS, and the drop in EF was diagnosed after CRS resolution. However, echocardiograms were only done as clinically indicated, and the actual proportion of patients with a decrease of LVEF is unknown. In the study of Ganatra et al., around 10% of patients develop cardiomyopathy in the context of high-grade CRS following CAR-T cell therapy.14 Among them, 50% had persistent cardiac dysfunction (median follow-up of 168.5 days). These results emphasize the need for baseline workup with echocardiogram before CAR-T cell infusion for every patient and repeat echocardiogram in case of clinical suspicion of cardiac dysfunction. Moreover, additional studies are needed to evaluate the long-term evolution of these post-CAR-T cell cardiomyopathies and the role of

earlier or more frequent use of tocilizumab to avoid some of these events. Finally, 30-d MACE was not significantly associated with hospital survival, PFS, or OS in our study. This may be related to relatively fewer non-arrhythmic MACE and the relatively higher competing risk of oncologic disease progression and related death. We acknowledge the limitations of our single-center study, including the retrospective nature of the analysis, the absence of long-term follow-up after MACE, and the lower frequency of follow-up cardiac imaging available in the follow-up period. We recognize the need for future prospective studies to confirm the findings of our study SOC cohorts. Another limitation to our study is that pretherapy and post-therapy cardiac serologic biomarkers were measured only as clinically indicated and did not yield specific predictive information regarding 30-d MACE.27 The strengths of our study are a large cohort, prospectively assigned toxicity grading, and comprehensive multidisciplinary evaluation, including EKG, echocardiogram, serologic markers, and telemetry. In conclusion, 30-d cardiac events were noted in 16% of patients after axi-cel and tisa-cel infusion in patients with LBCL, with CV death being rare. Age was the only independent predictor of MACE prior to infusion, although the development of grade 3 or higher CRS was associated with MACE after infusion. Excluding patients solely because of advanced age or cardiac comorbidities may be inappropriate since the majority of CV events appear to be non-life-threatening arrhythmias, treatable and/or reversible, and the prognosis of elderly patients is not inferior in regards to oncologic disease progression.28 Closer monitoring may be appropriate in elderly patients and those with underlying CV disease. More aggressive treatment of CRS with tocilizumab at our site may be a reason for lower comparable CV events compared to prior studies. Further studies are needed to predict response better, mitigate CRS, and increase response.29 In the future, machine learning may assist in CV event prediction in an initially asymptomatic population.30 This may be especially relevant to select patients able to receive CAR-T cell therapy in the outpatient setting. Disclosure RES has received research funding from Rafael Pharmaceuticals, BMS, and Seattle Genetics. CG is a consultant for Legend Biotech and received research support from Revimmune. PS is a consultant for Roche-Genentech and received research support from AstraZeneca-Acerta. CCP has received research funding from Merck Inc. SA has received research funding from Seattle Genetics, Tessa Therapeutics, Rapt Therapeutics; has a consulting/ advisory role for Seattle Genetics, Merck, Sanofi. SSN has re-

Haematologica | 107 July 2022

1564

ARTICLE - Cardiovascular events post T-cell therapy

R.E. Steiner et al.

ceived personal fees from Kite, a Gilead Company, Merck, Bristol Myers Squibb, Novartis, Celgene, Pfizer, Allogene Therapeutics, Cell Medica/Kuur, Incyte, Precision Biosciences, Legend Biotech, Adicet Bio, Calibr, and Unum Therapeutics; research support from Kite, a Gilead Company, Bristol Myers Squibb, Merck, Poseida, Cellectis, Celgene, Karus Therapeutics, Unum Therapeutics, Allogene Therapeutics, Precision Biosciences, and Acerta; and patents, royalties, or other intellectual property from Takeda Pharmaceuticals. ES is a consultant on scientific advisory boards of Bayer HealthCare Pharmaceuticals, Novartis, Magenta, Adaptimmune, Mesoblast, Axio; received honoraria from Magenta, Novartis, Bayer HealthCare Pharmaceuticals and other remuneration from license agreements or patents from Takeda. EJS is listed as a coinventor on a provisional patent application on Takeda owned by MD Anderson and licensed to Takeda. MW consults for AstraZeneca, Bayer Healthcare, BeiGene, CSTone, DTRM Biopharma (Cayman) Limited, Epizyme, Genentech, InnoCare, Janssen, Juno, Kite Pharma, Loxo Oncology, Miltenyi Biomedicine GmbH, Oncternal, Pharmacyclics, Veloso; has received research funding from Acerta Pharma, AstraZeneca, BeiGene, BioInvent, Celgene, Innocare, Janssen, Juno, Kite Pharma, Lilly, Loxo Oncology, Molecular Templates, Oncternal, Pharmacyclics, VelosBio; has received honorari from Acerta Pharma, Anticancer Association, AstraZeneca, BeiGene, CAHON, Chinese Medical Association, Clinical Care Options, Dava Oncology, Epizyme, Hebei Cancer Prevention Federation, Imbruvica, Imedex, Janssen, Kite Pharma, Miltenyi Biomedicine GmbH, Moffit Cancer Center, Mumbai Hematology Group, Newbridge Pharmaceuticals, OMI, Physicians Education Resources (PER), Scripps, The First Afflicted Hospital of Zhejiang University. FV received research funding from CRISP Therapeutics and Geron Corporation; received honoraria from i3Health, Else-

vier, America Registry of Pathology, Congressionally Directed Medical Research Program, Society of Hematology Oncology in the last 3 years. JW consults for BMS, Novartis, Kite, Genentech, Morphosys, AstraZeneca, ADC Therapeutics, Iksuda, Umoja; has received research funding from BMS, Novartis, Kite, Genentech, Morphosys, AstraZeneca, ADC Therapeutics, Curis, Unum. LJN has received honorarium and research funding from BMS/Celgene, Epizyme, Genentech, Gilead/KITE, Janssen, Novartis, Pfizer, TG Therapeutics, and Takeda and research funding from Caribou Biosciences and IGM Biosciences. All other authors have no conflicts of interest to disclose. Contributions RES designed the study, provided clinical care to patients, analyzed data, and wrote the paper; JB designed the study, reviewed echocardiograms, provided clinical care to patients, and co-authored the paper; EK designed the study, reviewed EKG, and co-authored the paper; AD, LN, JW, CG designed the study, analyzed the data, provided clinical care to patients, and co-authored the paper; LF reviewed the statistical analysis and co-authored the paper; MB, PS, CP, SA, SN, ES, GR collected clinical data, provided clinical care to patients and co-authored the paper; NP, KK, MW, F. provided clinical care to patients and co-authored the paper; CC collected clinical data and co-authored the paper. Funding This work was supported in part by the National Institutes of Health, National Cancer Institute, Cancer Center Support (CORE) grant CA 016672 to the University of Texas MD Anderson Cancer Center. PS salary is supported by the Lymphoma Research Foundation Career Development Award. AD is supported in part by the Ting Tsung and Wei Fong Chao Distinguished Chair.

References 1. Nastoupil LJ, Jain MD, Feng L, et al. Standard-of-care axicabtagene ciloleucel for relapsed or refractory large B-cell lymphoma: results from the US lymphoma CAR T Consortium. J Clin Oncol. 2020;38(27):3119-3128. 2. Locke FL, Ghobadi A, Jacobson CA, et al. Long-term safety and activity of axicabtagene ciloleucel in refractory large B-cell lymphoma (ZUMA-1): a single-arm, multicentre, phase 1-2 trial. Lancet Oncol. 2019;20(1):31-42. 3. Abramson JS, Palomba ML, Gordon LI, et al. Lisocabtagene maraleucel for patients with relapsed or refractory large B-cell lymphomas (TRANSCEND NHL 001): a multicentre seamless design study. Lancet. 2020;396(10254):839-852. 4. Neelapu SS, Locke FL, Bartlett NL, et al. Axicabtagene ciloleucel CAR T-cell therapy in refractory large B-cell lymphoma. N Engl J Med. 2017;377(26):2531-2544. 5. Schuster SJ, Bishop MR, Tam CS, et al. Tisagenlecleucel in adult relapsed or refractory diffuse large B-cell lymphoma. N Engl J Med. 2019;380(1):45-56.

6. https://www.fda.gov/news-events/press-announcements/fdaapproves-car-t-cell-therapy-treat-adults-certain-types-large-b -cell-lymphoma access date 01 21 2020 FDA approves CAR-T cell therapy to treat adults with certain types of large B-cell lymphoma. Yescarta (axicabtagene ciloleucel). 7. https://www.fda.gov/drugs/resources-information-approveddrugs/fda-approves-tisagenlecleucel-adults-relapsed-or-refract ory-large-b-cell-lymphoma access date 01 20 20. FDA approves tisagenlecleucel for adults with relapsed or refractory large Bcell lymphoma. 8. https://www.fda.gov/drugs/resources-information-approveddrugs/fda-approves-lisocabtagene-maraleucel-relapsed-or-refr actory-large-b-cell-lymphoma access date 07/11/2021. FDA approves lisocabtagene maraleucel for relapsed or refractory large B-cell lymphoma. 9. Strati P, Nastoupil LJ, Westin J, et al. Clinical and radiologic correlates of neurotoxicity after axicabtagene ciloleucel in large B-cell lymphoma. Blood Adv. 2020;4(16):3943-3951.

Haematologica | 107 July 2022

1565

ARTICLE - Cardiovascular events post T-cell therapy

R.E. Steiner et al.

10. Burns EA, Gentille C, Trachtenberg B, Pingali SR, Anand K. Cardiotoxicity associated with anti-CD19 chimeric antigen receptor T-cell (CAR-T) therapy: recognition, risk factors, and management. Diseases. 2021;9(1):20. 11. Ganatra S, Carver JR, Hayek SS, et al. Chimeric antigen receptor T-cell therapy for cancer and heart: JACC council perspectives. J Am Coll Cardiol. 2019;74(25):3153-3163. 12. Patel NP, Doukas PG, Gordon LI, Akhter N. Cardiovascular toxicities of CAR T-cell therapy. Curr Oncol Rep. 2021;23(7):78. 13. Pathan N, Hemingway CA, Alizadeh AA, et al. Role of interleukin 6 in myocardial dysfunction of meningococcal septic shock. Lancet. 2004;363(9404):203-209. 14. Ganatra S, Redd R, Hayek SS, et al. Chimeric antigen receptor Tcell therapy-associated cardiomyopathy in patients with refractory or relapsed non-Hodgkin lymphoma. Circulation. 2020;142(17):1687-1690. 15. Berro M, Arbelbide JA, Rivas MM, et al. Hematopoietic cell transplantation-specific comorbidity index predicts morbidity and mortality in autologous stem cell transplantation. Biol Blood Marrow Transplant. 2017;23(10):1646-1650. 16. Alvi RM, Frigault MJ, Fradley MG, et al. Cardiovascular events among adults treated with chimeric antigen receptor T-cells (CAR-T). J Am Coll Cardiol. 2019;74(25):3099-3108. 17. Neelapu SS, Tummala S, Kebriaei P, et al. Chimeric antigen receptor T-cell therapy - assessment and management of toxicities. Nat Rev Clin Oncol. 2018;15(1):47-62. 18. Lee DW, Santomasso BD, Locke FL, et al. ASTCT consensus grading for cytokine release syndrome and neurologic toxicity associated with immune effector cells. Biol Blood Marrow Transplant. 2019;25(4):625-638. 19. Cheson BD, Fisher RI, Barrington SF, et al. Recommendations for initial evaluation, staging, and response assessment of Hodgkin and non-Hodgkin lymphoma: the Lugano classification. J Clin Oncol. 2014;32(27):3059-3068. 20. Alexander M, Culos K, Roddy J, et al. Chimeric antigen receptor T cell therapy: a comprehensive review of clinical efficacy, toxicity, and best practices for outpatient administration. Transplant Cell

Ther. 2021;27(7):558-570. 21. Lefebvre B, Kang Y, Smith AM, Frey NV, Carver JR, ScherrerCrosbie M. Cardiovascular effects of CAR T cell therapy: a retrospective study. JACC CardioOncol. 2020;2(2):193-203. 22. Neelapu SS, Jacobson CA, Oluwole OO, et al. Outcomes of older patients in ZUMA-1, a pivotal study of axicabtagene ciloleucel in refractory large B-cell lymphoma. Blood. 2020;135(23):2106-2109. 23. Strati P, Ahmed S, Furqan F, et al. Prognostic impact of corticosteroids on efficacy of chimeric antigen receptor T-cell therapy in large B-cell lymphoma. Blood. 2021;137(23):3272-3276. 24. Kuznetsova T, Herbots L, López B, et al. Prevalence of left ventricular diastolic dysfunction in a general population. Circ Heart Fail. 2009;2(2):105-112. 25. Nagueh SF, Smiseth OA, Appleton CP, et al. Recommendations for the evaluation of left ventricular diastolic function by echocardiography: an update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging. Eur Heart J Cardiovasc Imaging. 2016;17(12):1321-1360. 26. Nagueh SF, Appleton CP, Gillebert TC, et al. Recommendations for the evaluation of left ventricular diastolic function by echocardiography. Eur J Echocardiogr. 2009;10(2):165-193. 27. Shalabi H, Sachdev V, Kulshreshtha A, et al. Impact of cytokine release syndrome on cardiac function following CD19 CAR-T cell therapy in children and young adults with hematological malignancies. J Immunother Cancer. 2020;8(2). 28. Wudhikarn K, Pennisi M, Garcia-Recio M, et al. DLBCL patients treated with CD19 CAR T cells experience a high burden of organ toxicities but low nonrelapse mortality. Blood Adv. 2020;4(13):3024-3033. 29. Strati P, Ahmed S, Kebriaei P, et al. Clinical efficacy of anakinra to mitigate CAR T-cell therapy-associated toxicity in large B-cell lymphoma. Blood Adv. 2020;4(13):3123-3127. 30. Ambale-Venkatesh B, Yang X, Wu CO, et al. Cardiovascular event prediction by machine learning: the Multi-Ethnic Study of Atherosclerosis. Circ Res. 2017;121(9):1092-1101.

Haematologica | 107 July 2022

1566

ARTICLE - Hemostasis

Renal function and clinical outcome of patients with cancer-associated venous thromboembolism randomized to receive apixaban or dalteparin. Results from the Caravaggio trial Cecilia Becattini,1 Rupert Bauersachs,2 Giorgio Maraziti,1 Laurent Bertoletti,3 Alexander Cohen,4 Jean M. Connors,5 Dario Manfellotto,6 Antonio Sanchez,7 Benjamin Brenner8 and Giancarlo Agnelli1 Internal, Vascular and Emergency Medicine - Stroke Unit, University of Perugia, Perugia, Italy; 2Klinikum Darmstadt GmbH, Darmstadt, Germany; 3Service de Médecine Vasculaire et Thérapeutique, CHU de St-Etienne, Saint-Etienne, France; 4Department of Haematology, St. Thomas' Hospital, King's College London, London, UK; 5Brigham and Women’s Hospital/Hematology Division, Harvard Medical School, Boston, MA, USA; 6Clinical Research Department, FADOI Foundation, Milan, Italy; 6Internal Medicine Department, Fatebenefratelli Foundation, San Giovanni Calibita Fatebenefratelli Hospital, Rome, Italy; 7Hospital Puerta de Hierro, Madrid, Spain and 8Institute of Hematology and BMT Rambam Health Care Campus Technion, Israel Institute of Technology Haifa, Haifa, Israel

Correspondence: Cecilia Becattini cecilia.becattini@unipg.it

Received: April 29, 2021. Accepted: July 30, 2021. Prepublished: August 12, 2021. https://doi.org/10.3324/haematol.2021.279072 ©2022 Ferrata Storti Foundation Haematologica material is published under a CC BY-NC license

Abstract The effect of renal impairment (RI) on risk of bleeding and recurrent thrombosis in cancer patients treated with direct oral anticoagulants for venous thromboembolism (VTE) is undefined. We ran a prespecified analysis of the randomized Caravaggio study to evaluate the role of RI as a risk factor for bleeding or recurrence in patients treated with dalteparin or apixaban for cancerassociated VTE. RI was graded as moderate (creatinine clearance between 30-59 mL/minute; 275 patients) and mild (between 6089 mL/minute; 444 patients). In the 1142 patients included in this analysis, the incidence of major bleeding was similar in patients with moderate vs. no or mild RI (HR 1.06-95% CI: 0.53-2.11), with no difference in the relative safety of apixaban and dalteparin. Recurrent VTE was not different in moderate vs. no or mild RI (HR=0.67, 95% CI: 0.38-1.20); in moderate RI, apixaban reduced recurrent VTE compared to dalteparin (HR=0.27, 95% CI: 0.08-0.96; P for interaction 0.1085). At multivariate analysis, no association was found between variation of renal function over time and major bleeding or recurrent VTE. Advanced or metastatic cancer was the only independent predictor of major bleeding (HR=2.84, 95% CI: 1.20-6.71), with no effect of treatment with apixaban or dalteparin. In our study, in cancer patients treated with apixaban or dalteparin, moderate RI was not associated with major bleeding or recurrent VTE. In patients with moderate renal failure, the safety profile of apixaban was confirmed with the potential for improved efficacy in comparison to dalteparin. ClinicalTrials.gov identifier: NCT03045406.

Introduction The treatment of venous thromboembolism (VTE) in cancer patients is challenging due to the high risk of recurrent VTE and bleeding.1,2 In these patients, low molecular weight heparin (LMWH) was shown to reduce the rate of recurrent VTE in comparison to vitamin K antagonists, without increasing bleeding complications.3,4 Randomized studies have shown that the direct oral anti-Xa agents edoxaban, rivaroxaban and apixaban are non-inferior to dalteparin in the treatment of VTE in cancer patients.5-8 Hence, major international guidelines have recently considered direct oral anti-Xa agents as an alternative to LMWH for the treatment of cancer-associated VTE.9-11

Renal impairment (RI) was independently associated with an increase of approximately 40% in the risk of major bleeding (4.6 vs. 2.4% person-years, adjusted hazard ratio= 1.40; 95% CI: 1.03-1.90) and recurrent thromboembolism (6.6 vs. 5.0% person-years, adjusted hazard ratio= 1.40; 95% CI: 1.10-1.77) in patients receiving anticoagulants for the treatment of VTE.12 The association between RI and the risk for recurrent VTE and bleeding was assessed in cancer patients receiving LMWH or vitamin K antagonists in observational studies and in sub-analyses of randomized studies.13-16 In these sub-analyses, the efficacy to safety profile of LMWH in comparison to warfarin was similar in cancer patients with and without RI.13-16 The effect of RI on the risk of recurrent VTE and bleeding

Haematologica | 107 July 2022

1567

ARTICLE - Apixaban and renal function in cancer-associated VTE in patients with cancer-associated VTE treated with direct oral anti-Xa agents is uncertain.17 As direct anti-Xa agents have a variable but substantial degree of renal excretion, RI may be associated with increased plasma levels of these agents with potential for increased bleeding risk. In phase III randomized controlled trials of direct oral anticoagulants for the treatment of VTE in the general population, no difference in recurrent VTE (RR=0.70, 95% CI: 0.43–1.15) but a significant reduction in major bleeding (RR= 0.51, 95% CI: 0.26–0.99) was seen in comparison with vitamin K antagonists in patients with creatinine clearance of 30-49 mL/min.18 The effect of RI on the efficacy and safety profile of the direct oral anticoagulants in the four randomized clinical studies of oral anti-Xa agents in patients with cancer associated VTE is unknown. In these studies, although patients with a creatinine clearance < 30 mL/min were excluded, a consistent proportion of patients had mild or moderate RI.5-8 In the Caravaggio study, oral apixaban was found to be non-inferior to subcutaneous dalteparin for the treatment of cancer-associated VTE.8,19 The rate of major bleeding was similar with apixaban and dalteparin. The aim of this pre-specified analysis in patients included in the Caravaggio study was to assess the association between RI and bleeding or recurrent VTE. The effect of both baseline renal function and its variations during treatment with apixaban or dalteparin was evaluated for the association with major bleeding and recurrent VTE.

Methods Caravaggio was a multinational, prospective, randomized, open-label, with blinded end-point evaluation (PROBE), non-inferiority study aimed at assessing whether oral apixaban was non-inferior to dalteparin for the treatment of newly diagnosed proximal deep vein thrombosis (DVT) and/or pulmonary embolism (PE) in patients with cancer. The rationale, design and results of the Caravaggio study were described previously.8,19 The trial was performed in accordance with the provisions of the Declaration of Helsinki and local regulations. The protocol and its amendments were approved by the institutional review board or ethics committee at each trial center. All the patients provided written informed consent. Consecutive adult patients with cancer who had symptomatic or incidental acute proximal DVT or PE were randomized to receive oral apixaban (10 mg twice daily for the first 7 days, followed by 5 mg twice daily) or subcutaneous dalteparin (200 IU per kilogram of body weight once daily for the first month, followed by 150 IU per kilogram once daily) for 6 months. Inclusion and exclusion

C. Becattini et al.

Table 1. Renal impairment stage classification.

Severity classification

eGFR*

Preserved

≥ 90 mL/min

Mild reduction

<90 mL/min and ≥60 mL/min

III

Moderate reduction

<60 mL/min and ≥30 mL/min

IIIa

<60 mL/min and ≥45 mL/min

IIIb

<45 mL/min and ≥30 mL/min

Severe reduction

< 30 mL/min and ≥15 mL/min

Pre-dialysis

<15 mL/min

RI stage

*eGFR based on the Cockcroft-Gault equation. RI: renal impairment; eGFR: estimated glomerular filtration rate.

criteria are reported in the Online Supplementary Data. Only patients with baseline creatinine assessment at randomization were included in this prespecified analysis. Patients were excluded in case of a creatinine clearance <30 mL/min based on the Cockcroft Gault equation. RI was classified into the conventional five stages as indicated in Table 1.20 Study outcomes This analysis has two co-primary outcomes: major bleeding and recurrent VTE defined according to Caravaggio criteria and occurring from randomization to day 180 (see Online Supplementary Data). Secondary study outcomes were clinically relevant nonmajor bleeding and a composite of major bleeding and recurrent VTE. Follow-up and measurements Renal function was calculated by three accepted methods, the Cockcroft-Gault, Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI), and Modification of Diet in Renal Disease (MDRD) equations (see Online Supplementary Data).21-23 Patients were also categorized on the basis of estimated glomerular filtration rate (eGFR) as having no or mild RI (eGFR of 60 mL/min or higher) vs. moderate RI (eGFR lower than 60 mL/min) and having eGFR of ≥50 mL/min vs. <50 mL/min. The management of study treatments according to creatinine clearance was dictated by the protocol. Statistical analysis To assess the effect of RI in the risk for study outcome events, two different analyses were performed: (i) a comparison of event rates in subgroups of patients randomized to apixaban or dalteparin, identified based on a specific cut-off level for eGFR (60 or 50 mL/min) at inclusion in the study; (ii) proportional hazards model for the time to study outcome events with eGFR (according to the Cockroft-Gault formula) as a time-varying covariate. We analysed log-

Haematologica | 107 July 2022

1568

ARTICLE - Apixaban and renal function in cancer-associated VTE

C. Becattini et al.

Table 2. Baseline renal function in patients randomized to apixaban or dalteparin.

Baseline eGFR Cockroft-Gault, mean ± SD CKD-EPI, mean ± SD MDRD, mean ± SD CKD stage according to Cockroft-Gault formula, n (%) I (eGFR ≥90 mL/min) II (eGFR <90 mL/min and ≥60 mL/min) IIIa (eGFR <60 mL/min and ≥45 mL/min) IIIb (eGFR <45 mL/min and ≥30 mL/min) CKD stage according to CKD-EPI, n (%) I (eGFR ≥90 mL/min) II (eGFR <90 mL/min and ≥60 mL/min) IIIa (eGFR <60 mL/min and ≥45 mLmin) IIIb (eGFR <45 mL/min and ≥30 mL/min) IV (eGFR <30 mL/min and ≥15 mL/min) CKD stage according to MDRD, n (%) I (eGFR ≥90 mL/min) II (eGFR <90 mL/min and ≥60 mL/min) IIIa (eGFR <60 mL/min and ≥45 mLmin) IIIb (eGFR <45 mL/min and ≥30 mL/min) IV (eGFR <30 mL/min and ≥15 mL/min)

All patients N=1142 (%)

Apixaban N=573 (50.2%)

Dalteparin N=569 (49.8%)

85.2 ± 33.9 78.3 ± 20.8 87.0 ± 33.9

84.8 ± 34.4 78.5 ± 20.7 86.7 ± 32.3

85.6 ± 33.5 78.1 ± 20.8 86.4 ± 30.6

423 (37.0) 444 (38.9) 181 (15.8) 94 (8.2)

201 (35.1) 235 (41.0) 97 (16.9) 40 (7.0)

222 (39.0) 209 (36.7) 84 (14.8) 54 (9.5)

364 (31.9) 524 (45.9) 189 (16.5) 62 (5.4) 3 (0.3)

176 (30.7) 270 (47.1) 97 (16.9) 29 (5.1) 1 (0.2)

188 (33.0) 254 (44.6) 92 (16.2) 33 (5.8) 2 (0.4)

439 (38.4) 479 (41.9) 176 (15.4) 46 (4.0) 2 (0.2)

219 (38.2) 242 (42.2) 92 (16.1) 19 (3.3) 1 (0.2)

220 (38.7) 237 (41.7) 84 (14.8) 27 (4.7) 1 (0.2)

Percentages are calculated relative to the total number of subjects in the modified intention-to-treat analysis set in each group. eGFR: estimated glomerular filtration rate; SD: standard deviation; CKD-EPI: Chronic Kidney Disease Epidemiology Collaboration; MDRD: Modified Diet and Renal Disease.

transformed eGFR (Y = log (eGFR) data throughout, as previously described for this kind of analyses.24 The final set of covariates for the multivariate analysis was selected among those with a P-value of 0.15 or less at univariate analyses. For comparison of proportions, deterioration of renal function was defined as a decrease in eGFR leading to a change of at least one stage from baseline values, according to the Cockroft-Gault formula. Because subgroup analyses in the present study were exploratory, the P values were not adjusted for multiple comparisons and should be interpreted with caution. Analyses were performed with SAS software (version 9.4). Other details are reported in the Online Supplementary Data.

Results Overall, 1142 patients were included in this analysis (Online Supplementary Data, Table S1). Thirteen patients from the original Caravaggio study were excluded due to lack of renal function assessment at inclusion in the study - one of whom reported major bleeding during the study - or for an eGFR lower than 30 mL/min at inclusion. Mean baseline eGFR according to the Cockroft-Gault, CKD-EPI and MDRD equations are reported in Table 2. At inclusion in the

study, 37% of patients had stage 1, 39% stage 2 and 24% stage 3 RI; 8.2% had stage 3b RI with eGFR between 30 and 44 mL/min according to the Cockroft-Gault formula. CKD-EPI classified a numerically lower proportion of patients in stage 1 and a numerically higher proportion in stage 2 RI compared to Cockroft-Gault or MDRD equations. Mean baseline eGFR, as well as distribution across different stages of RI, were similar in patients randomized to apixaban or dalteparin whatever the formula used for calculation of eGFR (Table 2). At inclusion in the study, 23.9% and 24.3% of patients randomized to receive apixaban and dalteparin had stage 3 RI according to the Cockroft-Gault formula, respectively. The distribution of RI stage according to cancer stage or site is reported in Online Supplementary Data, Table S2. Renal function and study outcome events The mean eGFR at baseline was similar in patients who experienced a major bleeding event compared to patients who did not experience one during the study (Table 3). At inclusion in the study, 25% and 24% of patients who experienced or did not experience a major bleed during the study had stage 3 RI, respectively. Event rates by CKD stage calculated by different formulas are reported in Online Supplementary Data, Table S3. The incidence of major bleeding was similar in patients with moderate RI vs. patients with no or mild RI (4.0 vs.

Haematologica | 107 July 2022

1569

ARTICLE - Apixaban and renal function in cancer-associated VTE

C. Becattini et al.

Table 3. Baseline characteristics of the study population according to study outcome groups.

All patients N= 1142 (%) Age, years Mean ± SD > 75 years, N (%) Range Female gender, N BMI, mean ± SD Baseline eGFR Cockroft Gault, mean ± SD CKD-EPI, mean ± SD MDRD, mean ± SD CKD stage according to Cockroft-Gault formula, N (%) I (eGFR ≥90 mL/min) II (eGFR <90 mL/min and ≥60 mL/min) IIIa (eGFR <60 mL/min and ≥45 mL/min) IIIb (eGFR <45 mL/min and ≥30 mL/min) CKD stage according to CKD-EPI, N (%) I (eGFR ≥90 mL/min) II (eGFR <90 mL/min and ≥60 mL/min) IIIa (eGFR <60 mL/min and ≥45 mL/min) IIIb (eGFR <45 mL/min and ≥30 mL/min) IV (eGFR <30 mL/min and ≥15 mL/min) CKD stage according to MDRD, N (%) I (eGFR ≥90 mL/min) II (eGFR <90 mLmin and ≥60 mL/min) IIIa (eGFR <60 mL/min and ≥45 mL/min) IIIb (eGFR <45 mL/min and ≥30 mL/min) IV (eGFR <30 mL/min and ≥15 mL/min) Locally advanced or metastatic cancer, N (%) ECOG score, N (%) 1 2

Patients Patients Patients Patients with without without with major major bleeding recurrent VTE recurrent VTE bleeding N= 78 (6.8%) N = 1098 N = 1064 N= 44 (3.9 %) (96.1 %) (93.2%)

67.7±11.1 296 (25.9) 21-93 579 26.7±5.2

67.9±8.2 9 (20.5) 51-86 20 25.6±4.1

67.7±11.2 287 (26.1) 21-93 559 26.8±5.2

65.5±10.2 14 (17.9) 42-87 38 26.5±5.2

67.9±11.2 282 (26.5) 21-93 541 26.8±5.2

85.2±33.9 78.3±20.8 86.7±32.3

81.5±29.1 79.3±22.0 89.2±36.6

85.4±34.1 78.2±20.7 86.6±32.1

96.1±39.4 84.8±21.1 98.4±50.2

84.4±33.4 77.8±20.7 85.8±30.4

423 (37.0) 444 (38.9) 181 (15.8) 94 (8.2)

18 (40.9) 15 (34.1) 6 (13.6) 5 (11.4)

405 (36.9) 429 (39.1) 175 (15.9) 89 (8.1)

39 (50.0) 25 (32.1) 10 (12.8) 4 (5.1)

405 (36.9) 429 (39.1) 175 (15.9) 89 (8.1)

364 (31.9) 524 (45.9) 189 (16.5) 62 (5.4) 3 (0.3)

15 (34.1) 18 (40.9) 8 (18.2) 2 (4.5) 1 (2.3)

349 (31.8) 506 (46.1) 181 (16.5) 60 (5.5) 2 (0.2)

32 (41.0) 32 (41.0) 13 (16.7) 1 (1.3) 0 (0.0)

332 (31.2) 492 (46.2) 176 (16.5) 61 (5.7) 3 (0.3)

439 (38.4) 479 (41.9) 176 (15.4) 46 (4.0) 2 (0.2)

19 (43.2) 15 (34.1) 8 (18.2) 1 (2.3) 1 (2.3)

420 (38.3) 464 (42.3) 168 (15.3) 45 (4.1) 1 (0.1)

39 (50.0) 29 (37.2) 10 (12.8) 0 (0.0) 0 (0.0)

400 (37.6) 450 (42.3) 166 (15.6) 46 (4.3) 2 (0.2)

779 (68.2)

38 (85.4)

741 (67.5)

38 (85.4)

741 (67.5)

553 (48.4) 235 (20.6)

20 (45.5) 14 (31.8)

533 (48.5) 221 (20.1)

20 (45.5) 14 (31.8)

533 (48.5) 221 (20.1)

MB: major bleeding. *according to Cockroft-Gault formula. VTE: venous thromboembolism: SD: standard deviation; GFR: estimated glomerular filtration rate; CKD-EPI: Chronic Kidney Disease Epidemiology Collaboration; MDRD: Modified Diet and Renal Disease; CKD: chronic kidney disease; ECOG: Eastern Cooperative Oncology Group.

3.8%; HR=1.06, 95% CI: 0.53-2.11) (Table 4). These results were confirmed in patients randomized to receive apixaban (3.6 vs. 3.7%) or dalteparin (4.3 vs. 3.9%). Rates of major bleeding were similar in patients randomized to apixaban or dalteparin in the two groups of moderate RI and no or mild RI (P-value for interaction 0.8819). These results were confirmed in patients with eGFR below or above 50 mL/min. A numerically lower rate of recurrent VTE was observed in patients with moderate RI as compared to patients with no or mild RI (HR=0.67, 95% CI: 0.38-1.20). Among patients randomized to apixaban, a not significant 69% reduction

of recurrent VTE was observed in those with moderate RI compared to those with no or mild RI (2.2 vs. 6.7%; HR= 0.31, 95% CI: 0.09-1.03). No difference in recurrent VTE was observed with dalteparin in patients with moderate RI vs. no or mild RI. A lower incidence of recurrent VTE was observed in patients with moderate RI randomized to apixaban compared to dalteparin (2.2 vs. 8.0%; HR=0.27, 95% CI: 0.08-0.96, P-value for interaction 0.1085). The reduction of recurrent VTE with apixaban compared to dalteparin was confirmed in patients with eGFR lower than 50 mL/min. A numerically higher rate of clinically relevant non-major

Haematologica | 107 July 2022

1570

ARTICLE - Apixaban and renal function in cancer-associated VTE

C. Becattini et al.

Table 4. Frequency of study outcome events based on baseline estimated glomerular filtration rate calculated according to the Cockroft-Gault formula. HR P for (95% CI) interaction

Major bleeding Overall eGFR <60*

Apixaban Dalteparin

Overall Apixaban Dalteparin

11/275

5/137

6/138

0·84

4.0

3.6

4.3

33/867

16/436

n/N, %

3.8

n/N, % eGFR ≥60*

95% CI

eGFR <50*

7/150

3/68

4/82

0·85

0.26-2.71

n/N %

4.7

4.4

4.9

0.19-3.80

17/431

0.92

eGFR ≥50*

37/992

18/505

19/487

0.91

3.7

3.9

0.47-1.83

n/N, %

3.7

3.6

3.9

0.48-1.73

1.06

1.02

1.11

1.26

1.24

1.28

0.53-2.11

0.37-2.79

0.43-2.83

95% CI

Overall

0·8819

HR (95% CI)

Recurrent VTE

eGFR <60*

HR P for (95% CI) interaction

Major bleeding

0.56-2.84 0.37-4.20

P for interaction

HR P for (95% CI) interaction

Overall

Apixaban

Dalteparin

eGFR <50*

6/150

0/68

6/82

14/275

3/137

11/138

0.27

5.1

2.2

8.0

0.08-0.96

n/N %

4.0

0.0

7.3

64/867

29/436

35/431

0.82

eGFR ≥50*

72/992

32/505

40/487

0.76

n/N, %

7.4

6.7

8.1

0.50-1.33

n/N (%)

7.3

6.3

8.2

0.48-1.22

0.67

0.31

0.97

0.53

N.A.

0.87

0.38-1.20

0.09-1.03

0.50-1.91

95% CI

0.23-1.22

n/N, % eGFR ≥60*

95% CI

Recurrent VTE or major bleeding Overall eGFR <60*

0.1085

0.43-3.79

Recurrent VTE

Apixaban Dalteparin

HR 95% CI

P for interaction

Apixaban

Dalteparin

eGFR <50*

12/150

3/68

9/82

0.39

n/N %

8.0

4.4

11.0

0.12-1.40

8.4

5.8

93/867

42/436

51/431

0.81

eGFR ≥50*

104/992

47/505

57/487

0.78

n/N, %

10.7

9.6

11.8

0.54-1.21

n/N %

10.5

9.3

11.7

0.53-1.15

0.76

0.59

0.91

0.74

0.45

0.92

0.48-1.21

0.27-1.26

0.51-1.61

95% CI

15/138 10.9 0.23-1.24

Clinically relevant non-major bleeding Overall eGFR <60*

0·3675

Overall

8/137

eGFR ≥60*

0.53

HR (95% CI)

0.41-1.34 0.14-1.46

P for interaction

HR P for (95% CI) interaction

Overall

Apixaban

Dalteparin

eGFR <50*

12/150

5/68

7/82

0.87

26/275

18/137

8/138

2.35

9.5

13.1

5.8

1.01-5.45

n/N, %

8.0

7.4

8.5

0.27-2.76

61/867

34/436

27/431

1.26

eGFR ≥50*

75/992

47/505

28/487

1.65

n/N, %

7.0

7.8

6.3

0.76-2.09

n/N, %

7.6

9.3

5.7

1.03-2.63

1.35

1.67

0.94

1.05

0.76

1.53

0.85-2.12

0.95-2.94

0.42-2.08

(95% CI)

n/N, % eGFR ≥60*

(95% CI)

Major bleeding or CRNMB Overall eGFR <60* n/N, %

0.2364

HR (95% CI)

P for interaction

23/137

13/138

1.83

0.2011

13.1

16.8

9.4

0.92-3.62

0.2682

0.66-3.56

Major bleeding or CRNMB

Apixaban Dalteparin

36/275

0.57-1.93 0.31-1.90

0.2947

0.46-1.86

Clinically relevant non-major bleeding

Apixaban Dalteparin

N.A.

HR P for (95% CI) interaction

23/275

n/N, %

15/138

N.A.

0.37-2.04

Recurrent VTE or major bleeding

Apixaban Dalteparin

0.9757

HR P for (95% CI) interaction

Overall

Apixaban

Dalteparin

eGFR <50*

18/150

8/68

10/82

0.96

n/N, %

12.0

11.8

12.2

0.38-2.40

0.5026

Continued on following page. Haematologica | 107 July 2022

1571

ARTICLE - Apixaban and renal function in cancer-associated VTE

eGFR ≥60*

C. Becattini et al.

89/867

46/436

43/431

1.06

eGFR ≥50*

107/992

61/505

46/487

1.29

n/N, %

10.3

10.6

10.0

0.70-1.61

n/N, %

10.8

12.1

9.4

0.88-1.89

1.28

1.61

0.95

1.11

0.95

1.34

0.87-1.89

0.98-2.64

0.51-1.78

95% CI

0.67-1.84 0.46-1.98

All patients in the modified intention-to-treat set with an available baseline value of estimated glomerular filtration rate (eGFR) are considered in this table. An eGFR cut-off of 50 mL per minute was used in some phase III studies with direct oral anticoagulants. The hazard ratios (HR) (last column of the table) are adjusted for the competing risk of death unrelated to event by resorting to the Fine & Gray regression model using eGFR group, symptomatic vs. unsuspected VTE and active cancer vs. history of cancer as covariates. The HR for comparison between dalteparin and apixaban are adjusted for the competing risk of death unrelated to event by resorting to the Fine & Gray regression model using treatment group, symptomatic vs. unsuspected venous thromboembolism and active cancer vs. history of cancer as covariates. n: number of patients with events; N: total number of patients in each category. HR: hazard ratio; 95% CI: 95% confidence interval; eGFR: estimated glomerular filtration rate; VTE: venous thromboembolism; CRNMB: clinically relevant non-major bleeding.

bleeding was observed in patients with moderate RI vs. no or mild RI, and this was mainly accounted for by patients randomized to receive apixaban (Table 4). In patients with moderate RI, a two-fold increase in the incidence of clinically relevant non-major bleeding was observed with apixaban in comparison to dalteparin (Pvalue for interaction 0.2364). Both major and clinically relevant non-major bleeding were more common at genito-urinary sites in patients with reduced eGFR (either lower than 60 or 50 mL/min) in comparison to patients with no or mild RI; both major and clinically relevant nonmajor bleeding were more common at gastrointestinal sites in patients with no or mild RI (eGFR higher than 60 or 50 mL/min) in comparison to patients with reduced eGFR (either lower than 60 or 50 mL/min) (Online Supplementary Data, Table S4). Renal function over time and study outcome During the six-month treatment period, 288 patients (25%) experienced a deterioration of eGFR leading to a change of at least one stage from baseline values. This deterioration occurred in similar proportions of patients, regardless of the baseline eGFR stage (Online Supplementary Data, Table S5). Age at inclusion was associated with a deterioration of eGFR over time; treatment for cancer at inclusion or within the previous 6 months was associated with increasing eGFR over time (Online Supplementary Data, Table S6). No association was found between ECOG or cancer that was locally advanced/metastatic or unsuspected vs. symptomatic VTE or active cancer vs. history of cancer and eGFR over time. Variation of renal function over time was similar in patients randomized to receive apixaban or dalteparin (Online Supplementary Data, Figure S1). A major bleed occurred in 2.8 and 4.2% of patients having and not having deterioration of eGFR leading to a change of at least one stage from baseline values, respectively (Online Supplementary Data, Table S7). Recurrent VTE occurred in 2.8% and 8.2% of patients with or without deterioration of eGFR leading to a change of at least one stage from baseline values, respectively.

Multivariate analyses using renal function as a time-varying covariate and with death unrelated to an event as a competing risk were performed for study outcome events. Recurrent or locally-advanced or metastatic cancer (HR= 2.84, 95% CI: 1.20-6.71) was an independent predictor of major bleeding (Table 5). No independent association was found between variation of renal function over time and major bleeding or recurrent VTE. A not significant association was found between recurrent or locally-advanced or metastatic cancer (HR=1.65, 95% CI: 0.95-2.86) or treatment with apixaban (HR=0.66, 95% CI: 0.42-1.05) and recurrent VTE. No association between variation of renal function over time and the risk for recurrent VTE was observed.

Discussion Renal insufficiency occurs frequently in patients with cancer-associated VTE, with more than 60% of patients enrolled in the Caravaggio study having mild or moderate RI. RI was found not to be a risk factor for major bleeding or recurrent VTE in this population. No association was found between either moderate RI (eGFR lower than 60 mL/min) or defined by eGFR lower than 50 mL/min and major bleeding in patients randomized to apixaban or dalteparin. Apixaban appeared to reduce recurrent VTE in patients with moderate RI in comparison to dalteparin, with no effect on the incidence of major bleeding but a not significant increase in clinically relevant non-major bleeding. The efficacy to safety ratio of apixaban was similar to that of dalteparin in patients with normal renal function or mild RI. Recurrent or locally advanced or metastatic cancer was the only independent predictor for major bleeding in this population of patients with cancer-associated VTE. RI is a common condition in the general adult population and is prevalent in patients with cancer. In a single-center cohort study, about two-thirds of patients with solid cancer had abnormal renal function (eGFR <90 mL/min/1.73 m2); 15% had moderate RI and 1% to 2% had severe RI.25 In the CLOT study, 24% of patients with

Haematologica | 107 July 2022

1572

ARTICLE - Apixaban and renal function in cancer-associated VTE

C. Becattini et al.

Table 5. Risk factors for major bleeding, recurrent venous thromboembolism or clinically relevant non-major bleeding according to a Cox model with time-varying covariates. Risk factors for major bleeding eGFR* over time Baseline age (years) Treatment for cancer at the time of inclusion or within previous 6 months Recurrent locally advanced or metastatic cancer Treatment (reference Dalteparin) / Apixaban Risk factors for recurrent VTE eGFR* over time Baseline age (years) Treatment for cancer at the time of inclusion or within previous 6 months Recurrent locally advanced or metastatic cancer Treatment (reference Dalteparin) / Apixaban

95%CI

0.57 0.99

0.21-1.55 0.97-1.01

0.26 0.41

1.45

0.67-3.14

0.35

2.84 0.96

1.20-6.71 0.53-1.75

0.02 0.89

1.70 0.99

0.82-3.52 0.97-1.01

0.15 0.48

0.74

0.33-1.62

0.45

1.65 0.66

0.95-2.86 0.42-1.05

0.08 0.08

1.18 1.03

0.61-2.28 1.00-1.05

0.63 0.03

1.07

0.59-1.98

0.83

1.25 1.52

0.77-2.03 0.99-2.35

0.36 0.08

1.17 0.99

0.62-2.22 0.97-1.01

0.63 0.23

1.08

0.62-1.86

0.80

1.99 0.74

1.24-3.21 0.51-1.07

0.004 0.11

Risk factors for clinically relevant non-major bleeding eGFR* over time Baseline age (years) Treatment for cancer at the time of inclusion or within previous 6 months Recurrent locally advanced or metastatic cancer Treatment (reference Dalteparin) / Apixaban Risk factors for major bleeding or recurrent VTE eGFR* over time Baseline age (years) Treatment for cancer at the time of inclusion or within previous 6 months Recurrent locally advanced or metastatic cancer Treatment (reference Dalteparin) / Apixaban

All patients in the modified intention-to-treat set with an available baseline value of estimated glomerular filtration rate (eGFR) are considered in this table. *eGFR included in the analyses as log-transformed eGFR (see methods section). HR: hazard ratio; 95% CI: 95% confidence interval; VTE: venous thromboembolism.

cancer-associated VTE had eGFR lower than 60 mL/min, with 22% having moderate RI and 2% severe RI.16 In the Catch study, 15% of patients with cancer-associated VTE had eGFR lower than 60 mL/min at inclusion;15 RI was more common in patients with gynecological and genitourinary malignancies. In the EINSTEIN VTE studies, more than half of the patients identified as having cancer had creatinine clearance <80 mL/min and 15% had creatinine clearance <50 mL/min.26 In the Amplify VTE study, patients with active cancer or history of cancer more commonly had creatinine clearance lower than 50 mL/min in comparison to non-cancer patients.27 Overall, our results showing a prevalence of moderate RI greater than 20% in patients with cancer-associated VTE are consistent with those from previous studies in this setting. The clinical relevance of our observation is related to the hypothesis that RI could reduce renal excretion of both LMWH and direct anti-Xa anticoagulants and potentially lead to anticoagulant overdosing and an increase in bleeding complications. Whether dose reduction or dose adjustment of LMWH based on anti-Xa activity are valid options to avoid an in-

crease in bleeding risk in patients with severe RI is uncertain.28,29 The evidence in favor of these approaches is limited both in the general population of patients with VTE and in cancer patients with VTE. To date, the optimal anticoagulant agent and regimen in patients with severe renal failure is undefined. Among LMWH, dalteparin seems to be associated with acceptable bleeding risk in patients with RI.16,30,31 Among direct oral anti-Xa anticoagulants, dose reduction for patients with RI was evaluated for edoxaban in the treatment of VTE. No dose adjustment has been tested in clinical trials for the treatment of VTE with apixaban or rivaroxaban. Our study shows that the current regimen of apixaban developed and approved for the treatment of VTE is effective and safe in cancer patients with moderate RI. We found no association between RI and bleeding or thromboembolic risk in patients with cancer-associated VTE randomized to receive apixaban or dalteparin. These findings are in contrast with those from previous studies in the general population of patients with acute VTE as well as in cancer patients included in a randomized study

Haematologica | 107 July 2022

1573

ARTICLE - Apixaban and renal function in cancer-associated VTE that compared tinzaparin or dalteparin with warfarin.12,15,16 These studies showed an increased risk of bleeding in patients with moderate RI in comparison with patients with no or mild RI. In two of these studies, RI was also associated with an increase in recurrent VTE.12,15 A comparison of recent studies with anti-Xa agents for the treatment of cancer-associated VTE is difficult as only about 7% of patients in the HOKUSAI VTE cancer study had moderate RI, with no RI data available for the Select-D study.5,6 This analysis shows similar safety profiles of apixaban and dalteparin across different stages of RI in terms of major bleeding. An increase in clinically relevant non-major bleeding was observed with apixaban compared to dalteparin in patients with moderate RI. This difference was mainly driven by an increase in apixaban-associated genito-urinary bleeding in patients with moderate RI. An increase in genito-urinary bleeding was reported in the main Caravaggio study and we have now added information that it is related to moderate RI.8 Whether this increase in genito-urinary bleeding is to be related to the presence of genito-urinary cancer, to the accumulation of study drugs at the urinary site or, rather, to the use of specific anticancer agents remains undefined.32 Indeed, as these findings derive from subgroup analyses and refer to limited numbers of events, they can only be regarded as hypothesis generating. In phase III trials with direct oral anticoagulants for the treatment of VTE, no effect of RI on efficacy was reported.17,18 Concerning the safety of direct oral anticoagulants, the favorable effect observed in the general population was confirmed across different stages of RI. In our study in patients with cancer-associated VTE, about 25% of patients had a significant variation of renal function leading to change in eGFR stage. A previous study in patients with cancer-associated VTE reported similar rates of variation of renal function over time.15 Anticancer agents, particularly platinum-based agents, may affect renal function, potentially leading to deterioration of RI. In this case, patients with RI may require surveillance for periodical reassessment of renal function and bleeding risk. In a subgroup meta-analysis on 1789 patients without cancer with creatinine clearance between 30 and 49 mL/min included in phase III trials in the treatment of VTE, no difference in terms of recurrent VTE (RR=0.70, 95% CI: 0.43–1.15) was reported in patients receiving direct oral anticoagulants vs. vitamin K antagonists (VKA); however, a significant reduction in major bleedings (RR=0.51, 95% CI: 0.26–0.99) was found.18 Similarly, in the four randomized clinical studies with oral anti-Xa agents for the treatment of cancer-associated VTE, no effect of renal function emerged on the efficacy and safety profile of these agents.5-8 Taken together, these data suggest that the oral anti-Xa agents could be used in patients with cancer-associated VTE and creatinine clearance 30-49 mL/min. Pa-

C. Becattini et al.

tients with a creatinine clearance <30 mL/min were excluded from phase III clinical trials on the treatment of cancer-associated VTE and no dose adjustment was scheduled for direct oral anticoagulants based on renal function in these patients, except for the studies with edoxaban. Data regarding safety and efficacy of direct oral anticoagulants in cancer patients with severe renal impairment are lacking. It is conceivable that, for patients with creatinine clearance <30 mL/min, treatment with unfractionated heparin may be preferred or, as an alternative, vitamin K antagonists can still be still an option. Despite being based on limited evidence, dose- and perhaps anti-Xa activity-adjusted LMWH might be considered.28,29,33,34 Further evidence is needed to define the optimal anticoagulant strategy for these patients. Several equations are currently in use to non-invasively estimate eGFR. eGFR is an accepted index of renal function, regardless of the formula used for the calculation. The Cockcroft-Gault equation is, at present, the most widely used in clinical practice and was employed in phase III clinical trials with direct oral anticoagulants to calculate eGFR and adjust the dose of the oral anti-Xa agents. For this reason, the Cockcroft-Gault equation is commonly used to adjust dosing of direct oral anticoagulants in patients with atrial fibrillation. However, as this formula may overestimate eGFR, the CKD-EPI equation was developed to more accurately estimate eGFR across all ranges of renal function and was tested in sub-analyses of studies with direct oral anticoagulants.35 Cockcroft–Gault, CKD-EPI and MDRD are all serum creatinine-based estimations of eGFR and depend on age and gender. Differences across equations emerged in our study between the Cockcroft-Gault and CKD-EPI equations concerning proportions of patients classified with stage I or stage II RI. The majority of reclassifications occurred in the group estimated as having normal renal function according to the Cockcroft-Gault equation. However, incidences of study outcome events in our study were similar in stages I and II RI, as calculated by different formulas. Our study has some limits. In particular, 13 patients included in the Caravaggio study were excluded from this analysis due to the lack of baseline creatinine value at the time of inclusion or to the violation of the inclusion requirement necessitating creatinine clearance greater than 30 mL/min. The study was designed as an open-label randomized study. Investigators were aware of study treatment assignment and of creatinine clearance values over time. The management of study treatments according to creatinine clearance was dictated by the protocol, demonstrating that accurate management of anticoagulant treatment may avoid complications due to deterioration of renal function. Finally, our analysis cannot provide information on the efficacy and safety of apixaban and

Haematologica | 107 July 2022

1574

ARTICLE - Apixaban and renal function in cancer-associated VTE

C. Becattini et al.

dalteparin in patients with severe RI (creatinine clearance lower than 30 mL/min), as these patients were excluded from the Caravaggio study. Our study benefits from its randomized design with prospective assessments of renal function in a large population of cancer patients and the blind adjudication of all the study outcome events by an independent committee. In conclusion, this analysis shows that a substantial proportion of cancer patients have RI when diagnosed with VTE, or experience a deterioration of renal function in the 6-month period beyond the index VTE. However, no effect of renal function at the initiation of anticoagulant treatment or of variation of renal function during anticoagulant treatment is associated with the risk for major bleeding or recurrent VTE in patients randomized to receive apixaban or dalteparin. Recurrent or locally-advanced or metastatic cancer was shown to be the only independent predictor of major bleeding in cancer patients with VTE. Clinically relevant non-major bleeding seems to be more common in patients with moderate RI receiving apixaban in comparison to dalteparin.

personal fees and non-financial support from Aspen, outside the submitted work. AC reports personal fees from AbbVie, Bayer AG and Boehringer Ingelheim, grants and personal fees from Bristol-Myers Squibb, and Daiichi-Sankyo Europe, personal fees from Exom and Janssen, grants and personal fees from Pfizer Limited and personal fees from Alexion Pharmaceuticals, outside the submitted work. JMC reports personal fees from Abbott, Bristol-Myers Squibb, Pfizer, Takeda, and Portola, outside the submitted work. DM has nothing to disclose. AS has nothing to disclose. Dr. Brenner reports personal fees from Bayer, ROVI Pharmaceuticals, Sanofi, and Leo Pharma, outside the submitted work. GA reports personal fees from Bristol-Myers-Squibb, Pfizer, and Bayer Healthcare, outside the submitted work.

Disclosures CB reports personal fees from Bayer HealthCare, Bristol Myers Squibb, and Daiichi Sankyo, outside the submitted work. RB reports personal fees from Bristol-Myers-Squibb, Pfizer, Bayer Healthcare, and LEO, outside the submitted work. GM has nothing to disclose. LB reports grants, personal fees and non-financial support from Bristol-Myers Squibb/Pfizer, grants and personal fees from Bayer, personal fees and non-financial support from Leo-Pharma,

Funding The Caravaggio study was funded by an unrestricted grant from the Bristol-Myers Squibb–Pfizer Alliance.

Contributions CB, GM, RB and GA were responsible for the conception and design of the study or analysis and interpretation of data, or both. All authors contributed to drafting the manuscript or revising it critically for important intellectual content and provided final approval of the manuscript submitted.

Data-sharing statement Data collected for the study may be shared after approval of potential proposals by the Steering Committee and with a signed data-access agreement, and at least 6 months after manuscript publication.

References 1.Prandoni P, Lensing AWA, Piccioli A, et al. Recurrent venous thromboembolism and bleeding complications during anticoagulant treatment in patients with cancer and venous thrombosis. Blood. 2002;100(10):3484-3488. 2.Weitz JI, Haas S, Ageno W, et al. Cancer associated thrombosis in everyday practice: perspectives from GARFIELD-VTE. J Thromb Thrombolysis. 2020;50(2):267-277. 3.Lee AYY, Kamphuisen PW, Meyer G, et al. Tinzaparin vs warfarin for treatment of acute venous thromboembolism in patients with active cancer. JAMA. 2015;314(7):677-686. 4.Lee AYY, Levine MN, Baker RI, et al. Low-molecular-weight heparin versus a coumarin for the prevention of recurrent venous thromboembolism in patients with cancer. N Engl J Med 2003;349(2):146-153. 5.Raskob GE, van Es N, Verhamme P, et al. Edoxaban for the treatment of cancer-associated venous thromboembolism. N Engl J Med. 2018;378(7):615-624. 6.Young AM, Marshall A, Thirlwall J, et al. Comparison of an oral factor Xa inhibitor with low molecular weight heparin in patients with cancer with venous thromboembolism: results of a randomized trial (SELECT-D). J Clin Oncol. 2018;36(20):2017-2023. 7.McBane RD, Wysokinski WE, Le-Rademacher J, et al. Apixaban,

dalteparin, in active cancer associated venous thromboembolism, the ADAM VTE Trial. Blood. 2018;132(Suppl 1):421. 8.Agnelli G, Becattini C, Meyer G, et al. Apixaban for the treatment of venous thromboembolism associated with cancer. N Engl J Med. 2020;382(17):1599-1607. 9.Key NS, Khorana AA, Kuderer NM, et al. Venous thromboembolism prophylaxis and treatment in patients with cancer: ASCO clinical practice guideline update. J Clin Oncol. 2020;38(5):496-520. 10.Konstantinides SV, Meyer G, Becattini C, et al. 2019 ESC guidelines for the diagnosis and management of acute pulmonary embolism developed in collaboration with the European Respiratory Society (ERS). Eur Heart J. 2020;41(4):543-603. 11.Lyman GH, Carrier M, Ay C, et al. American Society of Hematology 2021 guidelines for management of venous thromboembolism: prevention and treatment in patients with cancer. Blood Adv. 2021;5(4):927-974. 12.Goto S, Haas S, Ageno W, et al. GARFIELD-VTE Investigators. Assessment of outcomes among patients with venous thromboembolism with and without chronic kidney disease.

Haematologica | 107 July 2022

1575

ARTICLE - Total metabolic tumor volume predicts survival JAMA Netw open. 2020;3(10):e2022886. 13.Angelini DE, Radivoyevitch T, McCrae KR, Khorana AA. Bleeding incidence and risk factors among cancer patients treated with anticoagulation. Am J Hematol. 2019;94(7):780-785. 14.Trujillo-Santos J, Nieto JA, Tiberio G, et al. RIETE Registry. Predicting recurrences or major bleeding in cancer patients with venous thromboembolism. Findings from the RIETE Registry. Thromb Haemost. 2008;100(3):435-439. 15.Bauersachs R, Lee AYY, Kamphuisen PW, et al. CATCH Investigators. Renal impairment, recurrent venous thromboembolism and bleeding in cancer patients with acute venous thromboembolism-analysis of the CATCH study. Thromb Haemost. 2018;118(5):914-921. 16.Woodruff S, Feugère G, Abreu P, Heissler J, Ruiz MT, Jen F. A post hoc analysis of dalteparin versus oral anticoagulant (VKA) therapy for the prevention of recurrent venous thromboembolism (rVTE) in patients with cancer and renal impairment. J Thromb Thrombolysis. 2016;42(4):494-504. 17.Bavalia R, Middeldorp S, Weisser G, Espinola-Klein C. Treatment of venous thromboembolism in special populations with direct oral anticoagulants. Thromb Haemost. 2020;120(6):899-911. 18.van Es N, Coppens M, Schulman S, Middeldorp S, Büller HR. Direct oral anticoagulants compared with vitamin K antagonists for acute venous thromboembolism: evidence from phase 3 trials. Blood. 2014;124(12):1968-1975. 19.Agnelli G, Becattini C, Bauersachs R, et al. Apixaban versus Dalteparin for the treatment of acute venous thromboembolism in patients with cancer: the Caravaggio study. Thromb Haemost. 2018;118(9):1668-1678. 20.Inker LA, Astor BC, Fox CH, et al. KDOQI US commentary on the 2012 KDIGO clinical practice guideline for the evaluation and management of CKD. Am J Kidney Dis. 2014;63(5):713-735. 21.Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum creatinine. Nephron. 1976;16:31-41. 22.Levey AS, Stevens LA, Schmid CH, et al. CKD-EPI (Chronic Kidney Disease Epidemiology Collaboration). A new equation to estimate glomerular filtration rate. Ann Intern Med. 2009;150(9):604-612. 23.Levey AS, Coresh J, Greene T, et al. Chronic Kidney Disease Epidemiology Collaboration. Using standardized serum creatinine values in the Modification of Diet in Renal Disease study equation for estimating glomerular filtration rate. Ann Intern Med. 2006;145(4):247-254. 24.Asar Ö, Ritchie J, Kalra PA, Diggle PJ. Joint modelling of repeated measurement and time-to-event data: an introductory tutorial. Int J Epidemiol. 2015;44(1):334-344.

C. Becattini et al. 25.Janus N, Launay-Vacher V, Byloos E, et al. Cancer and renal insufficiency results of the BIRMA study. Br J Cancer. 2010;103(12):1815-1821. 26.Prins MH, Lensing AW, Brighton TA, et al. Oral rivaroxaban versus enoxaparin with vitamin K antagonist for the treatment of symptomatic venous thromboembolism in patients with cancer (EINSTEIN-DVT and EINSTEIN-PE): a pooled subgroup analysis of two randomised controlled trials. Lancet Haematol. 2014;1(01):e37-e46. 27.Agnelli G, Buller HR, Cohen A, et al. Oral apixaban for the treatment of venous thromboembolism in cancer patients: results from the AMPLIFY trial. J Thromb Haemost. 2015;13(12):2187-2191. 28.Farge D, Frere C, Connors JM, et al. International Initiative on Thrombosis and Cancer (ITAC) advisory panel. 2019 international clinical practice guidelines for the treatment and prophylaxis of venous thromboembolism in patients with cancer. Lancet Oncol. 2019;20(10):e566-e581. 29.Kearon C, Akl EA, Ornelas J, et al. Antithrombotic therapy for VTE disease: CHEST guideline and expert panel report. Chest. 2016;149(2):315-352. 30.Atiq F, van den Bemt PM, Leebeek FW, van Gelder T, Versmissen J. A systematic review on the accumulation of prophylactic dosages of low-molecular-weight heparins (LMWHs) in patients with renal insufficiency. Eur J Clin Pharmacol. 2015;71(8):921-929. 31.Park D, Southern W, Calvo M, et al. Treatment with dalteparin is associated with a lower risk of bleeding compared to treatment with unfractionated heparin in patients with renal insufficiency. J Gen Intern Med. 2016;31(2):182-187. 32.Verso M, Munoz A, Bauersachs R, et al. Effects of concomitant administration of anticancer agents and apixaban or dalteparin on recurrence and bleeding in patients with cancer-associated venous thromboembolism. Eur J Cancer. 2021;148:371-381. 33.Ortel TL, Neumann I, Ageno W, et al. ASH-2020 guidelines for management of venous thromboembolism: treatment of deep vein thrombosis and pulmonary embolism. Blood Adv. 2020;4(19):4693-4738. 34.McCormack T, Harrisingh MC, Horner D, Bewley S, Guideline Committee. Venous thromboembolism in adults: summary of updated NICE guidance on diagnosis, management, and thrombophilia testing. BMJ. 2020;369:m1565. 35.Hohnloser SH, Hijazi Z, Thomas L, et al. Efficacy of apixaban when compared with warfarin in relation to renal function in patients with atrial fibrillation: insights from the ARISTOTLE trial. Eur Heart J. 2012;33(22):2821-2830.

Haematologica | 107 July 2022

1576

ARTICLE - Iron Metabolism & its Disorders

Genetic modifiers of fetal hemoglobin affect the course of sickle cell disease in patients treated with hydroxyurea Pierre Allard,1 Nareen Alhaj,2 Stephan Lobitz,3,4 Holger Cario,4,5 Andreas Jarisch,4,6 Regine Grosse,4,7 Lena Oevermann,4,8 Dani Hakimeh,4,8 Laura Tagliaferri,1,4 Elisabeth Kohne,5 Annette Kopp-Schneider,2 Andreas E. Kulozik1,4 and Joachim B. Kunz1,4 for the German Sickle Cell Disease Study Group.§ 1

Department of Pediatric Oncology, Hematology and Immunology, Hopp-Children’s Cancer Center (KiTZ) Heidelberg, University of Heidelberg, Heidelberg; 2Abteilung Biostatistik, Deutsches Krebsforschungszentrum (DKFZ), Heidelberg; 3Gemeinschaftsklinikum Mittelrhein, Kemperhof, Pädiatrische Hämatologie und Onkologie, Koblenz; 4GPOH Konsortium Sichelzellkrankheit; 5Universitätsklinikum Ulm, Klinik für Kinder- und Jugendmedizin, Pädiatrische Hämatologie und Onkologie, Ulm; 6Klinikum der Johann-Wolfgang-GoetheUniversität, Zentrum für Kinder- und Jugendmedizin, Schwerpunkt Stammzelltransplantation und Immunologie, Frankfurt am Main; 7Universitätsklinikum Hamburg-Eppendorf, Zentrum für Geburtshilfe, Kinder- und Jugendmedizin, Klinik und Poliklinik für Pädiatrische Hämatologie und Onkologie, Hamburg and 8Charité - Universitätsmedizin Berlin, Campus VirchowKlinikum, Klinik für Pädiatrie m.S. Onkologie/Hämatologie/KMT, Berlin, Germany

Correspondence: Joachim Kunz joachim.kunz@med.uni-heidelberg.de Received: April 11, 2021. Accepted: July 29, 2021. Prepublished: October 28, 2021. https://doi.org/10.3324/haematol.2021.278952 ©2022 Ferrata Storti Foundation Haematologica material is published under a CC BY-NC license

A list of the members of the Sickle Cell Disease Study Group appears in the Appendix

Abstract The course of sickle cell disease (SCD) is modified by polymorphisms boosting fetal hemoglobin (HbF) synthesis. However, it has remained an open question how these polymorphisms affect patients who are treated with the HbF-inducing drug hydroxyurea/ hydroxycarbamide. The German SCD registry offers the opportunity to answer this question, because >90% of patients are treated according to national guidelines recommending the use of hydroxyurea in all patients above 2 years of age. We analyzed the modifying effect of HbF-related genetic polymorphisms in 417 patients with homozygous SCD >2 years old who received hydroxyurea. HbF levels were correlated with higher total hemoglobin levels, lower rates of hemolysis, a lower frequency of painful crises and of red blood cell transfusions. The minor alleles of the polymorphisms in the γ-globin promoter (rs7482144), BCL11A (rs1427407) and HMIP (rs66650371) were strongly associated with increased HbF levels. However, these associations did not translate into lower frequencies of vaso-occlusive events which did not differ between patients either carrying or not carrying the HMIP and BCL11A polymorphisms. Patients on hydroxyurea carrying the γ-globin promoter polymorphism demonstrated substantially higher hemoglobin levels (P<10-4) but also higher frequencies of painful crises and hospitalizations (P<0.01) when compared to patients without this polymorphism. Taken together, these data indicate that the γ-globin, HMIP and BCL11A polymorphisms correlate with increased HbF in SCD patients on hydroxyurea. While HbF is negatively correlated with the frequency of painful crises and hospitalizations, this was not observed for the presence of known HbF-boosting alleles.

Introduction Sickle cell disease (SCD) is a multiorgan disorder with a broad spectrum of clinical presentations. While some patients reach adulthood with few symptoms, others die early from complications such as acute anemia, infection or acute chest syndrome. Besides the β-globin and α-globin genotypes, the most important modifier of SCD is the persisting expression of fetal hemoglobin (HbF), a heritable quantitative trait determined mainly by the three loci BCL11A, HMIP and HBG.1 While α-thalassemia, if co-inherited with the HbS mutation, slows down HbS polymerization by reducing the cellular hemoglobin concentration,2 persisting HbF can interfere with the polymerization of

HbS.3 Genetic modifiers of HbF synthesis have generally shown a beneficial effect on the phenotype of SCD.1,2,4,5 However, some studies have yielded conflicting results and most exclusively included patients who did not receive the disease-modifying drug hydroxyurea.5 Several studies analyzing the effects of α-thalassemia and HbF-modifiers in patients treated with hydroxyurea suggested that the effect of hydroxyurea on laboratory and clinical outcomes largely supersedes the effect of the genetic modifiers.6-13 Both the effects of polymorphisms identified by genomewide association studies to modify the phenotype of SCD and the pharmacological effects of hydroxyurea were originally considered to be mediated via the expression and distribution of HbF.14,15 Genetic modifiers of HbF synthesis

Haematologica | 107 July 2022

1577

ARTICLE - Modifiers of SCD treated with hydroxyurea

P. Allard et al.

generally interfere with the expression or the binding activity of BCL11A. While this transcription factor is required for the perinatal “hemoglobin switch”, its actions are not limited to the β-globin locus,16 leaving room for effects on red blood cells and on the phenotype of SCD that are independent from HbF expression. Similarly, hydroxyurea does not exclusively target HbF expression but in addition exerts effects such as myelosuppression and nitric oxide release that may be equally important in ameliorating SCD.17 Considering the pleiotropic effects of both genetic modifiers and pharmacotherapy, the association of genetic modifiers with a certain phenotype may vary with the treatment that is applied. Thus, the actual importance of genetic modifiers for the clinical course of SCD needs to be assessed in the context of pharmacological treatment in addition to previous studies in treatment-naïve patients. In Germany, treatment guidelines18 encourage the use of hydroxyurea in all symptomatic patients with SCD, starting at the age of 2 years. Therefore, approximately 90% of patients with homozygous SCD registered in the German SCD registry are prescribed hydroxyurea.19 We made use of laboratory and clinical data collected in this registry to identify the effect of genetic modifiers on the phenotype of homozygous SCD. To this end, we selected polymorphisms known to modify either HbF expression or the phenotype of SCD, to be independent from each other and to be functionally relevant, for instance by altering transcription factor binding.15,20-23 We focused on the polymorphisms rs1427407 and rs7606173 in BCL11A, on the 3 bp deletion delCTA rs66650371 in HMIP and on the XmnI polymorphism in the γ-globin promoter, rs7482144. The allele BCL11A rs1427407 T was associated with increased HbF in genomewide association studies and was shown to reduce binding of GATA1 and TAL1 to a DNase-hypersensitive site that regulates BCL11A expression.20 After conditioning for the association with rs1427407, rs7606173 G is the BCL11A allele that remains most significantly associated with high HbF expression. It is located in a DNase hypersensitive site 7 kb closer to the BCL11A transcription start site than rs1427407.20 The deletion of three base pairs, CTA, in HMIP rs66650371 is considered to cause the strong association of the HMIP locus on chr6 with HbF levels by augmenting an enhancer-like activity located between the HBS1L and the MYB genes.21 The polymorphism in the γ-globin promoter that creates an XmnI restriction site, HGB2 rs7482144 A, has long been a candidate for being the HbFboosting sequence within the high-HbF β-globin haplotypes (‘Senegal’ and ‘Arab-Indian’).24,25 Together, polymorphisms in BCL11A, HMIP and the γ-globin promoter were estimated to explain approximately 22% of the variability in HbF expression.26 In addition, we analyzed the coinheritance of the α-thalassemia trait that has been shown in multiple studies to be associated with a reduced risk of

cerebrovascular complications, but an increased risk of painful crises.5 We restricted our analyses to patients at least 2 years of age with homozygous SCD who are treated with hydroxyurea, independently of the dose used.

Methods Patients’ recruitment and data collection Patients were recruited through the nationwide German SCD registry (NCT03327428) which collects prospective and retrospective data on patients with SCD in Germany. The study was performed according to the Declaration of Helsinki and approved by the institutional review board of the Medical Faculty of Heidelberg University (S 416/2014). Written informed consent was obtained from patients or legal guardians. The data collected included demographic information, diagnosis and genotype, treatment, laboratory parameters and clinical events. At the time of the data cutoff, May 13, 2020, 425 patients with homozygous SCD from 28 different institutions were enrolled in the registry. Data were analyzed by PA, JK, NA and AK-S. All co-authors had access to all registry data. Treatment guidelines implemented in 2014 recommend parental education, the use of penicillin prophylaxis at least until the age of 5 years, and annual screening with transcranial Doppler ultrasound starting from 2 until 18 years of age. The use of hydroxyurea is encouraged in all patients with SCD who have ever experienced a painful vaso-occlusive crisis, including mild ones. The recommended starting dose of hydroxyurea is 15 mg/kg/day for adults and 20 mg/kg/day for children. In the case of insufficient efficacy, a dose escalation up to the maximum tolerated dose or to 35 mg/kg/day is recommended. For the analysis of the frequency and distribution of genetic modifiers, patients with homozygous SCD of all ages and irrespective of treatment were included. Patients with compound heterozygous SCD (HbSC, HbS/β-thalassemia, others) were excluded. For the analysis of laboratory parameters and complications of SCD, only patients at least 2 years of age and with ongoing treatment with hydroxyurea were considered. Patients’ data collected after stem cell transplantation were excluded. Data on complications and treatment of SCD were documented annually, together with routine laboratory parameters (hemoglobin, mean corpuscular volume [MCV], reticulocytes, lactate dehydrogenase [LDH], bilirubin and HbF) (Table 1). The laboratory parameters were only considered if the patient had not received red blood cell transfusions within 100 days before the assessment. All laboratory parameters were determined while the patient was on hydroxyurea and followed for clinical complications. If more than one laboratory data point fulfilled these criteria, we used the average of all available data.

Haematologica | 107 July 2022

1578

ARTICLE - Modifiers of SCD treated with hydroxyurea

P. Allard et al.

Patients registered by May 13, 2020 were included in the analysis. For the correlation of HbF with laboratory parameters and clinical course, all patients with available data were analyzed (n=193 to n=202) (Figures 1 and 2). All patients for whom the complete set of genetic traits of interest (αthalassemia, HBG2-polymorphism rs7482144, BCL11A polymorphisms rs1427407 and rs7606173, HMIP polymorphism rs66650371) was available in combination with the respective laboratory parameter (n=121 for HbF to n=164 for MCV) (Table 2) were included in the analyses that correlated genetics with laboratory parameters and clinical course. In order to investigate geographic and ethnic differences in the phenotypic expression of SCD, we categorized patients according to the origins of their parents from one of the three regions Mediterranean Sea, Sub-Saharan Africa and “rest of the world”. The last included mainly patients from Iraq (n=13), all other countries contributed at most two patients. Pain crises were defined as pain requiring pharmacological treatment and hospitalization with no other obvious cause besides SCD. This definition does not include visits in the emergency department that did not result in hospital admission. Clinical events were considered from the first dose of hydroxyurea until last follow-up (median/mean observation period per patient 1.8/1.1 years; range, 0.5-5.3 years), independently of changes in the hydroxyurea dose. In 313 patients with at least 1 year of follow-up on hydroxyurea, the severity of SCD was graded as “severe” (n=83) if 1.5 or more pain crises requiring hospitalization per year or 0.5 or more episodes of acute chest syndrome (ACS) per year were documented. In addition, any stroke, sepsis, chronic pain or need of chronic red blood cell transfusions occurring on hydroxyurea defined a severe course. Genetic analysis The HBG2-polymorphism rs7482144 was identified by polymerase chain reaction (PCR) amplification (forward primer: ATA GCA CTT CTT ATT TGG AAA CCA A, reverse primer: TGT CTA AGT TGC CTC GAG ACT AAA G), XmnI digestion and restriction fragment length analysis.22 The BCL11A polymorphisms (rs1427407 and rs7606173) were analyzed using sequence-specific TaqMan genotyping.20 The 3 bp deletion in HMIP (rs66650371) was diagnosed by sequence-specific PCR and agarose gel electrophoresis.21 α-thalassemia deletions (-α3.7, -α4.2, -α20.5, --SEA and --MED) were detected by PCR and subsequent agarose gel electrophoresis.27 Statistical analysis To evaluate the effect of HbF levels on laboratory parameters we used linear model analysis. To obtain approximately normally distributed variables, the values of LDH, bilirubin and reticulocytes were log-transformed. The as-

sociation of HbF and clinical course was analyzed using a Poisson regression. The impact of the combination of single nucleotide polymorphisms on HbF and on the laboratory parameters was analyzed by linear multivariable regression analysis. The presence of a single nucleotide polymoprhism-variant was coded with 0 (no polymorphism - wildtype), 1 (heterozygous) and 2 (homozygous) for each studied polymorphism. A Poisson regression was used to estimate the contribution of each polymorphism on the clinical course. All statistical analyses were conducted using R version 4.0.2 (The R Foundation for Statistical Computing 2020).

Results Patients’ characteristics We restricted our analysis to patients at least 2 years of age on hydroxyurea, corresponding to 77.6% (n=330) of all patients with homozygous SCD in the registry. Of all patients with homozygous SCD (n=425), 19 were excluded because of young age (<2 years), 36 were excluded because they were not on hydroxyurea, 16 were excluded from the analysis of laboratory parameters because they had received red blood cell transfusions within 100 days before blood sampling, three were excluded because no data prior to allogeneic stem cell transplantation were documented and 13 did not have complete data on clinical complications. The age cut-off was set at 2 years because hydroxyurea is licensed for use in Europe starting at this age and because HbF levels were not correlated with age and sex in this group of patients (Online Supplementary Figures S1 and S2). The mean daily hydroxyurea dose was 23.2 mg/kg (standard deviation 5.9; range, 7.7-39.0). Mean HbF levels among these patients were higher than those reported in hydroxyurea-naïve patients of West African origin28,29 and comparable to those in hydroxyureanaïve patients from India.30 Most patients had initiated hydroxyurea treatment before they were enrolled in the registry, but in 25 patients paired HbF measurements from before and after initiation of hydroxyurea treatment were available. During hydroxyurea treatment, HbF was on average 1.8-fold higher than before (mean HbF 23.7% vs. 13.2%, P<10-4 (t-test) (Online Supplementary Figure S3). While α-thalassemia deletions and polymorphisms in BCL11A were present in at least a third of all patients, the HBG2-polymorphism rs7482144 and the polymorphism rs66650371 in HMIP affected only a minority of 10 to 15% of patients (Table 1). These less frequent traits were exceedingly rare among patients originating from sub-Saharan Africa but enriched in patients originating from the Mediterranean region (28% allele frequency for HMIP rs66650371) and from the rest of the world (47% allele frequency for the HBG2-polymorphism rs7482144) (Online

Haematologica | 107 July 2022

1579

ARTICLE - Modifiers of SCD treated with hydroxyurea

P. Allard et al.

Table 1. Patients’ characteristics.

Variables Sex Female Male Age at last observation under treatment (years) Hematologic parameters in patients on hydroxyurea HbF (%) Hb (g/dL) MCV (fL) Reticulocytes (‰) Bilirubin (mg/dL) LDH (U/L) Region of origin Sub-Saharan Africa Mediterranean Sea Rest of the world* α-globin-deletions** 0/1/2 BCL11A rs1427407 GG / GT / TT BCL11A rs7606173 CC / CG / GG HBG2-polymorphism rs7482144 WT / het. / hom. HMIP rs66650371 3 bp del WT / het. / hom. Clinical events while on hydroxyurea*** Pain crises /year ACS /year RBC transfusions/year Hospitalizations/year Severity of SCD while on hydroxyurea Not severe/severe

N of patients

Mean ± SD

Range

330

12.6±8.5

2.3-54.1

208 300 300 279 281 280

17.9±9.5 9.0±1.5 92.6±13.6 75.3±44.1 2.1±1.5 509.2±189.4

0-46.7 5-14.3 60.6-134.7 12.7-360 0.2-12.5 167-1545

0.7±1.2 0.1±0.4 0.4±1.1 0.9±1.4

0-10 0-3 0-10 0-11

208 209

285 62 25 243 / 107 / 23 135 / 88 / 15 47 / 117 / 74 207 / 14 / 24 211/ 23 / 6 299 300 298 298

230 / 83

n refers to the number of evaluable patients for each parameter. The data sets of some patients were incomplete, either because of incomplete documentation or because laboratory parameters could not be analyzed as the patient had received a red blood cell transfusion within 100 days before blood sampling. SD: standard deviation; MCV: mean corpuscular volume; LDH: lactate dehydrogenase; WT: wild-type; het.: heterozygous; hom: homozygous; ACS: acute chest syndrome; RBC: red blood cells; SCD: sickle cell disease, * Iraq n=13, all others n≤2. ** No patients with homozygous SCD and deletion of three α-globin genes were registered. *** Clinical events were only evaluated in patients with at least 6 months of follow up while on hydroxyurea, age >2 years.

Supplementary Table S1). Of note, the minor allele frequencies of these polymorphisms were paralleled by higher levels of HbF and total hemoglobin in patients not originating from sub-Saharan Africa (Online Supplementary Table S2). As expected for polymorphisms that are differentially enriched in specific ethnicities and, as is the case for the HBG2-polymorphism rs7482144, are linked to the HbS mutation, the frequency of homozygous carriers of the minor allele significantly (P<10-4) exceeded that predicted by the Hardy-Weinberg equilibrium. In contrast, αthalassemia deletions were detected significantly more frequently among patients originating from sub-Saharan Africa in comparison to all other patients (allele frequency 25.2% vs. 7.1%, P<10-4). Compatible with a linkage disequilibrium, both polymorphisms in BCL11A were signifi-

cantly associated with each other (P<10-4). As expected for independently inherited traits, we did not identify any significant associations between polymorphisms in BCL11A and in HMIP or HBG2. However, the polymorphisms that occurred preferentially in the non-sub-Saharan patients, HMIP and HGB2, were significantly associated with each other (P=0.008, c2 test). High HbF levels are associated with a milder phenotype of sickle cell disease All hematologic parameters tested were strongly associated with HbF levels (Figure 1). While patients with an HbF of 5% had a mean total hemoglobin of 7.9 g/dL (MCV 85 fL), patients with an HbF of 25% had a mean total hemoglobin of 9.4 g/dL (MCV 94 fL). Indicators of hemolysis

Haematologica | 107 July 2022

1580

ARTICLE - Modifiers of SCD treated with hydroxyurea

P. Allard et al.

Figure 1. Linear regression of HbF on laboratory parameters. (A) Linear regression of fetal hemoglobin (HbF) on total hemoglobin (n=202). (B) Linear regression of HbF on mean corpuscular volume (n=202). (C) Linear regression of HbF on ln lactate dehydrogenase (n=194). (D) Linear regression of HbF on ln bilirubin (n=195). (E) Linear regression of HbF on ln reticulocyte count (n=195). Hb: total hemoglobin; MCV; mean corpuscular volume; LDH: lactate dehydrogenase.

showed an inverse relation (patients with HbF 5% vs. 25%: LDH 560 vs. 441 U/L, bilirubin 2.34 vs. 1.51 mg/dL, reticulocytes 87‰ vs. 53 ‰) (Figure 1). Simultaneously, the frequency of pain crises and of red blood cell transfusions decreased with increasing HbF (Figure 2). The frequency of hospitalizations paralleled that of pain crises, reflecting our definition of “pain crisis” as requiring hospitalization and the fact that most admissions in patients with SCD are due to acute pain. In contrast, ACS was not correlated with HbF levels, indicating a differential effect of HbF on these complications.

In order to analyze whether adherence to hydroxyurea - reflected by increased MCV, but also HbF and total hemoglobin - is associated with the phenotype, we compared laboratory and clinical parameters in patients with severe SCD to those in patients with non-severe SCD (Online Supplementary Table S3). While severely affected patients were slightly older than non-severely affected patients and carried more HbF-modifying alleles, they did not differ in any other laboratory parameter, most importantly MCV, indicating that potential differences in adherence to hydroxyurea were too small to be detected.

Haematologica | 107 July 2022

1581

ARTICLE - Modifiers of SCD treated with hydroxyurea

P. Allard et al.

Table 2. Multivariable analysis of the effect of genetic traits on laboratory parameters and complications of sickle cell disease.

Coefficients (95% confidence intervals) of a linear regression model are reported for endpoints fetal hemoglobin, total hemoglobin, mean corpuscular volume, lactate dehydrogenase, bilirubin, and reticulocyte count. Positive values indicate a positive correlation, negative values indicate a negative correlation. Exponentiated coefficients (95% confidence intervals) of Poisson regression are reported for the endpoints hospitalization, pain crises, acute chest syndrome, and transfusions. Values greater than 1 indicate a positive correlation, values smaller than 1 indicate a negative correlation. Blue: positive correlation; red: negative correlation; dark blue/dark red1 P<0.01; blue/red2 P<0.05; light blue/ light red3 P<0.1; all other P>0.1; n: number of patients available for analysis HbF: fetal hemoglobin, Hb: total hemoglobin; MCV: mean corpuscular volume; LDH: lactate dehydrogenase; ACS: acute chest syndrome.

Effects of genetic modifiers can be either mediated by HbF or independent of HbF In order to assess whether the effects of genetic modifiers on SCD are mediated by boosting HbF expression, we next compared the correlation of genetic traits with HbF levels, with laboratory parameters and with complications of SCD (Table 2, Online Supplementary Figures S4-S11, Online Supplementary Tables S4-S7). As expected, the co-inheritance of α-thalassemia resulted in lower MCV and reduced rates of hemolysis without significantly affecting HbF or total hemoglobin. However, the α-thalassemia trait was not linked to the frequency of complications in these patients. The rarer alleles of the polymorphic loci BCL11A rs1427407, HMIP rs66650371 and HBG2-polymorphism rs7482144 were significantly associated with increased HbF levels. In addition, the BCL11A rs1427407 and HBG2 rs7482144 polymor-

phisms were associated with increased total hemoglobin concentrations. Consistent with this observation, patients carrying the HBG2-polymorphism rs7482144 received red blood cell transfusions less frequently than others. However, neither the increases in HbF nor those in total hemoglobin were reflected in a reduced frequency of complications of SCD. In contrast, pain crises and hospitalization were strikingly and highly significantly (P<0.01) more frequent in patients carrying the HBG2-polymorphism rs7482144 than in others (Table 2, Figure 3C, D). The coinheritance of two or more HbF-boosting alleles appears to be associated with an additive effect on HbF levels (Online Supplementary Figure S12). While small numbers of patients precluded definitive statistical analyses combining patients’ origins and genotypes, the association of rs7482144 with increased HbF and frequent complications appears to be independent of the region of origin (Online Supplementary Figure S13).

Haematologica | 107 July 2022

1582

ARTICLE - Modifiers of SCD treated with hydroxyurea

P. Allard et al.

Figure 2. Correlation between HbF level and complications of sickle cell disease. Poisson regression. (A-C) Numbers of patients available for analysis: 193 for hospitalization (A), 194 for pain crises (B), and red blood cell transfusion (C). Thick lines represent the predicted values, thin lines represent the 95% confidence interval. Haematologica | 107 July 2022

1583

ARTICLE - Modifiers of SCD treated with hydroxyurea

P. Allard et al.

The association of genetic traits with HbF does not correlate with the association of genetic traits and the rate of hemolysis or the frequency of complications of SCD (Table 2). While decreased LDH levels in patients with BCL11A rs1427407 and HMIP rs66650371 were concordant with increased HbF, BCL11A rs7606173 was not correlated with HbF levels but with increased LDH. At the same time, the HBG2polymorphism rs7482144, which was clearly associated with high HbF, was not significantly associated with changes in any of the parameters that are informative of hemolysis. Without having any detectable effect on HbF levels, BCL11A rs7606173 showed a trend towards higher LDH and more frequent hospitalizations and ACS. Although there was no association between increased HbF and a reduced frequency of ACS, patients carrying BCL11A rs1427407 T showed a significant reduction in the frequency of ACS (Table 2), suggesting an HbF-independent modulation of the SCD phenotype. We conclude that the induction of HbF expression is only one of several effects that modify the clinical phenotype of SCD in patients treated with hydroxyurea. As BCL11A, a key player that is influenced by polymorphisms in HMIP

and BCL11A itself, is a pleiotropic transcription factor, its effects will not be limited to the regulation of HbF expression. In contrast, the HBG2-polymorphism rs7482144 is localized in the γ-globin promoter and shows a strong association with HbF but was not correlated with a milder course of SCD in this group of patients on hydroxyurea.

The HBG2-polymorphism rs7482144 defines a group of patients with high risk of painful crises As has been observed in several studies of patients not on hydroxyurea, patients on hydroxyurea and with the HBG2polymorphism rs7482144 (heterozygous or homozygous) had significantly higher mean HbF (23.6±8.6% vs. 17.5±10%, P=0.0061) (Figure 3A) and mean total hemoglobin (10.1±1.3 g/dL vs. 8.8±1.4 g/dL, P<10-4) (Figure 3B) when compared to all other patients. At the same time, MCV, reticulocyte counts, bilirubin and LDH levels did not differ between patients who did or did not carry at least one HBG2-polymorphism rs7482144, indicating that the rate of hemolysis is not reduced by the HBG2-polymorphism rs7482144. While among patients who did not carry the HBG2-polymorphism rs7482144 higher levels of HbF and total hemoglobin were

Figure 3. Mean values of laboratory parameters and complications comparing patients positive or negative for the γ-globin promoter polymorphism rs7482144. (A) Mean HbF ± standard deviation (SD): negative patients (n=105) versus positive patients (n=24): 17.5±10% versus 23.6±8.6%; t-test P=0.0061. (B) Mean hemoglobin ± SD: negative patients (n=143) versus positive patients (n=29): 8.9±1.5 g/dL versus 10.1±1.3 g/dL; t-test P<10-4. (C) Mean frequency of hospitalizations per year ± SD: negative patients (n=142) versus positive patients (n=27): 0.8±1.3 versus 1.9±2.6; t-test P=0.0011. (D) Mean frequency of pain crises per year ± SD: negative patients (n=143) versus positive patients (n=27): 0.6±1.1 versus 1.7±2.5; t-test P=0.0003. HbF: fetal hemoglobin; Hb: total hemoglobin. Haematologica | 107 July 2022

1584

ARTICLE - Modifiers of SCD treated with hydroxyurea

P. Allard et al.

significantly associated with fewer hospitalizations (Online Supplementary Tables S8 and S9), such a correlation was not detected among carriers of the HBG2-polymorphism rs7482144 (Online Supplementary Tables S10 and S11). In contrast, among patients carrying the HBG2-polymorphism rs7482144, there was a slight trend towards a higher total hemoglobin in those who suffered from severe disease (10.8±1.6 g/dL vs. 10.0±1.1 g/dL, P=0.21). The comparison of hydroxyurea doses prescribed to patients with or without the HBG2-polymorphism rs7482144 A confirmed that hydroxyurea doses were similar in both groups, indicating that the higher rate of complications in patients carrying rs7482144 A was not related to low hydroxyurea doses (Online Supplementary Table S7). In contrast, patients who carried either BCL11A rs1427407 T or HMIP rs66650371 delCTA were prescribed lower doses of hydroxyurea (Online Supplementary Tables S4 and S6), compatible with the notion that the hydroxyurea dose may be titrated according to HbF response.

Discussion The negative correlation between HbF levels and the frequency of pain crises and of red blood cell transfusions confirms that the HbF level is a favorable prognostic marker in patients with SCD who are on treatment with hydroxyurea. At the same time, the polymorphisms rs1427407 in BCL11A, rs66650371 in HMIP and rs7482144 in the γ-globin promoter were strongly correlated with HbF levels in patients on hydroxyurea, indicating that γ-globin induction by hydroxyurea did not override the effects of genetic modifiers on HbF levels in this cohort of patients. This observation contrasts with results from trial cohorts that did not identify a significant effect of the genetic modifiers analyzed here on HbF levels in patients who had been treated with hydroxyurea.6-8,31 A possible explanation for this discrepancy may be the younger age of patients in these studies in comparison to our registry patients. Despite the strong effect of HbF on the frequency of complications of SCD, the correlation of genetic modifiers with HbF did not translate into a reduced frequency of pain crises and hospitalizations in those patients who carry HbF-boosting alleles. In contrast, the presence of rs7482144 in the γ-globin promoter was associated with higher total hemoglobin but, unexpectedly, also with increased frequencies of pain crises and hospitalizations. This is in contrast to several earlier series of patients with SCD not treated with hydroxyurea who carry rs7482144 in the γ-globin promoter. These patients were characterized by increased HbF levels9,22,32,33 and, in contrast to our findings, also less frequent vaso-occlusive events.1,34-37 Studies on the association of rs7482144 with complications of SCD in patients treated with hydroxyurea are scarce.38 Consist-

ent with our observations, the results of the BABY HUG trial showed a trend towards a higher frequency of vaso-occlusive events in patients who carry rs7482144 if treated with hydroxyurea.6 We can only speculate on the reason for the discordant effect of rs7482144 on HbF and on the frequency of pain crises. Possibly the protective effect of HbF induction is counterbalanced by the increase in total hemoglobin that results in high blood viscosity and precipitates vasoocclusion. Such an effect would result in an optimal dose level for hydroxycarbamide below the frequently used maximum tolerated dose. Patients presenting with frequent pain episodes despite high HbF and high total hemoglobin may benefit from a transient reduction of blood viscosity by cautious phlebotomy. Another possible explanation why high HbF levels and also the HbF-boosting polymorphism BCL11A rs1427407 are associated with a lower frequency of vaso-occlusive complications but rs7482144 in contrast is associated with a higher frequency of vaso-occlusive complications may involve the distribution of HbF. If HbF were distributed in a heterocellular manner in patients carrying the minor allele of rs7482144, the protection against vaso-occlusive crises would be inferior to that in patients with elevated HbF that is distributed pancellularly.39 As we do not have any data on the cellular distribution of HbF in our patients, we were not able to test this hypothesis. The major limitation of our study is the selection of few genetic markers that, with the exception of α-thalassemia trait,2 focus on the expression of HbF,20-22 but do not consider other mechanisms that may modify the phenotype of SCD.40 As HbF-boosting polymorphisms are enriched in patients who do not originate from sub-Saharan Africa, we cannot fully discriminate between the effects of single polymorphisms, of the genetic background or even of social and behavioral factors. For future studies, a genome-wide characterization of patients may help in the identification of genetic traits causally related to the phenotype of SCD even in ethnically heterogeneous groups of patients. In addition, we observed a relatively short period of a maximum of 4 years in young patients, who were enrolled in a registry, not in a controlled clinical trial. Because the registry design does not intend source data verification, we do not expect the laboratory parameters and the data on clinical complications to be as complete as in a clinical trial. Similarly, we do not have direct data monitoring patients’ adherence to hydroxyurea treatment. However, HbF levels and MCV in these patients were consistently higher than those in hydroxyurea-naïve patients,29,41 compatible with a good adherence to treatment in the majority of patients. The γ-globin promoter polymorphism rs7482144 was strongly associated with both pain crises and a reduced need for transfusions, indicating that the same genetic trait may have divergent effects on different phenotypic aspects of SCD. This is one reason why general conclusions on the

Haematologica | 107 July 2022

1585

ARTICLE - Modifiers of SCD treated with hydroxyurea

P. Allard et al.

course of SCD in an individual patient cannot be drawn from the presence of certain genetic markers. Second, while the genetic signature specifies relative risks in the subgroups, the overlap between subgroups is too large to allow for the prediction of individual risks for complications of SCD. Third, in the present study we only considered acute complications of SCD in a young group of patients but cannot yet evaluate the impact of the polymorphisms on long-term sequelae such as chronic kidney failure or pulmonary hypertension. Therefore, the decision as to whether hydroxyurea or even curative treatment options such as allogeneic stem cell transplantation or gene therapy should be offered cannot be confidently based on the genetic profile analyzed here. We conclude that polymorphisms in BCL11A, HMIP and HGB2 increase HbF levels in patients with SCD on hydroxyurea. However, the impact of these polymorphisms on the complications of SCD treated with hydroxyurea was not observed to be correlated with their effect on HbF. Disclosures No conflicts of interest to disclose. Contributions PA collected and analyzed data and wrote the manuscript; NA and AK-S performed statistical analyses; SL, HC, AJ, RG, LO, DH, LT, and AEK designed the study; EK performed genetic analyses; JBK designed and supervised the study, analyzed data and wrote the manuscript; all investigators of the German Sickle Cell Disease Study Group contributed patients’ data. Acknowledgments The authors thank Margit Happich and Gabriele Tolle for excellent support with genetic analyses. Funding The German SCD registry is supported by the grants DKS 2016.12 and 2020.03 to JBK from the German Childhood Cancer Foundation. Data-sharing statement The protocol of the SCD registry is available in German language at www.sichelzellkrankheit.info. For original data, please contact joachim.kunz@med.uni-heidelberg.de. As the informed consent does not allow disclosure of individual patients’ data, only aggregated data can be provided upon request. Appendix: Members of the German Sickle Cell Disease Study Group (in alphabetical order) Ferras Alashkar, Klinik für Hämatologie, Universitätsklinikum Essen, Essen; Carman Aramayo Singelmann, University Hospital Essen, Pediatric Haematology and Oncology,

Essen; Stefan Balzer, Children's Hospital Amsterdamer Straße Cologne; Clinic for Children and Youth Medicine, Pediatric Oncology/Hematology, Cologne; Ines Brecht, Universitätsklinikum Tübingen, Klinik für Kinder- und Jugendmedizin, Tübingen; Bastian Brummel, Universitätsklinikum Münster, Klinik für Kinder- und Jugendmedizin – Pädiatrische Hämatologie und Onkologie, Münster; Carl Friedrich Classen, Kinder- und Jugendklinik, Universitätsmedizin Rostock, Rostock; Alexander Claviez, Klinik für Kinder- und Jugendmedizin I, Pädiatrische Onkologie, Hämatologie, Stammzelltransplantation, Universitätsklinikum SchleswigHolstein, Campus Kiel; Selim Corbacioglu, Department of Pediatric Hematology, Oncology and Stem Cell Transplantation, University Hospital of Regensburg, Regensburg; Dagmar Dilloo, Pädiatrische Hämatologie und Onkologie, Universitätskinderklinik Bonn, Bonn; Matthias Dürken, Department of Pediatric Hematology and Oncology, University of Mannheim, Mannheim; Wolfgang Eberl, Institute for Clinical Transfusion Medicine and Children's Hospital, Klinikum Braunschweig GmbH; Sabine Ebert, Abteilung für Pädiatrische Hämatologie und Onkologie, Universitätsklinikum Hamburg-Eppendorf, Hamburg; Miriam Erlacher, Department of Pediatrics and Adolescent Medicine, Division of Pediatric Hematology and Oncology, University Medical Center Freiburg, Faculty of Medicine, University of Freiburg, Freiburg; Gabriele Escherich, Abteilung für Pädiatrische Hämatologie und Onkologie, Universitätsklinikum Hamburg-Eppendorf, Hamburg; Michael Frühwald, University Children’s Hospital Augsburg, University Hospital Augsburg, Augsburg; Hermann Full, SLK-Kliniken Heilbronn GmbH, Heilbronn; Ute Groß-Wieltsch, Pediatrics 5 (Oncology, Hematology, Immunology), Center for Pediatric, Adolescent and Women's Medicine, Klinikum Stuttgart - Olgahospital, Stuttgart; Sabine Heine, Department of Pediatric Hematology and Oncology, Saarland University Hospital, Homburg/Saar; Marc Hömberg, Pädiatrische Onkologie und Hämatologie, Universitätsklinikum Köln, Köln; Johannes Holzapfel, University Children’s Hospital Augsburg, University Hospital Augsburg, Augsburg; Ursula Holzer, Universitätsklinikum Tübingen, Klinik für Kinder- und Jugendmedizin, Tübingen; Claudia Khurana, Klinik für Kinder- und Jugendmedizin, Evangelisches Klinikum Bethel, Bielefeld; Udo Kontny, Klinik für Kinder- und Jugendmedizin, Universitätsklinikum RWTH Aachen, Aachen; Markus Metzler, Pediatric Oncology and Hematology, Department of Pediatrics and Adolescent Medicine, University Hospital Erlangen, Erlangen; Michaela Nathrath, Zentrum für Frauen- und Kindermedizin, Klinik für Pädiatrische Hämatologie und Onkologie, Kassel; Anna Partheil, Medizinische Hochschule Hannover, Pädiatrische Hämatologie und Onkologie, Hannover; Claudia Maria Pothoff, Pädiatrische Onkologie und Hämatologie, Universitätsklinikum Köln, Köln; Aram Prokop, Helios Kliniken Schwerin, Schwerin; Harald Reinhard, Asklepios Kinderklinik Sankt Augustin, St. Augustin; Daniela

Haematologica | 107 July 2022

1586

ARTICLE - Modifiers of SCD treated with hydroxyurea

P. Allard et al.

Schenk, Department of Pediatrics and Children's Cancer Research Center, Kinderklinik München Schwabing, Klinikum rechts der Isar, Fakultät für Medizin, Technische Universität München, Munich; Dominik Schneider, Clinic of Pediatrics, Dortmund Municipal Hospital, Dortmund; Natascha Ströter,

Universitätsklinikum Gießen und Marburg, Standort Gießen, Zentrum für Kinder- und Jugendmedizin, Abteilung für pädiatrische Hämatologie und Onkologie, Gießen; Thomas Wiesel, Children's Hospital, Vestische Youth Hospital, University of Witten/Herdecke, Datteln, Germany.

References 1. Lettre G, Sankaran VG, Bezerra MA, et al. DNA polymorphisms at the BCL11A, HBS1L-MYB, and beta-globin loci associate with fetal hemoglobin levels and pain crises in sickle cell disease. Proc Natl Acad Sci U S A. 2008;105(33):11869-11874. 2. Higgs DR, Aldridge BE, Lamb J, et al. The interaction of alphathalassemia and homozygous sickle-cell disease. N Engl J Med. 1982;306(24):1441-1446. 3. Singer K, Singer L. Studies on abnormal hemoglobins : VIII. The gelling phenomenon of sickle cell hemoglobin: its biologic and diagnostic significance. Blood. 1953;8(11):1008-1023. 4. Lettre G. The search for genetic modifiers of disease severity in the beta-hemoglobinopathies. Cold Spring Harb Perspect Med. 2012;2(10):a015032 5. Meier ER, Fasano RM, Levett PR. A systematic review of the literature for severity predictors in children with sickle cell anemia. Blood Cells Mol Dis. 2017;65:86-94. 6. Sheehan VA, Luo Z, Flanagan JM, et al. Genetic modifiers of sickle cell anemia in the BABY HUG cohort: influence on laboratory and clinical phenotypes. Am J Hematol. 2013;88(7):571-576. 7. Aleluia MM, Santiago RP, da Guarda CC, et al. Genetic modulation of fetal hemoglobin in hydroxyurea-treated sickle cell anemia. Am J Hematol. 2017;92(5):E70-E72. 8. Sheehan VA, Crosby JR, Sabo A, et al. Whole exome sequencing identifies novel genes for fetal hemoglobin response to hydroxyurea in children with sickle cell anemia. PloS One. 2014;9(10):e110740. 9. Ware RE, Despotovic JM, Mortier NA, et al. Pharmacokinetics, pharmacodynamics, and pharmacogenetics of hydroxyurea treatment for children with sickle cell anemia. Blood. 2011;118(18):4985-4991. 10. Friedrisch JR, Sheehan V, Flanagan JM, et al. The role of BCL11A and HMIP-2 polymorphisms on endogenous and hydroxyurea induced levels of fetal hemoglobin in sickle cell anemia patients from southern Brazil. Blood Cells Mol Dis. 2016;62:32-37. 11. Adekile A, Menzel S, Gupta R, et al. Response to hydroxyurea among Kuwaiti patients with sickle cell disease and elevated baseline HbF levels. Am J Hematol. 2015;90 (7):E138-139. 12. Green NS, Barral S. Genetic modifiers of HbF and response to hydroxyurea in sickle cell disease. Pediatr Blood Cancer. 2011;56(2):177-181. 13. Green NS, Ender KL, Pashankar F, et al. Candidate sequence variants and fetal hemoglobin in children with sickle cell disease treated with hydroxyurea. PloS One. 2013;8(2):e55709. 14. Dover GJ, Charache S, Boyer SH. Increasing fetal hemoglobin in sickle cell disease: comparisons of 5-azacytidine (subcutaneous or oral) with hydroxyurea. Trans Assoc Am Physicians. 1984;97:140-145. 15. Menzel S, Thein SL. Genetic modifiers of fetal haemoglobin in sickle cell disease. Mol Diagn Ther. 2019;23(2):235-244. 16. Bauer DE, Orkin SH. Hemoglobin switching's surprise: the versatile transcription factor BCL11A is a master repressor of fetal hemoglobin. Curr Opin Genet Dev. 2015;33:62-70.

17. McGann PT, Ware RE. Hydroxyurea for sickle cell anemia: what have we learned and what questions still remain? Curr Opin Hematol. 2011;18(3):158-165. 18. Cario H, Grosse R, Jarisch A, Kulozik AE, Kunz JB, Lobitz S. AWMF-Leitlinie 025/016 Sichelzellkrankheit. 2014. 19. Kunz JB, Lobitz S, Grosse R, et al. Sickle cell disease in Germany: results from a national registry. Pediatr Blood Cancer. 2019;67(4):e28130. 20. Bauer DE, Kamran SC, Lessard S, et al. An erythroid enhancer of BCL11A subject to genetic variation determines fetal hemoglobin level. Science. 2013;342(6155):253-257. 21. Farrell JJ, Sherva RM, Chen ZY, et al. A 3-bp deletion in the HBS1L-MYB intergenic region on chromosome 6q23 is associated with HbF expression. Blood. 2011;117(18):4935-4945. 22. Gilman JG, Huisman TH. DNA sequence variation associated with elevated fetal G gamma globin production. Blood. 1985;66(4):783-787. 23. Bhatnagar P, Purvis S, Barron-Casella E, et al. Genome-wide association study identifies genetic variants influencing F-cell levels in sickle-cell patients. J Hum Genet. 2011;56(4):316-323. 24. Labie D, Dunda-Belkhodja O, Rouabhi F, Pagnier J, Ragusa A, Nagel RL. The -158 site 5' to the G gamma gene and G gamma expression. Blood. 1985;66(6):1463-1465. 25. Kulozik AE, Kar BC, Satapathy RK, Serjeant BE, Serjeant GR, Weatherall DJ. Fetal hemoglobin levels and beta (s) globin haplotypes in an Indian populations with sickle cell disease. Blood. 1987;69(6):1742-1746. 26. Gardner K, Fulford T, Silver N, et al. g(HbF): a genetic model of fetal hemoglobin in sickle cell disease. Blood Adv. 2018;2(3):235-239. 27. Bowden DK, Vickers MA, Higgs DR. A PCR-based strategy to detect the common severe determinants of α thalassaemia. Br J Haematol. 1992;81(1):104-108. 28. Charache S, Terrin ML, Moore RD, et al. Effect of hydroxyurea on the frequency of painful crises in sickle cell anemia. Investigators of the Multicenter Study of Hydroxyurea in Sickle Cell Anemia. N Engl J Med. 1995;332(20):1317-1322. 29. Borba R, Lima CS, Grotto HZ. Reticulocyte parameters and hemoglobin F production in sickle cell disease patients undergoing hydroxyurea therapy. J Clin Lab Anal. 2003;17(2):66-72. 30. Kar BC, Satapathy RK, Kulozik AE, et al. Sickle cell disease in Orissa State, India. Lancet. 1986;2(8517):1198-1201. 31. Marahatta A, Flanagan JM, Howard TA, et al. Genetic variants that influence fetal hemoglobin expression from hydroxyurea treatment. Blood. 2020;136(Suppl 1):8-9. 32. Mtatiro SN, Makani J, Mmbando B, Thein SL, Menzel S, Cox SE. Genetic variants at HbF-modifier loci moderate anemia and leukocytosis in sickle cell disease in Tanzania. Am J Hematol. 2015;90(1):E1-4. 33. Nicolau M, Vargas S, Silva M, et al. Genetic modulators of fetal hemoglobin expression and ischemic stroke occurrence in African descendant children with sickle cell anemia. Ann Hematol. 2019;98(12):2673-2681.

Haematologica | 107 July 2022

1587

ARTICLE - Modifiers of SCD treated with hydroxyurea

P. Allard et al.

34. Pandey S, Pandey S, Mishra RM, Saxena R. Modulating effect of the -158 gamma (C-->T) Xmn1 polymorphism in Indian sickle cell patients. Mediterr J Hematol Infect Dis. 2012;4(1):e2012001. 35. Gueye Tall F, Martin C, Ndour EHM, et al. Combined and differential effects of alpha-thalassemia and HbF-quantitative trait loci in Senegalese hydroxyurea-free children with sickle cell anemia. Pediatr Blood Cancer. 2019;66(10):e27934. 36. Al-Allawi N, Qadir SMA, Puehringer H, Chui DHK, Farrell JJ, Oberkanins C. The association of HBG2, BCL11A, and HMIP polymorphisms with fetal hemoglobin and clinical phenotype in Iraqi Kurds with sickle cell disease. Int J Lab Hematol. 2018;41(1):87-93. 37. Dadheech S, Jain S, Madhulatha D, et al. Association of Xmn1 -158 gammaG variant with severity and HbF levels in beta-thalassemia

major and sickle cell anaemia. Mol Biol Rep. 2014;41(5):3331-3337. 38. Italia K, Jain D, Gattani S, et al. Hydroxyurea in sickle cell disease--a study of clinico-pharmacological efficacy in the Indian haplotype. Blood Cells Mol Dis. 2009;42(1):25-31. 39. Steinberg MH, Chui DH, Dover GJ, Sebastiani P, Alsultan A. Fetal hemoglobin in sickle cell anemia: a glass half full? Blood. 2014;123(4):481-485. 40. Steinberg MH, Sebastiani P. Genetic modifiers of sickle cell disease. Am J Hematol. 2012;87(8):795-803. 41. Brown AK, Sleeper LA, Miller ST, Pegelow CH, Gill FM, Waclawiw MA. Reference values and hematologic changes from birth to 5 years in patients with sickle cell disease. Cooperative Study of Sickle Cell Disease. Arch Pediatr Adolesc Med. 1994;148(8):796-804.

Haematologica | 107 July 2022

1588

ARTICLE - Iron Metabolism & its Disorders

Hepcidin regulation in Kenyan children with severe malaria and non-typhoidal Salmonella bacteremia Kelvin M. Abuga,1,2 John Muthii Muriuki,1 Sophie M. Uyoga,1 Kennedy Mwai,1,3 Johnstone Makale,1 Reagan M. Mogire,1,4 Alex W. Macharia,1,4 Shebe Mohammed,1 Esther Muthumbi,1 Salim Mwarumba,1 Neema Mturi,1 Philip Bejon,1,5 J. Anthony G. Scott,1,6 Manfred Nairz,7 Thomas N. Williams1,5,8 and Sarah H. Atkinson1,5,9 Kenya Medical Research Institute (KEMRI) Center for Geographic Medicine Research, KEMRI-Wellcome Trust Research Program, Kilifi, Kenya; 2Department of Public Health, School of Human and Health Sciences, Pwani University, Kilifi, Kenya; 3Epidemiology and Biostatistics Division, School of Public Health, University of the Witwatersrand, Johannesburg, South Africa; 4Open University, KEMRI-Wellcome Trust Research Program – Accredited Research Center, Kilifi, Kenya; 5Center for Tropical Medicine and Global Health, Nuffield Department of Clinical Medicine, University of Oxford, Oxford, UK; 6Department of Infectious Disease Epidemiology, London School of Hygiene and Tropical Medicine, London, UK; 7Department of Internal Medicine II, Medical University Innsbruck, Innsbruck, Austria; 8 Department of Infectious Diseases and Institute of Global Health Innovation, Imperial College, London, UK and 9Department of Pediatrics, University of Oxford, Oxford, UK 1

Correspondence: Sarah H. Atkinson satkinson@kemri-wellcome.org Kelvin M. Abuga kmokaya@kemri-wellcome.org Received: May 26, 2021. Accepted: September 1, 2021. Prepublished: September 9, 2021. https://doi.org/10.3324/haematol.2021.279316 ©2022 Ferrata Storti Foundation Haematologica material is published under a CC BY-NC license

Abstract Malaria and invasive non-typhoidal Salmonella (NTS) are life-threatening infections that often co-exist in African children. The iron-regulatory hormone hepcidin is highly upregulated during malaria and controls the availability of iron, a critical nutrient for bacterial growth. We investigated the relationship between Plasmodium falciparum malaria and NTS bacteremia in all pediatric admissions aged <5 years between August 1998 and October 2019 (n=75,034). We then assayed hepcidin and measures of iron status in five groups: (1) children with concomitant severe malarial anemia (SMA) and NTS (SMA+NTS, n=16); and in matched children with (2) SMA (n=33); (3) NTS (n=33); (4) cerebral malaria (CM, n=34); and (5) community-based children. SMA and severe anemia without malaria were associated with a 2-fold or more increased risk of NTS bacteremia, while other malaria phenotypes were not associated with increased NTS risk. Children with SMA had lower hepcidin/ferritin ratios (0.10; interquartile range [IQR]: 0.03-0.19) than those with CM (0.24; IQR: 0.14-0.69; P=0.006) or asymptomatic malaria (0.19; IQR: 0.09-0.46; P=0.01) indicating suppressed hepcidin levels. Children with SMA+NTS had lower hepcidin levels (9.3 ng/mL; IQR: 4.7-49.8) and hepcidin/ferritin ratios (0.03; IQR: 0.01-0.22) than those with NTS alone (105.8 ng/mL; IQR: 17.3-233.3; P=0.02 and 0.31; IQR: 0.06-0.66; P=0.007, respectively). Since hepcidin degrades ferroportin on the Salmonella-containing vacuole, we hypothesize that reduced hepcidin in children with SMA might contribute to NTS growth by modulating iron availability for bacterial growth. Further studies are needed to understand how the hepcidin-ferroportin axis might mediate susceptibility to NTS in severely anemic children.

Introduction Malaria and invasive non-typhoidal Salmonella (NTS) are major causes of illness and death among children living in sub-Saharan Africa. According to the World Health Organization (WHO), 94% of the 409,000 malaria-associated deaths in 2019 occurred in the sub-Saharan African region, with children under 5 years of age being disproportionately vulnerable.1 In this region, NTS bacteremia is also common accounting for 80% of the estimated 535,000 global cases in 2017.2 While NTS is commonly associated with self-limiting gastroenteritis in European populations, NTS infections in African children can cause life-threatening sepsis with case fatality rates of 20-25%.2,3 NTS bacteremia is highly prevalent in areas with concurrent malaria ende-

micity,4-6 and reduced malaria incidence has been associated with a decrease in NTS bacteremia.7,8 The association between NTS and malaria has been particularly observed in children with severe malarial anemia (SMA),5,9-11 but this has not been reported in all settings.12,13 SMA may increase susceptibility to NTS bacteremia via a number of contributory pathways including sustained hemolysis, accumulation of free heme from lysed red blood cells, increased gut permeability, disruption of immune responses, and upregulation of heme oxygenase-1 (Figure 1).14 Heme oxygenase-1 impairs neutrophil oxidative burst capacity, reduces neutrophil bactericidal activity, and promotes iron accumulation in macrophages.15,16 Recent in vitro and animal studies suggest that hepcidin, the master iron regulator,17 may also play an important role in deter-

Haematologica | 107 July 2022

1589

ARTICLE - Hepcidin, malaria and non-typhoidal Salmonella

K.M. Abuga et al.

Figure 1. The hepcidin-link between severe malarial anemia and non-typhoidal Salmonella bacteremia. Low hepcidin levels in children with severe malarial anemia (SMA) may contribute to the risk of non-typhoidal Salmonella (NTS) bacteremia. (A) During malaria infection, proinflammatory responses and parasitemia induce the expression of hepcidin, the main iron regulatory hormone. Hepcidin degrades ferroportin (FPN) on the macrophage membrane and the Salmonella-containing vacuole (SCV),21 resulting in decreased iron availability for NTS bacteria. The bacteria may also utilize other iron acquisition strategies such as transferrin through transferrin receptors (TfR) in early endosomes. Proinflammatory responses, including production of interleukin (IL)-6 and interferon-gamma (IFN-γ), mediate killing of NTS through formation of reactive oxygen species (ROS) and other pathways. (B) In SMA, increased hemolysis and erythropoietic drive induce production of erythroferrone (ERFE),36 a hormone that downregulates hepcidin synthesis. This results in increased expression of FPN on the surface of the macrophage and the SCV.21 Heme from hemolyzed parasitized red blood cells (pRBC) and the haptoglobin-hemoglobin (Hp-Hb) complex is broken down by heme oxygenase-1 (HO-1) into equimolar amounts of iron, biliverdin and carbon monoxide. HO-1 and hemebreakdown products downregulate immune responses and promote an anti-inflammatory environment.15 The net effect of low hepcidin, increased HO-1, SMA-induced anti-inflammatory cytokines such as IL-10 and increased intra-SCV iron levels is increased proliferation of NTS bacteria. DMT-1: divalent metal transporter 1; Cp: ceruloplasmin.

mining susceptibility to NTS by controlling the availability of iron,18-20 a nutrient critical for bacterial growth and proliferation.16,19 Hepcidin degrades ferroportin, the sole iron exporter, which was recently shown to transport iron across the Salmonella-containing vacuole (SCV).21,22 In murine studies, low hepcidin levels and increased ferroportin expression on the SCV are associated with increased susceptibility to Salmonella Typhimurium infections (Figure 1).18,22 However, there are no studies of hepcidin in NTS infection in humans. In this study, we investigated the relationship between malaria, anemia and NTS in 75,034 hospitalized Kenyan children over a 21-year period and then estimated levels of hepcidin and other iron biomarkers in children with NTS bacteremia and malaria.

line Supplementary Appendix). The study was conducted in Kilifi, a rural malaria-endemic area along the Kenyan coast. The estimated incidence rate of NTS bacteremia among children <5 years was 36.6 cases/100,000 person-years between 1998 to 2014.23 Our study included two parts: (i) We investigated the relationship between malaria and NTS bacteremia among all pediatric admissions (n=75,034) aged ≤60 months admitted between August, 1 1998 and October, 31 2019 with complete age, malaria and hemoglobin data. (ii) We then measured hepcidin, iron and inflammatory markers in five groups of children including those hospitalized with: 1) SMA and NTS coinfection (SMA+NTS); 2) SMA alone; 3) NTS alone; and 4) cerebral malaria (CM) and 5) community-based children with and without asymptomatic malaria using stored biobank samples over the 21year time period. Each child from group 1 was matched Methods with two from each of the other hospitalized groups based on age and sex (Figure 2). The community-based children Study design and participants Ethical approval was granted by the Scientific Ethics Review were part of an ongoing birth cohort evaluating malaria Unit of the Kenya Medical Research Institute and informed immunity,24 and their samples were selected from a single consent was provided by parents or guardians (see the On- cross-sectional bleed in May 2002. Haematologica | 107 July 2022

1590

ARTICLE - Hepcidin, malaria and non-typhoidal Salmonella

K.M. Abuga et al.

Figure 2. Selection of study participants. All children aged ≤60 months with complete age and hemoglobin data admitted between August 1998 and October 2019 were included in the retrospective epidemiological analysis. Children with concomitant severe malaria and non-typhoidal Salmonella (NTS), and whose specimens were available in the Kilifi biobank, were enrolled into the iron and hepcidin sub-study. Each child was then matched with two hospitalized children with NTS alone, severe malaria anemia (SMA) alone, and cerebral malaria (CM) alone based on age and sex.

Laboratory procedures Thick and thin blood films were stained with Giemsa and examined for Plasmodium falciparum using standard methods. Samples for bacterial culture were collected in BACTEC® Peds Plus bottles and processed with a BACTEC9050 automated blood-culture instrument (Becton-Dickson, UK). Positive samples were sub-cultured and serological tests and biochemical test kits (API, bioMérieux) were used to confirm suspected pathogens.23,25 Bacillus species, Micrococcus species, viridans group Streptococcus, coagulase-negative Staphylococcus, and coryneforms were considered contaminants. Rapid antibody tests were used for human immunodeficiency virus 1 (HIV-1) testing. Sickle cell disease was diagnosed using polymerase chain reaction (see the Online Supplementary Appendix). Iron and inflammatory biomarkers were assayed as previously described26 (Online Supplementary Appendix).

Statistical analyses All data were analyzed using STATA 15.1 (StatCorp. College Station, Texas, USA). For all pediatric admissions, we used univariable and multivariable logistic regression models to investigate for putative risk factors for NTS bacteremia. We used a causal directed acyclic graph to assess the suitability of covariates for multivariable analyses (Online Supplementary Figure S1), and a stepwise backward selection regression method to fit the final multivariable models (Online Supplementary Appendix). We also analyzed the relationship between SMA and risk of other bacterial organisms. In the hepcidin sub-study, continuous data were reported as medians and interquartile ranges (IQR) and compared using the Wilcoxon rank-sum test. Non-parametric Spearman’s correlation evaluated associations between variables. We normalized non-normallydistributed variables by loge-transformation and used multivariable linear regression models to adjust for inflammation (C-reactive protein) and year of admission.

Clinical definitions For children with Plasmodium falciparum malaria, we defined SMA as hemoglobin <5 g/dL and CM as Blantyre coma score <3 according to WHO criteria.27 Severe anemia Results was defined as hemoglobin <5 g/dL; fever as temperature >37.5⁰C; and NTS bacteremia as isolation of Salmonella Study of all hospital admissions enterica subspecies excluding Typhi or Paratyphi serovars A total of 75,034 children aged ≤60 months were adin blood cultures. mitted to Kilifi County Hospital during the 21-year study Haematologica | 107 July 2022

1591

ARTICLE - Hepcidin, malaria and non-typhoidal Salmonella

K.M. Abuga et al.

Table 1. Factors associated with non-typhoidal Salmonella bacteremia in all hospitalized children (n=75,034).

Characteristic Clinical features Age, years (IQR) Sex, male Fever Diarrhea2 Vomiting Severe pneumonia3 Very severe pneumonia4 Underweight5 Stunting6 Wasting7 Laboratory characteristics Malaria slide positive SMA8 Cerebral malaria8,9 Non-SMA malaria Sickle cell disease HIV positive10 Hb, g/dL (IQR) Hb <5 g/dL Hb 5-7 g/dL11 Hb 7-10 g/dL12 Severe anemia without malaria13

NTS, N (%)

Hospital controls, N (%)

OR (95% CI)1

1.07 (0.55, 1.90) 224/400 (56.0) 267/369 (72.4) 126/400 (31.5) 127/388 (32.7) 111/400 (27.8) 59/400 (14.8) 204/340 (60.0) 198/371 (53.4) 180/380 (47.4)

0.98 (0.18, 2.18) 42,226/74,633 (56.6) 37,886/61,979 (61.2) 14,187/74,615 (19.0) 18,061/73,242 (24.7) 21,068/74,610 (28.2) 7,480/74,606 (10.0) 28,479/68,661 (41.5) 27,812/69,510 (40.0) 19,493/67,493 (28.9)

1.01 (0.93, 1.08) 0.98 (0.80, 1.19) 1.66 (1.32, 2.09) 1.96 (1.58, 2.42) 1.49 (1.20, 1.84) 0.98 (0.78, 1.22) 1.55 (1.18, 2.05) 2.12 (1.70, 2.63) 1.72 (1.40, 2.11) 2.22 (1.81, 2.71)

0.87 0.82 <0.0001 <0.0001 0.0003 0.83 0.002 <0.0001 <0.0001 <0.0001

93/400 (23.3) 38/393 (9.7) 8/286 (2.7) 48/400 (12.0) 14/400 (3.7) 38/139 (27.3) 7.4 (5.2, 9.4) 89/400 (22.3) 86/311 (27.7) 147/225 (65.3) 44/400 (11.0)

16,370/74,632 (21.9) 2,253/74,223 (3.0) 1,719/63,244 (2.7) 13,708/74,632 (18.4) 1,115/74,608 (1.5) 1,756/35,327 (5.0) 9.8 (8.1, 11.6) 4,660/74,615 (6.2) 6,575/69,974 (9.4) 27,867/63,399 (44.0) 1,997/74,634 (2.7)

1.08 (0.85, 1.36) 3.42 (2.44, 4.79) 1.00 (0.50, 2.02) 0.61 (0.45, 0.82) 2.39 (1.40, 4.09) 7.19 (4.94, 10.48) 1.28 (1.24, 1.32) 4.30 (3.39, 5.45) 3.69 (2.87, 4.73) 2.40 (1.83, 3.16) 4.50 (3.28, 6.17)

0.52 <0.0001 0.99 0.001 0.002 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001

NTS: non-typhoidal Salmonella; n/N: number positive/number tested; OR: odds ratio; CI: confidence interval; SMA: severe malaria anemia; CM: cerebral malaria; Hb: hemoglobin; HIV: human immunodeficiency virus. 1Odds ratios and P values were derived from univariable logistic regression models; 2passage of three or more loose or liquid stools within 24 hours; 3history of cough or difficulty in breathing plus lower chest wall indrawing; 4cough or difficulty breathing plus either prostration, lethargy, hypoxia, loss of consciousness, or a history of convulsions; 5 weight-for-age z-score < –2; 6height-for-age z-score <–2; 7weight-for-height z-score <–2 or mid-upper arm circumference <12.5 cm in children >6 months of age using World Health Organization Child Growth Standards;50 8children with overlapping SMA and CM clinical syndromes were excluded from analysis; 9only children with Blantyre coma scale scores were included; 10data were available from February 2005 after routine HIV testing was introduced thus analyses included a limited number of children (n = 35,466); 11Excludes children with hemoglobin levels <5 g/dL; 12excludes children with hemoglobin levels <7 g/dL; 13severe anemia (hemoglobin <5 g/dL) and no malaria parasites on blood film.

period and had complete data for analysis. Median age was 11.8 months (IQR: 2.2-26.1) and 42,450 (56.6%) were male. P. falciparum malaria was identified in the blood films of 16,463 (21.9%) hospitalized children of whom 2,291 (13.9%) had SMA, 1,727 (10.5%) had CM, and 416 (2.5%) had concomitant SMA and CM. Pathogenic bacterial organisms were isolated from 3,792 (5.1%) blood cultures. NTS bacteremia was identified in 400 (10.5%) of the positive blood cultures. Of the NTS isolates, 309 were serotyped and 45.0% (139/309) were Salmonella enterica serovar Enteritidis and 44.3% (137/309) were serovar Typhimurium, while 10.7% (33/309) were not typeable. The prevalence of NTS bacteremia has decreased over the years as malaria has also decreased (Online Supplementary Figure S2). NTS bacteremia was identified in 93 of 16,463 (0.6%) hospitalized children with P. falciparum malaria, including 38 of 2,291 (1.7%) with SMA and eight of 1,727 (0.5%) with CM. SMA was associated with a 2-fold increased risk of NTS bacteremia in the final multivariable model (adjusted odds ratio [OR]: 2.17; 95% confidence interval [CI]: 1.44-3.28;

P=0.0002; Online Supplementary Table S2). However, a positive malaria slide (OR: 1.08; 95% CI: 0.85-1.36; P=0.52) and CM (OR: 1.00; 95% CI: 0.50-2.02; P=0.99) were not associated with increased risk of NTS bacteremia (Table 1). Children with malaria but without SMA had a 39% reduced risk of NTS bacteremia (OR: 0.61; 95% CI: 0.45-0.82; P=0.001 Table 1). Children with severe anemia without malaria parasitemia also had an increased risk of NTS bacteremia in final multivariable models (adj. OR: 4.03; 95% CI: 2.78-5.84; P<0.0001; Online Supplementary Table S2). The risk of NTS bacteremia increased by 26% for each 1 g/dL decrease in hemoglobin levels in all children (adj. OR: 1.26; 95% CI: 1.21-1.32; P<0.0001) and by 51% in children with malaria parasitemia (adj. OR: 1.51; 95% CI: 1.36-1.68; P<0.0001). In final multivariable models, other risk factors for NTS bacteremia were younger age, fever, diarrhea, sickle cell disease, very severe pneumonia, underweight and, in restricted models, HIV status (Online Supplementary Table S2). SMA was not associated with increased risk of other bacterial organisms causing bacteremia (Online Supplementary Figure S3).

Haematologica | 107 July 2022

1592

ARTICLE - Hepcidin, malaria and non-typhoidal Salmonella

K.M. Abuga et al.

Hepcidin sub-study We included 116 hospitalized children in the following groups: 1) 16 with SMA+NTS; 2) 33 with SMA alone; 3) 33 with NTS alone; and 4) 34 with CM (Figure 2); and 5) 291 community-based children with (n=49) and without (n=242) asymptomatic malaria parasitemia. The clinical characteristics of children in the sub-study are shown in the Online Supplementary Table S3.

IQR: 1.2-12.6, respectively). Hepcidin expression was suppressed in children with SMA as evidenced by a lower hepcidin/ferritin ratio (0.10; IQR: 0.03-0.19) compared to those with CM (0.24; IQR: 0.14-0.69; P=0.006), or asymptomatic parasitemia (0.19; IQR: 0.09-0.46; P=0.01) (Figure 3B). We then explored differences in putative regulators of hepcidin. Children with SMA had increased erythropoietic drive as indicated by higher soluble transferrin receptor (sTfR) levels (43.3 mg/L; IQR: 30.8-65.6) than those with CM (31.2 Hepcidin levels in children with malaria mg/L; IQR: 23.9-45.5; P=0.03), although ferritin and C-reative We first compared hepcidin levels among children with ma- protein (CRP) levels did not differ between the groups (Figlaria. Hepcidin levels were lower in children with SMA ure 3). Hospitalized children had higher levels of ferritin, (median 31.1 ng/mL, IQR: 5.5-61.2) compared to those with sTfR, and CRP and higher P. falciparum parasite densities CM (90.7 ng/mL; IQR: 38.7-176.1; P=0.002). However, both of than those living in the community (Figure 3C-F). these severe malaria groups had higher hepcidin levels than children with asymptomatic malaria parasitemia living in the Hepcidin levels in children with malaria and community (Figure 3A). We found similar hepcidin levels in non-typhoidal Salmonella community-based children with and without asymptomatic We then considered hepcidin levels in children with mamalaria parasitemia (6.5 ng/mL; IQR: 2.0-13.1 and 3.8 ng/mL; laria and NTS. Hepcidin levels were lower in children with

Figure 3. Iron and inflammatory biomarkers in children with malaria. Circulating levels of (A) hepcidin, (B) ferritin, (C) hepcidin/ferritin ratio, (D) soluble transferrin receptors (sTfR), (E) C-reactive protein (CRP), and (F) parasite density. P values from pairwise comparisons were determined by the Wilcoxon rank-sum test. NS’ (not significant) indicates P>0.05. AM: asymptomatic malaria; CM: cerebral malaria; SMA: severe malaria anemia. Haematologica | 107 July 2022

1593

ARTICLE - Hepcidin, malaria and non-typhoidal Salmonella

K.M. Abuga et al.

Table 2. Hepcidin and biomarkers of iron status and inflammation in a sub-study of hospitalized and community-based children.

Biomarker Hepcidin, ng/mL

Ferritin, mg/L

Hepcidin/ferritin ratio2

sTfR, mg/L

CRP, mg/L

Group

Medians (IQR)

SMA and NTS coinfection Severe malaria anemia NTS bacteremia Cerebral malaria Asymptomatic malaria Healthy controls SMA and NTS coinfection Severe malaria anemia NTS bacteremia Cerebral malaria Asymptomatic malaria Healthy controls SMA and NTS coinfection Severe malaria anemia NTS bacteremia Cerebral malaria Asymptomatic malaria Healthy controls SMA and NTS coinfection Severe malaria anemia NTS bacteremia Cerebral malaria Asymptomatic malaria Healthy controls SMA and NTS coinfection Severe malaria anemia NTS bacteremia Cerebral malaria Asymptomatic malaria Healthy controls

16 33 33 34 49 242 16 32 29 28 48 237 16 32 29 28 48 232 16 33 32 33 49 239 16 33 33 33 48 237

9.3 (4.7, 49.8) 31.1 (5.5, 61.2) 105.8 (17.3, 233.3) 90.7 (38.7, 176.1) 6.5 (2.0, 13.1) 3.8 (1.2, 12.6) 311.5 (241, 392) 348.5 (296, 384) 356.0 (203, 397) 370.0 (359, 393) 30.5 (17.0, 53.0) 16.0 (8.0, 26.0) 0.03 (0.01, 0.22) 0.10 (0.03, 0.19) 0.31 (0.06, 0.66) 0.24 (0.14, 0.69) 0.19 (0.09, 0.46) 0.27 (0.08, 0.66) 48.1 (36.8, 66.9) 43.3 (30.1, 61.3) 38.3 (30.9, 67.6) 31.2 (23.9, 45.5) 3.8 (2.7, 4.9) 3.6 (2.8, 4.8) 117.5 (79.0, 145.0) 120.0 (79.2, 152.7) 104.3 (35.1, 162.4) 120.9 (57.8, 150.9) 1.0 (0.3, 7.1) 0.3 (0.3, 2.0)

Reference 0.43 0.02 0.004 0.16 0.01 Reference 0.55 0.76 0.23 <0.0001 <0.0001 Reference 0.53 0.007 0.01 0.01 0.0006 Reference 0.64 0.50 0.02 <0.0001 <0.0001 Reference 0.80 0.67 0.93 <0.0001 <0.0001

IQR: interquartile range; SMA: severe malaria anemia; NTS: non-typhoidal Salmonella; CRP: C-reactive protein; sTfR: soluble transferrin receptor; and n/a: data not available. 1P values were derived using pairwise Wilcoxon rank sum test. 2Hepcidin/ferritin ratio was calculated by dividing hepcidin (ng/mL) by ferritin (mg/L).

SMA+NTS (9.3 ng/mL; IQR: 4.7-49.8) compared to those with NTS alone (105.8 ng/ml; IQR: 17.3-233.3) (Figure 4A; Table 2). Hepcidin/ferritin ratios were also lower in children with SMA+NTS (0.03; IQR: 0.01-0.22) compared to those with NTS alone (0.31; IQR: 0.06-0.66; P=0.007) (Table 2; Figure 4B). In a linear regression model controlled for CRP and year of admission, hepcidin levels were 2-fold higher in children with NTS compared to those with SMA+NTS (adj. b 1.99; 95% CI: 0.81-3.26; P=0.001) (Online Supplementary Table S4), although sTfR, ferritin, and CRP levels did not differ between the groups (Figure 4C-E). Only one participant in the sub-study had sickle cell disease and excluding the participant from the analysis did not alter our findings. Hepcidin levels were positively correlated with ferritin (r=0.38, P=0.0001), CRP (r=0.31; P=0.0007), hemoglobin (r=0.37; P<0.0001) and parasite density (r=0.44; P<0.0001), and negatively correlated with sTfR (r=-0.37; P<0.0001)

among the hospitalized children. The direction and strength of correlation between hepcidin and its predictors varied across individual groups as shown in the Online Supplementary Table S5.

Discussion Malaria and NTS are important causes of hospitalization and death among children living in sub-Saharan Africa.1,2 In this study, we analyzed retrospective data from 75,034 hospitalized children aged ≤60 months and found that SMA, but not CM or other malaria phenotypes, was associated with increased risk of NTS bacteremia. Children with severe anemia of all causes, both with or without malaria parasitemia, also had an increased risk of NTS bacteremia. In a sub-study investigating iron biomarkers, children with SMA had lower hepcidin levels than children

Haematologica | 107 July 2022

1594

ARTICLE - Hepcidin, malaria and non-typhoidal Salmonella

K.M. Abuga et al.

Figure 4. Iron and inflammatory biomarkers in children with severe malaria anemia and/or non-typhoidal Salmonella bacteremia. Circulating levels of (A) hepcidin, (B) ferritin, (C) hepcidin/ferritin ratio, (D) soluble transferrin receptors (sTfR), (E) Creactive protein (CRP), and (F) parasite density in children with severe malarial anemia (SMA) and/or non-typhoidal Salmonella (NTS) bacteremia. P values from pairwise comparisons were determined by the Wilcoxon rank-sum test. NS’ (not significant) indicates P>0.05.

with CM. We also found that children with SMA+NTS had lower hepcidin levels than children with NTS alone. We did not find differences in ferritin or CRP levels among hospitalized children, but children with SMA alone and SMA+NTS had lower hepcidin/ferritin ratios and higher sTfR levels. Children living in the community with asymptomatic parasitemia had lower levels of hepcidin, ferritin, CRP and sTfR and lower parasite densities than hospitalized children. Children with SMA had a 2-fold increased risk of NTS bacteremia compared to those without SMA. This risk was not observed in children with CM or other malaria phenotypes that excluded severe anemia. Moreover, each 1 g/dL decrease in hemoglobin concentrations in children with malaria was associated with a 51% increase in the risk of NTS bacteremia. SMA increased the risk of NTS, but not other

bacteria suggesting an NTS-specific effect rather than a generalized immunological failure to control bacteremia (Online Supplementary Figure S3). Previous studies across sub-Saharan Africa have also reported an increased risk of NTS bacteremia in children with SMA,5,9,10 but not CM.10,28 However, these observations have not been universal. A study in Mozambican children reported no clear-cut association between SMA and NTS bacteremia, although few children had NTS bacteremia (n=12).13 In agreement with the current study, previous studies have found no association between malaria and risk of NTS bacteremia,4,23 although other studies have reported mixed findings with malaria both reducing and increasing risk of NTS bacteremia.29-32 These differences might be explained by the prevalence of malarial anemia within the study populations or various other factors, including nutritional status. Taken to-

Haematologica | 107 July 2022

1595

ARTICLE - Hepcidin, malaria and non-typhoidal Salmonella gether, our findings suggest that SMA, rather than other malarial phenotypes, underlies the association between malaria and NTS bacteremia. Indeed, severe anemia due to all causes was strongly associated with NTS bacteremia, even after excluding children with malaria parasitemia, in agreement with a previous study in Malawian children.33 A number of pathways may contribute to increased risk of NTS bacteremia in children with SMA including hemolysis, iron overload and upregulation of heme oxygenase-1 (Figure 1).14 Hepcidin may also influence risk of NTS bacteremia in SMA by controlling the availability of iron for bacterial growth.18-20 We observed that hepcidin levels and hepcidin/ferritin ratios were lower in hospitalized children with SMA compared to those with CM. In agreement, a study in Kenyan children found lower hepcidin levels in malaria patients with severe anemia compared to those with higher hemoglobin levels.34 In contrast, a study in Nigerian children found no difference in hepcidin levels between children with SMA and CM and higher hepcidin levels in uncomplicated compared to severe malaria.35 Our findings may be explained by the low hepcidin/ferritin ratio and higher sTfR levels in SMA compared to CM, indicating increased erythropoietic activity. Severe anemia negatively regulates hepcidin production through the action of erythroferrone,36 even in the presence of inflammation/infection37,38 or sickle cell disease.39 Inflammation, as measured by ferritin and CRP, did not differ between the SMA and CM groups, although parasite density, known to correlate with hepcidin levels,40 was higher in CM. In agreement with previous studies,41,42 we found higher hepcidin levels in children with severe malaria compared to those with asymptomatic parasitemia. This is likely to be due to increased inflammation in severe malaria, rather than the older age of the community-based children, since older children would be expected to have higher hepcidin levels than younger children.26,43 Iron is an essential nutrient for bacterial growth and ex vivo studies suggest that increased serum iron levels may stimulate the growth of various bacteria including Salmonella Typhimurium.44,45 In mouse models, reduced hepcidin levels are associated with increased susceptibility to NTS infections,18 although little is known about the role of hepcidin during NTS and malaria infections in children. In the current study, children with SMA+NTS had lower hepcidin levels and hepcidin/ferritin ratios than those with NTS alone; although sTfR, CRP and ferritin levels did not differ between these groups. High hepcidin levels would be expected in children with NTS bacteremia since hepcidin is known to increase in response to inflammation and infection. A challenge infection study with Salmonella enterica Typhi in the United Kingdom identified higher hepcidin concentrations during acute infection.46 In vitro and murine studies also show that Salmonella Typhimurium may directly or indirectly upregulate hepcidin ex-

K.M. Abuga et al. pression and perturb iron regulatory pathways.20 Hepcidin concentrations may alter iron availability within the Salmonella-containing vacuole (SCV). Recent evidence indicates that hepcidin leads to increased degradation of ferroportin on the SCV, and limits the movement of iron into the SCV.21 However, this conflicts with an earlier report that ferroportin transports iron out of the SCV,47 and these contradictions may be based on differences in experimental systems used.48 It also remains controversial whether iron accumulation in the SCV may promote bacterial growth by increasing iron availability,19,22 or kill bacteria through the Fenton reaction.21 Low hepcidin levels in mice with severe hemolytic anemia were associated with increased susceptibility to Salmonella Typhimurium infection and hepcidin treatment improved survival.18 We hypothesize that low hepcidin levels in children with SMA, and non-malaria severe anemia, might contribute to NTS bacteremia by increasing iron availability in the SCV for NTS growth together with other mechanisms (Figure 1). Surprisingly, sTfR levels were elevated in children with NTS alone despite higher hemoglobin levels. It is not known whether NTS might induce transcription of transferrin receptors to increase transferrin iron acquisition, since transferrin receptors have been observed on the SCV during early phases of NTS endocytosis in murine models.49 To the best of our knowledge, this is the first study reporting hepcidin levels in children with NTS or with concomitant SMA and NTS bacteremia. The strengths of the study are that we utilized a very large 21-year dataset (n=75,034) with matching stored samples to identify and describe associations between severe malaria, NTS bacteremia and hepcidin in children. Our study also has important limitations. First, the study was observational, and as such, associations may be subject to unmeasured confounders and reverse causality. Second, we had few samples for children with SMA and NTS co-infection and no samples for those with CM and NTS co-infection since some samples were insufficient or missing, which may have introduced selection bias. Nonetheless, these are a unique sample set with accompanying clinical data collected over 21 years. Another limitation is that we did not measure additional parameters such as serum iron, transferrin saturation, and haptoglobin levels due to the volumes and availability of stored samples. Additionally, a few participants had sTfR concentrations above the cut-off values making it challenging to interpret findings from regression models for sTfR (Online Supplementary Table S3). Finally, our study was conducted in a single site. It is also possible that our study underestimated associations, considering the low sensitivity of blood cultures used to identify NTS. Nonetheless, this study complements the existing in vitro and animal data on the relationship between SMA and NTS bacteremia and provides preliminary evidence on the possible role of hepcidin in mediating this association.

Haematologica | 107 July 2022

1596

ARTICLE - Hepcidin, malaria and non-typhoidal Salmonella In conclusion, SMA was associated with a strongly increased risk of NTS bacteremia in children and reduced hepcidin levels were observed in children with SMA and SMA+NTS. The question of whether ferroportin transports iron into or out of the SCV remains an active area of research,21,47 and future findings may support our hypothesis or generate new ideas on how low hepcidin might mediate NTS susceptibility in children with SMA. Further studies are needed to understand the role of the hepcidin-ferroportin axis in susceptibility to NTS in human subjects, how hepcidin and iron disturbances might mediate susceptibility to bacteremia due to NTS or other organisms, and how P. falciparum, iron deficiency or other etiologies of severe anemia influence this relationship.

K.M. Abuga et al. Acknowledgments The authors would like to thank the children who participated in this study and their parents/guardians. This manuscript was submitted for publication with the permission of the Director of the Kenya Medical Research Institute (KEMRI).

Funding This study was funded by Wellcome (grant numbers 110255 to SHA, 212600 to KMA, 202800 to TNW, and a core award to the KEMRI-Wellcome Trust Research Program [203077]). KMA, RMM, EM and JMM were supported by the DELTAS Africa Initiative [DEL-15-003]. The DELTAS Africa Initiative is an independent funding scheme of the African Academy of Sciences (AAS)'s Alliance for Accelerating Excellence in Disclosures Science in Africa (AESA) and supported by the New PartNo conflicts of interest to disclose. nership for Africa's Development Planning and Coordinating Agency (NEPAD Agency) with funding from Wellcome Contributions [107769] and the UK government. The funders had no role SHA conceptualized and designed the study; SHA, MN, in study design, data collection and analysis, decision to JAGS, and TNW supervised the study; SMU, JM, SMwa- publish, or preparation of the manuscript. rumba and AWM performed laboratory analyses; EM serotyped non-typhoidal Salmonella isolates; KMA, JMM, and Data-sharing statement SHA analyzed and interpreted the data; KMA and SHA The data and analyses underlying this article are available wrote the original draft of the manuscript; KMA, JMM, SMU, in Harvard Dataverse at https://doi.org/ KM, JM, RM, AWM, MS, SMohammed, EM, SMwarumba, NM, 10.7910/DVN/KXZWN6 and applications for data access can PB, JAGS, MN, TNW, and SHA revised subsequent drafts be made through the Kilifi Data Governance Committee and approved the final draft for publication. cgmrc@kemri-wellcome.org.

References 1. World Health Organisation. World malaria report 2020: 20 years of global progress and challenges. 2020 [cited 14th May 2021]; Available from: https://www.who.int/publications/i/item/9789240015791 2. G.B.D. 2017 Non-Typhoidal Salmonella Invasive Disease Collaborators. The global burden of non-typhoidal salmonella invasive disease: a systematic analysis for the Global Burden of Disease Study 2017. Lancet Infect Dis. 2019;19(12):1312-1324. 3. Feasey NA, Dougan G, Kingsley RA, Heyderman RS, Gordon MA. Invasive non-typhoidal salmonella disease: an emerging and neglected tropical disease in Africa. Lancet. 2012;379(9835):2489-2499. 4. Park SE, Pak GD, Aaby P, et al. The relationship between invasive nontyphoidal Salmonella disease, other bacterial bloodstream infections, and malaria in Sub-Saharan Africa. Clin Infect Dis. 2016; 62(Suppl 1):S23-31. 5. Biggs HM, Lester R, Nadjm B, et al. Invasive Salmonella infections in areas of high and low malaria transmission intensity in Tanzania. Clin Infect Dis. 2014; 58(5):638-647. 6. Tabu C, Breiman RF, Ochieng B, et al. Differing burden and epidemiology of non-Typhi Salmonella bacteremia in rural and urban Kenya, 2006-2009. PLoS One. 2012;7(2):e31237. 7. Scott JA, Berkley JA, Mwangi I, et al. Relation between falciparum malaria and bacteraemia in Kenyan children: a population-based, case-control study and a longitudinal study. Lancet. 2011;378(9799):1316-1323.

8. Mackenzie G, Ceesay SJ, Hill PC, et al. A decline in the incidence of invasive non-typhoidal Salmonella infection in The Gambia temporally associated with a decline in malaria infection. PLoS One. 2010;5(5):e10568. 9. Bronzan RN, Taylor TE, Mwenechanya J, et al. Bacteremia in Malawian children with severe malaria: prevalence, etiology, HIV coinfection, and outcome. J Infect Dis. 2007;195(6):895-904. 10. Graham SM, Mwenechanya J, Tembo M, et al. The pattern of bacteraemia in children with severe malaria. Malawi Med J. 2002;14(1):11-15. 11. Nadjm B, Amos B, Mtove G, et al. WHO guidelines for antimicrobial treatment in children admitted to hospital in an area of intense Plasmodium falciparum transmission: prospective study. BMJ. 2010;340:c1350. 12. Falay D, Kuijpers LM, Phoba MF, et al. Microbiological, clinical and molecular findings of non-typhoidal Salmonella bloodstream infections associated with malaria, Oriental Province, Democratic Republic of the Congo. BMC Infect Dis. 2016;16:271. 13. Bassat Q, Guinovart C, Sigaúque B, et al. Severe malaria and concomitant bacteraemia in children admitted to a rural Mozambican hospital. Trop Med Int Health. 2009;14(9):1011-1019. 14. Abuga KM, Muriuki JM, Williams TN, Atkinson SH. How severe anaemia might influence the risk of invasive bacterial infections in African children. Int J Mol Sci. 2020;21(18):6976. 15. Cunnington AJ, de Souza JB, Walther M, Riley EM. Malaria

Haematologica | 107 July 2022

1597

ARTICLE - Hepcidin, malaria and non-typhoidal Salmonella impairs resistance to Salmonella through heme- and heme oxygenase-dependent dysfunctional granulocyte mobilization. Nat Med. 2011;18(1):120-127. 16. Lokken KL, Stull-Lane AR, Poels K, Tsolis RM. Malaria parasitemediated alteration of macrophage function and increased iron availability predispose to disseminated nontyphoidal Salmonella infection. Infect Immun. 2018;86(9):e00301-00318. 17. Pagani A, Nai A, Silvestri L, Camaschella C. Hepcidin and anemia: a tight relationship. Front Physiol. 2019;10:1294. 18. Yuki KE, Eva MM, Richer E, et al. Suppression of hepcidin expression and iron overload mediate Salmonella susceptibility in ankyrin 1 ENU-induced mutant. PLoS One. 2013;8(2):e55331. 19. Liu D, Gan ZS, Ma W, et al. Synthetic porcine Hepcidin exhibits different roles in Escherichia coli and Salmonella infections. Antimicrob Agents Chemother. 2017;61(10):e02638-16. 20. Kim DK, Jeong JH, Lee JM, et al. Inverse agonist of estrogenrelated receptor gamma controls Salmonella typhimurium infection by modulating host iron homeostasis. Nat Med. 2014;20(4):419-424. 21. Lim D, Kim KS, Jeong JH, et al. The hepcidin-ferroportin axis controls the iron content of Salmonella-containing vacuoles in macrophages. Nat Commun. 2018;9(1):2091. 22. Flannagan RS, Farrell TJ, Trothen SM, Dikeakos JD, Heinrichs DE. Rapid removal of phagosomal ferroportin in macrophages contributes to nutritional immunity. Blood Adv. 2021;5(2):459-474. 23. Muthumbi E, Morpeth SC, Ooko M, et al. Invasive Salmonellosis in Kilifi, Kenya. Clin Infect Dis. 2015;61(Suppl 4):S290-301. 24. Bejon P, Williams TN, Liljander A, et al. Stable and unstable malaria hotspots in longitudinal cohort studies in Kenya. PLoS Med. 2010;7(7):e1000304. 25. Berkley JA, Lowe BS, Mwangi I, et al. Bacteremia among children admitted to a rural hospital in Kenya. N Engl J Med. 2005;352(1):39-47. 26. Atkinson SH, Uyoga SM, Armitage AE, et al. Malaria and age variably but critically control Hepcidin throughout childhood in Kenya. EBioMedicine. 2015;2(10):1478-1486. 27. World Health Organisation. Severe malaria. 2014 [cited November 17, 2020]; Available from: https://www.who.int/malaria/publications/atoz/who-severemalaria-tmih-supplement-2014.pdf 28. Enwere G, Van Hensbroek MB, Adegbola R, et al. Bacteraemia in cerebral malaria. Ann Trop Paediatr. 1998;18(4):275-278. 29. Brent AJ, Oundo JO, Mwangi I, Ochola L, Lowe B, Berkley JA. Salmonella bacteremia in Kenyan children. Pediatr Infect Dis J. 2006;25(3):230-236. 30. Mandomando I, Bassat Q, Sigauque B, et al. Invasive Salmonella infections among children from rural Mozambique, 2001-2014. Clin Infect Dis. 2015;61(Suppl 4):S339-345. 31. Mabey DC, Brown A, Greenwood BM. Plasmodium falciparum malaria and Salmonella infections in Gambian children. J Infect Dis. 1987;155(6):1319-1321. 32. Walsh AL, Phiri AJ, Graham SM, Molyneux EM, Molyneux ME. Bacteremia in febrile Malawian children: clinical and microbiologic features. Pediatr Infect Dis J. 2000;19(4):312-318. 33. Calis JC, Phiri KS, Faragher EB, et al. Severe anemia in Malawian children. N Engl J Med. 2008;358(9):888-899. 34. Casals-Pascual C, Huang H, Lakhal-Littleton S, et al. Hepcidin

K.M. Abuga et al. demonstrates a biphasic association with anemia in acute Plasmodium falciparum malaria. Haematologica. 2012;97(11):1695-1698. 35. Burte F, Brown BJ, Orimadegun AE, et al. Circulatory hepcidin is associated with the anti-inflammatory response but not with iron or anemic status in childhood malaria. Blood. 2013;121(15):3016-3022. 36. Latour C, Wlodarczyk MF, Jung G, et al. Erythroferrone contributes to hepcidin repression in a mouse model of malarial anemia. Haematologica. 2017;102(1):60-68. 37. Jonker FA, Calis JC, Phiri K, et al. Low hepcidin levels in severely anemic malawian children with high incidence of infectious diseases and bone marrow iron deficiency. PLoS One. 2013;8(12):e78964. 38. Stoffel NU, Lazrak M, Bellitir S, et al. The opposing effects of acute inflammation and iron deficiency anemia on serum hepcidin and iron absorption in young women. Haematologica. 2019;104(6):1143-1149. 39. Mangaonkar AA, Thawer F, Son J, et al. Regulation of iron homeostasis through the erythroferrone-hepcidin axis in sickle cell disease. Br J Haematol. 2020;189(6):1204-1209. 40. Howard CT, McKakpo US, Quakyi IA, et al. Relationship of hepcidin with parasitemia and anemia among patients with uncomplicated Plasmodium falciparum malaria in Ghana. Am J Trop Med Hyg. 2007;77(4):623-626. 41. Oluboyo OA, Theodora I, Oluboyo A. Impact of malaria severity on serum levels of hepcidin and iron status in children. Online J Health Allied Sciences. 2019;18(1):1-4. 42. Mendonca VR, Souza LC, Garcia GC, et al. Associations between hepcidin and immune response in individuals with hyperbilirubinaemia and severe malaria due to Plasmodium vivax infection. Malar J. 2015;14(1):407. 43. Muriuki JM, Mentzer AJ, Webb EL, et al. Estimating the burden of iron deficiency among African children. BMC Med. 2020;18(1):31. 44. Prentice S, Jallow AT, Sinjanka E, et al. Hepcidin mediates hypoferremia and reduces the growth potential of bacteria in the immediate post-natal period in human neonates. Sci Rep. 2019;9(1):16596. 45. Cross JH, Bradbury RS, Fulford AJ, et al. Oral iron acutely elevates bacterial growth in human serum. Sci Rep. 2015;5:16670. 46. Darton TC, Blohmke CJ, Giannoulatou E, et al. Rapidly escalating hepcidin and associated serum iron starvation are features of the acute response to typhoid infection in humans. PLoS Negl Trop Dis. 2015;9(9):e0004029. 47. Chlosta S, Fishman DS, Harrington L, et al. The iron efflux protein ferroportin regulates the intracellular growth of Salmonella enterica. Infect Immun. 2006;74(5):3065-3067. 48. Nairz M, Weiss G. Iron in infection and immunity. Mol Aspects Med. 2020;75:100864. 49. Steele-Mortimer O, Meresse S, Gorvel JP, Toh BH, Finlay BB. Biogenesis of Salmonella typhimurium-containing vacuoles in epithelial cells involves interactions with the early endocytic pathway. Cell Microbiol. 1999;1(1):33-49. 50. World Health Organisation. WHO child growth standards based on length/height, weight and age. Acta Paediatr Suppl. 2006;450:76-85.

Haematologica | 107 July 2022

1598

ARTICLE - Myeloproliferative Disorders

Retrospective analysis of pacritinib in patients with myelofibrosis and severe thrombocytopenia Srdan Verstovsek,1 Ruben Mesa,2 Moshe Talpaz,3 Jean-Jacques Kiladjian,4 Claire N. Harrison,5 Stephen T. Oh,6 Alessandro M. Vannucchi,7 Raajit Rampal,8 Bart L. Scott,9 Sarah A. Buckley,10 Adam R. Craig,10 Karisse Roman-Torres10 and John O. Mascarenhas11 MD Anderson Cancer Center, Houston, TX, USA; 2Mayo Clinic, Scottsdale, AZ, USA; 3 University of Michigan, Comprehensive Cancer Center, Ann Arbor, MI, USA; 4Hôpital SaintLouis, Université Paris Diderot, Paris, France; 5Guy's and St Thomas' NHS Foundation Trust, London, UK; 6Washington University School of Medicine, St. Louis, MO, USA; 7University of Florence, Azienda Ospedaliera-Universitaria Careggi, Florence, Italy; 8Memorial Sloan Kettering Cancer Center, New York, NY, USA; 9Fred Hutchinson Cancer Research Center, Seattle, WA, USA; 10CTI BioPharma Inc, Seattle, WA, USA and 11Icahn School of Medicine at Mount Sinai, New York, NY, USA 1

Correspondence: Srdan Verstovsek sverstov@mdanderson.org Received: June 28, 2021. Accepted: September 15, 2021. Prepublished: September 23, 2021. https://doi.org/10.3324/haematol.2021.279415 ©2022 Ferrata Storti Foundation Haematologica material is published under a CC BY-NC license

Abstract Thrombocytopenia is common in patients with myelofibrosis (MF) and is a well-established adverse prognostic factor. Both of the approved Janus kinase (JAK) inhibitors, ruxolitinib and fedratinib, can worsen thrombocytopenia and have not been evaluated in patients with severe thrombocytopenia (platelet counts <50×109/L). Pacritinib, a novel JAK2/interleukin-1 receptor-associated kinase 1 inhibitor, has been studied in two phase III trials (PERSIST-1 and PERSIST2), both of which enrolled patients with MF and severe thrombocytopenia. In order to better characterize treatment outcomes for this population with advanced disease, we present a retrospective analysis of efficacy and safety data in the 189 patients with severe thrombocytopenia treated in the PERSIST studies. The proportion of patients in the pacritinib group meeting efficacy endpoints was greater than in the BAT group for ≥35% spleen volume reduction (23% vs. 2%, P=0.0007), ≥50% modified Total Symptom Score reduction (25% vs. 8%, P=0.044), and self-reported symptom benefit (“much” or “very much” improved; 25% vs. 8%, P=0.016) at the primary analysis time point (week 24). The adverse event profile of pacritinib was manageable, and dose modification was rarely required. There was no excess in bleeding or death in pacritinib-treated patients. These results indicate that pacritinib is a promising treatment for patients with MF who lack safe and effective therapeutic options due to severe thrombocytopenia.

Introduction Patients with myelofibrosis (MF) who have severe thrombocytopenia (platelet counts <50×109/L) comprise a subset of patients with cytopenic MF who generally have more advanced disease, including anemia, greater risk of bleeding, worse symptom burden, higher risk of leukemic transformation, and shorter survival (median 15 months) compared with patients with higher platelet counts.1-3 These patients lack effective treatment options and are often excluded from clinical trials. Neither ruxolitinib nor fedratinib, the only drugs currently approved for MF, has been studied in patients presenting with severe thrombocytopenia, and neither drug has a product label with a recommended starting dose for this population.4-9 Furthermore, both have been shown to cause treatment-related thrombocytopenia, which requires dose modification and may result in reduced efficacy. For example, patients treated with ruxolitinib at ≤10 mg twice a day (BID) were less likely to achieve

significant spleen volume responses.10 Development of cytopenias was the most common reason for patients discontinuing ruxolitinib.11 Ruxolitinib has been tested in patients with moderate thrombocytopenia (platelet counts 50-100×109/L). In the phase Ib EXPAND study, dose interruptions or reductions were required in 89% of patients who had platelet counts 50-74×109/L, and 78% experienced grade 3 or 4 thrombocytopenia as an adverse event.12 In the phase IIIb expanded-access JUMP study, 55% of patients who started on ruxolitinib 5 mg BID required further dose reduction, and grade 3 or 4 thrombocytopenia was a common adverse event.13 There is a significant unmet need for effective and safe therapies for patients living with MF and experiencing severe thrombocytopenia, who may comprise up to 35% of the MF population.14 Pacritinib is a novel inhibitor of Janus kinase 2 (JAK2) and interleukin-1 receptor-associated kinase 115 currently in development for patients with MF and thrombocytopenia. Two randomized controlled phase III trials, PERSIST-116 and

Haematologica | 107 July 2022

1599

ARTICLE - Pacritinib in thrombocytopenic myelofibrosis PERSIST-2,17 have compared the efficacy and safety of pacritinib versus best available therapy (BAT). Because earlier phase I-II studies had shown that pacritinib has limited myelosuppressive properties,18 neither phase III study had a lower limit on platelet counts, and both included patients with moderate and severe thrombocytopenia. The PERSIST studies represent the largest published data set of patients with MF and severe thrombocytopenia treated in randomized controlled trials. In order to better characterize treatment outcomes for this population with advanced disease, we present pooled data from patients with severe thrombocytopenia treated in PERSIST-1 and PERSIST-2 and describe the efficacy and safety profiles of pacritinib compared to BAT.

Methods The PERSIST-1 (clinicaltrials gov. Identifier: NCT01773187) and PERSIST-2 (clinicaltrials gov. Identifier: NCT02055781) study designs and methodology have been previously described.16,17 Key features of both studies are summarized below. The study protocols were approved by the Institutional Review Boards at each study site and the study was conducted in accordance with the Declaration of Helsinki. All patients provided written informed consent. Both studies included adult patients with either primary or secondary MF. Patients had intermediate-1, intermediate-2, or high-risk disease, as categorized by the Dynamic International Prognostic Scoring System, and palpable splenomegaly ≥5 cm below the left costal margin. PERSIST-1 enrolled patients regardless of platelet count, while PERSIST-2 was restricted to patients with platelet counts ≤100×109/L. Both studies included patients with severe thrombocytopenia (platelet counts <50x109/L) at baseline; this population comprised 16% of patients in PERSIST-1 and 45% in PERSIST-2. Prior use of JAK inhibitors was permitted only in PERSIST-2. PERSIST-1 randomized patients 2:1 to receive pacritinib 400 mg daily or BAT. PERSIST-2 randomized patients 1:1:1 to receive pacritinib 400 mg daily, pacritinib 200 mg BID, or BAT. Randomization was stratified by baseline platelet count. BAT included any available physicianselected treatment, including “watch and wait” (i.e., no active treatment). Ruxolitinib was included as an option only in PERSIST-2. Fedratinib was not available as BAT in either study. Patients randomized to receive BAT were allowed to cross over to pacritinib at 24 weeks or at disease progression. Safety and efficacy data were censored at the time of crossover. Statistical analysis Patients in PERSIST-1 and PERSIST-2 with baseline platelet counts <50×109/L were included in the analysis. Efficacy endpoints were assessed at week 24 and included

S. Verstovsek et al. the percentage of patients achieving ≥35% spleen volume response (SVR), the percentage achieving ≥50% reduction in the modified Total Symptom Score (TSS) v2.0,19 and the percentage reporting symptoms as “much” or “very much” improved on the Patient Global Impression of Change scale. Cardiac and hemorrhagic events were defined using Standardized MedDRA Queries. Since the PERSIST-2 study was terminated prematurely due to a clinical hold, intention-to-treat (ITT) efficacy analyses included all randomized patients in PERSIST-1 and the 71% in PERSIST-2 who were randomized at least 22 weeks prior to the hold. As the TSS instrument administered during PERSIST-1 was changed from v1.0 to v2.0 part-way through the study, only patients who had completed v2.0 at baseline were included in the ITT TSS analysis. Safety analyses included all treated patients (Online Supplementary Figure S1). Differences in baseline characteristics between groups were evaluated using the chi-square test (categorical variable) or the Wilcoxon rank-sum test (continuous variables). The Breslow and Day homogeneity test was performed to measure the magnitude of treatment effect on the response in the presence of a prior JAK or MF diagnosis. For efficacy outcomes, Fisher’s exact test was used to perform between-group categorical analysis; the Wilcoxon exact test was used for continuous outcome variables.

Results Patient characteristics In total, 192 patients (133 pacritinib, 59 BAT) with severe thrombocytopenia were enrolled in PERSIST-1 and PERSIST-2. Of these, 189 (132 pacritinib, 57 BAT) received at least one dose of study drug. There were 152 patients (104 pacritinib, 48 BAT) included in the ITT efficacy population, and 117 patients (80 pacritinib, 37 BAT) completed the TSS v2.0 at baseline (Online Supplementary Figure S1). As shown in Table 1, median age was 69 years (range, 50–91). The majority of patients (72%) had primary MF, with a median time from diagnosis of 2.0 years. Approximately one-third of patients (34%) had received prior treatment with a JAK2 inhibitor. Median platelet count at baseline was 28×109/L, 63.5% had hemoglobin <10 g/dL, and 48% had ≥1% peripheral blood blasts. Approximately half of the patients (49%) had grade 3 marrow fibrosis, and 38% had low or normal marrow cellularity (≤40%). The most common therapies selected as BAT were “watch and wait” (37%; with 25% receiving only “watch and wait” for the duration of the study), ruxolitinib (30%; only available for PERSIST-2), hydroxyurea (28%), and prednisone (12%). The duration of study drug exposure was similar for pacritinib and BAT (median 5.5 and 5.2 months, respectively), although 33% of patients on BAT cycled through multiple therapies on study. Shorter treatment durations were due,

Haematologica | 107 July 2022

1600

ARTICLE - Pacritinib in thrombocytopenic myelofibrosis

S. Verstovsek et al.

in part, to truncation at the time of the clinical hold, as mained 400 mg at both weeks 12 and 24, whereas patients 46% of patients were still on pacritinib at the time of the in PERSIST-2 who received ruxolitinib as BAT were preclinical hold. The median total daily dose of pacritinib re- scribed a median post-titration dose of 10 mg BID and

Table 1. Baseline patient and disease characteristics in patients treated with pacritinib or best available therapy

Age, median (range)

Pacritinib (N = 132)

BAT (N = 57)

P-value**

69 (50-91)

69 (50-84)

0.95

80 (61)

28 (49)

0.14

100/132 (76) 32/132 (24)

42/55 (76) 13/55 (24)

43 (33)

21 (37)

Male sex, N (%) ECOG PS, N* (%) 0-1 2-3

0.93

Prior JAK2 inhibitor, N (%) MF diagnosis, N (%) Primary MF PPV-MF PET-MF

0.17 98 (74) 20 (15) 14 (11)

38 (67) 8 (14) 22 (39)

2.0 (0–27)

2.6 (0–14)

26 (20) 63 (48) 43 (33)

5 (9) 30 (53) 22 (39)

Reticulin and collagen fibrosis staging, N* (%) MF 0-1 MF 2 MF 3

18/122 (15) 38/122 (31) 66/122 (54)

11/52 (21) 15/52 (29) 26/52 (50)

Bone marrow cellularity, N* (%) <20% 20-40% 41-100%

27/110 (25) 18/110 (16) 65/110 (59)

18/49 (37) 8/49 (16) 23/49 (47)

Bone marrow blast category, N* (%) ≥1% <1%

96/115 (84) 19/115 (17)

41/50 (82) 11/50 (18)

Time since MF diagnosis (years), median (IQR) DIPSS risk category, N (%) Intermediate-1 Intermediate-2 High

Peripheral blood blasts category, N* (%) ≥1% <1%

0.57

0.72 0.17

0.58

0.26

0.82

0.18 60/118 (51) 58/118 (49)

31/50 (62) 19/50 (38)

29 (6-49)

25 (5-49)

0.27

85/132 (64)

35/56 (63)

0.80

38 (29) 61 (46) 33 (25)

20 (35) 23 (40) 14 (25)

Spleen volume at baseline (cm3)‡, median (IQR)

2,566 (1,633-3,680)

2,466 (1,786-3,727)

0.87

Modified TSS score at baseline‡, median (IQR)

17 (12-29)

7 (12-27)

0.94

35 (27) 97 (73)

15 (26) 42 (74)

Platelet count (109/L), median (range) Hemoglobin <10 g/dL, N* (%) RBC transfusion dependence†, N (%) Dependent Independent Indeterminate

0.66

Study enrollment, N (%) PERSIST-1 PERSIST-2

0.98

BAT: best available therapy; DIPSS: Dynamic International Prognostic Scoring System; ECOG PS: Eastern Cooperative Oncology Group performance status; IQR: interquartile range; JAK2: Janus kinase 2; MF: myelofibrosis; PET: post-essential thrombocythemia; PPV: post-polycythemia vera; RBC: red blood cell; TSS: Total Symptom Score. *Denominators represent non-missing values. **Chi-square test for categorical variables and Wilcoxon rank-sum test for continuous variables. †Per Gale criteria.24 ‡Baseline values reported for efficacy population. Haematologica | 107 July 2022

1601

ARTICLE - Pacritinib in thrombocytopenic myelofibrosis

S. Verstovsek et al.

were on this treatment for a median duration of only 3.45 with higher response rates regardless of whether patients months. had received prior treatment with JAK2 inhibitors or whether they had primary versus secondary MF (Table 2). Efficacy The percentage of patients treated with pacritinib who Safety achieved a 24-week ≥35% SVR response was greater than The treatment-emergent adverse events (TEAE) that dethe percentage of patients treated with BAT (23.1% vs. 2.1%, veloped in patients with severe thrombocytopenia were P=0.0007) (Figures 1 and 2A). Higher response rates were consistent with the overall PERSIST study results and were observed in patients treated with pacritinib 200 mg BID generally grade 1 or 2 in severity (Table 3). The most comcompared with those on 400 mg daily (29.0% vs. 20.5%). mon non-hematologic TEAE in pacritinib-treated patients The median percentage change in spleen volume was were diarrhea (60.6%; 5.3% grade 3-4), nausea (30.3%; 1.5% greater for pacritinib-treated patients than for BAT-treated grade 3-4), and vomiting (26.5%; 0.8% grade 3-4). These patients (-29.4% vs. -1.3%, P<0.0001), and the percentage gastrointestinal events were observed more frequently in of patients who experienced any improvement (>0%) in the pacritinib group but rarely led to dose reduction (3.0% SVR was higher in the pacritinib compared with the BAT for diarrhea, 1.5% for nausea) or discontinuation (3.8% for group (56.7% vs. 33.3%, P=0.0089). Similarly, pacritinib- diarrhea, 0% for nausea). The most common hematologic treated patients were more likely to achieve ≥50% mod- TEAE in this thrombocytopenic population among pacritiified TSS reduction at week 24 compared to BAT-treated nib-treated patients were thrombocytopenia (34.8%) and patients (25% vs. 8.1%, P=0.0441) (Figures 1 and 2B), and anemia (31.8%). Hematologic TEAE were generally grade 3 the median reduction in modified TSS score was greater or 4 given the degree of cytopenias present at baseline, with pacritinib than with BAT (-30.3% vs 0%, P=0.0036). but these rarely led to dose reduction (4.5% and 2.3% for Response rates for TSS were similar between the two pa- thrombocytopenia and anemia, respectively) or discontinucritinib doses. The percentage of patients who experi- ation (3.8% and 3.8%, respectively). Among patients reenced any improvement (>0%) in TSS was higher in the maining on study, hemoglobin and platelet counts were pacritinib compared to the BAT group (53% vs. 32%, stable through week 24 (Figure 4). While thrombocytopenia P=0.049). Pacritinib-treated patients were significantly was observed more often on pacritinib, there was no exmore likely than patients treated with BAT to report that cess of hemorrhagic events (pacritinib vs. BAT: grade ≥1, their symptoms were “very much” or “much” improved at 51.5% vs. 59.6%; grade 3-4, 13.6% vs. 10.5%; fatal, 2% vs. week 24 (25.0% vs. 8.3%, respectively; P=0.016) (Figure 3). 0%). High-grade and fatal cardiac events were observed at Subgroup analyses showed that pacritinib was associated similar rates on pacritinib and BAT (grade 3-4, 9.1% vs. 14%;

Figure 1. Efficacy of pacritinib versus best available therapy based on 24-week response rates in patients with severe thrombocytopenia. Graph depicts the percentage of patients achieving ≥35% spleen volume reduction (SVR), achieving ≥50% reduction in modified Total Symptom Score (TSS), and reporting symptoms as being “much” or “very much” improved based on Patient Global Impression of Change (PGIC) at week 24. Percentages are based on all patients randomized at least 22 weeks prior to the termination of the PERSIST studies (intention-to-treat [ITT] population). BAT: best available therapy; CI: confidence interval; PAC: pacritinib. Haematologica | 107 July 2022

1602

ARTICLE - Pacritinib in thrombocytopenic myelofibrosis

S. Verstovsek et al.

Figure 2. Waterfall plots of percentage change from baseline. (A) Change in spleen volume and (B) change in modified Total Symptom Score (TSS) at week 24 in patients with severe thrombocytopenia. Data are shown for evaluable patients treated with pacritinib (pooled dose groups) or best available therapy (BAT) (including ruxolitinib, indicated with red asterisks). Gray horizontal lines indicate responder threshold (35% for spleen volume reduction [SVR], 50% for TSS).

Table 2. Spleen volume response and modified Total Symptom Score response rates among patients randomized to pacritinib versus best available therapy by subgroup: prior exposure to a JAK2 inhibitor (including ruxolitinib) and myelofibrosis subtype (primary vs. secondary after a prior diagnosis of polycythemia vera or essential thrombocythemia)

Response rate at week 24

Pacritinib

BAT

P-value*

Patients with ≥35% spleen volume reduction, % (n/N) Prior JAK2 inhibitor exposure Yes No MF diagnosis Primary Secondary

0.07 17.9 (5/28) 17.9 (5/28)

7.7 (1/13) 0 (0/35) 0.52

24.0 (18/75) 20.7 (6/29)

3.1 (1/32) 0 (0/16)

Patients with ≥50% reduction in modified TSS, % (n/N) Prior JAK2 inhibitors Yes No MF diagnosis Primary Secondary

17.9 (5/28) 28.8 (15/52) 30.4 (17/56) 12.0 (3/25)

15.4 (2/13) 4.2 (1/24) 12.5 (3/24) 0 (0/12)

0.14

0.43

Patients with “much” or “very much” improved PGIC scores, % (n/N) Prior JAK2 inhibitors Yes No MF diagnosis Primary Secondary

14.3 (4/28) 28.9 (22/76)

15.4 (2/13) 5.7 (2/35)

0.08

0.45 30.7 (23/75) 10.3 (3/29)

12.5 (4/32) 0 (0/16)

BAT: best available therapy; JAK2: Janus kinase 2; MF: myelofibrosis; PGIC: Patient Global Impression of Change; SVR: spleen volume reduction; TSS: Total Symptom Score. *Breslow and Day homogeneity test Haematologica | 107 July 2022

1603

ARTICLE - Pacritinib in thrombocytopenic myelofibrosis

S. Verstovsek et al.

fatal, 3% vs. 1.8%). Survival was similar between pacritinib- safe and effective therapies. Efficacy was observed irreand BAT-treated patients: hazard ratio (HR): 1.01 (95% con- spective of prior JAK2 inhibitor exposure or MF subtype fidence interval [CI]: 0.57-1.80). (primary or secondary), although SVR and symptom response rates were numerically higher in patients who did not have prior JAK inhibitor exposure. Efficacy in this subgroup of patients with severe thrombocytopenia was simiDiscussion lar to that observed in the PERSIST studies overall, In this retrospective analysis, the pacritinib group was as- including in patients with higher platelet counts. The pasociated with improved SVR and symptom response com- tient population described in this analysis had advanced pared to BAT in patients with MF and severe disease: in addition to severe thrombocytopenia (median thrombocytopenia, a population with an unmet need for platelet count 28x109/L), about half of the patients had

Figure 3. Self-reported symptoms in patients who completed the Patient Global Impression of Change at week 24 by treatment group. The percentage of evaluable patients with any improvement in disease symptoms was higher for patients randomized to pacritinib (84% [47/56]) than for those randomized to best available therapy (BAT) (48% [10/21]).

Figure 4. Median hemoglobin and platelet count over time through week 24. Among patients remaining on study, the median hemoglobin (A) and platelet count (B) remained stable over time in both pacritinib- and best available therapy (BAT)-treated patients. IQR: interquartile range. Haematologica | 107 July 2022

1604

ARTICLE - Pacritinib in thrombocytopenic myelofibrosis

S. Verstovsek et al.

Table 3. Most common treatment-emergent adverse events (≥10% all grade or ≥3% grade 3 or 4 in either group) in all treated patients.*

TEAE, N (%)

Pacritinib (N = 132)

BAT (N = 57)

All grade

Grade 3-4

All grade

Grade 3-4

Diarrhea

80 (60.6)

7 (5.3)

9 (15.8)

Thrombocytopenia

46 (34.8)

12 (21.1)

Anemia

42 (31.8)

12 (21.1)

11 (19.3)

Nausea

40 (30.3)

2 (1.5)

7 (12.3

1 (1.8)

Vomiting

35 (26.5)

1 (0.8)

4 (7.0)

1 (1.8)

Epistaxis

21 (15.9)

9 (6.8)

15 (26.3)

1 (1.8)

Peripheral edema

21 (15.9)

2 (1.5)

13 (22.8)

Fatigue

19 (14.4)

6 (4.5)

6 (10.5)

3 (5.3)

Dizziness

18 (13.6)

2 (1.5)

2 (3.5)

Pyrexia

17 (12.9)

6 (10.5)

Constipation

17 (12.9)

1 (0.8)

4 (7.0)

Abdominal pain

16 (12.1)

2 (1.5)

10 (17.5)

1 (1.8)

Dyspnea

14 (10.6)

2 (1.5)

5 (8.8)

2 (3.5)

Pneumonia

14 (10.6)

10 (7.6)

2 (3.5)

Decreased appetite

14 (10.6)

3 (2.3)

4 (7.0)

Upper respiratory tract infection

12 (9.1)

6 (10.5)

1 (1.8)

Contusion

13 (9.8)

6 (10.5)

Cough

10 (7.6)

1 (0.8)

7 (12.3)

Neutropenia

8 (6.1)

7 (5.3)

4 (7.0)

Leukopenia

7 (5.3)

5 (3.8)

2 (3.5)

Cardiac failure

5 (3.8)

3 (5.3)

2 (3.5)

Atrial fibrillation

2 (1.5)

1 (0.8)

4 (7.0)

2 (3.5)

General health deterioration

4 (3.0)

Lower respiratory tract infection

4 (3.0)

2 (3.5)

Sepsis

4 (3.0)

2 (1.5)

3 (5.3)

2 (3.5)

Abdominal pain, upper

7 (5.3)

1 (0.8)

3(5.3)

2 (3.5)

BAT: best available therapy; TEAE: treatment-emergent adverse event. *Events were counted regardless of whether they were considered related to study drug. Disease progression as an adverse event is not listed.

circulating blasts ≥1%, and two-thirds had significant anemia (hemoglobin <10 g/dL). These findings are consistent with previous reports describing the co-occurrence of these poor prognostic factors in patients with severe thrombocytopenia.2,3 Furthermore, the patients presented here had significant burden of disease, with spleen volumes consistent with those reported in patients with higher platelet counts.6,8,20 Interestingly, the majority of patients had hypo- or normocellular bone marrow, as opposed to the hypercellular marrow typically associated with myeloproliferative disease.

Pacritinib was tolerated at full doses in patients with severe thrombocytopenia, and the safety profile was consistent with that observed in the PERSIST studies overall, although rates of bleeding were higher in patients with severe thrombocytopenia regardless of whether they were treated with pacritinib or BAT. The most common adverse events were gastrointestinal, and these were predominantly low grade and manageable with anti-diarrheals. Despite severe thrombocytopenia at baseline, discontinuation due to myelosuppression was rare for patients on pacritinib. Furthermore, hemorrhagic events

Haematologica | 107 July 2022

1605

ARTICLE - Pacritinib in thrombocytopenic myelofibrosis were observed at similar frequencies for pacritinib and BAT, suggesting that bleeding is more likely associated with disease-related thrombocytopenia21 and platelet dysfunction.22 In summary, pacritinib was shown to be more effective than BAT in reducing splenomegaly and symptom burden in patients with MF and severe thrombocytopenia. While this analysis is post hoc and retrospective, the results highlight the important role that pacritinib may play in the future therapeutic landscape for patients living with cytopenic MF. In the recent PAC203 phase II dose-finding study conducted in patients with advanced and heavily pretreated MF, including those with severe thrombocytopenia, the SVR response rate for patients with severe thrombocytopenia was 17%,23 similar to that observed in this review (23%); differences could be attributed to the enrollment of patients with prolonged duration of prior ruxolitinib exposure on PAC203 (median duration of exposure was 2.1 years). The randomized phase III PACIFICA study is currently under way in patients with MF and severe thrombocytopenia comparing pacritinib 200 mg BID with physicians’ choice of therapy. Results from PACIFICA should confirm whether pacritinib will be a new therapeutic option for patients with cytopenic MF.

S. Verstovsek et al. RR reports consulting fees from Constellation, Incyte, Celgene/BMS, Promedior, CTI Biopharma, Jazz Pharmaceuticals, Blueprint, Stemline, Galecto, Pharmaessentia, Kartos Therapeutics, and AbbVie; and research funding from Incyte, Constellation, and Stemline. BLS reports serving on advisory committees for Celgene/BMS and Alexion; and research funding from Novartis. SAB, ARC, and KR-T are employees of CTI BioPharma, and ARC is a stockholder in CTI BioPharma, Inc. JOM reports research funding paid to his institution from Incyte, Roche, Forbius, CTI BioPharma, Merck, AbbVie, Novartis, Geron, Kartos Therapeutics, and PharmaEssentia; and serving in a consulting or advisory role for Celgene/BMS, Constellation, Incyte, Novartis, Roche, Sierra, CTI BioPharma, Genentech/Roche, Gilead Sciences, Promedior, PharmaEssentia, KartosTherapeutics, and AbbVie. Contributions RM, MT, J-JK, CNH, SV, and JOM contributed to the study design; SB and KR-T collected the data; all authors performed data analysis and interpretation; and all authors participated in manuscript preparation and approval of the final version.

Acknowledgments Disclosures Editing assistance was provided by Maxine Skipp, of Twist SV reports consulting for CTI BioPharma. RM reports con- Medical, LLC. This assistance was funded by TMAC Advisors. sulting for Novartis, Sierra Oncology, La Jolla Pharma, and AOP; and research support from Incyte, Gilead, CTI Bio- Data-sharing statement Pharma, Celgene, AbbVie, and Genentech. MT reports serv- CTI is committed to enhancing appropriate transparency of ing on advisory boards for Novartis, BMS, and research in a responsible manner and improving access to Constellation; and conducting clinical studies supported by clinical trial information in order to inform medical deciNovartis, Constellation, and BMS. J-JK reports serving on sion-making, advance scientific discovery and accelerate advisory boards for Novartis, BMS/Celgene, AbbVie, Incyte, development of new treatments to benefit patients. This AOP Orphan, and CTI BioPharma. CNH reports research includes data sharing, registration and results reporting, funding from Novartis, Celgene, and Constellation; and in- voluntary disclosures and lay/plain language summaries. stitutional and speaker funding from Novartis, Celgene, CTI Clinical trial results for PERSIST 2 (NCT02055781) and BioPharma, Gilead, Shire, Roche, Janssen, Promedior, PAC203 (NCT04884191) have been previously published Geron, Galacteo, and AOP. STO reports consulting/advisory (Mascarenhas et al. JAMA Oncol. 2018;4(5):652-659 and boards for Disc Medicine, Blueprint Medicines, PharmaEs- Gerds et al. Blood Adv. 2020;4(22): 5825-5835) and have sentia, Constellation, Geron, AbbVie, Sierra Oncology, In- been or will be posted to ClinicalTrials.gov and other public cyte, Kartos Therapeutics, CTI Biopharma, and registries as required. Inquiries regarding availability of Celgene/BMS. AMV reports serving on advisory boards for data and clinical trial documentation will be considered on Novartis, Incyte, AbbVie, and Celgene/BMS; and lecture fees a case-by-case basis and should be directed to: from Novartis, CTI BioPharma, Celgene/BMS, and AbbVie. Medinfo@ctibiopharma.com.)

References 1. Scotch AH, Kosiorek H, Scherber R, et al. Symptom burden profile in myelofibrosis patients with thrombocytopenia: Lessons and unmet needs. Leuk Res. 2017;63:34-40. 2. Hernandez-Boluda JC, Correa JG, Alvarez-Larran A, et al. Clinical characteristics, prognosis and treatment of myelofibrosis

patients with severe thrombocytopenia. Br J Haematol. 2018;181(3):397-400. 3. Masarova L, Alhuraiji A, Bose P, et al. Significance of thrombocytopenia in patients with primary and postessential thrombocythemia/polycythemia vera myelofibrosis. Eur J

Haematologica | 107 July 2022

1606

ARTICLE - Pacritinib in thrombocytopenic myelofibrosis

S. Verstovsek et al.

Haematol. 2018;100(3):257-263. 4. Jakafi (ruxolitinib) [package insert]. Wilmington, DE: Incyte; 2020. 5. Inrebic (fedratinib) [package insert]. Summit, NJ: Celgene Corporation; 2019. 6. Verstovsek S, Mesa RA, Gotlib J, et al. A double-blind, placebocontrolled trial of ruxolitinib for myelofibrosis. N Engl J Med. 2012;366(9):799-807. 7. Harrison CN, Vannucchi AM, Kiladjian JJ, et al. Long-term findings from COMFORT-II, a phase 3 study of ruxolitinib vs best available therapy for myelofibrosis. Leukemia. 2016;30(8):1701-1707. 8. Pardanani A, Harrison C, Cortes JE, et al. Safety and efficacy of fedratinib in patients with primary or secondary myelofibrosis: a randomized clinical trial. JAMA Oncol. 2015;1(5):643-651. 9. Harrison CN, Schaap N, Vannucchi AM, et al. Janus kinase-2 inhibitor fedratinib in patients with myelofibrosis previously treated with ruxolitinib (JAKARTA-2): a single-arm, open-label, non-randomised, phase 2, multicentre study. Lancet Haematol. 2017;4(7):e317-e324. 10. Center for Drug Evaluation and Research: Ruxolitinib Clinical Pharmacology and Biopharmaceutics Review (NDA 202192). https://www.accessdata.fda.gov/drugsatfda_docs/nda/2011/2021 92Orig1s000ClinPhar. 11. Kuykendall AT, Shah S, Talati C, et al. Between a rux and a hard place: evaluating salvage treatment and outcomes in myelofibrosis after ruxolitinib discontinuation. Ann Hematol. 2018;97(3):435-441. 12. Guglielmelli P, Kiladjian JJ, Vannucchi A, et al. The final analysis of EXPAND: a phase 1b, open-label, dose-finding study of ruxolitiminb in patients with myelofibrosis and low platelet count (50 x 109/L to <100 x 109/L) at baseline. Blood. 2020;136(Suppl 1):S4-5. 13. Al-Ali HK, Griesshammer M, Foltz L, et al. Primary analysis of JUMP, a phase 3b, expanded-access study evaluating the safety and efficacy of ruxolitinib in patients with myelofibrosis, including those with low platelet counts. Br J Haematol. 2020;189(5):888-903. 14. Masarova L, Mesa RA, Hernández-Boluda JC, Taylor JA. Severe

thrombocytopenia in myelofibrosis is more prevalent than previously reported. Leuk Res. 2020;91:106338. 15. Singer JW, Al-Fayoumi S, Ma H, Komrokji RS, Mesa R, Verstovsek S. Comprehensive kinase profile of pacritinib, a nonmyelosuppressive Janus kinase 2 inhibitor. J Exp Pharmacol. 2016;8:11-19. 16. Mesa RA, Vannucchi AM, Mead A, et al. Pacritinib versus best available therapy for the treatment of myelofibrosis irrespective of baseline cytopenias (PERSIST-1): an international, randomised, phase 3 trial. Lancet Haematol. 2017;4(5):e225-e236. 17. Mascarenhas J, Hoffman R, Talpaz M, et al. Pacritinib vs best available therapy, including ruxolitinib, in patients with myelofibrosis: a randomized clinical trial. JAMA Oncol. 2018;4(5):652-659. 18. Verstovsek S, Odenike O, Singer JW, Granston T, Al-Fayoumi S, Deeg HJ. Phase 1/2 study of pacritinib, a next generation JAK2/FLT3 inhibitor, in myelofibrosis or other myeloid malignancies. J Hematol Oncol. 2016;9(1):137. 19. Emanuel RM, Dueck AC, Geyer HL, et al. Myeloproliferative neoplasm (MPN) symptom assessment form total symptom score: prospective international assessment of an abbreviated symptom burden scoring system among patients with MPNs. J Clin Oncol. 2012;30(33):4098-4103. 20. Harrison C, Kiladjian JJ, Al-Ali HK, et al. JAK inhibition with ruxolitinib versus best available therapy for myelofibrosis. N Engl J Med. 2012;366(9):787-798. 21. Hasselbalch H. Idiopathic myelofibrosis: a clinical study of 80 patients. Am J Hematol. 1990;34(4):291-300. 22. Leoni P, Rupoli S, Lai G, et al. Platelet abnormalities in idiopathic myelofibrosis: functional, biochemical and immunomorphological correlations. Haematologica. 1994;79(1):29-39. 23. Gerds AT, Savona MR, Scott BL, et al. Determining the recommended dose of pacritinib: results from the PAC203 dose-finding trial in advanced myelofibrosis. Blood Adv. 2020;4(22):5825-5835. 24. Gale RP, Barosi G, Barbui T, et al. What are RBC-transfusiondependence and -independence? Leuk Res. 2011;35(1):8-11.

Haematologica | 107 July 2022

1607

ARTICLE - Non-Hodgkin Lymphoma

Phase I/II clinical trial of temsirolimus and lenalidomide in patients with relapsed and refractory lymphomas Ajay Major,1 Justin Kline,1 Theodore G. Karrison,1 Paul A. S. Fishkin,2 Amy S. Kimball,3,4 Adam M. Petrich,5,6 Sreenivasa Nattam,7 Krishna Rao,8 Bethany G. Sleckman,9 Kenneth Cohen,1 Koen van Besien,10 Aaron P. Rapoport3 and Sonali M. Smith1 University of Chicago, Chicago, IL; 2Illinois Cancer Care, Peoria, IL; 3University of Maryland School of Medicine and Marlene and Stewart Greenebaum Comprehensive Cancer Center, Baltimore, MD; 4Amgen Inc., Thousand Oaks, CA; 5Northwestern University, Chicago, IL; 6 Daiichi-Sankyo, Basking Ridge, NJ; 7Fort Wayne Oncology/Hematology, Fort Wayne, IN; 8 Southern Illinois University, Springfield, IL; 9Mercy Hospital, St. Louis, IL and 10Weill Cornell Medicine, New York, NY, USA

Correspondence: Sonali M. Smith smsmith@medicine.bsd.uchicago.edu

Received: April 18, 2021. Accepted: July 22, 2021. Prepublished: July 29, 2021. https://doi.org/10.3324/haematol.2021.278853 ©2022 Ferrata Storti Foundation Haematologica material is published under a CC BY-NC license

Abstract The PI3K/Akt/mTOR (PAM) axis is constitutively activated in multiple lymphoma subtypes and is a promising therapeutic target. The mTOR inhibitor temsirolimus (TEM) and the immunomodulatory agent lenalidomide (LEN) have overlapping effects within the PAM axis with synergistic potential. This multicenter phase I/II study evaluated combination therapy with TEM/LEN in patients with relapsed and refractory lymphomas. Primary endpoints of the phase II study were rates of complete (CR) and overall response (ORR). There were 18 patients in the phase I dose-finding study, and TEM 25 mg weekly and LEN 20 mg on day 1 through day 21 every 28 days was established as the recommended phase II dose. An additional 93 patients were enrolled in the phase II component with three cohorts: diffuse large B-cell lymphoma (DLBCL, n=39), follicular lymphoma (FL, n=15), and an exploratory cohort of other lymphoma histologies with classical Hodgkin lymphoma (cHL) comprising the majority (n=39 total, n=20 with cHL). Patients were heavily pretreated with a median of four (range, 1-14) prior therapies and one-third with relapse following autologous stem cell transplantation (ASCT); patients with cHL had a median of six prior therapies. The FL cohort was closed prematurely due to slow accrual. ORR were 26% (13% CR) and 64% (18% CR) for the DLBCL and exploratory cohorts, respectively. ORR for cHL patients in the exploratory cohort, most of whom had relapsed after both brentuximab vedotin and ASCT, was 80% (35% CR). Eight cHL patients (40%) proceeded to allogeneic transplantation after TEM/LEN therapy. Grade ≥3 hematologic adverse events (AE) were common. Three grade 5 AE occurred. Combination therapy with TEM/LEN was feasible and demonstrated encouraging activity in heavily-pretreated lymphomas, particularly in relapsed/refractory cHL (clinicaltrials gov. Identifier: NCT01076543).

Introduction Classical Hodgkin lymphoma (cHL) and non-Hodgkin lymphomas (NHL) are typically chemosensitive to early lines of therapy, but relapse is a frequent and often lifethreatening event. Development of novel non-chemotherapy agents for the treatment of lymphoma, including targeted and immunomodulatory agents, presents opportunities for disease control in a more rational approach. The PI3K/Akt/mTOR (PAM) signal transduction pathway is constitutively activated in lymphoma and is a promising therapeutic target that appears to be shared across biologically heterogeneous lymphoma subtypes.1-3 Mammalian target of rapamycin (mTOR) is a master regulator of growth and survival in normal and neoplastic cells.3 Activation of mTOR is regulated by upstream phosphatidylinositol-3,4,5 kinase (PI3K) and Akt signaling, which pro-

motes cell growth, cell survival and proliferation.4 Aberrant mTOR activation occurs via several mechanisms in NHL, including PTEN loss in mantle cell lymphoma (MCL), PIK3CA amplification in diffuse large B-cell lymphoma (DLBCL), and PKCd or Syk kinase activation in follicular lymphoma (FL).3 mTOR activation has also been demonstrated in cHL, likely mediated by mTORC1 and Akt.5,6 mTOR is a particularly attractive therapeutic target given its position as a common downstream regulator for several oncogenic pathways. The first generation mTOR inhibitor temsirolimus (TEM) is currently Food and Drug Administration approved for the treatment of metastatic renal cell carcinoma, and has demonstrated monotherapy activity in several lymphoma subtypes, including MCL, DLBCL, and FL.3,7,8,9 In addition to reliance on signal transduction pathways, the tumor microenvironment and immune composition contribute to lymphoma pathogenesis and may augment

Haematologica | 107 July 2022

1608

ARTICLE - Phase I/II TEM/LEN clinical trial

A. Major et al.

PAM axis deregulation.10,11 Lenalidomide (LEN) is an immune-modulatory agent which enables proteasomal degradation and downregulation of several transcription factors, which then function as oncogenes.12 Within the PAM axis, LEN inhibits Akt phosphorylation and VEGF translation.1,13 LEN is active in both NHL and cHL, and is frequently tested in combination regimens.14,15 Given the promising single-agent activity of both TEM and LEN and the potential for synergistic effects of the two agents on the PAM axis, we conducted a multicenter phase I/II study of combination TEM/LEN therapy in patients with relapsed and refractory lymphomas.

cytopenia for greater than 7 days (or associated with bleeding or requiring more than 1 platelet transfusion), ANC less than 500/mL for greater than 7 days despite growth factor administration, or any thromboembolic event. DLT was assessed after one cycle of TEM/LEN. The phase II study accrued patients into the three aforementioned cohorts: DLBCL, FL, and the exploratory cohort of other lymphoma histologies. Patients received therapy for up to 1 year, or until disease progression or development of toxicities requiring treatment cessation. Patients considered to be at high risk of developing venous thromboembolism received prophylactic aspirin or low molecular weight heparin.

Methods

Response and toxicity assessment criteria Eligible patients from the DLBCL, FL, and the exploratory cohort were assessed for response to therapy using the 2006 revised response criteria for lymphoma.16 Patients with Waldenström macroglobulinemia (WM) were assessed using the consensus recommendations for response.17 Response assessments were performed after cycle 2 (week 8), and then every 3 months thereafter. Confirmatory scans were recommended at least 4 weeks following initial documentation of an objective response. Toxicities were graded according to the National Cancer Institute Common Terminology Criteria for Adverse Events, Version 4. Patients were eligible for toxicity reporting if at least one dose of a study drug was administered. Patients were removed from the study if one of the following criteria applied: completion of 52 weeks of therapy, disease progression, unacceptable adverse events, study withdrawal, or eligibility for allogeneic transplantation. Patients who were candidates for allogeneic transplantation after progression on TEM/LEN proceeded directly to transplant without bridging therapy.

This study is an open-label phase I/II multicenter clinical trial of TEM/LEN combination therapy in patients with relapsed and refractory lymphomas. Weekly data and safety monitoring occurred through the University of Chicago Phase II consortium. This clinical trial was registered through the National Cancer Institute as protocol number 8309 (clinicaltrial gov. Identifier: NCT01076543). Study accrual occurred from 2010 to 2015. The protocol was approved by the University of Chicago Medical Center Institutional Review Board (IRB number 09-443-A). Patient selection and eligibility Patients with histologically-confirmed cHL and NHL treated with ≥ 1 prior cytotoxic regimen were eligible. There was no limit to the number of prior therapies allowed, and patients with prior autologous stem cell transplantation (ASCT) were eligible. Chronic lymphocytic leukemia/small lymphocytic lymphoma (CLL/SLL) was excluded due to poor efficacy of single-agent TEM observed previously in this disease.9 For the phase II component, patients were grouped into three cohorts: (i) DLBCL; (ii) FL; and (iii) an exploratory cohort of other lymphomas (including cHL, T-cell NHL [T-NHL], marginal zone lymphoma [MZL], lymphoplasmacytic lymphoma [LPL], and MCL). Full inclusion criteria are presented in the Online Supplementary Appendix. Study design and treatment plan For the phase I dose-finding study, TEM was administered intravenously weekly at a dose of 25 mg for all dose levels, and LEN was administered orally on day 1 through day 21 every 28 days at three dose levels: 15 mg, 20 mg, and 25 mg. One cycle was defined as 4 weeks, or 28 days. There was no intrapatient dose escalation. Treatment was administered on an outpatient basis. Patients were treated to intolerance, progression, or discontinuation at physician discretion. Dose-limiting toxicity (DLT) was defined as grade 3 or 4 non-hematologic toxicity, grade 4 thrombo-

Study endpoints and statistical analysis The phase I dose-finding study utilized a “3+3” design, and the phase II study accrued patients in a two-stage “minimax” design for each cohort.18 The primary endpoints of the phase II study were rates of complete (CR) and overall response (ORR), and secondary endpoints were duration of response (DOR), progression-free survival (PFS) and overall survival (OS), stratified by histology. PFS was defined as the time from study entry to progression or death from any cause. OS was defined as the time from study entry to death. DOR was defined as the time from the first documented date of response to the date of progression or death, whichever came first. PFS, OS, and DOR were estimated by the Kaplan-Meier method.19 Median time to event and associated 90% confidence intervals were determined using the procedure of Brookmeyer and Crowley.20 A full description of the null and alternative hypotheses is presented in the Online Supplementary Appendix.

Haematologica | 107 July 2022

1609

ARTICLE - Phase I/II TEM/LEN clinical trial

A. Major et al.

Results Phase I Of 18 patients enrolled in the phase I study, 13 were male and five were female, with a median age of 62 years (range, 41-80 years). Of these, 15 patients were evaluable for DLT assessment, with the remaining three patients non-evaluable due to one withdrawing consent before starting treatment, one withdrawing consent after one dose, and one dying of rapid disease progression after one dose. As shown in Table 1, there was one DLT at dose level 1 (grade 4 hypokalemia) and two DLT at dose level 3 (grade 3 diarrhea and grade 3 HSV mucositis). Other grade 3 or 4 adverse events not meeting DLT criteria are as follows, each occurring in one patient: hypokalemia, hypertriglyceridemia, vomiting, urinary tract infection, skin infection, nausea, hypoxia, hyponatremia, diarrhea, and hyperglycemia. Of the 18 patients, there were five with partial responses, three with stable disease, six with progressive disease, and four not adequately assessed. Per protocol, dose level 2 was thus established as the recommended phase II dose: TEM 25 mg weekly and LEN 20 mg on day 1 through day 21 every 28 days. Patients were treated for at least two consecutive 28-day cycles, and patients showing at least stable disease after two cycles were permitted to continue treatment for up to 52 weeks of therapy. Phase II The baseline characteristics for the 93 patients in the phase II study are displayed in Table 2, including 39 patients with DLBCL, 15 patients with FL, and 39 patients in the exploratory cohort. In the DLBCL cohort, six had a prior history of FL and three had a prior history of MZL. In the exploratory cohort, 20 had cHL, nine had T-NHL, five had MCL, four had MZL, and one had WM. Overall, there were 62 males and 31 females, with a median age of 57 years (range, 23-78 years). The cohort was very heavily pre-treated with four (range, 1-14) median prior treatments. A total of 31 patients (33%) had relapsed following ASCT. In the DLBCL and FL cohorts, all patients had previously received rituximab. In the exploratory cohort, all

patients with B-cell NHL had previously received rituximab. For the 20 cHL patients in the exploratory cohort, the median number of prior treatments was six (range, 314), 19 (95%) had progressed after previous treatment with brentuximab vedotin (BV), and 15 (75%) had progressed after previous ASCT. The FL cohort was closed prematurely due to slow accrual. The CONSORT diagram for the phase II study is displayed in Online Supplementary Figure S1. Primary and secondary endpoints Primary and secondary endpoints were analyzed for all 93 patients in the phase II study on an intention-to-treat basis. The rates of CR and ORR, and median DOR, PFS, OS, and follow-up for all three cohorts are displayed in Table 3. A waterfall plot of best responses from baseline is displayed in Figure 1. The DLBCL and exploratory cohorts achieved a sufficient number of responses in the first stage to proceed to the second stage of the phase II trial. The FL cohort was terminated due to lack of accrual. A swimmer’s plot of treatment duration, best responses, and follow-up for the DLBCL and cHL cohorts is displayed in Figure 2. In the DLBCL cohort, the ORR was 25.6% with 12.8% achieving CR. Twenty-two of the 39 DLBCL patients had germinal center B-cell-like (GCB) DLBCL, and three of those patients responded. Eight of the 39 DLBCL patients had activated B-cell-like (ABC) DLBCL, and seven of those patients responded. No patients with transformed lymphomas responded to TEM/LEN treatment. The total number of responders, ten of 39, was not sufficient to reject the null hypothesis of a 30% response rate. The median DOR of the DLBCL cohort was 13.8 months (90% confidence interval [CI]: 4.1-19.0 months). The median DOR was 4.1 months (90% CI: 2.6 months - not estimable) for GCB versus 13.8 months (90% CI: 11.3 months - not estimable) for ABC, which was not significant (P=0.09), although this comparison was based on few patients. The median PFS and OS were 7.0 months (90% CI: 3.5-8.0 months) and 9.1 months (90% CI: 6.0-16.0 months), respectively, as shown in Figure 3. At last follow-up assessment, six patients were alive, 30 had died, and three were

Table 1. Summary of dose levels, number of patients, and dose-limiting toxicities for the phase I study.

Dose level

Dose TEM (mg)

LEN (mg)

-1

25 mg

10 mg

25 mg

15 mg

25 mg

20 mg

25 mg

Number of patients

DLT

na 8 (2 inevaluable for DLT) 4 (1 inevaluable for DLT)

na grade 4 hypokalemia

DLT: dose-limiting toxicity; TEM: temsirolimus; LEN: lenalidomide; HSV: herpes simplex virus. Haematologica | 107 July 2022

1610

No DLT grade 3 diarrhea, grade 3 HSV mucositis

ARTICLE - Phase I/II TEM/LEN clinical trial

A. Major et al.

lost to follow-up. Five patients in the DLBCL cohort proceeded to allogeneic transplantation after TEM/LEN therapy, with three reported as alive and two lost to follow-up at last assessment. The FL cohort (n=15) was slow to accrue and closed pre-

maturely, and was therefore not included in the KaplanMeier survival analysis. The ORR was 46.6% with 33.3% achieving CR. The median DOR was 26.5 months (90% CI:17.6-35.2 months). The median PFS and OS were 27.7 months (90% CI: 6.5-35.8 months) and 35.8 months (90%

Table 2. Baseline patient characteristics for the phase II study.

Characteristics Number of patients Median age (y) Age range (y) Sex Male (%) Female (%) Race White Black or African American Asian Unknown or declined American Indian or Alaska Native More than one race Ethnicity Hispanic or Latino Non-hispanic Unknown or declined Histology DLBCL Follicular lymphoma Hodgkin lymphoma T-cell lymphoma Mantle cell lymphoma Marginal zone lymphoma Waldenström macroglobulinemia Lymphoma characteristics Germinal center subtype Non-germinal center subtype Double hit lymphoma Double expressor lymphoma Transformed lymphoma Number of prior regimens 1 2 3 4 >4 Median Range Type of prior therapy Multiagent chemotherapy (%) Radiation (%) Rituximab (%) Autologous stem cell transplantation (%) Brentuximab vedotin (%)

DLBCL

Exploratory cohort

39 65 25-78

15 61 43-76

39 49 23-72

28 (72) 11 (28)

9 (60) 6 (40)

25 (64) 14 (36)

33 4 1 1 0 0

12 2 1 0 0 0

29 6 1 1 1 1

1 37 1

1 14 0

7 31 1

39 15 20 9 5 4 1 22 8 3 4 9 4 12 5 7 11 3 1-11

2 2 4 2 5 3 1-6

2 1 9 10 17 4 1-14

39 (100) 9 (23) 39 (100) 8 (21) 0 (0)

15 (100) 1 (7) 15 (100) 2 (13) 0 (0)

39 (100) 17 (44) 10 (26) 21 (54) 20 (51)

FL: follicular lymphoma; DLBCL: diffuse large B-cell lymphoma; y:years. Haematologica | 107 July 2022

1611

ARTICLE - Phase I/II TEM/LEN clinical trial

A. Major et al.

Figure 1. Waterfall plot for best response for evaluable patients in the phase II study by histology (n=74). Of the 93 patients in the phase II study, 8 patients did not complete 2 cycles of temsirolimus/lenalidomide (TEM/LEN) for pre-specified response assessment, 10 patients did not have reported response data, and 1 patient had Waldenström macroglobulinemia. DLBCL: diffuse large B-cell lymphoma; FL: follicular lymphoma; cHL: classical Hodgkin lymphoma; T-NHL: non-Hodgkin lymphoma; MZL: marginal zone leukemia; MCL: mantle cell lymphoma.

CI: 18.8 months - not estimable), respectively. At last follow-up assessment, five patients were alive, six had died, and four were lost to follow-up. When evaluating all patients in the exploratory cohort, the ORR was 64.1% (CR 17.9%) and the median DOR was 5.5 months (90% CI: 2.6-23.7 months). Among all histologies in this cohort, the total number of responders, 25 of 39, was sufficient to reject the null hypothesis of a 30% response rate (P<0.10). The median PFS and OS were 7.0 months (90% CI: 4.6-9.9 months) and 25.5 months (90% CI: 10.8-60.6 months), respectively, as shown in Figure 3. At last follow-up assessment, ten patients were alive, 21 had died, and eight were lost to follow-up. A substantial portion of patients in the exploratory cohort had cHL (n=20). As displayed in Table 3, the ORR for cHL was 80% (CR 35%). The median DOR was 8.1 months (90% CI: 5.1-38.3 months). The median PFS and OS were 9.2 months (90% CI: 4.6 - 25.5 months) and 39.6 months (90% CI: 17.4 months - not reached), respectively, displayed in Figure 4. Eight cHL patients (40%) proceeded to allogeneic transplantation after TEM/LEN therapy. At last follow-up assessment, nine patients were alive, nine had died, and two were lost to follow-up. Notably, six of the eight patients who had received allogeneic transplantation after TEM/LEN were alive at last assessment. Of the 19 non-cHL patients in the exploratory cohort, nine achieved a PR and none achieved a CR (ORR 47.4%). Spe-

cifically, responses were observed in six of nine patients with T-NHL (67%) with a median DOR of 2.3 months (90% CI: 1.8 months - not reached), in two of four patients with MZL (50%), in the one patient with WM, and no patients with MCL. Treatment delivered The median number of TEM/LEN cycles delivered was four (range, 1-21). The CONSORT diagram in the Online Supplementary Figure S1 depicts the reasons for treatment discontinuation among all three cohorts. Twelve patients, three in the phase I study and nine in the phase II study, did not complete two cycles of TEM/LEN for the following reasons: adverse effects (n=6), progression of disease (n=3), withdrawal from study (n=1) or death (n=2). Reasons for discontinuing study treatment at any point beyond cycle 2 included toxicity (n=21, see safety and tolerability), progression of disease (n=36), death (n=3), or other (n=26). In this latter category, reasons for the discontinuation of therapy were either to pursue alternative treatment or due to physician or patient decisions. Fifty-one patients required dose reductions, primarily due to hematopoietic toxicities. Safety and tolerability Online Supplementary Table S1 summarizes the AE that occurred in greater than 10% of patients or that were grade 3

Haematologica | 107 July 2022

1612

ARTICLE - Phase I/II TEM/LEN clinical trial

A. Major et al.

Figure 2. Swimmer’s plot for patients in the diffuse large B-cell lymphoma (n=39) and classical Hodgkin lymphoma (n=20) cohorts. The plot includes treatment duration, duration of follow-up, best responses, time of progression, reason for treatment discontinuation, and time of subsequent transplantation. DLBCL: diffuse large B-cell lymphoma.

or 4 in severity in the phase I study, and Table 4 summarizes the AE in the phase II study. In the phase II study, grade 3 or 4 hematologic AE were common, and included anemia (n=27), lymphopenia (n=39), neutropenia (n=43), thrombocytopenia (n=40), and leukocytosis (n=37). Common grade 12 non-hematologic AE included alanine transaminase (ALT) and aspartate transaminase (AST) elevation, hypertriglyceridemia, hyperglycemia, hypocalcemia, hypokalemia, anorexia, fatigue, and rash. Grade 3 or 4 non-hematologic AE were uncommon, with only fatigue occurring in greater than 10% of patients in the phase II study. Three grade 5 AE were observed that were possibly related to TEM/LEN, and were colonic perforation, myocardial infarction, and sepsis. There was one case of grade 3 pneumonitis in the phase I study

and one case of grade 3 thromboembolism in the phase II study. There were no secondary malignancies identified. There were ten deaths on study: one in the phase I portion due to disease, and nine in the phase II study, three of which were the aforementioned grade 5 AE, four due to disease, and two which were unrelated to the study (1 seizure, 1 infectious pneumonia occurring several months after receiving a single dose of TEM/LEN).

Discussion Despite significant advances, there remains a need to identify safe, rational and efficacious regimens for re-

Haematologica | 107 July 2022

1613

ARTICLE - Phase I/II TEM/LEN clinical trial

A. Major et al.

Table 3. Primary and secondary outcomes: rates of complete response, overall response, and survival for all three cohorts, including the subset of Hodgkin lymphoma patients in the exploratory cohort.

DLBCL (n=39)

Primary outcomes Complete response (%) Overall response (%) Secondary outcomes PFS (median, 90% CI) OS (median, 90% CI) DOR (median, 90% CI) Follow-up (median, range)

FL (n=15)

Exploratory cohort (n=39) Entire cohort (N=39)

Hodgkin lymphoma (N=20)

5 (12.8) 10 (25.6)

5 (33.3) 7 (46.6)

7 (17.9) 25 (64.1)

7 (35) 16 (80)

7.0 months (3.5-8.0) 9.1 months 6.0-16.0) 13.8 months3 (4.1-19.0) 8.0 months (1.3-30.9)

27.7 months (6.5-35.8) 35.8 months (18.8-NE)1 26.5 months4 (17.6-35.2) 18.8 months (5.9-73.9)

7.0 months (4.6-9.9) 25.5 months (10.8-60.6) 5.5 months5 (2.6-23.7) 13.1 months (1.0-71.6)

9.2 months (4.6-25.5) 39.6 months (17.4-NR)2 8.1 months6 (5.1-38.3) 20.9 months (2.6-71.6)

NE: not estimable; 2NR: not reached; 3n=10; 4n=7; 5n=25; 6n=16; CI: confidence interval; FL: follicular lymphoma; DLBCL: diffuse large B-cell lymphoma; PFS: progression-free survival; OS: overall survival: DOR: duration of response. 1

Figure 3. Kaplan-Meier curves for progression-free survival, overall survival and duration of response in the diffuse large B-cell lymphoma (n=39) and exploratory cohorts (n=39). The duration of response curves are based on 10 and 25 responders, respectively. DLBCL: diffuse large B-cell lymphoma.

lapsed and refractory lymphomas. A promising target is the PAM signaling axis, with mTOR representing one of the penultimate components impacting mRNA translation, autophagy, and cell survival.5 LEN, with its pleiotropic effects on malignant and non-malignant cells, is a rational combination partner. This phase I/II clinical trial investigated the safety and efficacy of the first-generation mTOR

inhibitor TEM plus LEN across several lymphoma subtypes. The phase I component identified LEN 20 mg on days 1 through 21 of a 28-day cycle as the recommended phase II dose when combined with weekly TEM at 25 mg. Preliminary efficacy and acceptable toxicity in the phase I study prompted the phase II trial, which shows promising activity of TEM/LEN combination therapy in relapsed and

Haematologica | 107 July 2022

1614

ARTICLE - Phase I/II TEM/LEN clinical trial

A. Major et al.

refractory cHL. Among 20 cHL patients with very heavily pretreated disease, including a median of six prior lines of therapy and near universal BV exposure, we found an ORR of 80%, allowing many to be bridged to subsequent allogeneic stem cell transplantation. mTOR inhibitors, both alone and in combination, have been previously tested in lymphoid malignancies. Early research on mTOR inhibitors focused on relapsed MCL due to the putative role of mTOR inhibition in suppressing downstream overexpression of cyclin D1.5,7 Initial phase II trials of single-agent TEM in MCL demonstrated an ORR of 38-41%,7,21 culminating in a phase III study of TEM compared to investigator’s choice of therapy which demonstrated a superior PFS and ORR with TEM.8 Further mTOR-focused clinical research with the rapamycin analog everolimus demonstrated modest single-agent activity in relapsed aggressive lymphomas with an ORR of 30-38%,22 with an even higher ORR of 70% in WM.23 Given the encouraging single-agent activity of mTOR inhibitor monotherapy, there has been ongoing research into synergistic combinations.3 A phase II study of TEM in combination with rituximab for relapsed MCL found an improved ORR of 59%,24 and preliminary data have been presented on mTOR inhibitors in combination with the BCL2 inhibitor venetoclax25 as well as triplet therapy with mTOR inhibitors, BTK inhibitors and pomalidomide,26 all with encouraging early reports. Among NHL patients, our trial identified modest activity in DLBCL, with similar response rates compared to our previous phase II clinical trial of TEM monotherapy in DLBCL.9 Despite limited responses, we observed a median DOR of 13.8 months in heavily-pretreated patients with aggressive disease, a considerably longer DOR than only 2.4 months observed with single-agent TEM.9 These findings with TEM/LEN may be related to cell-of-origin; seven of ten DLBCL responders, including three of five complete responders, harbored an ABC phenotype where LEN is known to have preferential activity.29 Others have shown that upstream inhibition of PAM signaling may be active in DLBCL and related to PIK3CA amplification;3,27 to this point, the PI3K inhibitor copanlisib demonstrated single-agent activity in relapsed DLBCL with an ORR of 32% in ABC-type and 13% in GCB-type.28 Since our study was conducted, the management options for relapsed and refractory DLBCL have expanded substantially and now includes chimeric antigen receptor T (CAR-T) cell therapy, and the role of less aggressive regimens such as TEM/LEN is unclear. TEM/LEN may be an option for patients unable to tolerate CAR-T cells, particularly for patients with ABC DLBCL, or given the impact of PI3K inhibition on CAR-T cell activity and persistence in vivo, there may be a role to further explore PAM axis inhibition following CAR-T cells.30,31 The FL cohort was unfortunately closed due to low accrual, which may be related to the competitive landscape

Table 4. Summary of reported toxicities in the phase II study occurring in either greater than 10% of patients or grade ≥3 in severity.

Toxicity

All grades Grade 3-4 Grade 5

Non-hematologic toxicity ALT increased

39 (42%)

1 (1%)

Alk phos increased

31 (33%)

1 (1%)

Anorexia

37 (40%)

4 (4%)

AST increased

38 (41%)

Bleeding

10 (11%)

Chills

10 (11%)

Hypercholesterolemia

28 (30%)

1 (1%)

Constipation

32 (34%)

Cough

15 (16%)

Creatinine increased

23 (25%)

1 (1%)

Diarrhea

28 (20%)

2 (2%)

Dysgeusia

21 (23%)

Edema

29 (31%)

1 (1%)

Fatigue

65 (70%)

11 (12%)

Fever

22 (24%)

6 (6%)

Hypercalcemia

10 (11%)

Hyperglycemia

69 (74%)

8 (9%)

Hypernatremia

10 (11%)

Hypertension

10 (11%)

2 (2%)

Hypertriglyceridemia

35 (38%)

5 (5%)

Hypoalbuminemia

26 (28%)

1 (1%)

Hypocalcemia

50 (54%)

4 (4%)

Hypokalemia

56 (60%)

8 (9%)

Hypomagnesemia

25 (27%)

2 (2%)

Hyponatremia

19 (20%)

2 (2%)

Hypophosphatemia

22 (24%)

3 (3%)

Infection

11 (12%)

2 (2%)

Myalgia

9 (10%)

1 (1%)

Nausea

26 (28%)

2 (2%)

Neuropathy

16 (17%)

Pain

11 (12%)

1 (1%)

Pruritus

30 (32%)

1 (1%)

Rash

59 (63%)

7 (8%)

1 (1%)

22 (24%)

3 (3%)

4 (4%)

1 (1%)

Anemia

77 (83%)

27 (29%)

Lymphopenia

66 (71%)

39 (42%)

Neutropenia

63 (68%)

43 (46%)

Thrombocytopenia

76 (82%)

40 (43%)

Leukocytosis

76 (82%)

37 (40%)

Colonic perforation

Myocardial infarction

Sepsis Stomatitis Thromboembolism Hematologic toxicity

ALT: alanine transaminase; Alk phos: alkaline phosphatase; AST: aspartate transaminase.

Haematologica | 107 July 2022

1615

ARTICLE - Phase I/II TEM/LEN clinical trial

A. Major et al.

Figure 4. Kaplan-Meier curves for progression-free survival, overall survival and duration of response for the classical Hodgkin lymphoma patients (n=20) in the exploratory cohort. The duration of response curve is based on 16 responders.

of effective therapies both as part of clinical trials as well as routine clinical care. We are encouraged by the early activity of TEM/LEN in FL, but have insufficient data to comment further. In the exploratory cohort that enrolled other lymphoma histologies, we observed promising activity with TEM/LEN in T-NHL, with two-thirds of patients responding. Others have shown activity with both mTOR inhibitors and LEN in T-NHL, with ORR of 44% with single-agent everolimus32 and ORR of 26% with single-agent LEN.33 The activity of LEN in TNHL may be due to overexpression of Akt in T-NHL,34 with evidence of possible synergism between TEM and LEN in this study. There is preclinical rationale for combination targeting of the PAM axis in T-NHL, with dual mTOR and PI3K inhibition demonstrating activity in cutaneous T-NHL cell lines.35 The PI3K pathway appears to be particularly active in T-NHL,36 with a phase I trial of duvelisib demonstrating ORR of 50.0% in peripheral T-NHL and 31.6% in cutaneous T-NHL37 and a phase I trial of novel dual PI3K d/γ inhibitor tenalisib demonstrating ORR of 46% in relapsed T-NHL.38 Further investigation of PAM inhibition, with or without LEN, appears warranted. The most promising signal of activity in our study was observed in relapsed and refractory cHL. While our study was conducted prior to the era of checkpoint inhibitors (CPI), cHL patients enrolled in this trial were heavily pretreated, with a median of six prior regimens, with near universal

prior BV exposure and the majority having relapsed despite prior ASCT. We observed an ORR of 80% with a CR rate of 35%, which compares favorably with expected outcomes following either single-agent BV or CPI.39,40 This may be due to constitutive activation of the PAM axis in cHL with downstream activation of NF-kB promoting cell survival and proliferation.6,41 Others have explored mTOR inhibition as monotherapy and in combination with other agents in cHL. Based on preclinical work demonstrating cell cycle arrest and autophagy induced by TEM in cHL cell lines, a phase II trial of single-agent everolimus in 19 patients with relapsed cHL was conducted and found an ORR of 47%.42 Combined sirolimus and vorinostat had an ORR of 55% in relapsed cHL.43 However, combination therapy with TEM/LEN had higher response rates than LEN (ORR 19%)15 or mTOR inhibitors (ORR 47%) alone,42,44 supporting dual targeting of the PAM axis in cHL. Targeting other components of the PAM axis, such as PI3K inhibition, has demonstrated modest response rates, with an ORR of 20% with single-agent idelalisib in relapsed cHL.45 Additional areas to consider include combination with CPI, as mTOR inhibition may induce PD-L1 expression and encourage immune escape in preclinical models.46 Further, a recent study of the mTOR inhibitor everolimus in combination with ruxolitinib, an oral JAK inhibitor which targets the JAK/STAT pathway that is a putative escape mechanism to CPI,47 demonstrated an ORR of 79% in relapsed cHL that had progressed after CPI ther-

Haematologica | 107 July 2022

1616

ARTICLE - Phase I/II TEM/LEN clinical trial

A. Major et al.

apy.48 Overall, combined TEM/LEN had encouraging activity and supports further investigation in cHL. The combination of TEM/LEN therapy was feasible in this study, with hematologic AE being most commonly experienced. Pneumonitis is a previously-reported complication of mTOR inhibitor therapy,49 but we found only one case of grade 3 pneumonitis occurring in this study. LEN treatment has been associated with a risk for thromboembolism,50 and there was one grade 3 thromboembolism in the study. Although there is significant experience with weekly TEM dosing, it is clearly inconvenient for patients, and there is a suggestion that higher TEM dosing may be more efficacious than the 25 mg dose selected for evaluation in this study.8 Overall, the combination of TEM/LEN demonstrated encouraging activity in a heavily pretreated group of patients with relapsed and refractory cHL, and could be a platform for future investigations. In contrast, the addition of LEN to TEM did not show significant improvement over our prior TEM monotherapy study in either DLBCL or FL, although we did find encouraging response duration among the few responding patients and preliminary activity in a very small cohort of T-NHL patients. Two major unanswered questions are whether first-generation agents such as TEM or everolimus provide optimal mTOR inhibition,5 and whether upstream inhibition of PI3K should supplant our approach given the number of agents in this area. Future research exploring novel therapeutics or combinations of therapeutics acting on the PAM axis, particularly in patients who cannot tolerate transplantation or i.e., CAR-T cell therapy and have heavily pre-treated lymphomas, is warranted.

Disclosures AM received funding from the National Cancer Institute (NCI), Merck and a T32 institutional training grant (5T32CA009566); JK is a member on an entity's Board of Directors or advisory committees of SeattleGenetics, Verastem, Morphosys and Karyopharm, is part of the speakers bureau of Kite/Gilead and consults for Regeneron; AM received funding from a National Cancer Institute (NCI) T32 institutional training grant (5T32CA009566; ASK is currently employed by Amgen and is a current equity holder in a publicly-traded company (Amgen); AMP is currently employed by Daiichi-Sankyo and is a current equity holder in a publicly-traded company (AbbVie); TGK, PASF, SN, KR, BGS, KC, KvB and APR have no conflicts of interest to declare; SMS consults for Gilead, Celgene/BMS, Morphosys, Janssen, Bantam, Karyopharm, Genentech, TG Therapeutics,Bayer, Kite and Seattle Genetics. Contributions AM wrote the manuscript; JK, TGK, PASF, ASK, AMP,SN, KR, BGS, KC and APR performed research; KVB supervised the study and performed research; and SMS wrote the protocol, wrote the manuscript, supervised the study and performed research. Data Sharing Statement Within 6 months of publication, a de-identified dataset will be made publically available containing the data used to generate the primary and secondary analyses reported in this manuscript.

References 1. Westin JR. Status of PI3K/Akt/mTOR pathway inhibitors in lymphoma. Clin Lymphoma Myeloma Leuk. 2014; 14(5):335-342. 2. Curran E, Smith SM. Phosphoinositide 3-kinase inhibitors in lymphoma. Curr Opin Oncol. 2014;26(5):469-475. 3. Lee J-HS, Vo T-T, Fruman DA. Targeting mTOR for the treatment of B cell malignancies. Br J Clin Pharmacol. 2016;82(5):1213-1228. 4. Saxton RA, Sabatini DM. mTOR signaling in growth, metabolism, and disease. Cell. 2017;168(6):960-976. 5. Argyriou P, Economopoulou P, Papageorgiou S. The role of mTOR inhibitors for the treatment of B-cell lymphomas. Adv Hematol. 2012;2012:435342. 6. Márk Á, Hajdu M, Váradi Z, et al. Characteristic mTOR activity in Hodgkin-lymphomas offers a potential therapeutic target in high risk disease – a combined tissue microarray, in vitro and in vivo study. BMC Cancer. 2013;13(1):250. 7. Witzig TE, Geyer SM, Ghobrial I, et al. Phase II trial of singleagent temsirolimus (CCI-779) for relapsed mantle cell lymphoma. J Clin Oncol. 2005;23(23):5347-5356. 8. Hess G, Herbrecht R, Romaguera J, et al. Phase III study to evaluate temsirolimus compared with investigator’s choice therapy for the treatment of relapsed or refractory mantle cell lymphoma. J Clin Oncol. 2009;27(23):3822-3829.

9. Smith SM, van Besien K, Karrison T, et al. Temsirolimus has activity in non–mantle cell non-Hodgkin’s lymphoma subtypes: the University of Chicago Phase II Consortium. J Clin Oncol. 2010; 28(31):4740-4746. 10. Gribben JG, Fowler N, Morschhauser F. Mechanisms of action of lenalidomide in B-cell non-Hodgkin lymphoma. J Clin Oncol. 2015;33(25):2803-2811. 11. Montanari F, Diefenbach CSM. Hodgkin lymphoma: targeting the tumor microenvironment as a therapeutic strategy. Clin Adv Hematol Oncol. 2015;13(8):518-524. 12. Lu G, Middleton RE, Sun H, et al. The myeloma drug lenalidomide promotes the cereblon-dependent destruction of ikaros proteins. Science. 2014;343(6168):305-309. 13. Kotla V, Goel S, Nischal S, et al. Mechanism of action of lenalidomide in hematological malignancies. J Hematol Oncol. 2009;2(1):36. 14. Arora M, Gowda S, Tuscano J. A comprehensive review of lenalidomide in B-cell non-Hodgkin lymphoma. Therapeutic Adv Hematol. 2016;7(4):209-221. 15. Fehniger TA, Larson S, Trinkaus K, et al. A phase 2 multicenter study of lenalidomide in relapsed or refractory classical Hodgkin lymphoma. Blood. 2011;118(19):5119-5125. 16. Cheson BD, Pfistner B, Juweid ME, et al. Revised response

Haematologica | 107 July 2022

1617

ARTICLE - Phase I/II TEM/LEN clinical trial

A. Major et al.

criteria for malignant lymphoma. J Clin Oncol. 2007;25(5):579-586. 17. Treon SP. Update on treatment recommendations from the third international workshop on waldenstrom’s macroglobulinemia. Blood. 2006;107(9):3442-3446. 18. Simon R. Optimal two-stage designs for phase II clinical trials. Control Clin Trials. 1989;10(1):1-10. 19. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc. 1958;53:457-481. 20. Brookmeyer R, Crowley J. A confidence interval for the median survival time. Biometrics. 1982;38(1):29-41. 21. Ansell SM, Inwards DJ, Rowland KM, et al. Low-dose, singleagent temsirolimus for relapsed mantle cell lymphoma: a phase 2 trial in the North Central Cancer Treatment Group. Cancer. 2008;113(3):508-514. 22. Witzig TE, Reeder CB, LaPlant BR, et al. A phase II trial of the oral mTOR inhibitor everolimus in relapsed aggressive lymphoma. Leukemia. 2011;25(2):341-347. 23. Ghobrial IM, Gertz M, LaPlant B, et al. Phase ii trial of the oral mammalian target of rapamycin inhibitor everolimus in relapsed or refractory Waldenström macroglobulinemia. J Clin Oncol. 2010; 28(8):1408-1414. 24. Ansell SM, Tang H, Kurtin PJ, et al. Temsirolimus and rituximab in patients with relapsed or refractory mantle cell lymphoma: a phase 2 study. Lancet Oncol. 2011;12(4):361-368. 25. Tarantelli C, Gaudio E, Hillmann P, et al. The novel TORC1/2 kinase inhibitor PQR620 has anti-tumor activity in lymphomas as a single agent and in combination with venetoclax. Cancers. 2019;11(6):775. 26. Mato AR, Schuster SJ, Foss FM, et al. A once daily, oral, triple combination of BTK inhibitor, mTOR inhibitor and IMiD for treatment of relapsed/refractory Richter's transformation and de novo diffuse large B-cell lymphoma. Blood. 2020;136(Suppl 1):S21-22. 27. Godfrey J, Kang W, Huang L, et al. Macrophage activation by dual PI3K-d/γ inhibition enhances anti-CD47-mediated phagocytosis and prolongs survival in DLBCL. Blood. 2020;136(Suppl 1):40. 28. Lenz G, Hawkes E, Verhoef G, et al. Single-agent activity of phosphatidylinositol 3-kinase inhibition with copanlisib in patients with molecularly defined relapsed or refractory diffuse large B-cell lymphoma. Leukemia. 2020;34(8):2184-2197. 29. Zhang L-H, Kosek J, Wang M, Heise C, Schafer PH, Chopra R. Lenalidomide efficacy in activated B-cell-like subtype diffuse large B-cell lymphoma is dependent upon IRF4 and cereblon expression. Br J Haematol. 2013;160(4):487-502. 30. Zheng W, O’Hear CE, Alli R, et al. PI3K orchestration of the in vivo persistence of chimeric antigen receptor-modified T cells. Leukemia. 2018;32(5):1157-1167. 31. Dwyer CJ, Arhontoulis DC, Rangel Rivera GO, et al. Ex vivo blockade of PI3K gamma or delta signaling enhances the antitumor potency of adoptively transferred CD8+ T cells. Eur J Immunol. 2020;50(9):1386-1399. 32. Witzig TE, Reeder C, Han JJ, et al. The mTORC1 inhibitor everolimus has antitumor activity in vitro and produces tumor responses in patients with relapsed T-cell lymphoma. Blood. 2015;126(3):328-335. 33. Toumishey E, Prasad A, Dueck G, et al. Final report of a phase 2

clinical trial of lenalidomide monotherapy for patients with Tcell lymphoma. Cancer. 2015; 121(5):716-723. 34. Hong JY, Hong ME, Choi MK, et al. The clinical significance of activated p-AKT expression in peripheral T-cell lymphoma. Anticancer Res. 2015;35(4):2465-2474. 35. Bresin A, Cristofoletti C, Caprini E, et al. Preclinical evidence for targeting PI3K/mTOR signaling with dual-inhibitors as a therapeutic strategy against cutaneous T-cell lymphoma. J Invest Dermatol. 2020;140(5):1045-1053. 36. Huang D, Song TL, Nairismägi M, et al. Evaluation of the PIK3 pathway in peripheral T-cell lymphoma and NK/T-cell lymphoma. Br J Haematol. 2020;189(4):731-744. 37. Horwitz SM, Koch R, Porcu P, et al. Activity of the PI3K-d,γ inhibitor duvelisib in a phase 1 trial and preclinical models of Tcell lymphoma. Blood. 2018;131(8):888-898. 38. Huen A, Haverkos BM, Zain J, et al. Phase I/Ib study of tenalisib (Rp6530), a dual PI3K d/γ inhibitor in patients with relapsed/refractory T-cell lymphoma. Cancers. 2020;12(8):2293. 39. De Goycoechea D, Stalder G, Martins F, Duchosal MA. Immune checkpoint inhibition in classical Hodgkin lymphoma: from early achievements towards new perspectives. J Oncol. 2019;2019:9513701. 40. Mina A, Vakkalagadda C, Pro B. Novel therapies and approaches to relapsed/refractory HL beyond chemotherapy. Cancers. 2019;11(3):421. 41. Dutton A, Reynolds GM, Dawson CW, Young LS, Murray PG. Constitutive activation of phosphatidyl-inositide 3 kinase contributes to the survival of Hodgkin’s lymphoma cells through a mechanism involving Akt kinase and mTOR. J Pathol. 2005; 205(4):498-506. 42. Johnston PB, Inwards DJ, Colgan JP, et al. A phase II trial of the oral mTOR inhibitor everolimus in relapsed Hodgkin lymphoma. Am J Hematol. 2010;85(5):320-324. 43. Janku F, Park H, Call SG, et al. Safety and efficacy of vorinostat plus sirolimus or everolimus in patients with relapsed refractory Hodgkin lymphoma. Clin Cancer Res. 2020;26(21):5579-5587. 44. Johnston PB, Pinter-Brown LC, Warsi G, White K, Ramchandren R. Phase 2 study of everolimus for relapsed or refractory classical Hodgkin lymphoma. Exp Hematol Oncol. 2018;7(1):12. 45. Gopal AK, Fanale MA, Moskowitz CH, et al. Phase II study of idelalisib, a selective inhibitor of PI3Kd, for relapsed/refractory classical Hodgkin lymphoma. Ann Oncol. 2017;28(5):1057-1063. 46. Sun S-Y. Searching for the real function of mTOR signaling in the regulation of PD-L1 expression. Transl Oncol. 2020;13(12): 100847. 47. Marabelle A, Aspeslagh S, Postel-Vinay S, Soria J-C. JAK mutations as escape mechanisms to anti–PD-1 therapy. Cancer Discov. 2017;7(2):128-130. 48. Svoboda J, Barta SK, Landsburg DJ, et al. Everolimus plus itacitinib in relapsed/refractory classical Hodgkin lymphoma: results of a phase I/II investigator initiated trial (EVITA study). Blood. 2020; 136(Suppl1):S20-21. 49. Zhang X, Ran Y, Wang K. Risk of mTOR inhibitors induced severe pneumonitis in cancer patients: a meta-analysis of randomized controlled trials. Future Oncol. 2016; 12(12):1529-1539. 50. Al-Ani F, Bermejo JMB, Mateos M-V, Louzada M. Thromboprophylaxis in multiple myeloma patients treated with lenalidomide – a systematic review. Thromb Res. 2016;141:84-90.

Haematologica | 107 July 2022

1618

ARTICLE - Non-Hodgkin Lymphoma

Whole-genome profiling of primary cutaneous anaplastic large cell lymphoma Armando N. Bastidas Torres,1* Rutger C. Melchers,1* Liana van Grieken,1 Jacoba J. OutLuiting,1 Hailiang Mei,2 Cedrick Agaser,2 Thomas B. Kuipers,2 Koen D. Quint,1 Rein Willemze,1 Maarten H. Vermeer1 and Cornelis P. Tensen1 Department of Dermatology, Leiden University Medical Center and 2Sequencing Analysis Support Core, Leiden University Medical Center, Leiden, the Netherlands 1

*ANBT and RCM contributed equally as co-first authors.

Correspondence: Cornelis P. Tensen c.p.tensen@lumc.nl Received: June 26, 2020. Accepted: August 3, 2021. Prepublished: August 12, 2021. https://doi.org/10.3324/haematol.2020.263251 ©2022 Ferrata Storti Foundation Haematologica material is published under a CC BY-NC license

Abstract Primary cutaneous anaplastic large cell lymphoma (pcALCL), a hematological neoplasm caused by skin-homing CD30+ malignant T cells, is part of the spectrum of primary cutaneous CD30+ lymphoproliferative disorders. To date, only a small number of molecular alterations have been described in pcALCL and, so far, no clear unifying theme that could explain the pathogenetic origin of the disease has emerged among patients. In order to clarify the pathogenetic basis of pcALCL, we performed high-resolution genetic profiling (genome/transcriptome) of this lymphoma (n=12) by using whole-genome sequencing, whole-exome sequencing and RNA sequencing. Our study, which uncovered novel genomic rearrangements, copy number alterations and small-scale mutations underlying this malignancy, revealed that the cell cycle, T-cell physiology regulation, transcription and signaling via the PI-3-K, MAPK and G-protein pathways are cellular processes commonly impacted by molecular alterations in patients with pcALCL. Recurrent events affecting cancer-associated genes included deletion of PRDM1 and TNFRSF14, gain of EZH2 and TNFRSF8, small-scale mutations in LRP1B, PDPK1 and PIK3R1 and rearrangements involving GPS2, LINC-PINT and TNK1. Consistent with the genomic data, transcriptome analysis uncovered upregulation of signal transduction routes associated with the PI-3-K, MAPK and G-protein pathways (e.g., ERK, phospholipase C, AKT). Our molecular findings suggest that inhibition of proliferation-promoting pathways altered in pcALCL (particularly PI-3-K/AKT signaling) should be explored as potential alternative therapy for patients with this lymphoma, especially, for cases that do not respond to first-line skin-directed therapies or with extracutaneous disease.

Introduction Primary cutaneous anaplastic large cell lymphoma (pcALCL) is a hematologic neoplasm produced by malignant CD30+ large T cells that infiltrate the skin forming solitary or localized tumors.1,2 pcALCL is largely an anaplastic lymphoma kinase (ALK)-negative lymphoma with only ~2% of ALK positive (ALK+) cases.3 Patients with pcALCL have a favorable prognosis with a 5-year survival of 90%; yet, a minority of patients (~10% of cases) develop extracutaneous dissemination with a more aggressive clinical course.1,2,4 Unlike systemic anaplastic large cell lymphoma (sALCL), which appears to be driven by a variety of genetic alterations that overactivate STAT3 signaling,5-9 the pathogenetic basis of pcALCL has remained elusive. Previous studies using cytogenetic techniques, array-based plat-

forms (DNA, RNA and miRNA) and whole-exome sequencing (WES) have only identified a few recurrent molecular alterations in this lymphoma;10 thus, genetic abnormalities underlying pcALCL are still largely unknown. Chromosomal rearrangements involving ALK, DUSP22, TP63 and TYK2, common in sALCL, have also been detected in pcALCL, albeit at much lower frequencies.3,10 Frequent copy number alterations (CNA) include losses within chromosomes 6 and 13 and gains within chromosomes 7 and 17.11,12 Overexpressed genes with potential pathogenic and/or diagnostic relevance include EZH2 and IRF4.11,13,14 Oncomir miR-155, which is associated with other cutaneous T-cell lymphoma (CTCL) variants, is overexpressed in pcALCL too.15,16 Finally, single nucleotide variants (SNV) presumed to be pathogenic have been reported in JAK1, MSC and STAT3 in a few cases.6,9 Given the lack of a clear unifying theme between the few

Haematologica | 107 July 2022

1619

ARTICLE - Whole-genome profiling of pcALCL

A.N. Bastidas Torres et al.

reported molecular alterations in the disease, we performed a genome-wide analysis (genome/transcriptome) of pcALCL using whole-genome sequencing (WGS), WES and RNA sequencing (RNA-seq) to investigate genetic defects underlying this neoplasm. Our data show that molecular alterations in genes involved in the regulation of the cell cycle, T-cell physiology, transcription and signaling via the PI-3-K, MAPK and G-protein pathways underlie pcALCL. Hence, the inhibition of proliferation-promoting pathways operative in pcALCL (especially PI-3-K/AKT signaling) could be explored as an alternative therapy for pcALCL, in particular, for patients who do not respond to skin-directed therapies or with extracutaneous dissemination.

number values were determined using Bio-Rad Quantasoft software v1.7.4. (Online Supplementary Table S19). Select differentially expressed (DE) genes were validated by reverse transcriptase quantitative PCR (RT-qPCR) as described by van Doorn et al.17 Immunohistochemistry Formalin-fixed paraffin-embedded (FFPE) tumor sections from sequenced patients with available tissue material (i.e., cAL3-4/7-12) were immunohistochemically stained with primary antibodies against phospho-AKT (Abcam, Cat. No. ab38449) and phospho-STAT3 (Cell Signaling Technology, Cat. No. 9145) following the manufacturer’s intructions, counterstained in Mayer’s Hematoxylin solution and coverslipped using DPX-mountant (Sigma-Aldrich, Cat. No. 06522).

Methods Patient selection and sequencing Frozen tumor biopsies (≥75% tumor cells) from 12 patients with (ALK–) pcALCL (Online Supplementary Figure S1,; Online Supplementary Table S1) were subjected to WGS and RNA-seq. Seven matched tumor/germline (granulocytes) samples from this cohort (i.e., cAL2-5/9-11) were additionally subjected to WES. Details of sequencing, data processing and DNA/RNA analyses are provided in the Online Supplementary Appendix (Online Supplementary Figures S2 to S4; Online Supplementary Tables S2 to S18). Diagnosis was performed by an expert panel of pathologists and dermatologists according to the criteria of the World Health Organization-European Organisation for Research and Treatment of Cancer (WHO-EORTC) classification for primary cutaneous lymphomas. Patient material was approved by the Institutional Review Board of Leiden University Medical Center. Informed consent was obtained from patients in accordance with the Declaration of Helsinki. Validation of genetic alterations and differentially expressed genes Select rearrangements and SNV were validated by Sanger sequencing. Target sequences were polymerase chain reaction (PCR)-amplified, run on a 1% agarose gel, columnpurified and sequenced on the Applied Biosystems ABI3730xl platform (Applied Biosystems) (Online Supplementary Figures S5 and S6). Select CNA were validated by droplet digital PCR (ddPCR) using QX200 ddPCR system (Bio-Rad) following the manufacturer's guidelines. In short, genomic DNA (20-40 ng) was mixed with a frequentcutting restriction enzyme, ddPCR supermix and probes against the target/reference genes (FAM/HEX-labeled, respectively). The reaction mix was partitioned into 20,000 nanodroplets and PCR-amplified. FAM/HEX fluorescence was measured with Bio-Rad QX200 droplet reader. Copy

Results Landscape of genomic rearrangements The landscape of rearrangements of pcALCL showed considerable inter-patient heterogeneity. The number of rearrangements varied from one to 51 per patient (205 total events; mean/patient ± standard deviation [SD], 17 ± 17) (Figures 1 and 2A). Sixty-one percent of events were intrachromosomal (range/patient, 0–100%) (Figure 2B). We identified events fusing two annotated genes (30%), a gene with a non-genic region (42%), two non-genic regions (25%) or reordering sequences within a single gene (3%). Ten percent of rearrangements led to the expression of fusion transcripts (mean fusions/patient, 1.6; range, 0–4 fusions/patient) (Figure 2C). Complex rearrangements involving chromosome 3 and 7 were observed in patients cAL8 and cAL1, respectively (Online Supplementary Figure S7). We detected a total of 166 rearranged genes in our cohort (Online Supplementary Table S8), including a subset of 44 genes implicated in cancer (Figure 1; Online Supplementary Table S8). The vast majority of events were patient-specific. DUSP22 and TP63 rearrangements, reported in pcALCL before,10 were found in single cases (Online Supplementary Figure S6). Gene ontology annotation with DAVID revealed that rearranged genes were involved in multiple cellular processes/pathways, several of which have relevant roles in cancer onset and progression (e.g., transcription, cell cycle, chromatin modification, MAPK pathway) (Figure 2D; Online Supplementary Table S7). Analysis of rearranged genes with NCG 6.0 uncovered cancer genes principally associated with three cellular processes/pathways: the cell cycle (i.e., FMN2, FOXM1, HORMAD1, KMT2E, LINC-PINT, NUMA1, TP53, TP63, USP7), signaling via the PI-3-K and MAPK pathways (i.e., DGKI, ERBB2IP, GPS2, GRM8, MAP3K20, MET, PIK3C2B, PPM1L,

Haematologica | 107 July 2022

1620

ARTICLE - Whole-genome profiling of pcALCL

A.N. Bastidas Torres et al.

Figure 1. Landscape of genomic rearrangements in primary cutaneous anaplastic large cell lymphoma. Circos plot showing 205 genomic rearrangements detected in twelve pcALCL genomes by whole-genome sequencing. The outer ring displays rearranged genes with established roles in cancer. The area at the center of the plot contains arcs representing interchromosomal (blue) and intrachromosomal (red) events. The ring between the gene labels and the arcs contains human chromosome ideograms arranged circularly end to end.

PTPRK, TNK1) and the regulation of RNA expression/processing (i.e., ETS1, RBFOX1, TRERF1, ZNF208, ZNF28) (Online Supplementary Figure S6). A group of five genes were found to be recurrently rearranged in pcALCL: GPS2 (2 patients), LINC-PINT (3 patients), RBFOX1 (2 patients), TNK1 (2 patients) and VPS13D (2 patients). Notably, four of five of these genes encode products with roles in the cell cycle, transcription or signaling. LINC-PINT is a long non-coding RNA that acts as a TP53-dependent regulator of the cell cycle and a PRC2dependent silencer of gene expression.18 RBFOX1 is an RNA-binding protein that regulates alternative splicing.19 GPS2 and TNK1 are negative regulators of RAS/JNK1 and RAS/RAF1 signaling, respectively.20,21 Since deregulated RAS-MAPK signaling is a frequent driver of human cancers, we looked at potential detrimental effects of the rearrangements on the expression of GPS2 and TNK1. The expression data confirmed that both genes were functionally compromised in the affected patients. For instance, patient cAL4 expressed a truncated GPS2CHD3 transcript which rendered both fusion partners dysfunctional (Figure 2E, F). Similarly, patient cAL3 expressed a transcript encoding a truncated form of TNK1 which was

structurally analogous to a variant with proven oncogenic activity reported in Hodgkin lymphoma (Figure 2G, H).22 Landscape of copy number alterations pcALCL displayed a predominance of large numerical imbalances which included losses within 3q, 6q, 7p, 7q, 8p, 13q, and 16q as well as loss of chromosome Y. Also, gains within 1q, 2p, 2q, 7q and 12q (Figures 3A and 4). Our analysis detected 15 focal (≤ 3Mb) minimal common regions (MCR) between CNA found in our patients (9 deletions, 6 gains), six of which enclosed (putative) cancer genes (Figure 4; Online Supplementary Table S6). Half of these genes participate in the regulation of T-cell physiology. Deletion at 6q21, observed in six of 12 patients, was the most recurrent focal MCR in pcALCL. This region contained tumor suppressor PRDM1 (BLIMP1) (Figure 3B, C), which encodes a transcription factor that attenuates the proliferation and survival of T cells through suppression of IL-2 signaling.23 Three of 12 patients had deletions at 13q32.2 which enclosed STK24 (Figure 3D). This gene encodes a proapoptotic serine/threonine kinase that modulates phosphorylation of JNK and p38 kinases in response to stress stimuli.24 Two of 12 patients had deletions at 1p36.32 and

Haematologica | 107 July 2022

1621

ARTICLE - Whole-genome profiling of pcALCL

A.N. Bastidas Torres et al.

Figure 2. Rearrangements disrupt RAS-MAPK signaling inhibitors in primary cutaneous anaplastic large cell lymphoma. (A) Number of genomic rearrangements per patient. The distribution of inter- and intrachromosomal rearrangements per patient is shown. (B) Distribution of inter- and intrachromosomal rearrangements (cohort). (C) Distribution of genomic rearrangements based on the type of DNA sequences involved in the event (genic or nongenic) and the expression of fusion sequences determined through integration of whole-genome sequencing and RNA sequencing data (cohort). (D) Gene ontology annotation of rearranged genes. (E) Validation of GPS2 rearrangements in patients cAL3 and cAL4 by Sanger sequencing. (F) Diagram of fusion gene GPS2-CHD3 in patient cAL4 and Sashimi plots showing the expression of CHD3 in patients cAL4 (disrupted by rearrangement, purple) and cAL6 (with intact alleles, green). Besides disrupting the signaling-suppressing domain of GPS2, the fusion event also inactivated CHD3, the helicase subunit of chromatin remodeling complex Mi-2/NuRD. In patient cAL4, the GPS2-fused CHD3 transcript (5’-side of the gene) ends prematurely. A nonsense-mediated decay (NMD) transcript (3’-side of the gene) is expressed too. CDS: coding sequence. (G) Validation of TNK1 rearrangements in patients cAL3 and cAL4 by Sanger sequencing. (H) Sashimi plots showing the expression of TNK1 in patients cAL3/4 (disrupted by rearrangements, purple) and cAL7 (with intact alleles, green). Patient cAL3 expressed a transcript encoding a C-terminus truncated TNK1 protein which was structurally analogous to a known oncogenic variant. The expression of TNK1 was abolished in patient cAL4. Haematologica | 107 July 2022

1622

ARTICLE - Whole-genome profiling of pcALCL

A.N. Bastidas Torres et al.

16p13.3, which encompassed TNFRSF14 (HVEM) and RBFOX1, respectively. TNFRSF14 is a multifaceted receptor capable of sending stimulatory or inhibitory signals to T cells depending on the ligand it binds.25 Of note, loss of TNFRSF14 has been reported in pcALCL before.12 In addition, focal (≤ 3Mb) gains at 1p36.22 and 2p15 were observed in two and three of 12 patients, respectively. The former contained TNFRSF8 which encodes receptor CD30 (Figure 3D), whose overexpression is a hallmark of pcALCL. The latter enclosed XPO1, a gene that encodes a protein involved in the export of a plethora of proteins and RNA from the nucleus to the cytoplasm.26 Furthermore, six of 12 patients had gains at 7q36.1 (MCR ~ 6 Mb), which contained EZH2, the catalytic subunit of gene silencing complex PRC2 (Figure 3D).

Landscape of small-scale mutations and mutational signatures In order to detect somatic small-scale mutations (indels, SNV) in pcALCL, we performed WES on matched tumor/germline samples from seven patients. The number of somatic indels/SNV ranged from 44 to 246 (653 total events; mean/patient ± SD, 93 ± 70). The average somatic mutation rate was 15.8 mutations/Mb (range/patient, 9.2– 38.6 mutations/Mb). The majority of small-scale mutations were missense SNV (Figure 5A; Online Supplementary Table S9). The most frequent base substitution was C>T (mean/cohort, 56.2%), (Figure 5B), which was attributable to ultraviolet (UV) light exposure and aging-associated spontaneous deamination of 5-methylcytosine. DNA double-strand break (DSB) repair deficiency and activation-induced cytidine

Table 1. Fusion transcripts detected by RNA sequencing in primary cutaneous anaplastic large cell lymphoma.

Fusion transcript

Breakpoints (DNA)

Breakpoint type

Event class

WGS confirmed

cAL1

NLRP1–DERL2

chr17:5554176 - chr17:5477401

Genic - Genic

iDel

Yes

cAL1

ANKRD11–VPS9D1-AS1

chr16:89306818 - chr16:89713470

Genic - Genic

ITX

Yes

cAL1

INPP5B–GNL2

chr1:37926300 - chr1:37586211

Genic - Genic

iDel

Yes

cAL2

IGSF3–CD58

chr1:116578013 - chr1:116525260

Genic - Genic

iDel

Yes

cAL3

TNK1–GPS2

chr17:7386883 - chr17:7315061

Genic - Genic

ITX

Yes

cAL3

TP53–TNK1

chr17:7682311 - chr17:7382806

Genic - Genic

ITX

Yes

cAL4

GPS2–CHD3

chr17:7314379 - chr17:7886522

Genic - Genic

ITX

Yes

cAL5

LINC-PINT–RNASEH2B-AS1

chr7:130895998 - chr13:50895286

Genic - Genic

CTX

Yes

cAL5

SMARCAD1–TP63

chr4:94244727 - chr3:189786186

Genic - Genic

CTX

Yes

cAL5

TP63–SMARCAD1

chr3:189786186 - chr4:94244727

Genic - Genic

CTX

Yes

cAL5

ETS1–SLC37A4

chr11:128520859 - chr11:119027083

Genic - Genic

ITX

Yes

cAL7

NFKB2–SIRT3

chr10:102401366 - chr11:217595

Genic - Genic

CTX

Yes

cAL7

SIRT3–NFKB2

chr11:217564 - chr10:102401164

Genic - Genic

ITX

Yes

cAL7

GTPBP2–TRERF1

chr6:43623524 - chr6:42232607

Genic - Genic

CTX

Yes

cAL8

USP7–SAMD11

chr16:8943214 - chr1:942313

Genic - Genic

CTX

Yes

cAL8

MMS19–RRP12

chr10:97461940 - chr10:97410043

Genic - Nongenic

iDel

Yes

cAL9

NFAT5–WWP2

chr16:69595202 - chr16:69801863

Genic - Genic

iDel

Yes

cAL10

RABEPK–UBAC1

chr9:125226314 - chr9:135960125

Genic - Genic

ITX

Yes

cAL10

HIVEP3–MRPL43

chr1:41890425 - chr10:100998633

Genic - Genic

CTX

Yes

cAL10

DUSP22–UNC93B1

chr6:314945 - chr11:67992497

Genic - Genic

CTX

Yes

cAL11

AGTRAP–LST1

chr1:11738153 - chr6:31588590

Genic - Genic

CTX

Yes

cAL11

CLK2–NUMA1

chr1:155267627 - chr11:72044112

Genic - Genic

CTX

Yes

cAL11

KANSL1–ARL17B

chr17:46087894 - chr17:46356512

Genic - Genic

ITX

Yes

cAL12

ERBB2IP–SQRDL

chr5:66078502 - chr15:45654289

Genic - Genic

CTX

Yes

cAL12

RBPJ–TRAPPC12

chr4:26193798 - chr2:3436901

Nongenic - Genic

CTX

Yes

Sample

CTX; interchromosomal translocation; ITX: intrachromosomal translocation; iDel: interstitial deletion; WGS: whole-genome sequencing. Haematologica | 107 July 2022

1623

ARTICLE - Whole-genome profiling of pcALCL

A.N. Bastidas Torres et al.

deaminase (AID)-mediated somatic hypermutation were sociated with cancer (134 candidate cancer genes, 24 bona smaller contributors to the mutational landscape (Figure 5C). fide oncogenes/tumor suppressors) (Online Supplementary A total of 603 genes were found to carry somatic muta- Table S11). All small-scale mutations were patient-specific. tions across seven patients, 158 of which were genes as- Gene set analysis of mutated genes with Panther ident-

Figure 3. Landscape of copy number alterations reveals recurrent deletion of PRDM1 in primary cutaneous anaplastic large cell lymphoma. (A) Human chromosome ideograms showing regions of gain and loss detected by whole-genome sequencing in 12 primary cutaneous anaplastic large cell lymphoma (pcALCL) genomes. Blue bars to the right of the chromosomes depict regions of loss whereas red bars to the left of the chromosomes depict regions of gain. (B) Diagram representing deletions (light blue bars) in six patients with pcALCL. The minimal common region (MCR) of losses at 6q21 in the analyzed patients maps to PRDM1, a gene encoding a transcription factor that attenuates proliferation and survival of T cells. (C) Sashimi plots showing expression of PRDM1 in patients cAL6 (affected by a focal deletion, purple) and cAL9 (with intact alleles, green). Patient cAL6 expressed a PRDM1 transcript that lacked exons 5, 6 and 7 as a result of a monoallelic focal deletion (blue bar, 4.6 Kb). This structurally defective transcript encodes a protein lacking the Zn finger domains of PRDM1 which are required for its functionality. IDEL: interstitial deletion. (D) Recurrent copy number alterations affecting EZH2, PRDM1, RBFOX1, STK24 and TNFRSF8 in pcALCL were validated by droplet digital polymerase chain reaction. Ctrl: healthly CD4+ T cells. Haematologica | 107 July 2022

1624

ARTICLE - Whole-genome profiling of pcALCL

A.N. Bastidas Torres et al.

ified enrichment of genes involved in signaling mediated by G-proteins (β-adrenergic receptors, glutamate receptors, etc.) and PI-3-K signaling (Figure 5D; Online Supplementary Table S12). Analysis of mutated genes with NCG 6.0 database revealed a group of cancer genes with established roles in chromatin modification (i.e., ARID1B, BCOR, KDM5A, PRDM16, TRRAP), hematopoiesis (i.e., FLT3, FSTL3 and GATA2), signaling via the PI-3-K and MAPK pathways (i.e., ERBB4, MAP3K1, PDPK1, PIK3R1, RET and WNK2) and G-protein signaling (i.e., CREB3L1, DRD5, RGS7 and RGS12). Of note, although JAK-STAT mutations leading to STAT3 overactivation are common in sALCL, no mutations in JAK-STAT pathway genes were found in our pcALCL patients. We identified a group of 30 recurrently mutated genes (npa=2-5) in our cohort (Online Supplementary Table S10), tients which included (putative) oncogenes (i.e., CDK14, CNOT1,

LRP2, PDPK1) and tumor suppressors (i.e., CSMD1, LRP1B, PIK3R1) (Online Supplementary Figure S5). Within this group, PDPK1 and PIK3R1 were genes of special interest. PDPK1 encodes a master serine/threonine kinase that activates AKT, a kinase with a central role in the PI-3-K/AKT pathway. Two of 12 patients had mutations in the PH domain of PDPK1 (cAL4: p.T518M; cAL10: p.W535L) (Figure 5E and F). In contrast, PIK3R1 encodes the regulatory (inhibitory) subunit of class IA phosphoinositide 3-kinases (PI-3-K). Class IA PI-3-K mediate the production of PIP3, a membrane phospholipid that induces PI-3-K/AKT pathway activation by recruiting PDPK1 and AKT to the plasma membrane where AKT activation occurs. Activated AKT triggers downstream signaling events leading to cell proliferation and survival. Two of 12 patients had splice site mutations in PIK3R1 (cAL5: G>T at chr5:68,296,342; cAL9: G>A at chr5:68,293,835) (Figure 5E, F).

Figure 4. Distribution of recurrent genetic alterations in primary cutaneous anaplastic large cell lymphoma. First panel: recurrent chromosomal rearrangements impacting cancer genes. Second panel: recurrent large-scale copy number alterations (CNA) (>3 Mb). Third panel: focal minimal common regions (MCR) shared by copy number alterations (CNA). Fourth panel: cancer genes recurrently affected by indels/single nucleotide variants (SNV). CTX: interchromosomal rearrangement; ITX: intrachromosomal rearrangement; Del: deletion. Haematologica | 107 July 2022

1625

ARTICLE - Whole-genome profiling of pcALCL

A.N. Bastidas Torres et al.

Differentially expressed genes and fusion transcripts Given the fact that the CD4+ T-cell subset giving rise to pcALCL is unknown at present, we compared gene expression in pcALCL with gene expression in a control

group formed by several CD4+ T-cell subsets to find in this lymphoma abnormal expression profiles absent in various normal phenotypes of CD4+ T cells (Online Supplementary Figure S3). This comparison identified 3,162 DE genes (1,716

E B

Figure 5. Landscape of small-scale mutations reveals recurrent events affecting PI-3-K/AKT pathway genes in primary cutaneous anaplastic large cell lymphoma. (A) Number of small-scale mutations per patient. The distribution of missense, nonsense, frameshift and splice site mutations per patient is shown. (B) Distribution of nucleotide substitutions per patient. (C) Median contribution of COSMIC mutational signatures in primary cutaneous anaplastic large cell lymphoma (pcALCL) (n=7) and Sézary syndrome (n=30). AID SHM: activation-induced cytidine deaminase-mediated somatic hypermutation; DSBRD: doublestrand break repair deficiency; SS: Sézary syndrome; UV: ultraviolet. (D) Gene set analysis of genes affected by indels and single nucleotide variants (SNV) in pcALCL. Mutated genes were primarily involved in G-protein signaling and the PI-3-K pathway. (E) Diagram showing small-scale mutations affecting PIK3R1 and PDPK1 in patients with pcALCL. Splice site mutations in PIK3R1 are predicted to affect the three known protein isoforms (i.e., p85α, p55α and p50α) encoded by this gene. Missense mutations in PDPK1 impacted the PH domain of its encoded protein. (F) Sanger chromatograms showing validation of splice site and missense mutations in PIK3R1 and PDPK1, respectively. Haematologica | 107 July 2022

1626

ARTICLE - Whole-genome profiling of pcALCL

A.N. Bastidas Torres et al.

upregulated, 1,446 downregulated, false discovery rate [FDR] >0.01) (Figure 6A; Online Supplementary Table S15). The analysis not only confirmed upregulation of genes found to be overexpressed in pcALCL in prior studies (e.g., CCR8, CCR10, EZH2, IRF4, MIR155HG and TNFRSF8) (Figure 6B),11,14-16 but also revealed upregulation of a group of (proto)-oncogenes (e.g., CSF1R, EGFR, FGFR1/2/3, MET and PDGFRA/B) involved in signaling through the P-I-3K/AKT pathway and/or the MAPK pathway (Figure 6C). Of note, CSF1R overexpression has been reported in sALCL before.27 Downregulated cancer genes included inhibitors of the PI-3K/AKT pathway (i.e., PIK3IP1 and PIK3R1), inhibitors of the MAPK pathway (i.e., DUSP2, DUSP16, GPS2) and tumor suppressors implicated in other hematological cancers (i.e., FHIT, RHOH and TNFAIP3) (Figure 6C). Twenty-five fusion transcripts were identified in our cohort (Table 1), nine of which were formed by genes with established roles in cancer (i.e., ERBB2IP–SQRDL, ETS1– SLC37A4, GPS2–CHD3, NFAT5–WWP2, SIRT3–NFKB2, SMARCAD1–TP63, TNK1–GPS2, TP53–TNK1 and TP63–SMARCAD1). Deregulated cellular processes/pathways In order to investigate deregulated cellular processes/pathways in pcALCL, we analyzed up- and downregulated genes separately with GeneAnalytics. Upregulated canonical signal transduction routes were ERK signaling (enrichment score: 166.10), phospholipase C signaling (enrichment score: 86.47), PAK signaling (enrichment score: 59.21), AKT signaling (enrichment score: 49.09) and G-protein couple receptor (GPCR) signaling (enrichment score: 48.91). In addition, the analysis revealed upregulation of adhesion-related cellular processes (e.g., focal adhesion, integrin pathway) (Figure 6D). In contrast, all downregulated profiles were signaling processes leading to T-cell activation (T-cell receptor (TCR) signaling, Icos-IcosL pathway, etc.) (Figure 6E). Transcriptome analysis with Enrichr using the ARCHS4 database revealed that genes known to be co-expressed with receptor tyrosine kinases DDR2 and PDGFR (especially PDGFRβ) were the most overrepresented in the group of upregulated genes (Figure 6F), suggesting that these receptors and/or their associated signal transduction networks might play relevant roles in the disease. DDR2 binds to collagen and activates signaling pathways (e.g., MAPK pathway) implicated in cell adhesion, proliferation and extracellular matrix (ECM) remodeling. Similarly, PDGFR binds to its cognate ligands (PDGFs) and activates the PI-3-K/AKT and MAPK pathways, promoting cell proliferation and survival. In order to validate the upregulation of DDR2 and PDGFRB in pcALCL, transcript abundance of these genes was quantified by RT-qPCR in pcALCL tumors and a control group formed by benign inflammatory dermatoses (BID) and healthy CD4+ T cells (HC). Both genes were confirmed to be consistently upregulated in pcALCL in comparison to the control group (P<0.0001, Mann-Whitney U test) (Fig-

ure 6G). Additionally, we used RT-qPCR to assess in these two groups the expression of tumor suppressors GPS2 and PIK3R1, which displayed recurrent genetic defects and were downregulated in pcALCL according to the sequencing data. The analysis confirmed the downregulation of both tumor suppressors in pcALCL as well (P<0.05, MannWhitney U test) (Figure 6H). Since the PI-3-K/AKT pathway was the most consistently affected canonical signal transduction route in pcALCL (i.e., mutations in PDPK1/PIK3R1, generalized PIK3R1 downregulation, AKT signature), we investigated the presence of activated AKT (pAKT) by immunohistochemistry (IHC) in eight sequenced patients with available tumor tissue (i.e., cAL34/7-12). Robust AKT activation was observed in seven of the eight evaluated patients (Figure 7A; Online Supplementary Figure S8). Surprinsingly, although our pcALCL patients carried no mutations in JAK-STAT pathway genes, we detected STAT3 activation (pSTAT3) by IHC in three of eight evaluated patients. This suggests that STAT3 signaling might be induced in some cases by alternative non-JAK-STAT molecular alterations or via mutation-independent mechanisms (Figure 7B; Online Supplementary Figures S4 and S8).

Discussion In order to gain insight into the pathogenetic basis of pcALCL, we investigated rearrangements, CNA, smallscale mutations and gene expression in this lymphoma by analyzing genome (WGS, WES) and transcriptome (RNAseq) data. In agreement with previous studies, we observed solitary cases with DUSP22 and TP63 rearrangements in our cohort, which are events known to occur at low frequencies in pcALCL.10,28 We also confirmed that losses within 6q and gains within 7q were the most recurrent large-scale chromosomal imbalances. Moreover, similar to other CTCL variants (e.g., mycosis fungoides, Sézary syndrome, cutaneous γ/d T-cell lymphomas), we found that pcALCL displayed a predominant C>T mutational signature which was mainly attributable to UV light exposure and aging. Although UV-associated mutations appear to be a common feature among distinct CTCL variants, their significance in the lymphomagenesis of this group of neoplasms remains unclear. Even though pcALCL exhibited a manifest inter-patient heterogeneity in terms of number of genetic abnormalities and affected genes, our analysis showed that the cell cycle, T-cell physiology regulation, transcription and particularly signaling through the PI-3-K and MAPK pathways, were commonly affected cellular processes/pathways in our cohort. Recurrent genetic alterations involving cancer-associated genes included rearrangements involving GPS2, LINC-PINT and TNK1, deletion of PRDM1 and TNFRSF14, amplification of TNFRSF8 (CD30)

Haematologica | 107 July 2022

1627

ARTICLE - Whole-genome profiling of pcALCL

A.N. Bastidas Torres et al.

and EZH2, and small-scale mutations in CDK14, LRP1B, PDPK1 and PIK3R1. Deletion of PRDM1 and gain of EZH2, each observed in six of 12 patients, were the two most frequent genetic alter-

ations in our cohort. PRDM1, which is recurrently deleted in sALCL too, acts as a tumor suppressor.29 Prior research showed that reintroduction of PRDM1 in a sALCL cell line triggered cell proliferation impairment, cell cycle block

Continued on following page. Haematologica | 107 July 2022

1628

ARTICLE - Whole-genome profiling of pcALCL

A.N. Bastidas Torres et al.

Figure 6. RNA sequencing supports upregulation of signaling via the PI-3-K/AKT pathway, the MAPK pathway and G-proteins in primary cutaneous anaplastic large cell lymphoma. (A) Heatmap showing 3,162 differentially expressed genes (1,716 upregulated, 1,446 downregulated, false discovery rate [FDR] <0.01) in primary cutaneous anaplastic large cell lymphoma (pcALCL) when compared to CD4+ T cells. (B) Heatmap showing expression of genes with diagnostic and/or pathogenic relevance in pcALCL identified in prior studies. RNA sequencing confirmed the upregulation of several markers and oncogenes that characterize this lymphoma. (C) Heatmap showing (proto)-oncogenes (Proto-OG) and tumor suppressors (TSG) deregulated in pcALCL which are involved in the PI-3-K/AKT pathway and the MAPK pathway. (D) Upregulated cellular processes/pathways in pcALCL as determined by gene set pathway analysis with GeneAnalytics. (E) Downregulated cellular processes/pathways in pcALCL as determined by gene set pathway analysis with GeneAnalytics. (F) Top three human kinases for which known co-expressed genes were overrepresented in the transcriptome of pcALCL. (G) Validation of DDR2 and PDGFRB upregulation in pcALCL by reverse transcriptase quantitative polymerase chain reaction (RT-qPCR). (H) Validation of GPS2 and PIK3R1 downregulation in pcALCL by RT-qPCR. (*P<0.05, ****P<0.0001); pcALCL (n=12), BID (n=8), HC (n=4). BID: benign inflammatory dermatoses; HC: healthy CD4+ T cells.

and apoptosis.29 Notably, PRDM1 reintroduction also downregulated in these cells the expression of MIR155HG (miR155 precursor) and IRF4, two genes that are consistently overexpressed in sALCL and pcALCL.29 On the other hand, previous studies have shown that EZH2 is overexpressed in pcALCL.11,14 EZH2 appears to operate as an oncogene in the disease by epigenetically silencing TXNIP and CXCL10, two genes with anti-tumor effects in pcALCL.14 Although the molecular mechanisms leading to EZH2 overexpression in pcALCL require further study, our data suggest that gene amplification may be one of them. Four additional genes implicated in cancer underwent recurrent alterations in pcALCL (LINC-PINT, LRP1B, RBFOX1 and USH2A). LINC-PINT and RBFOX1 behave as tumor suppressors in gastric cancer and glioblastoma.30-33 LRP1B, an LDL family receptor involved in endocytosis, is inactivated by genetic and epigenetic mechanisms in several cancers (gastric, renal, thyroid, etc.).34-36 USH2A, one of the two most SNV-mutated genes in our cohort, is the 21st most mutated gene in breast cancer according to The Cancer Genome Atlas (TCGA)37 yet its function remains unclear. Genetic alterations impacting genes associated with the P-I-3K/AKT pathway, the MAPK family of pathways (via ERK, JNK and p38 kinases), and to a lesser extent G-protein signaling, were the most notable common denominator between patients with pcALCL. Nine of 12 patients had abnormalities in one or more genes linked to the aforementioned pathways (Online Supplementary Table S20). Patient-specific genetic alterations, which accounted for the vast majority of events, included rearrangements and small-scale mutations in CREB3L1, DGKI, DRD5, ERBB2IP, ERBB4, GRM8, MAP3K1, MAP3K20, MET, PIK3C2B, PPM1L, PTPRK, RET, RGS7, RGS12 and WNK2. In addition, recurrent molecular defects were found in two inhibitors of the MAPK pathway (GPS2 and TNK1) and two key regulators of the PI-3-K/AKT pathway (PDPK1 and PIK3R1). GPS2 and TNK1 had structural defects in two of 12 patients. GPS2 formed dysfunctional fusions with other cancer genes (CHD3 and TNK1), inactivating both fusion partners in a single event. GPS2 is lowly expressed in liposarcoma and rearranged in spindle cell sarcoma.38

In vitro knockdown of GPS2 has been shown to enhance proliferation, migration and dedifferentiation of liposarcoma cells, supporting its role as a tumor suppressor in this neoplasm.38 Similarly, rearrangements abolished TNK1 expression or rendered it dysfunctional in two of 12 patients. TNK1 is often deleted and lowly expressed in diffuse large B-cell lymphoma.39 Mouse models carrying monoallelic or biallelic TNK1 inactivation exhibit tissues with higher levels of RAS activation and develop spontaneous tumors at high rates, including lymphomas and carcinomas.21 Curiously, although wild-type TNK1 exhibits antiproliferative properties, C-terminus truncated forms of the protein, like the one observed in patient cAL3, appear to activate proliferation-promoting kinases such as AKT and STAT3/5.22 We observed small-scale mutations affecting PDPK1 or PIK3R1 in four of 12 patients. PDPK1 is frequently amplified or overexpressed in human cancers, but uncommonly affected by pathogenic SNV.40 Surprinsingly, we found two patients with SNV (p.T518M; p.W535L) affecting the PH domain of this kinase. PDPK1 becomes contitutively active as a monomer upon dissociation from its inactive homodimeric form.41 As activation of PDPK1 requires autophosphorylation of residues in the PH domain to destabilize the inactive homodimeric state, these mutations might impact the stability of the latter favoring an active monomeric state. This mechanism has been demonstrated for the PH domain mutation p.T513E.41 In contrast, loss-of-function mutations in PIK3R1 occur in glioblastoma, endometrial carcinoma and immune disorders.42 One of the two splice site mutations found in our cohort is causative of PASLI-like disease, a disorder characterized by immunodeficiency, lymphoproliferation, antibody defects and senescence of CD8+ T cells. This splice site mutation (G>A at chr5:68,293,835), which results in exon 11 skipping during mRNA maturation, leads to overactivation of the PI-3-K/AKT pathway in immune cells from these patients.42 The other observed splice site mutation (G>T at chr5:68,296,342) is predicted to result in exon 15 skipping, and consequently, deletion of the C-terminal SH2 domain of the protein which is essential for inhibiting the catalytic subunit of PI-3-K.

Haematologica | 107 July 2022

1629

ARTICLE - Whole-genome profiling of pcALCL

A.N. Bastidas Torres et al.

In line with the genomic data, transcriptome analysis revealed upregulation of several signal transduction routes associated with the PI-3-K/AKT pathway, the MAPK pathway and G-protein signaling (e.g., ERK, PLC and AKT). The PI-3K/AKT pathway, the signaling route most prominently altered in our cohort, was confirmed to be activated in our patients through IHC. Thus, suppresion of this pathway using PI-3-K inhibitors (e.g., duvelisib) could be explored as an alternative therapy for patients who do not respond to skin-directed therapies or with extracutaneous involvement. Upregulated genes in pcALCL matched expression patterns primarily associated with overexpression of receptor kinases DDR2 and PDGFRβ, suggesting a potential role of these receptors in the development of the disease. The upregulation of DDR2 and PDGFRB in pcALCL was con-

firmed by RT-qPCR. DDR2 is overexpressed in subsets of patients with sALCL and most Hodgkin lymphoma cases.43,44 Similarly, PDGFR is commonly overexpressed in sALCL and various other human cancers.45,46 Of note, PDGFR blockade with imatinib has been shown to be therapeutically effective in murine models of sALCL and a patient with refractory sALCL.45 Since our data suggest that PDGFR is overexpressed in pcALCL too, blockade of this receptor might also have a positive therapeutic effect on patients with pcALCL. However, validation of our molecular findings in large series of patients will be essential before the eventual clinical exploration of these potential therapies. In agreement with prior reports, our data showed that pcALCL shares some pathogenic features with sALCL such as the overexpression of several markers and oncogenes

Figure 7. Immunohistochemistry shows generalized AKT activation and ocassional STAT3 activation in primary cutaneous anaplastic large cell lymphoma. (A) Activation of the P-I-3K/AKT pathway (pAKT) was confirmed by immunohistochemistry (IHC) on tumor tissue from 7 of 8 evaluated patients (i.e., cAL3-4/7-12). Tumor cells exhibited activated AKT in the nucleus and the cytoplam. (B) Activation of STAT3 signaling (pSTAT3) was detected by IHC on tumor tissue from three of eight evaluated patients. Tumor cells displayed activated STAT3 in the nucleus. Normal skin (control) displayed AKT and STAT3 activation in keratinocytes and endothelial cells. NHS: normal human skin. Scale bar, 50 mm. Haematologica | 107 July 2022

1630

ARTICLE - Whole-genome profiling of pcALCL

A.N. Bastidas Torres et al.

(e.g., CSF1R, IRF4, MIR155HG, PDGFRB and TNFRSF8), the deletion of PRDM1, and the occasional presence of specific rearrangements (e.g., DUSP22 and TP63). However, our cohort displayed recurrent molecular alterations in cancer genes thus far not associated with sALCL (e.g., GPS2, LINCPINT, PDPK1, PIK3R1 and TNK1). In addition, our patients were devoid of alterations in JAK-STAT pathway genes which are common in sALCL. Still, our results should be interpreted with caution as future whole-genome studies involving larger cohorts are necessary to obtain a more precise estimate of the frequency of molecular alterations underlying pcALCL, including possible JAK-STAT mutations. Despite the absence of JAK-STAT alterations, some of our pcALCL patients exhibited STAT3 activation, suggesting that STAT3 signaling was either induced by non-JAK-STAT molecular alterations (e.g., oncogenic TNK1 variant in patient cAL3)22 or mediated by mutation-independent mechanisms in these individuals. In line with the latter, ALK– sALCL cell lines lacking JAK-STAT alterations have been found to induce STAT3 signaling through aberrant autocrine cytokine receptor stimulation.47 Yet, alternative activation mechanisms of JAK-STAT signaling (e.g., signaling crosstalk, epigenetics-related) cannot be excluded. In summary, our study has uncovered novel genetic alterations with pathogenic significance in pcALCL and established relevant molecular characteristics of this lymphoma. Firstly, pcALCL was found to be genetically heterogeneous. Secondly, pcALCL displayed a mutational signature mainly attributable to UV light and aging with smaller contributions from DSB repair deficiency and AID-mediated somatic hypermutation. Thirdly, some pcALCL cases displayed STAT3 activation even in the absence of canonical JAK-STAT alterations which generally underlie aberrant STAT3 signaling in other lymphomas. Finally, pcALCL appears to develop as a result of defects in genes with roles in the cell cycle, T-cell

physiology, transcription, and especially signaling via the PI3-K/AKT pathway, the MAPK pathway and G-proteins. Disclosures No conflicts of interest to disclose. Contributions ANBT, RM, KQ, RW, MV and CPT conceptualized and designed the project; ANBT, RM and CPT wrote the manuscript; ANBT, LG, HM, CA and TK performed the bioinformatic analyses; ANBT and JO performed the experiments; ANBT and RM analyzed the results, interpreted the data and produced figures and tables; KQ, RW and MV provided valuable biological specimens; ANBT, RM, LG, JO, HM, CA, TK, KQ, RW, MV and CPT revised and approved the final manuscript. Acknowledgments The authors thank Julian van Toledo for valuable bioinformatic support. Funding This study was funded by the Dutch Cancer Society (KWF, grant UL2013‐6104), the Netherlands Organization for Health Research and Development (ZonMw, grant 4043500-98-4027/435004503), Takeda Nederland B.V. and the Foundation for Pathological Research and Development (S.P.O.O., grant SPOO-2016003). Data sharing statement The sequencing data that support the findings of this study are openly available in the European Genome-Phenome Archive (EGA) at https://ega-archive.org, reference number [EGAS00001004429]. Additional data and protocols can be made available to other investigators upon request.

References lymphoproliferative disorders. Blood. 2014;124(25):3768-3771. 6. Luchtel RA, Zimmermann MT, Hu G, et al. Recurrent MSC (E116K) mutations in ALK-negative anaplastic large cell lymphoma. Blood. 2019;133(26):2776-2789. 7. Luchtel RA, Dasari S, Oishi N, et al. Molecular profiling reveals immunogenic cues in anaplastic large cell lymphomas with DUSP22 rearrangements. Blood. 2018;132(13):1386-1398. 8. Hu G, Dasari S, Asmann YW, et al. Targetable fusions of the FRK tyrosine kinase in ALK-negative anaplastic large cell lymphoma. Leukemia. 2017;32(2):565-569. 9. Crescenzo R, Abate F, Lasorsa E, et al. Convergent mutations and kinase fusions lead to oncogenic STAT3 activation in anaplastic large cell lymphoma. Cancer Cell. 2015;27(4):516-532. 10. Prieto-Torres L, Rodriguez-Pinilla SM, Onaindia A, Ara M, Requena L, Piris MA. CD30-positive primary cutaneous lymphoproliferative disorders: molecular alterations and targeted therapies. Haematologica. 2019;104(2):226-235.

1. Willemze R, Jaffe ES, Burg G, et al. WHO-EORTC classification for cutaneous lymphomas. Blood. 2005;105(10):3768-3785. 2. Bekkenk MW, Geelen FA, van Voorst Vader PC, et al. Primary and secondary cutaneous CD30(+) lymphoproliferative disorders: a report from the Dutch Cutaneous Lymphoma Group on the longterm follow-up data of 219 patients and guidelines for diagnosis and treatment. Blood. 2000;95(12):3653-3661. 3. Melchers RC, Willemze R, van de Loo M, et al. Clinical, histologic, and molecular characteristics of anaplastic lymphoma kinasepositive primary cutaneous anaplastic large cell lymphoma. Am J Surg Pathol. 2020;44(6):776-781. 4. Melchers RC, Willemze R, Vermaat JSP, et al. Outcomes of rare patients with a primary cutaneous CD30+ lymphoproliferative disorder developing extracutaneous disease. Blood. 2020;135(10):769-773. 5. Velusamy T, Kiel MJ, Sahasrabuddhe AA, et al. A novel recurrent NPM1-TYK2 gene fusion in cutaneous CD30-positive

Haematologica | 107 July 2022

1631

ARTICLE - Whole-genome profiling of pcALCL

A.N. Bastidas Torres et al.

11. van Kester MS, Tensen CP, Vermeer MH, et al. Cutaneous anaplastic large cell lymphoma and peripheral T-cell lymphoma NOS show distinct chromosomal alterations and differential expression of chemokine receptors and apoptosis regulators. J Invest Dermatol. 2010;130(2):563-575. 12. Laharanne E, Oumouhou N, Bonnet F, et al. Genome-wide analysis of cutaneous T-cell lymphomas identifies three clinically relevant classes. J Invest Dermatol. 2010;130(6):1707-1718. 13. Wada DA, Law ME, Hsi ED, et al. Specificity of IRF4 translocations for primary cutaneous anaplastic large cell lymphoma: a multicenter study of 204 skin biopsies. Mod Pathol. 2011;24(4):596-605. 14. Yi S, Sun J, Qiu L, et al. Dual role of EZH2 in cutaneous anaplastic large cell lymphoma: promoting tumor cell survival and regulating tumor microenvironment. J Invest Dermatol. 2018;138(5):1126-1136. 15. Benner MF, Ballabio E, van Kester MS, et al. Primary cutaneous anaplastic large cell lymphoma shows a distinct miRNA expression profile and reveals differences from tumor-stage mycosis fungoides. Exp Dermatol. 2012;21(8):632-634. 16. Sandoval J, Diaz-Lagares A, Salgado R, et al. MicroRNA expression profiling and DNA methylation signature for deregulated microRNA in cutaneous T-cell lymphoma. J Invest Dermatol. 2015;135(4):1128-1137. 17. van Doorn R, Slieker RC, Boonk SE, et al. Epigenomic analysis of Sezary syndrome defines patterns of aberrant DNA methylation and identifies diagnostic markers. J Invest Dermatol. 2016;136(9):1876-1884. 18. Marin-Bejar O, Mas AM, Gonzalez J, et al. The human lncRNA LINC-PINT inhibits tumor cell invasion through a highly conserved sequence element. Genome Biol. 2017;18(1):202. 19. Carreira-Rosario A, Bhargava V, Hillebrand J, Kollipara RK, Ramaswami M, Buszczak M. Repression of Pumilio protein expression by Rbfox1 promotes germ cell differentiation. Dev Cell. 2016;36(5):562-571. 20. Spain BH, Bowdish KS, Pacal AR, et al. Two human cDNAs, including a homolog of Arabidopsis FUS6 (COP11), suppress Gprotein- and mitogen-activated protein kinase-mediated signal transduction in yeast and mammalian cells. Mol Cell Biol. 1996;16(12):6698-6706. 21. Hoare S, Hoare K, Reinhard MK, Lee YJ, Oh SP, May WS Jr. Tnk1/Kos1 knockout mice develop spontaneous tumors. Cancer Res. 2008;68(21):8723-8732. 22. Gu TL, Cherry J, Tucker M, Wu J, Reeves C, Polakiewicz RD. Identification of activated Tnk1 kinase in Hodgkin's lymphoma. Leukemia. 2010;24(4):861-865. 23. Martins G, Calame K. Regulation and functions of Blimp-1 in T and B lymphocytes. Annu Rev Immunol. 2008;26:133-169. 24. Ling P, Lu TJ, Yuan CJ, Lai MD. Biosignaling of mammalian Ste20-related kinases. Cell Signal. 2008;20(7):1237-1247. 25. Steinberg MW, Cheung TC, Ware CF. The signaling networks of the herpesvirus entry mediator (TNFRSF14) in immune regulation. Immunol Rev. 2011;244(1):169-187. 26. Kim J, McMillan E, Kim HS, et al. XPO1-dependent nuclear export is a druggable vulnerability in KRAS-mutant lung cancer. Nature. 2016;538(7623):114-117. 27. Mathas S, Kreher S, Meaburn KJ, et al. Gene deregulation and spatial genome reorganization near breakpoints prior to formation of translocations in anaplastic large cell lymphoma. Proc Natl Acad Sci U S A. 2009; 106(14):5831-5836. 28. Schrader AM, Chung YY, Jansen PM, et al. No TP63 rearrangements in a selected group of primary cutaneous CD30+ lymphoproliferative disorders with aggressive clinical course. Blood. 2016;128(1):141-143.

29. Boi M, Rinaldi A, Kwee I, et al. PRDM1/BLIMP1 is commonly inactivated in anaplastic large T-cell lymphoma. Blood. 2013;122(15):2683-2693. 30. Zhang M, Zhao K, Xu X, et al. A peptide encoded by circular form of LINC-PINT suppresses oncogenic transcriptional elongation in glioblastoma. Nat Commun. 2018;9(1):4475. 31. Feng H, Zhang J, Shi Y, Wang L, Zhang C, Wu L. Long noncoding RNA LINC-PINT is inhibited in gastric cancer and predicts poor survival. J Cell Biochem. 2019;120(6):9594-9600. 32. Hu J, Ho AL, Yuan L, et al. Neutralization of terminal differentiation in gliomagenesis. Proc Natl Acad Sci U S A. 2013;110(36):14520-14527. 33. Sengupta N, Yau C, Sakthianandeswaren A, et al. Analysis of colorectal cancers in British Bangladeshi identifies early onset, frequent mucinous histotype and a high prevalence of RBFOX1 deletion. Mol Cancer. 2013;12:1. 34. Lu YJ, Wu CS, Li HP, et al. Aberrant methylation impairs low density lipoprotein receptor-related protein 1B tumor suppressor function in gastric cancer. Genes Chromosomes Cancer. 2010;49(5):412-424. 35. Ni S, Hu J, Duan Y, et al. Down expression of LRP1B promotes cell migration via RhoA/Cdc42 pathway and actin cytoskeleton remodeling in renal cell cancer. Cancer Sci. 2013;104(7):817-825. 36. Prazeres H, Torres J, Rodrigues F, et al. Chromosomal, epigenetic and microRNA-mediated inactivation of LRP1B, a modulator of the extracellular environment of thyroid cancer cells. Oncogene. 2011;30(11):1302-1317. 37. Cancer Genome Atlas N. Comprehensive molecular portraits of human breast tumours. Nature. 2012;490(7418):61-70. 38. Huang XD, Xiao FJ, Wang SX, et al. G protein pathway suppressor 2 (GPS2) acts as a tumor suppressor in liposarcoma. Tumour Biol. 2016;37(10):13333-13343. 39. May WS, Hoare K, Hoare S, Reinhard MK, Lee YJ, Oh SP. Tnk1/Kos1: a novel tumor suppressor. Trans Am Clin Climatol Assoc. 2010;121:281-292. 40. Gagliardi PA, Puliafito A, Primo L. PDK1: at the crossroad of cancer signaling pathways. Semin Cancer Biol. 2018;48:27-35. 41. Ziemba BP, Pilling C, Calleja V, Larijani B, Falke JJ. The PH domain of phosphoinositide-dependent kinase-1 exhibits a novel, phospho-regulated monomer-dimer equilibrium with important implications for kinase domain activation: singlemolecule and ensemble studies. Biochemistry. 2013;52(28):4820-4829. 42. Lucas CL, Zhang Y, Venida A, et al. Heterozygous splice mutation in PIK3R1 causes human immunodeficiency with lymphoproliferation due to dominant activation of PI3K. J Exp Med. 2014;211(13):2537-2547. 43. Willenbrock K, Kuppers R, Renne C, et al. Common features and differences in the transcriptome of large cell anaplastic lymphoma and classical Hodgkin's lymphoma. Haematologica. 2006;91(5):596-604. 44. Renne C, Willenbrock K, Kuppers R, Hansmann ML, Brauninger A. Autocrine- and paracrine-activated receptor tyrosine kinases in classic Hodgkin lymphoma. Blood. 2005;105(10):4051-4059. 45. Laimer D, Dolznig H, Kollmann K, et al. PDGFR blockade is a rational and effective therapy for NPM-ALK-driven lymphomas. Nat Med. 2012;18(11):1699-1704. 46. Papadopoulos N, Lennartsson J. The PDGF/PDGFR pathway as a drug target. Mol Aspects Med. 2018;62:75-88. 47. Chen J, Zhang Y, Petrus MN, et al. Cytokine receptor signaling is required for the survival of ALK- anaplastic large cell lymphoma, even in the presence of JAK1/STAT3 mutations. Proc Natl Acad Sci U S A. 2017;114(15):3975-3980.

Haematologica | 107 July 2022

1632

ARTICLE - Non-Hodgkin Lymphoma

Total metabolic tumor volume as a survival predictor for patients with diffuse large B-cell lymphoma in the GOYA study Lale Kostakoglu,1 Federico Mattiello,2 Maurizio Martelli,3 Laurie H. Sehn,4 David Belada,5 Chiara Ghiggi,6 Neil Chua,7 Eva González-Barca,8 Xiaonan Hong,9 Antonio Pinto,10 Yuankai Shi,11 Yoichi Tatsumi,12 Christopher Bolen,13 Andrea Knapp,2 Gila Sellam,2 Tina Nielsen,2 Deniz Sahin,2 Umberto Vitolo14 and Marek Trněný15 Department of Radiology and Medical Imaging, University of Virginia, Charlottesville, VA, USA; 2F. Hoffmann-La Roche Ltd, Basel, Switzerland; 3Department of Translational and Precision Medicine, Sapienza University, Rome, Italy; 4BC Cancer Center for Lymphoid Cancer and the University of British Columbia, Vancouver, British Columbia, Canada; 5Fourth Department of Internal Medicine-Hematology, Charles University, Hospital and Faculty of Medicine, Hradec Králové, Czech Republic; 6Universitaria San Martino, Genoa, Italy; 7Cross Cancer Institute, University of Alberta, Edmonton, Alberta, Canada; 8Institut Català d’Oncologia, Institut d’Investigació Biomédica de Bellvitge, Universitat de Barcelona, Hospitalet de Llobregat, Barcelona, Spain; 9Fudan University Shanghai Cancer Center, Shanghai, China; 10Hematology-Oncology, Istituto Nazionale Tumori, Fondazione G. Pascale, IRCCS, Naples, Italy; 11Department of Medical Oncology, Beijing Key Laboratory of Clinical Study on Anticancer Molecular Targeted Drugs, National Cancer Center/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China; 12 Department of Patient Safety and Management, Kindai University Hospital and Department of Hematology and Rheumatology, Kindai University Faculty of Medicine, Osaka, Japan; 13 Genentech, Inc., South San Francisco, CA, USA; 14Multidisciplinary Oncology Outpatient Clinic, Candiolo Cancer Institute, FPO-IRCCS, Candiolo, Turin, Italy and 15First Department of Medicine, Charles University General Hospital, Prague, Czech Republic. 1

Correspondence: Lale Kostakoglu lk3qf@virginia.edu Received: March 1, 2021. Accepted: August 4, 2021. Prepublished: August 19, 2021. https://doi.org/10.3324/haematol.2021.278663 ©2022 Ferrata Storti Foundation Haematologica material is published under a CC BY-NC license

Abstract This retrospective analysis of the phase III GOYA study investigated the prognostic value of baseline metabolic tumor volume parameters and maximum standardized uptake values for overall and progression-free survival (PFS) in treatment-naïve diffuse large B-cell lymphoma. Baseline total metabolic tumor volume (determined for tumors >1 mL using a threshold of 1.5 times the mean liver standardized uptake value +2 standard deviations), total lesion glycolysis, and maximum standardized uptake value positron emission tomography data were dichotomized based on receiver operating characteristic analysis and divided into quartiles by baseline population distribution. Of 1,418 enrolled patients, 1,305 had a baseline positron emission tomography scan with detectable lesions. Optimal cut-offs were 366 cm3 for total metabolic tumor volume and 3,004 g for total lesion glycolysis. High total metabolic tumor volume and total lesion glycolysis predicted poorer PFS, with associations retained after adjustment for baseline and disease characteristics (high total metabolic tumor volume hazard ratio: 1.71, 95% confidence interval [CI]: 1.35– 2.18; total lesion glycolysis hazard ratio: 1.46; 95% CI: 1.15–1.86). Total metabolic tumor volume was prognostic for PFS in subgroups with International Prognostic Index scores 0–2 and 3–5, and those with different cell-of-origin subtypes. Maximum standardized uptake value had no prognostic value in this setting. High total metabolic tumor volume associated with high International Prognostic Index or non-germinal center B-cell classification identified the highest-risk cohort for unfavorable prognosis. In conclusion, baseline total metabolic tumor volume and total lesion glycolysis are independent predictors of PFS in patients with diffuse large B-cell lymphoma after first-line immunochemotherapy.

Introduction Current standard of care for patients with previously untreated diffuse large B-cell lymphoma (DLBCL) is immunochemotherapy with the anti-CD20 antibody rituximab

in combination with cyclophosphamide, doxorubicin, vincristine, and prednisone (CHOP).1,2 DLBCL is a heterogeneous disease that is reflected in transcriptionally defined, molecularly distinct subtypes.3,4 Consequently, 10–15% of patients have primary refractory disease, and 20–30% of

Haematologica | 107 July 2022

1633

ARTICLE - Total metabolic tumor volume predicts survival patients experience relapse after initial response to firstline therapy.5 Identification of high-risk patients is challenging with the currently available risk-stratification models.6,7 Gene expression profiling and other molecular signatures can predict outcome, independently of clinical International Prognostic Index (IPI) score.3,4 18 F-fluorodeoxyglucose (FDG) positron emission tomography (PET)/computed tomography (CT) is an important tool for the diagnosis, staging, and early evaluation of response to therapy in patients with DLBCL. FDG-PET/CT together with baseline quantitative measures of FDG uptake, which act as indicators of metabolically active whole-body disease burden, have demonstrated good prognostic value for the prediction of long-term survival outcomes in a number of small studies of patients with previously untreated DLBCL.8-15 Examples of quantitative PET variables currently being explored for potential prognostic value include total metabolic tumor volume (TMTV), total lesion glycolysis (TLG), and maximum standardized uptake value (SUVmax). Other studies have investigated the prognostic value of interim response assessment by PET, with this having limited predictive value,16,17 or end-oftreatment complete response by PET, which was shown to be associated with improved survival outcomes.18-20 The phase III GOYA study (clinicaltrials gov. Identifier: NCT01287741) compared the glycoengineered type II antiCD20 antibody, obinutuzumab, plus CHOP with rituximab plus CHOP in patients with previously untreated DLBCL.21 Although improvements in progression-free survival (PFS) have been demonstrated with obinutuzumab- versus rituximab-based therapy in follicular lymphoma22 and chronic lymphocytic leukemia,23 no significant PFS difference was found between the obinutuzumab and rituximab arms in the GOYA study (hazard ratio [HR]: 0.92; 95% confidence interval [CI]: 0.76–1.11; P=0.39).21 Here, we report a post hoc analysis of data from the GOYA study, evaluating the prognostic value of baseline quantitative PET parameters (TMTV, TLG and SUVmax), independently and combined with other risk factors, as potential prognostic markers for PFS and overall survival (OS) in this large population of patients with DLBCL. In GOYA, no statistically significant difference in PFS was found between treatment arms21 so the two arms were pooled for this post hoc analysis.

L. Kostakoglu et al. provided written informed consent. FDG-PET/CT scans were performed at baseline (≤35 days before the first dose of study treatment) and at end of treatment (6–8 weeks after the last antibody treatment), or in the event of early discontinuation (4–8 weeks after the last dose of antibody treatment). TMTV, TLG, and SUVmax were centrally determined by three experienced nuclear medicine physicians from each baseline PET/CT scan collected in real time during the study. TMTV was calculated using a tumor threshold of 1.5 times the mean SUV of the liver +2 standard deviations. Only those tumors that measured >1 mL were included in the TMTV calculation. Gene expression profiling using the NanoString Lymphoma Subtyping Research-Use-Only assay (NanoString Technologies, Inc., Seattle, WA, USA) was used for cell-oforigin classification. The primary objective of this retrospective analysis was to determine the prognostic value of baseline quantitative PET parameters (TMTV, TLG, and SUVmax) for OS and PFS in the overall patient population. Secondary objectives were to investigate their prognostic value according to IPI risk category and DLBCL cell-of-origin subtype.

Statistical analysis Statistical analyses were performed on the PET intent-totreat (ITT) population, which included all randomized patients with a baseline PET scan with detectable lesions. PFS and OS from randomization were estimated using the Kaplan–Meier method. In order to determine the optimal cut-offs, baseline TMTV and TLG were dichotomized based on time-dependent receiver operating characteristic (ROC) curves, selecting the 24-month landmark (i.e., considering investigator-assessed PFS events that occurred between randomization and 2 years). PFS and OS were evaluated according to these cut-offs and also an a priori cut-off of 300 cm3 (not validated) for TMTV.12 Covariate effects on PFS were estimated using multivariable Cox models, including treatment and study stratification factors. Hazard ratios were stratified for IPI score and number of planned chemotherapy cycles. Baseline TMTV values were dichotomized according to the ROC-derived cut-off, while for TLG, data were dichotomized according to the median TLG value at baseline. Quartiles of TMTV and TLG were defined based on their distribution in the available population at baseline. Associations between PFS and OS, and baseline TMTV and TLG Methods quartiles, were assessed using the Kaplan–Meier method. GOYA was a phase III, open-label, multicenter, randomized The association between PFS and baseline TMTV and TLG study. The study design and patient disposition have been quartiles was also evaluated separately in patients with described previously (see the Online Supplementary Ap- germinal center B cell (GCB) and activated B cell (ABC) pendix).21 GOYA was conducted in accordance with the In- cell-of-origin subtypes, and in patients with IPI scores of ternational Conference on Harmonisation guidelines for 0–2 and 3–5. Impact of PET parameters on PFS was also Good Clinical practice, and the protocol was approved by evaluated, adjusting for IPI score as well as cell-of-origin the ethics committees of participating centers. Patients subtype, to determine the prognostic value they provide Haematologica | 107 July 2022

1634

ARTICLE - Total metabolic tumor volume predicts survival

L. Kostakoglu et al.

in addition to other known factors. TMTV group also had a significantly greater median sum of Statistical analyses were performed using SAS v9.4 (SAS products of diameter (SPD) of up to six target lesions (Table Institute, Inc., Cary, NC, USA); P values <0.05 were con- 1). TMTV was associated with median time to randomizasidered significant. tion, with a shorter median time in the high versus low TMTV group (0.69 vs. 0.92 months). Dichotomized TLG data and both TMTV and TLG quartiles showed a similar overall pattern to dichotomized TMTV data, with respect to the two Results study populations, clinical characteristics and median time Patient population and baseline demographics to randomization (Online Supplementary Tables S2 and S3). The GOYA study enrolled 1,418 patients between July 2011 There was no association between quantitative PET and June 2014, with 1,414 patients included in the final measures and cell-of-origin. analysis population (GOYA ITT population). Of these, 1,305 Based on the data cut-off of January 31, 2018 for the final (92.3%) had a PET scan with detectable lesions at baseline analysis of GOYA, the median duration of follow-up from from which quantitative data could be obtained (PET ITT randomization to last contact or death was 48 months population; see the Online Supplementary Appendix). Base- (range, 0.1–78 months). line demographics and disease characteristics according to dichotomized baseline TMTV values in the PET ITT popu- Association between baseline quantitative positron lation are displayed in Table 1. Demographics and baseline emission tomography parameters and characteristics were similar in the GOYA ITT and PET ITT investigator-assessed progression-free and populations (Online Supplementary Table S1). The propor- overall survival tion of patients with an IPI score of 3–5, Ann Arbor stage Total metabolic tumor volume III/IV, Eastern Cooperative Oncology Group performance ROC analysis identified 366 cm3 as the optimal cut-off status (ECOG PS) ≥2, extranodal involvement and elevated for determining the association between TMTV and PFS. serum lactate dehydrogenase (LDH) was significantly Using this cut-off, Kaplan–Meier analysis showed that greater in the high versus low TMTV group, and the high PFS was shorter for patients with high versus low TMTV Table 1. Demographics and baseline characteristics according to high (≥366 cm3) or low (<366 cm3) baseline total metabolic tumor volume in the positron emission tomography intent-to-treat population (n=1,305).

Median age, years (range) Male, N (%) Median time from diagnosis to randomization, months (range) Ann Arbor stage, N (%) I or II III or IV IPI score, N (%) 0–2 3 4–5 ECOG PS, N (%) 0–1 2–3 Any extranodal involvement, N (%) Serum LDH elevated, N (%) Median SPD, mm (range) Cell-of-origin subtype, N (%) GCB ABC Unclassified

High TMTV (N = 636)

Low TMTV (N = 669)

61.0 (18-86) 358 (56.3) [N = 632] 0.69 (0.1-36.3)

62.0 (18-85) 333 (49.8) [N = 668] 0.92 (0.0-8.7)

P* 0.003 0.0198 0 0

91 (14.3) 545 (85.7)

223 (33.3) 446 (66.7) 0

251 (39.5) 234 (36.8) 151 (23.7) [N = 635] 521 (82.0) 114 (18.0) 458 (72.0) [N = 633] 501 (79.1) 8,198 (160-510,000) [N = 410] 239 (58.3) 103 (25.1) 68 (16.6)

484(72.3) 140 (20.9) 45 (6.7) – 623 (93.1) 46 (6.9) 418 (62.5) [N = 668] 244 (36.5) [N = 666] 2,570 (0-194,400) [N = 451] 255 (56.5) 124 (27.5) 72 (16.0)

0.0003 0 0 0.740

*Kruskal-Wallis rank sum test (numeric variables) or Fisher's exact test (categorical variables). ABC: activated B-cell-like; ECOG PS: Eastern Cooperative Oncology Group performance status; GCB: germinal center B-cell-like; IPI: International Prognostic Index; LDH: lactate dehydrogenase; SPD: sum of products of diameter of up to six target lesions; TMTV: total metabolic tumor volume. Haematologica | 107 July 2022

1635

ARTICLE - Total metabolic tumor volume predicts survival

L. Kostakoglu et al.

Figure 1. Kaplan–Meier analysis of investigator-assessed progression-free survival according to baseline total metabolic tumor volume cut-offs. Total metabolic tumor volume (TMTV) cut-off of (A) 366 cm3 and (B) 300 cm3.

Figure 2. Univariate Cox regression model evaluating factors associated with investigator-assessed progression-free survival (positron emission tomography intent-to-treat population; n=1,305). Total metabolic tumor volume (TMTV) high defined as ≥366 cm3 and TMTV low defined as <366 cm3; TLG high defined as ≥ median total lesion glycolysis (TLG) and TLG low defined as <median TLG. CI: confidence interval; ECOG PS: Eastern Cooperative Oncology Group performance status; IPI: International Prognostic Index; LDH: lactate dehydrogenase; SPD: sum of products of diameters of up to 6 target lesions.

(Figure 1A), with a stratified HR for PFS of 1.73 (95% CI: 1.40–2.13) (P<0.0001). The same association was identified when the a priori cut-off of 300 cm3 was evaluated (stratified HR: 1.53; 95% CI: 1.24–1.90; P<0.0001; Figure 1B). Four-year PFS rates were lower for patients with baseline TMTV ≥366 cm3 compared with <366 cm3 (4-year rate: 59.7% vs. 74.7%), and univariate followed by multivariable Cox regression analysis confirmed the association between high TMTV and shorter PFS (multivariable analysis: HR: 1.71; 95% CI: 1.35–2.18; P<0.0001; Figure 2; Online Supplementary Figure S1A). Inferior PFS was also independently associated with other characteristics in

the multivariable model, including Asian geographic region (Western European vs. Asian region: HR: 0.61; 95% CI: 0.48–0.78; P<0.0001; North American vs. Asian region: hazard ratio: 0.63; 95% CI: 0.47–0.86; P=0.004). The association between high TMTV and poorer PFS was stronger for the subgroup of patients with an IPI risk score of 3–5 (HR: 1.93; 95% CI: 1.41–2.65) than for those with a score of 0–2 (HR: 1.59; 95% CI: 1.18–2.14), when adjusted for treatment group, geographic region, sex and SPD (P=0.002 for the TMTV HR adjusted for interaction with IPI score; Online Supplementary Table S4). When TMTV was divided into quartiles according to base-

Haematologica | 107 July 2022

1636

ARTICLE - Total metabolic tumor volume predicts survival

L. Kostakoglu et al.

Figure 3. Kaplan–Meier analysis of investigator-assessed progression-free survival according to a baseline total lesion glycolysis cut-off of 3,004 g. PFS: progression-free survival; TLG: total lesion glycolysis.

line values, the highest quartile was associated with worse prognosis for investigator-assessed PFS, compared with the lowest quartile in both IPI categories 0–2 and 3–5 (quartile 4 vs. quartile 1, stratified HR: 1.95; 95% CI: 1.45– 2.61; P<0.0001; Online Supplementary Figure S2). Kaplan– Meier graphs for PFS according to baseline TMTV quartiles in patients with low and high IPI scores are shown in the Online Supplementary Figure S3A and B. OS was also found to be poorer in patients with high versus low TMTV, based on the tumor volume cut-off of 366 cm3 (stratified HR: 1.76; 95% CI: 1.34–2.30; P<0.0001; 4-year OS rate: 74.5% vs. 86.6%) and the 300 cm3 cut-off (Online Supplementary Figure S4).

(HR: 1.64; 95% CI: 1.21–2.23), compared with those who had a score of 0–2 (HR: 1.44; 95% CI: 1.07–1.94), when adjusted for treatment group, geographic region, sex and SPD (P=0.016 for the TLG hazard ratio adjusted for interaction with IPI score) (Online Supplementary Table S4). When TLG was divided into quartiles according to baseline values, the highest quartile was associated with worse prognosis for investigator-assessed PFS, compared with the lowest quartile (quartile 4 vs. quartile 1, stratified HR: 1.73; 95% CI: 1.30–2.31; P<0.0001). Kaplan–Meier graphs for PFS according to baseline TLG quartiles in patients with low and high IPI scores are shown in Online Supplementary Figure S3C and D. Maximum standardized uptake value

A utilizable SUVmax ROC curve for PFS could not be generROC analysis identified a cut-off of 3,004 g as the optimal ated, and it was therefore impossible to define a reasonvalue for assessing the association with PFS. Using this able cut-off for this parameter. An alternative ROC curve, cut-off, Kaplan–Meier analysis showed that patients with generated for 2-year PFS, was also not prognostic for surhigh TLG had significantly poorer PFS compared with pa- vival (Online Supplementary Figure S5). tients who had low TLG (P<0.0001) (Figure 3), with 4-year PFS rates of 61.3% and 73.5%, respectively. Univariate fol- Prognostic value of baseline quantitative positron lowed by multivariable Cox regression analysis confirmed emission tomography parameters according to the independent association with high TLG (multivariable cell-of-origin subtype analysis: HR: 1.46; 95% CI: 1.15–1.86; P=0.002) and other DLBCL cell-of-origin subtype information was available characteristics in the model, including Asian geographic for 861 patients in the PET-evaluable population (GCB, region (Western European vs. Asian region: HR: 0.62; 95% n=494 [57.4%]; ABC, n=227 [26.4%]; unclassified, n=140 CI: 0.48–0.79; P=0.0001; North American vs. Asian region: [16.3%]). The baseline characteristics for patients with HR: 0.64; 95% CI: 0.47–0.88; P=0.005) (Figure 2; Online available cell-of-origin subtype information (data not Supplementary Figure S1B). This association was stronger shown) were similar to those for the overall PET ITT in the subgroup of patients with an IPI risk score of 3–5 population. Total lesion glycolysis

Haematologica | 107 July 2022

1637

ARTICLE - Total metabolic tumor volume predicts survival

L. Kostakoglu et al.

Figure 4. Kaplan–Meier analysis of investigator-assessed progression-free survival according to baseline International Prognostic Index status and a total metabolic tumor volume cut-off of 366 cm3. Differences in progression-free survival (PFS) probability were significant between each subgroup (P<0.0001): IPI: International Prognostic Index; TMTV: total metabolic tumor volume.

High TMTV was prognostic for poorer PFS in patients with both ABC/unclassified DLBCL and in those with GCB DLBCL. Additionally, when assessed according to quartiles, higher TMTV was prognostic for poorer investigator-assessed PFS in patients with ABC/unclassified DLBCL (quartile 4 vs. quartile 1, stratified HR: 2.46; 95% CI: 1.49–4.07; P=0.0003) and in those with GCB DLBCL (stratified HR: 2.51; 95% CI: 1.44–4.38; P=0.0003) (Online Supplementary Figure S6A and B). Similar findings were observed for higher TLG (quartile 4 vs. quartile 1, ABC/unclassified: stratified HR: 2.10; 95% CI: 1.28–3.45; P=0.0022; GCB: stratified HR: 2.16; 95% CI: 1.27–3.68; P<0.0001; Online Supplementary Figure S6C, D). Similar to findings obtained for the PET ITT population, in the cell-of-origin subtype analysis, high TMTV was prognostic independently of IPI status based on the Cox multivariable analysis (Online Supplementary Table S5). The results for TLG were very similar to those obtained for TMTV (data not shown). Multivariable analysis with IPI and cell-of-origin classification The multivariable analysis with dichotomized TMTV and IPI score or cell-of-origin subtype further stratified the risk groups and identified the highest risk group with the worst prognosis. Kaplan–Meier analysis with dichotomized IPI score (0–2 and 3–5) and TMTV (≥366 and <366

groups) showed that patients with low IPI score and low TMTV (n=484) had a 74.3% 5-year PFS compared with 49.0% for patients with high IPI score and high TMTV (n=385); values for patients with high IPI score and low TMTV (n=185) were 71.1% compared with 65.5% for low IPI score with high TMTV (n=251; Figure 4). Kaplan–Meier analysis with dichotomized cell-of-origin subtype (GCB vs. non-GCB) and TMTV (≥366 and <366 groups) showed that patients with GCB subtype and low TMTV (n=255) had an 80.0% 5-year PFS, compared with 48.8% for patients with non-GCB subtype and high TMTV (n=171); values for patients with non-GCB subtype and low TMTV (n=196) were 65.1%, compared with 59.7% for GCB subtype and high TMTV (n=239) (Figure 5).

Discussion This post hoc analysis of the phase III GOYA study confirms that baseline quantitative PET parameters are prognostic for both PFS and OS in patients with DLBCL receiving first-line immunochemotherapy. Considerably worse prognosis was observed within the highest quartile (quartile 4) compared with the lowest for baseline TMTV (PFS and OS) and TLG (PFS). In general, TLG appeared slightly less prognostic than TMTV, regardless of whether it was split into quartiles or dichotomized.

Haematologica | 107 July 2022

1638

ARTICLE - Total metabolic tumor volume predicts survival ROC analysis identified 366 cm3 as the optimal tumor volume cut-off for TMTV in the GOYA PET ITT population. Despite the ROC cut-off being close to the a priori one and both, in turn, being quite close to the median (366 cm3 represents the 49th percentile), the methodology and the data-driven nature of the 366 cm3 cut-off should be acknowledged, along with the potential for slight overfitting. However, confidence in the prognostic value of the ROC cut-off can be taken from the finding that the stratified HR for PFS and OS were highly similar to those for the a priori cut-off. Thus, a TMTV cut-off of 300–400 cm3 can be suggested for further investigation in future trials with a similar population of DLBCL patients undergoing first-line standard-of-care immunochemotherapy. This cut-off range is similar to values of 300 cm3 and 396 cm3 reported by Cottereau et al.12 and Mikhaeel et al.,9 respectively, but somewhat different to values of 220–261 cm3 reported in other studies.13,15,24 The variation in reported values is likely due to differences in patient characteristics and the methodology used to measure TMTV.15,25,26 The optimal cut-off for TLG was calculated to be 3,004 g, which represented the 49th percentile. This value was within the range of those previously published,9,11-14,24 although there was substantial variation across studies. As for the related parameter of TMTV, differences in pa-

L. Kostakoglu et al. tient characteristics and methodology could explain this variation in reported values.15,25,26 The current analysis involves the largest DLBCL population to date in which the prognostic value of baseline quantitative PET measures has been assessed and confirmed. Our findings are consistent with other retrospective studies in patients with DLBCL, which have shown high TMTV to be a strong predictor of reduced survival.9-15,24,27-30 Furthermore, a meta-analysis reported a highly significant association between high TMTV and poor PFS (pooled HR: 2.93; 95% CI: 2.29–3.73) and OS (pooled HR: 3.52; 95% CI: 2.67–4.64) in patients with DLBCL.31 Mikhaeel et al. demonstrated that the combination of baseline TMTV and response by PET/CT after two immunochemotherapy cycles (rituximab-CHOP) improved the predictive power of interim PET for PFS, allowing the definition of a poor-prognosis group of DLBCL patients, and was more prognostic than either early response or IPI score in this setting.9 For TLG, in agreement with the present analysis, previous retrospective studies have also shown a significant association between high baseline TLG and reduced survival in patients with DLBCL.8,9,24,27,29,32,33 Notably, the cut-offs defined for TMTV and TLG in the current study added to the prognostic power of the dichotomized IPI categories (0–2 and 3–5). The association between high TMTV and poorer PFS was stronger for the

Figure 5. Kaplan–Meier analysis of investigator-assessed progression-free survival according to germinal center B-cell status and a baseline total metabolic tumor volume cut-off of 366 cm3. Differences in progression-free survival (PFS) probability were significant between each subgroup (P<0.0001). GCB: germinal center B-cell-like; TMTV: total metabolic tumor volume. Haematologica | 107 July 2022

1639

ARTICLE - Total metabolic tumor volume predicts survival subgroup of patients with an IPI risk score of 3–5, than for those with a lower score. This has been previously demonstrated for TMTV, which retained an association with poorer PFS9,11,13,29 and poorer OS11-13 after adjustment for IPI score and other baseline parameters. Furthermore, TMTV provided additional stratification for patients with a high IPI score (≥4) into two groups with differing PFS and OS rates at 3 years.24 This finding has lent further credence for including TMTV in management decisions in both low and high IPI groups to offer a more individualized treatment approach. We found that TMTV and TLG were prognostic in both GCB and non-GCB subgroups of DLBCL, suggesting that these quantitative parameters are widely applicable across the different DLBCL subgroups. The Kaplan–Meier analysis with dichotomized TMTV and IPI score or cell-of-origin subtype further stratified risk groups with different outcomes and identified a patient subgroup with the highest risk for treatment failure and unfavorable PFS. Patients with a high TMTV and non-GCB phenotype or high TMTV and a high IPI score (5-year PFS rates of 48.8% and 49.0%, respectively) were at the highest risk of relapse or progression, compared with those who had a low TMTV and a GCB subtype or low TMTV and low IPI score (5-year PFS rates of 80.0% and 74.3%, respectively). These findings confirmed similar reports from other retrospective studies performed in smaller numbers of patients.12,13,24 This integrated risk modeling could lead to better selection of high-risk patients in whom management should be modified before initiation of first line therapy. Baseline SUVmax was not a suitable parameter for predicting survival outcome in the current study. Previous studies that also used ROC analysis to define an SUV cut-off corroborate our finding, indicating that SUV has little prognostic value in patients with DLBCL.9,13,29,30,34 Prognosis appears to be driven by metabolic tumor burden and maximum FDG avidity; although two patients may have widely different tumor burdens, they can have a similar SUVmax for the entire body, as their most active lesions are equally FDG-avid. One more important finding that should be highlighted is the association between TMTV and median time to treatment randomization. This finding emphasizes the recognition and need for avoidance of any potential delays in initiation of treatment, particularly in patients with a high tumor burden. Greater optimization and standardization of quantitative PET parameters and methodology would markedly improve their utilization in clinical practice. Furthermore, automation of TMTV assessment using image processing could enable rapid availability of reproducible results following scans.26,35,36 Other factors requiring further investigation include identification of patient subpopulations who would benefit most from applying PET-based tumor

L. Kostakoglu et al. metrics; combination of PET-based tumor metrics with other predictive or prognostic tools (e.g., gene expression profiling, circulating tumor DNA, end-of-treatment PET or minimal residual disease); the role of TMTV/TLG in personalization of treatment; and the use of more specific PET radiotracers. Currently, assessment of TMTV alone or in combination with other risk factors is not regularly used in clinical practice. Such assessments may become more routine with the growing interest in individualized treatment strategies driven by the increasing availability of targeted treatments. The current analysis is limited by the fact that it was post hoc, rather than a prospectively defined analysis. However, the prospective nature of the trial with standardized PET procedures and the large sample size of PET-CT scans available at baseline provide confidence in the conclusions drawn from the data. Although the treatment arms from GOYA were combined, there is no evidence of a difference in PFS or OS between obinutuzumab-CHOP and rituximab-CHOP after long-term follow up of the study.37 In conclusion, these data from the phase III GOYA study further demonstrate the prognostic value of baseline quantitative PET metrics, as well as their incremental prognostic value in combination with IPI score or cell-oforigin classification, for predicting survival in a previously untreated DLBCL patient population receiving immunochemotherapy. The identification of patients with the worst prognosis prior to initiation of first-line treatment may lead to timely changes, from planned therapy to alternative treatment strategies, tailored to provide greater clinical benefit to high-risk patients. However, integration of PET-derived metrics with molecular and clinical data will likely be necessary to exclude confounding variables and improve their prognostic/predictive value in the future. Disclosures LK reports consultancy fees from F. Hoffmann-La Roche Ltd, Genentech, Inc. and travel expenses from F. HoffmannLa Roche Ltd; FM is an employee of F. Hoffmann-La Roche Ltd; MM reports consulting, advisory board role and speaker’s bureau for F. Hoffmann-La Roche Ltd, Janssen, Novartis, Gilead Sciences and Sandoz; and travel expenses from F. Hoffmann-La Roche Ltd; LS reports research funding from Teva; and consultancy and honoraria fees from F. Hoffmann-La Roche, Genentech, Inc., AbbVie, Amgen, Apobiologix, Acerta, AstraZeneca, Celgene, Debiopharm, Gilead, Incyte, Janssen, Kite Pharma, Karyopharm, Lundbeck, Merck, MorphoSys, Novartis, Seattle Genetics, Sandoz, Takeda, Teva, TG Therapeutics and Verastem; DB reports research funding from and consulting or advisory role for F. Hoffmann-La Roche Ltd, Gilead Sciences, Janssen-Cilag, Seattle Genetics, Morphosys; NC reports research funding

Haematologica | 107 July 2022

1640

ARTICLE - Total metabolic tumor volume predicts survival from F. Hoffmann-La Roche Ltd and honoraria, consultancy and advisory board fees from F. Hoffmann-La Roche Ltd, Merck, Bristol-Meyers Squibb, AstraZeneca, Amgen and Lundbeck; EGB reports consultancy or advisory role for AbbVie, Celgene, EUSA Pharma, Gilead Sciences, Janssen and Koiwa; speakers’ bureau from AbbVie, Janssen and Takeda; and travel expenses from AbbVie, Janssen and F. Hoffmann-La Roche Ltd.; AP reports honoraria from and speaker’s bureau for Bristol-Myers Squibb, F. Hoffmann-La Roche Ltd; Janssen, Celgene, Gilead Sciences and Mundipharma EDO; consultancy or advisory fees from BristolMyers Squibb, F. Hoffmann-La Roche Ltd, Celgene, Gilead Sciences, Mundipharma EDO; patents, royalties, other intellectual property from Mundipharma EDO and travel expenses from F. Hoffmann-La Roche Ltd. CG; XH; YS and YT declare no conflicts of interest; CB is an employee of Genentech, Inc. and has equity ownership in F. Hoffmann-La Roche Ltd; AK is an employee of F. Hoffmann-La Roche Ltd; GS is an employee and stockholder in F. Hoffmann-La Roche Ltd and reports travel expenses from F. HoffmannLa Roche Ltd.; TN and DS are employees and stockholders of F. Hoffmann-La Roche Ltd; UV reports consulting or advisory role for Janssen, Celgene, Juno Therapeutics, Kite Pharma; speaker’s bureau for F. Hoffmann-La Roche Ltd, Janssen, Celgene, Gilead Sciences, Servier, Abbvie; research funding from Celgene; travel expenses from Celgene, F. Hoffmann-La Roche Ltd, AbbVie and advisory board session for Genmab; MT reports honoraria and consultancy fees from Janssen, Gilead Sciences, Bristol-Meyers Squibb, Amgen, AbbVie, Takeda, F. Hoffmann-La Roche Ltd and MorphoSys and honoraria from Incyte.

L. Kostakoglu et al. LS, CG, NC, MT and XH recruited and followed up patients; TN, GS, DS and AK collected data; FM, LK and DS analyzed data; FM, LK, TN, UV, CB, E-GB, AP, DB, AK, YT, YS, LS, CG, NC, MT and XH interpreted data. All authors were involved in the writing of the manuscript, provided their final approval of the manuscript and are accountable for all aspects of the work. Acknowledgments The authors wish to thank the GOYA study team investigators, coordinators, nurses, and patients and are grateful to Will Harris and Joseph Paulson for providing statistical support for the manuscript. Funding The GOYA study was supported by F. Hoffmann-La Roche Ltd, with scientific support from the Fondazione Italiana Linfomi. Editorial support (in the form of writing assistance, collating author comments, assembling tables/figures, grammatical editing, and referencing) under the direction of Lale Kostakoglu was provided by Katie Smith, PhD, and Molly Heitz, PhD, of Ashfield MedComms, an Ashfield Health company, and funded by F. Hoffmann-La Roche Ltd.

Data-sharing statement Qualified researchers may request access to individual patient level data through the clinical study data request platform (https://vivli.org/). Further details on Roche's criteria for eligible studies are available here (https://vivli.org/members/ourmembers/). For further details on Roche's Global Policy on the Sharing of Clinical Information and how to request access to related clinical Contributions study documents, see here (https://www.roche. com/reLK, MM, DS and TN designed the study; LK, MM, TN and DS search_and_development/who_we_are_how_we_work/clini conducted the study; AP, MM, UV, CB, EG-B, AP, DB, YT, YS, cal_trials/our_commitment_to_data_sharing.htm).

References 1. National Comprehensive Cancer Network (NCCN). NCCN clinical practice guidelines in oncology (NCCN Guidelines): B-cell lymphomas. 2020 Version 4.2020. Available from: https://www.nccn.org/professionals/physician_gls/pdf/bcell.pdf. 2. Tilly H, Gomes da Silva M, Vitolo U, et al. Diffuse large B-cell lymphoma (DLBCL): ESMO clinical practice guidelines for diagnosis, treatment and follow-up. Ann Oncol. 2015;26(Suppl 5):S116-125. 3. Alizadeh AA, Eisen MB, Davis RE, et al. Distinct types of diffuse large B-cell lymphoma identified by gene expression profiling. Nature. 2000;403(6769):503-11. 4. Chapuy B, Stewart C, Dunford AJ, et al. Molecular subtypes of diffuse large B cell lymphoma are associated with distinct pathogenic mechanisms and outcomes. Nat Med. 2018;24(5):679-690. 5. Chaganti S, Illidge T, Barrington S, et al. Guidelines for the management of diffuse large B-cell lymphoma. Br J Haematol.

2016;174(1):43-56. 6. Sehn LH, Berry B, Chhanabhai M, et al. The Revised International Prognostic Index (R-IPI) is a better predictor of outcome than the standard IPI for patients with diffuse large B-cell lymphoma treated with R-CHOP. Blood. 2007;109(5):1857-1861. 7. El-Galaly TC, Villa D, Alzahrani M, et al. Outcome prediction by extranodal involvement, IPI, R-IPI, and NCCN-IPI in the PET/CT and rituximab era: a Danish-Canadian study of 443 patients with diffuse-large B-cell lymphoma. Am J Hematol. 2015;90(11):1041-1046. 8. Kim TM, Paeng JC, Chun IK, et al. Total lesion glycolysis in positron emission tomography is a better predictor of outcome than the International Prognostic Index for patients with diffuse large B cell lymphoma. Cancer. 2013;119(6):1195-1202. 9. Mikhaeel NG, Smith D, Dunn JT, et al. Combination of baseline metabolic tumour volume and early response on PET/CT improves progression-free survival prediction in DLBCL. Eur J Nucl Med Mol Imaging. 2016;43(7):1209-1219.

Haematologica | 107 July 2022

1641

ARTICLE - Total metabolic tumor volume predicts survival 10. Song MK, Yang DH, Lee GW, et al. High total metabolic tumor volume in PET/CT predicts worse prognosis in diffuse large B cell lymphoma patients with bone marrow involvement in rituximab era. Leuk Res. 2016;42:1-6. 11. Sasanelli M, Meignan M, Haioun C, et al. Pretherapy metabolic tumour volume is an independent predictor of outcome in patients with diffuse large B-cell lymphoma. Eur J Nucl Med Mol Imaging. 2014;41(11):2017-2022. 12. Cottereau AS, Lanic H, Mareschal S, et al. Molecular profile and FDG-PET/CT total metabolic tumor volume improve risk classification at diagnosis for patients with diffuse large B-cell lymphoma. Clin Cancer Res. 2016;22(15):3801-3809. 13. Toledano MN, Desbordes P, Banjar A, et al. Combination of baseline FDG PET/CT total metabolic tumour volume and gene expression profile have a robust predictive value in patients with diffuse large B-cell lymphoma. Eur J Nucl Med Mol Imaging. 2018;45(5):680-688. 14. Chang CC, Cho SF, Chuang YW, et al. Prognostic significance of total metabolic tumor volume on (18)F-fluorodeoxyglucose positron emission tomography/ computed tomography in patients with diffuse large B-cell lymphoma receiving rituximab-containing chemotherapy. Oncotarget. 2017;8(59):99587-99600. 15. Vercellino L, Cottereau A-S, Casasnovas O, et al. High total metabolic tumor volume at baseline predicts survival independent of response to therapy. Blood. 2020;135(16): 1396-1405. 16. Mamot C, Klingbiel D, Hitz F, et al. Final results of a prospective evaluation of the predictive value of interim positron emission tomography in patients with diffuse large B-cell lymphoma treated with R-CHOP-14 (SAKK 38/07). J Clin Oncol. 2015;33(23):2523-2529. 17. Pregno P, Chiappella A, Bellò M, et al. Interim 18-FDG-PET/CT failed to predict the outcome in diffuse large B-cell lymphoma patients treated at the diagnosis with rituximab-CHOP. Blood. 2012;119(9):2066-2073. 18. Bishton MJ, Hughes S, Richardson F, et al. Delineating outcomes of patients with diffuse large b cell lymphoma using the national comprehensive cancer network-international prognostic index and positron emission tomography-defined remission status; a population-based analysis. Br J Haematol. 2016;172(2):246-254. 19. Kanemasa Y, Shimoyama T, Sasaki Y, et al. Analysis of prognostic value of complete response by PET–CT and further stratification by clinical and biological markers in DLBCL patients. Med Oncol. 2017;34(2):29. 20. Kostakoglu L, Martelli M, Sehn LH, et al. End-of-treatment PET/CT predicts PFS and overall survival in DLBCL after firstline treatment: results from GOYA. Blood Adv. 2021;5(5):1283-1290. 21. Vitolo U, Trněný M, Belada D, et al. Obinutuzumab or rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisone in previously untreated diffuse large B-cell lymphoma. J Clin Oncol. 2017;35(31):3529-3537. 22. Marcus R, Davies A, Ando K, et al. Obinutuzumab for the firstline treatment of follicular lymphoma. N Engl J Med. 2017;377(14):1331-1344. 23. Goede V, Fischer K, Busch R, et al. Obinutuzumab plus chlorambucil in patients with CLL and coexisting conditions. N Engl J Med. 2014;370(12):1101-1110.

L. Kostakoglu et al. 24. Shagera QA, Cheon GJ, Koh Y, et al. Prognostic value of metabolic tumour volume on baseline (18)F-FDG PET/CT in addition to NCCN-IPI in patients with diffuse large B-cell lymphoma: further stratification of the group with a high-risk NCCN-IPI. Eur J Nucl Med Mol Imaging. 2019;46(7):1417-1427. 25. Gormsen LC, Vendelbo MH, Pedersen MA, et al. A comparative study of standardized quantitative and visual assessment for predicting tumor volume and outcome in newly diagnosed diffuse large B-cell lymphoma staged with 18F-FDG PET/CT. Eur J Nucl Med Mol Imaging Research. 2019;9(1):36. 26. Ilyas H, Mikhaeel NG, Dunn JT, et al. Defining the optimal method for measuring baseline metabolic tumour volume in diffuse large B cell lymphoma. Eur J Nucl Med Mol Imaging. 2018;45(7):1142-1154. 27. Xie M, Zhai W, Cheng S, Zhang H, Xie Y, He W. Predictive value of F-18 FDG PET/CT quantization parameters for progression-free survival in patients with diffuse large B-cell lymphoma. Hematology. 2016;21(2):99-105. 28. Xie M, Wu K, Liu Y, Jiang Q, Xie Y. Predictive value of F-18 FDG PET/CT quantization parameters in diffuse large B cell lymphoma: a meta-analysis with 702 participants. Med Oncol. 2015;32(1):446. 29 Ceriani L, Milan L, Martelli M, et al. Metabolic heterogeneity on baseline 18FDG-PET/CT scan is a predictor of outcome in primary mediastinal B-cell lymphoma. Blood. 2018;132(2):179-186. 30. Delaby G, Hubaut MA, Morschhauser F, et al. Prognostic value of the metabolic bulk volume in patients with diffuse large B-cell lymphoma on baseline (18)F-FDG PET-CT. Leuk Lymphoma. 2020;61(7):1584-1591. 31. Guo B, Tan X, Ke Q, Cen H. Prognostic value of baseline metabolic tumor volume and total lesion glycolysis in patients with lymphoma: a meta-analysis. PLoS One. 2019;14(1):e0210224. 32. Esfahani SA, Heidari P, Halpern EF, Hochberg EP, Palmer EL, Mahmood U. Baseline total lesion glycolysis measured with (18)F-FDG PET/CT as a predictor of progression-free survival in diffuse large B-cell lymphoma: a pilot study. Am J Nucl Med Mol Imaging. 2013;3(3):272-281. 33. Zhou M, Chen Y, Huang H, Zhou X, Liu J, Huang G. Prognostic value of total lesion glycolysis of baseline 18Ffluorodeoxyglucose positron emission tomography/computed tomography in diffuse large B-cell lymphoma. Oncotarget. 2016;7(50):83544-83553. 34. Zhang YY, Song L, Zhao MX, Hu K. A better prediction of progression-free survival in diffuse large B-cell lymphoma by a prognostic model consisting of baseline TLG and %DSUVmax. Cancer Med. 2019;8(11):5137-5147. 35. Jemaa S, Fredriskson J, Coimbra Aea. A fully automated measurement of total metabolic tumor burden in diffuse large B-cell lymphoma and follicular lymphoma. Blood. 2019;134(Suppl 1):S4666. 36. Burggraaff CN, Rahman F, Kaßner I, et al. Optimizing workflows for fast and reliable metabolic tumor volume measurements in diffuse large C-cell lymphoma. Mol Imaging Biol. 2020;22(4):1102-1110. 37. Sehn LH, Martelli M, Trněný M, et al. Final analysis of GOYA: a randomized, open-label, Phase III study of obinutuzumab or rituximab plus CHOP in patients with previously untreated diffuse large B-cell lymphoma. Blood. 2019;134(Suppl 1):S4088.

Haematologica | 107 July 2022

1642

ARTICLE - Platelet Biology and its Disorders

Platelet dysfunction in platelet-type von Willebrand disease due to the constitutive triggering of the Lyn-PECAM1 inhibitory pathway Loredana Bury, Emanuela Falcinelli, Anna Maria Mezzasoma, Giuseppe Guglielmini, Stefania Momi and Paolo Gresele Department of Medicine and Surgery, Section of Internal and Cardiovascular Medicine, University of Perugia, Perugia, Italy

Correspondence: Loredana Bury loredana.bury@gmail.com Received: March 18, 2021. Accepted: July 29, 2021. Prepublished: August 19, 2021. https://doi.org/10.3324/haematol.2021.278776 ©2022 Ferrata Storti Foundation Haematologica material is published under a CC BY-NC license

Abstract Platelet-type von Willebrand disease (PT-VWD) is an inherited platelet disorder. It is characterized by macrothrombocytopenia and mucocutaneous bleeding, of variable severity, due to gain-of-function variants of GP1BA conferring to glycoprotein Ibα (GPIbα) enhanced affinity for von Willebrand factor (VWF). The bleeding tendency is conventionally attributed to thrombocytopenia and large VWF-multimer depletion. However, while some indications suggest that platelet dysfunction may contribute to the bleeding phenotype, no information on its characteristics and causes are available. The aim of the present study was to characterize platelet dysfunction in PT-VWD and shed light on its mechanism. Platelets from a PT-VWD patient carrying the p.M239V variant, and from PT-VWD mice carrying the p.G233V variant, showed a remarkable platelet function defect, with impaired aggregation, defective granule secretion and reduced adhesion under static and flow conditions. VWFbinding to GPIbα is known to trigger intracellular signaling involving Src-family kinases (SFK). We found that constitutive phosphorylation of the platelet SFK Lyn induces a negative-feedback loop downregulating platelet activation through phosphorylation of PECAM1 on Tyr686 and that this is triggered by the constitutive binding of VWF to GPIbα. These data show, for the first time, that the abnormal triggering of inhibitory signals mediated by Lyn and PECAM1 may lead to platelet dysfunction. In conclusion, our study unravels the mechanism of platelet dysfunction in PT-VWD caused by deranged inhibitory signaling. This is triggered by the constitutive binding of VWF to GPIbα which may significantly contribute to the bleeding phenotype of these patients.

Introduction Gain-of-function variants in the GP1BA gene conferring to GPIbα enhanced affinity for VWF cause platelet-type von Willebrand disease (PT-VWD), a rare inherited bleeding disorder.1,2 Patients with PT-VWD have mild thrombocytopenia with increased platelet volume, enhanced ristocetin-induced platelet agglutination (RIPA) and prolonged bleeding time associated with mucocutaneous bleeding, which can be severe.3 The cause of the apparently counterintuitive bleeding phenotype of patients with platelets displaying enhanced affinity for VWF is incompletely understood and is thought to be due, principally, to thrombocytopenia and to the consumption of large VWF-multimers bound to platelets.1 The possibility that platelet dysfunction may contribute

to the bleeding phenotype of PT-VWD has attracted little attention, but defective fibrinogen binding, delayed aggregation in response to ADP and thrombin, impaired thrombus formation on a damaged carotid artery and unstable clot formation have been reported in a mouse model of PT-VWD,4,5 raising the hypothesis that platelet function may be altered in PT-VWD patients. However, the mechanism of the platelet function defect is unknown and no functional studies have been performed with human PT-VWD platelets. The first step in primary hemostasis is the interaction between platelets and subendothelial collagen, mediated by the binding of von Willebrand factor (VWF) to the exposed collagen and to the GPIbα subunit of the platelet GPIb-IX-V complex. This slows down platelets onto the damaged vessel wall thus allowing the direct interaction

Haematologica | 107 July 2022

1643

ARTICLE - Impaired platelet function in PT-VWD

L. Bury et al.

of collagen with its receptors on platelets. However, GPIbIX-V engagement by VWF also triggers intraplatelet signaling with the sequential activation of the Src-family kinases (SFK) Src and Lyn. These are associated with the cytoplasmic tail of GPIbα, which triggers a cytoplasmic Ca2+ increase and PI3-kinase/PKC activation, ultimately leading to inside-out αIIbβ3 activation.6-8 Here, for the first time, we show defective platelet function in a PT-VWD patient carrying the p.M239V variant and in PT-VWD mice carrying the p.G233V variant, due to impaired Rap1b and αIIbβ3 activation. Mechanistic studies revealed a constitutive phosphorylation of Lyn inducing a negative-feedback loop, downregulating platelet activation through the phosphorylation of PECAM1 on Tyr686. Our results, showing that constitutive binding of VWF to GPIbα triggers PECAM1-mediated inhibitory signaling downregulating platelet activation, suggest that defective platelet function in PT-VWD may significantly contribute to the bleeding phenotype associated with this disorder.

Methods

Light transmission aggregometry (LTA) LTA was carried out using platelet-rich plasma (PRP) employing a range of agonists, as described previously.13 For the study of shape change, PRP was pretreated with the Arg-Gly-Asp-Ser peptide (RGDS) αIIbβ3 blocker.14,15 For details see the Online Supplementary Data. αIIbβ3 activation αIIbβ3 activation was assessed using the PAC-1 MoAb (BD Bioscience, Franklin Lakes, NJ, USA) for human platelets and the JON/A MoAb (Emfret Analytics, Eibelstadt, Germany) for murine platelets, as described previously.13,17 For details see the Online Supplementary Data. Granule content and secretion Platelet granule content was assessed by electron microscopy, as described previously.10,13 Alpha-granule secretion was measured as P-selectin expression18-20 and dense granule content and release using the green fluorescent dye mepacrine21 by flow cytometry, as described previously.19 For details see the Online Supplementary Data.

The control and PT-VWD terms in this manuscript refer to human platelets while the respective terms used for murine platelets are TgWT and TgG233V. All human and animal studies were approved by the competent institutional review boards (human: CEAS Umbria, approval n. 2663/15; animal: Italian Ministry of Public Health, authorization n. 561/2015-PR).

Spreading assay We resuspended washed platelets22,23 (20x106/mL) in Tyrode’s buffer and spreading on fibrinogen and type I collagen was assessed, as described previously.24 For details see the Online Supplementary Data.

Mouse strains and blood sampling The generation of PT-VWD mice expressing human GP1BA carrying the p.G233V variant (TgG233V), and of control mice expressing a wild-type human GP1BA (TgWT), has been previously described (Online Supplementary Figure S1).4,5,10,12 Mice were bred and housed in the animal facility at the University of Perugia, Italy, and all experiments were performed using 3- to 6- month-old mice. The number of males and females, as well as platelet counts, are reported under each figure legend. For details see the Online Supplementary Data.

Rap-1b pull-down assay Rap-1b activity (Rap-1b-GTP) was assessed using an active Rap-1b pull-down and detection kit (Pierce Biotechnology, Rockford, IL, USA) in washed platelets stimulated for 30 seconds with ADP 10 mM, thrombin 0.1 U/mL or their vehicle.27 For details see the Online Supplementary Data.

Platelet adhesion under flow conditions Platelet adhesion under flow conditions over fibrillar type Human studies and blood sampling I collagen was assessed using citrated human or murine The PT-VWD patient carrying the M239V GP1BA variant (to blood, as described previously.20,25 For details see the Onour knowledge the only patient so far diagnosed in Italy) line Supplementary Data. has been previously reported;9,10 she had an ISTH-BAT score of 11.11 Each experiment was repeated three to five Measurement of intracellular calcium (Ca2+) times from blood samples collected on different days (in PRP was loaded with the Ca2+-sensitive dye FLUO 3-acethe text indicated as independent experiments, eight in toxymethyl ester (FLUO 3-AM; Molecular Probes) and Ca2+ total, Online Supplementary Table S1). Age-and sex- mobilization induced by various agonists was assessed by matched healthy controls were studied in parallel. For de- flow cytometry. as described previously.17,26 For details see tails see the Online Supplementary Data. the Online Supplementary Data.

Phosphorylation of signaling proteins Human washed platelets (300x109/L) were stimulated for 30 seconds under continuous stirring with ADP 2 mM, CVX 60 ng/mL, ristocetin 0.3 mg/mL or their vehicle. To exclude

Haematologica | 107 July 2022

1644

ARTICLE - Impaired platelet function in PT-VWD

L. Bury et al.

αIIbβ3-mediated outside-in signaling as a cause of signaling protein phosphorylation, platelets were treated with the αIIbβ3 inhibitor tirofiban 0.4 mM for ten minutes before stimulation. Phosphorylation of Src (Tyr416), Lyn (Tyr397 and Tyr507), PLC β3 (Ser537), Akt (Ser473), Syk (Tyr525/526) PKC substrate, PECAM1 (Tyr686 and Ser702),28 was analyzed by western blotting, as described previously.10 For details see the Online Supplementary Data.

second wave in response to ADP and epinephrine (Figure 1A, Online Supplementary Figure S2). Given that the analysis of LTA curves suggested a shape change defect, LTA was repeated after pre-incubation with RGDS, which prevents aggregation. This allowed a more thorough observation of the initial phase of the platelet activation response to stimuli; indeed shape change turned out to be significantly reduced in response to all agonists (Figure 1B).

cAMP and cGMP production cAMP production in response to Iloprost (100 ng/mL) and cGMP production in response to the nitric oxide (NO)donor s-nitroso-n-acetyl penicillamine (SNAP, 0.1-1-10 mM) was assessed using the dual range cAMP Enzymeimmunoassay Biotrak (EIA) System and the dual range cGMP Enzymeimmunoassay Biotrak (EIA) System (Amersham, GE Healthcare, Milan, Italy).

Platelet αIIbβ3 activation and granule secretion are defective in PT-VWD PAC-1 binding to human PT-VWD platelets (Figure 2A, Online Supplementary Figure S3A) and JON/A binding to TgG233V mouse platelets (Figure 2B) were impaired in response to all tested stimuli. PAC-1 and JON/A binding in resting conditions did not differ between control and PT-VWD platelets and between TgWT and TgG233V platelets, suggesting that the mildly increased volume of PT-VWD and TgG233V platelets did not influence flow cytometry results. Results The defect of second wave aggregation suggested a granPlatelet aggregation and shape change are defective ule secretion defect. Granule content, evaluated by transin PT-VWD mission electron microscopy, was normal in human The aggregation of human PT-VWD platelets was strongly PT-VWD platelets compared to controls (α-granules: impaired in response to TRAP-6, defective in response to 5.8±0.4 vs. 6.4±1.3/platelet; dense granules: 0.7±0.4 vs. collagen, convulxin and arachidonic acid and lacking the 0.9±0.3/platelet, P=ns) (Figure 2C), while secretion of both

Figure 1. Platelet aggregation and shape change are defective in PT-VWD. (A) Human platelet aggregation in response to TRAP6 20 mM (i), ADP 10 mM (ii) and collagen 1 mg/mL (iii) in platelet-rich plasma from the PT-VWD patient and a parallel healthy control. Traces are representative of four independent experiments (Online Supplementary Table S1, samples A, C, D, E). *=P<0.05 vs. control, unpaired Student’s t test. (B) Platelet shape change assessed by LTA after pretreatment with RGDS (120 mg/mL) in response to TRAP-6 2 mM (i), ADP 10 mM (ii) and collagen 1 mg/mL (iii) in platelet-rich plasma from the PT-VWD patient and a parallel healthy control. Tracings are representative of four independent experiments (Online Supplementary Table S1, samples A, C, D, E) *P<0.05 vs. control, unpaired Student’s t test. Haematologica | 107 July 2022

1645

ARTICLE - Impaired platelet function in PT-VWD

L. Bury et al.

Figure 2. αIIbβ3 activation and granule secretion are defective in PT-VWD. (A) Integrin αIIbβ3 activation (PAC-1 binding) as assessed by flow cytometry in response to TRAP-6 20 mM (i), ADP 10 mM (ii) and CVX 20 ng/mL (iii) in human whole blood from the PT-VWD patient and a parallel healthy control. PAC-1 binding is reported as percentage of positive platelets, calculated as the percentage of platelets (gated for their forward scatter and side scatter values and for their positivity to the platelet marker CD42b) that bound PAC-1 over the total platelet population after setting for aspecific binding. Data are means ± SEM from at least five independent experiments (Online Supplementary Table S1, samples A, B, C, D, E). **P<0.01 vs. control, Two-way ANOVA. (B) Integrin αIIbβ3 activation (JON-A binding) as assessed by flow cytometry in response to thrombin 0.05 U/mL + GPRP 2 mM and convulxin 25 ng/mL in murine whole blood from TgG233V and TgWT mice. Jon-A binding is reported as percentage of positive platelets, calculated as the percentage of platelets (gated for their forward scatter and side scatter values and for their positivity to the platelet marker CD42b) that bound Jon-A over the total platelet population, after setting for aspecific binding. Data are means ± SEM, n=5 (TgWT: three females and two males, mean platelet count: 627850±21566/mL; TgG233V: three females and two males; mean platelet count: 330600±17677/mL). **P<0.01 vs. TgWT, two-way ANOVA. (C) Representative images of the ultrastructure of human control and PT-VWD platelets used to assess granular content (original magnification 13000X). Specimens were observed with a Philips Electron Optics EM208 transmission electron microscope at 80 kV at room temperature. Granules were counted from at least 50 platelets (α-granules: 5.8±0.4 vs. 6.4±1.3/platelet; dense granules: 0.7±0.4 vs. 0.9±0.3/platelet, P=ns). Black arrowheads indicate examples of α-granules; white arrowheads indicate examples of dense granules. Dense granules were defined by their highly electron-dense core surrounded by a clear space enclosed by a single Haematologica | 107 July 2022

1646

Continued on following page.

ARTICLE - Impaired platelet function in PT-VWD

L. Bury et al.

membrane. D-E) α-granule secretion in human (D) and murine (E) platelets as assessed by the measurement of P-selectin expression by flow cytometry. Agonists were ADP 10 mM, TRAP6 10 mM or CVX 20 ng/mL for human platelets and thrombin 0.05 U/mL + GPRP 2 mM or CVX 25 ng/mL for murine platelets. P-selectin expression is reported as a percentage of positive platelets, calculated as the percentage of platelets (gated for their forward scatter and side scatter values and for their positivity to the platelet marker CD41) that bound CD62P over the total platelet population, after setting for aspecific binding. Data are means ± SEM from at least five independent experiments (Online Supplementary Table S1, samples A, B, C, D, E) or from n=5 mice (TgWT: three females and two males, mean platelet count: 627850±21566/mL; TgG233V: three females and two males; mean platelet count: 330600±17677/mL). **P<0.01; two-way ANOVA. (F,G) Dense granule secretion in human (F) and murine (G) platelets as assessed by the measurement of mepacrine release by flow cytometry. Mepacrine release was calculated by the following formula: 1-(residual mepacrine content after thrombin stimulation with 0.05 U/mL/ total mepacrine incorporated)x100. “Total mepacrine incorporated” is the percentage of platelets (gated for their forward scatter and side scatter values and for their positivity to the platelet markers CD41 or CD42b) that incorporated mepacrine over the total platelet population, after setting for aspecific binding; “residual mepacrine content” is the percentage of platelets stimulated with 0.05 U/mL of thrombin that incorporated mepacrine over the total platelet population, after setting for aspecific binding. Data are means ± SEM from at least 5 independent experiments (Online Supplementary Table S1, samples A, B, C, D, E) or from n=5 mice (TgWT: three females and two males, mean platelet count: 627850±21566/mL; TgG233V: three females and two males; mean platelet count: 330600±17677/mL). *P<0.05, **P<0.01; unpaired Student’s t test.

α- and dense granules was defective with both human PTVWD and murine TgG233V platelets, as shown by impaired agonist-induced platelet surface P-selectin expression (Figure 2D and 2E, Online Supplementary Figure S3B) and mepacrine release (Figure 2F, G). P-selectin expression (Figure 2C) and mepacrine incorporation did not differ between control and PT-VWD resting platelets (mepacrine incorporation: 85.9±9 % vs. 94.7±4%, P=ns) and between TgWT and TgG233V platelets (mepacrine incorporation: 76.5±14.2 % vs. 84.3±11.3%, P=ns), here too suggesting that the mildly increased volume of PT-VWD and TgG233V platelets did not affect flow cytometry results. Platelet surface expression of αIIbβ3 and GPIb/IX/V receptors was comparable between PT-VWD and control platelets (data not shown). Spreading and adhesion under flow conditions are defective in PT-VWD Platelet spreading on fibrinogen and type I collagen was defective both with human PT-VWD (Figure 3A, B) and murine TgG233V platelets (Figure 3C), compared with controls. Also, total surface area covered by platelets upon perfusion of blood over a collagen-coated surface was significantly reduced as compared with controls, both at high (αIIbβ3-dependent) and low (αIIbβ3-independent) shear rates (Figure 3D, E).29 Given that granule secretion, shape change, platelet spreading on type I collagen and adhesion at low shear rate do not depend on αIIbβ3 activation, a global platelet function defect in PT-VWD was suspected, therefore signal transduction mechanisms were explored. Platelet Ca2+ mobilization and Rap-1b activation are impaired in PT-VWD We first studied cytoplasmic free calcium movements which are involved in shape change, adhesion, aggregation and the release of platelet granules,30 all processes defective in PT-VWD. Indeed, ADP-, TRAP-6- and CVX-induced calcium mobilization was impaired in human

PT-VWD platelets (Figure 4A). The activation of the small GTPase Rap1b, which is Ca2+-mediated31 and is required for αIIbβ3 activation, was also impaired in response to ADP and thrombin in human PT-VWD (Figure 4B) and TgG233V (Online Supplementary Figure S4A) platelets. The phosphorylation of PLCβ3, activator of Rap1b,32 was comparable between PT-VWD and control platelets upon activation with ADP and thrombin (Online Supplementary Figure S4B). SFK are crucial steps in the pathways leading to the cytoplasmic Ca2+ increase, and reciprocally, Ca2+ rises modulate SFK activation and activate Rap1b.32,33 Another key pathway regulating Rap1b is triggered by PKC activation34 thus we explored SFK and PKC activation. Src-family kinases, Lyn and PECAM1 are constitutively activated in PT-VWD platelets SFK are phosphorylated upon agonist binding to a wide repertoire of platelet surface receptors and thus play a central role in transducing activatory signals;35 we therefore assessed their phosphorylation. Phosphorylation of SFK was markedly enhanced in resting PT-VWD and TgG233V platelets compared to controls (Figure 5A, Online Supplementary Figure S5A). Interestingly, control and TgWT platelets pre-incubated with ristocetin, which triggers the binding of VWF to GPIbα, showed enhanced SFK phosphorylation (Figure 5A, Online Supplementary Figure S5A). We focused on Lyn because it is the member of SFK which, besides playing a role in platelet activatory signaling, also triggers a signaling that dampens platelet activation by phosphorylating immunotyrosine-based inhibitory motif (ITIM)–containing receptors.35 Lyn was phosphorylated at Tyr397, which means activated, in resting PT-VWD and TgG233V, but not in control platelets (Figure 5B, Online Supplementary Figure S5B). Interestingly, here too ristocetin triggered the phosphorylation of Lyn at Tyr397 of control and TgWT platelets. Accordingly, the negative regulatory site of Lyn - Tyr507 - was less phosphorylated in PT-VWD and TgG233V platelets than in control and TgWT platelets, while the incubation with ristocetin triggered

Haematologica | 107 July 2022

1647

ARTICLE - Impaired platelet function in PT-VWD

L. Bury et al.

the de-phosphorylation at Tyr507 of control platelets (Figure 5C, Online Supplementary Figure S5C). Moreover, the ITIM-containing platelet receptor PECAM1, a substrate of Lyn,36 was phosphorylated at Tyr686 and Ser702 (Ser702 is phosphorylated after residue 686), thus activated, in resting PT-VWD and TgG233V but not in control platelets. Here too, ristocetin triggered the PECAM1 phosphorylation of control platelets and TgWT platelets (Figure 5D, 5E, Online Supplementary Figure S5D, E). Upon platelet stimulation, SFK and Lyn (Tyr397) phosphorylation was only mildly increased in PT-VWD platelets while it was markedly increased in control platelets (Online Supplementary Figure S6A, B). On the contrary, Lyn phosphorylation at Tyr507 was significantly increased in activated PT-VWD platelets compared to control platelets (Online Supplementary Figure S6C). Finally, upon platelet stimulation, PECAM1 phosphorylation of PT-VWD platelets was comparable to control platelets (Online Supplementary Figure S6D, E). We then focused on the activation of the main Lyn substrates, Akt and Syk: Akt phosphorylation was comparable in resting PT-VWD and control platelets while it was diminished in PT-VWD platelets upon stimulation with ADP and CVX (Online Supplementary Figure S7); Syk phosphorylation was comparable both in resting and in stimulated PT-VWD and control platelets (Online Supplementary Figure S7). On the other hand, PKC substrate phosphorylation was increased in resting PT-VWD platelets. Here too, ristocetin triggered the phosphorylation of PKC substrates of control platelets (Figure 5F, Online Supplementary Figure S5F). Upon platelet stimulation, PKC substrate phosphorylation was only mildly increased in PT-VWD platelets while it was increased twofold in control platelets, with a significant difference between control and PT-VWD platelets (Online Supplementary Figure S6F). cAMP and cGMP production were comparable between control and PT-VWD platelets (Online Supplementary Figure S8).

Discussion The bleeding diathesis of PT-VWD patients has been conventionally attributed to the clearance of platelet-VWF complexes from the circulation.1 Recently, we showed that defective proplatelet formation and ectopic platelet release by megakaryocytes in the bone marrow contribute to thrombocytopenia in PT-VWD.10 However, the mild to moderate thrombocytopenia of PT-VWD does not seem sufficient to explain a bleeding phenotype that may sometimes be severe, especially after surgery and delivery.1,37-41 Previous studies in TgG233V mice showed defective fibrinogen binding,4 decreased pro-coagulant activity, delayed

aggregation in response to ADP and thrombin and impaired in vivo thrombus formation,5 suggesting platelet dysfunction in these mice. Similarly, studies on platelets from a patient with type-2B von Willebrand disease (2B-VWD), which is the VWF counterpart of PT-VWD, and from 2B-VWD mice, both carrying the p.V1316M VWF variant, showed that the constitutive binding of mutant VWF to platelets causes impaired platelet function with reduced αIIbβ3 activation, contributing to the bleeding tendency of 2B-VWD.42,43 However, no studies on platelet function in PT-VWD patients have ever been performed. Here we show that a remarkable platelet function defect is present in PT-VWD patients and we discovered that it is caused by the constitutive triggering of platelet inhibitory signal transduction negatively regulating platelet activation through the hyper-activation of Lyn and PECAM1. Our study focused on two different well known GP1BA gain-of-function variants: p.M239V, carried by a PT-VWD patient, and p.G233V, expressed in PT-VWD mice. PT-VWD platelets showed defective αIIbβ3 activation and, consequently, impaired platelet aggregation, spreading on fibrinogen and adhesion to collagen at high shear rates, all processes dependent on αIIbβ3 activation. Moreover, PTVWD platelets showed defective α- and d-granule secretion, shape change and adhesion to collagen at low shear rate, which are all independent of αIIbβ3 activation but strictly dependent on Ca2+ mobilization.1649 Indeed, we found that Ca2+ mobilization induced by various agonists is defective in PT-VWD platelets. Platelet function was impaired in response to agonists stimulating both G-protein-coupled and ITAM-coupled receptors. In platelets, the generation of the Ca2+ and DAG second messengers leads to the activation of PKC. This activates Rap1b31,34 which, in turn, activates αIIbβ3 through the recruitment of Talin1.44 Therefore, we assessed Rap-1b activation and we found it impaired not due to a dysfunction of its main activator PLCβ3, but to a defective PKC pathway which contributes to its activation.34,45 This finding, together with our finding that ristocetin triggers PKC substrate phosphorylation in control platelets, suggests that enhanced VWF-GPIbα interaction leads to the activation and subsequent exhaustion of the PKC pathway, as recently described in 2B-VWD platelets.43 This, in turn, reduces Rap-1b activation, causing αIIbβ3 dysfunction. The binding of VWF to GPIbα activates PKC through a signaling cascade that involves Lyn,46 a SFK that we found to be activated in resting PT-VWD platelets. Thus, it can be hypothesized that the cause of baseline PKC upregulation, and the relative refractoriness of this pathway to stimulation, is Lyn activation. In addition, the activation of Lyn negatively regulates Ca2+ flux, and, consequently, platelet activation, through the activation of phosphatases.47 Taken together, our data show a global defect of platelet

Haematologica | 107 July 2022

1648

ARTICLE - Impaired platelet function in PT-VWD

L. Bury et al.

Figure 3. Platelet spreading and adhesion under flow conditions are defective in PT-VWD. (A) Representative images of human control and patient (PT-VWD) platelets 30 minutes after layering onto fibrinogen or collagen. Scale bar is 10 mm. Platelets were stained with FITC-conjugated phalloidin and analyzed using a Carl Zeiss Axio Observer.A1 fluorescence microscope (Carl Zeiss Inc., Oberkochen, Germany) with a 100x/1.4 Plan-Apochromat oil-immersion objective at room temperature. (B, C) Human (B) and murine (C) platelet spreading on glass coverslips coated with fibrinogen (100 mg/mL) or type I collagen (25 mg/mL). Spreading is expressed as percentage of surface covered by platelets. Data are means ± SEM from five independent experiments (Online Supplementary Table S1, samples A, B, C, D, E) or from n=5 mice (TgWT: two females and three males, mean platelet count: 661600±99984/mL; TgG233V: two females and three males; mean platelet count: 305000±26728/mL). **P<0.01; two-way ANOVA. The number of platelets that adhered to the surface was comparable between control and PT-VWD platelets (number of platelets adhering to fibrinogen per microscopic field: controls 93±27, PT-VWD 87±18, P=ns; number of platelets adhering to collagen per microscopic field: controls 78±17, PT-VWD 83±12, P=ns) and between TgWT and TgG233V platelets (number of platelets adhering to fibrinogen per microscopic field: TgWT 101±21, TgG233V 105±24, P=ns; number of platelets adhering to collagen per microscopic field: TgWT 98±9, TgG233V 91±13, P=ns). (D-E) Human (D) and murine (E) platelet adhesion to type I collagen under flow conditions. Whole blood was perfused at different shear rates in glass microcapillary tubes coated with type I collagen (30 mg/cm2). Specimens were observed under an optical microscope after fixation with 0.25% glutaric-dialdehyde and MayGrünwald/Giemsa staining. The total surface covered by platelets was calculated with ImageJ software. Data are means ± SEM of three independent experiments (Online Supplementary Table S1, samples B, D, F) (D) or from n=3 mice (TgWT: three males, mean platelet count: 631550±21142/mL; TgG233V: three males; mean platelet count: 350800±94469/mL) (E). *P<0.05, **P<0.01; twoway ANOVA.

function in PT-VWD suggestive of the triggering of a negative feedback regulatory system. The binding of VWF to the extracellular region of GPIbα induces the association of Src and Lyn with the cytoplasmic tail of GPIbα, starting the downstream phosphorylation of a number of substrates such as PI3K, Akt, p38, Syk and PLC, involved in inside-out signaling leading to Rap-1b, Talin-1 and, thus, αIIbβ3 activation.6,35,48 However, in activated platelets, Lyn also initiates a negative feedback regulatory pathway by phosphorylating PECAM1, providing docking sites for SH2 domains of phosphatases such as the SHP1/SHP2 tyrosine phosphatases and the SHIP1/SHIP2 inositol phosphatase that mediate the termination of platelet activation signals.36,49-54

In our study, we show that Lyn and PECAM1 are activated in resting PT-VWD platelets identifying, to the best of our knowledge, the first platelet function defect due to the hyper-activation of Lyn and PECAM1 in humans. cAMP and cGMP production were comparable in control and PT-VWD platelets, indicating that these two negative platelet regulatory pathways are not involved in the platelet dysfunction of PT-VWD. Moreover, we show that Lyn and PECAM1 are activated in control platelets stimulated with ristocetin, confirming that the binding of VWF to GPIbα phosphorylates Lyn and PECAM1.35,54 In support of our findings is the observation that platelets from PECAM1-knock-out, Lyn-knock-out, or PECAM1/Lyn

Haematologica | 107 July 2022

1649

ARTICLE - Impaired platelet function in PT-VWD

L. Bury et al. Figure 4. Platelet Ca2+ mobilization and Rap-1b activation are defective in PT-VWD. (A) Changes in cytosolic free Ca2+ triggered by TRAP-6 10 mM, ADP 5 mM and CVX 50 ng/mL, as assessed in FLUO 3-AM (8 mM)-labeled platelets. After the measurement of baseline fluorescence of PRP samples, agonists were added and changes in green were recorded in fluorescence function of time. The graphs show data from one representative experiment. Points represent the ratio with respect to baseline mean fluorescence intensity (MFI). The columns show maximal Ca2+ rise after stimulation (fluorescence ratio) with different agonists. Data are means ± SEM from four independent experiments (Online Supplementary Table S1, samples D, E, F, H) (P<0.05; unpaired t-test). (B) Rap1b activation (Rap1b-GTP) of human control and patient (PT-VWD) platelets after stimulation with thrombin (0.1 U/mL) or ADP (10 mM) under stirring conditions was assessed by western blotting by loading 500 mg of proteins. Densitometric analysis was performed using Image J software. Quantification of Rap1b-GTP is relative to total RAP1b expression and is expressed in arbitrary units (AU). Data are means ± SEM from three independent experiments (Online Supplementary Table S1, samples B, D, F) (*P<0.05; two-way ANOVA).

double-knock-out mice are hyper-responsive to stimuli, confirming that PECAM1 and Lyn suppress platelet reactivity.36,51,53,54 To assess whether the blockade of Lyn could restore normal platelet reactivity, we measured granule secretion and αIIbβ3 activation by control and PT-VWD platelets after incubation with the Lyn inhibitor, bafetinib. However, bafetinib, suppressing the platelet activating function of Lyn, inhibited αIIbβ3 activation and granule secretion35 (Online Supplementary Figure S9). Conversely, human and murine platelets in which PECAM1 activation was triggered by a PECAM1 homophilic ligand50 or by cross-linking antibodies51,52 showed decreased aggregation,50,51 secretion,51,52 Ca2+ mobilization51,52 and fibrinogen binding,52 all functions that we found to be impaired in PTVWD platelets. Indeed, PECAM1 has recently been identified as a potential novel target for antiplatelet therapy.55

Interestingly, PECAM1-knockout mice also show reduced trabecular bone volume, an increased number of osteoclasts and enhanced bone resorption56 while PT-VWD mice show enhanced trabecular bone volume, a decreased number of osteoclasts and decreased bone resorption, findings so far unexplained and possibly related to the role of PECAM1 as a negative regulator of osteoclastogenesis.12 When we assessed the activation of Lyn substrates, namely Akt and Syk, we found decreased Akt phosphorylation after stimulation with ADP and convulxin in PT-VWD platelets. The same Akt dysfunction was previously reported in 2BVWD platelets43 and is in line with the key role played by Akt in GPIb/IX/V-mediated αIIbβ3 integrin-dependent adhesion, spreading, and aggregation.57 In conclusion, our study shows that the constitutive binding of VWF to GPIbα in PT-VWD platelets triggers deranged sig-

Haematologica | 107 July 2022

1650

ARTICLE - Impaired platelet function in PT-VWD

L. Bury et al.

naling, leading to the inhibition of αIIbβ3 activation, defective secretion and impaired platelet adhesion. Together, these generate a platelet function defect that may significantly contribute to the bleeding phenotype of these patients.

This mechanism may also account for the worsening of the bleeding diathesis of PT-VWD patients in conditions associated with an increase of circulating VWF, and therefore of its enhanced binding to GPIbα, such as surgery or preg-

Figure 5. SFK, Lyn and PECAM-1 are constitutively phosphorylated in PT-VWD platelets. (A) SFK phosphorylation at Tyr 416 in human control and patient (PT-VWD) washed platelets unstimulated (-) or stimulated with ristocetin (0.3 mg/mL) for 30 sec under stirring conditions, was assessed by western blotting by loading 30 µg of proteins. Densitometric analysis of the pSFK/total SFK ratio was performed using Image J software and is expressed in arbitrary units (AU). Data are means ± SEM from six independent experiments (Online Supplementary Table S1, samples C, D, E, F, G, H) (*P<0.05 vs. control; #P<0.05 vs. resting; two-way ANOVA). The bands shown for resting platelets and platelets stimulated with ristocetin belong to the same gel. (B) Lyn phosphorylation at Tyr397 in human control and patient (PT-VWD) washed platelets unstimulated (-) or stimulated with ristocetin (0.3 mg/mL) for 30 sec under stirring conditions, was assessed by western blotting by loading 30 mg of proteins. Densitometric analysis of the pLyn/total Lyn ratio was performed using Image J software and is expressed in arbitrary units (AU). Data are means ± SEM from six independent experiments (Online Supplementary Table S1, samples C, D, E, F, G, H) (*P<0.05 vs. control; #P<0.05 vs. resting; two-way ANOVA). The bands shown for resting platelets and platelets stimulated with ristocetin belong to the same gel. (C) Lyn phosphorylation at Tyr507 in human control and patient (PT-VWD) washed platelets unstimulated (-) or stimulated with ristocetin (0.3 mg/mL) for 30 sec under stirring conditions, was assessed by western blotting by loading 30 mg of proteins. Densitometric analysis of the pLyn/total Lyn ratio was performed using Image J software and is expressed in Continued on following page. Haematologica | 107 July 2022

1651

ARTICLE - Impaired platelet function in PT-VWD

L. Bury et al.

arbitrary units (AU). Data are means ± SEM from three independent experiments (Online Supplementary Table S1, samples F, G, H) (*P<0.05 vs. control; #P<0.05 vs. resting; two-way ANOVA). The bands shown for resting platelets and platelets stimulated with ristocetin belong to the same gel. (D) PECAM1 phosphorylation at Tyr686 in human control and patient (PT-VWD) washed platelets unstimulated (-) or stimulated with ristocetin (0.3 mg/mL) for 30 sec under stirring conditions, was assessed by western blotting by loading 30 µg of proteins. Densitometric analysis of the pPECAM1/total PECAM1 ratio was performed using Image J software and is expressed in arbitrary units (AU). Data are means ± SEM from six independent experiments (Online Supplementary Table S1, samples C, D, E, F, G, H) (*P<0.05 vs. control; #P<0.05 vs. resting; two-way ANOVA). The bands shown for resting platelets and platelets stimulated with ristocetin belong to the same gel. (E) PECAM1 phosphorylation at Ser702 in human control and patient (PT-VWD) washed platelets unstimulated (-) or stimulated with ristocetin (0.3 mg/ml) for 30 sec under stirring conditions, was assessed by western blotting by loading 30 mg of proteins. Densitometric analysis of the pPECAM1/total PECAM1 ratio was performed using Image J software and is expressed in arbitrary units (AU). Data are means ± SEM from three independent experiments (Online Supplementary Table S1, samples F, G, H) (*P<0.05 vs control; #P<0.05 vs. resting; two-way ANOVA). The bands shown for resting platelets and platelets stimulated with ristocetin belong to the same gel. (F) PKC substrate phosphorylation in human control and patient (PT-VWD) washed platelets unstimulated (-) or stimulated with ristocetin (0.3 mg/mL) for 30 sec under stirring conditions, was assessed by loading 30 mg of proteins. Densitometric analysis of the PKC substrate/actin ratio was performed using Image J software and is expressed in arbitrary units (AU). Data are means ± SEM from four independent experiments (Online Supplementary Table S1, samples E, F, G, H) (*P<0.05 vs. control; #P<0.05 vs. resting; two-way ANOVA). The bands shown for resting platelets and platelets stimulated with ristocetin belong to the same gel.

nancy.35-37 Our data obtained with platelets carrying two different GP1BA gain-of-function variants suggest that the platelet function defect in PT-VWD is independent of the GP1BA variant type and is due to the enhanced affinity of GPIbα for VWF. However, it should be noted that the variants we studied affect the same domain of GPIbα, therefore we cannot predict the effects on platelet function of variants in other domains.58 A similar mechanism for platelet dysfunction may play a role in the platelet dysfunction of 2B-VWD. Indeed, results of studies with human and murine 2B-VWD platelets carrying the p.V1316M VWF variant are similar to results of the present study. Both disorders share the same platelet function defect, characterized by αIIbβ3 dysfunction, α- and d-granule secretion defect, defective Ca2+ signaling and adhesion under flow conditions.5,42 Defective Rap1b and Akt activation and PKC upregulation with consequent desensitization have been shown for both disorders,43 together with a dysregulation of the RhoA pathway.59 It would be interesting to assess Lyn and PECAM1 phosphorylation in 2B-VWD platelets to check whether the same negative-feedback loop downregulating platelet activation here described is present in 2B-VWD. Finally, our results imply that inhibitors of PECAM1 might be explored to restore platelet function in PT-VWD. However, the PECAM1 blockers currently available are antibodies that block PECAM1 adhesive interactions60-63 and not PECAM1 phosphorylation or downstream signaling, therefore they do not allow assessment of whether PECAM1 blockade might restore normal platelet function. The development of PECAM1 inhibitors selectively suppressing intracellular phosphorylation might represent a

novel approach to the antihemorrhagic treatment of PTVWD. Disclosures No conflicts of interest to disclose. Contributions LB, EF, AMM, GG and SM performed experiments, analyzed and interpreted data; PG designed and supervised the study; PG contributed the patient for the study; LB and PG wrote the manuscript; PG critically revised the manuscript. Acknowledgments The continued collaboration of our PT-VWD patient is gratefully acknowledged. Funding This work was supported by a Telethon grant (GGP15063) to PG and by a fellowship by Fondazione Umberto Veronesi to LB and EF. The authors thank Prof. Jerry Ware (University of Arkansas, USA) and Dr. Maha Othman (Queen’s University, Canada) for kindly providing the TgWT and TgG233V mice and Prof. Debra K. Newman (Medical College of Wisconsin, USA) for the kind gift of the anti PECAM1 antibody. Data-sharing statement All data generated or analyzed during this study are included in this article and its supplementary material files. Further enquiries can be directed to the corresponding author.

References 1. Othman M, Kaur H, Emsley J. Platelet-type von Willebrand disease: new insights into the molecular pathophysiology of a unique platelet defect. Semin Thromb Hemost. 2013;39(6):663-673.

2. Othman M, Kaur H, Favaloro EJ, et al. Platelet type von Willebrand disease and registry report: communication from the SSC of the ISTH. J Thromb Haemost. 2016;14(2):411-414. 3. Gresele P, Bury L, Falcinelli E. Inherited platelet function

Haematologica | 107 July 2022

1652

ARTICLE - Impaired platelet function in PT-VWD

L. Bury et al.

disorders: algorithms for phenotypic and genetic investigation. Semin Thromb Hemost. 2016;42(3):292-305. 4. Guerrero JA, Kyei M, Russell S, et al. Visualizing the von Willebrand factor/glycoprotein Ib-IX axis with a platelet-type von Willebrand disease mutation. Blood. 2009;114(27):5541-5546. 5. Kaur H, Corscadden K, Ware J, Othman M. Thrombocytopathy leading to impaired in vivo haemostasis and thrombosis in platelet type von Willebrand disease. Thromb Haemost. 2017;117(3):543-555. 6. Kasirer-Friede A, Cozzi MR, Mazzucato M, De Marco L, Ruggeri ZM, Shattil SJ. Signaling through GP Ib-IX-V activates alpha IIb beta 3 independently of other receptors. Blood. 2004;103(9):3403-3411. 7. Li Z, Zhang G, Feil R, Han J, Du X. Sequential activation of p38 and ERK pathways by cGMP-dependent protein kinase leading to activation of the platelet integrin alphaIIb beta3. Blood. 2006;107(3):965-972. 8. Yin H, Liu J, Li Z, Berndt MC, Lowell CA, Du X. Src family tyrosine kinase Lyn mediates VWF/GPIb-IX-induced platelet activation via the cGMP signaling pathway. Blood. 2008;112(4):1139-1146. 9. Giannini S, Cecchetti L, Mezzasoma AM, Gresele P. Diagnosis of platelet-type von Willebrand disease by flow cytometry. Haematologica. 2010;95(6):1021-1024. 10. Bury L, Malara A, Momi S, Petito E, Balduini A, Gresele P. Mechanisms of thrombocytopenia in platelet-type von Willebrand disease. Haematologica. 2019;104(7):1473-1481. 11. Gresele P, Orsini S, Noris P, et al. Validation of the ISTH/SSC bleeding assessment tool for inherited platelet disorders: a communication from the Platelet Physiology SSC. J Thromb Haemost. 2020;18(3):732-739. 12. Suva LJ, Hartman E, Dilley JD, et al. Platelet dysfunction and a high bone mass phenotype in a murine model of platelet-type von Willebrand disease. Am J Pathol. 2008;172(2):430-439. 13. Gresele P, Falcinelli E, Giannini S, et al. Dominant inheritance of a novel integrin beta3 mutation associated with a hereditary macrothrombocytopenia and platelet dysfunction in two Italian families. Haematologica. 2009;94(5):663-669. 14. Vezza R, Spina D, Tallarida RJ, Nathan M, Page CP, Gresele P. Antivasoconstrictor and antiaggregatory activities of picotamide unrelated to thromboxane A2 antagonism. Thromb Haemost. 1997;78(5):1385-1391. 15. Riondino S, Lotti LV, Cutini L, Pulcinelli FM. Collagen-induced platelet shape change is not affected by positive feedback pathway inhibitors and cAMP-elevating agents. J Biol Chem. 2005;280(8):6504-6510. 16. DiMinno G, Silver MJ. Mouse antithrombotic assay: a simple method for the evaluation of antithrombotic agents in vivo. Potentiation of antithrombotic activity by ethyl alcohol. J Pharmacol Exp Ther. 1983;225(1):57-60. 17. Sebastiano M, Momi S, Falcinelli E, Bury L, Hoylaerts MF, Gresele P. A novel mechanism regulating human platelet activation by MMP-2-mediated PAR1 biased signaling. Blood. 2017;129(7):883-895. 18. Mumford AD, Frelinger AL 3rd, Gachet C, et al. A review of platelet secretion assays for the diagnosis of inherited platelet secretion disorders. Thromb Haemost. 2015;114(1):14-25. 19. Canobbio I, Cipolla L, Consonni A, et al. Impaired thrombininduced platelet activation and thrombus formation in mice lacking the Ca(2+)-dependent tyrosine kinase Pyk2. Blood. 2013;121(4):648-657. 20. Cozzi MR, Guglielmini G, Battiston M, et al. Visualization of nitric oxide production by individual platelets during adhesion in flowing blood. Blood. 2015;125(4):697-705.

21. Wall JE, Buijs-Wilts M, Arnold JT, et al. A flow cytometric assay using mepacrine for study of uptake and release of platelet dense granule contents. Br J Haematol. 1995;89(2):380-385. 22. Mustard JF, Perry DW, Ardlie NG, Packham MA. Preparation of suspensions of washed platelets from humans. Br J Haematol. 1972;22(2):193-204. 23. 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. 24. Bury L, Falcinelli E, Chiasserini D, Springer TA, Italiano JE Jr., Gresele P. Cytoskeletal perturbation leads to platelet dysfunction and thrombocytopenia in variant forms of Glanzmann thrombasthenia. Haematologica. 2016;101(1):46-56. 25. Momi S, Impagnatiello F, Guzzetta M, et al. NCX 6560, a nitric oxide-releasing derivative of atorvastatin, inhibits cholesterol biosynthesis and shows anti-inflammatory and anti-thrombotic properties. Eur J Pharmacol. 2007;570(1-3):115-124. 26. do Ceu Monteiro M, Sansonetty F, Goncalves MJ, O'Connor JE. Flow cytometric kinetic assay of calcium mobilization in whole blood platelets using Fluo-3 and CD41. Cytometry. 1999;35(4):302-310. 27. Amison RT, Momi S, Morris A, et al. RhoA signaling through platelet P2Y(1) receptor controls leukocyte recruitment in allergic mice. J Allergy Clin Immunol. 2015;135(2):528-538. 28. Paddock C, Lytle BL, Peterson FC, et al. Residues within a lipidassociated segment of the PECAM-1 cytoplasmic domain are susceptible to inducible, sequential phosphorylation. Blood. 2011;117(22):6012-6023. 29. Guglielmini G, Appolloni V, Momi S, et al. Matrix metalloproteinase-2 enhances platelet deposition on collagen under flow conditions. Thromb Haemost. 2016;115(2):333-343. 30. Lopez E, Bermejo N, Berna-Erro A, et al. Relationship between calcium mobilization and platelet alpha- and delta-granule secretion. A role for TRPC6 in thrombin-evoked delta-granule exocytosis. Arch Biochem Biophys. 2015;585:75-81. 31. Franke B, Akkerman JW, Bos JL. Rapid Ca2+-mediated activation of Rap1 in human platelets. EMBO J. 1997;16(2):252-259. 32. Li Z, Delaney MK, O'Brien KA, Du X. Signaling during platelet adhesion and activation. Arterioscler Thromb Vasc Biol. 2010;30(12):2341-2349. 33. Anguita E, Villalobo A. Src-family tyrosine kinases and the Ca(2+) signal. Biochim Biophys Acta Mol Cell Res. 2017;1864(6):915-932. 34. Cifuni SM, Wagner DD, Bergmeier W. CalDAG-GEFI and protein kinase C represent alternative pathways leading to activation of integrin alphaIIbbeta3 in platelets. Blood. 2008;112(5):1696-1703. 35. Senis YA, Mazharian A, Mori J. Src family kinases: at the forefront of platelet activation. Blood. 2014;124(13):2013-2024. 36. Ming Z, Hu Y, Xiang J, Polewski P, Newman PJ, Newman DK. Lyn and PECAM-1 function as interdependent inhibitors of platelet aggregation. Blood. 2011;117(14):3903-3906. 37. Noris P, Schlegel N, Klersy C, et al. Analysis of 339 pregnancies in 181 women with 13 different forms of inherited thrombocytopenia. Haematologica. 2014;99(8):1387-1394. 38. O'Connor D, Lester W, Willoughby S, Wilde JT. Pregnancy in platelet-type VWD: a case series. Thromb Haemost. 2011;106(2):386-387. 39. Orsini S, Noris P, Bury L, et al. Bleeding risk of surgery and its prevention in patients with inherited platelet disorders. Haematologica. 2017;102(7):1192-1203. 40. Sanchez-Luceros A, Woods AI, Bermejo E, et al. PT-VWD posing diagnostic and therapeutic challenges - small case series. Platelets. 2017;28(5):484-490.

Haematologica | 107 July 2022

1653

ARTICLE - Impaired platelet function in PT-VWD

L. Bury et al.

41. Kaur H, Ozelo M, Scovil S, James PD, Othman M. Systematic analysis of bleeding phenotype in PT-VWD compared to type 2B VWD using an electronic bleeding questionnaire. Clin Appl Thromb Hemost. 2014;20(8):765-771. 42. Casari C, Berrou E, Lebret M, et al. von Willebrand factor mutation promotes thrombocytopathy by inhibiting integrin alphaIIbbeta3. J Clin Invest. 2013;123(12):5071-5081. 43. Casari C, Paul DS, Susen S, et al. Protein kinase C signaling dysfunction in von Willebrand disease (p.V1316M) type 2B platelets. Blood Adv. 2018;2(12):1417-1428. 44. Guidetti GF, Torti M. The small GTPase Rap1b: a bidirectional regulator of platelet adhesion receptors. J Signal Transduct. 2012;2012:412089. 45. Harper MT, Poole AW. Diverse functions of protein kinase C isoforms in platelet activation and thrombus formation. J Thromb Haemost. 2010;8(3):454-462. 46. Liu J, Pestina TI, Berndt MC, Jackson CW, Gartner TK. Botrocetin/VWF-induced signaling through GPIb-IX-V produces TxA2 in an alphaIIbbeta3- and aggregation-independent manner. Blood. 2005;106(8):2750-2756. 47. Maxwell MJ, Yuan Y, Anderson KE, Hibbs ML, Salem HH, Jackson SP. SHIP1 and lyn kinase negatively regulate integrin αIIbβ3 signaling in platelets. J Biol Chem. 2004;279(31):32196-32204. 48. Wu Y, Asazuma N, Satoh K, et al. Interaction between von Willebrand factor and glycoprotein Ib activates Src kinase in human platelets: role of phosphoinositide 3-kinase. Blood. 2003;101(9):3469-3476. 49. Newman PJ, Newman DK. Signal transduction pathways mediated by PECAM-1: new roles for an old molecule in platelet and vascular cell biology. Arterioscler Thromb Vasc Biol. 2003;23(6):953-964. 50. Jones KL, Hughan SC, Dopheide SM, Farndale RW, Jackson SP, Jackson DE. Platelet endothelial cell adhesion molecule-1 is a negative regulator of platelet-collagen interactions. Blood. 2001;98(5):1456-1463. 51. Cicmil M, Thomas JM, Leduc M, Bon C, Gibbins JM. Platelet endothelial cell adhesion molecule-1 signaling inhibits the activation of human platelets. Blood. 2002;99(1):137-144. 52. Jones CI, Garner SF, Moraes LA, et al. PECAM-1 expression and activity negatively regulate multiple platelet signaling pathways.

FEBS Lett. 2009;583(22):3618-3624. 53. Crockett J, Newman DK, Newman PJ. PECAM-1 functions as a negative regulator of laminin-induced platelet activation. J Thromb Haemost. 2010;8(7):1584-1593. 54. Rathore V, Stapleton MA, Hillery CA, et al. PECAM-1 negatively regulates GPIb/V/IX signaling in murine platelets. Blood. 2003;102(10):3658-3664. 55. Soriano Jerez EM, Gibbins JM, Hughes CE. Targeting platelet inhibition receptors for novel therapies: PECAM-1 and G6b-B. Platelets. 2021;32(6):761-769. 56. Wu Y, Tworkoski K, Michaud M, Madri JA. Bone marrow monocyte PECAM-1 deficiency elicits increased osteoclastogenesis resulting in trabecular bone loss. J Immunol. 2009;182(5):2672-2679. 57. Yin H, Stojanovic A, Hay N, Du X. The role of Akt in the signaling pathway of the glycoprotein Ib-IX induced platelet activation. Blood. 2008;111(2):658-665. 58. Othman M, Notley C, Lavender FL, et al. Identification and functional characterization of a novel 27-bp deletion in the macroglycopeptide-coding region of the GPIBA gene resulting in platelet-type von Willebrand disease. Blood. 2005;105(11):4330-4336. 59. Kauskot A, Poirault-Chassac S, Adam F, et al. LIM kinase/cofilin dysregulation promotes macrothrombocytopenia in severe von Willebrand disease-type 2B. JCI Insight. 2016;1(16):e88643. 60. Qing Z, Sandor M, Radvany Z, et al. Inhibition of antigen-specific T cell trafficking into the central nervous system via blocking PECAM1/CD31 molecule. J Neuropathol Exp Neurol. 2001;60(8):798-807. 61. Bogen S, Pak J, Garifallou M, Deng X, Muller WA. Monoclonal antibody to murine PECAM-1 (CD31) blocks acute inflammation in vivo. J Exp Med. 1994;179(3):1059-1064. 62. Dasgupta B, Chew T, deRoche A, Muller WA. Blocking platelet/endothelial cell adhesion molecule 1 (PECAM) inhibits disease progression and prevents joint erosion in established collagen antibody-induced arthritis. Exp Mol Pathol. 2010;88(1):210-215. 63. Woodfin A, Voisin MB, Nourshargh S. PECAM-1: a multi-functional molecule in inflammation and vascular biology. Arterioscler Thromb Vasc Biol. 2007;27(12):2514-2523.

Haematologica | 107 July 2022

1654

ARTICLE - Platelet Biology and its Disorders

Establishment of a Bernard-Soulier syndrome model in zebrafish Qing Lin,1,2* Riyang Zhou,1,2* Panpan Meng,2* Liangliang Wu,2 Lian Yang,2 Wenyu Liu,2 Jiaye Wu,2 Yuhuan Cheng,1 Linjuan Shi1 and Yiyue Zhang1,2 Key Laboratory of Zebrafish Modeling and Drug Screening for Human Diseases of Guangdong Higher Education Institutes, Department of Developmental Biology, School of Basic Medical Sciences, Southern Medical University and 2Division of Cell, Developmental and Integrative Biology, School of Medicine, South China University of Technology, Guangzhou, China *QL, RZ and PM contributed equally as co-first authors. 1

Correspondence: Yiyue Zhang mczhangyy@scut.edu.cn Received: March 30, 2021. Accepted: August 6, 2021. Prepublished: August 19, 2021. https://doi.org/10.3324/haematol.2021.278893 ©2022 Ferrata Storti Foundation Haematologica material is published under a CC BY-NC license

Abstract Platelets play an essential role in thrombosis and hemostasis. Abnormal hemostasis can cause spontaneous or severe post-traumatic bleeding. Bernard-Soulier syndrome (BSS) is a rare inherited bleeding disorder caused by a complete quantitative deficiency in the GPIb-IX-V complex. Multiple mutations in GP9 lead to the clinical manifestations of BSS. Understanding the roles and underlying mechanisms of GP9 in thrombopoiesis and establishing a proper animal model of BSS would be valuable to understand the disease pathogenesis and to improve its medical management. Here, by using CRISPR-Cas9 technology, we created a zebrafish gp9SMU15 mutant to model human BSS. Disruption of zebrafish gp9 led to thrombocytopenia and a pronounced bleeding tendency, as well as an abnormal expansion of progenitor cells. The gp9SMU15 zebrafish can be used as a BSS animal model as the roles of GP9 in thrombocytopoiesis are highly conserved from zebrafish to mammals. Utilizing the BSS model, we verified the clinical GP9 mutations by in vivo functional assay and tested clinical drugs for their ability to increase platelets. Thus, the inherited BSS zebrafish model could be of benefit for in vivo verification of patient-derived GP9 variants of uncertain significance and for the development of potential therapeutic strategies for BSS.

Introduction Platelets play an essential role in thrombosis and hemostasis. Without nuclei, they are the smallest formed elements of the blood in mammals, and are produced from giant polyploid precursors, the megakaryocytes.1 Megakaryocytes differentiate from hematopoietic stem cells in the hematopoietic sites (bone marrow, yolk sac, fetal liver and spleen). It is well accepted that hematopoietic stem cells give rise to megakaryocyte-erythrocyte progenitors in the first instance, and these progenitors commit to both erythroid and megakaryocytic lineage cells.2-4 During thrombopoiesis, megakaryocytes then arise from committed megakaryocytic precursors, undergo several cycles of endomitosis to become polyploid and release platelets as cell fragments into the bloodstream.5 When a blood vessel ruptures, adherent platelets become activated and initiate the platelet aggregation process, so the circulating platelets adhere to different components of vascular subendothelial structures to stop the bleeding.3 Functional platelets require a series of platelet membrane receptors properly expressed, as these receptors are es-

sential for platelet adhesion to form clots on the damaged vessel wall and trigger transmembrane signaling leading to cell aggregation and activation. The glycoprotein Ib-IXV complex (GPIb-IX-V), expressed in platelets and megakaryocytes, belongs to the leucine-rich repeat family of membrane proteins. The complex consists of four distinct transmembrane polypeptide subunits,6 of which GPIbα, GPIbβ, and GPIX are all necessary for efficient biosynthesis of the receptor, whereas GPV is more loosely associated.7,8 When blood vessels rupture causing bleeding, the GPIbIX-V complex mediates platelet attachment to the sites of blood vessel wall injury and activates the platelets.9 Bernard-Soulier syndrome (BSS), a rare autosomal recessively inherited bleeding disorder also known as hemorrhagiparous thrombocytic dystrophy,7-9 is caused by quantitative or qualitative defects within the membrane GPIb-IX-V complex.6 The syndrome is characterized by thrombocytopenia, giant platelets with the absence of platelet aggregation in response to ristocetin, and a range of sometimes life-threatening mucocutaneous bleeding disorders.7,9 Mutations in GP1BA (GPIbα), GP1BB (GPIbβ), and GP9 (GPIX) have all been reported in BSS patients,10

Haematologica | 107 July 2022

1655

ARTICLE - Modeling Bernard-Soulier syndrome in zebrafish with the reported mutation frequency being higher in GP9 than in the other two genes in a study of 211 BSS families.10 GPIX is a subunit of the platelet membrane glycoprotein complex, and multiple point mutations in the GP9 gene are reported to be associated with BSS in patients.10,11 Several GP9 point mutations, such as GP9 70T>C and GP9 182A>G, have been reported to affect protein conformation and consequently attenuate the GPIb-IX-V complex component expression, but there is still a lack of in vivo functional evidence on whether these mutations affect thrombocytes.10,12-15 Disruption of GPIX in mice results in thrombocytopenia with giant platelets and prolonged bleeding phenotypes similar to human BSS, while intrauterine embryogenesis makes mice difficult to manipulate during early developmental stages.16 Whether and how GP9 participates in the regulation of embryonic thrombocyte development remains incompletely known. Zebrafish (Danio rerio) have been used as powerful vertebrate models for developmental biology and for modeling many human diseases. Their high fecundity, external development and their optical transparency during early development facilitate high-throughput genetic and small-molecule screening, especially for in vivo phenotype fast-reading in early developmental stages. Importantly, thrombocytes in early vertebrates are equivalent to mammalian platelets in functional and regulatory terms.17,18 Thrombocytopoiesis has been described in zebrafish and has demonstrated conservation of regulatory factors and similar developmental processes as those in mammals.19,20 Thrombocyte disorders including congenital amegakaryocytic thrombocytopenia, inherited thrombocytopenia, essential thrombocythemia and several platelet functional disorders have been modeled successfully in zebrafish.2125 In addition, thrombocytic lineage-specific transgenic reporter strains of zebrafish are available, such as the Tg(cd41:eGFP) and Tg(mpl:eGFP) lines.19,21 Thrombocyte function can be detected by the efficiency of clotting, measured as the time to occlusion or the thrombus surface area, both having been well-established in zebrafish.21,26 As multiple mutations in GP9 are associated with clinical manifestations of BSS, understanding the functions and underlying mechanisms of GP9 in thrombocytopoiesis and establishing a proper zebrafish model of BSS would be valuable to understand the disease pathogenesis and to verify the clinical significance of patient-derived GP9 variants. In the current study, we targeted the gp9 gene using CRISPR-Cas9 technology to generate an inherited BSS zebrafish model. The gp9SMU15 zebrafish modeled BSS well, as it displayed thrombocytopenia from early development to adult stage, thrombocytic progenitor expansion in adult hematopoiesis, and exhibited bleeding disorders. We further demonstrated that human GP9 mirrored the thrombocyte phenotypes in the zebrafish model, and pa-

Q. Lin et al. tient-derived GP970T>C and GP9182A>G mutations are thrombocytopoietic deficient mutations responsible for human BSS. Moreover, we found that recombinant human thrombopoietin (rhTPO), used clinically for promoting thrombocytopoiesis, was also effective for relapsing BSS thrombocytopenia phenotypes in the model. Interestingly, we found that decitabine, used for treating leukemia and idiopathic thrombocytopenia purpura in the clinic, could also effectively alleviate thrombocytopenia in BSS zebrafish. Thus, gp9SMU15 zebrafish could serve as an ideal model for understanding the deficiency of thrombocytopoiesis and hemostasis associated with BSS, and it could be applied as a useful tool for fast clarifying human patient-derived GP9 variants of unknown clinical significance, as well as for the development of therapeutic strategies for BSS.

Methods Generation of gp9SMU15 mutant lines We generated gp9SMU15 mutants using the CRISPR/Cas9 method. The zebrafish gp9 target within exon 2 is 5’GGGCAAAGTCACGCACCTGC-3’. gp9 guide RNA was transcribed in vitro by using mMESSAGE mMACHINE kit (Ambion). We generated gp9 mutants by injecting one-cell stage embryos with Cas9 protein (EnGen Cas9 NLS, NEB) and gRNA. F0 fish were crossed with wild-type (WT) fish to produce F1 progeny, which were then genotyped and sequenced to identify the inheritable mutation. We acquired the main frameshift mutations with the deletion of 17 bp (F1). We crossed the heterozygotes F1 gp9SMU15/+ to generate homozygous mutant F2 gp9SMU15. All work involving zebrafish was reviewed and approved by the Animal Ethics Committee or the Animal Research Advisory Committee of Southern Medical University and South China University of Technology Analysis of thrombocyte function We monitored the time to stop bleeding or the bleeding area size in adult zebrafish and the time to occlusion in zebrafish larvae after injury. Comparable cuts were made in WT siblings and gp9SMU15 mutants with a scalpel blade at the junction region between the torso and the tail fin to destroy the vessels near the tail part; the cuts were performed in a single-blind manner to exclude the subjective factor of operation, and the fish were genotyped after the injury. The time to stop bleeding (the time the injury was made to the time the bleeding stopped) was measured under a microscope and recorded by a video camera. Sodium hydroxide (NaOH) treatment to induce gill bleeding was performed as previously described with some modifications.27,28 We placed zebrafish in a glass dish containing 15 mL of 20 mM NaOH solution, and the process was recorded under a microscope (Carl Zeiss Meditec AG, Jena,

Haematologica | 107 July 2022

1656

ARTICLE - Modeling Bernard-Soulier syndrome in zebrafish Germany). FeCl3 treatment was performed as previously described with some modifications.26,29 In detail, 6-day post-fertilization (dpf) larvae, with various genotypes Tg(cd41:eGFP);WT siblings and Tg(cd41:eGFP);gp9SMU15 were

Q. Lin et al. immobilized in 0.8% low-melt agarose gel and 0.02% anesthetic tricaine (Sigma-Aldrich). A drop (2.5 mL) of 1% FeCl3 was placed on the tail region. From the time the first drop appeared on the tail region to the accumulation of the

Figure 1. Generation and characterization of gp9SMU15 zebrafish using CRISPR/Cas9. (A) The zebrafish gp9 gene structure. The start coding region is shown with ATG. The targeting sequence aimed at the coding region is shown in the purple dashed box. (B) Sanger sequencing identified the wild-type (WT) sequence (top) and 17-bp deletion (red arrowhead) in gp9SMU15. Frameshift mutation of gp9 creates premature stop codons indicated with the red asterisk. (C) Agarose gel pictures of the polymerase chain reaction product from WT, mutant, and heterozygote zebrafish. Lane 1: the wild type bands; lane 2: the mutant bands; lanes 3: the heterozygote bands. (D) Gp9 protein structures in WT and mutant fish. The Gp9 protein analyzed with the SMART program contains one leucine-rich repeat N-terminal (LRRNT) domain, one leucine-rich repeat C-terminal (LRRCT) domain and one transmembrane domain. Orange ovals indicate a LRRNT or LRRCT domain, the green box indicates the transmembrane domain. Red slashes indicate the premature stop of Gp9 protein. (E) gp9 17,+0 mutated transcripts generated in gp9SMU15 mutants. Primer pairs F1 (wt_FP) F2 (mut_FP) F3 (common_FP) and R1 (common_RP) were utilized for detecting the WT form and gp9 17,+0 form, respectively. Statistical significance was determined using a two-sample Student t-test, n≥10 per group, data were combined from four biological replicates, mean ± standard error of mean. **P<0.01; *P<0.05; ns: not significant. Haematologica | 107 July 2022

1657

ARTICLE - Modeling Bernard-Soulier syndrome in zebrafish

Q. Lin et al.

thrombocytes at the injury site, the time to occlusion was measured at the microscopic stage monitored by a fluorescence microscope.

mutant zebrafish were utilized. Embryos were soaked into egg water containing drugs in appropriate concentration at 36 hours post-fertilization (hpf) using 12-well plates and scored at 4 dpf by counting the cd41:eGFPhigh-labeled Drug treatment thrombocytes to determine whether drugs were effective. SMU15 Tg(cd41:eGFP) sibling and Tg(cd41:eGFP);gp9 transgenic Fifteen embryos were put in each well with a volume of 3

Figure 2. gp9SMU15 adult fish display thrombocytopenia and abnormal precursor expansion. (A) Flow cytometry analysis of peripheral blood (PB) cells in wild-type (WT) and gp9SMU15 fish. FITC was directly proportional to 488-GFP cells and side scatter (SSC) was indicative of cellular granularity. (B) Percentage of cd41:eGFP+ thrombocytes in PB determined by flow cytometry (Student t-test, n=6; mean ± standard error of mean [SEM]; ns: not significant; *P<0.05). (C) May-Grünwald-Giemsa staining of kidney marrow (KM) cells in WT and gp9SMU15 fish. The scale bar represents 20 mm. Arrowheads are colored yellow for lymphocytes or thrombocytes and green for precursors. Asterisks are colored blue for myelomonocytes and red for erythrocytes. (D) Blood cell counts of WT and gp9SMU15 KM by May-Grünwald-Giemsa staining. L&T: lymphocytes and thrombocytes; P: precursors; M: myelomonocytes; E: erythrocytes. (Student t-test, n=6; mean ± SEM; ns: not significant; *P<0.05). (E) Flow cytometry analysis of KM cells in WT and gp9SMU15 fish. FITC was directly proportional to 488-GFP cells and SSC was indicative of cellular granularity. (F) Percentage of cd41:eGFPlow and cd41:eGFPhigh thrombocytes in KM cells determined by flow cytometry (Student t-test, n=6; mean ± SEM; ns: not significant; **P<0.01). Haematologica | 107 July 2022

1658

ARTICLE - Modeling Bernard-Soulier syndrome in zebrafish mL solution. To obtain a proper working concentration of rhTPO (rhTPO injection; 3SBIOINC, Shenyang, China) and decitabine (Selleck S1200; Selleck, Houston, TX, USA), we treated 1.5-dpf WT embryos with gradient concentrations of rhTPO (50, 100 and 500 U/mL) or decitabine (10, 20, 50 and 100 mM) until 4 dpf to see whether zebrafish thrombocytes could respond to the concentrations. Doses of 100 U/mL rhTPO and 20 mM of decitabine were chosen as the concentrations that could elevate WT thrombocytes efficiently but not cause developmental retardation in larvae.

Results The gp9SMU15 adult fish display thrombocytopenia and hematopoietic precursor expansion To investigate the role of gp9 in thrombocytopoiesis, we generated a gp9-deficient zebrafish mutant by targeting exon 2 of the gp9 gene using CRISPR-Cas9 technology (Figure 1A). We obtained an F0 mutant line containing gp9 (-17 bp, +0) (Figure 1B, C), and outcrossed this line with WT zebrafish to generate F1 heterozygous and F2 homozygous progenies. The gp9 (-17 bp, +0) mutant was predicted to encode a truncated Gp9 protein lacking the two-leucine rich repeat domains and the transmembrane domain (Figure 1D). Using quantitative real-time polymerase chain reaction to detect gp9 mRNA, we found that the mutant fish produced only the mutated mRNA (Figure 1E). Therefore, gp9 (-17 bp, +0) (gp9SMU15 hereafter) was predicted to be a loss-of-function mutant. To characterize the effects of gp9 disruption on thrombocytopoiesis in zebrafish, we determined thrombocyte numbers in gp9SMU15 mutants and their siblings by monitoring cd41:eGFP expression in the Tg(cd41:eGFP) line, as the vast majority of circulating thrombocytes could be recognized by flow cytometry analysis (FACS) with bright cd41:GFPhigh fluorescence in adult fish.19 Notably, although gp9SMU15 mutants were viable and able to survive to adulthood, the gp9SMU15 adult mutants displayed thrombocytopenia as the number of circulating cd41:eGFPhigh thrombocytes in adult mutants was about 78% of the number in adult WT sibling fish (Figure 2A, B). These data indicate that gp9SMU15 zebrafish displayed thrombocytopenia in the adult stage, resembling BSS in human patients. To directly examine BSS-like hematologic disorders in gp9SMU15 adult fish, kidney-marrow (KM) cells were collected from gp9SMU15 adult fish or siblings and subjected to cytological analyses and blood cell count. We found a significant expansion of the precursor cell population in adult gp9SMU15 fish (Figure 2C, D). Recent single-cell RNA-sequencing of zebrafish adult KM revealed the continuous nature of thrombocyte development,30 and FACS could recognize hematopoietic stem and progenitor cells (HSPC) to differentiated thrombocytes with weak to bright cd41:eGFP

Q. Lin et al. (GFPlow to GFPhigh) fluorescence in the adult KM cells.19,31,32 To compare the thrombocyte population of gp9SMU15 fish with that of their WT siblings in KM, cd41:eGFP+ cells were isolated by FACS. In accordance with the cytological results (Figure 2C, D), the FACS results also showed an increase in the cd41:eGFPlow cell population from 73% in the adult WT sibling fish to 82% in adult gp9SMU15 fish (Figure 2E, F), suggesting that gp9SMU15 fish had a higher percentage of cd41:eGFPlow HSPC than had their WT siblings. The morphological analysis revealed that there were no obvious cell size changes within the cd41:eGFPlow HSPC population (Online Supplementary Figure S1A, C). However, the populations of cd41:eGFPhigh thrombocytes in gp9SMU15 mutant KM and peripheral blood were both larger than those in their WT siblings (medium vs. small) (Online Supplementary Figure S1A-D). The above data demonstrate that gp9SMU15 zebrafish mutants probably have a developmental block, with expansion of larger thrombocytes and reduced small-nucleated thrombocytes in adulthood. BSS patients display increased bone marrow megakaryocytes but reduced mature platelets.34 In our study, zebrafish gp9SMU15 mutants showed a similar expansion of the progenitor population but reduced circulating thrombocytes. The gp9SMU15 mutants display impaired thrombocyte function BSS patients generally have moderate thrombocytopenia, giant platelets and a bleeding tendency. They may suffer severe hemorrhage following injury or during surgery.33 Under normal conditions, WT zebrafish rarely bleed severely, whereas ~21% (4/19) of gp9SMU15 mutants had blood spots on the body or near the tail fins (Figure 3A), indicating that the gp9 mutation led to a pronounced bleeding tendency. Thrombocyte function can be assessed by the ability to stop bleeding and to clot, which has been wellestablished in the zebrafish model.21,29 By calculating the time to stop bleeding after a tail-cut injury, we found that the average time in gp9SMU15 adult mutants was ~50 s, much longer than in the WT group (~20 s) (Figure 3B, C). We also used NaOH as a substance to create chemical damage to induce gill bleeding.27,28 We found more severe gill bleeding in gp9SMU15 fish than in WT fish (Figure 3D, E). We then monitored the accumulation of cd41:eGFPhigh thrombocytes following administration of FeCl3 to induce an oxidative injury to the vascular endothelium,26,29 and found that the time to accumulation of thrombocytes at the injured venule in gp9SMU15 zebrafish larvae was significantly longer (mean ± standard error of mean [SEM], 36.4 ± 5.2 s) than in the WT sibling controls (mean ± SEM, 22.2 ± 2.4 s) (Figure 3F). Taken together, these results demonstrated that the thrombocytopenic gp9SMU15 zebrafish exhibited bleeding disorders similar to the extended clinical manifestations of clotting and prolonged bleeding time in BSS patients.

Haematologica | 107 July 2022

1659

ARTICLE - Modeling Bernard-Soulier syndrome in zebrafish

Q. Lin et al.

B C

D E

Figure 3. gp9SMU15 mutants display impaired thrombocyte function. (A-D) Clotting in adult wild-type (WT) siblings and gp9SMU15 mutants. Representative images, under normal conditions, of blood spots on the zebrafish body or near the tail fins (A). The time to stop bleeding in adult WT siblings (upper) and gp9SMU15 mutants (lower) (B). (C) Quantitative data showing the time to stop bleeding in WT and gp9SMU15 siblings (D) Representative images of gill bleeding after injury in the sibling and gp9SMU15 mutants. The representative pictures were captured at 90 s. (E) Quantitative data showing the red pixels around the gill in WT and gp9SMU15 fish The red color pixels, indicating the extent of bleeding, were counted by Adobe photoshop CC software. (F) Zebrafish after injury induced by 1% ferric chloride: quantitative data showing the time to accumulation of thrombocytes at the injured venule. Statistical significance was determined by a two-sample Student t-test, n≥10, mean ± standard error of mean, *P<0.05. Scale bars: 500 mm. Haematologica | 107 July 2022

1660

ARTICLE - Modeling Bernard-Soulier syndrome in zebrafish The gp9-disruption led to thrombocytopenia in zebrafish larvae To characterize the effects of gp9 disruption on thrombocytopoiesis in zebrafish embryos or larvae, we determined thrombocyte numbers in gp9SMU15 mutants and their siblings by monitoring cd41 mRNA and cd41:eGFP expression. In zebrafish larvae, two distinct populations of cd41:eGFP cells could be recognized, cd41:eGFPlow precursor and progenitor cells and the cd41:eGFPhigh thrombocytes, by FACS or by whole mount observation under a fluorescence microscope.19,32,34 We isolated GFP+ cells from 3-dpf embryos by FACS, and found that the cd41:eGFPhigh cell population was significantly decreased but that the cd41:eGFPlow cell population was increased in 3-dpf gp9SMU15 embryos (Figure 4A, B), suggesting that thrombocyte differentiation was affected by the gp9 mutation. Furthermore, quantification analysis showed that, at 4 dpf, signals of cd41:eGFPhigh cells in the caudal hematopoietic tissue (CHT) region were markedly reduced in gp9SMU15 mutant larvae compared with the signals from WT sibling embryos (Figure 4C, D), indicating that gp9SMU15 mutants exhibited thrombocytopenia during embryonic development. This conclusion was further supported by the downregulation of several thrombocyte-related genes, such as cd41 and genes required for thrombocytopoiesis (fog1 and nfe2) (Figure 4E). Notably, the aorta-gonad-mesonephros (AGM) cd41:eGFPlow cells, mainly represented by HSPC, were unaltered in 2-dpf gp9SMU15 mutants (Figure 4F), and cmyb expression was unaltered in the AGM and CHT regions (Figure 4G), suggesting that HSPC were unaffected by gp9 mutations. Other definitive blood lineages were also unaffected, as indicated by unaltered mpo+ neutrophils at 3-dpf, βe1globin+ and O-dianisdine+ erythrocytes, and rag1+ lymphocytes at 5-dpf (Online Supplementary Figure S2A-E). Furthermore, primitive blood lineages including embryonic erythroid markers (αe1-globin and gata1) and a primitive myeloid marker (pu.1) were unaffected in gp9SMU15 mutants (Online Supplementary Figure S3A-D). Taken together, these results show that gp9SMU15 mutants exhibit a lineage-specific thrombocytic deficiency without other blood lineage cells being affected. To rule out the possibility of off-target mutations, we overexpressed Gp9-eGFP fusion protein under the control of the heat-shock promoter (hsp:gp9-egfp) and showed that heat-shock-induced gp9 overexpression could protect gp9SMU15 mutants from thrombocytopenia (Online Supplementary Figure S4A-C). These data demonstrate that the loss of gp9 function leads to thrombocytopenia in zebrafish larvae.

Q. Lin et al. cyte proliferation is reduced or apoptosis increased in gp9-deficient mutants. Thus we monitored the CHT region cd41:eGFP positive cell growth or death changes in the Tg(cd41:eGFP);gp9SMU15 transgenic line.33 A 5-bromo-2'-deoxyuridine (BrdU) pulse labeling incorporation assay revealed that the percentage of BrdU-positive cd41:eGFP+ cells in total cd41:eGFP+ cells in the CHT was markedly reduced in gp9SMU15 mutants at 3 dpf, and the decreased population were BrdU-labeled cd41:GFPlow cells rather than cd41:GFPhigh cells, suggesting that the proliferation of cd41:eGFPlow hematopoietic precursors or thrombocyte progenitor cells was affected by the gp9 mutation (Figure 5A, B). To investigate the apoptosis of thrombocytes in gp9SMU15 mutants, we performed a terminal deoxynucleotidyl transferase dUTP nick-end labeling (TUNEL) assay, and found almost no cd41:eGFP/TUNEL double-positive cells in the CHT at 3 dpf in either WT siblings or gp9SMU15 mutants (Figure 5C), indicating that thrombocytopenia was not likely caused by apoptosis. These results demonstrated that thrombocytopenia in gp9SMU15 mutants could be mainly attributed to reduced proliferation.

Validation of the zebrafish model using human BernardSoulier syndrome-related GP9 mutations The phylogenetic tree of Gp9 suggests the evolutional conservation from zebrafish to mammals (Online Supplementary Figure S5A). Multiple sequence alignment demonstrates the similarity of GPIX between zebrafish and mammals (Online Supplementary Figure S5B), especially in the ectodomain (with ~37% identity).35 Human BSS is reported to be associated with several GP9 mutations, such as GP9 c.70T>C (C24R) and GP9 c.182A>G (N61S) (Online Supplementary Figure S5B), which have been noted to change the GPIX conformation and mildly impair the protein expression of itself and the GPIB-IX-V complex at the molecular level.10,12,13 However, there is still a lack of functional evidence for the two mutations affecting thrombocytopoiesis in animal models. To verify the mutation’s effects in models, we first studied gp9SMU15 zebrafish to see whether human GP9 (hsGP9WT hereafter) could rescue zebrafish BSS. We found that thrombocytopenia of zebrafish gp9SMU15 mutants could be recovered by overexpression of hsGP9WT (Figure 6A-C), as cd41:eGFPhigh thrombocyte counts in gp9SMU15 zebrafish injected with hsGP9WT mRNA (34.2 ± 3.2 thrombocytes) increased to almost the normal level in WT siblings (34.6 ± 1.9 thrombocytes) (Figure 6B, C). These results further confirmed the functional conservation of thrombocytopoiesis in zebrafish and human GP9, and suggested that gp9SMU15 zebrafish could be utilized as a BSS model for Thrombocytopenia in gp9SMU15 mutants is caused by fast-validating the functional significance of hsGP9 variants. reduced proliferation of thrombocyte precursors We further overexpressed the human BSS isoform GP9c.70T>C To explain the cellular basis of the thrombocytopenia in and c.182A>G mRNA (hereafter hsGP970T>C and hsGP9182A>G, regp9SMU15 mutants, we further explored whether thrombo- spectively, as shown in Online Supplementary Figure S5A) Haematologica | 107 July 2022

1661

ARTICLE - Modeling Bernard-Soulier syndrome in zebrafish

Q. Lin et al.

Figure 4. Thrombocytopenia in gp9SMU15 zebrafish embryos. (A-D) cd41:eGFPhigh thrombocytes were decreased in gp9SMU15 embryos. Flow cytometry analysis of cd41:eGFP+ cells in 3-day post fertilization (dpf) wild-type (WT) and gp9SMU15 embryos. GFP area was directly proportional to 488-GFP cells and side scatter (SSC) was indicative of cellular granularity (A). Percentage of cd41:eGFPlow and cd41:eGFPhigh thrombocytes in 3-dpf embryo cells determined by flow cytometry (Student t-test, n>100 embryos per group, and data were combined from three biological replicates, mean ± standard error of mean [SEM], ns: not significant, **P<0.01) (B). Representative images for staining of cd41:eGFP protein in 4 dpf Tg(cd41:eGFP);WT and mutant Tg(cd41:eGFP);gp9SMU15 embryos. Images showed GFPhigh signals (C). Statistical significance was determined using a two-sample Student t-test, n>10, mean ± SEM, ***P<0.001 (D). (E) Relative expressions of thrombocytic markers (cd41, nfe2 and fog1) in 3dpf WT siblings (blue column) and gp9SMU15 mutants (red column) by quantitative real-time polymerase chain reaction. Statistical significance was determined using a two-sample Student t-test; n≥10 per group, and data were combined from three biological replicates, mean ± SEM, *P<0.05, ***P<0.001. (F) Hematopoietic stem and progenitor cells (HSPC) were not affected in gp9SMU15 mutants. Quantification of the aorta-gonad-mesonephros (AGM)-localized cd41:eGFPlow HSPC in gp9SMU15 mutants and WT embryos at 2 dpf. Statistical significance was determined using a two-sample Student t-test, n>10, mean ± SEM, ns: not significant. (G) Whole-mount in situ hybridization (WISH) of cmyb expression in control (left panel) and gp9SMU15 (right panel) embryos at 36 hours post fertilization (hpf) and 3 dpf. Haematologica | 107 July 2022

1662

ARTICLE - Modeling Bernard-Soulier syndrome in zebrafish

Q. Lin et al.

Figure 5. Abated thrombocyte precursor proliferation in gp9SMU15 mutants. (A) Profiling of BrdU incorporation by caudal hematopoietic tissue (CHT) cd41:eGFP+ cells in 3-day post-fertilization (dpf) Tg(cd41:eGFP);WT (left) and Tg(cd41:eGFP);gp9SMU15 (right) by double antibody staining of BrdU (red signals) and GFP (green signals). White triangular arrowheads indicate the yellow color merged signals represent the BrdU+cd41low cells. Gray triangular arrowheads indicate the yellow color merged signals represent the BrdU+cd41hiigh cells. The images of the stained samples were captured by setting the confocal parameter as pinhole size 35 mm and 488 laser gain 700 to filter most cd41:eGFPlow signals, and the cd41:eGFPhigh thrombocytes fluorescent signals were counted. Scale bars: 50 mm. (B) Statistical data showing the percentages of CHT-localized cd41:eGFP+ cells that incorporated BrdU. Asterisks indicate statistical difference, determined using a two-sample Student t-test, n≥10, mean ± standard error of mean, **P<0.01. (C) No TUNEL incorporation with 3-dpf cd41:eGFP+ cells in the CHT of Tg(cd41:eGFP);gp9SMU15 mutants (right panels) and their wild-type (WT) siblings (left panels). Green: GFP; red: TUNEL. The green arrowheads indicate the cd41:eGFP+ cells and the red arrowheads indicate the TUNEL+ cells, No cd41:eGFP+/TUNEL+ cells were detected. Scale bars: 50 mm. Haematologica | 107 July 2022

1663

ARTICLE - Modeling Bernard-Soulier syndrome in zebrafish

Q. Lin et al.

in gp9SMU15 zebrafish embryos to evaluate the function of the two isoforms on thrombocytopoiesis. Unlike hsGP9WT overexpression, hsGP970T>C and hsGP9182A>G overexpression did not restore cd41:eGFPhigh thrombocyte counts in gp9SMU15 mutants (Figure 6B, C), suggesting that two human GP9 mutations affect GP9 function on thrombocytopoiesis. The above data demonstrate that the gp9SMU15 zebrafish model could serve as a good BSS model for functional validation of human GP9 mutations, and that hsGP970T>C and hsGP9182A>G are both loss-of-function mutations.

toms of thrombocytopenia, we wondered whether drugs clinically used for the treatment of thrombocytopenia would be effective to relieve BSS. We therefore utilized the Tg(cd41:eGFP);gp9SMU15 BSS zebrafish to test drug responses. rhTPO is a commercially available thrombocytopoietic agent approved by the US Food and Drug Administration (FDA) for the treatment of human thrombocytopenia.39 We treated 1.5-dpf WT sibling embryos with rhTPO until 4 dpf to see whether zebrafish thrombocytes could respond to rhTPO (Figure 7A) and found that 100 U/mL rhTPO could effectively increase cd41:eGFPhigh thrombocyte numbers in both WT Thrombopoietin and decitabine effectively modify sibling embryos and gp9-deficient mutants. This suggested Bernard-Soulier syndrome phenotypes in the model a conserved response from zebrafish thrombocytes to In clinical settings, the primary treatment for BSS is bone mammalian cytokines, and that rhTPO could relieve thrommarrow transplantation. Since BSS patients show symp- bocytopenia in BSS fish (Figure 7B, C).

Figure 6. Functional validations of the human-derived GP9 mutations associated with Berndard-Soulier syndrome in the gp9SMU15 zebrafish model. (A) Schematic diagram of the validation of the Bernard-Soulier syndrome (BSS) zebrafish model using human BSS-related GP9 mutations. hsGP9WT, hsGP970T>C and hsGP9182A>G mRNA were injected into one cell of the Tg(cd41:eGFP);gp9SMU15 embryos. (B) Representative images of staining of cd41:eGFP protein in 4 days post-fertilization (dpf) Tg(cd41:eGFP);WT and mutant Tg(cd41:eGFP);gp9SMU15 embryos. Images show GFPhigh signals. (C) Quantification of GFPhigh cell numbers at the caudal hematopoietic tissue (CHT) region. Statistical significance was determined by a Student t-test, n>10, mean ± SEM, ***P<0.001, ****P<0.0001. Scale bars: 50 mm. Haematologica | 107 July 2022

1664

ARTICLE - Modeling Bernard-Soulier syndrome in zebrafish Decitabine, approved by the FDA for the treatment of leukemia and myelodysplastic syndrome,36 is now being used for the treatment of idiopathic thrombocytopenia purpura in clinical trials.37 Whether it could be used to

Q. Lin et al. treat BSS is unknown. We found that 20 mM decitabine could effectively increase cd41:eGFPhigh thrombocytes in WT sibling embryos, and also expand thrombocytes in gp9SMU15 embryos (Figure 7D, E), suggesting that decitabine

Figure 7. Therapeutic responses of Tg(cd41:eGFP);gp9SMU15. (A) Schematic diagram of the drug treatment. The embryos were exposed to recombinant human thrombopoietin (rhTPO) and decitabine at 1.5 days post-fertilization (dpf) and fixed at 4 dpf. (B) Thrombocytopenia indicated by decreased cd41:eGFPhigh cells in Tg(cd41:eGFP);gp9SMU15 mutant embryos and thrombocyte expansion by rhTPO. Images show GFPhigh signals. Representative images of Tg(cd41:eGFP);WT and Tg(cd41:eGFP);gp9SMU15 embryos treated with sodium chloride saline control and rhTPO (B). (C) Statistical significance was determined using a paired Student ttest, n≥5, mean ± standard error of mean (SEM), *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001. (D) Representative images of Tg(cd41:eGFP);WT and Tg(cd41:eGFP);gp9SMU15 embryos treated with dimethylsulfoxide (DMSO) control and decitabine. Images show GFPhigh signals. (E) Quantification of cd41:eGFPhigh numbers in the caudal hematopoietic tissue (CHT). Statistical significance was determined by a paired Student t-test, n>10, mean ± SEM. *P<0.05, **P<0.01, ***P<0.001,****P<0.0001. Scale bars: 50 mm. Haematologica | 107 July 2022

1665

ARTICLE - Modeling Bernard-Soulier syndrome in zebrafish could increase the number of zebrafish thrombocytes and may relieve thrombocytopenia in the BSS model. The above results demonstrate that both rhTPO and decitabine could effectively relieve thrombocytopenia in BSS zebrafish (although thrombocyte numbers were not fully rescued), suggesting that these drugs, clinically used for thrombocytopenia, could also be applied in the treatment of BSS. The response of the BSS zebrafish model to these therapeutic agents suggests that the model could be useful for the evaluation and screening of thrombopoietic drugs.

Discussion BSS, characterized by prolonged bleeding times, enlarged platelets, an inability to clot and thrombocytopenia, are bleeding disorders that result from mutations in genes of the glycoprotein Ib complex, including GP1bA, GP1bb and GP9.10 GP9 in particular is the most frequently mutated gene reported in a group of BSS patients from 211 families.10 In this study, we generated and characterized a CRISPR-Cas9 targeted gp9SMU15 zebrafish mutant, which is the first established inherited BSS zebrafish model. The gp9-deficient zebrafish were viable and displayed thrombocytopenia, with a bleeding disorder, thus resembling Gp9-knockout mice and human BSS patients. The gp9SMU15 zebrafish mutant provides a complementary model to aid our understanding of the various roles of gp9 in thrombocytopoiesis and in BSS. Thrombocytopenia is an important feature reported in both human BSS and Gp9 knockout mice. Our data strengthened our understanding of the role of gp9 in thrombocytopoiesis in vertebrates, as gp9-deficient zebrafish showed a consistent thrombocytopenia phenotype from embryonic stages to adulthood. We also demonstrated that gp9SMU15 fish showed reduced thrombocytes but increased precursor cells from embryonic to adult stages. Interestingly, the 3 dpf whole-mount in situ hybridization showed unaltered cmyb expression but cd41:eGFPlow cells were increased, suggesting that the increased cd41:eGFPlow cells were likely thrombocyte-erythroid progenitors rather than cmyb+ hematopoietic stem cells or other lineage progenitors. In particular, their KM cd41:eGFP+ cells were larger than their siblings, similar to the larger thrombocytic cell volumes in mice and humans.9,16 As regards the cellular mechanisms for the thrombocytopenia caused by gp9 mutations, we further clarified that thrombocytopenia may be attributed to proliferation defects in thrombocyte lineage cells. It is probable that proliferation blocked precursors or thrombocyte progenitors had consequent problems in differentiating into thrombocytes. Thrombocytes in zebrafish are the functional equivalent of mammalian platelets, playing major roles in clotting and preventing bleeding.18 Patients with thrombocytopenia or

Q. Lin et al. bleeding disorders can have severe hemorrhage following injury or surgery. Similarly to BSS patients, gp9-deficient zebrafish also display a spontaneous bleeding syndrome, accompanied by a prolonged bleeding time after injury. Taken together, these results confirm that gp9 deficiency increases the tendency to bleeding, which is conserved from zebrafish to humans. The bleeding disorders in gp9deficient animals and patients could be attributed to insufficiently functional thrombocytes. The activation of thrombocytes may also be affected in GP9-mutated individuals, according to several reports.38,39 Taken together, these results suggested that zebrafish with inherited gp9deficiency, displaying a bleeding disorder resembling human BSS, could serve as a complementary model to broaden our understanding of the various roles of GP9 in thrombocytopoiesis, as well as in thrombosis and hemostasis. Based on the conservation of thrombocytopoiesis regulatory factors and strong similarity of GPIX across species, we verified that human GP9 could compensate zebrafish gp9, as human GP9WT mRNA is able to rescue thrombocytopenia in gp9SMU15 zebrafish. For functional validation of the BSS patient-derived mutations, previous studies were mainly restricted to cell lines.40,41 Here we provide an example of fast verification of patient-derived mutations using zebrafish, as we showed that gp9SMU15 zebrafish could serve as a cost-effective model for validating human GP9 mutations associated with BSS. The two GP9 mutations (hsGP970T>C and hsGP9182A>G) associated with BSS were evaluated by simply overexpressing them in the zebrafish model, and the failure to rescue thrombocytopenia confirmed that the two mutations are loss-of-function mutations in thrombocytopoiesis. Thus, the BSS zebrafish model with a high-throughput screening capability could help to accelerate the discovery of potential genetic variants of unknown clinical significance in vivo. The primary clinical treatment for BSS is supportive with platelet transfusions, anti-fibrinolytics particularly for mucocutaneous bleeds and recombinant activated factor VII in attempts to shorten bleeding times.42-44 More effective ways to treat BSS are much needed. Previous studies have shown that rhTPO rapidly increases platelet counts in patients with idiopathic thrombocytopenia purpura,45 and lowdose decitabine promotes thrombocytopoiesis in idiopathic thrombocytopenia purpura and healthy controls.46 Here, we showed that rhTPO and decitabine could efficiently relieve BSS phenotypes in zebrafish larvae (Figure 7). Given that the thrombocytopenia deficiency could be consistently observed in both larval and adult stages of zebrafish (Figures 2 and 4), we think that the two drugs may still relieve thrombocytopenia in adult BSS zebrafish if administered at a proper dose using a suitable way of drug delivery. The BSS zebrafish model will be beneficial for drug discovery, as evidenced by the demonstration that rhTPO and decitabine,

Haematologica | 107 July 2022

1666

ARTICLE - Modeling Bernard-Soulier syndrome in zebrafish used for promoting thrombocytopoiesis, also effectively relieved BSS phenotypes in the zebrafish larvae, which may also shed light on the treatment of BSS by promoting thrombocytopoiesis. In summary, we established a gp9-deficient zebrafish model displaying thrombocytopenia and a bleeding disorder, which resemble the clinical features of BSS in patients. We confirmed the conservation of the roles of GP9 in thrombocytopoiesis as well as in thrombosis and hemostasis from zebrafish to mammals. The BSS zebrafish model could be utilized as a cost-effective model to evaluate the function of human related mutations, and this line may also be applied in in vivo screening for novel drugs to treat BSS. The zebrafish BSS model, as a proofof-principle example, together with other established thrombocyte related zebrafish models, will allow us to examine the effect of clinically discovered mutations with uncertain clinical significance in thrombocyte development and function, as well as to conduct a robust drug evaluation or establish a screening platform to assess the therapeutic response of BSS, particularly against mutations with thrombocytopenia.

Q. Lin et al. Contributions YZ conceived and supervised the study and wrote, reviewed and edited the article; QL conceived the study, performed investigations and wrote, reviewed and edited the article; RZ and PM performed investigations, wrote, reviewed and edited the article; LW, LY, WL, JW, YC and LS performed investigations, analyzed the data and wrote the original draft of the manuscript. Acknowledgments The authors thank Dr. Jin Xu for his constructive suggestions and Mr. Xiaohui Chen for helping with the maintenance of zebrafish lines. Funding This work was supported by the National Natural Science Foundation of China (81870100 and 31871475), the National Key R&D Program of China (2018YFA0800200), Guangdong Province Universities and Colleges Pearl River Scholar Funded Scheme (2019), and Fundamental Research Funds for the Central Universities (2019ZD54 and 2018MS69). Data-sharing statement The data supporting the findings of this study are available within the article and its supplementary material.

Disclosures No conflicts of interest to disclose.

References 1. Nakao K, Angrist AA. Membrane surface specialization of blood platelet and megakaryocyte. Nature. 1968;217(5132):960-961. 2. Lordier L, Jalil A, Aurade F, et al. Megakaryocyte endomitosis is a failure of late cytokinesis related to defects in the contractile ring and Rho/Rock signaling. Blood. 2008;112(8):3164-3174. 3. Machlus KR, Italiano JJ. The incredible journey: from megakaryocyte development to platelet formation. J Cell Biol. 2013; 201(6):785-796. 4. Woolthuis CM, Park CY. Hematopoietic stem/progenitor cell commitment to the megakaryocyte lineage. Blood. 2016;127(10):1242-1248. 5. Bianchi E, Norfo R, Pennucci V, Zini R, Manfredini R. Genomic landscape of megakaryopoiesis and platelet function defects. Blood. 2016;127(10):1249. 6. Lanza F. Bernard-Soulier syndrome (hemorrhagiparous thrombocytic dystrophy). Orphanet J Rare Dis. 2006;1:46. 7. Bernard J, Soulier JP. Sur une nouvelle variété de dystrophie thrombocytaire-hémorragipare congénitale. Sem Hop Paris. 1948;24(Spec. No.):3217-3222 8. de la Salle C, Lanza F, Cazenave JP. Biochemical and molecular basis of Bernard-Soulier syndrome: a review. Nouv Rev Fr Hematol. 1995;37(4):215. 9. Lopez J, Andrews R, Afshar-Kharghan V, Berndt M. BernardSoulier syndrome. Blood. 1998;91:4397-4418. 10. Savoia A, Kunishima S, De Rocco D, et al. Spectrum of the mutations in Bernard-Soulier syndrome. Hum Mutat. 2014;35(9):1033-1045. 11. McEwan PA, Yang W, Carr KH, et al. Quaternary organization of GPIb-IX complex and insights into Bernard-Soulier syndrome revealed by the structures of GPIbβ and a GPIbβ/GPIX chimera.

Blood. 2011;118(19):5292-5301. 12. Clemetson JM, Kyrle PA, Brenner B, Clemetson KJ. Variant Bernard-Soulier syndrome associated with a homozygous mutation in the leucine-rich domain of glycoprotein IX. Blood. 1994;84(4):1124-1131. 13. Rivera CE, Villagra J, Riordan M, Williams S, Lindstrom KJ, Rick ME. Identification of a new mutation in platelet glycoprotein IX (GPIX) in a patient with Bernard–Soulier syndrome. Br J Haematol. 2001;112(1):105-108. 14. Dağıstan N, Kunishima S. First Turkish case of Bernard-Soulier syndrome associated with GPIX N45S. Acta Haematol. 2007;118(3):146-148. 15. Ali S, Ghosh K, Shetty S. Novel genetic abnormalities in Bernard-Soulier syndrome in India. Ann Hematol. 2014;93(3):381-384. 16. Meehan TF, Conte N, West DB, et al. Disease model discovery from 3,328 gene knockouts by The International Mouse Phenotyping Consortium. Nat Genet. 2017; 49(8):1231-1238. 17. Jagadeeswaran P, Liu YC, Sheehan JP. Analysis of hemostasis in the zebrafish. Methods Cell Biol. 1999;59:337-357. 18. Jagadeeswaran P, Sheehan JP, Craig FE, Troyer D. Identification and characterization of zebrafish thrombocytes. Br J Haematol. 1999;107(4):731-738. 19. Lin H-F, Traver D, Zhu H, et al. Analysis of thrombocyte development in CD41-GFP transgenic zebrafish. Blood. 2005;106(12):3803-3810. 20. Khandekar G, Kim S, Jagadeeswaran P. Zebrafish thrombocytes: functions and origins. Adv Hematol. 2012;2012:857058. 21. Lin Q, Zhang Y, Zhou R, et al. Establishment of a congenital amegakaryocytic thrombocytopenia model and a thrombocyte-

Haematologica | 107 July 2022

1667

ARTICLE - Modeling Bernard-Soulier syndrome in zebrafish specific reporter line in zebrafish. Leukemia. 2017;31(5):1206-1216. 22. Rost MS, Shestopalov I, Liu Y, et al. Nfe2 is dispensable for early but required for adult thrombocyte formation and function in zebrafish. Blood Adv. 2018;2(23):3418-3427. 23. Marconi C, Di Buduo CA, LeVine K, et al. Loss-of-function mutations in PTPRJ cause a new form of inherited thrombocytopenia. Blood. 2019;133(12):1346-1357. 24. Ma AC, Cheung AM, Ward AC, et al. The study of Jak2 V617F mutation in polycythemia vera with zebrafish model. Cell Res. 2008;18(S1):S141-S141. 25. Albers CA, Cvejic A, Favier R, et al. Exome sequencing identifies NBEAL2 as the causative gene for gray platelet syndrome. Nat Genet. 2011;43(8):735-737. 26. Gregory M, Hanumanthaiah R, Jagadeeswaran P. Genetic analysis of hemostasis and thrombosis using vascular occlusion. Blood Cells Mol Dis. 2002;29(3):286-295. 27. Kim S, Carrillo M, Kulkarni V, Jagadeeswaran P. Evolution of primary hemostasis in early vertebrates. PLoS One. 2009;4(12):e8403. 28. Deebani A, Iyer N, Raman R, Jagadeeswaran P. Effect of MS222 on hemostasis in zebrafish. J Am Assoc Lab Anim Sci. 2019;58(3):390-396. 29. Zheng L, Abdelgawwad MS, Zhang D, et al. Histone-induced thrombotic thrombocytopenic purpura in adamts13−/− zebrafish depends on von Willebrand factor. Haematologica. 2020;105(4):1107-1119. 30. Macaulay IC, Svensson V, Labalette C, et al. Single-cell RNAsequencing reveals a continuous spectrum of differentiation in hematopoietic Cells. Cell Rep. 2016;14(4):966-977. 31. Ma D, Zhang J, Lin HF, Italiano J, Handin RI. The identification and characterization of zebrafish hematopoietic stem cells. Blood. 2011;118(2):289-297. 32. Svoboda O, Stachura DL, Machoňová O, et al. Dissection of vertebrate hematopoiesis using zebrafish thrombopoietin. Blood. 2014;124(2):220-228. 33. Savoia A, Pastore A, De Rocco D, et al. Clinical and genetic aspects of Bernard-Soulier syndrome: searching for genotype/phenotype correlations. Haematologica. 2011;96(3):417-423. 34. Kissa K, Murayama E, Zapata A, et al. Live imaging of emerging hematopoietic stem cells and early thymus colonization. Blood. 2008;111(3):1147-1156.

Q. Lin et al. 35. Mo X, Nguyen NX, Mcewan PA, et al. Binding of platelet glycoprotein Ibβ through the convex surface of leucine-rich repeats domain of glycoprotein IX. J Thromb Haemost. 2009;7(9):1533-1540. 36. Jabbour E, Issa JP, Garcia-Manero G, Kantarjian H. Evolution of decitabine development: accomplishments, ongoing investigations, and future strategies. Cancer. 2008;112(11):2341-2351. 37. Zhou H, Qin P, Liu Q, et al. A prospective, multicenter study of low dose decitabine in adult patients with refractory immune thrombocytopenia. Am J Hematol. 2019;94(12):1374-1381. 38. Michelson AD, Benoit SE, Furman MI, Barnard MR, Nurden P, Nurden AT. The platelet surface expression of glycoprotein V is regulated by two independent mechanisms: proteolysis and a reversible cytoskeletal- mediated redistribution to the surface-connected canalicular system. Blood. 1996;87(4):1396-1408. 39. Canobbio I, Balduini C, Torti M. Signalling through the platelet glycoprotein Ib-V–IX complex. Cell Signal. 2004;16(12):1329-1344. 40. Zmajkovic J, Lundberg P, Nienhold R, et al. A gain-of-function mutation in EPO in familial erythrocytosis. N Engl J Med. 2018;378(10):924-930. 41. Kim AR, Ulirsch JC, Wilmes S, et al. Functional selectivity in cytokine signaling revealed through a pathogenic EPO mutation. Cell. 2017;168(6):1053. 42. Grainger JD, Thachil J, Will AM. How we treat the platelet glycoprotein defects; Glanzmann thrombasthenia and Bernard Soulier syndrome in children and adults. Br J Haematol. 2018;182(5):621-632. 43. Almeida AM, Khair K, Hann I, Liesner R. The use of recombinant factor VIIa in children with inherited platelet function disorders. Br J Haematol. 2003;121(3):477-481. 44. Ozelo MC, Svirin P, Larina L. Use of recombinant factor VIIa in the management of severe bleeding episodes in patients with Bernard-Soulier syndrome. Ann Hematol. 2005;84(12):816-822. 45. Wang S, Yang R, Zou P, et al. A multicenter randomized controlled trial of recombinant human thrombopoietin treatment in patients with primary immune thrombocytopenia. Int J Hematol. 2012;96(2):222-228. 46. Zhou H, Hou Y, Liu X, et al. Low-dose decitabine promotes megakaryocyte maturation and platelet production in healthy controls and immune thrombocytopenia. Thromb Haemost. 2015;113(5):1021-1034.

Haematologica | 107 July 2022

1668

ARTICLE - Platelet Biology and its Disorders

Single platelet and megakaryocyte morpho-dynamics uncovered by multicolor reporter mouse strains in vitro and in vivo Leo Nicolai,1,2 Rainer Kaiser,1,2 Raphael Escaig,1,2 Marie-Louise Hoffknecht,1,2 Afra Anjum,1,2 Alexander Leunig,1,2 Joachim Pircher,1,2 Andreas Ehrlich,1 Michael Lorenz,1 Hellen IshikawaAnkerhold,1 William C. Aird,3 Steffen Massberg1,2 and Florian Gaertner4 Department of Medicine I, University Hospital, LMU Munich, Munich, Germany; 2German Center for Cardiovascular Research (DZHK), partner site Munich Heart Alliance, Munich, Germany; 3Department of Medicine, Center for Vascular Biology Research, Beth Israel Deaconess Medical Center, Boston, MA, USA and 4Institute of Science and Technology (IST) Austria, Klosterneuburg, Austria

Correspondence: Leo Nicolai leo.nicolai@med.uni-muenchen.de

Florian Gaertner florian.gaertner@ist.ac.at Received: April 9, 2021. Accepted: September 9, 2021. Prepublished: September 16, 2021. https://doi.org/10.3324/haematol.2021.278896 ©2022 Ferrata Storti Foundation Haematologica material is published under a CC BY-NC license

Abstract Visualizing cell behavior and effector function on a single cell level has been crucial for understanding key aspects of

mammalian biology. Due to their small size, large number and rapid recruitment into thrombi, there is a lack of data on fate and behavior of individual platelets in thrombosis and hemostasis. Here we report the use of platelet lineage restricted multi-color reporter mouse strains to delineate platelet function on a single cell level. We show that genetic labeling allows for single platelet and megakaryocyte (MK) tracking and morphological analysis in vivo and in vitro, while not affecting lineage functions. Using Cre-driven Confetti expression, we provide insights into temporal gene expression patterns as well as spatial clustering of MK in the bone marrow. In the vasculature, shape analysis of activated platelets recruited to thrombi identifies ubiquitous filopodia formation with no evidence of lamellipodia formation. Single cell tracking in complex thrombi reveals prominent myosin-dependent motility of platelets and highlights thrombus formation as a highly dynamic process amenable to modification and intervention of the acto-myosin cytoskeleton. Platelet function assays combining flow cytrometry, as well as in vivo, ex vivo and in vitro imaging show unaltered platelet functions of multicolor reporter mice compared to wild-type controls. In conclusion, platelet lineage multicolor reporter mice prove useful in furthering our understanding of platelet and MK biology on a single cell level.

Introduction Platelets are anucleate cells of 1-2 µm diameter and are the second most abundant blood cell type. They are rapidly recruited upon vascular injury, undergo an adhesion and activation cascade and subsequently form a hemostatic plug. This process is highly coordinated and tightly interconnected with the coagulation system.1 In atherosclerosis, erosion or rupture of plaques leads to unwanted platelet recruitment and thrombus formation, causing vessel occlusion and ischemia.2 In vivo imaging of murine thrombosis models has considerably advanced our understanding of this process.3 For example, recent work has contributed to understanding thrombus architecture, defining layers of differently activated platelets in growing thrombi.4 Fluorescent tracking of key components like tissue factor, procoagu-

lant surface and thrombin generation have shed light on thrombus formation in detail.4-6 Nevertheless, tracking and analysis of individual platelets has remained elusive because of their small size, large number and rapid recruitment.7 Platelets originate from megakaryocytes (MK) in the bone marrow, spleen and lung.8,9 Maturation of MK into platelet-producing cells occurs through differentiation from hematopoietic stem cells (HSC) to multipotent progenitor, common myeloid progenitor, MK-erythroid progenitor, and MK progenitor.10 Upon final differentiation, MK progenitors undergo DNA replication without cell division, a process termed endomitosis.11 This process leads to accumulation of DNA content of 4n, 8n, 16n, 32n, 64n, and even up to 128n in a single polylobulated nucleus.11,12 This process results in functional gene amplification most likely necessary for a ramp-up in protein synthesis

Haematologica | 107 July 2022

1669

ARTICLE - Platelet and MK function on the single cell level necessary for platelet production.13 Genetic targeting after/during endomitosis can lead to stochastic recombination events in each set of chromosomes, leading to distinct genotypes within one cell. In a final step, MK start to produce proplatelets in a shear-dependent manner by protrusion of cytoplasmic extensions into the blood flow. Despite their large size, MK turnover, behavior and positioning in the bone marrow niche are not well understood, warranting in vivo and in vitro imaging approaches.14-16 Multicolor reporter mouse strains have proven to be very useful in understanding cellular dynamics and cell fate in a wide range of cell types in vivo.17-21 The Brainbow imaging technique, originally developed to visualize and distinguish individual neurons and their fine axonal processes in the brain, allowed for the first-time multicolor tracking of individual cells.22 Recently, a Cre-reporter mouse, termed R26R-Confetti was developed to achieve tissue-specific Brainbow expression.23 Here we report the use of a platelet lineage restricted R26R-Confetti multicolor reporter mouse strain which allows for single platelet and MK tracking and morphological analysis in vivo and in vitro, while not affecting lineage functions.

Methods Mouse strains PF4-Cre24 and Rosa26-Confetti23 were purchased from The Jackson Laboratory and maintained and cross-bred at our animal facility (stock no: 008535 and 017492). vWFCre mice were a gift of W. Aird.25 Mhy9 fl/fl mice were a gift from Dr. Gachet.26 All strains were backcrossed to and maintained on C57BL/6J-background. If not otherwise stated, animals of same sex and age were randomly assigned to experimental groups, and PF4-Cre negative control animals were used. We used Rosa26Confetti/+ animals expressing Cre for Online Supplementary Figure S1. The other experiments were performed with homozygous Rosa26Confetti/Confetti mice expressing Cre. Cre negative animals were used as controls. All procedures performed on mice were approved by the local legislation on the protection of animals (Regierung von Oberbayern, Munich). Confocal intravital microscopy of the mesentery After induction of narcosis, animals received 15 µg of X488 antibody (Emfret) and 80 µg of Fibrinogen-AlexaFluor594 conjugate to visualize platelets and fibrin(ogen), respectively. Confocal microscopy of the mesentery was performed as previously described.27 Briefly, the skin and peritoneum were opened along the midline, and 200 µL of warm phosphate-buffered saline was pipetted into the abdominal cavity. Next, the prox-

L. Nicolai et al. imal bowel was carefully exteriorized onto a glass coverslip and the mesenteric vessels was exposed. Wet tissue paper was used to stabilize the bowel and prevent artifacts from bowel movements. A 1x1 mm piece of filter paper was saturated with 10% ferric chloride solution and was carefully placed in direct contact with a mesenteric vein for 3 minutes (min). Next, the ensuing thrombus formation was imaged on an inverted Zeiss LSM 880 in AiryScan Fast Mode (20x/0.8 obj., 990 ms/frame, laser power: 0.96%). Bone marrow whole mount stains Bone marrow from PF4cre-Confetti animals was harvested and processed as described previously. For whole mounts, bones were snap frozen after paraformaldehyde (PFA) fixation (4%, 60 min) and sucrose treatment (30%, overnight). Bones were then cut longitudinally with a cryotome to expose marrow, and were stained with CD41-AF647, (1:100), CD144-AF647 (1:100), FarRed Nuclear Stain (1:1000) and were imaged with a Zeiss LSM 880 confocal microscope (20x/0.8 obj.) in AiryScan mode. Static thrombus formation Static thrombi were generated by activating platelet-rich plasma (PRP) with thrombin (0.1 U) for 5 min, dilution of these thrombi 1:10 and assessment of platelets was done with a LSM 880 confocal microscope in AiryScan mode. Histology For mesentery, lung and cremaster muscle histology, organs were harvested and kept in 1% PFA for 1 hour, before transfer in 20% sucrose overnight. After embedding in TissueTek OCT mounting medium (Sakura), organs were snap frozen and cut using a cryotome (10 µm slices). Staining was performed using antibodies as indicated. Flow chamber experiments As previously described,28 heparinized whole blood was perfused over a collagen coated surface at arterial shear rate (1,000/s) for 5 min. Formed thrombi were assessed on a LSM 880 confocal microscope in AiryScan mode (20x/0.8 obj., 2x zoom for overviews, and 5x zoom for detailed views, laser power: 0.96%). Data analysis Data analysis was realized using FIJI (ImageJ) and ZenBlack (Zeiss). 4D in vivo data was stabilized with the “ImageStabilizer” plugin and dimensions were reduced by maximum intensity projection (MIP). Cells were tracked with “Manual tracking” FIJI plugins. For secondary analysis, IBIDI Chemotaxis tool was used. For form analysis, cells were manually outlined, masked and converted to binary images. Solidity also known as convexity, was computed as the proportion of the pixels in the convex

Haematologica | 107 July 2022

1670

ARTICLE - Platelet and MK function on the single cell level hull that are also in the object (Area/ConvexArea). Aspect ratio as a maker of polarization was computed by major_axis divided by minor_axis. Number of filopodia were quantified manually. For quantification, FIJI shape descriptors were used. Data were post-processed in Excel (Microsoft), Prism (GraphPad) and Illustrator (Adobe). Statistical analysis and reproducibility Data are shown as means ± standard error of the mean (SEM). Statistical parameters, including the exact value of replicates (‘‘n’’) for individual experiments can be found within the figure legends. In order to evaluate statistical differences between groups, t-tests and analysis of variance (ANOVA) were performed. A P value of <0.05 was considered statistically significant. Analyses were performed with Prism (GraphPad Software) and Excel (Microsoft).

Results A multicolor platelet lineage reporter mouse Livet et al. introduced the Cre-recombinase dependent Brainbow 2.1 construct (loxP-STOP-loxP-GFP-PFY-PxolloxP-RFP-PFC-Pxol) to allow genetic fluorescent labeling of neuronal networks.22 The construct consists of two color tandems, yellow fluorescent protein (YFP) and inverted green fluorescent protein (GFP) and red green fluorescent protein (RFP) and inverted cyan green fluorescent protein (CFP), creating YFP-, GFP-, RFP- or CFP-positive populations (Figure 1A). GFP is expressed in the nucleus, CFP is membrane bound, whereas YFP and RFP show cytosolic expression. Flanking loxP sites enable Cre driven recombination, leading to stochastically driven removal and/or inversion of tandem constructs. Repeated/ongoing expression of the Cre recombinase can lead to expression of the anti-sense fluorescent protein by inversion of the tandem after initial recombination. Snippert et al. recently developed a Cre-reporter mouse, termed R26R-Confetti-stop-flox, allowing tissue specific Brainbow 2.1 expression under the control of a CreDeleter.23 We crossed R26R-Confetti-stop-flox mice to a platelet factor 4 (PF4)-Cre transgenic mice to limit recombination events to the platelet/MK lineage (Figure 1A).24 Confocal microscopy of bone marrow whole mounts revealed large, brightly labeled cells that were positive for platelet/MK marker Gp1b, confirming that these cells were indeed MK (Figure 1B; Online Supplementary Figure S1A). Also, we detected proplatelet formation in a subpopulation of Confetti-expressing cells indicating intact thrombopoiesis (Figure 1B). Multicolor expression enabled precise size and shape analyses of labeled MK (Figure 1C, D). 29

L. Nicolai et al. Confetti expression in polyploid megakaryocytes generates unique color patterns and reveals late expression of platelet factor 4 in the megakaryocyte lineage In contrast to previously reported Confetti-expression patterns in other cell types showing a single XFP expressed per cell, individual MK expressed one up to all four possible XFP (Figure 1E).19,23 The majority of MK expressed two to three XFP (Figure 1F). MK are unique as they undergo endomitosis during their maturation. Therefore, genetic targeting after endomitosis can theoretically lead to multiple genotypes in one cell. This could explain the recombination of multiple Confetti-constructs in PF4crepositive MK. Indeed, Confetti-positive cells contained multilobulated nuclei and correlation analysis of color expression to MK size revealed that color expression diversity increased with cell size, a parameter that strongly correlates with ploidy and DNA content (Online Supplementary Figure S1B; Figure 1G, H).30 In order to confirm this observation, we generated von Willebrand factor (vWF)Cre; R26RConfetti/+ mice. vWF is expressed early in the MK lineage, at the level of platelet-biased stem cells, and should therefore lead to recombination prior to endomitosis - and thus also prior to polyploidization.31 Indeed, in contrast to PF4-Cre; R26RConfetti/+ mice, vWF-Cre; R26RConfetti/+ MK expressed either one color (mainly YFP) or two colors (RFP and CFP) (Figure 2A-C). The simultaneous coexpression of RFP and CFP is possible since their two open reading frames are positioned in a head-to-tail tandem dimer and are therefore able to permanently flip during megakaryopoiesis due to a persistent Cre recombinase expression under the control of the PF4 or vWF promoter.22 We confirmed this finding in peripheral blood platelets of these mice (Figure 3A-C). Together, our data provide new insights into the expression pattern of the PF4-Cre transgene in the MK lineage: robust expression in MK seems to occur rather late in MK development, with important implications regarding the use of PF4-Cre in studying MK biology. As reported previously by others, we also detected PF4-Cre-driven expression of XFP in cells beyond the MK lineage for example in the mesenterial fat and cremaster muscle tissue (Online Supplementary Figure S1C, D).32 These cells appear to be mainly of macrophage origin as evidenced by their morphology.33 Analysis of the megakaryocyte niche in Confetti mice reveals clusters of mature megakaryocytes The MK bone marrow niche is still only partially understood, and analysis/differentiation of single or clustered MK is hampered by their unconventional shape in different states of thrombopoiesis, forming long protrusions or extending pro-platelets. Gp1b staining revealed multiple MK that could not be clearly defined as single or clustered cells (Figure 2D). The Confetti signal enabled separation of

Haematologica | 107 July 2022

1671

ARTICLE - Platelet and MK function on the single cell level single and clustered MK, and unexpectedly revealed that almost 50% of MK were found in clusters of two or more cells (Figure 2D, E). Large, mature MK (PF4+) were found in close vicinity to CD144+ blood vessels (Figure 2F). In contrast, small cells expressing vWF-Cre-driven Confetti (a mixed cell population containing both immature MK progenitors and non-MK cells) localized further away from vessels, highlighting the preferential localization of mature MK near vessels (Figure 2F).34 In conclusion, multicolor reporter mice may provide a useful tool to investigate clonality and expansion of the MK within the bone marrow niche.

L. Nicolai et al. Exclusion of functional effects on the megakaryocyte/platelet lineage in PFR4-Cre-Confetti mice Confocal imaging of isolated platelets from PF4-Cre-Rs26Confetti mice revealed multicolor expression in YFP, RFP and CFP, but no nGFP labeling, as this fluorescent protein is restricted to the nucleus (Figure 3D). Next, we excluded effects of this construct on the function of the platelet/MK lineage as this would affect generalization of findings drawn from using this mouse line. Platelet counts, mean platelet volume and hemoglobin content in peripheral blood did not show significant changes between Cre+ and

Figure 1. Rs26-Confetti as a tool to study megakaryocytes in vivo. (A) Rs26-Confetti gene product and Cre-dependent recombination; (B) representative endogenous Confetti fluorescence (YFP, RFP, GFP and CFP) and Gp1b (AlexaFluor647) staining of a bone marrow whole mount. *Indicates pro-platelet formation; (C and D) size and shape analysis of megakaryocytes (MK) in bone marrow whole mounts, n=93 cells in 8 bones; (E) representative MK expressing 1-4 colors; (F) percentage of MK expressing 1,2,3 or 4 colors, n=8 bones, ANOVA across color expression groups; (G to H) MK size related to color expression (1, 2, 3, 4 colors), n=93 cells in 8 bones; (G) ANOVA across color expression groups; (H) correlation analysis using linear regression. YFP: yellow fluorescent protein; GFP: green fluorescent protein; RFP: red green fluorescent protein; CFP: cyan green fluorescent protein. Haematologica | 107 July 2022

1672

ARTICLE - Platelet and MK function on the single cell level Cre- animals (Online Supplementary Figure S2A). Flow cytometric analysis of isolated platelets confirmed strong fluorescence in PE and FITC channels in Cre+ mice, which was absent in Cre- mice (Online Supplementary Figure S2B), while morphology was unaltered (Online Supplementary Figure S2C). Flow cytometric assessment of quiescent platelets showed no alterations in size, surface expression of Gp1b and GpIIbIIIa, activation status (P-selectin expression) or fibrinogen binding in response to activation (Online Supplementary Figure S2D). Upon stimulation with ADP, collagen or thrombin, markers of activation did not differ significantly between Cre+ animals and Cre- littermates (Online Supplementary Figure S2E-G). Efficiency of

L. Nicolai et al. platelet activation was independent of XFP-expression as gating on XFP (YFP+, RFP+, and YFP+RFP+) revealed no difference in fibrinogen binding upon thrombin stimulation (Online Supplementary Figure S2H). Next, we assessed hemostatic function in reporter and littermate mice. Time to initial hemostasis and total bleeding time were unaltered in tail bleeding assays (Online Supplementary Figure S3A). In vitro assessment of clot retraction revealed a similar degree of retraction (Online Supplementary Figure S3B). A ferric-chloride induced thrombosis model of the carotid artery revealed similar thrombus formation kinetics in Cre+ and Cre- animals (Online Supplementary Figure S3C; Online

Figure 2. Rs26-Confetti expression in the platelet lineage reveals megakaryocyte clustering in the bone marrow. (A-C) Analysis of megakaryocytes (MK) in bone marrow whole mounts from von Willebrand factor (vWF)-Cre and PF4-Cre dependent Rs26-Confetti reporter mice. (A) Representative micrographs showing YFP, CFP, RFP and GFP overlay, arrows: MK, scale bar: 50 µm. Of note, vWF is expressed early during megkaryopoiesis in pluripotent platelet biased hematopoietic stem cells, leading to vWF-Credriven expression of Confetti in a large proportion of small, multi-lineage bone marrow cells in addition to large MK. (B) Quantification of color expression n=3 mice per group, t-test; (C) detailed analysis of color expression. (D) Confetti fluorescence (YFP, RFP, GFP and CFP) and Gp1b (DyLight647) staining of representative single/grouped cells. Image was cropped from Figure 1B and shown at higher magnification. Compare Figure 1B and E percentage of single and grouped MK, n=10 bones, ANOVA. (F) Mean distance of Confetti positive cells in the bone marrow of PF4-cre (n=50) and vWF-cre (n=157) dependent Rs26-Confetti reporter mice from CD144+ blood vessels. Analyzed in n=3 bones from individual mice per group, t-test. YFP: yellow fluorescent protein; GFP: green fluorescent protein; RFP: red green fluorescent protein; CFP: cyan green fluorescent protein. Haematologica | 107 July 2022

1673

ARTICLE - Platelet and MK function on the single cell level

L. Nicolai et al.

Figure 3. Rs26-Confetti expression in platelets. (A-C) Analysis of peripheral blood platelets of von Wllebrand factor (vWF)-Cre and PF4-Cre dependent Rs26-Confetti reporter mice. (A) Representative micrographs of spread platelets showing YFP, CFP, RFP overlay; (B) quantification of color expression n=3 mice per group, t-tests; (C) detailed analysis of color expression. (D) Fluorescence expression in PF4-Cre-Rs26-Confetti platelets. Scale bars= 5 µm. YFP: yellow fluorescent protein; GFP: green fluorescent protein; RFP: red green fluorescent protein; CFP: cyan green fluorescent protein.

Supplementary Video S1). Simultaneous visualization of XFP (RFP and YFP) and intravenously (i.v.) injected plateletlabeling antibody (Far Red; Dylight 649) revealed the same platelet recruitment pattern indicating that the endogenous fluorescence signal in Cre+ mice, which was absent in Cre- mice allowed reliable tracking of thrombus formation in vivo (Online Supplementary Figure S3D). In addition, similar amounts of isolated platelets were recruited to collagen and fibrinogen coated surfaces and showed comparable spreading and lamellipodium formation (Online Supplementary Figure S3E-H). Analysis of multicolor platelets in reporter mice Reporter platelets spreading on fibrinogen revealed homogenous distribution of fluorescence signal in individual cells (Figure 4A). As expected from our data on fluorescence expression in MK, platelets revealed a range of colors depending on (i) the expression of one to three XFP and (ii) intensity of expression of each XFP (Figure 4A). Mapping color and size, aspect ratio (marker of polarization) and solidity (marker of irregular shape) and number of filopodia did not reveal overt clustering of cells dependent on color (Figure 4B-E). Similarly, analysis based on expressed color (YFP, RFP and CFP) did not reveal differences in size or shape or migration behavior (Online Supplementary Figure S4A-C).

Morphology analysis of platelets within thrombi shows prominent filopodia formation Thrombosis and hemostasis trigger rapid recruitment of activated platelets in large numbers and therefore complicate single cell tracking. In order to test if multi-color labeling by R26R-Confetti allows segmentation of individual platelets within the cell-packed environment of a thrombus we generated fibrin-rich platelet thrombi under static conditions in vitro. Confocal microscopy revealed reorganization of the fibrin network by platelets, leading to tight platelet-fibrin aggregates (Figure 4F). Stochastic expression of XFP in R26R-Confetti platelets still allowed identification of individual platelets in dense aggregates (Figure 4G). Next, we generated static microthrombi from isolated Confetti platelets (Figure 5A). Confocal analysis enabled us to discriminate single platelets in these tightly packed aggregates, as well as protrusions formed by individual cells classified as filopodia (Online Supplementary Video S2). Recruited platelets did not form discernible sheet-like lamellipodia (Figure 5A). In addition, the morphology of the individual cells could be quantified and shape descriptors of aggregated platelets were determined. Over 80% of recruited cells formed filopodia of various lengths (Figure 5B). Flow chamber-based assays are an important tool to

Haematologica | 107 July 2022

1674

ARTICLE - Platelet and MK function on the single cell level

L. Nicolai et al.

Figure 4. Rs26-Confetti expression in activated platelets in vitro. (A-E) Spreading of Rs26-Confetti platelets on fibrinogen. (A) Representative micrograph; (B) size; (C), aspect ratio and (D) solidity in relation to color expression, n=48 cells; (E) filopodia formation. (F, G) In vitro generated fibrin clots using Rs26-Confetti platelet-rich plasma, (G) revealing cell aggregates.

study thrombus formation in vitro. By using heparinized blood from R26R-Confetti reporter mice, we were able to visualize multicolor aggregate formation under blood shear conditions, and higher magnification revealed single platelet morphology (Figure 5C, D). Similar to static thrombi generated from isolated platelets, the majority of recruited platelets formed filopodia, but no lamellipodia (Figure 5D, E). In vitro approaches are limited when studying complex biological events. Therefore, we assessed the applicability of platelet R26R-Confetti expression in vivo. In order to study thrombus formation, we utilized a ferric-chloride induced mesenteric thrombosis model. Similar to our in vitro findings, RFP, YFP and CFP fluorescence expression allowed for intravital discrimination of individual cells in growing thrombi (Figure 6A; Online Supplementary Video S3). Moreover, analysis of single platelet morphology revealed formation of filopodia of the majority of recruited individual platelets (Figure 6B).

Single cell tracking of individual cells in complex thrombi reveals myosin dependent motility in vivo and in vitro For a long time, activated recruited platelets were considered stationary cells. Recently, platelets were shown to actively migrate at sites of vascular injury.7,35 In addition, myosin IIa-dependent pulling forces trigger collective platelet motion in a process termed clot retraction.26 We utilized the Confetti model to track individual platelets in static microthrombi over time using time lapse microscopy and identified single motile platelets (Figure 6C; Online Supplementary Video S2). Similarly, single cell tracking in vivo showed relocation dynamics of individual platelets in complex thrombi (Figure 6D). Particularly in the consolidation phase of thrombi, collective motility of recruited platelets was observed (Online Supplementary Video S3). We tracked >100 platelets in a single thrombus, which on the one hand revealed platelet repositioning in all parts of the thrombus. On the other hand, the observed motility patterns were

Haematologica | 107 July 2022

1675

ARTICLE - Platelet and MK function on the single cell level

L. Nicolai et al.

Figure 5. Rs26-Confetti enables studying of platelet dynamics on a single cell level in vitro. (A, B) In vitro generated static platelet thrombi. (A) Representative micrograph and (B) shape descriptors and filopodia of single platelets; n=33 cells in two thrombi. (C-E) Flow chamber experiments with arterial shear rate (1,000/s) over collagen; (C, D) representative micrographs and (E) shape descriptors and filopodia of single platelets; n=33 cells in two thrombi; n=71 cells from three micrographs.

highly heterogenous, as accumulated distance, velocity and directionality varied between cells (Figure 6E). Finally, we bred PF4cre; R26RConfetti mice to Mhy9-floxed mice to generate a multicolor reporter mouse of myosin heavy chain deficient platelets.26 Mhy9-disorders are characterized by increased platelet size and reduced number – macrothrombocytopenia. In vitro, Mhy9-deficient platelets spread normally with no apparent morphological difference from control platelets (Online Supplementary Figure S4D). By intravital microscopy of injured mesenteric

vessels in PF4cre; R26RConfetti; Myh9flox mice we were able to perform size measurements of individual platelets in vivo confirming enlarged Mhy9-/- platelets (Figure 6F). Single cell tracking within thrombi revealed reduced velocity and accumulated distances of Myh9-deficient platelets compared to control mice (Online Supplementary Figure S4E).26 The associated lack of collective contractile behavior led to increased embolization and less consolidation of the thrombus as the clot was not retracted (Online Supplementary Video S4; Figure 6F).

Haematologica | 107 July 2022

1676

ARTICLE - Platelet and MK function on the single cell level

L. Nicolai et al.

Figure 6. Rs26-Confetti enables studying of platelet dynamics on a single cell level in vivo. (A, B) In vivo FeCl3-dependent thrombus formation in the mesentery. (A) Representative micrograph and (B) shape descriptors and filopodia of single platelets; n=48 cells in two thrombi. (C) Tracking of individual platelets over time (arrows) in static in vitro thrombi. Micrograph shows higher magnification of a region of interest cropped from (A). (D, E) Tracking of thrombus recruited platelets in vivo; (D) exemplary micrograph of individual motile platelet (arrow) and (E) tracking of n=118 individual platelets in single thrombus allows for computation of single cell velocity, covered distance and directionality. (F, left) Size differences of control and Mhy9-deficient platelets on the Rs26-Confetti background are evident in vivo and (F, right) differences of individual platelet behavior in Mhy9deficient animals in Fe Cl3-dependent thrombus formation in the mesentery. Arrow marks exemplary platelet moving through thrombus with flow, unable to retract. (G) Visualization of CD45+ leukocytes in Rs26-Confetti thrombi. Scale bar =10 mm. Haematologica | 107 July 2022

1677

ARTICLE - Platelet and MK function on the single cell level Thrombus formation is a complex process involving the concerted action of additional cell types, for example neutrophils and monocytes. It is therefore important to be able to label additional cell populations in deployed fluorescence reporter strains, and to exclude passive platelet motility patterns depending on leukocyte migration.35 We injected CD45 antibody conjugated to a Far Red emitter (Alexa Fluor 647) to stain circulating leukocytes. Excitation/emission spectra revealed no overlap with Confetti signal and enabled side-by-side visualization of leukocytes and platelets in multicolor thrombi (Figure 6G; Online Supplementary Figure S4F).

Discussion Here, we describe in detail a novel mouse model that can be used to assess single cell dynamics of the platelet/MK lineage. We previously utilized this model to study single cell behavior mainly under inflammatory conditions in vivo, which helped to reveal the platelet migratory phenotype.7 We further extended our understanding of the employed mouse model, and ruled out a major effect of Confetti fluorescence expression/ Cre recombination on platelet and MK biology. In addition to tracking of individual cells, this model allows to discern cell shape in vivo and in vitro, which is important for understanding platelet effector functions.35 We demonstrate the versatility of this model by showing that it also allows for additional staining of populations/antigens using far red emitters. We substantiate the relevance of the reporter model for a better understanding of the biology of MK and platelets by revealing previously unkown aspects of this lineage. First, we identified pairs and clusters of MK in the bone marrow. These clusters may indicate clonal expansion of MK precursors within the MK-specific bone marrow niche, where MK precursors proliferate before differentiating and forming platelets.31,34 Additional work will be required to resolve megakaryopoiesis spatially and temporally in greater detail. PF4 is expressed in mature MK where it has been shown to act as negative regulator of megakarypoesis and hematopoiesis.15,36 Thus, late PF4 expression possibly serves as a paracrine negative feedback loop to control MK differentiation. Expression of multiple XFP following PF4-Cre driven recombination of the Confetti construct supports the notion that PF4 is mainly expressed after initiatiation of MK endomitosis in mature MK. Therefore, it might serve as a useful late genetic marker for studying megakaryopoesis, however, with the drawback of potenial off-target effects in monocytic cells.32,33 In contrast, using the vWF promotor for specific and robust Cre expression in early MK progenitors, might prove useful in studying the con-

L. Nicolai et al. tribution of cytoskeletal proteins or transcription factors in early megakaryopoiesis. Despite great progress in recent years, it remains difficult to assess individual platelet behavior in vivo.37 Especially data on platelet cytoskeletal rearrangements during activation and recruitment are almost exclusively derived from in vitro studies. For example, platelet lamellipodia formation, termed “spreading”, is a key element of in vitro assessment of platelet function.38 However, there is no evidence that platelets form lamellipodia in hemostatic plugs in vivo.39 Nesbitt et al. have demonstrated that platelets might not undergo dramatic shape changes in thrombosis.40 Recently, using mouse lines in which lamellipodia formation is disrupted, we identified lamellipodia as a key cytoskeletal feature of immune responsive platelets with a neglible role in classical thrombosis and hemostasis.35 Along these lines, we here consolidate these data by excluding lamellipodia formation but showing prominent filopodia formation of activated platelets in three different thrombosis models in vivo and in vitro. These needle-like protrusions might serve as anchor points as well as sensors registering the microenviroment.41 In addition, they might increase platelet surface area and redistribute adhesion receptors,42 affecting the platelets propensity to activate. This fits very well with data showing normal hemostasis and thrombosis in mice unable to form lamellipodia.35,39,43,44 As interference with the cytoskeleton might prove an interesting pharmacological target, our data highlight the need to further investigate the regulation of filopodia formation in platelets. Regarding platelet motility within complex thrombi, our experiments show ubiquitous and global repositioning of platelets in thrombus formation. Single cell tracking in our model revealed a collective motion of platelets towards the thrombus core largely depending on force generation by myosinIIa, as genetic ablation of Mhy9 drastically alters clot formation and reduces platelet re-location, confiming previous reports.26 In addition to consolidation of the clot, the centripedal, collective motion of platelets may generate forces required for the mechanical extrusion of procoagulant platelets to reinforce thrombin generation and fibrin formation at the thrombus surface as recently proposed by others.45 Thus, multicolor tracking of platelets highlights thrombus formation as a highly dynamic process amenable to modification and intervention of the acto-myosin cytoskeleton. Combination of platelet-specific multicolor reporter mouse lines with genetic/pharmacological targeting and labeling of neutrophils or other immune cells might help to better understand clot initiation, consolidation and lysis. Future studies could take advantage of this model to understand single platelet and MK dynamics in a wide range of disease models and conditions.

Haematologica | 107 July 2022

1678

ARTICLE - Platelet and MK function on the single cell level Disclosures No conflicts of interest to disclose. Contibutions LN and FG initiated the study; LN conceived the study design; LN, RK, JP, HI-A and FG developed the study methodology; LN, RK, RE, M-LH, AA, AL, JP, AE, ML, HI-A and FG carried out the investigation; WA, SM and FG provided resources; LN, RK and RE performed the final analysis; LN visualized the study; LN and FG supervised the study; LN performed project administration; LN, FG and SM acquired funding; LN wrote the original draft. All authors edited and approved the final version of the manuscript.

L. Nicolai et al. Funding This study was supported by the Deutsche Forschungsgemeinschaft (DFG) SFB 914 ( to SM [B02 and Z01]), the DFG SFB 1123 (to SM [B06]), the DFG FOR 2033 (to SM), the German Center for Cardiovascular Research (DZHK) (Clinician Scientist Programme), MHA 1.4VD (to SM), Postdoc Start-up Grant, 81X3600213 (to FG), 81X3600222 (to LN), the FP7 program (project 260309, PRESTIGE [to SM]). This project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No. 83344, ERC-2018-ADG “IMMUNOTHROMBOSIS” [to SM] and the Marie Skłodowska Curie Individual Fellowship (EU project 747687, LamelliActin [to FG]).

Acknowledgments Data-sharing statement The authors thank Anna Titova, Sebastian Helmer, Nicole Original data can be made available on reasonable request Blount and Beate Jantz for technical assistance. to the authors.

References 1. Versteeg HH, Heemskerk JWM, Levi M, Reitsma PH. New fundamentals in hemostasis. Physiol Rev. 2013;93(1):327-358. 2. Jackson SP. Arterial thrombosis--insidious, unpredictable and deadly. Nat Med. 2011;17(11):1423-1436. 3. Furie B, Furie BC. Mechanisms of thrombus formation. N Engl J Med. 2008;359(9):938-949. 4. Stalker TJT, Wu E, Wannemacher J, et al. Hierarchical organization in the hemostatic response and its relationship to the platelet-signaling network. Blood. 2013;121(10):1875-1885. 5. Stalker TJ, Welsh JD, Brass LF. Shaping the platelet response to vascular injury. Curr Opin Hematol. 2014;21(5):410-417. 6. Armstrong PC, Hoefer T, Knowles RB, et al. Newly formed reticulated platelets undermine pharmacokinetically short-lived antiplatelet therapies. Arterioscler Thromb Vasc Biol. 2017;37(5):949-956. 7. Gaertner F, Ahmad Z, Rosenberger G, et al. Migrating platelets are mechano-scavengers that collect and bundle bacteria. Cell. 2017;171(6):1368-1382.e23. 8. Lefrançais E, Ortiz-Muñoz G, Caudrillier A, et al. The lung is a site of platelet biogenesis and a reservoir for haematopoietic progenitors. Nature. 2017;544(7648):105-109. 9. Junt T, Schulze H, Chen Z, et al. Dynamic visualization of thrombopoiesis within bone marrow. Science. 2007;317(5845):1767-1770. 10. Haas S, Hansson J, Klimmeck D, et al. Inflammation-induced emergency megakaryopoiesis driven by hematopoietic stem cell- like megakaryocyte progenitors.. Cell Stem Cell. 2015;17(4):422-434. 11. Machlus KR, Italiano JE, Jr. The incredible journey: from megakaryocyte development to platelet formation. J Cell Biol. 2013;201(6):785-796. 12. Zimmet J, Ravid K. Polyploidy: occurrence in nature, mechanisms, and significance for the megakaryocyte-platelet system. Exp Hematol. 2000;28(1):3-16. 13. Machlus KR, Italiano JE. 2 - Megakaryocyte development and platelet formation. In: Michelson AD, editor. Platelets (Fourth Edition). Academic Press; 2019. p. 25-46. 14. Italiano JE. Megakaryopoiesis and Platelet Biogenesis. Molecular and Cellular Biology of Platelet Formation. Springer; 2016. p. 3-22.

15. Bruns I, Lucas D, Pinho S, et al. Megakaryocytes regulate hematopoietic stem cell quiescence through CXCL4 secretion. Nat Med. 2014;20(11):1315-1320. 16. Zhang L, Urtz N, Gaertner F, et al. Sphingosine kinase 2 (Sphk2) regulates platelet biogenesis by providing intracellular sphingosine 1-phosphate (S1P). Blood. 2013;122(5):791-802. 17. Abe T, Fujimori T. Reporter mouse lines for fluorescence imaging. Dev Growth Diff. 2013;55(4):390-405. 18. Chappell J, Harman JL, Narasimhan VM, et al. Extensive proliferation of a subset of differentiated, yet plastic, medial vascular smooth muscle cells contributes to neointimal formation in mouse injury and atherosclerosis models. Circ Res. 2016;119(12):1313-1323. 19. Tas JMJ, Mesin L, Pasqual G, et al. Visualizing antibody affinity maturation in germinal centers. Science. 2016;351(6277):1048-1054. 20. Peng T, Tian Y, Boogerd CJ, et al. Coordination of heart and lung co-development by a multipotent cardiopulmonary progenitor. Nature. 2013;500(7464):589-592. 21. Kretzschmar K, Watt FM. Lineage tracing. Cell. 2012;148(1-2):33-45. 22. Livet J, Weissman TA, Kang H, et al. Transgenic strategies for combinatorial expression of fluorescent proteins in the nervous system. Nature. 2007;450(7166):56-62. 23. Snippert HJ, Van Der Flier LG, Sato T, et al. Intestinal crypt homeostasis results from neutral competition between symmetrically dividing Lgr5 stem cells. Cell. 2010;143(1):134-144. 24. Tiedt R, Schomber T, Hao-Shen H, Skoda RC. Pf4-Cre transgenic mice allow the generation of lineage-restricted gene knockouts for studying megakaryocyte and platelet function in vivo. Blood. 2007;109(4):1503-1506. 25. Yuan L, Chan GC, Beeler D, et al. A role of stochastic phenotype switching in generating mosaic endothelial cell heterogeneity. Nat Commun. 2016;7:10160. 26. Leon C, Eckly A, Hechler B, et al. Megakaryocyte-restricted MYH9 inactivation dramatically affects hemostasis while preserving platelet aggregation and secretion. Blood. 2007;110(9):3183-3191. 27. Samson AL, Alwis I, Maclean JA, et al. Endogenous fibrinolysis facilitates clot retraction in vivo. Blood. 2017;130(23):2453-2462.

Haematologica | 107 July 2022

1679

ARTICLE - Platelet and MK function on the single cell level 28. Petzold T, Thienel M, Konrad I, et al. Oral thrombin inhibitor aggravates platelet adhesion and aggregation during arterial thrombosis. Sci Transl Medi. 2016;8(367):367ra168. 29. Levine RF, Hazzard KC, Lamberg JD. The significance of megakaryocyte size. Blood. 1982;60(5):1122-1131. 30. Tomer A. Human marrow megakaryocyte differentiation: multiparameter correlative analysis identifies von Willebrand factor as a sensi 2004;104(9):2722-2727. 31. Sanjuan-Pla A, Macaulay IC, Jensen CT, et al. Platelet-biased stem cells reside at the apex of the haematopoietic stem-cell hierarchy. Nature. 2013;502(7470):232-236. 32. Nagy Z, Vögtle T, Geer MJ, et al. The Gp1ba-Cre transgenic mouse: a new model to delineate platelet and leukocyte functions. Blood. 2019;133(4):331-343. 33. Pertuy F, Aguilar A, Strassel C, et al. Broader expression of the mouse platelet factor 4-cre transgene beyond the megakaryocyte lineage. J Thromb Haemost. 2015;13(1):115-125. 34. Stegner D, Judith MM, Angay O, et al. Thrombopoiesis is spatially regulated by the bone marrow vasculature. Nat Commun. 2017;8(1):1-11. 35. Nicolai L, Schiefelbein K, Lipsky S, et al. Vascular surveillance by haptotactic blood platelets in inflammation and infection. Nat Commun. 2020;11(1):5778. 36. Lambert MP, Wang Y, Bdeir KH, Nguyen Y, Kowalska MA, Poncz M. Platelet factor 4 regulates megakaryopoiesis through low-density lipoprotein receptor-related protein 1 (LRP1) on megakaryocytes.

L. Nicolai et al. Blood. 2009;114(11):2290-2298. 37. Jackson SP, Nesbitt WS, Westein E. Dynamics of platelet thrombus formation. J Thromb Haemost. 2009;7(Suppl 1):S17-20. 38. Aslan JE, Itakura A, Gertz JM, McCarty OJ. Platelet shape change and spreading. Methods Mol Biol. 2012;788:91-100. 39. Schurr Y, Sperr A, Volz J, et al. Platelet lamellipodium formation is not required for thrombus formation and stability. Blood. 2019;134(25):2318-2329. 40. Nesbitt WS, Westein E, Tovar-Lopez FJ, et al. A shear gradientdependent platelet aggregation mechanism drives thrombus formation. Nat Med. 2009;15(6):665-673. 41. Mattila PK, Lappalainen P. Filopodia: molecular architecture and cellular functions. Nat Rev Mol Cell Biol. 2008;9(6):446-454. 42. Galbraith CG, Yamada KM, Galbraith JA. Polymerizing actin fibers position integrins primed to probe for adhesion sites. Science. 2007;315(5814):992-995. 43. Paul DS, Casari C, Wu C, et al. Deletion of the Arp2/3 complex in megakaryocytes leads to microthrombocytopenia in mice. Blood Adv. 2017;1(18):1398-1408. 44. Stenberg PE, Barrie RJ, Pestina TI, et al. Prolonged bleeding time with defective platelet filopodia formation in the Wistar Furth rat. Blood. 1998;91(5):1599-1608. 45. Nechipurenko DY, Receveur N, Yakimenko AO, et al. Clot contraction drives the translocation of procoagulant platelets to thrombus surface. Arterioscler Thromb Vasc Biol. 2019;39(1):37-47.

Haematologica | 107 July 2022

1680

LETTER TO THE EDITOR

Outcomes of refractory or relapsed Hodgkin lymphoma patients with post-autologous stem cell transplantation brentuximab vedotin maintenance: a French multicenter observational cohort study The majority of patients with Hodgkin lymphoma (HL) are cured with first-line therapy, but 10-20% of patients still experience refractory or relapsing (R/R) disease. The current standard of care for R/R HL is salvage chemotherapy, followed by autologous hematopoietic stem cell transplantation (ASCT) and brentuximab vedotin (BV) maintenance, based on the results of AETHERA. This study demonstrated that R/R HL patients with refractory disease, or experiencing early (less than 12 months from chemotherapy completion) or extranodal relapse (at any time) have a lower risk of progression or death when receiving BV maintenance compared to placebo.1,2 These results led to the approval of post-transplant BV maintenance for high-risk R/R HL patients in 2017. In AMAHRELIS (Adcetris Maintenance after Autologous stem cell transplantation in Hodgkin lymphoma: a Real-Life Study), a retrospective nationwide French cohort study, we investigated the real-life outcome of patients with R/R HL who received post-transplant BV maintenance. Notably, most patients received BV during salvage, in contrast to the AETHERA cohort in which prior BV exposure represented an exclusion criterion. We also performed a central review of 18-fluorodeoxyglucose-positron emission tomography (FDG-PET) imaging at relapse and before transplantation, by two independent experts, with a complete evaluation of 79% of the cohort. We included patients 16 years and older with R/R HL who received at least two infusions of BV maintenance after ASCT. Patients who received BV for progression after transplant were excluded. Among 1,134 patients included in the French Society of Bone Marrow Transplantation databases who underwent ASCT for R/R HL between 2012 and 2017 in France, we received responses for 835 patients (73%) from 25 centers. Finally, 115 patients met eligibility criteria for our study (Figure 1A). The patients’ characteristics are summarized in Table 1. The median age was 34 years (range, 16-68 years), and 62 (54%) were male. Sixty-nine (60%) patients had stage III or IV disease at diagnosis. ABVD was the first line of treatment for 64 patients (56%), escalated BEACOPP was administered to 42 (37%) and nine patients received other regimens. Fifty (43%) patients had primary refractory disease, 32 (28%) experienced early relapse (before 12 months) and 33 (29%) relapsed later than 12 months. At relapse, histological confirmation was obtained for half of

the patients. Sixty-seven (58%) patients had stage III-IV disease, 19 (17%) had B symptoms and extranodal disease occurred in half of them. A BV-based salvage regimen was used in 34 (29.5%) patients during the first salvage and 29 of them (85%) achieved a complete response (CR), while 81 did not receive BV and 24 (29%) of them achieved a CR. The difference in CR rate between patients who did or did not receive a BV-based salvage regimen was highly significant (Figure 1B). Among 57 patients who did not receive a BV-based regimen at first salvage, 46 (81%) were given BV during the second salvage and 37 of them (80%) achieved CR. Pre-transplant FDG-PET status was reported for 111 (97%) patients and among them, 93 (84%) were reported to be in CR. Among 91 patients (79% of the cohort) with centrally reviewed FDG-PET data, 82.4% (75/91) achieved metabolic CR (defined as a Deauville score 1-3) before ASCT. According to AETHERA, 95% of patients met inclusion criteria for BV maintenance due to primary refractory disease (43%), early relapse (28%) or extranodal involvement (49%). The mean number of BV injections after ASCT was 11 (range, 3-18), without difference between patients who did or did not receive salvage BV. The median time between ASCT and the first BV maintenance cycle was 70 days (range, 18-223), and 88 (77%) patients were treated within 3 months from ASCT. The main reported adverse event was neuropathy, which occurred in 50 (43%) patients, with complete resolution in half of them. Treatment-related events led to BV maintenance discontinuation in 10% of patients, and included neuropathy (6 patients), infections (3 patients), thrombocytopenia (1 patient), and pancreatitis (1 patient). Neuropathy was more frequent in patients who received pre-transplant BV, without impact on treatment discontinuation rate. The median follow-up period was 35 months. The 2-year progression-free survival (PFS) and overall survival (OS) for the whole cohort were 75.3% (95% confidence interval: 68.4-84.3%) and 96.4% (95% confidence interval: 94.2%100%), respectively (Figure 1C, D). Seven patients died: three from disease progression, two from a second cancer (1 acute myeloid leukemia and 1 pancreatic cancer), and two from infection (meningitis due to Streptococcus pneumoniae in both cases). The non-relapse mortality rate was 3.5% (4 patients) during the follow-up of our study. During or after BV maintenance, 30 (26%) patients relapsed and among them, 21 (70%) received an immune checkpoint

Haematologica | 107 July 2022

1681

LETTER TO THE EDITOR

Figure 1. Impact of post-transplant brentuximab vedotin maintenance therapy in real-life practice: the AMAHRELIS study. (A) Flow chart of patients entering the study. (B) Proportion of patients who did or did not receive brentuximab vedotin during the first and second lines of salvage therapy. The percentage of patients in complete remission is indicated. ***P<0.001 (Fisher test). (C, D) Progression-free and overall survival of the 115 patients of the AMAHRELIS cohort since transplant. The 95% confidence intervals are shown by the pale shaded areas on both sides of the survival curves. (E, F) Progression-free and overall survival probabilities dependent on the achievement of a complete metabolic response among 91 patients of the AMAHRELIS cohort after central review of the 18-fluorodeoxyglucose-positron emission tomography data. R/R HL: relapsed-refractory Hodgkin lymphoma; ASCT: autologous hematopoietic stem cell transplantation; BV: brentuximab vedotin; CR: complete remission; noCR: not in complete remission; PFS: progression-free survival; OS: overall survival; mCR: metabolic complete response; ≠mCR: not in metabolic complete response. Haematologica | 107 July 2022

1682

LETTER TO THE EDITOR blocker of whom 15 (71%) had a response, including 13 CR. Using a univariate Cox regression model, we tested several variables listed in Table 2 for correlation with survival. We found that refractory status, early relapse (less than 12

months), high-risk LYSA prognostic score (primary refractory disease, or early relapse and disseminated disease),3 and absence of pre-transplant metabolic CR (after central FDG-PET review) were significantly predictive of reduced

Table 1. Characteristics of the 115 patients from the AMAHRELIS cohort and the 165 patients from AETHERA.

AMAHRELIS N=115 Male Age in years (mean, min-max) Frontline chemotherapy ABVD Escalated BEACOPP other Time to relapse Primary refractory disease (≤ 3 months) Early relapse (> 3 or ≤ 12 months) Late relapse (> 12 months) Histological confirmation at relapse Yes No Stage at relapse I-II III-IV Unknown B symptoms at time of relapse Yes No Unknown Bulky disease at relapse Yes No Unknown Extranodal relapse Yes No Unknown LYSA score Low Intermediate High Salvage lines (n) 1 2 ≥3 Pre-transplant BV Yes No Pre-transplant FDG-PET* No metabolic CR Metabolic CR (DS 1,2,3) Not centrally reviewed Time to BV in days (median, min-max)

AETHERA N=165

N 62 34

% 54 16-68

N 76

% 46

64 42 9

56 37 8

119 26 20

72 16 12

50 32 33

43 28 29

99 53 13

60 32 8

67 48

58 42

45 67 3

39 58 3

19 85 11

17 74 10

11 90 14

10 78 12

56 54 5

49 47 4

9 38 68

8 33 59

56 50 9

49 43 8

81 34

70 30

16 75 24 70

17.6 82.4 18-223

165 Unknown 45 FDG positive 64 FDG negative 56

100 27 39 34 100 28-49

Data are number and percentage unless otherwise indicated. *Tomography percentages are based on the 91 patients for whom central review of imaging was available. ABVD: adriamycin, bleomycin, vinblastine, dacarbazine; BEACOPP: bleomycin, etoposide, adriamycin, cyclophosphamide, vincristine, procarbazine, prednisone; LYSA: Lymphoma Study Association; FDG: 18-fluorodeoxyglucose; PET: positron emission tomography; CR: complete remission; DS: Deauville score; BV: brentuximab vedotin. Haematologica | 107 July 2022

1683

LETTER TO THE EDITOR Table 2. Univariate and multivariate Cox regression analyses on AMAHRELIS.

Progression-free survival Univariate

Overall survival

Multivariate

95% CI

Sex

115

0.51-2

0.98

Age*

115

0.98-1

0.58

Refractory

115

2.1

1.1-4.2

0.033

0.68

Rel.<12

115

3.4

1.2-9.8

0.021

ECOG*

0.63-1.8

Stage*

112

B-symptoms*

95% CI

Univariate P

95% CI

0.35 0.068-1.8

Multivariate P

0.98-1.1

0.18

0.378

0.59

0.11-3

0.53

3.84 0.75-19.69 0.107

2.3

0.28-19

0.44

0.85

1.4

0.51-3.9

0.5

0.74-1.4

0.91

1.6

0.69-3.6

0.28

104

0.56 0.19-1.6

0.28

0.62 0.074-5.2

0.66

Bulk*

101

1.5

0.51-4.2

0.48

1.9

0.22-17

0.55

Extra.*

110

0.82 0.41-1.6

0.57

0.38

0.074-2

0.25

Irrad.

112

1.1

0.42-2.8

0.85

2.5

0.48-13

0.28

AETHERA

104

1.3

0.89-2

0.16

1.9

0.81-4.2

0.14

LYSA

115

0.32 0.14-0.75 0.008

0.58

0.11-3

0.51

Pre-BV

115

0.89 0.43-1.8

0.75

0.6

0.12-2.4

0.43

Salvage

115

1.1

0.63-1.8

0.84

1.8

0.75-4.3

0.18

FDG-PET

2.9

1.3-6.6

0.013

7.7

1,7-35

0.67

3.34

0.18-2.54

1.41-7.9

0.557

0.006

95% CI

0.21

1 0.29-1.61

0.0079 7.68 1.71-34.51 0.008

HR: hazard ratio; 95% CI: 95% confidence interval (min-max); *: at the time of relapse; Rel<12: relapse before 12 months; Extra.: extranodal relapse; Irrad: irradiated field relapse; Pre-BV: pre-transplant use of brentuximab vedotin; FDG-PET: results of central 18-fluorodeoxyglucose positron emission tomography analysis with Deauville scores 1, 2 and 3 classified as metabolic complete remission, and scores 4 and 5 classified as no metabolic complete remission.

PFS, while FDG-PET status was the only variable significantly correlated with OS (Table 2). Notably, survival probability was similar between patients who did or did not receive a BV-based regimen before transplantation (Table 2). PFS and OS probabilities at 24 months dependent on significant variables, including LYSA prognostic score, refractory status or relapse timing, are provided in Online Supplementary Figure S1. Using a multivariate Cox model for significant variables identified in univariate analysis, we found that only absence of metabolic CR (i.e., Deauville score 4 and 5) before transplantation correlated significantly with reduced PFS and OS (Table 2), which was confirmed by the log-rank test (Figure 1E, F). Currently, evaluation of pre-transplant response to salvage therapy by FDG-PET is recommended,4–6 although not allowing post-transplant therapeutic guidance. Our results showed that FDG-PET response after salvage is strongly associated with survival, and thus a next step could be to assess FDG-PET-driven post-transplant strategies in clinical trials. In our study, R/R HL patients treated with post-transplant BV maintenance had a 2-year PFS of 75%, similar to the results of AETHERA, although a direct comparison between AETHERA and our current study would not be cor-

rect, since the patients’ characteristics were different (Table 1). More patients received first-line escalated BEACOPP in our cohort. Moreover, a majority of patients in our cohort (70%) received off-label BV-based salvage regimens, while patients with pre-transplant exposure to BV were excluded from AETHERA. Notably, pre-transplant BV use had no impact on the completion of BV maintenance in our study, and was associated with a high pretransplant CR rate (82% metabolic CR rate according to FDG-PET central review). We observed that achievement of metabolic CR before transplant was predictive of improved PFS and OS, also after multivariate analysis. Indeed, recent studies have attempted to increase metabolic CR rate by incorporating BV into initial salvage therapy, including bendamustine, DHAP, ESHAP and gemcitabine.7–10 In particular, the BRAVE study,9 a phase II trial of BV-DHAP without post-transplant BV maintenance, resulted in a 2-year PFS rate of 74%, similar to our results. These observations suggest that the optimal timing of BV use, as salvage or as maintenance, remains to be determined in future prospective clinical trials. Despite accurate risk stratification and generalization of BV use, the prognosis of high-risk R/R HL patients remains a matter of concern. However, the excellent OS results ob-

Haematologica | 107 July 2022

1684

LETTER TO THE EDITOR served in our cohort highlight the generalization of use of efficient salvage therapies in post-transplant relapse. In particular, 70% of patients who relapsed during BV maintenance received an immune checkpoint blocker, and 71% of them responded. We may thus hypothesize that selected patients may benefit from immune checkpoint blockade earlier on, as currently being investigated in clinical trials using immune checkpoint blockers as part of salvage or post-transplant maintenance therapy.11–13 During first-line salvage, a combination of BV and nivolumab resulted in a CR rate of 61% after four cycles, without unexpected toxicity,11 and pembrolizumab combined with gemcitabine, vinorelbine and liposomal doxorubicin led to CR in 95% of patients.12 In the post-transplant setting, consolidation with eight cycles of pembrolizumab resulted in an 82% PFS rate at 18 months.13 Thus, incorporating immune checkpoint blockers into salvage and/or post-transplant strategies represents a promise for R/R HL patients at high risk of treatment failure or progression which should be investigated further in clinical trials. On the other hand, identification of a subgroup of patients with a more favorable profile in the context of these new therapies may enable omission of consolidative ASCT thereby avoiding the risk of early and late toxicities. In this perspective, FDG-PET-based risk stratification at relapse could benefit from quantitative analysis and the assessment of the dynamic evolution of metabolic tumor volume.14 In conclusion, our real-life nationwide study confirmed the improved survival of R/R HL patients receiving posttransplant BV compared to historical cohorts. The exact timing of BV administration, and the place of new therapies such as immune checkpoint blockers in current salvage strategies remain to be determined in future clinical trials.

France; 6Lymphoma Academic Research Organization (LYSARC) Lymphoma Study Association Imaging, Hôpital Henri Mondor, Créteil, France; 7Paris Est University, Créteil, France; 8Hematology Department and INSERM U1151, Institut Necker Enfants Malades, Necker University Hospital, AP-HP, Paris, France; 9Department of Hematology, Nantes University Hospital, Nantes, France; 10

Department of Hematology, Institut Universitaire du Cancer

Toulouse- Oncopole, Toulouse, France; 11Lymphoid Malignancies Unit, Hôpital Henri Mondor, Créteil, France; 12Department of Hematology, Aix-Marseille University, Marseille, France; 13

Department of Medical Oncology, Centre Léon Bérard, Lyon,

France; 14Department of Hematology, Centre Hospitalier Universitaire (CHU) Estaing, Clermont-Ferrand, France; 15Department of Hematology, Centre Henri Becquerel, Rouen, France; 16 17

Department of Hematology, Clinique Louis Pasteur, Nancy, France;

Department of Hematology, Institut Gustave-Roussy, Villejuif,

France; 18Department of Hematology, CHU Besançon, Besançon, France; 19Strasbourg University Hospital, Strasbourg, France; 20

INSERM S-1113, Strasbourg, France; 21Strasbourg University, Faculty

of Medicine, Strasbourg, France;

Department of Hematology,

Hôpital Saint Vincent de Paul, Lille, France;

Department of

Hematology, Hôpital Lyon Sud, Pierre-Bénite, France;

Société

Francophone de Greffe de Moelle et de Thérapie Cellulaire, France; 25

Service de pathologie, Hôpital Cochin, AP-HP, Paris, France;

Centre de recherche des Cordeliers, Sorbonne University, INSERM,

Paris University, Paris, France;

Department of Hematology, CHRU

Nancy, Hôpital Brabois, Nancy, France;

CNRS UMR 7365, Équipe 6,

Biopôle de L'Université de Lorraine, Vandœuvre-les-Nancy, France; 29

Institut Cochin, INSERM U1016, Paris, France;

Department of

Hematology, University of Montpellier, Montpellier, France; 31

Department of Hematology, Dijon University Hospital, Dijon,

France;

INSERM UMR 1231 CHU Dijon, Dijon, France;

Department

of Hematology, CHU Paris-GH St-Louis Lariboisière F-Widal Hôpital Saint-Louis, Paris, France;

Department of Hematology,

Centre Hospitalier Lyon Sud, Pierre-Bénite, France and 35

Authors

Translational Research Center in Onco-hematology, Faculty of

Medicine, University of Geneva, Geneva 4, Switzerland.

Amira Marouf,1,2,3 Anne Segolene Cottereau,2,4 Salim Kanoun,5 Paul Deschamps,1,2 Michel Meignan,6,7 Patricia Franchi,1,2 David Sibon,8 Clara Antoine,6 Thomas Gastinne,9 Cecile Borel,10 Mohammad Hammoud,11 Guillaume Sicard,12 Romane Gille,13 Doriane Cavalieri,14 Aspasia Stamatoullas,15 Lauriane Filliatre-Clement,16 Julien Lazarovici,17 Adrien Chauchet,18 Luc-Matthieu Fornecker,19,20,21 Sandy Amorin,22 Mathieu Rocquet,1 Nicole Raus,23,24 Barbara Burroni,25,26 Marie Therese Rubio,24,27,28 Didier Bouscary,1,2,29 Philippe Quittet,30 Rene Olivier Casasnovas,31,32 Pauline Brice,33 Herve Ghesquieres,34 Jérôme Tamburini1,2,29,35 and Benedicte Deau1,2 on behalf of the SFGM-TC and LYSA groups.

Correspondence: BENEDICTE DEAU: benedicte.deau-fischer@aphp.fr https://doi.org/10.3324/haematol.2021.279564 Received: July 5, 2021. Accepted: November 11, 2021. Prepublished: December 30, 2021. Disclosures AM has no disclosures to make; BD has received financial support from Roche and Takeda, not related to the current manuscript.

Department of Hematology, Hôpital Cochin, Assistance Publique -

Hôpitaux de Paris (AP-HP), Paris, France; 2Université de Paris, Paris, 3

Contributions

France; INSERM UMR 1163, Institut Imagine, Paris, France;

AM, PD, PF, DS, TG, CB, MH, GS, RG, DC, AS, LF, JL, AC, LF, SA, MR, NR,

BB, MTR, DB, PQ, OC, PB, HG, JT and BD performed the research, BD

Department of Nuclear Medicine, Hôpital Cochin, Assistance 5

Publique-Hôpitaux de Paris (AP-HP), Paris, France; Cancer Research

designed the study, ASC, SK, MM, CA performed the PET review, AM, JT

Center of Toulouse (CRCT), Team 9, INSERM UMR 1037, Toulouse,

and BD analyzed the data, JT and BD wrote the paper.

Haematologica | 107 July 2022

1685

LETTER TO THE EDITOR

Acknowledgments The authors would like to thank all the patients for their cooperation.

Data-sharing statement Original data can be made available on reasonable request to the authors.

Funding Takeda France provided a grant to support the central PET review for this study.

References 1. Moskowitz CH, Nademanee A, Masszi T, et al. Brentuximab vedotin as consolidation therapy after autologous stem-cell transplantation in patients with Hodgkin’s lymphoma at risk of relapse or progression (AETHERA): a randomised, double-blind, placebo-controlled, phase 3 trial. Lancet. 2015;385(9980):1853-1862. 2. Moskowitz CH, Walewski J, Nademanee A, et al. Five-year PFS from the AETHERA trial of brentuximab vedotin for Hodgkin lymphoma at high risk of progression or relapse. Blood. 2018;132(25):2639-2642. 3. Neste EVD, Casasnovas O, André M, et al. Classical Hodgkin’s lymphoma: the Lymphoma Study Association guidelines for relapsed and refractory adult patients eligible for transplant. Haematologica. 2013;98(8):1185-1195. 4. Moskowitz AJ, Schöder H, Yahalom J, et al. PET-adapted sequential salvage therapy with brentuximab vedotin followed by augmented ifosamide, carboplatin, and etoposide for patients with relapsed and refractory Hodgkin’s lymphoma: a non-randomised, open-label, single-centre, phase 2 study. Lancet Oncol. 2015;16(3):284-292. 5. Moskowitz CH, Matasar MJ, Zelenetz AD, et al. Normalization of pre-ASCT, FDG-PET imaging with second-line, non–crossresistant, chemotherapy programs improves event-free survival in patients with Hodgkin lymphoma. Blood. 2012;119(7):1665-1670. 6. Hoppe RT, Advani RH, Ai WZ, et al. Hodgkin lymphoma, version 2.2020, NCCN clinical practice guidelines in oncology. J Natl Compr Canc Netw. 2020;18(6):755-781. 7. Michallet AS, Guillermin Y, Deau B, et al. Sequential combination of gemcitabine, vinorelbine, pegylated liposomal doxorubicin and brentuximab as a bridge regimen to transplant in relapsed

or refractory Hodgkin lymphoma. Haematologica. 2015;100(7):e269-271. 8. LaCasce AS, Bociek RG, Sawas A, et al. Brentuximab vedotin plus bendamustine: a highly active first salvage regimen for relapsed or refractory Hodgkin lymphoma. Blood. 2018;132(1):40-48. 9. Kersten MJ, Driessen J, Zijlstra JM, et al. Combining brentuximab vedotin with dexamethasone, high-dose cytarabine and cisplatin as salvage treatment in relapsed or refractory Hodgkin lymphoma: the phase II HOVON/LLPC Transplant BRaVE study. Haematologica. 2021;106(4):1129-1137. 10. Garcia-Sanz R, Sureda A, de la Cruz F, et al. Brentuximab vedotin and ESHAP is highly effective as second-line therapy for Hodgkin lymphoma patients (long-term results of a trial by the Spanish GELTAMO group). Ann Oncol. 2019;30(4):612-620. 11. Herrera AF, Moskowitz AJ, Bartlett NL, et al. Interim results of brentuximab vedotin in combination with nivolumab in patients with relapsed or refractory Hodgkin lymphoma. Blood. 2018;131(11):1183-1194. 12. Moskowitz AJ, Shah G, Schöder H, et al. Phase II trial of pembrolizumab plus gemcitabine, vinorelbine, and liposomal doxorubicin as second-line therapy for relapsed or refractory classical Hodgkin lymphoma. J Clin Oncol. 2021;39(28):3109-3117. 13. Armand P, Chen Y-B, Redd RA, et al. PD-1 blockade with pembrolizumab for classical Hodgkin lymphoma after autologous stem cell transplantation. Blood. 2019;134(1):22-29. 14. Moskowitz AJ, Schöder H, Gavane S, et al. Prognostic significance of baseline metabolic tumor volume in relapsed and refractory Hodgkin lymphoma. Blood. 2017;130(20):2196-2203.

Haematologica | 107 July 2022

1686

LETTER TO THE EDITOR

Vaccine-induced immune thrombotic thrombocytopenia: a possible pathogenic role of ChAdOx1 nCoV-19 vaccineencoded soluble SARS-CoV-2 spike protein Vaccine-induced immune thrombotic thrombocytopenia (VITT) is a rare catastrophic syndrome, occurring 5 to 30 days after the first dose of the adenoviral vector-based vaccines, ChAdOx1 nCoV-19 and coronavirus disease 2019 (COVID-19) Janssen vaccine, both encoding the full length severe acute respiratory syndrome coronavirus-2 (SARSCoV-2) spike protein (SP).1-3 Although VITT resembles autoimmune heparin-induced thrombotic thrombocytopenia, which is caused by the development of antibodies against platelet factor 4 (PF4)-polyanion complexes,4,5 the etiopathogenesis, in particular what triggers the initial platelet activation, is still poorly understood. Recently published data indicate that ChAdOx1 nCoV-19 vaccine constituents6 and the adenovirus per se7 could be the polyanion that, binding to PF4, supports the formation of immunocomplexes which, in turn, activate platelets and stimulate the release of procoagulant neutrophil extracellular traps (NET).6 The excipient EDTA in ChAdOx1 nCoV-19 vaccine has been shown to increase microvascular permeability, favoring dissemination of vaccine components into the bloodstream.6 Interesting data from a preprinted paper show that adenoviral DNA can undergo alternative splicing, which could lead to the synthesis of soluble SP (sSP) variants lacking the membrane anchor but able to activate platelets and endothelial cells.8 However, a case of VITT has been described after a second dose of the Moderna mRNAbased SARS-CoV-2 vaccine,9 suggesting that other factors besides the adenoviral vector may be implicated in the pathogenesis of VITT. In the present study, we tested the hypothesis that a sSP, possibly the product of alternative splicing, is the first trigger of VITT. The study was approved by the Ethics Committee of the University “La Sapienza” of Rome (study number 6305). We studied three patients with VITT and seven vaccinated healthy controls within 3 weeks after the first dose of the ChAdOx1 nCoV-19 vaccine. The main demographics, clinical and laboratory findings of the studied cohort are summarized in Table 1. VITT patients 1 and 2, both suffering from malignant middle cerebral artery stroke and extensive venous splanchnic and arterial pulmonary thrombosis, have been described previously.10 Patient 3 was a 59-year-old male who developed complete intra-extrahepatic portal system and partial superior mesenteric vein thrombosis 11 days after vaccination. The diagnosis of VITT was based on the detection of high

plasma levels of antibodies (IgG/IgM/IgA) against PF4-polyanion complexes, quantified by enzyme-linked immunosorbent assay (ELISA) (Immucor, Lifecodes, Waukesha, WI, USA), and positivity of the serum-induced platelet function test performed according to Guarino et al.11 None of the volunteer healthy controls tested positive for anti-PF4 antibodies. A hypercoagulable state was observed in VITT patients with high plasma levels of von Willebrand factor (VWF), D-dimer, and coagulation factor VIII. The reduction in factor XIII suggests an excessive consumption within thrombi. All VITT patients showed evidence of endothelial activation with significantly elevated VWF:RCo, the active form of VWF able to bind to platelet glycoproteins GPIb-IX-V (Online Supplementary Figure S1). Serum markers of NET, quantified by measuring DNA-myeloperoxidase complexes (Roche, cat. #11774425001) according to Kessenbrock et al.,12 were elevated in all patients. The thrombus retrieved from the middle cerebral artery of patient 1 was composed mainly of platelets and massively infiltrated by neutrophils10 (Online Supplementary Figure S2). Most importantly we observed SP within the thrombus. Double staining with anti-CD61 (a platelet marker) and antiSARS-CoV-2 SP antibodies showed co-localization of the two antigens. Immunohistochemistry, performed using two different antibodies: anti-angiotensin-converting enzyme 2 (mouse monoclonal anti-ACE2 - Cell Signaling Technology, Boston, MA, USA, cat. #74512, dilution 1:200) and antiSARS-CoV-2 SP (rabbit polyclonal anti-SARS-CoV-2 SP Cell Signaling Technology, Boston, MA, USA, cat. #56996, dilution 1:100), showed weak but diffuse SP staining which co-localized with ACE2 staining. A thrombus retrieved from a pre-pandemic age- and sex-matched stroke patient showed stronger diffuse staining for ACE2 but no evidence of staining for SP (Figure 1). To test if signaling by sSP is related to the pathogenesis of VITT we investigated whether sSP and the soluble ACE2 (sACE2), the shed form of the primary receptor for SARSCoV-2, were present in the serum of VITT patients by ELISA (My BioSource, San Diego, CA, USA, cat. #MBS7608267 and BioVision, Milpitas, CA, USA, cat. #E4528-100, respectively). All VITT serum samples and two of seven samples from volunteer healthy controls showed detectable levels of sSP and sACE2. Only one volunteer healthy control without evident sSP showed measurable sACE2. Serum from patient 1 was collected and analyzed at different intervals after the

Haematologica | 107 July 2022

1687

1688

Haematologica | 107 July 2022

V-HC 3

V-HC 4

V-HC 5

V-HC 6

V-HC 7

none

HTN

HT Previous breast cancer

12 (median)

10 (median)

> 4309

5441

> 4318

177

336

366

133

0.724

1.696

0.231

0.18

0.12

0.15

0.41

0.09

0.32

1.76

1.29

Neg*

< 2%

1.88

2.6

1.165

0.868

0.869

2.561

1.0094

0.533

32.876

2.843

1.1929

Died

Bilateral MCA + PV+ Pulmonary IVIG Steroid arteries

Intra- and extraIVIG Steroid hepatic PV + Fondaparinux Recovering Partial superior PE PLT transf. mesenteric vein

Died

Treatment for Outcome VITT

Right MCA + IVIG Steroid PV+ Pulmonary Fondaparinux arteries PE PLT transf.

Site of thrombosis

Soluble SARS-CoV-2 spike protein, soluble ACE2, and anti-PF4 antibodies were assayed in serum collected within 72 h of hospital admission for patients with vaccine-induced immune thrombotic thrombocytopenia and within 16 days after the first dose of ChAdOx1 nCov-19 vaccination in healthy controls. *In patient 1, high levels of anti-PF4 antibodies were found at day 24 after vaccination (OD405: 1.68). Platelet activation was not inhibited efficiently by high-dose heparin in patients 1 and 2 (ATP release after 20 min: 19% and 9%, respectively) whereas it was inhibited efficiently in patient 3 (ATP release after 20 min: 1%). RV: reference values; FBG: fibrinogen; NET: neutrophil extracellular traps; ELISA: enzyme-linked immunosorbent assay; OD405: optical density; PF4: platelet factor 4; sSP: soluble spike protein; sACE2: soluble angiotensin-converting enzyme 2; VITT: vaccine-induced immune thrombotic thrombocytopenia; V-HC: vaccinated healthy control; F: female; M: male; HT: hypothyroidism; MCA: middle cerebral artery; PV: portal vein; IVIG: intravenous immunoglobulin (1 g/kg for 2 days); PE: plasma exchange; PLT transf.: platelet transfusion; HTN: hypertension; NA: not available. Value 0 means that the protein was not detected with this methodology.

37 (median)

V-HC 2

V-HC 1

57 (median)

VITT Patient 2

VITT Patient 3

VITT Patient 1

Platelet Time from D-dimer FBG Platelets activation Anti-PF4 Age CosSP sACE2 vaccine to (µg/L) mg/dL (x109/L) NET assay Subjects Sex (ELISA) RV: (years) morbidities symptoms RV: (ELISA) RV: RV: (% ATP (pg/mL) (ng/mL) OD405 <0.5 (days) 0-500 200-400 150-450 release) RV: <2%

Table 1. Summary of the demographics, clinical and laboratory data of patients with vaccine-induced immune thrombotic thrombocytopenia and of ChAdOx1 nCov-19 vaccinated healthy controls.

LETTER TO THE EDITOR

LETTER TO THE EDITOR stroke (Online Supplementary Table S1). Notably, sSP was still detectable at 35 days after vaccination with values comparable to those detected at days 9 and 11. Although a consistent reduction in ELISA reactivity for anti-PF4 antibodies was observed 20 days after plasma exchange, the platelet functional activity test showed that serum from VITT patient 1 was still able to induce platelet ATP release (20%) after 20 minutes suggesting that a stimulus other than anti-PF4 antibodies persisted in the blood and pro-

moted platelet activation. To evaluate whether the sSP detected in the serum of the VITT patients is implicated in platelet activation, serum obtained from VITT patients was used to stimulate washed platelets from three healthy donors in the absence or presence of an antibody directed against the S1 subunit of the SP (Invitrogen, clone P06DHuRb). For comparison we also tested the effect of a FcγIIa receptor-blocking antibody (Boster Biological Technology, clone IV.3) that blocks pla-

Figure 1. A thrombus from patient 1 with vaccine-induced immune thrombotic thrombocytopena was rich in platelets and stained positive for SARS-CoV-2 spike protein and ACE2. (A) Double immunofluorescence of thrombotic material retrieved from the right middle cerebral artery of patient 1 during the first mechanical thrombectomy. Staining with hematoxylin-eosin showed that the thrombus was made up almost exclusively of platelets, with abundant granulocytes. Platelets within the area encircled in (A) are stained in red with CD61 antibodies, the SARS-CoV-2 spike protein (SP) is stained in green, while the nuclei of the inflammatory cells are stained in blue with DAPI. The overlap of SARS-CoV-2 SP and platelets is shown in yellow (merge). (B) Immunohistochemistry highlights the presence of SARS-CoV-2 SP associated with decreased amounts of ACE2, within the thrombus of patient 1 (P1) as compared with a thrombus retrieved from a patient in the pre-pandemic era (CTL). H&E: hematoxylin-eosin; SARS-CoV-2: after severe acute respiratory virus coronavirus-2; ACE2: angiotensin-converting enzyme-2. Haematologica | 107 July 2022

1689

LETTER TO THE EDITOR telet activation mediated by anti-PF4/polyanion complexes. After 15 minutes of incubation (at 37°C) with the respective antibodies, platelet activation was assessed by measuring the binding of PAC1-FITC (BD Bioscience, San Jose, CA, USA), which detects the active form of the platelet integrin αIIbβ3, and the binding of α-CD62P-PE, a marker of platelet degranulation, on a BD Accuri C6 flow cytometer. Sera from all VITT patients induced robust activation of platelets from healthy donors (Figure 2A, B). As expected, the anti-FcγRIIA blocking antibody inhibited the observed activation. Interestingly we also found that the antibody against the SP partially inhibited serum-induced platelet activation (Figure 2A, B); in detail, it significantly reduced the activation of the integrin by 60%, supporting our working hypothesis that SP may have a direct effect on platelet activation (Figure 2C). During SARS-CoV-2 infection, SP can be shed and free S1 subunit can be released from both the virus and infected cells.13 The concentration of SP in the plasma was shown to correlate with the severity of COVID-19. Although the presence of free S1 in the bloodstream could be the effect of tissue damage due to viral invasion, the potential, harmful effects of this circulating protein are still uncertain. In-

deed, anomalous thrombogenic activity (with platelet activation and fibrinolytic impairment) has been observed in plasma from both COVID-19 patients and healthy individuals after SP addiction, suggesting a dangerous effect of SP per se on clot formation.14 The mechanism by which the SP could be triggering an effect on platelets is not known. Studies by Zhang and colleagues15 have demonstrated that both SARS-CoV-2 and the SP alone can directly activate platelets via ACE2/SP interaction (although evidence of ACE2 receptors on platelets is controversial) and that platelet activation is suppressed by recombinant human ACE2 protein and by anti-SP monoclonal antibody. We found a partial suppression of platelet integrin activation in all VITT patients by using an antibody directed against the S1 subunit. Moreover, we observed that the level of ACE2 expression was lower on platelets inside the thrombus of a VITT patient than on platelets from the thrombus of a pre-pandemic control patient. A progressive reduction of ACE2 expression on platelets of critically ill COVID-19 patients and on platelets incubated with SARSCoV-2 has been described. This suggests a possible internalization or shedding of the ACE2 receptor induced by SARS-CoV-2. A similar ACE2 downregulation could be hap-

Figure 2. The SARS-CoV-2 spike protein contributes to ChAdOx1 nCoV-19 vaccine-induced platelet activation. (A, B) An antibody against the S1 domain of the SARS-CoV-2 spike protein (α-Spike) decreases vaccine-induced immune thrombotic thrombocytopenia (VITT) serum-induced platelet activation of washed platelets from healthy donors. Platelets were washed by serial centrifugation and resuspended in a solution of Tyrode’s buffer and sera (3:1) at a final concentration of 5x107 cells/mL. The blocking antibodies against the spike protein (SP) and FcγRIIA were incubated for 15 min at 4 mg/ml. Platelet activation was assessed by measuring (A) the binding of PAC1-FITC, an antibody that binds the active form integrin αIIbβ3, and (B) the binding of α-CD62P-PE, which is a marker of granule secretion, on a BD Accuri C6 Plus. The bar graph shows the mean ± standard deviation of the response of the platelets of the three patients (the response for each patient is the average of 3 technical replicates). In each experiment we always included negative controls with buffer alone and with sera from healthy donors who had been vaccinated with ChAdOx1 nCoV-19 but who did not experience any unusual side-effect after the injection. Statistical analyses were performed using ordinary one-way analysis of variance and the Hold-Sidak multiple comparison test. *P<0.05; **P<0.01, ***P<0.001, ****P<0.0001. (C) Working model of the mechanism of vaccine-induced platelet activation. We postulate a multiplehit model for platelet activation in the etiopathogenesis of VITT. The first hit is platelet activation by the SP. (a) The interaction between the SP and the ACE2 receptors on endothelial cells induces endothelial cell activation, (b) which results in platelet recruitment and activation through exposure of adhesion receptors and release of VWF. (c) The direct interaction between the SP and platelets would also activate platelets directly. (d) Activated platelets then release their granular contents which include large amounts of PF4 that, binding to polyanions, generate new antigens, which leads in some individuals to the production of anti-PF4/polyanion autoantibodies. (e) The second hit is the stimulation of platelets via the FcγRIIA by IgG/PF4 and IgG/PF4/polyanion immune-complexes. (f) IgG/PF4/polyanion immune-complexes also stimulate neutrophils that, (g) when co-stimulated by platelets, (h) release neutrophil extracellular traps (NET). (i) Thus, the third hit is NET, which support the coagulation cascade and further support platelet activation. These multiple stimuli amplify platelet activation and lead to thrombosis and/or thrombocytopenia. MFI: mean fluorescent intensity; HD: healthy donor. Haematologica | 107 July 2022

1690

LETTER TO THE EDITOR pening in platelets of VITT patients, stimulated by the SP. Based on these preliminary findings we postulate a multiple-hit model for platelet activation in the etiopathogenesis of VITT (Figure 2C), which would explain why we could inhibit platelet activation by blocking either FcγRIIA or the sSP. The first hit would be the direct interaction between sSP and the ACE2 receptors on endothelial cells and, possibly, on platelets. Activated endothelial cells would induce platelet recruitment and adhesion by exposing adhesion receptors and releasing VWF. Activated platelets would release their granular contents, which include large amounts of PF4. Through the interaction of PF4 with polyanions, new antigens are generated with a consequent production of anti-PF4/polyanion autoantibodies. The second hit would be the stimulation of FcγRIIA by IgG/PF4 and IgG/PF4/polyanion immune-complexes resulting in the amplification of platelet activation. IgG/PF4/polyanion immune-complexes also stimulate neutrophils that, when co-stimulated by platelets, can release NET. Thus, the third hit is these traps, which support the coagulation cascade and further support platelet activation. Interestingly, patient 2 who had a rapidly fatal clinical course showed the highest level of NET. We found SP in the serum of two volunteer healthy controls, but neither of them presented anti-PF4 antibodies, nor did their sera activate platelets from healthy donors. We can hypothesize that sSP variants are produced rarely from alternative nuclear splicing events but that, even more rarely, sSP variants capable of activating platelets can be transcribed. A lower ACE2 receptor expression on platelets, possibly genetically-determined, could also justify the different chance of platelets being activated by sSP. Experiments on more subjects, both vaccinated and non-vaccinated, and on VITT patients are mandatory to confirm the present data and our hypotheses.

Emergency Department, Stroke Unit, Sapienza University of Rome;

Department of Neuroscience, Istituto Superiore di Sanità;

Department of Radiology, Oncology and Pathological Science,

Sapienza University of Rome; 4Department of Medico-Surgical Sciences and Biotechnologies, Sapienza University of Rome; Department of Translational and Precision Medicine, Sapienza

University of Rome; 6Department of Experimental Medicine, Sapienza University of Rome; 7Hematology, Department of Translational and Precision Medicine, Sapienza University of Rome; Neuroradiology Unit, Department of Human Neurosciences,

Sapienza University of Rome and 9Department of Human Neurosciences, Sapienza University of Rome, Rome, Italy Correspondence: MANUELA DE MICHELE - m.demichele@policlinicoumberto1.it https://doi.org/10.3324/haematol.2021.280180 Received: October 15, 2021. Accepted: February 9, 2022. Prepublished: February 17, 2022. Disclosures No conflicts of interest to disclose. Contributions MDM conceived and designed the study, enrolled the patients, interpreted the results, and prepared the original manuscript. PP, AC, and RR performed the ELISA on sera from all subjects included in this study, contributed to data interpretation and critically revised the manuscript. GD and ML performed the histopathological and immunohistochemical examinations of the retrieved thrombus and edited Figure 1. LS performed functional platelet activation assays, edited Figure 2 and critically revised the manuscript. FP performed the functional and serological platelet analyses and contributed to the interpretation of the results. AC is the consultant hematologist who made treatment decisions regarding the VITT patients and analyzed coagulation parameters. MI is the interventional neuroradiologist who retrieved the thrombus from

Authors

patient 1, reviewed the manuscript and edited the figures and references. OGS, IB, EN, LP and MTD participated in data collection, edited the tables and revised the manuscript. DT critically reviewed

Manuela De Michele,1 Paola Piscopo,2 Alessio Crestini,2 Roberto

and edited the manuscript.

Rivabene,2 Giulia d’Amati,3 Martina Leopizzi,4 Lucia Stefanini,5 Fabio Pulcinelli,6 Antonio Chistolini,7 Marta Iacobucci,8 Oscar G. Schiavo,1

Data-sharing statement

Irene Berto, Ettore Nicolini, Luca Petraglia, Maria Teresa Di Mascio

Data will be made available to researchers upon reasonable

and Danilo Toni

request.

References 1. Pavord S, Scully M, Hunt BJ, et al. Clinical features of vaccineinduced immune thrombocytopenia and thrombosis. N Engl J Med. 2021;385(18):1680-1689. 2. Greinacher A, Thiele T, Warkentin TE, Weisser K, Kyrle PA, Eichinger S. Thrombotic thrombocytopenia after ChAdOx1 nCov19 vaccination. N Engl J Med. 2021;384(22):2092-2101.

3. See I, Su JR, Lale A, et al. US case reports of cerebral venous sinus thrombosis with thrombocytopenia after Ad26.COV2.S vaccination, March 2 to April 21, 2021. JAMA. 2021;325(24):2448-2456. 4. Greinacher A, Selleng K, Warkentin TE. Autoimmune heparininduced thrombocytopenia. J Thromb Haemost.

Haematologica | 107 July 2022

1691

LETTER TO THE EDITOR 10. De Michele M, Iacobucci M, Chistolini A, et al. Malignant cerebral infarction after ChAdOx1 nCov-19 vaccination: a catastrophic variant of vaccine-induced immune thrombotic thrombocytopenia. Nat Commun. 2021;12(1):4663. 11. Guarino ML, Massimi I, Mardente S, et al. New platelet functional method for identification of pathogenic antibodies in HIT patients. Platelets. 2017;28(7):728-730. 12. Kessenbrock K, Krumbholz M, Schönermarck U, et al. Netting neutrophils in autoimmune small-vessel vasculitis. Nat Med. 2009;15(6):623-625. 13. Ogata AF, Cheng CA, Desjardins M, et al. Circulating severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) vaccine antigen detected in the plasma of mRNA-1273 vaccine recipients. Clin Infect Dis. 2022;74(4):715-718. 14. Grobbelaar LM, Venter C, Vlok M, et al. SARS-CoV-2 spike protein S1 induces fibrin(ogen) resistant to fibrinolysis: implications for microclot formation in COVID-19. Biosci Rep. 2021;41(8):BSR20210611. 15. Zhang S, Liu Y, Wang X, et al. SARS-CoV-2 binds platelet ACE2 to enhance thrombosis in COVID-19. J Hematol Oncol. 2020;13(1):120.

2017;15(11):2099-2114. 5. Althaus K, Möller P, Uzun G, et al. Antibody-mediated procoagulant platelets in SARS-CoV-2-vaccination associated immune thrombotic thrombocytopenia. Haematologica. 2021;106(8):2170-2179. 6. Greinacher A, Selleng K, Palankar R, et al. Insights in ChAdOx1 nCov-19 vaccine-induced immune thrombotic thrombocytopenia (VITT). Blood. 2021;138(22):2256-2268. 7. Baker AT, Boyd RJ, Sarkar D, et al. ChAdOx1 interacts with CAR and PF4 with implications for thrombosis with thrombocytopenia syndrome. Sci Adv. 2021;7(49):eabl8213. 8. Kowarz E KL, Reis J, et al. “Vaccine-induced Covid-19 mimicry” syndrome: splice reactions within the SARS-CoV-2 Spike open reading frame result in Spike protein variants that may cause thromboembolic events in patients immunized with vectorbased vaccines. Research Square website. Accessed Sept. 10, 2021. DOI: https://doi.org/10.21203/rs.3.rs-558954/v1 [preprint, not peer-reviewed]. 9. Sangli S, Virani A, Cheronis N, et al. Thrombosis with thrombocytopenia after the messenger RNA-1273 vaccine. Ann Int Med. 2021;174(10):1480-1482.

Haematologica | 107 July 2022

1692

LETTER TO THE EDITOR

All-oral triplet combination of ixazomib, lenalidomide, and dexamethasone in newly diagnosed transplanteligible multiple myeloma patients: final results of the phase II IFM 2013-06 study High-dose therapy with autologous stem cell transplantation (ASCT) is considered as a standard of care for patients with transplant-eligible symptomatic newly-diagnosed multiple myeloma (NDMM).1 The benefit of ASCT in such patients has been recently confirmed by two phase III randomized trials demonstrating better progression-free survival and/or overall survival in the transplant arm.2-4 In the past decades, induction therapy before ASCT has been improved dramatically, resulting in deeper responses and prolonged progression-free survival. The triplet combination of bortezomib, lenalidomide and dexamethasone (VRD) is one of the standard-of-care induction regimens in the context of transplantation.1,2,5 Ixazomib is the first-in-class oral proteasome inhibitor approved for patients with relapsed/refractory multiple myeloma in combination with lenalidomide and dexamethasone.6 Here, we report the results of the multicenter, open-label, phase II study by the Intergroupe Francophone du Myelome (IFM), 2013-06, evaluating the efficacy and safety of ixazomib, lenalidomide and dexamethasone (IRD) used as an induction and consolidation regimen followed by ixazomib maintenance in transplanteligible patients with NDMM. This study included transplant-eligible patients with previously untreated symptomatic NDMM. Key selection criteria are indicated in Online Supplementary Figure S1. All patients provided written informed consent to participation in the study, which was approved by relevant national health authorities and ethics committees and was conducted in accordance with the International Conference on Harmonization of Good Clinical Practice guidelines and the principles of the Declaration of Helsinki. This clinical trial is registered at www.clinicaltrials.gov as NCT01936532 and with EUDRACT number 2013-001443-3. Induction therapy comprised three 28-day cycles of oral ixazomib (4 mg on days 1, 8, and 15), oral lenalidomide (25 mg on days 1–21) and oral dexamethasone (40 mg on days 1, 8, 15, and 22). Stem cell harvest was planned for all patients after high-dose cyclophosphamide (3 g/m2) plus granulocyte colony-stimulating factor). Patients proceeded to transplant using melphalan 200 mg/m2 as the conditioning regimen. Patients whose disease did not progress then proceeded to early consolidation therapy with two 28-day cycles of IRD, followed by late consoli-

dation with six 28-day cycles of ixazomib (4 mg on days 1, 8, and 15) and lenalidomide (25 mg on days 1–21) without dexamethasone. Patients subsequently received maintenance therapy with ixazomib (4 mg/day on days 1, 8, and 15) for 1 year. The primary endpoint was stringent complete response (sCR) rate at the completion of extended consolidation. Secondary endpoints included response at each step of the program, time to response, quality of stem cell harvest, progression-free survival, overall survival, and safety. Myeloma response assessment was based on the International Myeloma Working Group uniform response criteria.7 sCR was defined as complete response (CR) with the addition of normal serum free light chain ratio and absence of clonal plasma cells in the bone marrow, as assessed by flow cytometry analysis. All patients were followed until death or end of the study (June 2020). Forty-two eligible patients were enrolled between November 2014 and May 2015. The patients’ characteristics are summarized in Table 1. Their median age was 60 years. Eight (19%) patients had a high-risk cytogenetic profile. Table 1. Patients’ characteristics.

demographic

and

baseline

Characteristic

N=42

Gender: male/female, N

21/21

Median age, years (range)

disease

60 (43-66)

ECOG PS, N (%) 0, 1, 2

23 (55), 15 (36), 4 (9)

Isotype, N (%) IgG, IgA, light chain only

27 (64), 9 (22), 6 (14)

ISS stage, N (%) I, II, III

12 (29), 23 (55), 7 (17)

Median creatinine, μmol/L, (range) Cytogenetic risk profile, N (%) High-risk* Standard Stem-cell collection Median CD34+ cell yield (x 106/kg)

72 (48-134) 8 (19) 34 (81) 7,2 (1.4-14.6).

*High-risk cytogenetics was defined by the presence of t(4;14) (with a positive cut-off at 30%) and/or 17p deletion (with a positive cut-off at 50%). ECOG: Eastern Cooperative Oncology Group; PS: performance status; Ig: immunoglobulin; ISS: International Staging System.

Haematologica | 107 July 2022

1693

LETTER TO THE EDITOR The patients’ disposition in the study program is summarized in Online Supplementary Figure S1. Overall, 40 (95%) patients completed induction and 37 (88%) underwent ASCT. Stem cell collection failed in one patient. Plerixafor was needed for stem cell mobilization in five patients. By the end of induction (n=42), the overall response rate was 80% (n=33), including 30% (n=10) very good partial responses and 12% (n=5) CR/sCR (Figure 1A). At the end of consolidation (primary endpoint) (n=37), the sCR rate was 41% (33% in an intention-to-treat analysis). The median time to partial response and CR was 1 and 8 months, respectively. As of June 2020, the median follow-up from the start of therapy was 62.6 months. Twenty-nine patients had progressive disease and seven patients died due to myeloma progression. The median progression-free survival was 41.8 months (95% confidence interval: 33.262) and the 3-year overall survival was 92.8% (95% confidence interval: 85.3-100) (Figure 1B, C). There were no IRD-related deaths. Overall, seven (16.6%) patients discontinued treatment permanently due to treatment-related toxicity: one patient during induction (skin rash), three during consolidation (1 skin rash, 2 thrombocytopenia) and three patients during maintenance (colon cancer, thrombocytopenia, pneumonia). For these patients, the median time to ixazomib discontinuation was 227 days. Overall, 33 (79%) patients had at least one dose modification of one of the study drugs. A dose reduction of ixazomib, lenalidomide or dexamethasone

occured in 60%, 67% and 29% of patients, respectively. Adverse events reported for at least 10% of patients are described in Table 2. During induction, grade 3-4 neutropenia was the most frequent treatment-related adverse event, occurring in eight (19%) patients. Skin rash was reported in 12 (29%), including 5% with grade 3-4. During consolidation, neutropenia and thrombocytopenia were the most frequent adverse events, with grade 3-4 events occuring in 14 (38%) and eight (22%) patients, respectively. During maintenance, thrombocytopenia and lung infection were the most frequent adverse events, occurring in ten (32%) and 12 (39%) patients, respectively. Grade 1-2 sensory peripheral neuropathy was reported in 12 (29%) patients, including two with grade 2 peripheral neuropathy. Deep-vein thrombosis occurred in one patient. The primary objective of this phase II study was to evaluate the efficacy of a transplant program with the oral triplet IRD as induction and consolidation in NDMM patients. In the intention-to-treat population (n=42), the overall response rate was 92.3%, including 70.3% with a very good partial response or better. Our study showed that responses continuously deepened throughout the program. At the completion of extended consolidation, the per protocol CR/sCR rate was 44% (37% in intention-to-treat analysis). These response rates are close to those obtained with VRD as the induction/consolidation regimen in the IFM-2009 and GEM2012 trials.2,5 However, patients

Figure 1. Efficacy of the all-oral ixazomib, lenalidomide, and dexamethasone transplant program. (A) Response rate. (B) Progression-free survival. (C) Overall survival. The median duration of the follow-up was estimated using the reverse Kaplan-Meier method. Progression-free survival was calculated as the time from the start of treatment to the first documentation of progressive disease, or death if the patient died due to any cause before progression. Overall survival was calculated as the time from the start of treatment to death. The Kaplan-Meier method was used to estimate the survival distribution. All analyses were conducted using R version 4.0. ASCT: autologous stem cell transplantation; CR: complete remisison; sCR: stringent complete remission; VGPR: very good partial response; PR: partial response. ITT: intention-to-treat; PP: per protocol. Haematologica | 107 July 2022

1694

LETTER TO THE EDITOR Table 2. Adverse events reported through induction, consolidation and maintenance.

Induction (N=42) Adverse event

Consolidation (N=37)

Maintenance (N=31)

Any grade Patients (%)

Grade 3/4 Patients (%)

Any grade Patients (%)

Grade 3/4 Patients (%)

Any grade Patients (%)

Grade 3/4 Patients (%)

Neutropenia

8 (19)

14 (38)

2 (6)

Thrombocytopenia

1 (2)

11 (30)

8 (22)

10 (32)

7 (23)

Constipation

8 (19)

7 (19)

1 (3)

Diarrhea

8 (19)

6 (16)

1 (3)

4(13)

Nausea

10 (24)

6 (16)

2 (5)

5(16)

Pneumonia/bronchitis

8 (19)

5 (12)

8 (22)

12 (39)

Skin rash

12 (29)

2 (5)

10 (27)

3(9)

Peripheral neuropathy

6 (13)

8(22)

1(3)

6(19)

Hematologic

Non-hematologic

Safety was monitored until 30 days after the last dose of study drug, except for secondary malignancies (monitored continuously during follow-up). Toxicities were graded according to the National Cancer Institute Common Toxicity Criteria of Adverse Events (version 4.03; Bethesda, MD, USA).

in the present study received a higher number of cycles (induction, n=3; early consolidation, n=2; late consolidation, n=6) in comparison with patients from IFM-2009 (5 cycles of VRD) or GEM2012 (8 cycles of VRD). In the present study, patients received ixazomib maintenance for 1 year, with no significant improvement in CR/sCR rates. The phase III, placebo-controlled TOURMALINE-MM3 trial demonstrated a modest progression-free survival benefit in NDMM patients receiving post-ASCT ixazomib maintenance for 2 years (26.5 vs. 21.3 months at the start of maintenance).8 After a median follow-up of nearly 5 years, the median progression-free survival observed in the present study was 41.8 months with a 3-year overall survival of 92.8%. Continuous ixazomib therapy following ASCT in combination with lenalidomide and dexamethasone maintenance has been evaluated (and compared with lenalidomide and dexamethasone) in the randomized phase III trial GEM2014 (n=332). After a median follow-up of 56 months, the addition of ixazomib did not result in a progression-free survival benefit.9 At the time the study was designed, continuous lenalidomide maintenance after ASCT did not demonstrate a benefit on overall survival and was not appoved.10 In the present study, a fixed-duration maintenance with ixazomib alone appears to be a suboptimal approach for transplant-eligible NDMM patients. Safety was an important objective of the present phase II trial. The strategy was feasible, with seven (16.6%) patients discontinuing therapy due to treatment toxicity and no IRD-related mortality. Overall, 33 (79%) patients had at least one dose modification of one study drug. The hematologic toxicities were predictable and manageable.

The most common hematologic toxicity related to IRD was thrombocytopenia with grade >3 occuring in 29 (69%) patients. Thrombocytopenia related to the IRD combination was expected and has been described previously.6,8,11 Considering non-hematologic toxicities, skin rash occurred in 23 (54%) patients, with only two (5%) grade 3/4 adverse events. Grade ≥2 peripheral neuropathy occurred in two patients and one patient had grade ≥3. These results compare favorably with those of VRD strategies with a rate of grade >3 peripheral neuropathy of 12% and 4% in the IFM-2009 and GEM2012 studies, respectively.2,5 The triplet combination of carfilzomib, lenalidomide and dexamethasone (KRD) with transplantation demonstrated strong efficacy results but is associated with substantial cardiac events.12 In the present study, no patient developed treatment-related cardiac failure. To conclude, a transplant program with all-oral IRD as induction and consolidation, followed by 1 year of maintenance with ixazomib is effective in NDMM patients and has a favorable safety profile. However, these results are inferior, with respect to progression-free survival, to those achieved with VRD ASCT and lenalidomide maintenance. To date, ixazomib-based combinations have failed to significantly improve the outcome of transplant-eligible patients with NDMM.8,9 This suboptimal efficacy can be partially explained by inferior in vitro proteasome inhibition with ixazomib in comparison with other proteasome inhibitors.13 In NDMM patients, ixazomib could however be suitable for a specific subset of frail patients with comorbidities (e.g., pre-existing neuropathy, cardiac insufficiency), thought not to be able to tolerate bortezomib or carfilzomib-based combinations. Recently, daratu-

Haematologica | 107 July 2022

1695

LETTER TO THE EDITOR mumab in combination with bortezomib, thalidomide and dexamethasone was approved for transplant-eligible NDMM patients and is now considered as a standard of care.1,14 The phase II randomized study GRIFFIN also demonstrated strong efficacy results (without a safety signal) with daratumumab in combination with VRD in transplant-eligible NDMM patients.15 One way of improving IRD efficacy in the context of transplantation could be the addition of anti-CD38 antibody. The convenience and efficacy profile of IRD in transplant-eligible NDMM patients led to the design of the IFM phase II study 2018-01 (NCT03669445) to evaluate the efficacy and safety of IRD in combination with daratumumab in such patients. Based on their efficacy/safety profiles, bortezomib or carfilzomib-based induction regimens with anti-CD38 should be considered as a standard of care for transplant-eligible NDMM patients.

Correspondence: CYRILLE TOUZEAU - cyrille.touzeau@chu-nantes.fr https://doi.org/10.3324/haematol.2021.280394 Received: December 6, 2021. Accepted: February 9, 2022. Prepublished: February 17, 2022. Disclosures CT, AP, LK, CJ, MM, CH, XL and PM have received honoraria from Takeda, Amgen, BMS/Celgene and Janssen; MLC has received honoraria from Celgene and Takeda. Contributions The following centers and investigators from the IFM participated in this study: Dijon, Centre Hospitalier du Bocage (D. Caillot, J.N. Bastié, I. Lafon, M.L. Chrétien, E. Ferrant); Lille, Hôpital Claude Huriez (E. Boyle, T. Facon, S. Manier, X. Leleu, C. Herbaux); Nantes, Hôpital Hôtel Dieu (P. Moreau, N. Blin, P. Chevallier, J. Delaunay, V. Dubruille, T. Gastinne, T. Guillaume, S. Le Gouill, B. Mahé, C.

Authors

Touzeau); Paris, Hôpital St Antoine (M. Mohty, A. Verkhoff); Toulouse, IUC Oncopole (M. Attal, M. Roussel, B. Hebraud); Tours, Hôpital Bretonneau (L. Benbouker, C. Dartigeas, M. Ertault de la

Cyrille Touzeau,

1,2,3

Aurore Perrot, Murielle Roussel, Lionel Karlin, 4

Bretonnière, E. Gyan, S. Lissandre, H. Monjanel, M. Renaud, A. Iltis);

Lotfi Benboubker, Caroline Jacquet, Mohamad Mohty, Thierry

Lyon, Centre Hopitalier de Lyon Sud (L. KArlin, E. Bachy; F.

Facon,10 Salomon Manier,10 Marie-Lorraine Chretien,11 Mourad Tiab,12

Bouafia-Sauvy, D. Espinouse, L. Lebras, A-S Michallet, C. Sarkozy,

Cyrille Hulin,13 Xavier Leleu,14 Hervé Avet Loiseau,4 Thomas Dejoie,15

F. Broussais-Guillaumot, B. Coiffier, G. Salles); Bordeaux Hopital

Lucie Planche, Michel Attal and Philippe Moreau

Haut Lévêque (C.Hulin, G. Marit, A. Lascaux, S.Dimicoli-Salazar, M.

1,2,3

Robles); Nancy, CHU Brabois (A. Perrot, C. Bonmati, D. Ranta, L. Service d’Hématologie, Centre Hospitalier Universitaire (CHU) Hotel

Clement-Filiatre, G. Roth Guepin, M. D’Aveni); La Roche Sur Yon

Dieu, Nantes; CRCINA, INSERM, CNRS, Université d’Angers, Université

CHD les Oudairies (M. Tiab, H. Maisonneuve, B.Villemagne, M.

de Nantes, Nantes; Site de Recherche Intégrée sur le Cancer (SIRIC)

Voldoire); Paris, Hopital St Louis (J.P. Fermand, B. Arnulf, M. Baron,

«ILIAD», INCA-DGOS-INSERM 12558, Nantes; CHU de Toulouse, IUCT-

M. Malphettes, B. Asli).

O, Université de Toulouse, UPS, Service d’Hématologie, Toulouse; Acknowledgments

Hématologie Clinique et Thérapie Cellulaire, CHU Limoges; 6Hôpital

Lyon Sud, Pierre-Benite; Service Hématologie et Thérapies

The authors gratefully acknowledge the work performed by

Cellulaires, CHRU Bretonneau, Tours; Service d’Hématologie, CHU

individual research teams at all participating study sites; they are

Nancy, Vandoeuvre-lès-Nancy; Hématologie Clinique et Thérapie

also indebted to Dr C. Mathiot, Dr C. Boccaccio, and L. Biron for

Cellulaire, Hôpital Saint Antoine, Sorbonne Université, INSERM UMRs

their administrative and material support.

938, Paris; Maladies du Sang, CHRU de Lille, Lille; Hématologie 10

Clinique, CHU Dijon Bourgogne, Dijon; 12Service d’Hématologie, Centre

Funding

Hospitalier Departemental, La Roche sur Yon; 13Service

The study was funded by Takeda Pharmaceuticals.

d’Hématologie, Hôpital Haut-Lévêque, CHU de Bordeaux, Pessac; Service d’hématologie, CHU de Poitiers, Portiers; 15Service de

Data-sharing statement

Biochimie, CHU Hotel Dieu, Nantes and Département de Recherche

The authors will make original data and the study protocol

Clinique, CHU Hotel Dieu, Nantes, France.

available to other investigators without unreasonable restrictions.

References 1. Dimopoulos MA, Moreau P, Terpos E, et al. Multiple myeloma: EHA-ESMO clinical practice guidelines for diagnosis, treatment and follow-up. HemaSphere. 2021;5(2):e528. 2. Attal M, Lauwers-Cances V, Hulin C, et al. Lenalidomide, bortezomib, and dexamethasone with transplantation for myeloma. N Engl J Med. 2017;376(14):1311-1320.

3. Cavo M, Gay F, Beksac M, et al. Autologous haematopoietic stemcell transplantation versus bortezomib–melphalan–prednisone, with or without bortezomib–lenalidomide–dexamethasone consolidation therapy, and lenalidomide maintenance for newly diagnosed multiple myeloma (EMN02/HO95): a multicentre, randomised, open-label, phase 3 study. Lancet Haematol.

Haematologica | 107 July 2022

1696

LETTER TO THE EDITOR 2020;7(6):e456-e468. 4. Gay F, Musto P, Rota-Scalabrini D, et al. Carfilzomib with cyclophosphamide and dexamethasone or lenalidomide and dexamethasone plus autologous transplantation or carfilzomib plus lenalidomide and dexamethasone, followed by maintenance with carfilzomib plus lenalidomide or lenalidomide alone for patients with newly diagnosed multiple myeloma (FORTE): a randomised, open-label, phase 2 trial. Lancet Oncol. 2021;22(12):1705-1720. 5. Rosiñol L, Oriol A, Rios R, et al. Bortezomib, lenalidomide, and dexamethasone as induction therapy prior to autologous transplant in multiple myeloma. Blood. 2019;134(16):1337-1345. 6. Moreau P, Masszi T, Grzasko N, et al. Oral ixazomib, lenalidomide, and dexamethasone for multiple myeloma. N Engl J Med. 2016;374(17):1621-1634. 7. Kumar S, Paiva B, Anderson KC, et al. International Myeloma Working Group consensus criteria for response and minimal residual disease assessment in multiple myeloma. Lancet Oncol. 2016;17(8):e328-346. 8. Dimopoulos MA, Gay F, Schjesvold F, et al. Oral ixazomib maintenance following autologous stem cell transplantation (TOURMALINE-MM3): a double-blind, randomised, placebocontrolled phase 3 trial. Lancet. 2018;393(10168):253-264. 9. Rosinol L, Oriol A, Ríos Tamayo R, et al. Ixazomib plus lenalidomide/dexamethasone (IRd) versus lenalidomide /dexamethasone (Rd) maintenance after autologous stem cell transplant in patients with newly diagnosed multiple myeloma: results of the Spanish GEM2014MAIN trial. Blood. 2021;138(Suppl 1):466.

10. McCarthy PL, Holstein SA, Petrucci MT, et al. Lenalidomide maintenance after autologous stem-cell transplantation in newly diagnosed multiple myeloma: a meta-analysis. J Clin Oncol. 2017;35(29):3279-3289. 11. Dimopoulos M, Laubach J, Gutierrez M, Richardson PG. Efficacy and safety of long-term ixazomib maintenance therapy in patients (pts) with newly diagnosed multiple myeloma (NDMM) not undergoing transplant: an integrated analysis of four phase 1/2 studies. Blood. 2017;130(Suppl 1):902. 12. Roussel M, Lauwers-Cances V, Robillard N, et al. Frontline therapy with carfilzomib, lenalidomide, and dexamethasone (KRd) induction followed by autologous stem cell transplantation, Krd consolidation and lenalidomide maintenance in newly diagnosed multiple myeloma (NDMM) patients: primary results of the Intergroupe Francophone Du MyéLome (IFM) Krd phase II study. Blood. 2016;128(22):1142. 13. Besse A, Besse L, Kraus M, et al. Proteasome inhibition in multiple myeloma: head-to-head comparison of currently available proteasome inhibitors. Cell Chem Biol. 2019;26(3):340-351. 14. Moreau P, Attal M, Hulin C, et al. Bortezomib, thalidomide, and dexamethasone with or without daratumumab before and after autologous stem-cell transplantation for newly diagnosed multiple myeloma (CASSIOPEIA): a randomised, open-label, phase 3 study. Lancet. 2019;394(10192):29-38. 15. Voorhees PM, Kaufman JL, Laubach JP, et al. Daratumumab, lenalidomide, bortezomib, & dexamethasone for transplanteligible newly diagnosed multiple myeloma: GRIFFIN. Blood. 2020;136(8):936-945.

Haematologica | 107 July 2022

1697

LETTER TO THE EDITOR

Complement C1s inhibition with sutimlimab results in durable response in cold agglutinin disease: CARDINAL study 1-year interim follow-up results Sutimlimab (BIVV009; TNT009) is a humanized monoclonal antibody designed to target C1s, the C1 complex serine protease responsible for activating the classical complement pathway which triggers hemolysis in cold agglutinin disease (CAD).1 Results from the pivotal CARDINAL study showed sutimlimab to be an efficacious and well-tolerated therapy during 26 weeks of treatment among CAD patients.2 We present 1-year interim results of the ongoing 2-year CARDINAL extension which show that sutimlimab has a sustained treatment effect in CAD via long-term complement inhibition. Sutimlimab maintained mean hemoglobin levels ≥11 g/dL with sustained improvement in quality of life; no new safety concerns were identified. CAD is a rare autoimmune hemolytic anemia characterized by chronic hemolysis mediated entirely by activation of the classical complement pathway.3 CAD is a low-grade clonal lymphoproliferative disorder with no underlying overt malignancy or infection.4 Clinical manifestations of CAD include classical complement pathway-mediated chronic hemolytic anemia, profound fatigue, acute hemolytic crises, as well as transient, cold-induced, red blood cell agglutination-mediated circulatory symptoms including acrocyanosis.5 There are currently no approved therapies for CAD.6 Unapproved pharmacological modalities offer varying efficacy and safety, as well as different response rates and response durations.69 Rituximab depletes B cells and induces a partial response in approximately 50% of patients, with a median delay of 1.5 months and relapse within 1 year. The addition of cytotoxic agents (bendamustine or fludarabine) to rituximab, although associated with increased response rates, is accompanied by more serious toxicity, including severe neutropenia.8,9 Blood transfusions are reserved for acute hemolytic anemia and are only a transient temporizing measure; other treatment is required to restrain CAD-associated hemolysis, to which transfused red blood cells are also subjected.4 Thus, an alternative treatment that is non-cytotoxic, rapid, and durable in controlling classical complement-mediated hemolysis and its clinical manifestations in CAD patients is needed. CARDINAL is a prospective, open-label, single-arm, multicenter trial comprising 16 sites from eight countries (ClinicalTrials.gov identifier: NCT03347396). This two-part study had a 26-week treatment period (part A; completed 11 July, 2019) and an ongoing extension (part B) for 2 years after the last patient had completed part A. Data for the combined study period up to a minimum of 53 weeks of follow-up for

all ongoing patients are presented here (data cut: 16 January, 2020). Patients ≥18 years of age with a confirmed diagnosis of CAD, baseline hemoglobin ≤10 g/dL, and a history of recent transfusion (≥1 blood transfusions in the preceding 6 months) were enrolled.2 Patients were treated with sutimlimab intravenously on days 0 and 7, followed by biweekly dosing for 2 years. Patients weighing <75 kg or ≥75 kg received sutimlimab 6.5 g or 7.5 g fixed dose, respectively. The complete study design and results of part A were reported previously.2 Patients had to be vaccinated against encapsulated bacterial pathogens (Neisseria meningitis, including serogroup B meningococcus, Haemophilus influenzae, and Streptococcus pneumoniae) within 5 years before enrollment. Efficacy endpoints for part B included a change from baseline in hemoglobin levels, hemolytic markers (total bilirubin), blood transfusions up to 53 weeks, and quality of life assessed using the Functional Assessment of Chronic Illness Therapy (FACIT)-Fatigue Scale up to 51 weeks (last data recording within the 1-year treatment period). Safety endpoints included incidence of treatmentemergent adverse events (TEAE) and treatment-emergent serious adverse events (TESAE) as well as changes in systemic lupus erythematosus panel parameters up to 53 weeks. Of 42 patients screened, 24 patients were enrolled and received one or more doses of sutimlimab in part A.2 Twentytwo patients (91.7%) completed part A and entered part B. Most patients were female (62.5%) and ≥65 years of age (79.2%). Patients had reduced mean (standard deviation [SD]) hemoglobin 8.6 (1.6) g/dL and elevated mean (SD) bilirubin 53.3 (24.0) µmol/L levels at baseline.2 After the first sutimlimab dose, mean (standard error [SE]) hemoglobin levels improved rapidly from baseline by 1.2 (0.3) g/dL within the first week, improved by 2.3 (0.3) g/dL during the third week, and were maintained thereafter at ≥11 g/dL from week 5 through week 53 (Figure 1). Overall, there was a sustained increase in hemoglobin level of ≥2 g/dL from week 3 to week 53, and 55.0% of patients (11/22) had normalized hemoglobin level (≥12 g/dL) at week 53. Mean total bilirubin decreased rapidly by week 1 after sutimlimab, was normalized (<upper limit of normal: 20.5 µmol/L) by week 3 and remained normalized through to week 53 (Figure 1). Normalization of bilirubin level (<20.5 µmol/L) was achieved in 63.6% of patients (14/22) at week 53, at which time the mean (SE) change in bilirubin from baseline was a reduction of −35.3 (4.2) µmol/L. Seventeen (70.8%) and 19 (86.4%) pa-

Haematologica | 107 July 2022

1698

LETTER TO THE EDITOR tients remained transfusion-free from week 5 to 26 (part A) and week 27 to 53 (part B), respectively (Online Supplementary Table S1). The baseline mean (SE) FACIT-Fatigue score of 32.5 (2.3) points indicated that quality of life was significantly affected (Figure 1). Following sutimlimab treatment, mean (SE) FACIT-Fatigue score increased to 39.7 (1.8) by week 1 (a 7point improvement) and remained >40 from weeks 3 to 51

(last data recording). Overall, there was a mean (SE) improvement of 10.4 (3.5) points by week 26, which was sustained at 11.4 (2.8) points at week 51, consistent with a clinically meaningful improvement (≥3-point increase).10 Sutimlimab led to near-complete inhibition of classical complement pathway activity. Mean (SE) classical complement pathway activity declined from 20.0% (3.4%) at baseline to 3.0% (0.7%) at week 25 and 3.0% (0.7%) at week 51, along-

Figure 1. Mean hemoglobin, total bilirubin, and FACIT-Fatigue score from baseline to weeks 51–53 after sutimlimab treatment in patients with cold agglutinin disease (full analysis set). Three patients with a presumed Gilbert’s syndrome test result were excluded from the total bilirubin data (n = 21). The range of the FACIT-Fatigue score was 0–52. Patients with a completed questionnaire were included (a questionnaire was completed if ≥7 of 13 items were answered by the patient). FACIT-Fatigue data were not collected at week 53. The number of patients varied by study visit and by analyte in this interim analysis as the study is ongoing. The data cutoff date was 16 January, 2020. The upper limit of normal (ULN) bilirubin level was defined as 20.5 µmol/L. Bsl: baseline; FACIT: Functional Assessment of Chronic Illness Therapy; Hb: hemoglobin; SE: standard error.

Figure 2. Mean classical complement pathway activity from baseline up to 51 weeks after sutimlimab treatment in patients with cold agglutinin disease (full analysis set). Classical pathway activity and C4 levels are shown. The number of patients varied by study visit and by analyte in this interim analysis as the study is ongoing. The data cutoff date was 16 January, 2020. Classical pathway activation was determined using an enzyme-linked immunosorbent assay that measures the functional capacity of the classical pathway (Wieslab® classical complement pathway assay; normal range in serum is 69–129%.). The standard international reference range for serum C4 is 0.18 to 0.45 g/L. Pharmacodynamic assessments were performed at 3-month intervals during the first year of treatment in part B and then at 6-month intervals. Bsl: baseline; SE: standard error. Haematologica | 107 July 2022

1699

LETTER TO THE EDITOR side normalization of complement C4 levels (defined as <normal range: 0.18–0.45 g/L). Mean (SE) total C4 level at baseline was 0.04 (0.02) g/L, at week 25 it was 0.29 (0.02) g/L, and at week 51 it was 0.24 (0.11) g/L (Figure 2). This coincided with improvements seen in hemoglobin, bilirubin, and FACIT-Fatigue score. From baseline to week 53, all 24 patients experienced ≥1 TEAE; nine patients (37.5%) had ≥1 TEAE related to sutimlimab (Table 1), with the most common being acrocyanosis/cyanosis and infusion-related reaction (two

patients each [8.3%]). Of the 281 TEAE, the most common were diarrhea and nasopharyngitis (20.8% each), and anemia, nausea, and hypertension (16.7% each). In part B, one event of device-related thrombosis was reported and considered by the investigator related to the indwelling catheter and not related to sutimlimab. Overall, 57 infections were reported as TEAE in all 24 patients and nine TEAE grade ≥3 infections were reported in six patients (25.0%). Four TEAE (grade 1 or 2) suggestive of potential hypersensitivity to sutimlimab were reported in three patients; all resolved and

Table 1. Summary of treatment-emergent adverse events (safety analysis set).

Parts A and B Total (N = 24) TEAE, N Patients with ≥1 TEAE, N (%) Most common TEAE (>10%) Diarrhea, N (%) Nasopharyngitis, N (%) Anemia, N (%) Hypertension, N (%) Nausea, N (%) Confusional state, N (%) Constipation, N (%) Contusion, N (%) Cough, N (%) Cyanosis, N (%) Cystitis, N (%) Dizziness, N (%) Fatigue, N (%) Gastroenteritis, N (%) Hemorrhoids, N (%) Upper abdominal pain, N (%) Upper respiratory tract infection, N (%)

281 24 (100.0)

Patients with ≥1 related TEAE,* N (%) Total number of related TEAE, N Most common related TEAE (>1 patient) Acrocyanosis/cyanosis, N (%) Infusion-related reaction, N (%)

9 (37.5) 25†

Patients with ≥1 TEAE grade 3 or higher, N (%)

14 (58.3)

Patients with ≥1 TEAE infection grade 3 or higher, N (%) Total number of TEAE infections, N Total number of TESAE infections, N Total number of TEAE infection grade 3 or higher, N

6 (25.0) 57 9 9

TESAE, N Patients with ≥1 TESAE, N (%) Patients with ≥1 related TESAE,* N (%)

30 12 (50.0) 1 (4.2)

5 (20.8) 5 (20.8) 4 (16.7) 4 (16.7) 4 (16.7) 3 (12.5) 3 (12.5) 3 (12.5) 3 (12.5) 3 (12.5) 3 (12.5) 3 (12.5) 3 (12.5) 3 (12.5) 3 (12.5) 3 (12.5) 3 (12.5)

2 (8.3) 2 (8.3)

Total number of TEAE thromboembolic events, N Patients with ≥1 TESAE thromboembolic event, N (%)

1‡ 1 (4.2)

Patients who discontinued treatment and/or study owing to a TEAE, N (%)

2 (8.3)§

Deaths, N (%)

1 (4.2)‖

Data cutoff date was 16 January; 2020.2 *Adverse events with missing causality assessment were included in the related treatment-emergent adverse events (TEAE)/treatment-emergent serious adverse events (TESAE); adverse events with investigator causality assessment of “possible” or “probable” were considered related. †Events comprised pain in both hands/legs related to acrocyanosis/cyanosis, dysphagia, application-site hemorrhage, fatigue, peripheral edema, temperature intolerance, cystitis, upper respiratory tract infection, viral infection, contusion, infusion-related reaction, hypertension, arthralgia, tendonitis, dyspnea, rhinorrhea, and erythema. ‡Deep vein thrombosis in the right arm due to an indwelling catheter. §One treatment/study discontinuation due to the patient’s death (as per footnote‖); one treatment/study discontinuation due to multiple non-serious TEAE, including acrocyanosis due to cold agglutinin disease and dysphagia, which were both assessed as related to sutimlimab. A third patient discontinued treatment/study because of a pretreatment serious adverse event of polymyalgia rheumatica. ‖One patient died of progressive carcinoma (unrelated to study treatment). Haematologica | 107 July 2022

1700

LETTER TO THE EDITOR did not recur, and none led to study discontinuation. Three patients experienced TEAE of acrocyanosis/cyanosis, of whom one had a history of Raynaud’s phenomenon. One patient had a TESAE of disabling circulatory symptoms with livid foot discoloration, not considered related to sutimlimab treatment by the investigator. No patients developed systemic lupus erythematosus, nor were there any TEAE consistent with the development of other autoimmune diseases. Thirty TESAE were reported in 12 patients (50%). Serious infections with encapsulated bacteria (Streptococcus pyogenes, Streptococcus pneumoniae, Escherichia coli, and Staphylococcus species) were reported, but no meningococcal infections or TEAE of meningitis were identified. No patient discontinued the study due to an infection. Two patients discontinued the treatment and/or study during the 53 weeks for reasons unrelated to study treatment; post-treatment hemolytic parameters were not collected for these patients. One treatment/study discontinuation occurred due to multiple non-serious TEAE, including acrocyanosis and dysphagia; 9 weeks after treatment, hemoglobin levels remained stable, however all other hemolytic parameters assessed were not within normal range; FACIT-Fatigue score was 7 points below baseline. The 1-year interim CARDINAL study follow-up further demonstrates that continued classical complement pathway inhibition with sutimlimab results in sustained remission of hemolysis in CAD patients, with durably increased hemoglobin levels, normalized bilirubin levels, and improved FACIT-Fatigue scores. After sutimlimab treatment, the number of blood transfusions decreased with time and most patients remained transfusion-free from weeks 5 to 53, which translates into reduced patient burden and utilization of healthcare resource. Patients with CAD in this study had comparable baseline mean FACIT-Fatigue scores to patients with other serious chronic conditions, including rheumatoid arthritis, advanced cancer-related anemia, and paroxysmal nocturnal hemoglobinuria.11-14 Within 1 week of sutimlimab treatment, there was a mean increase of 7.24 points from baseline in the FACIT-Fatigue score, which was improved further and sustained throughout the follow-up period, indicating improved quality of life. These improvements coincided with reduced classical complement pathway activity. In addition to anemia driven by hemolytic activity, inflammation associated with classical complement pathway activation may be a key driver of fatigue in patients with CAD.15 In summary, 1-year interim results of the ongoing CARDINAL study demonstrate continued inhibition of the classical complement pathway at C1s with sutimlimab and sustained treatment effects in CAD. Sutimlimab demonstrated an acceptable safety profile at 1 year; no new safety signals were identified. No TEAE suggestive of serious hypersensitivity or anaphylactic reactions associated

with sutimlimab were identified. Other than one devicerelated thrombosis, no other vascular thromboembolic TEAE were reported. These data reinforce the positive riskbenefit profile of sutimlimab as an effective long-term therapy with an acceptable safety profile for management of patients with chronic CAD, particularly with symptoms influenced predominantly by activation of the classical complement pathway (e.g., chronic hemolysis, anemia, and fatigue).

Authors Alexander Röth,1 Wilma Barcellini,2 Shirley D’Sa,3 Yoshitaka Miyakawa,4 Catherine M. Broome,5 Marc Michel,6 David J. Kuter,7 Bernd Jilma,8 Tor Henrik Anderson Tvedt,9 Ilene C. Weitz,10 Parija Patel,11 Xiaoyu Jiang,11 Caroline Reuter,11 Jun Su,11 Frank Shafer,11 Michelle Lee11 and Sigbjørn Berentsen12 Department of Hematology and Stem Cell Transplantation, West

German Cancer Center, University Hospital Essen, University of Duisburg-Essen, Essen, Germany; 2Fondazione IRCCS Ca’ Granda Ospedale Maggiore Policlinico, Milan, Italy; 3UCLH Centre for Waldenström’s Macroglobulinemia and Related Conditions, University College London Hospitals NHS Foundation Trust, London, UK; Thrombosis and Hemostasis Center, Saitama Medical University

Hospital, Saitama, Japan; 5Division of Hematology, MedStar Georgetown University Hospital, Washington, DC, USA; 6Henri-Mondor University Hospital, Assistance Publique-Hôpitaux de Paris, UPEC, Créteil, France; 7Division of Hematology, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA; 8Department of Clinical Pharmacology, Medical University of Vienna, Vienna, Austria; 9Section for Hematology, Department of Medicine, Haukeland University Hospital, Bergen, Norway; 10Keck School of Medicine of USC, Los Angeles, CA, USA; 11Sanofi, Cambridge, MA, USA and 12Department of Research and Innovation, Haugesund Hospital, Haugesund, Norway Correspondence: ALEXANDER RÖTH - alexander.roeth@uk-essen.de https://doi.org/10.3324/haematol.2021.279812 Received: August 19, 2021. Accepted: February 9, 2022. Prepublished: February 17, 2022. Data were first presented at the 62nd American Society of Hematology Virtual Annual Congress, 5–8 December, 2020. Disclosures AR has received research support from Roche, received honoraria, and provided consultancy to Alexion Pharmaceuticals, Inc., Apellis Pharmaceuticals, Novartis, Roche, Kira, Bioverativ, a Sanofi company, Sanofi and Sobi. SD has received grant funding, honoraria, and/or speaker’s fees from Janssen, BeiGene, and Sanofi. YM has provided

Haematologica | 107 July 2022

1701

LETTER TO THE EDITOR consultancy for Bioverativ and Sanofi. WB has received research

Contributions

support from Alexion Pharmaceuticals, Inc., and Novartis; participated

AR, FS, PP, ML, JS, and CR analyzed/reviewed data. XJ reviewed the

in advisory boards for Agios, Alexion Pharmaceuticals, Inc., Bioverativ,

results of the statistical analysis. All authors had access to primary

and Incyte; and has been an invited speaker for Alexion

clinical trial data, had full editorial control of the manuscript, and

Pharmaceuticals, Inc., and Novartis. CMB has received honoraria

provided their final approval of all content.

and/or research funding from Alexion Pharmaceuticals, Inc., Bioverativ, Cellphire, Incyte, Rigel, and Sanofi Genzyme. MM has provided

Acknowledgments

consultancy to Alexion Pharmaceuticals, Inc., Rigel, and Bioverativ.

We would like to thank the investigators, healthcare providers,

DJK has provided consultancy to Actelion (Syntimmune), Agios,

research staff, and patients who participated in the CARDINAL study.

Alnylam, Amgen, Argenx, Bristol-Myers Squibb, Caremark, Daiichi

The authors are grateful for the assistance provided by Jennifer Wang

Sankyo, Dova, Kyowa Kirin, Merck Sharp & Dohme, Momenta, Novartis,

and Katarina Kralova of Sanofi.

Pfizer, Platelet Disorder Support Association, Principia, Protalex, Protalix Biotherapeutics, Rigel, Sanofi, Shionogi, Shire, Takeda

Funding

(Bioverativ), UCB, Up-To-Date, and Zafgen; and received research

This study was funded by Sanofi (Waltham, MA, USA). Medical writing

funding from Actelion (Syntimmune), Agios, Alnylam, Amgen, Argenx,

and editing support were provided by Rhutika Dessai and Santo

Bristol-Myers Squibb, Kezar Life Sciences, Inc., Principia, Protalex,

D’Angelo of Fishawack Communications Ltd., part of Fishawack

Rigel, and Takeda (Bioverativ). BJ has received reimbursement for

Health, and by Lisa Buttle and Sally Ratcliffe of Lucid Group, and

travel costs for scientific presentations and consultancy to True North

funded by Sanofi.

Therapeutics, Bioverativ and Sanofi. THAT has participated in advisory boards for Alexion Pharmaceuticals, Inc., Novartis, and Ablynx. ICW

Data-sharing statement

has received honoraria from Alexion Pharmaceuticals Inc. and Sanofi.

Qualified researchers may request access to patient-level data and

XJ, ML, and FS are employees of Sanofi and may hold shares and/or

related study documents such as the clinical study report, study

stock options in the company. PP, JS, and CR were employees of

protocol (with amendments), statistical analysis plan, and dataset

Sanofi at the time of the study. SB has received research funding

specifications. Of note, patient-level data will be anonymized, and

from Mundipharma; lecture honoraria from Apellis Pharmaceuticals,

study documents will be redacted in order to protect the privacy of

Alexion Pharmaceuticals, Inc., Bioverativ, Janssen-Cilag, and Sanofi;

trial participants. Further information related to Sanofi’s data sharing

and provided consultancy to Apellis Pharmaceuticals, Bioverativ,

criteria, eligible studies, and process for requesting access can be

Sanofi, and True North Therapeutics.

found at: https://www.clinicalstudydatarequest.com/.

References 1. Bartko J, Schoergenhofer C, Schwameis M, et al. A randomized, first-in-human, healthy volunteer trial of sutimlimab, a humanized antibody for the specific inhibition of the classical complement pathway. Clin Pharmacol Ther. 2018;104(4):655-663. 2. Röth A, Barcellini W, D'Sa S, et al. Sutimlimab in cold agglutinin disease. N Engl J Med. 2021;384(14):1323-1334. 3. Berentsen S, Ulvestad E, Langholm R, et al. Primary chronic cold agglutinin disease: a population based clinical study of 86 patients. Haematologica. 2006;91(4):460-466. 4. Jäger U, Barcellini W, Broome CM, et al. Diagnosis and treatment of autoimmune hemolytic anemia in adults: recommendations from the First International Consensus Meeting. Blood Rev. 2020;41:100648. 5. Berentsen S. Cold agglutinin disease. Hematology Am Soc Hematol Educ Program. 2016;2016(1):226-231. 6. Jia MN, Qiu Y, Wu YY, et al. Rituximab-containing therapy for cold agglutinin disease: a retrospective study of 16 patients. Sci Rep. 2020;10(1):12694. 7. Röth A, Bommer M, Hüttmann A, et al. Eculizumab in cold agglutinin disease (DECADE): an open-label, prospective, bicentric, nonrandomized phase 2 trial. Blood Adv. 2018;2(19):2543-2549. 8. Berentsen S, Randen U, Vågan AM, et al. High response rate and durable remissions following fludarabine and rituximab combination therapy for chronic cold agglutinin disease. Blood. 2010;116(17):3180-3184. 9. Berentsen S, Randen U, Oksman M, et al. Bendamustine plus

rituximab for chronic cold agglutinin disease: results of a Nordic prospective multicenter trial. Blood. 2017;130(4):537-541. 10. Nordin A, Taft C, Lundgren-Nilsson A, Dencker A. Minimal important differences for fatigue patient reported outcome measures-a systematic review. BMC Med Res Methodol. 2016;16:62. 11. Cella D, Yount S, Sorensen M, Chartash E, Sengupta N, Grober J. Validation of the Functional Assessment of Chronic Illness Therapy Fatigue Scale relative to other instrumentation in patients with rheumatoid arthritis. J Rheumatol. 2005;32(5):811-819. 12. Cella D, Eton DT, Lai JS, Peterman AH, Merkel DE. Combining anchor and distribution-based methods to derive minimal clinically important differences on the Functional Assessment of Cancer Therapy (FACT) anemia and fatigue scales. J Pain Symptom Manage. 2002;24(6):547-561. 13. Escalante CP, Chisolm S, Song J, et al. Fatigue, symptom burden, and health-related quality of life in patients with myelodysplastic syndrome, aplastic anemia, and paroxysmal nocturnal hemoglobinuria. Cancer Med. 2019;8(2):543-553. 14. Schrezenmeier H, Röth A, Araten DJ, et al. Baseline clinical characteristics and disease burden in patients with paroxysmal nocturnal hemoglobinuria (PNH): updated analysis from the International PNH Registry. Ann Hematol. 2020;99(7):1505-1514. 15. Weitz IC, Ueda Y, Shafer F, et al. Inflammation and fatigue in patients with cold agglutinin disease (CAD): analysis from the phase 3 Cardinal study. Blood. 2020;136(suppl 1):7-8, abstract 759.

Haematologica | 107 July 2022

1702

LETTER TO THE EDITOR

Clonal hematopoiesis of indeterminate potential-related epigenetic age acceleration correlates with clonal hematopoiesis of indeterminate potential clone size in patients with high morbidity Clonal hematopoiesis of indeterminate potential (CHIP) is defined as the presence of myeloid cancer-associated, somatic mutations with a variant allele frequency (VAF) of ≥2% in the hematopoietic cells of individuals without hematologic malignancy.1 Its prevalence steeply increases with age, affecting about 15-20% of individuals of ≥60 years, indicating an association with aging processes. DNMT3A, TET2, and ASXL1 are the most frequently mutated genes and can all be classified as epigenetic modifiers.2-4 CHIP is associated with various age-related, adverse health conditions, e.g., cardiovascular diseases and stroke.2 Additionally, CHIP bares the inherent risk of progression to hematologic malignancy.1-3 Together, these adverse effects translate into an increased overall mortality.2,3 There are nine well-accepted hallmarks of aging, which comprise genomic instability and epigenetic alterations.5 As CHIP can be interpreted as genomic instability, matching CHIP state with other established markers of aging is of high interest. DNA methylation patterns change over an individual’s life, allowing the establishment of “epigenetic clocks” that estimate epigenetic age through regression of methylation ratios at predefined CpG sites on chronological age.6-8 Epigenetic age acceleration has been shown to correlate with various age-associated conditions and overall mortality (reviewed by Horvath and Raj6). Due to its interrelatedness with older age and aging-associated conditions, we asked whether the presence of CHIP itself is a surrogate marker of an enhanced biological aging process. We therefore determined epigenetic age measures using an established adaption of the epigenetic clock9 originally published by Vidal-Bralo and colleagues8 in a cohort of 381 individuals with previously determined CHIP status.10 Blood samples were collected from 417 inpatients from the Departments of Cardiology, Nephrology, Musculoskeletal Surgery, and Oncology at Charité – Universitätsmedizin Berlin as reported previously.10 Tumor patients were included before starting (radio-)chemotherapy. Individuals with hematologic malignancy were excluded. The study was conducted in accordance with the Declaration of Helsinki and patients gave informed consent. Whole blood DNA of the study participants was screened for CHIP (VAF ≥2%) as performed and published previously.10 Methylation status of CpG sites of interest

was measured according to a protocol for the methylation-sensitive single-nucleotide-primer-extension method (ms-SNuPE) developed by Vidal-Bralo and colleagues8 and adapted by Vetter and colleagues.9 DNA methylation (DNAm) age was calculated with the 7-CpG clock using a regression model from multiple linear regression of chronological age on methylation fractions of the Berlin Aging Study II cohort. Age acceleration (AA) was determined as the residuals from regressing DNAm age on chronological age. Intrinsic epigenetic age acceleration (IEAA) was determined as the residuals from regressing DNAm age on chronological age and blood cell counts.9 Alpha-level was set at 0.05, and IBM SPSS Statistics software (IBM, USA) was used. Methylation ratios were successfully determined for 381 of 417 individuals (=91%) (Online Supplementary Figure S1A). Mean chronological age was 74.7 years (standard deviation [SD]= ±7.9 years, range, 55-98 years) and 45% of individuals were female. Our cohort was characterized by a high prevalence of clinically relevant morbidities as detailed in Table 1. Of the 381 individuals, 106 (=28%) had evidence of CHIP. Median VAF was 6% (Figure 1A), and an overview of all mutations is presented in Figure 1B and the Online Supplementary Table S2. The mean DNAm age was 76.2 years (±9.7 years), mean DNAm age acceleration was ±8.4 years, and mean IEAA was ±8.1 years. AA (n=381) and IEAA (n=297) were analyzed with regard to baseline and clinical characteristics. In line with existing data6, AA and IEAA were significantly higher in males than in females (AA: P=0.006; IEAA: P=0.001) (Figure 1C). No statistically relevant differences in AA could be noticed with regard to medical conditions, except for hyperuricaemia (AA: P=0.001; IEAA: P<0.001) and kidney failure requiring hemodialysis (AA: P<0.001; IEAA: P=0.018) (Online Supplementary Table S1). In our study, the general presence of CHIP was not associated with age acceleration (AA: P= 0.183; IEAA: P=0.513), nor were individual CHIP mutations (e.g., DNMT3A, TET2, ASXL1), functional CHIP groups or the presence of more than one mutation when compared to individuals without CHIP (Figure 2A to C; Online Supplementary Figure S1A to C). In order to obtain a more differentiated view, we investigated the relationship between CHIP clone size and epigenetic age acceleration. First, patients with CHIP were subdivided into three groups according to clone size.

Haematologica | 107 July 2022

1703

LETTER TO THE EDITOR Table 1. Baseline characteristics of 381 individuals analyzed in our study. Cases are reported as N (%) unless otherwise specified.

Variable Demographics Female Male Age, years (mean [SD]) Age, <61 Age, 61-70 Age, 71-80 Age, >80 Clinical data Malignant tumor Coronary heart disease Cardiac insufficiency Cardiac arrhythmia (including atrial fibrillation) Cerebrovascular disease Peripheral artery disease Arterial hypertension Type 2 diabetes mellitus Hyperlipidemia Obesity (BMI ≥ 30) Hyperuricemia Kidney failure (all) ESRD Chronic obstructive pulmonary disease

Missing data

Cases

CHIP+

0 0 0

172 (45%) 209 (55%) 74.7 (±7.9) 3 (0.8%) 129 (33.9%) 171 (44.9%) 78 (20.5%)

46 (27%) 60 (29%)

69 (18%) 183 (49%) 150 (40%) 193 (51%) 67 (18%) 69 (18%) 307 (81%) 113 (30%) 207 (55%) 87 (23%) 45 (12%) 149 (40%) 39 (10%) 50 (13%)

22 (32%) 47 (26%) 42 (28%) 54 (28%) 17 (25%) 24 (35%) 83 (27%) 32 (28%) 62 (30%) 16 (18%) 15 (33%) 44 (30%) 8 (21%) 12 (24%)

0 4 4 4 4 4 4 4 4 4 4 4 4 4

1 (33%) 18 (14%) 57 (33%) 30 (38%)

SD: standard deviation; BMI: body mass index; ESRD: end stage renal disease. CHIP: clonal hematopoiesis of indeterminate potential.

Clones with a VAF <10% were considered “small to intermediate”, while clones with a VAF of ≥10% were considered “large”. A third group was defined by the absence of CHIP. Using ANOVA, significant differences between the individual groups were detected (P=0.023). This finding remained significant after correction for sex and age using ANCOVA (P=0.022). Differences in AA and IEAA were most pronounced when comparing individuals without CHIP to those with large clones (contrast analysis: P=0.01) (Figure 2D; Online Supplementary Figure S2D). The difference in mean AA amounted to 4.5 years. A comparison of AA between individuals without CHIP and individuals with clones of small to intermediate size showed no differences (P=0.718). Next, we sought to identify to which extent smaller clones still affect AA/IEAA. The lowest VAF with significant differences in AA/IEAA compared to AA/IEAA in individuals without CHIP was ≥5% (KruskalWallis-Test: P=0.008; Mann-Whitney-Test: P=0.008). This effect was likewise significant for IEAA (Kruskal-WallisTest: P=0.016; Mann-Whitney-Test: P=0.044). In order to further explore the relation between CHIP clone size and DNAm AA while accounting for the continuous character of VAF, we performed Spearman’s correlations for DNAm age acceleration measures and VAF of individuals with CHIP. For individuals with more than one CHIP mutation, VAF of the largest clone was used. Our

analysis revealed a statistically significant positive correlation between VAF and AA/IEAA (AA: RS=0.328, P<0.001, IEAA: RS=0.420, P<0.001) (Figure 2E and F). This association remained significant after correction for sex and age (AA: RS=0.323, P<0.001, IEAA: RS=0.411, P<0.001), further supporting a connection between CHIP clone size and accelerated aging processes. Two previous studies consistently described a significant association of CHIP and epigenetic age acceleration based on methylation array data.11,12 However, the whole genome sequencing (WGS) methodology applied for detection of CHIP in these investigations is characterized by low coverage, prohibiting the reliable detection of small to intermediate sized CHIP clones. This is reflected by comparably low CHIP frequencies of about 6% in both studies and no CHIP clones with a VAF <6% were included in Nachun and colleagues.11 However, clones with a VAF of ≥2 and <6% account for a substantial part of CHIP.13 In our cohort, 54% of CHIP carriers had CHIP mutations with VAF solely between 2-6%, which is within the expected frequency range. Our key finding is the correlation of CHIP clone size with epigenetic age acceleration, showing that an increase in clone size is associated with increased epigenetic age acceleration. In contrast to the two previous studies, CHIP clones were identified by targeted sequencing with robust

Haematologica | 107 July 2022

1704

LETTER TO THE EDITOR coverage (average depth: 2,400x), allowing reliable detection of small- and intermediate-sized CHIP clones. We could therefore for the first time demonstrate that a significant epigenetic age acceleration can only be detected with CHIP clones of a VAF of 5% or greater, while smaller clones – contributing to a relevant proportion of CHIP in general – do not seem to be associated with acceleration of the epigenetic age. This association matches the results of multiple studies that report an increasing likelihood of

CHIP-associated adverse conditions with ascending VAF.2,13 To the best of our knowledge, the CpG of the clock implemented here are no major targets of the frequently mutated DNMT3A14 and TET2.15 Accelerated epigenetic age in CHIP carriers can therefore be interpreted as a reflection of actual accelerated biological aging rather than being an epiphenomenon of altered methylation processes caused by mutations. However, large clones might result from a predominant pool of HSC, possibly implicat-

D C

Figure 1. Clonal hematopoiesis of indeterminate potential (CHIP) characteristics and distribution of epigenetic age acceleration according to sex and CHIP status. (A) Density plot depicting the distribution of variant allele frequency (VAF) of largest CHIP mutation per individual. Dotted lines show 1st, 2nd and 3rd quartile. (B) Plot depicting number and type of mutations for the 106 individuals with present CHIP as determined by Arends et al.10 (C and D) Density plots depicting the distribution of age acceleration according to sex (C) and CHIP status (D). Dashed lines show respective means. P-values were determined by MannWhitney-test. Haematologica | 107 July 2022

1705

LETTER TO THE EDITOR ing higher cycle numbers and mitotic aging. Even though mitotic and epigenetic aging probably do not address the same aging processes,9,16 we acknowledge that further research is needed to explore these relations. Compared to the two previous studies that used methylation array based epigenetic clocks (e.g., by Horvath), our

epigenetic age estimations relied on relatively few CpG sites. Nevertheless, our clock performs comparably well and high intercorrelation with Horvath clock was shown.17 Unlike the previous studies, we could not detect an association of individual mutations with a particularly accelerated epigenetic age. However, due to rather small sample sizes

Figure 2. Clonal hematopoiesis of indeterminate potential characteristics and epigenetic age acceleration. (A) Jittered dot plot of epigenetic age acceleration (AA) by affected gene. (B) Jittered dot plot of AA by affected gene group. (C) Jittered dot plot of AA by number of clonal hematopoiesis of indeterminate potential (CHIP) mutations. (D) Jittered dot plot of AA by CHIP clone size. Mean and standard deviation are shown in black. (E and F) Dot plots of AA against CHIP variant allele frequency (VAF). Spearman’s correlation of VAF against AA (E) or intrinsic epigenetic age acceleration (IEAA) (F). For patients with multiple CHIP mutations, largest VAF was used for calculation. Haematologica | 107 July 2022

1706

LETTER TO THE EDITOR in our subgroups, we cannot fully exclude type II errors in this place. While the study population is heterogeneous with respect to the underlying medical conditions, the overall disease burden of the study cohort is above the population average. Therefore, the study design is not suited to evaluate the impact of individual medical conditions on epigenetic aging. While a generally high level of comorbidities likely restores comparability with regard to CHIP as an independent factor, our study is not powered to investigate associations between medical conditions and combined CHIP/AA status. In conclusion, our study revealed a correlation of CHIP clone size and accelerated epigenetic aging. This finding matches and extends our knowledge on the role of CHIP clone size and sets the foundation for future investigations exploring the interrelatedness of CHIP and aging.

Authors Jasper David Feldkamp,1 Valentin Max Vetter,2,3 Christopher Maximilian Arends,1 Tonio Johannes Lukas Lang,1 Lars Bullinger,1,4 Frederik Damm,1,4 Ilja Demuth2,5 and Mareike Frick1,4 Charité – Universitätsmedizin Berlin, corporate member of Freie Universität Berlin, Humboldt-Universität zu Berlin, and Berlin Institute of Health, Medical Department, Division of Hematology, Oncology, and Cancer Immunology; 2Charité – Universitätsmedizin Berlin, corporate member of Freie Universität Berlin, HumboldtUniversität zu Berlin, Department of Endocrinology and Metabolic Disease (including Division of Lipid Metabolism), Biology of Aging Working Group; 3Department of Psychology, Humboldt-Universität zu Berlin; 4German Cancer Consortium (DKTK) and German Cancer Research Center (DKFZ), Heidelberg and 5Charité – Universitätsmedizin Berlin, BCRT - Berlin Institute of Health Center for Regenerative Therapies, Berlin, Germany 1

Correspondence: MAREIKE FRICK - mareike.frick@charite.de https://doi.org/10.3324/haematol.2021.280021 Received: October 14, 2021. Accepted: February 21, 2022. Prepublished: March 3, 2022. Disclosures JF, VV, CMA, TL, ID, and MF declare no conflicts of interest. LB received honoraria from Seattle Genetics, Sanofi, Astellas, Amgen, consultancy fee from Gilead, Hexal, and Menarini, consultancy fee and Honoraria Abbvie, BMS/Celgene, Daiichi Sankyo, Janssen, Jazz Pharmaceuticals, Novartis and Pfizer, and research funding from Bayer and Jazz Pharmaceuticals. FD reports personal fees from AbbVie, Astra Zeneca, Gilead, Novartis, and Roche outside the submitted work. Contributions JF and VV performed experiments, analyzed and interpreted data; MF, ID, and FD designed and supervised research and experiments, CMA collected samples and clinical data, performed experiments, analyzed and interpreted data; JF and MF wrote the manuscript; VV, CMA, TL, LB, FD and ID revised the manuscript. All authors approved the final version. Funding This study was supported by DKTK and institutional funding both awarded to MF and a grant of the Deutsche Forschungsgemeinschaft (grant # DE 842/7-1) to ID. FD was supported by the Deutsche Forschungsgemeinschaft (grant #DA1787/1-1), the DKMS Giving Life Foundation, the Else KrönerFresenius-Stiftung (grant #2017_EKES.33), and the Deutsche Krebshilfe (#70113643). CMA was supported by the BIH clinician scientist program. Data sharing statement The data that support the findings of this study are available from the corresponding author, upon reasonable request.

References 1. Steensma DP, Bejar R, Jaiswal S, et al. Clonal hematopoiesis of indeterminate potential and its distinction from myelodysplastic syndromes. Blood. 2015;126(1):9-16. 2. Jaiswal S, Fontanillas P, Flannick J, et al. Age-related clonal hematopoiesis associated with adverse outcomes. N Engl J Med. 2014;371(26):2488-2498. 3. Genovese G, Kähler AK, Handsaker RE, et al. Clonal hematopoiesis and blood-cancer risk inferred from blood DNA sequence. N Engl J Med. 2014;371(26):2477-2487. 4. Frick M, Chan W, Arends CM, et al. Role of donor clonal hematopoiesis in allogeneic hematopoietic stem-cell transplantation. J Clin Oncol. 2019;37(5):375-385. 5. Lopez-Otin C, Blasco MA, Partridge L, Serrano M, Kroemer G. The hallmarks of aging. Cell. 2013;153(6):1194-1217. 6. Horvath S, Raj K. DNA methylation-based biomarkers and the epigenetic clock theory of ageing. Nat Rev Genet. 2018;19(6):371-384. 7. Horvath S. DNA methylation age of human tissues and cell types. Genome Biol. 2013;14(10):R115. 8. Vidal-Bralo L, Lopez-Golan Y, Gonzalez A. Simplified assay for

epigenetic age estimation in whole blood of adults. Front Genet. 2016;7:126. 9. Vetter VM, Meyer A, Karbasiyan M, Steinhagen-Thiessen E, Hopfenmuller W, Demuth I. Epigenetic clock and relative telomere length represent largely different aspects of aging in the Berlin Aging Study II (BASE-II). J Gerontol A Biol Sci Med Sci. 2019;74(1):27-32. 10. Arends CM, Galan-Sousa J, Hoyer K, et al. Hematopoietic lineage distribution and evolutionary dynamics of clonal hematopoiesis. Leukemia. 2018;32(9):1908-1919. 11. Nachun D, Lu AT, Bick AG, et al. Clonal hematopoiesis associated with epigenetic aging and clinical outcomes. Aging Cell. 2021;20(6):e13366. 12. Robertson NA, Hillary RF, McCartney DL, et al. Age-related clonal haemopoiesis is associated with increased epigenetic age. Curr Biol. 2019;29(16):R786-R787. 13. Abelson S, Collord G, Ng SWK, et al. Prediction of acute myeloid leukaemia risk in healthy individuals. Nature. 2018;559(7714):400-404. 14. Qu Y, Lennartsson A, Gaidzik VI, et al. Differential methylation in

Haematologica | 107 July 2022

1707

LETTER TO THE EDITOR CN-AML preferentially targets non-CGI regions and is dictated by DNMT3A mutational status and associated with predominant hypomethylation of HOX genes. Epigenetics. 2014;9(8):1108-1119. 15. Fazila A, Vasu P, Jesper C, et al. Genome-wide profiling identifies a DNA methylation signature that associates with TET2 mutations in diffuse large B-cell lymphoma. Haematologica. 2013;98(12):1912-1920. 16. Marioni RE, Harris SE, Shah S, et al. The epigenetic clock and

telomere length are independently associated with chronological age and mortality. Int J Epidemiol. 2018;45(2):424-432. 17. Vetter VM, Kalies CH, Sommerer Y, et al. Relationship between five epigenetic clocks, telomere length and functional capacity assessed in older adults: cross-sectional and longitudinal analyses. J Gerontol A Biol Sci Med Sci. 2022 Jan 15. [Epub ahead of print]

Haematologica | 107 July 2022

1708

LETTER TO THE EDITOR

Decitabine salvage for TP53-mutated, relapsed/refractory acute myeloid leukemia after cytotoxic induction therapy TP53-mutated acute myeloid leukemia (AML) represents a therapeutic challenge due its chemotherapy-refractoriness, and to an uncertain role for hematopoietic cell transplant (HCT) intensification.1 We initiated a trial to determine whether decitabine might salvage TP53-mutated AML after failure of cytarabine-based induction. Seventeen patients were enrolled before the trial was closed due to slow accrual. Decitabine was well tolerated in this pretreated population and allowed transition to HCT in seven of 17 patients (41%). The 1-year overall survival (OS) was 29% (median 244 days; 95% confidence interval [CI]: 116-390). Survival was longer in patients receiving HCT (median 354 days), and two long-term survivors were transplanted in molecular remission. Detection of TP53 clonal response by bone marrow (BM) immunohistochemistry (IHC) or peripheral blood (PB) exome sequencing was associated with improved survival, suggesting the utility of these secondary endpoints in future clinical trials. This single-arm, open-label, prospective clinical trial (clinicaltrials gov. Identifier: NCT03063203) was approved by the Institutional Review Board at Washington University in St. Louis. The study enrolled 17 patients between October 2017 and September 2020 before closing due to slow accrual during the SARS-CoV-2 pandemic, and to shifting treatment practices towards the use of venetoclax combinations.2 Eligible patients had TP53-mutated relapsed/refractory AML following cytarabine-based induction chemotherapy, and at least one of the following: BM blasts >5%, flow-based measurable residual disease (MRD) >0.5%, persistent cytogenetic abnormality by fluorescence in situ hybridization (FISH) or karyotyping, or persistent TP53 mutation at ≥5% variant allele frequency (VAF) with exome sequencing. Decitabine was administered at 20 mg/m2/day on days 1-10 of 28-day cycles and could be reduced to days 1-5 once BM aspirate blasts were <5%. Granulocyte colony-stimulating factor (G-CSF) use was allowed during the treatment of sepsis and neutropenic fevers, but not to support neutrophil recovery. The primary objective was to determine the 1-year OS in patients with TP53-mutated AML compared to historic controls (1-year OS 25%).1,3,4 OS was defined from the time of enrollment to death from any cause. Secondary endpoints included determination of: i) proportion of morphologic responses, as defined by the European LeukemiaNet 2017 (ELN)5; ii) time to transplant and the number of patients able to undergo HCT; iii) 2-year eventfree survival after transplant compared to historical controls (18-22%);6 and iv) average number of hospital days during cycles 1-2, as a surrogate of toxicity.

The average time from the initiation of induction chemotherapy to trial enrollment was 42 days (median 25 days), consistent with primary refractory disease, persistent MRD, or rapid relapses after induction chemotherapy (Table 1). Performance status was 0 or 1 in 16 of 17 patients, reflecting a population that had been fit for cytotoxic chemotherapy (Table 1). Sixteen of the 17 patients had complex cytogenetics and nine of 17 had cytogenetic loss of chromosome 17p (Online Supplementary Table S1). As expected in relapsed/refractory AML and TP53-mutated AML, BM aspirates were commonly hypoplastic,7 with a blast count mean of 18% at trial enrollment and of 37% at diagnosis. The mean number of hospital days during combined cycles 1 and 2 was 21 (median 14 days; Online Supplementary Table S1), including the inpatient decitabine administration days. Observed grade 3-4 serious adverse events (SAE) reflected typical complications associated with decitabine therapy, including anemia (1), febrile neutropenia (6), heart failure (1), gastrointestinal pain (1), infections (2), abnormal liver function test (LFT) (2), troponin elevation (1), lymphopenia (6), neutropenia (6), thrombocytopenia (4), acidosis (1), hyperglycemia (2), electrolyte imbalance (2), acute kidney injury (2), dyspnea (1), respiratory failure (3), hypertension (1), and hypotension (1). The median survival was 244 days (95% CI: 116-390; Table 1; Figure 1A). Twelve patients died within 1 year of presentation, providing a 29% 1-year OS. An interim analysis noted 27% predictive power to reject the null hypothesis if the study were to include 60 patients8 (calculated using PASS v.15.0.5). All non-transplanted patients eventually relapsed and died of disease progression. Seven patients underwent HCT, at a mean of 106 days (median 117 days, Online Supplementary Table S1). Three patients died in remission from complications of transplantation, two died after relapse/progression (Online Supplementary Table S1). Two patients remain alive at 26 and 18 months; both were transplanted in molecular remission. Overall, HCT was associated with longer survivals (median 354 days; Figure 1B). No patient achieved a complete morphologic response by ELN criteria (Table 1). Robust neutrophil recovery was not noted, though G-CSF was not used to treat asymptomatic neutropenia. Five patients displayed platelet counts normalization (Figure 1C). BM and PB samples were collected at enrollment (day 0) and at the end of cycles 1, 2, and 3. For 16 patients, samples at AML diagnosis (pre-induction) were also available for correlative studies. TP53 IHC (antibody clone DO7) was performed on 4 mm BM sections using the Benchmark

Haematologica | 107 July 2022

1709

LETTER TO THE EDITOR XT automated stainer (Ventana Medical Systems, Tucson, AZ, USA). Quantitative scoring was performed on nuclear staining in 500 hematopoietic cells. Based on published cutoffs,9 IHC response was defined as a reduction of

TP53+ cells to <10% of total cells on a core biopsy sample. Many TP53 missense mutations are associated with IHCdetectable TP53, via protein stabilization.10,11 TP53 IHC showed staining in all cases with TP53 missense muta-

Table 1. Clinical characteristics and treatment responses. Variable Sex Age in years (range) ECOG Performance Status

Days from induction to enrollment (range) Induction chemotherapy

1-year OS Transplant (day of treatment) (n=7) Survival post transplant (days) ELN responses

Level

N=17

F M Median 0 1 2

6 11 61 (43-74) 4 12 1

35.3 64.7

Median

25 (16-176)

7+3 Vyxeos (CPX-351) FLAG died before 1 year survived > 1 year Median Mean Median Mean CRi mLFS SD PD NA

6 10 1 12 5 117 106 254 236 5 1 6 2 3

23.5 70.5 6

71% 29%

29.4% 5.9% 35.3% 11.8% 17.6%

7+3: 7 days cytarabine, 3 days anthracycline. FLAG: fludarabine, cytarabine, idarubicin and granulocyte colony stimulating factor; CRi: complete remission with incomplete count recovery; mLFS: morphologic leukemia free state; SD: stable disease; PD: progressive disease; NA: not possible to evaluate.

Figure 1. Summary of clinical responses. (A) Kaplan-Meier curve describing the overall survival of the 17 patients enrolled in the trial (median survival 244 days, 95% confidence interval: 116-390 days). (B) Kaplan-Meier curve describing the overall survival of the seven transplanted patients (median survival 354 days). (C) Line plots showing the platelet (Plt) count trends at different time points for the 17 the patients enrolled in the study. On the right side of the plot are separately displayed five patients (003, 016, 019, 024 and 029) with platelet recovery. SD: stable disease; NA: not possible to evaluate; CRi: complete remission with incomplete count recovery. Haematologica | 107 July 2022

1710

LETTER TO THE EDITOR

Figure 2. Molecular responses trends and their association with overall survival. (A and B) Line plots showing the % TP53-positive bone marrow cells by immunohistochemistry (IHC) at different time points of response assessment for: responder cases (IHC TP53 <10% reduction during therapy) vs. non-responder cases (IHC TP53 >10% during therapy). (C and D) Line plots of TP53 variant allele frequency (VAF) during therapy for responder cases ((E) reduction of copy number (CN) adjusted TP53 VAF below 0.05%) vs. nonresponder cases ((F) VAF stable or progressing over time). (E) Overall survival curves of the 17 cases stratified by responses assessed with IHC (median OS 345 days vs. 116 days, P<0.002). (F) Overall survival stratified by responses assessed with exome sequencing (median OS 390 days vs. 165 days; P<0.001).

tions; however, three of four cases with nonsense mutations did not have detectable TP53 by IHC (R213*, Y107*, F54Sfs*69; Online Supplementary Figure S1A to C). Unexpectedly, the C-terminal nonsense mutation (R342*) led to elevated TP53 protein at three separate time points (Online Supplementary Figure S1D). TP53 IHC protein levels correlated between pre-induction and day 0 (R2=0.56). Only two cases (WUDAC015 and WUDAC021) were associ-

ated with a significant reduction of TP53 IHC staining after cytotoxic induction therapy, confirming the limited efficacy of standard induction in this cohort (Online Supplementary Figure S1E and F). Serial assessment of response by IHC noted eight patients with responses (Figure 2A), and five patients without responses (Figure 2B). Four patients could not be evaluated: three had nonsense variants (WUDAC001, WUDAC002, and WUDAC029; Online

Haematologica | 107 July 2022

1711

LETTER TO THE EDITOR Supplementary Figures S1A to C), and one (WUDAC014) had del17p, all resulting in absent IHC TP53 staining. Exome sequencing was performed in parallel using PB samples to circumvent sampling variation due to hemodilute collections in this hypocellular disease (Online Supplementary Table S1). Exome capture utilized an IDT exome reagent, and was resolved on an Illumina NovaSeq S4 300XP to a median depth of 200x for pre-induction and day 0 samples, and 148x (range, 76 – 200x) for other time points. This provided >100x coverage for >90% of targets in 61 of 67 samples. Molecular response was defined as a reduction in the copy number adjusted TP53 to a VAF of <0.0512. The computational pipeline is available at https://github.com/genome/analysisworkflows/blob/968d7d80c3cec865c7fa58b4dc24561a4d bfd9ad/definitions/pipelines/somatic_exome.cwl. Mutation burden assessment at day 0 and at pre-induction showed correlation (R2=0.8). Eight patients achieved a molecular response, and seven patients displayed persistence of TP53 mutations after therapy (Figure 2C and D). For one patient (WUDAC016), no follow-up PB samples were available (Online Supplementary Figure S1G). The absolute TP53 tumor burden quantified by BM IHC, either at pre-induction or at day 0, did not consistently correlate with PB exome results (R2=0.15 and 0.27, respectively). However, qualitative response trends were concordant in 11 of 16 evaluable patients (Online Supplementary Figures S1D to F and S1H to O). One case (WUDAC005; Online Supplementary Figure S1P) showed stable disease by TP53 IHC, but progressive disease by PB exome sequencing, suggesting peripheralization of AML cells during therapy. Exome analyses revealed that the global molecular and clinical response was dictated by the TP53 clonal response trend (Online Supplementary Figure S2). Discordance between the TP53 clone and an alternate clone was only observed in WUDAC001, who progressed with a different clone during TP53 clonal response. Survival outcomes were longer in patients with molecular responses identified by TP53 IHC (median OS 345 days vs. 116 days, P<0.002; Figure 2E) or by exome sequencing (median OS 390 days vs. 165 days, P<0.001; Figure 2F). These results are consistent with data from other studies,13 and suggest that IHC and exome sequencing could be useful adjunctive strategies to quantify responses in future clinical trials. However, each approach has limitations: IHC is applicable only to cases with mutations that stabilize TP53 protein (typically missense variants) and lacks specificity below tumor burden of 10%, due to background staining that occurs in a small number of non-malignant cells. Sequencing of PB samples qualitatively reflected the measurement of TP53 levels in the BM in this study, however this approach is affected by the proportion of circulating malignant cells.

TP53-mutated AML has dismal outcomes and is commonly associated with chemotherapy resistance. Although we found that decitabine is tolerated after intensive chemotherapy, and that molecular responses are achievable in a subset of relapsed/refractory TP53 patients, long-term survival remained poor. These results are consistent with prior studies reporting lower responses to decitabine in relapsed/refractory disease versus untreated cases.14,15 Therefore, novel therapies, upfront combination and consolidation strategies should be considered. The hypoplastic BM in many patients makes accurate response determination challenging, due to hemodilute aspirate collections. The integration of molecular endpoints into clinical trials may improve response quantification, and increase the ability to identify significant differences between treatment arms.

Authors Francesca Ferraro,1 Agata Gruszczynska,1 Marianna B. Ruzinova,2 Christopher A. Miller,1 Mary Elizabeth Percival,3,4 Geoffrey L. Uy,1 Iskra Pusic,1 Meagan A. Jacoby,1 Mathew J. Christopher,1 Miriam Y. Kim,1 Peter Westervelt,1 Amanda F. Cashen,1 Mark A. Schroeder,1 John F. DiPersio,1 Camille N. Abboud,1 Lukas D. Wartman,1 Feng Gao,5 Daniel C. Link,1 Timothy J. Ley1 and John S. Welch1 1

Department of Medicine, Division of Oncology, Washington

University School of Medicine, St. Louis, MO; 2Department of Pathology and Immunology, Washington University School of Medicine, St. Louis, MO; 3Department of Internal Medicine, Division of Hematology, University of Washington School of Medicine, Seattle, WA; 4Clinical Research Division, Fred Hutchinson Cancer Research Center, Seattle, WA and 5Department of Surgery, Division of Public Health Sciences, Washington University School of Medicine, St. Louis, MO, USA. Correspondence: John S. Welch - jwelch@dom.wustl.edu https://doi.org/10.3324/haematol.2021.280153 Received: October 7, 2021. Accepted: February 21, 2022. Prepublished: March 3, 2022. Disclosures No conflicts of interest to disclose. Contributions FF and JSW designed the trial, wrote the manuscript and oversaw the study; MEP, GLU, IP, MAJ, MYK, PW, AFC, MAS, JFD, CNA and LDW contributed to clinical enrollment and data analysis; AG, MBR, CAM, DCL, and TJL contributed to molecular analysis; FG contributed to statistical analysis. All authors had access to primary data, reviewed and approved the manuscript.

Haematologica | 107 July 2022

1712

LETTER TO THE EDITOR Acknowledgments

Research Excellence in AML of the National Cancer Institute (P50

We thank Megan Haney and Jeff King for assistance in patient

CA171963 to DCL), the Genomics of AML Program Project (P01

enrollment, sample collection, and data processing; Sharon Heath,

CA101937, to TJL), a K12 Program grant (CA167540, to FF) and by

Nicole Helton, and the Tissue Procurement Core for assistance in

Janssen Pharmaceuticals.

sample collection and processing; the McDonnell Genome Institute at Washington University in St. Louis for support in sequencing; and

Data sharing statement

Anh Vu for help with mutation manual review.

Data sharing is subject to Institutional policies governing clinical research data. Data can be made available upon request to the

Funding

corresponding author stating dbGaP study ID phs000159.

The study was supported by grants from the Specialized Program of

References 1. Rucker FG, Schlenk RF, Bullinger L, et al. TP53 alterations in acute myeloid leukemia with complex karyotype correlate with specific copy number alterations, monosomal karyotype, and dismal outcome. Blood. 2012;119(9):2114-2121. 2. Maiti A, Qiao W, Sasaki K, et al. Venetoclax with decitabine vs intensive chemotherapy in acute myeloid leukemia: a propensity score matched analysis stratified by risk of treatment-related mortality. Am J Hematol. 2021;96(3):282-291. 3. Papaemmanuil E, Gerstung M, Bullinger L, et al. Genomic classification and prognosis in acute myeloid leukemia. N Engl J Med. 2016;374(23):2209-2221. 4. Stengel A, Kern W, Haferlach T, et al. The impact of TP53 mutations and TP53 deletions on survival varies between AML, ALL, MDS and CLL: an analysis of 3307 cases. Leukemia. 2017;31(3):705-711. 5. Dohner H, Estey E, Grimwade D, et al. Diagnosis and management of AML in adults: 2017 ELN recommendations from an international expert panel. Blood. 2017;129(4):424-447. 6. 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. 7. Welch JS. Patterns of mutations in TP53 mutated AML. Best Pract Res Clin Haematol 2018;31(4):379-383. 8. Jennison CaT, Bruce W. Group sequential methods with applications to clinical trials. Chapman & Hall; 2000.

9. Sallman DA, DeZern AE, Garcia-Manero G, et al. Eprenetapopt (APR-246) and azacitidine in TP53-mutant myelodysplastic syndromes. J Clin Oncol. 2021;39(14):1584-1594. 10. Ruzinova MB, Lee YS, Duncavage EJ, Welch JS. TP53 immunohistochemistry correlates TP53 mutation status and clearance in decitabine-treated patients with myeloid malignancies. Haematologica. 2019;104(8):e345-e348. 11. McGraw KL, Nguyen J, Komrokji RS, et al. Immunohistochemical pattern of p53 is a measure of TP53 mutation burden and adverse clinical outcome in myelodysplastic syndromes and secondary acute myeloid leukemia. Haematologica. 2016;101(8):e320-e323. 12. Klco JM, Miller CA, Griffith M, et al. Association between mutation clearance after induction therapy and outcomes in acute myeloid leukemia. JAMA. 2015;314(8):811-822. 13. Hunter AM, Komrokji RS, Yun S, et al. Baseline and serial molecular profiling predicts outcomes with hypomethylating agents in myelodysplastic syndromes. Blood Adv. 2021;5(4):1017-1028. 14. DiNardo CD, Maiti A, Rausch CR, et al. 10-day decitabine with venetoclax for newly diagnosed intensive chemotherapy ineligible, and relapsed or refractory acute myeloid leukaemia: a singlecentre, phase 2 trial. Lancet Haematol. 2020;7(10):e724-e736. 15. Welch JS, Petti AA, Miller CA, et al. TP53 and decitabine in acute myeloid leukemia and myelodysplastic syndromes. N Engl J Med. 2016;375(21):2023-2036.

Haematologica | 107 July 2022

1713

LETTER TO THE EDITOR

Tricuspid-valve regurgitant jet velocity as a risk factor for death in β-thalassemia Pulmonary hypertension (PH) has been recognized as a frequent complication in patients with β-thalassemia, owing to the underlying chronic anemia, endothelial damage, and hypercoagulable state.1,2 However, there are mounting concerns that reported prevalence rates using echocardiographic thresholds overestimate the true epidemiology of the disease. In 2014, we published results from a large multicenter study of 1,309 Italian β-thalassemia patients which evaluated PH prevalence through a dedicated protocol for right heart catherization (RHC), as part of the Italian Webthal® project (clinicaltrials gov. Identifier: NCT01496963).3 In the original study, adult (≥18 years) patients with a diagnosis of β-thalassemia major or intermedia and without chronic restrictive lung disease or a left ventricular ejection fraction (LVEF) ≤50% were recruited between January 2012 and January 2013. Patients first underwent a screening transthoracic echocardiography using continuous-wave Doppler sampling of the peak tricuspid-valve regurgitant jet velocity (TRV) to calculate systolic pulmonary artery pressure (sPAP), based on which they were divided into three

groups: PH unlikely (n=1,234), sPAP ≤36 mm Hg or TRV ≤3.0 m/s; PH possible (n=28), sPAP >36 to <40 mm Hg or TRV >3.0 to <3.2 m/s; and PH likely (n=47), sPAP ≥40 mm Hg or TRV ≥3.2 m/s. Patients with PH likely underwent RHC and the prevalence of confirmed PH in the entire study population was 2.1%. The positive predictive value for the TRV ≥3.2 m/s threshold for the diagnosis of PH was 93.9%.3 This latter finding was adopted by β-thalassemia international management guidelines which now recommend RHC in patients with a TRV ≥3.2 m/s.2,4 In this study, we provide long-term mortality data for patients who had a TRV ≥3.2 m/s in the original study to further confirm the clinical value of this threshold to flag patients who require RHC and PH-directed management. The Institutional Review Boards of all participating centers approved the study protocol and all participants signed a written informed consent before inclusion in the study. We followed 42 of the 47 patients with TRV ≥3.2 m/s (PH likely) at the initial echocardiographic assessment (baseline) until March 2021, death, or loss to follow-up. Five patients had transitioned care to other

Table 1. Comparison of study parameters between baseline tricuspid-valve regurgitant jet velocity groups.

Baseline TRV Parameter Baseline characteristics Age in years, median (IQR) Male, % Splenectomized, % Thalassemia diagnosis, % β-thalassemia intermedia β-thalassemia major Hemoglobin in g/dL, median (IQR) Serum ferritin in ng/mL, median (IQR) LVEF in %, median (IQR) TRV in m/s, median (IQR) Management PH-directed therapy, %† None Single agent Double agent Outcome Death from any cause, % Death from cardiopulmonary disease, %

P-value*

≥3.2 m/s (N = 42)

>3.0 to <3.2 m/s (N = 26)

48 (39.8-57.8) 38.1 88.1

44.5 (39.3-48.3) 46.2 76.9

66.7 33.3 9.1 (8-9.6) 970 (476.8-1,538.5) 60 (60-66.5) 3.43 (3.31-4.09)

50 50 9.1 (8.5-9.7) 767.5 (500-1,097) 62 (53.5-66) 3.04 (3.04-3.08)

0.097 0.614 0.312 0.208

0.495 0.466 0.843 <0.001 <0.001

35.7 42.9 21.4

100 0 0

47.6 33.3

15.4 0

0.009 0.001

*Comparisons made using the Mann Whitney U test for continuous variables and Fisher’s Exact test for categorical variables. †Bosentan, ambrisentan, sildenafil, tadalafil, macitentan, riociguat, and other angiotensin converting enzyme inhibitors, calcium channel and betablockers, and anticoagulants. IQR: interquartile range; LVEF: left ventricular ejection fraction; TRV: tricuspid-valve regurgitant jet velocity; PH: pulmonary hypertension. Haematologica | 107 July 2022

1714

LETTER TO THE EDITOR institutions immediately after the original study and were not included in this analysis. For comparisons, we also followed 26 of the 28 patients (2 patients transitioned care) originally grouped as PH possible (TRV >3.0 to <3.2 m/s). For each patient, we retrieved data on baseline demographics (age and sex), thalassemia diagnosis, splenectomy status, hemoglobin and serum ferritin levels (mean over previous 10 years), and LVEF. The use of PHdirected therapies during the period of observation and

last observation TRV values, if available, were also retrieved. Patients with a baseline TRV ≥3.2 m/s (n=42) and TRV >3.0 to <3.2 m/s (n=26) had comparable age, hemoglobin level, serum ferritin level, and LVEF (Table 1). They also had comparable proportions of males, β-thalassemia intermedia diagnosis, and splenectomy (Table 1). None of the patients with a TRV >3.0 to <3.2 m/s received PH-directed therapy, while 18 (42.9%) patients received single

Figure 1. Kaplan-Meier survival curves. (A) All-cause mortality, and (B) cardiopulmonary disease-related mortality according to baseline TRV group. TRV: tricuspid-valve regurgitant jet velocity. Haematologica | 107 July 2022

1715

LETTER TO THE EDITOR

Figure 2. Kaplan-Meier survival curves. (A) All-cause mortality, and (B) cardiopulmonary disease-related mortality according to TRV improvement. TRV: tricuspid-valve regurgitant jet velocity.

agent and nine (21.4%) patients received double agent therapy in the TRV ≥3.2 m/s group (Table 1). The median follow-up time was 6 years (interquartile range [IQR]: 2.3-8). The crude all-cause mortality rate was 47.6% (95% confidence interval [CI]: 32.0-63.6; cardiopulmonary disease n=14, hepatic failure n=2, post fracture n=1, severe hemolytic anemia n=1, unknown n=2) in patients with a baseline TRV ≥3.2 m/s compared with 15.4% (95% CI: 4.4-34.9; hepatic failure n=2, sepsis n=1,

unknown = 1) in patients with a TRV >3.0 to <3.2 m/s (P=0.009; Table 1). The crude cardiopulmonary diseaserelated mortality rate was 33.3% (95% CI: 19.6-49.6; right-sided heart failure n=12, pulmonary embolism n=2) in patients with a baseline TRV ≥3.2 m/s while there were no cardiopulmonary disease-related deaths in patients with a TRV >3.0 to <3.2 m/s (P=0.001; Table 1). On KaplanMeier curves, cumulative all-cause mortality-free survival estimates at 5 and 9 years for patients with a baseline

Haematologica | 107 July 2022

1716

LETTER TO THE EDITOR TRV ≥3.2 m/s were 59% and 47% compared with 84% and 84% in patients with a TRV >3.0 to <3.2 m/s (log-rank chi-square: 5.944, P=0.015; Figure 1A). Cumulative cardiopulmonary disease-related mortality-free survival estimates at 5 and 9 years for patients with a baseline TRV ≥3.2 m/s were 71% and 56% compared with 100% and 100% in patients with a TRV >3.0 to <3.2 m/s (log-rank chi-square: 9.413, P=0.002; Figure 1B). Of note, 29 of the 42 patients with a baseline TRV ≥3.2 m/s underwent RHC in the original study, two patients did not have confirmed PH on RHC and remained alive; while 27 patients had confirmed PH on RHC, out of which 15 died during the period of observation (12 due to cardiopulmonary disease). On Cox regression analysis, the unadjusted hazard ratio (HR) for death from any cause in patients with a baseline TRV ≥3.2 m/s compared with TRV >3.0 to <3.2 m/s was 3.402 (95% CI: 1.157-10.005; P=0.026). In a multivariate, forward stepwise Cox regression model including baseline TRV groups and adjusting for age, sex, splenectomy status, thalassemia diagnosis, hemoglobin level, serum ferritin level, LVEF, and PH-directed therapy; only baseline TRV was significantly and independently associated with an increased risk of death (adjusted HR for TRV ≥3.2 m/s vs. TRV >3.0 to <3.2 m/s: 3.337; 95% CI: 1.122-9.923; P=0.030). On receiver operating characteristic curve analysis, a baseline TRV of 3.2 m/s had the highest sum of sensitivity and specificity for the outcomes of death from any cause (area under the curve [AUC]: 0.67±0.07; P=0.024) and death from cardiopulmonary disease (AUC: 0.74±0.07; P=0.007) among the range of TRV values >3.0 m/s included in this study. Thirty-two out of the 42 patients with a baseline TRV ≥3.2 m/s had a TRV at the last observation. TRV values improved to <3.2 m/s in 12 (37.5%) patients of whom three (25%) were receiving single agent and five (41.7%) were receiving double therapy. TRV values remained ≥3.2 m/s in 20 (62.5%) patients of which 14 (70%) were receiving single agent and one (5%) was receiving double therapy. Survival was significantly better in patients who had TRV improvement than those who did not for both all-cause mortality (log-rank chi-square: 7.876, P=0.005; Figure 2A) and cardiopulmonary disease-related mortality (log-rank chi-square: 5.237, P=0.020; Figure 2B). Our study further highlights the relevance of a TRV ≥3.2 m/s threshold to inform PH-related management in βthalassemia by establishing an increased risk of mortality, especially from cardiopulmonary disease. Our original study supported a recommendation to undergo RHC at this TRV threshold through the finding of a high positive predictive rate, as well as more recent data on high mortality in those patients with confirmed PH on RHC.2-5 However, not all patients were able to undergo RHC,

which makes clinical correlates with adverse outcomes identified in the full subset of patients with a TRV ≥3.2 m/s in this study very informative. This may also support management decisions in resource-restricted settings or when RHC is not feasible. The key limitations in our work that need to be addressed in future studies include inability to evaluate lower TRV thresholds for long-term mortality outcomes or RHC correlation; however, the lack of cardiopulmonary disease-related deaths in the subset of patients with TRV >3.0 to <3.2 m/s, further supports our original choice of a TRV ≥3.2 m/s for further evaluation. Indeed, low mortality rates (n=3, heart failure, cancer, unknown) were previously reported in thalassemia cohorts with lower TRV thresholds (n=148, mean TRV: 2.3±0.4 m/s, range: 0.2-3.5; 5% with TRV ≥3.0 m/s).6 On the contrary, data in patients with sickle cell disease have consistently reported an increased risk of mortality in patients with comparatively lower TRV thresholds ≥2.5 m/s.7,8 The reasons behind such difference remain unclear. They may be attributed to observational study design and adjustment for cofounders, the underlying mechanism of PH (hemolysis vs. hypercoagulability), or the more severe anemia in patients with thalassemia which may contribute to TRV elevation beyond that attributed to PH. Although this study was not designed to evaluate the role of intervention, it does suggest a protective role for TRV improvements below the 3.2 m/s threshold. There is a high unmet need for randomized clinical trials evaluating the role of intervention in β-thalassemia patients with PH, to avoid the detrimental effects of this complication.

Authors Giorgio Derchi,1* Khaled M. Musallam,2* Valeria Maria Pinto,3 Giovanna Graziadei,4 Marianna Giuditta,5 Susanna Barella,6 Raffaella Origa,6 Gavino Casu,7 Annamaria Pasanisi,8 Filomena Longo,9 Maddalena Casale,10 Roberta Miceli,11 Pierluigi Merella,7 Barbara Gianesin,12 Pietro Ameri,13,14 Immacolata Tartaglione,10 Silverio Perrotta,10 Antonio Piga,15 Maria Domenica Cappellini4,16 and Gian Luca Forni3 on behalf of the Webthal® project Department of Cardiology, High Specialty Ligurian Clinical Institute (ICLAS), Genoa, Italy; 2Thalassemia Center, Burjeel Medical City, Abu Dhabi, UAE; 3Center for Microcythemia, Congenital Anemia and Iron Dysmetabolism, Galliera Hospital, Genoa, Italy; 4Department of Medicine and Medical Specialties, IRCCS Ca' Granda Foundation, Maggiore Policlinico Hospital, Milan, Italy; 5Cardiovascular Disease Unit, IRCCS Ca' Granda Foundation, Maggiore Policlinico Hospital, Milan, Italy; 6Microcitemico Pediatric Hospital " Antonio Cao ", ARNAS G. Brotzu, Cagliari, Italy; 7Cardiology Unit, AO University of Sassari, Sassari, Italy; 8Hematology Unit, A. Perrino Hospital, Brindisi, Italy; 9Reference Centre for Hemoglobinopathies, AOU San Luigi Gonzaga Hospital, Orbassano, Italy; 10Department of Women, Child and General and Specialized Surgery, University Luigi Vanvitelli, Naples, Italy; 11Cardiology Unit, Galliera Hospital, Genoa, Italy; 12ForAnemia Foundation, Genoa, Italy; 13Cardiology Unit, IRCCS 1

Haematologica | 107 July 2022

1717

LETTER TO THE EDITOR Ospedale Policlinico San Martino, Genoa, Italy; 14Department of Internal Medicine, University of Genoa, Genoa Italy; 15Department of Clinical and Biological Sciences, University of Turin, Turin, Italy and 16 Department of Clinical Sciences and Community Health, University of Milan, Milan, Italy

References

*GD and KMM contributed equally as co-first authors. Correspondence: GIAN LUCA FORNI - gianluca.forni@galliera.it https://doi.org/10.3324/haematol.2021.280389 Received: November 21, 2021. Accepted: February 22, 2022. Prepublished: March 3, 2022. Disclosures No conflicts of interest to disclose. Contributions GD, KMM, VMP and GLF conveived and designed the study; GD, VMP, GG, MG, SB, RO, GC, APa, FL, MC, RM, PM, BG, PA, IT, SP, APi, MDC and GLF collected data; KMM perfomed statistical analysis. All authors reviewed and interpreted the results. KMM drafted the manuscript. All authors reviewed the manuscript review for important intellectual content and approved the manuscript before submission. Data sharing statement Data can be made available upon request to the corresponding author.

1. Taher AT, Musallam KM, Cappellini MD. Beta-thalassemias. N Engl J Med. 2021;384(8):727-743. 2. Taher AT, Cappellini MD. How I manage medical complications of beta-thalassemia in adults. Blood. 2018;132(17):1781-1791. 3. Derchi G, Galanello R, Bina P, et al. Prevalence and risk factors for pulmonary arterial hypertension in a large group of betathalassemia patients using right heart catheterization: a Webthal study. Circulation. 2014;129(3):338-345. 4. Taher A, Musallam K, Cappellini MD. Guidelines for the Management of Non Transfusion Dependent Thalassaemia (NTDT). Vol 2. Nicosia, Cyprus 2017. 5. Pinto VM, Musallam KM, Derchi GE, et al. Mortality in betathalassemia patients with confirmed pulmonary arterial hypertension on right heart catheterization. Blood. 2022;139(13):2080-2083. 6. Morris CR, Kim HY, Trachtenberg F, et al. Risk factors and mortality associated with an elevated tricuspid regurgitant jet velocity measured by Doppler-echocardiography in thalassemia: a Thalassemia Clinical Research Network report. Blood. 2011;118(14):3794-3802. 7. Maitra P, Caughey M, Robinson L, et al. Risk factors for mortality in adult patients with sickle cell disease: a metaanalysis of studies in North America and Europe. Haematologica. 2017;102(4):626-636. 8. Gladwin MT, Sachdev V, Jison ML, et al. Pulmonary hypertension as a risk factor for death in patients with sickle cell disease. N Engl J Med. 2004;350(9):886-895.

Haematologica | 107 July 2022

1718

LETTER TO THE EDITOR

BRAF V600E-positive cells as molecular markers of bone marrow disease in pediatric Langerhans cell histiocytosis Langerhans cell histiocytosis (LCH) is an inflammatory myeloid hematologic malignancy in which genetic alterations in the MAPK pathway promote the differentiation of hematopoietic stem cells (HSC) into mononuclear phagocytic lineages such as monocytes and dendritic cells, resulting in apoptosis resistance, local inflammation, and impaired lymph node migration.1-9 The BRAF V600E mutation is detected in various differentiation stages of myeloid cells, including HSC or nearby progenitors and mononuclear phagocytic cells, especially in high-risk systemic LCH.1-9 Bone marrow (BM) involvement in LCH, which causes cytopenia, is characterized by CD1a-positive immunohistochemical staining.10-15 However, the BRAF mutation status and clinical impact of BM disease (BMD) at the molecular level are not fully understood. In order to clarify the clinical impact of measurable BMD in LCH, we investigated somatic mutations in paired tumor and BM samples using a sensitive detection system. We retrospectively performed mutational analyses of 59 LCH tumors by targeted amplicon sequencing using custom-designed primers and subsequently analyzed somatic mutations in 41 paired BM samples using allelespecific droplet digital polymerase chain reaction (ddPCR). BRAF V600E was identified in 25 of 41 (61%) tumor samples. Measurable BMD was detected in 21 of the 25 (84%) cases positive for BRAF V600E, but was not detected in cases positive for other mutations. High mutational loads in BM were significantly associated with multisystem LCH with risk organ involvement, younger age, and disease progression. BRAF V600E in BM was detectable in patients who were refractory to treatment, and the mutational load in BM cells was higher than those in peripheral blood mononuclear cells. A high mutational burden in the BM defines a distinct clinical phenotype of high-risk, young-age patients with multisystem LCH and potential alternative risk factors. Between 1996 and 2020, 65 Japanese pediatric patients diagnosed with LCH by a central or local pathology review were enrolled. The patients were treated with various protocols, including those of the Japan LCH Study Group -96 and -02 and the Japanese Pediatric Leukemia/Lymphoma Study Group LCH-12 in the following 13 domestic centers: Hirosaki University Hospital, Miyagi Children’s Hospital, Nihon University Hospital, Chiba Children’s Hospital, Tokyo Medical and Dental University Hospital, Tohoku University Hospital, Niigata University Hospital, Kyoto University Hospital, Fujita Medical University Hospital, National Center for Child Health Hospital, Hyogo

Children’s Hospital, Shizuoka Children's Hospital, and Yamaguchi University Hospital.16 The parents/guardians of the study participants, who were minors, provided informed consent. The study was conducted according to the principles of the Declaration of Helsinki and approved by the ethics committee of Hirosaki University Graduate School of Medicine. We collected 65 biopsied LCH tumor specimens and 41 paired BM samples. We extracted genomic DNA from formalin-fixed, paraffin-embedded (FFPE) tumor specimens (50–70-mm thick) using the GeneRead DNA FFPE Kit (Qiagen, Hilden, Germany) according to the manufacturer’s protocol. In the BM samples, mononuclear cells were separated by erythrocyte hemolysis, and DNA was extracted. In the peripheral blood samples, plasma and peripheral blood mononuclear cells (PBMC) were separated by centrifugation, and erythrocyte hemolysis was performed before DNA extraction. DNA from frozen or fresh BM and PBMC was extracted using the QIAamp DNA Blood Mini kit (Qiagen, Hilden, Germany). We evaluated 59 LCH cases using PCRbased targeted next-generation sequencing (NGS), excluding six samples with poor PCR amplification. The amplicon libraries were analyzed for mutations in exons 12 and 15 of BRAF and exons 2 and 3 of MAP2K1 (Online Supplementary Table S1). For original data, please contact kkudo@hirosaki-u.ac.jp. Of the 59 LCH tumor specimens analyzed, we detected somatic mutations in 54 (92%) at various allele frequencies (2.3–35%, median 10%) by targeted NGS. BRAF and MAP2K1 were mutated in 46 (78%) and eight (14%) samples, respectively. BRAF V600E was identified in 33 (56%) samples. In-frame deletions in exon 12 of BRAF were found in ten (17%) samples (Online Supplementary Table S2). We also identified a novel mutation, BRAF V600delinsDL (c.1798_1799insATT, p.Val600delinsAspLeu), which was confirmed to be an activating mutation (data not shown). Next, we conducted mutational analysis of the 41 paired BM samples using allele-specific ddPCR to detect the following mutations: BRAF V600E, BRAF exon 12 deletion, MAP2K1 exon 2/3 deletion, BRAF V600D, and BRAF V600delinsDL (Online Supplementary Table S1; Online Supplementary Figure S1). In this study, only patients who underwent BM examination at the physician's decision because of fever, inflammatory reaction, mild cytopenia, or staging were included. Most patients underwent staging such as computed tomography (CT), magnetic resonance imaging (MRI), X-ray, bone scintigraphy, and BM examination instead of positron emission tomography

Haematologica | 107 July 2022

1719

LETTER TO THE EDITOR (PET)-CT, and none of the patients changed their disease type during the course. Measurable BMD was defined as molecularly detectable BM cells with an allele frequency of 0.01% or higher based on the ddPCR analysis of somatic mutations in BM samples. The clinical significance of BMD was examined in 38 patients, excluding three patients for whom quantitative mutational analysis was not applicable. These 38 patients consisted of 12 patients with multisystem LCH with risk organ involvement (MSRO+), 13 with multisystem LCH without risk organ involvement (MSRO-), eight with multifocal bone disease (MFB), and five with single-system LCH with a single lesion (SS) (Table 1). We detected BMD before treatment at various allele frequencies (0.03– 7.0%, median 0.83%) in 21 of the 38 (55%) patients, consisting of 12 MSRO+, seven MSRO-, two MFB, and no SS patients. Contrarily, the remaining 17 BMD-negative patients comprised six MSRO-, six MFB, and five SS patients, and no patient with MSRO+. Among them, there were 11 patients with mutations other than BRAF V600E, including four MSRO-, three MFB, and four SS patients (Table 1). The BRAF V600E mutation was detected in 21 of 25 (84%) BM samples from BRAF V600E-positive LCH, whereas no other variants were identified, indicating notable BMD in BRAF V600E-positive LCH cases (Figure 1A). The rate of BM involvement in the MSRO+, MSRO-, MFB, and SS cases was 100%, 54%, 25%, and 0%, respectively (Figure 1B). In the 25 cases of BRAF V600E-positive LCH, the median variant allele frequency (VAF) of BRAF V600E in the BM of the MSRO+, MSRO-, and MFB patients was 1.0% (range, 0.20–7.0%), 0.030% (range, 0–2.2%), and 0% (range, 0– 0.51%), respectively (Figure 1B). However, no somatic mutations were detected in the 13 patients with other types of mutations. Mutational burden varied among the four clinical phenotypes (Kruskal–Wallis, P<0.001). Notably, MSRO+ disease showed the highest mutational burden (Mann–Whitney test, P<0.001) (Figure 1B). We also investigated the relationship between BMD and age, disease progression, and disease relapse. BMD was detected only in patients younger than 3 years of age, except for one patient aged 4.6 years. BMD-positive patients had a lower median age than BMD-negative patients (0.9 years vs. 5.4 years, Mann–Whitney test, P=0.012) (Figure 1C). The cumulative incidence of reactivation and relapse in the 21 patients who were positive for BMD at diagnosis was higher than that in the 17 patients who were negative for BMD (79% vs. 36%, P=0.006 by log-rank test) (Figure 1D). This suggests that BMD positivity is a risk factor for disease progression. In order to identify the origin of the mutation-positive cells, we measured the somatic mutational load in paired BM and peripheral blood mononuclear cells (PBMC) from 13 patients at 31 time points, before and after initial chemotherapy. There were patients in whom mutated

cells were present only in BM cells after chemotherapy or early relapse. The median VAF of BRAF V600E was 0.65% (range, 0.016–6.7%) in the BM cells and 0.16% (range, 0–4.0%) in the PBMC, indicating a higher mutational load in the former (paired t-test, P<0.001) (Figure 1E). These results suggest that the precursors in BRAF V600E-positive LCH are BM cells rather than PBMC. In order to assess BMD, we also measured the mutational load of BRAF V600E in 12 patients whose samples were available at multiple time points during treatment and disease progression. Refractory patients showed a high mutational load during the treatment (Figure 1F). Furthermore, we examined CD34-positive cells in BM samples of four patients and detected CD34-positive cells harboring BRAF V600E in three MSRO+ patients. These results indicate that BM is a source of immature precursor cells harboring BRAF V600E, causing residual disease even during chemotherapy, and that BMD is a potential target for the treatment of BRAF V600E-positive LCH. Hematopoietic involvement is defined based on clinical criteria, such as anemia and/or thrombocytopenia. The clinical significance of CD1a positivity remains to be confirmed because of its limited sensitivity.10-15 Measurable BMD was detected in 84% of BRAF V600E-positive cases, whereas cytopenia and CD1a positivity were detected in only 32% and 11% of cases, respectively. A comprehensive pathology review, including immunostaining using with BRAF V600E antibody clone VE1, was performed on patients from whom additional BM pathology specimens were obtained. The results showed no specific findings in the BM clots. Furthermore, BRAF V600E was detected by immunostaining in only one of 25 patients. These results suggest that allele-specific ddPCR is more reliable than CD1a or BRAF V600E immunostaining for BMD identification (Table 2). The initial treatment comprised prednisolone, cytarabine, and vincristine for 6 weeks for all patients with systemic LCH, and in most progressive patients, the combination of cytarabine and cladribine was administered as second-line therapy.16 In addition, refractory patients P20, P21, and P41 were treated with vemurafenib, a BRAF inhibitor. Transfusion-dependent cytopenia was improved rapidly in all three patients, showing clinical efficacy. However, only one of the three patients showed a rapid decrease in BMD, whereas the remaining two patients showed an increase in BMD, contrary to the clinical efficacy (Figure 1F). Further studies are needed to elucidate this mechanism. We acknowledge that this was a retrospective study with different treatments and is biased toward high-risk patients receiving BM aspiration. Interestingly, BMD was identified not only in the MSRO+ patients but also in the MSRO- and MFB patients, suggesting that BMD is potentially a risk factor in these groups. Five patients, P2, P4,

Haematologica | 107 July 2022

1720

LETTER TO THE EDITOR Table 1. Characteristics of 38 patients subjected to allele-specific digital droplet polymerase chain reaction of bone marrow samples.

Case

Sample Age, Y

Sex

Disease type

Progression/ relapse

Observation period, Y

VAF of BM, %

VAF of PBMC, %

Somatic variants

P1 P2

BM BM

1.8 5.4

M F

MSROMSRO-

relapse relapse

3.7 3.9

0.25 0.00

NA NA

BRAF V600E BRAF V600E

16.1

neg

2.4

0.00

BRAF N486_P490del

P4 P5

BM BM

0.8 0.9

F M

MSRO+ MSRO-

relapse relapse

3.2 3.2

4.46 0.94

NA 0.30

1.7

MSRO-

relapse

3.1

0.00

P7 P8

BM BM

2.3 1.7

M F

MFB MFB

relapse neg

2.4 2.2

0.26 0.51

0.00 0.09

FFPE

0.8

MSRO-

neg

22.4

0.00

P10 P11 P12

FFPE FFPE FFPE

0.4 1.4 0.6

M F F

MSROMSROMSRO+

neg relapse neg

14.0 12.1 10.6

0.00 2.24 0.98

NA NA NA

P13

FFPE

5.4

MSRO-

relapse

9.9

0.00

P14

FFPE

1.7

MFB

neg

4.8

0.00

P15 P16

FFPE FFPE

0.4 1.5

F F

MSROMSRO+

neg neg

3.0 2.8

0.13 6.96

NA NA

P17

FFPE

0.1

MSRO-

neg

2.4

0.00

P18 P19 P20 P21

BM BM BM BM

7.2 6.9 0.1 0.3

M F M F

MFB MFB MSRO+ MSRO+

neg neg progression progression

1.5 3.8 1.5 1.7

0.00 0.00 0.42 0.54

0.00 0.00 0.20 0.27

P22

8.1

neg

1.4

0.00

P23

MSRO+

relapse

1.3

1.07

0.26

P24

0.8

MFB

relapse

1.3

0.00

P25

1.4

neg

1.1

0.00

P26 P28

BM BM

14.1 2.5

M F

SS MSRO-

neg neg

1.2 1.2

0.00 0.03

0.00 0.00

P29

5.4

neg

1.0

0.00

P30

MFB

neg

1.1

0.00

P31 P32 P35 P36 P37 P38 P39 P40 P41

BM BM FFPE FFPE BM BM BM FFPE BM

1.5 0.7 4.6 0.8 3 1.2 0.4 0.3 0.5

M M F M F M M F M

MSRO+ MSRO+ MSRO+ MSRO+ MSROMFB MSRO+ MSROMSRO+

progression progression progression neg neg neg progression relapse progression

1.8 1.0 0.7 0.3 0.6 0.5 0.4 5.4 0.3

0.83 0.20 5.39 0.21 0.03 0.00 4.10 0.84 3.40

0.59 NA NA 0.00 0.02 0.00 3.50 NA 1.20

BRAF V600E BRAF V600E BRAF N486_P490del BRAF V600E BRAF V600E MAP2K1 Q56_Q61>R BRAF V600D BRAF V600E BRAF V600E BRAF N486_P490del BRAF V600delinsDL BRAF V600E BRAF V600E BRAF N486_P490del BRAF V600E BRAF V600D BRAF V600E BRAF V600E MAP2K1 Q56_V60del BRAF V600E MAP2K1 Q56_Q61>R MAP2K1 E102_I103del BRAF V600E BRAF V600E MAP2K1 Q56_G61>R MAP2K1 E102_I103del BRAF V600E BRAF V600E BRAF V600E BRAF V600E BRAF V600E BRAF V600E BRAF V600E BRAF V600E BRAF V600E

BM: bone marrow; FFPE: formalin-fixed paraffin-embedded; M: male; F: female; MSRO-: multisystem disease; MSRO+: multisystem disease with risk organ involvement; MFB: multifocal bone disease; VAF: variant allele frequency; PBMC: peripheral mononuclear cell. cells; NA: not applicable. Haematologica | 107 July 2022

1721

LETTER TO THE EDITOR

Figure 1. Bone marrow disease in pediatric Langerhans cell histiocytosis. (A) Distribution of somatic mutations in 41 paired tumor and bone marrow (BM) samples in patients with Langerhans cell histiocytosis (LCH). We performed polymerase chain reaction (PCR)-based targeted next generation sequencing using custom-designed primers. Genomic DNA (40–100 ng) was amplified using KOD FX Neo DNA polymerase (Toyobo Co., Ltd., Osaka, Japan) under the following conditions: 98 °C for 30 seconds (s), followed by 42 cycles at 95 °C for 5 s, 55–60 °C for 5 s, and 72 °C for 30 s. The amplicon target and primer sequences are shown in the Online Supplementary Table S1. All somatic mutations were identified using CLC Genomic Workbench v.4.9 (CLC bio, Aarhus, Denmark) for sequence alignment with the human genome reference genome (hg19). (B) Measurable bone marrow disease (BMD) and clinical phenotypes in 38 cases. We conducted droplet digital PCR assay using PrimePCR ddPCR Mutation Assay (dHsaCP2000027, dHsaCP2000028; Bio-Rad, Hercules, CA, USA) according to the manufacturer’s protocol for validation of the BRAF V600E variant; the sequences, annealing temperature, and two-dimensional plots of each custom primer are shown in the Online Supplementary Table S1. A total of 80−200 ng DNA was used for the assay. All samples were analyzed in duplicates, and only those that were positive in both were defined as positive. The outcome data were analyzed using QuantaSoft version 1.7.4.0917 (Bio-Rad). MSRO+; multisystem with risk organs; MSRO-: multisystem without risk organs; MFB: multiple focal bone disease; SS: single system, single Continued on following page. Haematologica | 107 July 2022

1722

LETTER TO THE EDITOR lesion. (C) BMD and age. (D) The cumulative incidence curves of progression and relapse. BMD-positive patients (red line) have a higher LCH progression risk than BMD-negative patients (blue line). (E) Comparison of the mutational burdens in paired mononuclear cells of the BM and peripheral blood. The variant allele frequency (VAF) of BRAF V600E was analyzed in 31 paired cell samples. (F) Persistent detection of BMD in LCH during treatment. Serial quantitative mutation analyses at multiple time points were performed in 12 applicable cases (8 MSRO+, 2 MSRO-, and 2 MFB). Nine patients, including 7 MSRO+, 1 MSRO-, and 1 MFB case, relapsed or had disease progression, representing active diseases that required a change in treatment. In 3 patients, somatic mutations were no longer detectable.

Table 2. CD1a positivity and cytopenia in BRAF V600E-positive Langerhans cell histiocytosis cases.

Negative Negative Negative Negative Negative Negative Negative Negative NA NA NA Positive Negative Negative Negative Negative NA Negative Positive NA Negative NA Negative Negative Negative

BRAF V600E positivity (VE1) of BM Negative Negative Negative Negative Negative Negative Negative Negative Negative Negative NA NA Positive Negative NA Negative Negative NA Negative Negative Negative NA NA NA NA

VAF of BRAF V600E in BM, % 0.25 0.00 4.46 0.94 0.26 0.51 2.24 0.98 0.13 6.96 0.00 0.42 0.54 1.07 0.00 0.03 0.83 0.20 5.39 0.21 0.03 0.00 4.10 0.84 3.40

2/19 cases (11%)

1/17cases (6%)

21/25 cases (84%)

Case

Cytopenia

CD1a positivity of BM

P1 P2 P4 P5 P7 P8 P11 P12 P15 P16 P18 P20 P21 P23 P26 P28 P31 P32 P35 P36 P37 P38 P39 P40 P41

No No No No No No No No No Yes No Yes Yes No No No Yes No Yes Yes No No Yes No Yes 8/25 cases (32%)

BM: bone marrow; NA: not applicable; VAF: variant allele frequency.

P5, P9, and P37, were found to have complications of central diabetes insipidus, two of which were negative for BMD. Two patients, P4 and P31, developed neurodegenerative diseases, and both were positive for BMD. However, the association between BMD and late complications was not clear. The effect of BMD should be evaluated prospectively. Our results confirm previous mutational analysis results in BM samples in a small number of patients.6-9 Furthermore, we present novel findings that BMD positivity correlates with prognosis and that quantitative analysis defines the disease type. These results support the current theory of BRAF mutation acquisition in HSC or nearby progenitor cells in systemic LCH.1-4,6

In conclusion, we found that BMD was frequently detected at the molecular level in pediatric BRAF V600Epositive LCH. A high mutational load in the BM was correlated with distinct clinical features of young-age patients with high-risk multisystem LCH, suggesting that measurable BMD is a novel risk factor that could provide an alternative to cytopenia or CD1a positivity. Further prospective studies with larger cohorts are required.

Authors Ko Kudo,1 Tsutomu Toki,1 Rika Kanezaki,1 Tatsuhiko Tanaka,1 Takuya Kamio,1 Tomohiko Sato,1 Shinya Sasaki,1 Masaru Imamura,2 Chihaya

Haematologica | 107 July 2022

1723

LETTER TO THE EDITOR Imai,2 Kumiko Ando,3 Harumi Kakuda,3 Takehiko Doi,4 Hiroshi

Disclosures

Kawaguchi,4 Masahiro Irie,5 Yoji Sasahara,5 Akihiro Tamura,6

PM receives honoraria from AbbVie, Amgen, Celgene, Janssen, and

Daiichiro Hasegawa, Yosuke Itakura, Kenichiro Watanabe, Kenichi 8

Sakamoto, Yoko Shioda, Motohiro Kato, 11

8,9

Pharmaceuticals', and Takeda's Board of Directors or advisory

Kazuko Kudo, Reiji

committees, and is a member of AbbVie's, Amgen's, Celgene's, and

Fukano, Atsushi Sato, Hiroshi Yagasaki, Hirokazu Kanegane, 15

Janssen Pharmaceuticals Speakers Bureau. JL is employed by

Itaru Kato, Katsutsugu Umeda, Souichi Adachi, Tatsuki 17,18

Kataoka,

Etsuro Ito

Akira Kurose, Atsuko Nakazawa,

Takeda, is a member of AbbVie's, Amgen's, Celgene's, Janssen

Janssen Pharmaceuticals. JJ received honoraria from Amgen and

Kiminori Terui and

1,21

Mundipharma. MD received speaker honoraria from Amgen, BMS Celgene, Janssen, Takeda, Sanofi. NWCJvdD is a consultant for

Amgen, Bayer, Bristol-Myers Squibb, Celgene, Janssen

Department of Pediatrics, Hirosaki University Graduate School of 2

Medicine, Hirosaki; Department of Pediatrics, Niigata University

Pharmaceuticals, Novartis, Servier, and Takeda and receives

Graduate School of Medical and Dental Sciences, Niigata;

research funding from Amgen, Bristol-Myers Squibb, Celgene,

Janssen Pharmaceuticals, and Novartis. SZ receives research

Department of Hematology and Oncology, Chiba Children's Hospital, 4

funding from and is a member of the Board of Directors or

Chiba; Department of Pediatrics, Hiroshima University Graduate 5

School of Biomedical and Health Sciences, Hiroshima; Department of

advisory committees for Celgene, Janssen Pharmaceuticals, and

Pediatrics, Tohoku University Graduate School of Medicine, Sendai;

Takeda. JV is employed by and owns equity in Janssen

Pharmaceuticals. PS receives honoraria and research funding from

Hyogo Prefectural Kobe Children's Hospital, Kobe; 7Department of

Amgen, Celgene, Janssen Pharmaceuticals, Karyopharm, and

Hematology and Oncology, Shizuoka Children's Hospital, Shizuoka;

Takeda and receives researching funding from Skyline. AB receives

honoraria from Amgen, Bristol-Myers Squibb, Celgene, and Janssen

Department of Hematology and Oncology, Children's Cancer Center,

Children’s Cancer Center, National Center for Child Health and 9

Pharmaceuticals. All other authors declare no competing interests.

Development, Tokyo; Department of Pediatrics, Graduate School of 10

Medicine, The University of Tokyo, Tokyo; Department of Pediatrics, Fujita Health University School of Medicine, Aichi; 11Department of

Contributions

Pediatrics, Yamaguchi University Graduate School of Medicine,

KK, TTo and EI designed the study; KK, TT, RK and TT performed

Yamaguchi; Department of Hematology and Oncology, Miyagi

the experiments; KK, TT and RK analyzed and interpreted the data;

Children's Hospital, Sendai; 13Department of Pediatrics, Nihon

KK and EI wrote the manuscript; KK, TS, TK, SS, MI, CI, KA, HK, TD,

University Graduate School of Medicine, Tokyo; Department of Child

HK, MI, YS, AT, DH, YI, KW, KS, YS, MK, KaK, IK, KU, SA, AS, HY, HK,

Health and Development, Graduate School of Medical and Dental

RF and KT evaluated the patients and collected the clinical data;

Sciences, Tokyo Medical and Dental University, Tokyo; Department of

KK, TK, AK and AN collected formalin-fixed paraffin-embedded

Pediatrics, Graduate School of Medicine, Kyoto University, Kyoto;

samples and pathological data. All authors have read and approved

the final manuscript.

Department of Human Health Sciences, Graduate School of 17

Medicine, Kyoto University, Kyoto; Department of Diagnostic Pathology, Graduate School of Medicine, Kyoto University, Kyoto;

Acknowledgments

The authors would like to thank R. Shirai, H. Kudo, and M.

Department of Pathology, Iwate Medical University, Iwate, Japan;

Department of Anatomic Pathology, Hirosaki University Graduate

School of Medicine, Hirosaki;

Department of Clinical Research, 21

Saitama Children's Medical Center, Saitama and Department of

Hashimoto for their technical assistance. The ddPCR analysis was supported by the Scientific Research Facility Center of Hirosaki University Graduate School of Medicine.

Community Medicine, Hirosaki University Graduate School of Funding

Medicine, Hirosaki, Japan

This study was partially supported by the Japan Society for the Correspondence:

Promotion of Science through a Grant-in-Aid for Scientific

Ko Kudo - kkudo@hirosaki-u.ac.jp

Research (KAKENHI: 17K10095, 21K07813) and Japan Cancer

Etsuro Ito - eturou@hirosaki-u.ac.jp

Research Project 21ck0106605h0002 from the Japan Agency for

https://doi.org/10.3324/haematol.2021.279857

Medical Research and Development (AMED).

Received: August 23, 2021.

Data-sharing statment For original data, please contact kkudo@hirosaki-u.ac.jp

Accepted: March 8, 2022. Prepublished: March 17, 2022.

References 1. Bigenwald C, Le Berichel J, Wilk CM, et al. BRAFV600E-induced senescence drives Langerhans cell histiocytosis pathophysiology. Nat Med. 2021;27(5):851-861. 2. McClain KL, Bigenwald C, Collin M, et al. Histiocytic disorders.

Nat Rev Dis Primers. 2021;7(1):73. 3. Allen CE, Merad M, McClain KL. Langerhans-cell histiocytosis. N Engl J Med. 2018;379(9):856-868. 4. Rodriguez-Galindo C, Allen CE. Langerhans cell histiocytosis.

Haematologica | 107 July 2022

1724

LETTER TO THE EDITOR Blood. 2020;135(16):1319-1331. 5. Badalian-Very G, Vergilio JA, Degar BA, et al. Recurrent BRAF mutations in Langerhans cell histiocytosis. Blood. 2010;116(11):1919-1923. 6. Berres ML, Lim KPH, Peters T, et al. BRAF-V600E expression in precursor versus differentiated dendritic cells defines clinically distinct LCH risk groups. J Exp Med. 2014;211(4):669-683. 7. Durham BH, Roos-Weil D, Baillou C, et al. Functional evidence for derivation of systemic histiocytic neoplasms from hematopoietic stem/progenitor cells. Blood. 2017;130(2):176-180. 8. Milne P, Bigley V, Bacon CM, et al. Hematopoietic origin of Langerhans cell histiocytosis and Erdheim-Chester disease in adults. Blood. 2017;130(2):167-175. 9. Milne P, Bigley V, Gunawan M, Haniffa M, Collin M. CD1c+ blood dendritic cells have Langerhans cell potential. Blood. 2015;125(3):470-473. 10. Haupt R, Minkov M, Astigarraga I, et al. Langerhans cell histiocytosis (LCH): guidelines for diagnosis, clinical work-up, and treatment for patients till the age of 18 years. Pediatr Blood Cancer. 2013;60(2):175-184. 11. Minkov M, Pötschger U, Grois N, Gadner H, Dworzak MN. Bone marrow assessment in Langerhans cell histiocytosis. Pediatr Blood Cancer. 2007;49(5):694-698.

12. Kumar M, Sachdeva MUS, Naseem S, et al. Bone marrow infiltration in Langerhan's cell histiocytosis - An unusual but important determinant for staging and treatment. Int J Hematol Oncol Stem Cell Res. 2015;9(4):193-197. 13. Kim H-K, Park C-J, Jang S, et al. Bone marrow involvement of Langerhans cell histiocytosis: immunohistochemical evaluation of bone marrow for CD1a, Langerin, and S100 expression. Histopathology. 2014;65(6):742-748. 14. Galluzzo ML, Braier J, Rosenzweig SD, Garcia De Dávila MT, Rosso D. Bone marrow findings at diagnosis in patients with multisystem langerhans cell histiocytosis. Pediatr Dev Pathol. 2010;13(2):101-106. 15. Tzankov A, Kremer M, Leguit R, et al. Histiocytic cell neoplasms involving the bone marrow: summary of the workshop cases submitted to the 18th Meeting of the European Association for Haematopathology (EAHP) organized by the European Bone Marrow Working Group, Basel 2016. Ann Hematol. 2018;97(11):2117-2128. 16. Morimoto A, Shioda Y, Imamura T, et al. Intensified and prolonged therapy comprising cytarabine, vincristine and prednisolone improves outcome in patients with multisystem Langerhans cell histiocytosis: results of the Japan Langerhans Cell Histiocytosis Study Group-02 Protocol Study. Int. J. Hematology. 2016;104(1):99-109.

Haematologica | 107 July 2022

1725

LETTER TO THE EDITOR

Treatment emergent peripheral neuropathy in the CASSIOPEIA trial Peripheral neuropathy (PNP) is one of the most common adverse events of multiple myeloma (MM) treatment. Thalidomide and bortezomib are particularly prone to induce treatment-emergent peripheral neuropathy (TEPN).1-4 Both agents are part of standard treatment regimens for newly diagnosed transplant-eligible MM patients. Recently, daratumumab was added to the combination of thalidomide and bortezomib, introducing a quadruplet regimen, which has significantly improved progression-free survival (PFS) in newly diagnosed MM patients, as demonstrated in the CASSIOPEIA trial.5 Following these positive results, this quadruplet regimen was approved by the European Medicines Agency (EMA) in 2019 for use in newly diagnosed MM. TEPN however remains a concern, and whether daratumumab impacts TEPN, was analysed in this trial. TEPN varies from mild symptoms to severe disability and is characterized by mainly sensory symptoms, such as paresthesia and neuropathic pain.6,7 The most frequent used grading system to grade the severity of the TEPN are the common terminology criteria for adverse events (CTCAE) criteria, as used in the CASSIOPEIA trial. TEPN has a significant impact on quality of life (QoL) from grade 2 and higher, defined as “limiting instrumental ac-

tivities of daily living (IADL)”. Until now, the incidence of TEPN in quadruplet regimens has not been extensively analyzed. This is the first analysis which focuses on the cumulative incidence of TEPN, the impact on PFS, effect of dose adjustments and potential risk factors for TEPN in a daratumumab-based regimen. The phase 3 CASSIOPEIA trial investigated the efficacy of adding daratumumab to bortezomib, thalidomide and dexamethasone (VTD). The trial design can be found in the Online Supplementary Figure S1. We analyzed PNP grade 2 to 4, scored according to CTCAE version 4. PNP that occurred after the start of induction therapy is defined as therapy-emergent neuropathy. Patients with PNP grade ≥2 at baseline were excluded from the trial. PNP was defined as peripheral sensory neuropathy and peripheral motor neuropathy and was graded and reported by investigators. The TEPN assessment was performed from start of induction until end of maintenance. The complete clinical trial report of part 1 of the study was published in 2019.5 The cumulative incidence of PNP ≥ grade 2 was calculated. Associations of possible risk factors with the cumulative incidence of PNP were evaluated using the method of Fine and Gray, using univariate as well as multivariate

Figure 1. Cumulative incidence of treatment-emergent peripheral neuropathy grade ≥ 2 and grade ≥3 in the CASSIOPEIA trial per arm. PNP: peripheral neuropathy; VTD: bortezomib, thalidomide and dexamethasone; D-VTD: VTD plus daratumumab. Haematologica | 107 July 2022

1726

LETTER TO THE EDITOR analyses. In order to evaluate the impact of PNP on subsequent PFS, two analyses were performed. First, a univariate Cox regression was performed with development of PNP ≥ grade 2 as a time-dependent covariate. This analysis included all patients and takes into account that some patients never developed PNP ≥ grade 2, and other patients at some time point developed PNP ≥ grade 2, which could imply an increased risk to progression or death (i.e., the events for PFS) from that moment on. The hazard ratio (HR) and 95% confidence interval (CI) were calculated. Second, a landmark analysis was performed in which PFS was calculated from start of consolidation, comparing patients with or without PNP ≥ grade 2 during induction treatment. While HR and 95% CI were determined again, PFS for both groups could now also be illustrated by Kaplan-Meier survival curves. In addition, the impact of dose modification of bortezomib and thalidomide on PFS from start consolidation was evaluated in patients who developed PNP ≥ grade 2 until consolidation treatment. PFS was calculated up to 3 years after randomization. A flowdiagram of patients included in the analysis is provided in the Online Supplementary Figure S2. A two-sided P-value lower than 0.05 was considered to be statistically significant. Overall, 1,085 newly diagnosed MM patients were randomized in the CASSIOPEIA study, of which 1,074 initiated treatment in the dara-VTD arm (n=536) or VTD arm (n=538). Baseline characteristics in the two arms were similar and have been reported before.5 At baseline 20 patients (2%) had a medical history of PNP grade 1 (D-VTD n=15; VTD n=5). During the trial, 394 patients (37%) de-

veloped grade 2-4 PNP: maximum grade 2 occurred in 289 (27%) patients, maximum grade ≥3 occurred in 105 (10%) patients. In the VTD arm 38% developed grade ≥2 PNP versus 35% in the dara-VTD arm. The cumulative incidence of grade ≥2 PNP was similar between both arms at 6 months 28% and 22% in the VTD and D-VTD arm respectively, grade≥3 PNP at 6 months was 7% and 4% (Figure 1). We observed that among predefined risk factors patients with a higher body mass index (BMI) had a greater risk for the development of PNP grade ≥2. Per BMI group this risk increased: in the multivariate analysis, subdistribution HR (sHR)=1.39 (95% CI: 1.09-1.77, P=0.008) for BMI 25-30, sHR=1.57 (95% CI: 1.12-2.20, P=0.008) and sHR=2.07 (95% CI: 1.30-3.30, P=0.002) for BMI>35, compared to the group BMI<25 (Table 1). Another risk factor associated with a higher cumulative incidence was PNP grade 1 at baseline (sHR=2.64, 95% CI: 1.39-4.99, P=0.003) and an older age (sHR=1.02, 95% CI: 1.00-1.03, P=0.04). Multivariate analysis also indicated that the cumulative incidence of PNP grade ≥2 was significantly lower in the dara-VTD arm (33% at 12 months) when compared to the VTD arm (37%) (sHR=0.77, 95% CI: 0.63-0.95, P=0.01). A significant difference in the incidence of PNP grade ≥2 between the France and the Netherlands was also observed (33% vs. 49%) (sHR=1.73, 95% CI: 1.30-2.30, P<0.001). A biological explanation for this difference has not been found. In order to evaluate the impact of PNP during induction on subsequent PFS, two analyses were performed. First, the development of grade ≥2 PNP during any time point

Figure 2. Kaplan-Meier curves of progression-free survival from start consolidation. Progression-free survival of patients who had experienced treatment-emergent peripheral neuropathy during induction vs. patients who did not experience treatment emergent peripheral neuropathy (Yes vs. No). Haematologica | 107 July 2022

1727

LETTERS TO THE EDITOR Table 1. Risk factors for treatment-emergent peripheral neuropathy grade 2-4 by uni- and multivariate analysis.

Univariate analysis Cumulative Risk of PNP grade 2-4 incidence* (%) sHR (95% CI) Sex M F Age group < 50 ≥ 50-65 Age Arm No Dara Dara Cytogenetics Std risk High risk ISS-stage I II III Country France Netherlands Belgium Creat Clearance < 60 60 – 90 > 90 DM No Yes BMI < 25 25-30 30-35 >35 ECOG PS 0 1 2 PNP gr I at baseline No Yes Liver function Normal Impaired Disease type IgG IgA Urine only FLC Other Baseline M-prot Baseline % PC

P-value

37 33

0.89 (0.73-1.09)

0.27

30 36 n.a.

1.26 (0.95-1.68) 1.02 (1.00-1.04)

37 33

Multivariate analysis Risk of PNP grade 2-4 sHR (95% CI)

P-value

0.90 (0.72-1.13) Not included

0.37

0.11 0.01

1.02 (1.00-1.03)

0.04

0.86 (0.71-1.05)

0.13

0.77 (0.63-0.95)

0.01

36 34

0.96 (0.72-1.27)

0.76

1.07 (0.79-1.44)

0.67

40 33 30

0.76 (0.61-0.93) 0.69 (0.51-0.94)

0.01 0.02

0.83 (0.64-1.06) 0.78 (0.54-1.13)

0.14 0.19

33 49 38

1.67 (1.28-2.17) 1.19 (0.83-1.71)

<0.001 0.34

1.73 (1.30-2.30) 1.23 (0.86-1.78)

<0.001 0.26

26 36 36

1.48 (0.93-2.36) 1.48 (0.94-2.33)

0.10 0.09

1.29 (0.78-2.12) 1.10 (0.66-1.85)

0.32 0.70

35 44

1.37 (0.98-1.93)

0.067

1.23 (0.88-1.73)

0.23

29 40 41 48

1.47 (1.18-1.83) 1.60 (1.18-2.17) 2.04 (1.31-3.17)

0.001 0.002 0.002

1.39 (1.09-1.77) 1.57 (1.12-2.20) 2.07 (1.30-3.30)

0.008 0.008 0.002

38 34 28

0.87 (0.71-1.07) 0.73 (0.50-1.06)

0.19 0.10

0.90 (0.73-1.12) 0.76 (0.51-1.12)

0.34 0.17

35 75

3.08 (1.83-5.21)

<0.001

2.64 (1.39-4.99)

0.003

35 37

1.04 (0.75-1.44)

0.83

1.16 (0.83-1.63)

0.38

36 34 36 35 31 n.a. n.a.

0.95 (0.72-1.26) 0.97 (0.71-1.31) 0.99 (0.68-1.43) 0.91 (0.51-1.63) 0.97 (0.93-1.01) 1.00 (0.99-1.00)

0.71 0.83 0.96 0.75 0.12 0.18

0.96 (0.71-1.29) 0.97 (0.63-1.48) 0.90 (0.57-1.43) 0.86 (0.48-1.57) 0.99 (0.92-1.07) 1.00 (0.99-1.00)

0.79 0.87 0.66 0.63 0.80 0.16

*Cumulative incidence at 1 year. ISS: International Staging System; DM: diabetes mellitus; BMI: body mass index; ECOG PS: Eastern Cooperative Oncology Group Performance Status; PNP: peripheral neuropathy; FLC: free light chain; Std: standard; DARA: daratumumab; prot: protein; gr: grade; Ig: immunoglobulin; CI: confidence interval; sHR: subdistribution hazard ratio; PC: Plasma cells, DM: is already described in the legend, n.a.: not applicable. Haematologica | 107 July 2022

1728

LETTER TO THE EDITOR of the trial was included as a time-dependent covariate in a univariate Cox regression analysis (HR=0.83, 95% CI: 0.67-1.04, P=0.10). Second, a landmark analysis showed that PFS from start consolidation was similar, 80% versus 82% at 2 years (HR=0.86, 95% CI: 0.661.13, P=0.27) for patients developing grade ≥2 PNP during induction therapy versus patients who did not develop grade ≥2 PNP (Figure 2). Dose reduction and/or discontinuation of bortezomib or thalidomide due to PNP grade ≥2 did not impact PFS in comparison to patients who did not experience PNP. PFS at 2 years from start consolidation in patients who received dose modification of bortezomib and/or thalidomide was 83% versus 78% in patients receiving the full dosage of both (HR=1.05, 95% CI: 0.62-1.78, P=0.87) (Online Supplementary Figure S3A). There was also no significant difference in patients receiving either a reduced-dosage thalidomide or bortezomib in response to PNP grade ≥2 versus patients receiving the full dosage (Online Supplementary Figures 3B and C). With the introduction of quadruplet regimens the impact of these regimens on QoL remains important. One of the most debilitating adverse events is peripheral neuropathy. Dara-VTD will be used widely in many countries and in this sub-analysis of the CASSIOPEIA trial we present a detailed analysis of TEPN in the two regimens used in this trial. We observed a clinically relevant cumulative incidence of 37% grade ≥2 PNP. We report an incidence of TEPN which is similar to other trials with subcutaneous bortezomib-based regimens.8-11 When comparing triplet regimens, a high incidence of TEPN is reported in VRD and VTD treatment, whereas VCD regimens seem associated with lower incidence of TEPN.10 Studies including daratumumab in one of the two arms reported a PNP grade ≥3 incidence which was slightly lower in the daratumumab arm8,9 consistent with the data we present here. This suggests a possible beneficial effect of daratumumab on TEPN. However, an explanation for a possible beneficial mechanism has to our knowledge not been described. With other large phase III trials analyzing the possible beneficial effect of quadruplet therapy still ongoing, such as the Perseus (VRD vs. Dara-VRD) and Iskia (KRD vs. Isatuximab-KRD), data on the incidence of TEPN will need to be closely monitored. The concern remains that modification of the treatment doses lead to a worse treatment response and reduced PFS.12,13 Here, PFS at 2 years was similar in patients who did or did not receive dose modification of bortezomib and/or thalidomide. Previous studies confirm these findings, reporting no difference in PFS, response rates and OS between the patients with or without PNP.13,14 Although PFS is not influenced by TEPN, it is known to impact other features: such as, QoL, inclusion in subsequent clinical trials - often excluding patients with a PNP grade 2 or higher - and the use of subsequent drugs. Unfor-

tunately due to the retrospective nature of this analysis we could not generate data on these issues. This study and a recently published prespecified secondary analysis on QoL in the CASSIOPEIA trial15 are of great importance. The incidence of TEPN in patients receiving myeloma treatment with bortezomib and thalidomide remains clinically significant. In addition, identifying the patients at risk is essential. Risk factors for the development of grade ≥2 PNP included PNP at baseline, older age and BMI >25. As Dara-VTD has been approved by EMA, this regimen will soon become the standard treatment in newly diagnosed myeloma patients in many European countries. These results highlight the need for increased awareness of TEPN.

Authors Cathelijne Fokkema,1 Phillipe Moreau,2 Bronno van der Holt,3 Jérôme Lambert,4 Mark van Duin,1 Ruth Wester,1 Joost L.M. Jongen,5 Pieter A. van Doorn,5 Sophie Godet,6 KonSiong Jie,7 Olivier Fitoussi,8 Michel Delforge,9 Awa Keita-Manta,10 Odile Luycx,11 Tom Cupedo,1 Niels W.C.J. van de Donk,12 Sonja Zweegman,12 Jessica T. Vermeulen,13 Pieter Sonneveld1 and Annemiek Broijl1 1

Department of Hematology, Erasmus MC Cancer Institute,

Rotterdam, the Netherlands; 2Hematology Department, University Hospital Hôtel-Dieu, Nantes, France; 3HOVON Data Center, Department of Hematology, Erasmus MC Cancer Institute, Rotterdam, the Netherlands; 4Biostatistical Department, Hospital Saint Louis, Paris, France; 5Department of Neurology, Erasmus MC, Rotterdam, the Netherlands; 6Department of Hematology, University Hospital of Reims and UFR Médicine, Reims, France; 7

Department of Hematology, Zuyderland MC, Sittard, the

Netherlands; 8Department of Hematology, Polyclinique Bordeaux Nord Aquitaine, Bordeaux, France; 9Department of Hematology, Universitaire Ziekenhuizen Leuven, Leuven, Belgium; 10Premier Research, CRO, Paris, France; 11Department of Hematology, Hospital Scorff Hospital Group Bretagne Sud, Lorient, France; 12

Department of Hematology, Amsterdam UMC, Cancer Center

Amsterdam, Amsterdam, the Netherlands and 13Janssen Research & Development, LLC, Leiden, the Netherlands Correspondence: A. Broijl - a.broyl@erasmusmc.nl https://doi.org/10.3324/haematol.2021.280567 Received: January 18, 2022. Accepted: March 9, 2022. Prepublished: March 17, 2022. Disclosures PM receives honoraria from AbbVie, Amgen, Celgene, Janssen, and Takeda, is a member of AbbVie's, Amgen's, Celgene's, Janssen Pharmaceuticals', and Takeda's Board of Directors or advisory committees, and is a member of AbbVie's, Amgen's, Celgene's, and

Haematologica | 107 July 2022

1729

LETTER TO THE EDITOR Janssen Pharmaceuticals Speakers Bureau. JL is employed by Janssen Pharmaceuticals. JJ received honoraria from Amgen and Mundipharma. MD received speaker honoraria from Amgen, BMS Celgene, Janssen, Takeda, Sanofi. NWCJvdD is a consultant for Amgen, Bayer, Bristol-Myers Squibb, Celgene, Janssen Pharmaceuticals, Novartis, Servier, and Takeda and receives research funding from Amgen, Bristol-Myers Squibb, Celgene, Janssen Pharmaceuticals, and Novartis. SZ receives research funding from and is a member of the Board of Directors or advisory committees for Celgene, Janssen Pharmaceuticals, and Takeda. JV is employed by and owns equity in Janssen Pharmaceuticals. PS receives honoraria and research funding from Amgen, Celgene, Janssen Pharmaceuticals, Karyopharm, and Takeda and receives researching funding from Skyline. AB receives honoraria from Amgen, BristolMyers Squibb, Celgene, and Janssen Pharmaceuticals. All other authors declare no competing interests. Contributions AC, PM, BvdH, JJ, PvD, TC, PS and AB developed the study concept; AC, BvdH, JL, and AB set up the study methology; AC, BvdH and JL carried out the investigation; AC, BvdH and JL performed the formal analysis; PM, MvD, RW and SG provided resources; KSJ, OF, MD, AK, OL, NvdD, SZ and AB collected data; AC, BvdH and JL were in charge of data curation; AC and BvdH visualized the project; AC and AB wrote the original draft. All authors contributed to writing,

reviewing and editing of the manuscript. PS acquired funding; TC, AB and PS supervised the project; AB was in charge of project administration. Acknowledgments We thank members of Myeloma Research Rotterdam and the Department of Hematology for critical discussions and reading of the manuscript. We thank the patients who participated in this study, the staff members at the study sites, the data and safety monitoring committee, and the staff members involved in data collection and analyses. Funding This study was supported by Intergroupe Francophone du Myélome (IFM), the Dutch-Belgian Cooperative Trial Group for Hematology Oncology (HOVON) in collaboration with Janssen Research & Development and the Dutch Cancer Society (KWF). Data-sharing statment The data sharing policy of Janssen Pharmaceutical Companies of Johnson & Johnson is available online. Requests for access to the study data can be submitted through the Yale Open Data Access Project site. For the Janssen data sharing policy see https://www.janssen. com/clinical-trials/transparency For the Yale Open Data Access Project site see http://yoda.yale.edu.

References 1. Luczkowska K, Litwinska Z, Paczkowska E, et al. Pathophysiology of drug-induce peripheral neuropathy in patients with multiple myeloma. J Physiol Pharmacol. 2018;69(2). 2. Richardson PG, Xie W, Mitsiades C, Chanan-Khan AA, Lonial S, Hassoun H, et al. Single-agent bortezomib in previously untreated multiple myeloma: efficacy, characterization of peripheral neuropathy, and molecular correlations with response and neuropathy. J Clin Oncol. 2009;27(21):3518-3525. 3. Jongen JLM, Broijl A, Sonneveld P. Chemotherapy-induced peripheral neuropathies in hematological malignancies. J Neuroncol. 2015;121(2):229-237. 4. Sonneveld P, Jongen JLM. Dealing with neuropathy in plasmacell dyscrasias. Hematology. 2010;2010(1):423-430. 5. Moreau P, Attal M, Hulin C, et al. Bortezomib, thalidomide, and dexamethasone with or without daratumumab before and after autologous stem-cell transplantation for newly diagnosed multiple myeloma (CASSIOPEIA): a randomised, open-label, phase 3 study. Lancet. 2019;394(10192):29-38. 6. Argyriou AA, Iconomou G, Kalofonos HP. Bortezomib-induced peripheral neuropathy in multiple myeloma: a comprehensive review of the literature. Blood. 2008;112(5):1593-1599. 7. Badros A, Goloubeva O, Dalal JS, et al. Neurotoxicity of bortezomib therapy in multiple myeloma: a single-center experience and review of the literature. Cancer. 2007;110(5):1042-1049. 8. Voorhees PM, Kaufman JL, Laubach J, et al. Daratumumab, lenalidomide, bortezomib, and dexamethasone for transplanteligible newly diagnosed multiple myeloma: the GRIFFIN trial.

Blood. 2020;136(8):936-945. 9. Mateos M-V, Dimopoulos MA, Cavo M, et al. Daratumumab plus bortezomib, melphalan, and prednisone for untreated myeloma. N Engl J Med. 2017;378(6):518-528. 10. Moreau P, Hulin C, Macro M, et al. VTD is superior to VCD prior to intensive therapy in multiple myeloma: results of the prospective IFM2013-04 trial. Blood. 2016;127(21):2569-2574. 11. Merz M, Salwender HJ, Haenel M, et al. Clinical risk factors for peripheral neuropathy in patients treated with subcutaneous or intravenous bortezomib for newly diagnosed multiple myeloma. Blood. 2015;126(23):4233. 12. Chaudhry V, Cornblath DR, Polydefkis M, et al. Characteristics of bortezomib- and thalidomide-induced peripheral neuropathy. J Peripher Nerv Syst. 2008;13(4):275-282. 13. Richardson PG, Sonneveld P, Schuster MW, et al. Reversibility of symptomatic peripheral neuropathy with bortezomib in the phase III APEX trial in relapsed multiple myeloma: impact of a dose-modification guideline. Br J Haematol. 2009;144(6):895-903. 14. Tacchetti P, Terragna C, Galli M, et al. Bortezomib- and thalidomide-induced peripheral neuropathy in multiple myeloma: clinical and molecular analyses of a phase 3 study. Am J Hematol. 2014;89(12):1085-1091. 15. Roussel M, Moreau P, Hebraud B, et al. Bortezomib, thalidomide, and dexamethasone with or without daratumumab for transplantation-eligible patients with newly diagnosed multiple myeloma (CASSIOPEIA): health-related quality of life outcomes of a randomised, open-label, phase 3 trial. Lancet Haematol. 2020;7(12):e874-e883.

Haematologica | 107 July 2022

1730

LETTER TO THE EDITOR

Prolonged viral replication in patients with hematologic malignancies hospitalized with COVID-19 Severe acute respiratory syndrome coronavirus 2 (SARSCoV-2) has emerged as a major cause of mortality worldwide. Immunocompromised hosts, especially patients with hematologic malignancies, have suffered greatly due to this pandemic, with reported mortality rates reaching between 28-44%.1–3 Multiple studies have described clinical features of coronavirus disease 2019 (COVID-19) to be similar in both immunocompromised hosts and other patients.1,4 Similarly, the broad spectrum of COVID-19 clinical phenotypes described in immunocompetent patients, including complications due to immunological hyperactivation, co-infection, and/or coagulopathy, has been also observed in hematologic patients. However, information about viral evolution is scarce. Only a few case reports have documented that immunosuppressed patients may have prolonged viral replication.5– 12 Yet, such viral shedding poses a major threat to both patients and public health. It can delay treatment for hematologic malignancies, increase duration of a patient’s infectivity, and potentially favor selection of mutant strains. This series of patients would, therefore, require a personalized approach in response. We aimed to describe the incidence, clinical features, risk factors and outcomes of prolonged SARS-CoV-2 viral shedding in a cohort of consecutive patients with hematologic malignancies admitted to our hospital. For this purpose, we performed a prospective study including all consecutive adults with hematological malignances hospitalized ≥48 hours for COVID-19 at Hospital Clinic (Barcelona, Spain) between March 2020 and March 2021. Real-time reverse transcription polymerase chain reaction (rRT-PCR) testing performed on nasal and oropharyngeal swab specimens confirmed COVID-19 diagnoses. Hematologic patients with a prolonged positive rRT-PCR underwent further testing for viral sub-genomic mRNA (sgRNA) identification, which is better correlated with active viral replication.13 Prolonged viral replication was defined as patients with a prolonged positive rRT-PCR and positive sgRNA at ≥3 weeks since initial positive rRT-PCR. Only these patients who underwent rRT-PCR testing at ≥3 weeks since diagnosis were eligible for evaluation in this study. In this sense, we discarded all those patients who either died within the first 3 weeks of diagnosis or did not undergo follow-up rRT-PCR testing. High-quality data on characteristics of all patients hospitalized for COVID-19 were directly collected from electronic health records with an intelligent system (SILDv1.0 system). We performed rRT-PCR in most samples (85%) using Cobas®6800 (Roche Diagnostic, Germany) which detected the E and ORF1b genes. In the remaining samples, rRT-PCR was performed using the LightMix ModularDx SARS-CoV (COVID19) E-gene kit (Tib Molbiol, Roche Diagnostics). A

positive result was defined when the cycle threshold (Ct) value of the E gene was ≤38. Presence of sgRNA was tested with primers and probes targeting sequence downstream of the start codons of the E gene with a specific forward leader primer, using SuperScript™ III Platinum™ One-Step qRT-PCR Kit (Invitrogen) in the thermocycler StepOne (Applied Biosystems). A positive result was defined when the Ct value of the E gene was ≤39. Factors associated with prolonged viral shedding were evaluated using a multivariate analysis (step-forward procedure), including all significant variables (P<0.05) obtained from univariate analysis. A two-tailed P<0.05 was considered as significant. A total of 3,216 consecutive adults with COVID-19 were admitted to our hospital for ≥48 hours during the study period, of whom 124 (3.9%) had hematologic malignancies. Of those, 67 (54%) were eligible for viral persistence assessment due to survival status and follow-up rRT-PCR at ≥3 weeks since initial diagnosis. Prolonged viral replication was documented in 17 (25.4%) evaluable patients, representing 17.3% (17 of 98) of all hematologic patients alive after 3 weeks. Online Supplementary Figure S1 details the study flowchart, while Online Supplementary Table S1 compares patient characteristics on the basis of eligibility for viral persistence assessment. The most common hematologic malignancies in the study population were lymphoma and chronic lymphocytic leukemia. Overall, 20.9% of patients had received a hematopoietic stem cell transplant (HSCT), and median C-RP at admission was 6.1 (interquartile range [IQR], 2.2-11.2) mg/dL. Table 1 details epidemiological and clinical characteristics of the cohort. Online Supplementary Table S2 summarizes the main characteristics, clinical features, and outcomes of those 17 patients with prolonged viral replication. In 12 (70.6%) of those patients, sgRNA was negative before genomic rRT-PCR. Median time of sgRNA positivity was 54 (IQR, 32.5-100) days compared to 77 (IQR, 56.5-107) days for rRT-PCR positivity (P<0.001). Six (35.3%) patients died after a median time of 72 (IQR, 53-144) days of receiving a COVID-19 diagnosis. Table 2 shows univariate and multivariate risk factors for prolonged viral replication. In the multivariate analysis, lymphoma (odds ratio [OR] 5.44, 95% confidence interval [CI]: 1.24-23.84); hypogammaglobulinemia (OR 4.64, 95% CI: 1.10-19.60); and prior chemotherapy (OR 27.21, 95% CI 2.88-257.40) were independently associated with prolonged viral replication. The discriminatory power of the score, as evaluated by the area under the receiver operating characteristic curve, was 0.884 (95% CI: 0.8030.965), demonstrating a good ability to predict prolonged viral replication. Some case reports have described the possibility of pro-

Haematologica | 107 July 2022

1731

LETTER TO THE EDITOR Table 1. Main epidemiological and clinical characteristics of hematologic patients.

Patients N=67 (%) Patient's characteristics Median (IQR) age, in years

65 (54-77)

Age > 65 years (%)

35 (52.2)

Sex male, N (%)

42 (62.7)

Hematologic malignancy Lymphoma*

30 (44.8)

Chronic lymphocytic leukemia

10 (14.9)

Multiple myeloma

7 (10.4)

Acute leukemia Myelodysplastic syndrome Others Prior HSCT Allogenic HSCT# Autologous HSCT# Prior CAR-T cell therapy

6 (9) 5 (7.5) 6 (9) 14 (20.9) 4 (6) 12 (17.9) 3 (4.5)

Other comorbidities Hypertension

26 (38.8)

Diabetes mellitus

13 (19.4)

Chronic heart disease

21 (31.3)

Chronic lung disease

7 (10.4)

Chronic renal failure

12 (17.9)

Chronic liver disease

9 (13.4)

Solid organ transplant

5 (7.5)

Other clinical features Prior corticosteroid use (3 months)

35 (52.2)

Prior chemotherapy (3 months)

36 (53.7)

Prior rituximab use (12 months)

15 (22.4)

Current neutropenia (< 500/mm3) Long-term lymphopenia (> 1 month; < 900 lymphocytes/mm3) Hypogammaglobulinemia Active hematologic disease Median (IQR) days from symptom onset to hospital admission

6 (9.0) 32 (47.8) 23/62 (34.3) 40 (59.7) 4 (2-6)

Vital signs at admission; median (IQR) Temperature -Median (ºC)

IQR: interquartile range; HSCT: hematopoietic stem cell transplant; CAR: chimeric antigen receptor; rpm: respirations per minute; C-RP: C-reactive protein; LDH: lactate dehydrogenase. *Diffuse large B-cell lymphoma, 11 patients; follicular lymphoma, 7; Hodgkin lymphoma, 5; T-cell lymphoma, 2; other, 4. #Two allogenic HSCT recipients had undergone a prior autologous transplant.

37.0 (36.2-37.9)

Respiratory rate -Median (rpm)

20 (18-24)

Oxygen saturation -Median

95 (94-97)

Laboratory values at admission; median (IQR) Ferritin (ng/mL)

207 (487-973)

C-RP (mg/dL)

6.1 (2.2-11.2)

D-dimer (ng/mL)

800 (400-1800)

LDH (U/L)

291 (207-365)

Lymphocyte count (cells/mm3)

800 (400-1350)

longed COVID-19 infection in immunosuppressed patients.5–12 We documented that such a prolonged SARSCoV-2 infection is quite frequent. It affected more than 25% of patients with hematologic malignancies, at least in our cohort of patients requiring hospitalization due to COVID-19 and surviving more than 3 weeks. We agree with prior case reports that the clinical spectrum of these infections range from chronic asymptomatic infection to severe presentation and death. This fact, therefore, implies that physicians should actively screen for this complication by performing repeated rRT-PCR after a COVID-19 diagnosis until the patient tests negative, particularly among those with risk factors for prolonged viral replication. In immunocompetent patients, viral culture was positive only in samples with a cycle threshold value of 28 or less.14 Therefore, to preclude the diagnosis of a persistent viable virus, it seems reasonable that physicians perform a sgRNA and/or viral culture in all immunosuppressed patients in whom the rRT-PCR value is persistently positive with Ct values below 28.13 Hospitalized hematologic patients with COVID-19 and hypogammaglobulinemia, recent chemotherapy, and/or lymphoma, especially after rituximab use, face a higher risk of viral persistence. It is understandable that these risk factors are present among patients with prolonged SARSCoV-2 infection. Such factors deeply impair B-cell response and the production of antibodies with neutralizing activity against the virus, which comprise the main factors of the host immune response against the virus.15,16 In order to compensate this deficit, physicians could consider administering convalescent plasma or specific monoclonal antibodies on a repeated basis. Conversely, in this setting, administering dexamethasone as COVID-19 treatment without the presence of an inflammatory phenotype may prove deleterious. Prolonged SARS-CoV-2 replication represents a huge challenge in patient care, public health and medicine. First, viral load persistence could cause delays in required treatment for hematologic malignancies and thereby, significantly contribute to a worsening in a patient’s prognosis. Moreover, mandatory prolonged isolation due to such persistence impacts a patient’s daily social life and decreases quality of life. Second, from a public health perspective, prolonged viral replication is a tremendous problem since it is conceptually feasible that prolonged infections may facilitate the progressive emergence of mutants. However, this is rather theoretical and mainly supported by single case reports. Further studies demonstrating this association are needed. Fi-

Haematologica | 107 July 2022

1732

LETTER TO THE EDITOR Table 2. Univariate and multivariate risk factors for prolonged viral shedding.

Risk factor

Non-prolonged Prolonged viral shedding viral shedding N=50 (%) N=17 (%)

Univariate odds ratio (95% CI)

P-value

Multivariate odds ratio (95% CI)

P-value

Age > 65 years

27 (54.0)

4 (23.5)

0.26 (0.08-0.91)

0.047

0.47 (0.08-2.61)

0.387

Male sex

33 (66.0)

9 (52.9)

0.58 (0.19-1.77)

0.336

Lymphoma

18 (36.0)

12 (70.6)

4.27 (1.30-14.06)

0.013

5.44 (1.24-23.84)

0.025

Acute leukemia

5 (10.0)

1 (5.9)

0.56 (0.61-5.19)

1.000

Myelodysplastic syndrome

5 (10.0)

0 (0)

0.319

Multiple myeloma

4 (8.0)

3 (17.6)

2.46 (0.49-12.35)

0.358

Chronic lymphocytic leukemia

9 (18.0)

1 (5.9)

0.56 (0.07-4.51)

0.432

Chronic myeloid leukemia

3 (6.0)

0 (0)

0.565

Others

6 (12.0)

0 (0)

0.325

Hematopoietic stem cell transplant

10 (20.0)

4 (23.5)

1.23 (0.33-4.60)

0.740

0 (0)

3 (17.6)

0.014

Hypertension

22 (44.0)

4 (23.5)

0.39 (0.11-1.37)

0.160

Diabetes mellitus

11 (22.0)

2 (11.8)

0.47 (0.09-2.39)

0.490

Chronic heart disease

17 (34.0)

4 (23.5)

0.60 (0.17-2.12)

0.550

Chronic lung disease

6 (12.0)

1 (5.9)

0.46 (0.05-4.11)

0.669

Chronic renal failure

11 (22.0)

1 (5.9)

3.18 (1.35-7.49)

0.270

Chronic liver disease

5 (10.0)

4 (23.5)

2.77 (0.65-11.83)

0.216

Solid organ transplant

4 (8.0)

1 (5.9)

0.72 (0.08-6.92)

1.000

Prior corticosteroid use

22 (44.0)

13 (76.5)

4.14 (1.18-14.47)

0.026

1.39 (0.23-8.31)

0.721

Hypogammaglobulinemia

11 (22.0)

11 (64.7)

6.50 (1.96-21.56)

0.001

4.64 (1.10-19.60)

0.037

Active hematologic malignancy

25 (50.0)

15 (88.2)

7.50 (1.55-36.27)

0.009

3.61 (0.53-24.36)

0.188

Prior rituximab (12 months)

5 (10.0)

10 (58.8)

12.86 (3.3848.94)

<0.001

2.52 (0.26-24.60)

0.427

Prior chemotherapy (3 months)

20 (40.0)

16 (94.1)

24.0 (2.95195.60)

<0.001

27.21 (2.88257.40)

0.004

4 (8.0)

2 (11.8)

1.53 (0.26-9.23)

0.639

20 (40.0)

12 (70.6)

3.60 (1.10-11.80)

0.029

2.87 (0.52-15.85)

0.227

CAR-T cell therapy

Current neutropenia Long-term lymphopenia

nally, guidelines regarding diagnostic approaches and treatments in this particular setting are lacking. In our practice, we observed three patterns: i) some of our patients had no response to standard treatment with a 5day course of remdesivir, and convalescent plasma and prolonged antiviral treatment were necessary at times for improvement; ii) some patients experienced a clinical response after receiving remdesivir but either fell ill again or had an increase in SARS-CoV-2 load after antiviral therapy completion; and iii) some patients did not respond to antiviral strategies at all. That stated, each of these situations requires a personalized approach that

would perhaps necessitate the inclusion of prolonged antiviral treatment with remdesivir, combined with monoclonal antibodies and/or plasma. Our study has some limitations that should be noted. We retrospectively described the incidence of prolonged viral replication. However, we may have missed some patients with prolonged viral replication, especially at the beginning of the pandemic when we did not perform follow-up testing. Another limitation is the lack of genetic tests performed in some patients to confirm the same linage. However, we would like to highlight that we have this information for four patients and were able to confirm pro-

Haematologica | 107 July 2022

1733

LETTER TO THE EDITOR longed infection. Finally, viral culture was only available (and positive) in two of the 17 patients; we used sgRNA as a surrogate marker of viral viability. Although there is an open debate, our experience has shown that sgRNA has a good correlation with viral culture.15 In summary, prolonged viral replication is frequent in patients with hematologic malignancies, especially in those with recent chemotherapy, lymphoma and hypogammaglobulinemia. This represents a major threat for both, patients and the public health. A personalized diagnostic approach and specific therapeutic strategies are necessary to address this concerning clinical scenario.

Contributions CG-V, PP-A, M-AM and AS developed the concept; CG-V, PP-A, AS, and M-AM developed the methodology; GC-C, MS-B, M-AM performed the software analysis.; CC-V, PP-A and AS performed the formal analysis; PP-A, MC, NG-P, CG-V, EG carried out the investigation; all authors provided resources and took care of data; CG-V, PP-A and AS wrote the original draft; all authors reviewed and edited the manuscript; CG-V, M-AM, AS supervised the research; PP-A and CG-V were responsible for project adminstration. Acknowledgements Hospital Clinic of Barcelona COVID-19 Research Group: Infectious Disease Research Group: Daiana Agüero, Sabina Herrera,

Authors

Rodrigo Alonso, Laia Serra, Ana Camón, Juan Ambrosioni, Celia Cardozo, Marta Hernandez, Veronica Rico, Laura Morata, Ignacio 1*

Carolina Garcia-Vidal, Pedro Puerta-Alcalde, Aina Mateu,

Grafia, Jose Luis Blanco, Josep Mallolas, Alexy Inciarte, Esteban

Genoveva Cuesta-Chasco,3 Fernanda Meira,1 Carlos Lopera,1 Patricia 1

Martínez, María Martínez, Jose María Miró, Montse Solà, Ainhoa

Monzó, Marta Santos-Bravo, Gerard Dueñas, Mariana Chumbita, 1

Ugarte, Lorena de la Mora and all the staff members.

Nicole Garcia-Pouton, Anna Gaya, Marta Bodro, Sabina Herrera, 3

Medical Intensive Care Unit: Fernando Fuentes, Adrian Téllez, Sara

Mar Mosquera, Francesc Fernández-Avilés, José Antonio Martínez, 1

Fernández, and all the staff members.

Josep Mensa, Eva Giné, Maria Ángeles Marcos, Alex Soriano, and

Department of International Health: Daniel Camprubi, Maria Teresa

COVID-19-researcher group.

de Alba, Marc Fernandez, Elisabet Ferrer, Berta Grau, Helena Marti, Maria Jesus Pinazo, Natalia Rodríguez, Montserrat Roldan, Isabel

Department of Infectious Diseases, Hospital Clinic of Barcelona-

Vera, Nana Williams, Alex Almuedo-Riera, Jose Muñoz, and all the

IDIBAPS, University of Barcelona, Barcelona; 2Department of

staff members.

Internal Medicine, Hospital Universitari Mútua Terrassa, Terrassa;

Department of Internal Medicine: Miquel Camafort, Julia Calvo, Aina

Capdevila, Francesc Cardellach, Emmanuel Coloma, Ramon

Department of Microbiology, Hospital Clinic of Barcelona, 4

Barcelona and Department of Hematology, Hospital Clinic, IDIBAPS,

Estruch, Joaquim Fernández-Solá, Alfons López-Soto, Ferran

University of Barcelona, Barcelona, Spain

Masanés, Jose Milisenda, Pedro Moreno, Jose Naval, David Nicolás, Omar Oberoi, Sergio Prieto-González, Olga Rodríguez-Núnez, Emili Secanella, Cristina Sierra and all the staff members.

*CG-V, PP-A, AM and AS contributed equally to this manuscript.

Department of Microbiology: Manel Almela, Míriam Alvarez, Jordi Correspondence:

Bosch, Josep Costa, Julià Gonzàlez, Francesc Marco, Sofia Narvaez,

CAROLINA GARCIA-VIDAL

Cristina Pitart, Elisa Rubio, Andrea Vergara, Mª Eugenia Valls, Jordi

carolgv75@hotmail.com - cgarciav@clinic.cat

Vila and all the staff members.

PEDRO PUERTA-ALCALDE

Department of Pharmacy: Esther López, Montse Tuset and all the

pedro.puerta84@gmail.com - puerta@clinic.cat

staff members.

https://doi.org/10.3324/haematol.2021.280407 Received: November 23, 2021.

Funding

Accepted: March 10, 2022.

This work has been financed by funds for research ad hoc COVID-

Prepublished: March 17, 2022.

19 from citizens and organizations patronage to Hospital Clínic de Barcelona-Fundació Clínic per a la Recerca Biomèdica. This

Disclosures

research forms part of an activity that has received funding from

CG-V has received honoraria for talks on behalf of Gilead Science,

EIT Health. EIT Health is supported by the European Institute of

MSD, Novartis, Pfizer, Janssen, Lilly as well as a grant from Gilead

Innovation and Technology (EIT), a body of the European Union

Science and MSD. AS has received honoraria for talks on behalf of

that receives support from the European Union´s Horizon 2020

Merck Sharp and Dohme, Pfizer, Novartis, Angellini, as well as grant

Research and Innovation Program. PP-A [JR20/00012 and

support from Pfizer. PC has received honoraria for talks on behalf of

PI21/00498], NG-P [FI19/00133], and CG-V [PI21/01640] have

Merck Sharp and Dohme, Pfizer, Gilead and Alexion. JM has received

received research grants funded by Instituto de Salud Carlos III

honoraria for talks on behalf of Merck Sharp and Dohme, Pfizer,

(ISCIII) and co-funded by the European Union. The funders had

Novartis and Angellini. PP-A has received honoraria for talks on behalf

neither a specific role in study design or collection of data, nor in

of Merck Sharp and Dohme, Lilly, ViiV Healthcare and Gilead Science.

writing of the paper or decision to submit.

Haematologica | 107 July 2022

1734

LETTER TO THE EDITOR

Other acknowledgements We would like to thank Anthony Armenta for providing medical editing assistance for the manuscript at hand.

Data-sharing statement: Database used for this study will be shared by the corresponding authors under reasonable request.

References 1. Kuderer NM, Choueiri TK, Shah DP, et al. Clinical impact of COVID-19 on patients with cancer (CCC19): a cohort study. Lancet. 2020;395(10241):1907-1918. 2. Yang K, Sheng Y, Huang C, et al. Clinical characteristics, outcomes, and risk factors for mortality in patients with cancer and COVID-19 in Hubei, China: a multicentre, retrospective, cohort study. Lancet Oncol. 2020;21(7):904-913. 3. Mehta V, Goel S, Kabarriti R, et al. Case fatality rate of cancer patients with COVID-19 in a New York hospital system. Cancer Discov. 2020;10(7):935-941. 4. Fung M, Babik JM. COVID-19 in Immunocompromised hosts: what we know so far. Clin Infect Dis. 2021;72(2):340-350. 5. Avanzato VA, Matson MJ, Seifert SN, et al. Case study: prolonged infectious SARS-CoV-2 shedding from an asymptomatic immunocompromised individual with cancer. Cell. 2020;183(7):1901-1912. 6. Choi B, Choudhary MC, Regan J, et al. Persistence and evolution of SARS-CoV-2 in an immunocompromised host. N Engl J Med. 2020;383(23):2291-2293. 7. Lancman G, Mascarenhas J, Bar-Natan M. Severe COVID-19 virus reactivation following treatment for B cell acute lymphoblastic leukemia. J Hematol Oncol. 2020;13(1):1-3. 8. Lodi L, Moriondo M, Pucci A, et al. Chronic asymptomatic SARSCoV-2 infection in the immunocompromised patient: new challenges and urgent needs. Clin Infect Dis. 2022;74(3):553. 9. Hensley MK, Bain WG, Jacobs J, et al. Intractable coronavirus

disease 2019 (COVID-19) and prolonged severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) replication in a chimeric antigen receptor-modified T-cell therapy recipient: a case study. Clin Infect Dis. 2021;73(3):e815-e821. 10. D’Abramo A, Vita S, Maffongelli G, et al. Prolonged and severe SARS-CoV-2 infection in patients under B-cell-depleting drug successfully treated: a tailored approach. Int J Infect Dis. 2021;107:247-250. 11. Kemp SA, Collier DA, Datir RP, et al. SARS-CoV-2 evolution during treatment of chronic infection. Nature. 2021;592(7853):277-282. 12. Guetl K, Moazedi-Fuerst F, Rosskopf K, et al. SARS-CoV-2 positive virus culture 7 weeks after onset of COVID-19 in an immunocompromised patient suffering from X chromosomelinked agammaglobulinemia. J Infect. 2021;82(3):414-451. 13. Bravo MS, Berengua C, Marín P, et al. Viral culture confirmed SARS-CoV-2 subgenomic RNA value as a good surrogate marker of infectivity. J Clin Microbiol. 2022;60(1):e0160921. 14. Kim M-C, Cui C, Shin K-R, et al. Duration of culturable SARSCoV-2 in hospitalized patients with Covid-19. N Engl J Med. 2021;384(7):671-673. 15. Robbiani DF, Gaebler C, Muecksch F, et al. Convergent antibody responses to SARS-CoV-2 in convalescent individuals. Nature. 2020;584(7821):437-442. 16. Ju B, Zhang Q, Ge J, et al. Human neutralizing antibodies elicited by SARS-CoV-2 infection. Nature. 2020;584(7819):115-119.

Haematologica | 107 July 2022

1735