Towards targeting of adaptive immune responses in rheumatoid arthritis Anne Musters
Towards targeting of adaptive immune responses in rheumatoid arthritis Anne Musters
Colofon Author: Anne Musters Cover design: Elisa Calamita, www.elisacalamita.com Provided by thesis specialist Ridderprint, ridderprint.nl Printing: Ridderprint Layout and design: Michèle Duquesnoy, persoonlijkproefschrift.nl ISBN: 978-94-6483-260-0 Copyright 2023 © Anne Musters The Netherlands. All rights reserved. No parts of this thesis may be reproduced, stored in a retrieval system or transmitted in any form or by any means without permission of the author. This thesis is printed on 100% recycled paper.
Voor mijn lieve ouders
Contents Chapter 1 General introduction and outline 9 Part I Adaptive immune responses at sites of inflammation 21 Chapter 2 In rheumatoid arthritis, synovitis at different inflammatory sites is dominated by shared but patient-specific T cell clones 23 Chapter 3 In rheumatoid arthritis inflamed joints share dominant patient-specific B-cell clones 41 Part II Treatment options in early stages of rheumatoid arthritis 57 Chapter 4 Prevention of rheumatoid arthritis: A systematic literature review of preventive strategies in at-risk individuals 59 Part III Dynamics in the of B-cell repertoire after B-cell depletion in different phases of RA 93 Chapter 5 Sensitive B-cell receptor repertoire analysis shows repopulation correlates with clinical response to rituximab in rheumatoid arthritis 95 Chapter 6 Dynamics of the B cell receptor repertoire during the preclinical phase of rheumatoid arthritis: Longitudinal studies in untreated RA-risk individuals (DOMINO study) and the effects of rituximab treatment (PRAIRI study) 117 Chapter 7 General discussion and future perspectives 141 Appendices English summary 162 Nederlandse samenvatting 166 PhD portfolio 170 List of publications 174 Contributing authors 178 About the author 183 Dankwoord (acknowledgements) 184
General introduction and outline 1
10 CHAPTER 1 Rheumatoid arthritis (RA) is a chronic autoimmune disease that affects 0.5-1% of the population worldwide [1,2]. The disease can occur at any age, but on average patients are 40-50 years of age when diagnosed. RA is characterized by symmetrical peripheral polyarthritis, commonly, but not solely, affecting the hands, feet, knees, or ankles. Signs and symptoms of arthritis are usually accompanied by systemic inflammation and other clinical manifestations. Furthermore, RA is associated with an increased risk of cardiovascular disease (i.e. atherosclerosis and vasculitis) and interstitial lung disease [3,4]. If left untreated the inflammatory process eventually results in joint destruction, in some cases resulting in severe disability [5,6]. RA has a big impact on both the affected individual and society. While some patients can reach remission, not all patients respond to treatment. In addition to the disease itself, this can result in loss of work productivity and impairment of social activities. Even those who do respond are subjected to life-long treatment with costly therapies, which poses a substantial (financial) burden on patients and society. Autoantibodies RA is a syndrome rather than one uniform disease and two main subtypes of RA can be distinguished based on the presence or absence of autoantibodies against specific antigens. This is referred to as seropositive and seronegative RA, respectively. The most notable autoantibodies in RA are IgM-rheumatoid factor (RF), which is directed against the Fc tail of IgG, and anti-modified protein antibodies (AMPAs), autoantibodies directed against various post-translationally modified proteins. A variety of AMPAs has been identified, of which anti-citrullinated protein antibodies (ACPAs), are the most prevalent. Between seropositive and seronegative RA some remarkable differences can be observed. For instance, in ACPA-positive patients disease onset occurs at a younger age and in these patients a more severe clinical disease can be observed than in ACPA-negative patients [7]. In line with this, both RF-positive and ACPA-positive patients also have lower remission rates after initiation of disease-modifying anti-rheumatic drug (DMARD) treatment and more joint damage, although clinical manifestations at diagnosis are often indistinguishable between seropositive and seronegative patients [8,9]. Stages In the evolution from healthy to full-blown RA, different disease phases can be discriminated (see also Figure 1): - an initial at-risk period (with genetic and/or environmental risk factors)
11 General introduction and outline - a phase of clinically silent autoimmunity (e.g. autoantibodies present) - clinically suspect arthralgia (CSA); i.e. inflammatory joint pain without arthritis - undifferentiated arthritis (UA); clinically overt arthritis but RA diagnosis not established yet - RA (early and established) [10,11] Nonetheless, some phases might not be apparent in all patients [12]. Figure 1 | overview of stages in RA development As one can imagine, the identification of seronegative at-risk individuals in the pre-clinical stage, thus before the onset of arthritis, is quite challenging. Therefore, most studies investigating this phase focus on seropositive at-risk individuals. In seropositive individuals, this pre-clinical at-risk phase is characterized by immune system activation, including autoantibody production and non-specific musculoskeletal symptoms, mostly arthralgia [13]. Individuals may also experience fatigue, pain, and transient swelling of the joints [14]. Interestingly, signs of systemic inflammation, such as circulating antibodies and high levels of C-reactive protein (CRP), can be found years before the onset of RA and signify an elevated risk of developing RA in the near future [15,16]. Before developing overt clinical arthritis both ACPA-positive and ACPA-negative patients may have a symptomatic stage, which is characterized by the presence of arthralgia and/or subclinical inflammation. In this CSA phase, ACPA-positive and ACPA-negative patients have somewhat different clinical manifestations, with fewer tender joints but more rapid progression to RA in ACPA-positive patients [17]. Diagnosing and classification of rheumatoid arthritis In daily practice, the diagnosis of RA is made by the judgement of a rheumatologist, rather than by specific diagnostic criteria. This is usually based on the recognition of the clinical symptoms, radiological features, and autoantibody status. For research purposes, the 2010 American College of Rheumatology/European League Against Rheumatism Classification Criteria for RA have been developed (Figure 2) [18]. 1
12 CHAPTER 1 Figure 2 | 2010 American College of Rheumatology/European League Against Rheumatism Classification Criteria for RA [18] RF: IgM-rheumatoid factor, ACPA: antibodies against citrullinated proteins, CRP: C-reactive protein, ESR: erythrocyte sedimentation rate. Adaptive immune response Genetic and immunological studies show that cells of the adaptive immune response are involved in the pathogenesis of RA [19–26]. The specificity of this immune response is encoded by rearranged T- and B-cell receptors (TCR and BCR, respectively) expressed by clones of T- and B-lymphocytes, plasmablasts, and plasma cells. In the time preceding disease manifestation (pre-RA) and in the early stages of clinically overt disease, T-cells shift towards a pro-inflammatory phenotype, with an accumulation of expanded T-cells in the synovium and increased levels of serum pro-inflammatory cytokines, including IL-2 [26–29]. B-lineage cells are also altered in pre-RA individuals, with high levels of IgA plasmablasts in the peripheral blood [30]. Moreover, besides ACPA, RA patients often display other AMPAs, for instance against acetylated and carbamylated proteins, closer to disease onset [31,32]. These findings suggest that the immune system is already derailed years before RA onset, which would potentially allow for targeted interventions to prevent or at least delay disease onset. To investigate the adaptive immune response thoroughly and on a genomic level next-generation sequencing (NGS) technology was developed. This technique makes it able to analyze the repertoire of T- and B-cell receptors individually on the RNA level, in any given bodily compartment, at any given time. Since clones of activated T- and B-cells present identical TCRs and BCRs at their surface, expanded clones can be identified as a deviation in the repertoire, also known as dominant clones or
13 General introduction and outline highly expanded clones. Simply said; the larger the clone, the higher the frequency of the cells that have the same TCR/BCR. Arbitrarily, we marked a clone larger than 0.5% of the total TCR/BCR repertoire as expanded (Figure 3). Figure 3 | Next-generation sequencing technology to identify expanded TCR- or BCR-clones During the first step TCR/BCR mRNA is isolated from a patient sample (i.e. blood, synovial fluid or tissue). From there the next-generation sequencing (NGS) pipeline is started with cDNA synthesis, linear amplification, and sequencing. During bioinformatic analysis unique clones are identified, characterized by fingerprints of their TCR/BCR. In the last step of the bioinformatic process counting of the unique fingerprints yields an estimate of the frequency of each clone. Expanded clones can be identified as a deviation in the repertoire, also known as dominant clones or highly expanded clones (i.e. a clone larger than 0.5% of the total repertoire). Window of opportunity In recent years, substantial advances in the treatment of RA have been seen, including the introduction of several new classes of drugs, including biologic Disease-Modifying Antirheumatic Drugs (DMARDs), such as anti-TNF, and targeted synthetic DMARDs (i.e. JAK/STAT-inhibitors). Targeting the adaptive immune response using these targeted therapies, for instance with abatacept (CTLA4-Ig) or rituximab, have been proven to be of clinical benefit in RA patients [24,33]. However, these new therapies are as yet not curative, only effective in 60% of RA patients, and often only induce a partial clinical response [34]. At the moment, no formal treatment recommendations exist for individuals in the pre-clinical stage of RA after presenting with arthralgia; patients are usually monitored over time, but as a rule, they do not receive any (DMARD) treatment until clinical signs of arthritis or a formal diagnosis of RA is established. 1
14 CHAPTER 1 However, it has been shown that early treatment initiation in RA patients improves disease signs and symptoms, with lower disease scores and improved physical functioning as well as reduced structural damage detected by radiography [35]. For instance, in the “Prevention of RA by Rituximab” study (PRAIRI study) rituximab significantly delayed RA development, but was unable to prevent the risk of RA development [36]. Early treatment with abatacept showed improvement of subclinical arthritis and reduced RA development [37]. Also, early intervention with methotrexate was only partly successful as it did not prevent RA, but was able to decrease joint inflammation and reduce disease-related symptoms [38]. Many patients also experience a rapid decrease in their workability before treatment initiation and although treatment was unable to reverse disease-associated disability, it stabilized the need for sick leave and prevented incapacity for work [39]. Treatment in the preclinical phase of the disease could lead to fewer complaints in the arthralgia phase, prevention of joint damage, and improved ability to work. Aim and outline of the thesis The overarching aim of this thesis is to gain more knowledge on the adaptive immune response in different phases of RA and in various locations; ranging from the early at-risk phase to clinically apparent RA, from studies in blood-only to other bodily compartments, and from T-cells to B-cells. This thesis consists of three parts: Part I of this thesis describes the behaviour of the adaptive immune responses at various sites of inflammation during RA. We aimed to find shared characteristics of the inflammatory process to could give us insight whether development of selective targeting would be an alternative to generalized immunosuppressive strategies. In Chapter 2, we use the earlier-mentioned NGS technology to quantitatively assess whether different T-cell clones dominate the inflammatory infiltrate at various sites of inflammation in RA, i.e. in blood, synovial tissue, and synovial fluid. In addition, different joints and different locations within one joint are compared. Furthermore, we analyze to what extent these different compartments share the same dominant T-cell clones. Since T- and B-cells closely interact in adaptive responses, in Chapter 3 we analyze to what extent different joints also share dominant B-cell clones. This research builds upon the research performed in Chapter 2, by investigating the same compartments but now for the distribution of B-cells. Part II of this thesis focuses on the earliest stages of RA (i.e. the at-risk phase, phase of clinically silent autoimmunity, CSA, and UA), as early intervention in at-risk
15 General introduction and outline individuals has the theoretical potential to delay or even prevent disease onset. In Chapter 4, we perform a systematic literature review in order to provide an overview of all preventive strategies applied to at-risk individuals, taking into account all studies that have hitherto been performed and ongoing clinical trials, as well as patient perspectives to understand the feasibility of these types of interventions. In Part III of this thesis, the behaviour of B-cells before and after B-cell depletion is investigated. To gain more insight in the pathophysiology of B-cells in both RA and at-risk individuals. In Chapter 5, we investigate the depletion and repopulation of B-cells after B-cell depletion with rituximab treatment in RA patients. Although B-cell depleting therapy in RA is clearly effective, response is variable and does not always correlate with B-cell depletion itself. Time points of achieved depletion and repopulation are defined based on the percentage of unmutated BCR-clones in the repertoire. Furthermore, the predictive value of early depletion and early repopulation on clinical response is assessed to gain more insight into how this correlates with clinical response. In Chapter 6, we focus on B-cells in individuals at-risk of developing RA. We investigate changes in the BCR repertoire over time and the effects of a single dose of rituximab in at-risk individuals. For this, we use data from the earlier mentioned randomized controlled trial, the PRAIRI study, in combination with data from a longitudinal cohort, the “DOMINant clones in the Onset of RA” (DOMINO) study. In addition, a phenotypic analysis of B lineage cells is performed in a similar cohort of at-risk-individuals. Finally, in Chapter 7 a summary of the studies presented in this thesis is provided and discussed in light of current literature, with an outlook to future perspectives. 1
16 CHAPTER 1 References 1 Doran MF, Pond GR, Crowson CS, O’Fallon WM, Gabrie SE. 2002; Trends in incidence and mortality in rheumatoid arthritis in Rochester, Minnesota, over a forty-year period. Arthritis Rheum. 46(3):625–31. 2 Rossini M, Rossi E, Bernardi D, Viapiana O, Gatti D, Idolazzi L, Caimmi C, Derosa M, Adami S. 2014; Prevalence and incidence of rheumatoid arthritis in Italy. Rheumatol Int. 34(5):659–64. 3 Semb AG, Ikdahl E, Wibetoe G, Crowson C, Rollefstad S. 2020; Atherosclerotic cardiovascular disease prevention in rheumatoid arthritis. Nat Rev Rheumatol. 16:361–79. 4 Pappas DA, Nyberg F, Kremer JM, Lampl K, Reed GW, Horne L, Ho M, Onofrei A, Malaviya AN, Rillo OL, Radominski SC, Gal J, Gibofsky A, Popkova TV, Laurindo L, Kerzberg EM, Zahora R, Pons-Estel BA, Curtis JR, et al. 2018; Prevalence of cardiovascular disease and major risk factors in patients with rheumatoid arthritis: a multinational cross-sectional study. Clin Rheumatol. 37:2331–40. 5 van de Sande MG, de Hair MJ, van der Leij C, Klarenbeek PL, Bos WH, Smith MD, Maas M, de Vries N, van Schaardenburg D, Dijkmans BA, Gerlag DM, Tak PP. 2011; Different stages of rheumatoid arthritis: features of the synovium in the preclinical phase. Ann Rheum Dis. 70(5):772–7. 6 Scherer HU, Haupl T, Burmester GR. 2020; The etiology of rheumatoid arthritis. J Autoimmun. 110:102400. 7 Boeters DM, Mangnus L, Ajeganova S, Lindqvist E, Svensson B, Toes REM, Trouw LA, Huizinga TWJ, Berenbaum F, Morel J, Rantapaa-Dahlqvist S, van der Helm-van Mil AHM. 2017; The prevalence of ACPA is lower in rheumatoid arthritis patients with an older age of onset but the composition of the ACPA response appears identical. Arthritis Res Ther. 19(1):115. 8 van der Helm-van Mil AH, Verpoort KN, Breedveld FC, Toes RE, Huizinga TW. 2005; Antibodies to citrullinated proteins and differences in clinical progression of rheumatoid arthritis. Arthritis Res Ther. 7(5):R949. 9 Wu C-Y, Yang H-Y, Luo S-F, Lai J-H. 2021; From Rheumatoid Factor to Anti-Citrullinated Protein Antibodies and Anti-Carbamylated Protein Antibodies for Diagnosis and Prognosis Prediction in Patients with Rheumatoid Arthritis. Int J Mol Sci. 22(2):686. 10 Plenge RM, Padyukov L, Remmers EF, Purcell S, Lee AT, Karlson EW, Wolfe F, Kastner DL, Alfredsson L, Altshuler D, Gregersen PK, Klareskog L, Rioux JD. 2005; Replication of Putative Candidate-Gene Associations with Rheumatoid Arthritis in >4,000 Samples from North America and Sweden: Association of Susceptibility with PTPN22, CTLA4, and PADI4. Am J Hum Genet. 77(6):1044–60. 11 Lundstrom E, Kallberg H, Alfredsson L, Klareskog L, Padyukov L. 2009; Gene-environment interaction between the DRB1 shared epitope and smoking in the risk of anti-citrullinated protein antibody-positive rheumatoid arthritis: all alleles are important. Arthritis Rheum. 60(6):1597–603. 12 Woude D, Mil A. 2018; Update on the epidemiology, risk factors, and disease outcomes of rheumatoid arthritis. Best Pract Res Clin Rheumatol. 32:174–87. 13 Rakieh C, L Nam J, Hunt L, Hensor EMA, Das S, Bissell LA, Villeneuve E, McGonagle D, Hodgson R, Grainger A, Wakefield RJ, Conaghan PG, Emery P. 2015; Predicting the development of clinical arthritis in anti-CCP positive individuals with non-specific musculoskeletal symptoms: A prospective observational cohort study. Ann Rheum Dis. 74:1659–66.
17 General introduction and outline 14 Bemis EA, Demoruelle MK, Seifert JA, Polinski KJ, Weisman MH, Buckner JH, Gregersen PK, Mikuls TR, ODell JR, Keating RM, Deane KD, Holers VM, Norris JM. 2021; Factors associated with progression to inflammatory arthritis in first-degree relatives of individuals with RA following autoantibody positive screening in a non-clinical setting. Ann Rheum Dis. 80(2):154–61. 15 Nielen MM, van Schaardenburg D, Reesink HW, Twisk JW, van de Stadt RJ, van der Horst-Bruinsma IE, de Gast T, Habibuw MR, Vandenbroucke JP, Dijkmans BA. 2004; Increased levels of C-reactive protein in serum from blood donors before the onset of rheumatoid arthritis. Arthritis Rheum. 50(8):2423–7. 16 Rantapää-Dahlqvist S, de Jong B a W, Berglin E, Hallmans G, Wadell G, Stenlund H, Sundin U, van Venrooij WJ. 2003; Antibodies against cyclic citrullinated peptide and IgA rheumatoid factor predict the development of rheumatoid arthritis. Arthritis Rheum. 48(10):2741–9. 17 Burgers L, van Steenbergen HW, ten Brinck RM, Huizinga T, van der Helm-van Mil AHM. 2017; Differences in the symptomatic phase preceding ACPA-positive and ACPA-negative RA: a longitudinal study in arthralgia during progression to clinical arthritis. Ann Rheum Dis. 76:1751–4. 18 Aletaha D, Neogi T, Silman AJ, Funovits J, Felson DT, Bingham III CO, Birnbaum NS, Burmester GR, Bykerk VP, Cohen MD, Combe B, Costenbader KH, Dougados M, Emery P, Ferraccioli G, Hazes JMW, Hobbs K, Huizinga TWJ, Kavanaugh A, et al. 2010; 2010 Rheumatoid arthritis classification criteria: An American College of Rheumatology/European League Against Rheumatism collaborative initiative. Arthritis Rheum. 62(9):2569–81. 19 Okada Y, Wu D, Trynka G, Raj T, Terao C, Ikari K, Kochi Y, Ohmura K, Suzuki A, Yoshida S, Graham RR, Manoharan A, Ortmann W, Bhangale T, Denny JC, Carroll RJ, Eyler AE, Greenberg JD, Kremer JM, et al. 2014; Genetics of rheumatoid arthritis contributes to biology and drug discovery. Nature. 506(7488):376–81. 20 Kruglyak L. 2008; The road to genome-wide association studies. Nat Rev Genet. 9(4):314–8. 21 Hindorff LA, Sethupathy P, Junkins HA, Ramos EM, Mehta JP, Collins FS, Manolio TA. 2009; Potential etiologic and functional implications of genome-wide association loci for human diseases and traits. Proc Natl Acad Sci U S A. 106(23):9362–7. 22 Hu X, Kim H, Stahl E, Plenge R, Daly M, Raychaudhuri S. 2011; Integrating autoimmune risk loci with gene-expression data identifies specific pathogenic immune cell subsets. Am J Hum Genet. 89(4):496–506. 23 Yarwood A, Han B, Raychaudhuri S, Bowes J, Lunt M, Pappas DA, Kremer J, Greenberg JD, Plenge R, Rheumatoid Arthritis Consortium International (RACI), Worthington J, Barton A, Eyre S. 2015; A weighted genetic risk score using all known susceptibility variants to estimate rheumatoid arthritis risk. Ann Rheum Dis. 74(1):170–6. 24 Kremer JM, Westhovens R, Leon M, Giorgio E Di, Alten R, Steinfeld S, Russell A, Dougados M, Emery P, Nuamah IF, Williams GR, Becker J-C, Hagerty DT, Moreland LW. 2003; Treatment of rheumatoid arthritis by elective inhibition of T-cell activation with fusion protein CTLA4Ig. N Engl J Med. 349(20):1907–15. 25 de Hair M, van de Sande MVD, Ramwadhdoebe T, Hansson M, Landewé R, van der Leij C, Maas M, Serre G, van Schaardenburg D, Klareskog L, Gerlag D, van Baarsen LGM, Tak P. 2014; Features of the synovium of individuals at risk of developing rheumatoid arthritis. Arthritis Rheumatol (Hoboken, N.j). 66:513–22. 26 Klarenbeek PL, de Hair MJH, Doorenspleet ME, van Schaik BDC, Esveldt REE, van de Sande MGH, Cantaert T, Gerlag DM, Baeten D, van Kampen a HC, Baas F, Tak PP, de Vries N. 2012; Inflamed target tissue provides a specific niche for highly expanded T-cell clones in early human autoimmune disease. Ann Rheum Dis. 71(6):1088–93. 1
18 CHAPTER 1 27 Kokkonen H, Soderstrom I, Rocklov J, Hallmans G, Lejon K, Rantapaa Dahlqvist S. 2010; Up-regulation of cytokines and chemokines predates the onset of rheumatoid arthritis. Arthritis Rheum. 62(2):383–91. 28 Hunt L, Hensor EM, Nam J, Burska AN, Parmar R, Emery P, Ponchel F. 2016; T cell subsets: an immunological biomarker to predict progression to clinical arthritis in ACPA-positive individuals. Ann Rheum Dis. 75(10):1884–9. 29 Thomas R, McIlraith M, Davis LS, Lipsky PE. 1992; Rheumatoid synovium is enriched in CD45RBdim mature memory T cells that are potent helpers for B cell differentiation. Arthritis Rheum. 35(12):1455–65. 30 Kinslow JD, Blum LK, Deane KD, Demoruelle MK, Okamoto Y, Parish MC, Kongpachith S, Lahey LJ, Norris JM, Robinson WH, Holers VM. 2016; Elevated IgA Plasmablast Levels in Subjects at Risk of Developing Rheumatoid Arthritis. Arthritis Rheumatol. 68(10):2372–83. 31 Verheul MK, Bohringer S, van Delft MAM, Jones JD, Rigby WFC, Gan RW, Holers VM, Edison JD, Deane KD, Janssen KMJ, Westra J, Brink M, Rantapaa-Dahlqvist S, Huizinga TWJ, van der Helm-van Mil AHM, van der Woude D, Toes REM, Trouw LA. 2018; Triple Positivity for Anti-Citrullinated Protein Autoantibodies, Rheumatoid Factor, and Anti-Carbamylated Protein Antibodies Conferring High Specificity for Rheumatoid Arthritis: Implications for Very Early Identification of At-Risk Individuals. Arthritis Rheumatol. 70(11):1721–31. 32 Volkov M, Kampstra ASB, van Schie KA, Kawakami A, Tamai M, Kawashiri S, Maeda T, Huizinga TWJ, Toes REM, van der Woude D. 2021; Evolution of anti-modified protein antibody responses can be driven by consecutive exposure to different post-translational modifications. Arthritis Res Ther. 23(1):298. 33 Edwards JC, Szczepanski L, Szechinski J, Filipowicz-Sosnowska A, Emery P, Close DR, Stevens RM, Shaw T. 2004; Efficacy of B-cell-targeted therapy with rituximab in patients with rheumatoid arthritis. N Engl J Med. 350(25):2572–81. 34 Bykerk V. 2009; Unmet needs in rheumatoid arthritis. J Rheumatol Suppl. 36(Suppl 82):42–6. 35 Huizinga TW, Landewé RB. 2005; Early aggressive therapy in rheumatoid arthritis: a “window of opportunity”? Nat Clin Pr Rheumatol. 1(1):2–3. 36 Gerlag DM, Safy M, Maijer KI, Tang MW, Tas SW, Starmans-Kool MJF, van Tubergen A, Janssen M, de Hair M, Hansson M, de Vries N, Zwinderman AH, Tak PP. 2019; Effects of B-cell directed therapy on the preclinical stage of rheumatoid arthritis: the PRAIRI study. Ann Rheum Dis. 78(2):179–85. 37 Rech J, Ostergaard M, Tascilar K, Hagen M, Mendez L V, Kleyer A, Kroenke G, Simon D, Schoenau V, Hueber A, Kleinert S, Baraliakos X, Fleck M, Rubbert-Roth A, Behrens F, Kofler D, Feuchtenberger M, Zaenker M, Voll R, et al. 2021; Abatacept Reverses Subclinical Arthritis in Patients with High-risk to Develop Rheumatoid Arthritis -results from the Randomized, Placebo-controlled ARIAA Study in RA-at Risk Patients. Arthritis Rheumatol. 73:939–41. 38 Krijbolder DI, Verstappen M, van Dijk BT, Dakkak YJ, Burgers LE, Boer AC, Park YJ, de Witt-Luth ME, Visser K, Kok MR, Molenaar ETH, de Jong PHP, Bohringer S, Huizinga TWJ, Allaart CF, Niemantsverdriet E, van der Helm-van Mil AHM. 2022; Intervention with methotrexate in patients with arthralgia at risk of rheumatoid arthritis to reduce the development of persistent arthritis and its disease burden (TREAT EARLIER): a randomised, double-blind, placebo-controlled, proof-of-concept trial. Lancet. 400(10348):283–94.
19 General introduction and outline 39 Neovius M, Simard JF, Klareskog L, Askling J, Group AS. 2011; Sick leave and disability pension before and after initiation of antirheumatic therapies in clinical practice. Ann Rheum Dis. 70(8):1407–14. 1
PART I
Adaptive immune responses at sites of inflammation
In rheumatoid arthritis, synovitis at different inflammatory sites is dominated by shared but patient-specific T cell clones Anne Musters, Paul L. Klarenbeek, Marieke E. Doorenspleet, Giulia Balzaretti, Rebecca E.E. Esveldt, Barbera D.C. van Schaik, Aldo Jongejan, Sander W. Tas, Antoine H.C. van Kampen, Frank Baas, Niek de Vries Journal of Immunology 2018 Jul 15;201(2):417-422 2
24 CHAPTER 2 Abstract Genetic and immunological evidence clearly points to a role for T cells in the pathogenesis of rheumatoid arthritis (RA). Selective targeting of such disease-associated T cell clones might be highly effective while having few side effects. However, such selective targeting may only be feasible if the same T cell clones dominate the immune response at different sites of inflammation. We leveraged high-throughput technology to quantitatively assess whether different T cell clones dominate the inflammatory infiltrate at various sites of inflammation in this prototypic autoimmune disease. In 13 RA patients, we performed quantitative next-generation sequencing–based human TCRβ repertoire analysis in simultaneously obtained samples from inflamed synovial tissue (ST) from distinct locations within one joint, from multiple joints, and from synovial fluid (SF) and peripheral blood (PB). Identical TCRβ clones dominate inflammatory responses in ST samples taken from different locations within a single joint and when sampled in different joints. Although overall ST–SF overlap was comparable to higher ST–ST values, the overlap in dominant TCRβ clones in ST–SF comparisons was much lower than ST–ST and comparable to the low ST–PB overlap. In individual RA patients, a limited number of TCRβ clones dominate the immune response in the inflamed ST regardless of the location within a joint and which joint undergoes biopsy; in contrast, there is limited overlap of ST with SF or PB TCR repertoires. This limited breadth of the T cell response in ST of the individual RA patient indicates that development of immunotherapies that selectively modulate dominant T cell responses might be feasible.
25 Synovitis is dominated by shared T-cell clones Introduction In recent years, we have seen substantial advances in the treatment of rheumatoid arthritis (RA), including the introduction of several new drugs. However, these new therapies are not curative, are effective in only 60% of the RA patients, and often only induce partial clinical response [1]. Thus, there is a clear need to identify novel, targeted, more effective therapies. Genetic and immunological studies show that cells of the adaptive immune response are involved in the pathogenesis of RA [2–9]. Specificity of this immune response is encoded by rearranged T and B cell receptors expressed by clones of T and B lymphocytes, plasmablasts, and plasma cells. Targeting the adaptive immune response via biologics, such as abatacept (CTLA4-Ig) and rituximab, has been proven to be of clinical benefit in RA patients [7,10]. Recent advances in selective immunomodulation of disease-associated T and B cell clones resulted in novel, more selective, intensive, and effective therapies in the field of oncology [11,12]. Such Ag receptor–directed therapies would also hold promise for more effective treatment in RA provided that a common signature for RA can be found. However, so far, analyses on Ag receptor characteristics in RA synovial tissue (ST) have yielded two seemingly paradoxical findings: 1) several studies observed sharing of T cell clones between different joints, whereas 2) other studies show that different joints show a huge variation in T cell clones, even within a single patient [9,13–25]. As a consequence, it is unclear whether T cell clones in RA are homogeneous and whether they can be used for targeted therapies. In this study, we aim to shed more light on this paradox, taking a high-throughput quantitative approach to TCR repertoire analysis in a unique cohort of 13 RA patients in whom synovial biopsy specimens were taken from multiple locations within the same joint, in another (contralateral) joint, as well as in synovial fluid (SF) and peripheral blood (PB) samples. In doing so, we aimed to answer the following three questions: 1) do various clones dominate the T cell response at different locations within one single inflamed joint? 2) do different T cell clones dominate the TCRβ repertoire in multiple joints? and 3) are TCRβ repertoires in ST and SF dominated by different T cell clones? 2
26 CHAPTER 2 Materials and methods Patients We included 13 RA patients meeting the 2010 American College of Rheumatology/ European League Against Rheumatism Classification Criteria for RA who had active disease (disease activity score evaluated in 28 joints >3.2) [26]. All but one patient were autoantibody positive (anticyclic citrullinated peptide test >25 kAU/l and/or IgM rheumatoid factor >12.5 kU/l). All patients were typed for HLA class II alleles (Supplemental Table 1). Two patients were treated with a biological at the time of arthroscopy (infliximab and rituximab; last infusion 1 month before sampling). We did not observe a significant difference in the number of TCRβ clones, the number of highly expanded TCRβ clones (HECs), or the impact of these HECs on the total TCR repertoire between the different types of treatment (data not shown). More details on patient characteristics are shown in Table 1. From 10 patients, ST biopsy specimens were taken from either one (n = 1) or two inflamed joints (n = 9), all of which were paired with PB. In seven of these patients, we also collected SF from the same joint on which biopsy was performed prior to the arthroscopy. Three additional patients were included for paired SF and PB analysis. The study was approved by the independent Medical Ethics Committee of the Academic Medical Center/University of Amsterdam and performed according to the Declaration of Helsinki. All patients gave written informed consent. Sampling of synovial biopsy specimens, SF, and PB To obtain ST biopsy specimens, a minimally invasive arthroscopy was performed from a clinically inflamed knee or ankle, as described previously [27]. ST biopsy specimens from 11 inflamed knee joints were taken from two locations: the infrapatellar (IP) and the suprapatellar (SP) regions. If biopsy was performed on multiple joints, this was done within the same day. SF was obtained by arthrocentesis. In case of a combined collection of ST and SF, the SF was collected prior to the arthroscopy to avoid contamination of SF by hemorrhagic fluid. In all patients, PB was drawn at the time of the arthroscopy and/or arthrocentesis.
27 Synovitis is dominated by shared T-cell clones Table 1 | Patient characteristics (n=13) Age (mean (SD)), years 52 (13) Male (n (%)) 3 (23%) IgM-RF positive (n (%)) 12 (92%) IgM RF level (median (IQR)), kU/L 68 (26-276) ACPA positive (n (%)) 7 (54%) ACPA level (median (IQR)), kAU/L 84 (8-644) IgM-RF and ACPA both pos. (n (%)) 7 (54%) Disease duration (median (IQR)), months 7 (3-159) DAS28 (mean (SD)) 5.2 (1.1) Patients without therapy (n (%)) 7 (54%) Patients treated with csDMARD (n (%)) 4 (31%) Patients treated with bDMARD (n(%)) 2 (15%) IgM-RF, IgM rheumatoid factor; ACPA, anti-citrullinated peptide antibodies; csDMARD, conventional synthetic disease-modifying antirheumatic drugs; bDMARD, biological disease-modifying antirheumatic drugs. Linear amplification and next-generation sequencing The protocol used for linear amplification and next-generation sequencing has been described before [9,28,29]. Sequencing was performed on the MiSeq Gene and Small Genome Sequencer (Illumina). Based on earlier studies, TCRβ clones with a frequency ≥0.5% were considered dominant and therefore called HECs [9,30]. Statistics Values are either expressed as mean or median depending on the presence of a normal or nonnormal distribution of the data. To test for similarity between different sites, the Chao-modified Sørensen index was used. This index, originating from the field of biodiversity, measures dispersion and gives a value between 0 and 1. Values near 0 indicate no overlap between two locations, whereas values close to 1 indicate that the two locations are identical [31–34]. Differences between groups were analyzed using the unpaired Mann–Whitney U test, paired t test, one-way ANOVA, or Tukey multiple comparison test if appropriate. The p values <0.05 (two-sided) were considered statistically significant. R (version 3.1.0), package SpadeR (version 0.1.1), and GraphPad Prism (version 6.0) were used to perform the analyses. 2
28 CHAPTER 2 Results Do various clones dominate the T cell response at different locations within one single inflamed joint? In large joints, ST biopsy specimens can be obtained from multiple anatomic locations. For example, in the knee, biopsy specimens can be taken from either the SP or IP region. It is unknown whether the T cell repertoires are different at these distinctive locations. To investigate this, we included seven RA patients (see Table 1 for characteristics) with inflamed knee joints and took biopsy specimens from both the SP and IP regions from both inflamed joints (in total, n = 14) and analyzed the TCRβ repertoires. For comparison, we used TCR repertoires from the paired PB samples. We used 26,866 quality-filtered, randomly selected TCRβ sequences per sample. The number of TCRβ clones found in the SP and IP regions (mean 6679 [SD 2612] versus 7869 [SD 2550], respectively) and the number of HECs (mean 8.0 [SD 3.8] versus 7.6 [SD 3.8]) were comparable (Fig. 1A, 1B). The HECs accounted for 29.9% (mean, SD 23.4) and 18.2% (SD 12.8) of the total repertoire, respectively (Fig. 1C). PB showed a similar number of TCRβ clones (9982, SD 5898) with a mean of 6.9 HECs (SD 6.2) that accounted for 40.7% (SD 32.4) of the total repertoire (Fig. 1A–C). Subsequently, using PB as control, we analyzed to what extent the 25 most expanded (top 25) TCRβ clones in the SP and IP regions showed overlap. Of the 25 most expanded TCRβ clones in the SP region, 54.3% (mean, SD 17.7) were also present among the 25 most expanded TCRβ clones in the IP region. This overlap was significantly lower when comparing the IP samples with PB (mean 20.0%, SD 8.0, p < 0.0001; Fig. 1F), leading us to believe that dominant TCRβ clones from the ST are hardly present in the PB. Comparable results were found if we restricted the analysis to the top 10 or top 100 TCRβ clones (Supplemental Fig. 1). Of the top 25 SP TCRβ clones, 86% could be retrieved among the top 1000 of the IP TCRβ clones (Supplemental Fig. 2), showing that the vast majority of the expanded SP TCRβ clones is also present in the IP region. If we do not focus on the most dominant clones, but instead look at the total measured TCRβ repertoire, we used the Chao-modified Sørensen index to test for similarity between the different regions. For the measured TCRβ repertoires in SP and IP regions, we thus observed a score of 0.49 (mean, SD 0.11), which is significantly higher than that for the comparison of ST and PB (mean 0.15, SD 0.07; p < 0.0001) (Fig. 1G). For these analyses, we only included the patients from whom we had a complete dataset, and therefore paired analyses were possible.
29 Synovitis is dominated by shared T-cell clones Collectively, these findings demonstrate that in the inflamed RA ST, the TCR repertoire shows substantial similarity at different regions within one joint and is dominated by the same TCRβ clones. Such overlap is not observed when comparing ST to PB. Figure 1 | Comparing T-cell receptor repertoires within one joint Bar charts of (A) the number of TCRβ-clones, (B) number of highly expanded TCRβ-clones (HECs) and (C) impact of HECs on total repertoire per compartment (bars show mean and SD; using a Tukey’s multiple comparison test). (D) Example of overlap-plot from one patient comparing the suprapatellar (SP) to the infrapatellar (IP) synovial tissue (ST) region, showing clear overlap of dominant TCRβ-clones in the upper right quadrant; (E) Comparison of ST to peripheral blood (PB), showing little overlap. Scatter plots of (F) percentage of overlapping top-25 TCRβ-clones and (G) Chao-modified Sørensen indices of the total TCRβ-clones repertoire when comparing different compartments (n=14; lines at mean and SD; **** p<0.0001 using a paired t test). Do different T cell clones dominate the TCRβ repertoire in multiple joints? Polyarthritis is a hallmark feature of RA. To the best of our knowledge, no quantitative analysis exists on whether different T cell clones dominate the TCR repertoire in multiple joints. To test this hypothesis, we compared the TCRβ repertoires from ST biopsy specimens simultaneously taken from two inflamed contralateral joints (left and right; either knee or ankle) from nine RA patients (see Table 1 for characteristics). We analyzed paired PB samples as a control. The general features of the ST TCR repertoires from contralateral joints were not significantly different: in the left joint, we identified 7124 TCRβ clones (mean, SD 2665) with 7.1 HECs (mean, SD 2.8); in the right joint, we observed 6735 TCRβ clones (SD 3452) and 9.7 HECs (SD 5.2) 2
30 CHAPTER 2 (Fig. 2A, 2B). The impact of the HECs on the total repertoire was 18.9% (mean, SD 14.5) and 18.5% (SD 10.6), respectively, which was also not significantly different (Fig. 2C). Next, we compared the overlap between the 25 most expanded TCRβ clones in both joints. In line with the findings on different locations within the same joint, we again observed that 50.2% (mean, SD 14.9) of the 25 most dominant ST TCRβ clones were identical between different joints (p = 0.30, Fig. 2E). Last, the Chao-modified Sørensen index as a measurement for similarity was assessed. In contrast to what was hypothesized, the Chao-modified Sørensen index also demonstrated large overlap, with a mean score of 0.43 (SD 0.11) (Fig. 2F). Both ST–ST overlap analyses showed significantly higher overlap when compared with overlap between ST and PB (p < 0.001, p < 0.0001, respectively, Fig. 2E, 2F). The Chao-modified Sørensen index between contralateral joints did not differ from the ST overlap observed when comparing two regions within one joint (i.e., IP versus SP [p = 0.12]). In summary, these data support the notion that TCR repertoires in ST biopsy specimens taken simultaneously from two different (contralateral) inflamed joints show substantial overlap. Figure 2 | Comparing T-cell receptor repertoires in two contralateral inflamed joints Bar charts of (A) the number of TCRβ-clones, (B) number of highly expanded TCRβ-clones (HECs) and (C) impact of HECs on total repertoire per joint (bars show mean and SD; using a one-tailed Mann-Whitney test). (D) Example of overlap-plots from one patient when comparing the ST of the left (L) joint to right (R) joint, showing substantial overlap. Scatter plot of (E) percentage of overlapping top-25 TCRβ-clones and (F) Chao-modified Sørensen indices of the total TCRβ-clones repertoire when comparing different compartments (n=9; lines at mean and SD; *** p < 0.001, **** p < 0.0001 using a paired t test).
31 Synovitis is dominated by shared T-cell clones Are TCR repertoires in ST and SF dominated by different T cell clones? Besides synovitis, joint effusion is also a feature of arthritis. As SF can be acquired with a less invasive procedure than ST, fluid is often used to study certain aspects of the pathogenesis of arthritis. However, it is not known whether the T cell repertoires from ST are similar to those from SF in the same joint. To address this, we included seven RA patients (see Table for characteristics) from whom we obtained paired samples: ST biopsy specimens and SF from the same joint as well as PB as control. In ST, we detected 6781 TCRβ clones (mean, SD 2759) compared with 4923 TCRβ clones in SF (SD 3048, Fig. 3A). We identified 8.2 HECs in ST (mean, SD 4.0) and 18.7 HECs in SF (SD 15.6, p < 0.001, Fig. 3B). The HECs accounted, respectively, for 23.0% (mean, SD 18.5) and 45.3% (SD 27.2) of the total repertoire (p < 0.05, Fig. 3C). ST and SF differed significantly from PB regarding the number of TCRβ clones (features described earlier in this article; p < 0.05, p < 0.01, respectively), whereas the number of HECs in ST and SF was significantly different from that in PB (both p < 0.01, Fig. 3A–C). Comparing ST to SF from the same joint, the top 25 overlap was 26.0% (mean, SD 14.8). The top 25 overlap between SF and PB was 18.0% (SD 8.8, Fig. 3F). Thus, the top 25 analysis did not differ significantly between the ST–SF, SF–PB, and ST–PB comparisons. Comparing all TCRβ clones using the Chao-modified Sørensen index for ST to SF resulted in an index of 0.41 (mean, SD 0.13), whereas we noted an index of 0.13 for the SF–PB comparison (SD 0.06, Fig. 3G). The Chao-modified Sørensen index of the ST–SF comparison did significantly differ from the SF– PB and ST–PB comparisons (p < 0.001, p < 0.0001, respectively). Moreover, within one joint, both results from the top 25 overlap and Chao-modified Sørensen index differed significantly when comparing the intrajoint ST–SF overlap to the intrajoint ST–ST (IP versus SP regions) overlap as described earlier (Fig. 3F, 3G). In summary, these data show that the dominant TCRβ clones from ST are not fully reflected in SF and even less so in PB. Moreover, significantly less similarity between ST and SF/PB was observed compared with the high similarity seen within ST biopsy specimens taken from different locations in the same joint. 2
32 CHAPTER 2 Figure 3 | Comparing T-cell receptor repertoires in synovial tissue, synovial fluid and peripheral blood Bar charts of (A) the number of TCRβ-clones, (B) number of highly expanded TCRβ-clones (HECs) and (C) impact of HECs on total repertoire per compartment (bars show mean and SD; * p < 0.05, ** p < 0.01, *** p < 0.001, using a Tukey’s multiple comparison test). (D) Overlap-plot comparing synovial tissue (ST) to synovial fluid (SF) in one patient; (E) Overlap plot SF to peripheral blood (PB) in one patient. Scatter plot of (F) percentage of overlapping top-25 TCRβ-clones and (G) Chao-modified Sørensen indices of the total TCRβ-clones repertoire when comparing different compartments (n=8; lines at mean and SD; * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001, using a paired t test). Discussion This quantitative, comprehensive, whole-repertoire TCR analysis shows that RA synovitis is not dominated by variable, local T cell responses but rather by uniform, systemic T cell responses. Within a single patient, synovial inflammation in multiple joints was dominated by a limited number of expanded TCRβ clones, even when these clones were not dominantly present in PB. This observation suggests that immunotherapy, selectively targeting a limited number of shared, expanded T cell clones, might be effective and feasible [35]. Therefore, to develop this kind of targeted immunotherapy, further characterization of the overlapping “clones” is indicated (e.g., regarding phenotype, TCR α- and β-chain pairing, Ag specificity, and genomic profiles). This meets substantial challenges because it requires harvesting of (enough) cells from the synovium in a phe-
33 Synovitis is dominated by shared T-cell clones notypically unchanged state, which is currently difficult to perform routinely on a large scale. Hopefully, novel technologies for single-cell characterization will rapidly develop to allow this in the near future. A striking observation is the difference in TCR repertoire when comparing SF to ST. Previous literature showed an overlap of clones between ST, SF, and PB, but the exact quantitative relation between clones in the different compartments could not be found [36]. An early pilot study already showed that the overlap of TCRβ clones between ST and PB is low (4%) [9]. The present study clearly validates this, but the apparent quantitative difference between TCRβ clones (especially when focusing on the expanded clones) of the SF and ST that we determined was not shown before. Thus, it seems that the ST and SF are two separate compartments instead of one. This is supported by the similarity index, which confirms that the SF shows significantly more overlap with the ST than the PB does. In contrast, the top 25 overlap between ST and SF is as low as the top 25 overlap between ST and PB, indicating that although the general overlap is quite high, this is not the case for the highly expanded ST TCRβ clones. Hence, for future T cell studies in RA, it would be recommended to regard ST and SF as separate compartments and use caution when extrapolating T cell findings from SF to ST. Our data show clear oligoclonal expansions in the ST and show that the same TCRβ clones dominate the repertoires in biopsy specimens from different regions in the same joint. Even with a very stringent top 25 overlap analysis, we show a large number of overlapping TCRβ clones. These findings imply that single-locus ST biopsies can be used instead of multilocus biopsies when studying T cells. One can therefore speculate that ultrasound (US)-guided or even blind needle biopsies might be just as informative on the T cell repertoire as biopsies taken via arthroscopy. This would improve accessibility because US-guided or blind needle biopsies can be performed at the outpatient clinic, whereas arthroscopy requires theater time. Moreover, these procedures are minimally invasive, well tolerated, and take less time to perform. An earlier study already showed that the quality and RNA yield is preserved in US-guided procedures [37]. Ideally, a follow-up study comparing specimens from US-guided or blind needle biopsies to biopsy specimens obtained via arthroscopy (from the same joint) should be performed to test this. It should be noted that we sequenced the TCR β-chain only; thus, it remains possible that the identified sequences correspond to “public” sequences (i.e., clones that have the same TCR β-chain but have a different TCR α-chain). This might or might not drive a similar Ag specificity of the receptor depending on the TCR α-chain it is 2
34 CHAPTER 2 paired with. Because our studies are based on next-generation sequencing analysis of TCR mRNA, our results might be influenced quantitatively by differences in expression among different types of T cells [9]. However, in vitro studies showed that TCR surface expression and TCR mRNA levels are equal for naive and memory CD4 and CD8 T cells, whereas effector T cells were shown to have at most 2-fold higher levels of TCR mRNA and TCR surface expression, indicating that potential bias in RNA-based TCR studies is probably minimal [9,38,39]. Theoretically, there is a risk of amplification bias due to the fact that some TCR primers might be more efficient than others. However, our system uses an initial linear amplification procedure to prevent this. Furthermore, recent analysis using unique molecular identifiers in our protocol showed excellent correspondence with the results of our linear amplification protocol (data not shown). The reported findings may potentially be extended to other forms of arthritis. In a pilot study in psoriatic arthritis (PsA) patients, we earlier showed substantial overlap in TCR repertoire between ST in both knees in one PsA patient, whereas overlap between synovial repertoires and PB was limited in two PsA patients [40]. These preliminary results are comparable to the results shown in this study and support the notion that TCRβ clones in ST are different compared with PB in both diseases. Clearly, the findings in PsA need confirmation in larger patient groups. Follow-up studies on B cell receptor clonality in both diseases and in other immune-mediated inflammatory diseases may shed more light on the composition of the B cell repertoire in different compartments and different phases of the disease. In RA, this would follow up on an earlier study that showed dominant B cell clones shared among multiple joints of the same patient but not in PB [30]. In conclusion, we show substantial overlap of the TCR repertoires between inflamed ST biopsy specimens from different joints or from different sites within one joint in the same patient, whereas SF does not fully reflect the T cell repertoire in the inflamed ST. This implies that for T cell (receptor) analysis, tissue biopsies are required but that tissue collection might be simplified using less invasive procedures [e.g., through high-quality, US-guided, or blind needle biopsies [27,37]. More importantly, it shows that underlying T cell responses in the ST share a uniform specificity, which suggests that Ag- and/or receptor-specific therapies targeting a limited set of TCRβ clones in individual patients may be feasible in RA and possibly other inflammatory joint diseases, including PsA or even the immune-mediated inflammatory disease group at large.
www.ridderprint.nlRkJQdWJsaXNoZXIy MTk4NDMw