Citation for published version (APA): Paramarta, J. E. (2014). Spondyloarthritis: From disease phenotypes to novel treatments

Transcription

1 UvA-DARE (Digital Academic Repository) Spondyloarthritis: From disease phenotypes to novel treatments Paramarta, Jacky Link to publication Citation for published version (APA): Paramarta, J. E. (2014). Spondyloarthritis: From disease phenotypes to novel treatments General rights It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons). Disclaimer/Complaints regulations If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Ask the Library: or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You will be contacted as soon as possible. UvA-DARE is a service provided by the library of the University of Amsterdam ( Download date: 31 Dec 2017

2 SPONDYLOARTHRITIS: FROM DISEASE PHENOTYPES TO NOVEL TREATMENTS Jacqueline Elizabeth Paramarta

3 ISBN: Printing of thesis was financially supported by AbbVie, Novartis Pharma, Pfizer B.V., UCB Pharma B.V., ChipSoft, ABN AMRO Bank N.V., AMC-UvA, and the Dutch Arthritis Association. Cover design: J.E. Paramarta Layout & printing: Off Page, Copyright 2014 by J.E. Paramarta. All rights reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, without permission of the author.

4 SPONDYLOARTHRITIS: FROM DISEASE PHENOTYPES TO NOVEL TREATMENTS ACADEMISCH PROEFSCHRIFT ter verkrijging van de graad van doctor aan de Universiteit van Amsterdam op gezag van de Rector Magnificus prof. dr. D.C. van den Boom ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Agnietenkapel op woensdag 15 oktober 2014, te 14:00 uur door Jacqueline Elizabeth Paramarta geboren te Amsterdam

5 PROMOTIECOMMISSIE Promotores Overige leden Prof. dr. D.L.P. Baeten Prof. dr. P.P. Tak Prof. dr. J.W.J. Bijlsma Prof. dr. R.B.M. Landewé Prof. dr. M.A. de Rie Prof. dr. P.M. van Hagen Dr. A.W.R. van Kuijk Faculteit der Geneeskunde

6

7 CONTENTS Chapter 1 General introduction 9 Spondyloarthritis: from unifying concepts to improved treatment Rheumatology (Oxford). 2014;53: Chapter 2 Outline of this thesis 31 Part I Disease phenotypes in spondyloarthritis Chapter 3 Peripheral joint inflammation in early onset spondyloarthritis 37 is not specifically related to enthesitis Ann Rheum Dis. 2014;73: Chapter 4 Undifferentiated spondyloarthritis vs ankylosing spondylitis 51 and psoriatic arthritis: a real-life prospective cohort study of clinical presentation and response to treatment Rheumatology (Oxford). 2013;52: Chapter 5 Prevalence and burden of peripheral disease in spondyloarthritis: 63 proposal for modification of the Assessment of SpondyloArthritis international Society (ASAS) classification criteria Submitted for publication Part II Novel treatments in spondyloarthritis Chapter 6 Efficacy and safety of adalimumab for the treatment of peripheral 81 arthritis in spondyloarthritis patients without ankylosing spondylitis or psoriatic arthritis Ann Rheum Dis. 2013;72: Chapter 7 Fast relapse upon discontinuation of tumour necrosis factor 97 blocking therapy in patients with peripheral spondyloarthritis Ann Rheum Dis. 2013;72: Chapter 8 Adalimumab serum levels and antidrug antibodies towards 103 adalimumab in peripheral spondyloarthritis: No association with clinical response to treatment or with disease relapse upon treatment discontinuation Arthritis Res Ther. 2014;16:R160 [Epub ahead of print] Chapter 9 The il-23/il-17 immune axis as a promising new target in 117 the treatment of spondyloarthritis Curr Opin Rheumatol. 2014;26:361-70

8 Chapter 10 Anti-interleukin-17a monoclonal antibody secukinumab in 133 treatment of ankylosing spondylitis: a randomised, double-blind, placebo-controlled trial Lancet. 2013;382: Chapter 11 Targeting synovial mast cells in spondyloarthritis: a proof-of-concept 159 study with the tyrosine kinase inhibitor nilotinib Submitted for publication Chapter 12 General discussion and summary 179 Appendices Nederlandse samenvatting 198 Dankwoord 202 PhD portfolio 206 Curriculum vitae 209 List of publications 210

9

10 General introduction Spondyloarthritis: from unifying concepts to improved treatment Jacqueline E. Paramarta, Dominique Baeten Department of Clinical Immunology and Rheumatology, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands Rheumatology (Oxford) Sep;53(9):

11 ABSTRACT Spondyloarthritis (SpA) is a chronic immune-mediated inflammatory disease with diverse phenotypic manifestations including spondylitis, arthritis, enthesitis and extra-articular manifestations (psoriasis, uveitis, inflammatory bowel disease). The common genetic risk factors, the strong familial aggregation and the overlapping immunopathology suggest that these different phenotypic manifestations share common pathogenic pathways. This concept is further strengthened by the good clinical response of all different SpA manifestations to TNF-blocking therapies. However, the phenotypic diversity of SpA is still a major challenge in properly diagnosing, classifying and monitoring the disease and may lead to undertreatment of less typical SpA cases such as undifferentiated SpA. The optimal use of current treatments and the development of novel therapies, including compounds targeting the IL-23/IL-17 axis, thus requires a detailed understanding of both the clinical presentation and the underlying pathogenic pathways in SpA. 10

12 INTRODUCTION SpA is one of the most prevalent forms of chronic inflammatory arthritis, affecting approximately % of the Western population. 1,2 The clinical presentation of the disease is dominated by inflammation that can affect the axial skeleton (spine and sacroiliac joints), peripheral joints (especially the large joints of the lower extremities) and extra-articular sites such as the eye (uveitis), skin (psoriasis) and gut (IBD), resulting in highly heterogeneous phenotypes of disease in different patients. 3 Moreover, this inflammation can be accompanied by different degrees and types of structural damage, ranging from extensive bone destruction to new bone formation and even complete ankylosis of axial and peripheral joints. In combination, the different types and degrees of tissue inflammation and structural damage lead to a variety of different phenotypes in SpA. Nevertheless, the strong common genetic background, as evidenced by its high heritability, suggests a common origin of the different SpA phenotypes. 4,5 The most important genetic risk factor, HLA-B27, has been known now for 40 years. 6 More recently, genome-wide association studies have revealed additional genetic links in SpA, e.g. genetic risk variants of interleukin-23 receptor (IL-23R) and endoplasmic reticulum aminopeptidase 1 (ERAP-1), which may give further clues to the pathogenesis of this disease. 7 Besides the genetic background, the different SpA manifestations also share the same key environmental triggers such as microorganisms and mechanical stress. 8.9 Therefore the key question arising here is whether the different disease manifestations of SpA are driven by the same pathogenic mechanisms and, if so, whether therapeutic targeting of these public pathways can lead to clinical benefit in the different SpA subforms. On the contrary, if the different disease manifestations are not driven by the same pathogenic pathways, one should delineate sharply which private pathway is pathophysiologically and therapeutically relevant in which subset of SpA patients. 1 General introduction WHERE DOES IT ALL START? McGonagle et al. 10 proposed a unifying pathophysiological concept for SpA based on the idea that specific tissues may be particularly sensitive to mechanical triggers. They proposed that the primary lesion of SpA is enthesitis (inflammation at the attachment sites of ligaments, tendons and joint capsules to the bone) and that this could explain all rheumatic signs of SpA, including synovitis, which would be secondary to the release of proinflammatory mediators from the enthesis. This hypothesis was based on an MRI study in SpA and RA patients with recent-onset knee effusion, which showed that entheseal abnormalities were more prominent in SpA than in RA. 11 Testing the validity of the proposed hypothesis 10 is of relevance for both pathophysiological and clinical research since, if the hypothesis is correct, pathophysiological studies should focus on experimental models of enthesitis rather than synovitis, translational research should aim to obtain and study entheseal biopsies rather than synovial biopsies and clinical trials should focus on specific treatments targeting primary enthesitis rather than compounds directed towards secondary synovitis. However, several remarks arise when critically assessing this hypothesis. 10 First, the aforementioned MRI study suggests that enthesitis is more frequently seen during arthritis in established SpA than RA, 11 but the cross-sectional design does not allow us to conclude that 11

13 enthesitis is primary to synovitis in SpA. Second, subsequent imaging studies comparing SpA with other inflammatory controls yielded conflicting results, questioning the specificity of enthesitis for SpA Third, in patients with SpA-associated diseases without clinical arthritis, not only subclinical enthesitis, but also subclinical synovitis was shown to be present in a higher percentage compared with control patients. 22,23 Fourth, the few studies examining the share of enthesitis and synovitis in sacroiliac joint biopsies and MRIs of the sacroiliac joints showed that synovitis and subchondral bone marrow changes (osteitis) were the most prominent features while enthesitis was neither the earliest nor the principal pathologic change. 24,25 Finally, animal models are not very useful in addressing this issue, as the ankylosing enthesitis model in DBA/1 mice clearly demonstrates that entheseal stress can lead to osteoproliferation and mild inflammation reminiscent of heel enthesitis in human SpA 26 while, in contrast, both peripheral arthritis and axial spondylitis in human HLA-B27/β2 microglobulin transgenic rats are characterized by pronounced synovitis in the absence of enthesitis. 27 An additional issue with the enthesitis hypothesis is that this concept, which is based on imaging data, still lacks supportive information on the underlying cellular and molecular mechanisms. The few studies that have managed to obtain entheseal tissue samples from human SpA patients have yielded conflicting results on the major infiltrating cell types. 28,29 Experimental studies suggested that TNF overexpression in the TNFΔARE model may drive enthesitis, but the same model shows extensive erosive polyarthritis phenocopying human RA rather than human SpA. 30 Similarly, a ground-breaking mechanistic study recently provided convincing evidence that IL-23 overexpression induces enthesitis (including involvement of the aortic valve, an enthesis-like structure) by acting on double-negative, enthesis-resident T lymphocytes, 31 but the same overexpression also leads to a severely destructive polyarthritis. 32 Whereas these studies provide evidence that the TNF and IL-23/IL-17 cytokine axes may be involved in enthesitis in SpA, and thus that targeting these pathways may be an effective treatment strategy for this disabling SpA feature, they question at the same time the concept that SpA inflammation would primarily or uniquely target entheses. These experimental findings are consistent with the histological studies in established as well as early peripheral SpA demonstrating that the synovial inflammation is as pronounced but pathologically distinct from RA Collectively these data challenge the concept that enthesitis is the primary lesion in SpA and rather suggest that different tissues of the skeleton (including the enthesis, the synovial membrane and the bone marrow) can be affected by TNF and IL-23/IL-17-mediated inflammation (Figure 1). TOWARDS A UNIFYING CLINICAL CLASSIFICATION Besides the discussed efforts to capture the different SpA manifestations into a unifying pathophysiological concept, increasing attempts have been made to define more clearly the clinical classification of SpA subtypes. SpA is classically subdivided into different phenotypic forms based on aetiology (reactive arthritis caused by a bacterial trigger), outcome (AS), associated extra-articular symptoms (PsA and IBD-related SpA) and a residual group called undifferentiated spondyloarthritis (USpA). 3,41 Although this classification allows us to capture the clinical presentation in daily clinical practice reliably, it also presents a number of major 12

14 1 General introduction Figure 1. Hypotheses on the development of arthritis in spondyloarthritis. The first hypothesis is the enthesitis hypothesis where the primary lesion is enthesitis, and synovitis is secondary to the release of pro-inflammatory mediators from the enthesis (A). The second hypothesis shows that different tissues of the skeleton (including the enthesis, the synovial membrane, and the bone marrow) can be affected by TNF and IL-23/IL-17 mediated inflammation already at the onset of arthritis (B). disadvantages. First, it does not reflect the increasing evidence that these phenotypes represent different presentations of a single disorder rather than a spectrum of distinct, though related, disease entities. 42 The concept of SpA as a single disorder is not only based on the overlap between different subtypes in terms of genetic background, 7,43,44 familial aggregation, immunopathology, 37,40,48 and pathophysiology, 49 but also on the clinical observation that a single patient can display several SpA phenotypes at once (e.g. a single patient can have AS plus PsA) or can evolve from one phenotype to another over time. For example, a substantial proportion of USpA patients with axial symptoms will evolve over time to full-blown AS, which relates to the fact that sacroiliitis on conventional radiographs, the hallmark of AS, 50 can take years to develop. 51 Second, the phenotypic subclassification favours clinical research in the major subtypes, AS and PsA, at the expense of less prevalent subtypes. For example, TNF blockers have been well studied and broadly implemented in AS and PsA but are still not registered for other SpA subforms. Finally, this phenotypic classification recognizes established, full-blown forms of SpA but fails to adequately capture less typical presentations of the disease. In the early 1990s, two different classification criteria for SpA were developed: the Amor criteria and the European Spondyloarthropathy Study Group criteria. 41,52 Recently the Assessment of SpondyloArthritis international Society (ASAS) developed novel classification criteria for SpA. 53,54 Key aspects of these criteria are that (I) SpA is subdivided into axial SpA and peripheral SpA, (II) imaging abnormalities are not only defined by X-ray, but also by MRI, as new bone formation is a slow process that may take years to become visible on X-rays and (III) HLA-B27 positivity is an important entry criterion, allowing SpA patients without imaging abnormalities to be identified. The distinction between axial and peripheral disease in the ASAS 13

15 criteria makes a lot of sense, as there is pathophysiological and therapeutic evidence that axial and peripheral disease may be driven by slightly distinct cellular and molecular mechanisms (e.g. the major cellular source of IL-17 in peripheral SpA is mast cells and to a lesser degree neutrophils, 40 whereas neutrophils are the major IL-17-expressing cell population in axial SpA, 55 and DMARDs might be effective in peripheral SpA, but not in axial SpA) Also, different outcome measurements are used for axial and peripheral disease. These novel criteria open new possibilities for clinical research, including therapeutic trials, in established as well as early SpA, as they are broader and more inclusive than the classical phenotypic classification. LIMITATIONS OF THE ASAS CLASSIFICATION CRITERIA Despite the aforementioned advantages of the new ASAS classification criteria, they also confront us with new challenges. First, these criteria were developed for classification, thereby to facilitate clinical research, but not for diagnosis. For example, a recent study compared the prevalence of axial SpA in at-risk patients with chronic back pain according to the ASAS axial SpA criteria with the diagnosis of axial SpA by the rheumatologist s expert clinical diagnosis and concluded that the concordance yielded a kappa of only 0.41 (low to moderate agreement), suggesting that the ASAS axial SpA criteria and rheumatology experts captured somewhat different patient populations. According to the ASAS axial SpA criteria, 24% of the patients were undiagnosed and 21% were misdiagnosed in clinical practice, 60 which could be interpreted as a false-positive diagnosis (in comparison with the gold standard, the rheumatologist s expert opinion) in 24% of patients when applied as a diagnostic tool. This is probably related to the fact that MRI and HLA-B27 are major entry criteria in the ASAS classification but are diagnostically only useful in patients with a moderately high a priori chance of SpA. 61 A second limitation of the ASAS criteria is that the subdivision into either axial SpA or peripheral SpA is not a true reflection of the clinical reality since approximately 30% of the SpA patients have combined axial and peripheral disease manifestations. 54 The ASAS proposed that the axial SpA criteria should be applied in patients with combined axial and peripheral disease. 54 With this definition, however, there is a risk that a substantial number of SpA patients with peripheral disease and some degree of back pain will fail to be classified as axial SpA, and thus as SpA all together, despite fulfilling the peripheral SpA criteria. 62 For example, an HLA-B27-negative patient with Crohn s disease with a monoarthritis of the knee, active inflammatory back pain for 6 months, but normal imaging of the sacroiliac joints would not classify as SpA, since neither the imaging nor the HLA-B27 criterion are met, which are required for axial SpA. However, the patient would be classified as SpA if the peripheral SpA criteria had been used. This questions whether patients with active axial (back pain) as well as peripheral symptoms (arthritis, enthesitis or dactylitis) should not enter the ASAS criteria by either one of the two criteria arms (Figure 2). A third challenge of the new criteria is that we need to reassess carefully specific subgroups of axial and/or peripheral SpA (e.g. HLA-B27-positive vs -negative patients, and patients fulfilling the imaging vs clinical arm of the axial SpA criteria) to see whether they are different in terms of disease activity, disease burden, long-term outcome and response to treatment. Many studies addressing the latter issue are being conducted and results are awaited. 14

16 1 General introduction Figure 2. Classification of patients with axial, peripheral or combined symptoms. Patients with pure active axial symptoms should be classified according to the axial SpA criteria, and patients with pure peripheral symptoms according to the peripheral SpA criteria. However, according to ASAS the patients with combined axial and peripheral symptoms could only classify as having SpA if they fulfil the axial SpA criteria. We propose a small modification to these criteria so patients with combined symptoms could enter either of the two criteria arms to reduce the chance that these patients unfairly misclassify as not having SpA. HOW TO MEASURE DISEASE ACTIVITY AND OUTCOME An additional issue arising here is that the heterogeneity of the disease manifestations of SpA not only make it a challenge to properly classify the disease, but also to measure the disease activity and outcome of therapy. Most disease activity and outcome parameters either capture only a single disease manifestation (axial, peripheral or extra-articular) or are only validated in a single phenotypic SpA subtype (AS or PsA). This hampers the evaluation of SpA as a whole, especially in patients with combined axial and peripheral disease manifestations, where wellvalidated AS parameters such as the BASDAI 63 and Ankylosing Spondylitis Disease Activity Score (ASDAS) 64,65 may underestimate peripheral disease activity and where swollen and tender joint counts as used in PsA miss aspects of axial disease. It remains to be determined whether global measurements such as patient s and physician s global assessments of disease activity are perhaps the most appropriate measurements to monitor disease activity in SpA patients with a combined phenotype. 62,66,67 Alternatively, it appears that measurements initially developed for axial disease, such as the BASDAI and ASDAS, may also be responsive to effective therapy in SpA patients with peripheral and/or combined disease. 62 Besides the challenge to capture the different aspects of the disease in a single clinical disease activity measurement, an additional complication is that CRP and ESR are often normal in SpA. Therefore several groups have attempted to identify additional blood and/or target tissue biomarkers that reflect disease activity, such as MMP-3 in the blood 68 and CD163 + macrophages in the synovial tissue. 34 Unfortunately, these biomarkers are not sufficiently robust and reliable to be used as a surrogate marker in individual patients and thus their use remains restricted to 15

17 clinical trials. The same issue applies to biomarkers of structural damage in general and new bone formation in particular: measuring structural outcome by the gold standard of X-rays requires long-term observation 69 and may be inadequate to assess new bone formation in peripheral disease, 70,71 but biochemical biomarkers of structural outcome lack sensitivity and specificity. Collectively there is still a very high unmet need for reliable biomarkers to monitor disease activity (inflammation) and outcome (structural damage) in SpA. OPTIMAL USE OF TNF BLOCKERS One of the major drivers behind the current efforts to unify the SpA concept was the successful introduction of TNF blocking therapies in SpA >10 years ago. 66,67,72-74 First, the major rationale for their use in SpA was the clinical and immunopathological overlap between SpA and Crohn s disease Second, the original trials were not restricted to a single subtype but included different forms of axial and peripheral SpA. 66,67 And third, the subsequent phase III trials demonstrated clinical efficacy for axial disease, arthritis and enthesitis, as well as extraarticular manifestations such as uveitis, psoriasis and IBD ,78-80 Collectively these data pointed towards TNF as an overarching inflammatory principle in SpA. Surprisingly, however, the clinical development programs of the previous decade were completely focused on AS and PsA and TNF blockers were neither studied nor registered in other SpA subforms. It has been only recently that the SpA community regained interest in exploring the potential use of TNF inhibitors in non-prototypical forms of SpA. Several clinical trials have shown that TNF blockers are not only effective in AS, but also in the non-radiographic forms of axial SpA And several small studies have also reported that TNF inhibitors are equally efficacious in patients with non-as, non-psa peripheral SpA, which was recently confirmed in two randomized placebo-controlled clinical trials. 62,91 Not only the clinical response, but also the immunopathological response in the synovial tissue upon effective treatment appeared to be similar across the different phenotypic subtypes in peripheral SpA. 87,92 Taken together, these studies strongly support the notion that TNF inhibition should not be restricted to specific SpA subtypes but, on the contrary, should be made broadly available to treat the different disease manifestations in patients with diverse phenotypic presentations. WHY DO WE NEED NEW THERAPIES? TNF inhibitors have dramatically improved the treatment and care of SpA patients, but there is still a high unmet need for other and better therapeutic compounds. First, up to 40% of patients do not respond well to anti-tnf treatment, either due to adverse reactions and intolerance or due to inefficacy. One of the proposed mechanisms behind the inefficacy of TNF blockers is the development of anti-drug antibodies that can either increase the clearance of the TNF inhibitor 93 or directly neutralize the functional part of the biologic drug. 94 Whether this is also a real clinical issue in the treatment of SpA is still to be determined. If immunogenicity is a real issue, it could be useful to try a second or third TNF inhibitor, but registry studies indicate that the response rate and drug survival decrease with each TNF inhibitor. 95,96 16

18 Second, TNF blockade does not induce long-lasting remission. In axial SpA, almost all patients relapse after interruption of treatment, often after a couple of months Moreover, recently we also found similar relapse rates in peripheral SpA. 103 Third, TNF blockade halts joint destruction 104 but fails to slow down new bone formation. The lack of effect on osteoproliferation has been previously reported for axial disease in AS and more recently also for peripheral disease in PsA. 71 It remains highly debated whether osteoproliferation might be halted if anti-tnf treatment is started earlier in the disease course 108 or whether bone remodelling is really independent from inflammation and has to be targeted separately. 109 Although there is emerging evidence that high-dose NSAIDs may have an inhibitory effect on new bone formation, developing adequate treatments to halt new bone formation is still a major unmet medical need in SpA as, e.g. physical function is independently determined by both disease activity and radiographic damage of the spine. 114 Taken together, the lack of efficacy and/or intolerance in a significant proportion of patients, the rapid relapse upon treatment interruption and the failure to halt new bone formation emphasize the need for other effective treatments besides TNF blockade in SpA. 1 General introduction WHICH EXISTING COMPOUNDS HAVE BEEN TESTED SO FAR? Several compounds have been tested in SpA (mainly AS and PsA) based on their efficacy in diseases such as RA and psoriasis, but none of these compounds has proven to be highly effective (Table 1). First, T cell- and B cell-targeted therapies such as abatacept (CTLA4-Ig), alefacept (anti-lfa3), efalizumab (anti-cd11a) 120 and rituximab (anti-cd20) only showed modest, if any, therapeutic efficacy in SpA, which is consistent with the concept that SpA belongs to the class of autoinflammatory rather than autoimmune disorders. 126 Second, biologic therapies targeting key proinflammatory cytokines other than TNF have not been successful, as blockade of IL-1 with anakinra and inhibition of IL-6 with both tocilizumab 130 and sarilumab 131 failed to show clinical efficacy in AS. Third, other small molecule compounds such as thalidomide 132,133 and pamidronate (a bisphosphonate) 134 also failed, as these drugs were respectively too toxic for widespread use and only moderately effective in AS. NOVEL TREATMENT OPTIONS ON THE HORIZON In contrast to the previously mentioned treatments, a number of novel treatment strategies have recently yielded promising results in SpA (Table 1). Apremilast, an orally available small molecule inhibitor of phosphodiesterase 4 (PDE4), which increases intracellular cyclic adenosine monophosphate (camp) and thus modulates multiple pro- and anti-inflammatory mediators, has demonstrated clinical efficacy in PsA 135 and a possible trend towards clinical effect in AS 136 in phase II trials. The major interest, however, has recently focused on therapies targeting the IL-23/IL-17 axis, as there is an extensive rationale for targeting this pathway in SpA. First, AS as well as the SpA-associated disorders psoriasis and Crohn s disease are associated with a single nucleotide polymorphism of IL-23R Second, in vitro evidence indicates that the unfolded protein response, a cellular stress programme that can be initiated by HLA-B27 misfolding, 17

19 Tabel 1. Summary of clinical trials in SpA with compounds other than NSAIDs, DMARDs or TNF inhibitors Drug Mechanism of action Study design SpA subtype T-cell and B-cell targeted biological therapies Abatacept Inhibits T-cell co-stimulation Randomized, double-blind, placebo-controlled, phase II trial PsA Abatacept Inhibits T-cell co-stimulation Open-label study AS Abatacept Inhibits T-cell co-stimulation Open-label study Axial SpA Alefacept Inhibits T-cell activation Randomized, double blind, placebo-controlled, phase II trial Alefacept Inhibits T-cell activation Open label extension study PsA Efalizumab Inhibits T-cell activation Randomized, double blind, placebo-controlled, phase II trial Rituximab B-cell depletion Open-label study AS PsA PsA Rituximab B-cell depletion Open-label study PsA Rituximab B-cell depletion Open-label study PsA Anticytokine biological therapies Anakinra IL-1R inhibitor Open-label study AS Anakinra IL-1R inhibitor Open-label study AS Anakinra IL-1R inhibitor Open-label study PsA Tocilizumab IL-6R inhibitor Randomized, double blind, placebo-controlled, phase II trial Sarilumab IL-6R inhibitor Randomized, double blind, placebo-controlled, phase II trial AS AS Secukimumab IL-17A inhibitor Randomized, double blind, placebo-controlled, phase II trial Secukimumab IL-17A inhibitor Randomized, double blind, placebo-controlled, phase II trial Ustekinumab IL-23 and IL-12 inhibitor Randomized, double blind, placebo-controlled, cross-over, phase II trial Ustekinumab IL-23 and IL-12 inhibitor Randomized, double blind, placebo-controlled, phase III trial Ustekinumab IL-23 and IL-12 inhibitor Randomized, double blind, placebo-controlled, phase III trial AS PsA PsA PsA PsA 18

20 Treatment arms Primary outcome Main results Reference 1 Placebo (n=42) Abatacept 3 mg/kg (n=45) Abatacept 10 mg/kg (n=40) Abatacept 30/10 mg/kg (n=42) Anti-TNF naive (n=15) Anti-TNF IR (n=15) ACR20 at day 169 Placebo: 19% 3 mg/kg: 33% 10 mg/kg: 48% 30/10 mg/kg: 42% ASAS40 at week 24 Anti-TNF naive: 13% Anti-TNF IR: 0% Abatacept 10 mg/kg (n=7) BASDAI50 at week 24 14% [117] Placebo + MTX (n=62) Alefacept + MTX (n=123) ACR20 at week 24 Placebo + MTX: 23% Alefacept + MTX: 54% Alefacept + MTX (n=160) ACR20 at week 24 54% [119] Placebo (n=53) Efalizumab (n=54) Anti-TNF naive (n=10) Anti-TNF IR (n=10) ACR20 week 24 Placebo: 19% Efalizumab: 28% ASAS20 at week 24 Anti-TNF naive: 30% Anti-TNF IR: 50% Rituximab day 0 and 14 (n=9) PsARC at week 24 56% [123] Rituximab day 0 and 15 (n=20) ACR20 at week 24 35% [124] [115] [116] [118] [120] [121] General introduction Anakinra 100 mg/day (n=9) ASAS20 at 3 months 67% [127] Anakinra 100 mg/day (n=20) ASAS20 at week 24 25% [128] Anakinra 100 mg/day (n=20) PsARC at week 24 30% [129] Placebo (n=51) Tocilizumab 8 mg/kg (n=51) Placebo (n=50) Sarilumab 100 mg/week (n=52) Sarilumab 150 mg/week (n=50) Sarilumab 100 mg/2 weeks (n=49) Sarilumab 150 mg/2 weeks (n=50) Sarilumab 200 mg/2 weeks (n=50) Placebo (n=6) Secukinumab (n=24) Placebo (n=14) Secukinumab (n=28) Placebo till week 12 (n=70) Ustekinumab till week 12 (n=76) Placebo (n=206) Ustekinumab 45 mg (n=205) Ustekinumab 90 mg (n=204) Placebo (n=104) Ustekinumab 45 mg (n=103) Ustekinumab 90 mg (n=105) ASAS20 at week 12 Placebo: 28% Tocilizumab: 37% ASAS20 at week 12 Placebo: 24% Sarilumab 100 mg/week: 19% Sarilumab 150 mg/week: 38% Sarilumab 100 mg/2 weeks: 19% Sarilumab 150 mg/2 weeks: 30% Sarilumab 200 mg/2 weeks: 30% ASAS20 at week 6 Placebo: 17% Secukinumab: 61% ACR20 at week 6 Placebo: 23% Secukinumab: 39% ACR20 at week 12 Placebo: 14% Ustekinumab: 42% ACR20 at week 24 Placebo: 23% Ustekinumab 45 mg: 42% Ustekinumab 90 mg: 50% ACR20 at week 24 Placebo: 20% Ustekinumab 45 mg: 44% Ustekinumab 90 mg: 44% [130] [131] [146] [147] [148] [149] [150] 19

21 Tabel 1. Continued Drug Mechanism of action Study design SpA subtype Small molecules Thalidomide Inhibition of TNF and IL-12 Open-label study AS Thalidomide Inhibition of TNF and IL-12 Open-label study AS Pamidronate Bisphosphonate Randomized, double-blind, dose-response study Apremilast PDE4 inhibitor Randomized, double blind, placebo-controlled, phase II trial Apremilast PDE4 inhibitor Randomized, double blind, placebo-controlled, unpowered, phase II trial Imatinib Tyrosine kinase inhibitor Open-label study 4 AS, 2 PsA AS PsA AS Anti-TNF IR = anti-tnf inadequate response; PsARC = psoriatic arthritis response criteria; BASFI = Bath Ankylosing Spondylitis Functional Index; BASMI = Bath Ankylosing Spondylitis Metrology Index; strongly increases the production of IL-23, 140 and macrophages of AS patients tend to produce increased amounts of IL Third, the number of circulating CD4 + IL-17 + T cells is increased in AS, 142,143 especially the KIR3DL2-expressing T cells that respond to cell surface HLA-B27 homodimers 144 and IL-17-producing γδ T cells. 145 Fourth, we and others have demonstrated increased IL-17 expression by innate immune cells such as neutrophils and mast cells in SpA target tissues such as the facet joints and synovial tissue. 40,55 Finally, IL-23 overexpression in mice was recently shown to induce enthesitis, which may partially phenocopy human SpA via ROR-γt + CD3 + CD4 - CD8 - T cells, which elaborate proinflammatory cytokines including IL-17 and IL More importantly, however, a proof-of-concept randomized placebo-controlled trial with secukinumab, a fully human anti-il-17a monoclonal antibody, in AS demonstrated high response rates, with >60% of the patients reaching a 20% improvement in ASAS criteria (ASAS20) as early as week 6 of treatment. 146 A similar trial with secukinumab showed a trend towards clinical efficacy in PsA, 147 whereas ustekinumab, a monoclonal antibody directed against the p40 chain common to both IL-23 and IL-12, was superior to placebo in several clinical trials in PsA Ongoing clinical trials will have to determine the exact efficacy and safety of several compounds targeting the IL-23/IL-17 axis for treatment of the different disease manifestations of SpA, including the potential effect on structural damage. Such compounds may not only be directed towards key cytokines of this axis, but may also directly target a specific pathogenic IL-17-producing cell population. For example, a small open-label trial has shown some positive clinical effects of treatment with imatinib, a tyrosine kinase inhibitor that may target c-kitpositive mast cells and innate lymphoid cells as potential sources of IL-17 in SpA. 40,151 Similarly, small molecules targeting key transcription factors of the IL-17 axis, such as RORγ, are currently in preclinical development and may be further explored for the treatment of SpA. 20

22 Treatment arms Primary outcome Main results Reference 1 Thalidomide 200 mg/day (n=30) 20% improvement in 4/7 indices* at 12 months 70% [132] Thalidomide 200 mg/day (n=13) ASAS20 at 6 months 62% [133] 10 mg/month (n=43) 60 mg/month (n=41) Placebo (n=68) Apremilast 2x20 mg/day (n=69) Apremilast 1x40 mg/day (n=67) Placebo (n=19) Apremilast 2x30 mg/day Imatinib mg/day ΔBASDAI at 6 months 10 mg: 0.93 (15%) 60 mg: 2.22 (35%) ACR20 at week 12 Placebo: 12% Apremilast 2x20 mg/day: 44% Apremilast 1x40 mg/day: 36% ΔBASDAI, ΔBASFI, ΔBASMI, ΔBASG Several at week 12, e.g. BASDAI50 and ASAS40 Numerically greater improvement apremilast vs. placebo [134] [135] [136] Both 50% [151] General introduction BASG = Bath Ankylosing Spondylitis Global score; * the 7 indices consisted of: BASDAI, BASFI), duration of morning stiffness, total body pain, spinal pain, patient s global assessment of disease activity and physician s global assessment of disease activity CONCLUSION The global management of SpA has dramatically improved over the last 15 years, not only by the introduction of TNF blockers for this disease, but also by ongoing efforts to develop unifying concepts and to understand pathogenic pathways underlying the different phenotypic SpA manifestations. Further progress is expected from studies focusing on early disease recognition, proper classification and adequate assessment and monitoring of disease activity of not only AS and PsA, but also other SpA forms. This needs to be combined with clinical trials focusing on the optimization of existing therapies, such as TNF blockade for early and/or atypical disease, as well as clinical development of novel drugs, in particular those targeting the IL-23/IL-17 pathways. The aim is not only to discover novel anti-inflammatory alternatives for TNF inhibition in SpA, but to develop genuine diseasemodifying and remission-inducing therapies. Moreover, the ongoing proof-of-concept studies with novel targeted therapies will also provide further clues regarding the pathophysiologic mechanisms behind the different disease manifestations of SpA and are thereby likely to contribute to the development of unifying translational concepts for this disabling disorder. RHEUMATOLOGY KEY MESSAGES -- SpA is a disease with heterogeneous manifestations that share a common pathophysiologic background. -- Promising novel therapeutic options for SpA, such as those targeting the IL-23/IL-17 axis, are pending. -- Future compounds have to be tested in SpA as a whole, regardless of subphenotype. 21

23 Acknowledgements Prof Dr D. Baeten was supported by a VIDI grant from The Netherlands Organization for Scientific Research (NWO) and by a grant from the Dutch Arthritis Foundation (Reumafonds). Both authors were involved in drafting the article or revising it critically for important intellectual content and both authors approved the final version for publication. REFERENCES 1 Akkoc N. Are spondyloarthropathies as common as rheumatoid arthritis worldwide? A review. Curr Rheumatol Rep 2008;10: Reveille JD, Witter JP, Weisman MH. Prevalence of axial spondylarthritis in the United States: estimates from a cross-sectional survey. Arthritis Care Res (Hoboken) 2012;64: Dougados M, Baeten D. Spondyloarthritis. Lancet 2011;377: Chandran V, Schentag CT, Brockbank JE et al. Familial aggregation of psoriatic arthritis. Ann Rheum Dis 2009;68: Dernis E, Said-Nahal R, D Agostino MA et al. Recurrence of spondylarthropathy among firstdegree relatives of patients: a systematic crosssectional study. Ann Rheum Dis 2009;68: Schlosstein L, Terasaki PI, Bluestone R, Pearson CM. High association of an HL-A antigen, W27, with ankylosing spondylitis. N Engl J Med 1973;288: Reveille JD. Genetics of spondyloarthritis beyond the MHC. Nat Rev Rheumatol 2012;8: Girschick HJ, Guilherme L, Inman RD et al. Bacterial triggers and autoimmune rheumatic diseases. Clin Exp Rheumatol 2008;26:S McGonagle D, Lories RJ, Tan AL, Benjamin M. The concept of a synovio-entheseal complex and its implications for understanding joint inflammation and damage in psoriatic arthritis and beyond. Arthritis Rheum 2007;56: McGonagle D, Gibbon W, Emery P. Classification of inflammatory arthritis by enthesitis. Lancet 1998;352: McGonagle D, Gibbon W, O Connor P et al. Characteristic magnetic resonance imaging entheseal changes of knee synovitis in spondylarthropathy. Arthritis Rheum 1998;41: D Agostino MA, Said-Nahal R, Hacquard-Bouder C et al. Assessment of peripheral enthesitis in the spondylarthropathies by ultrasonography combined with power Doppler: a cross-sectional study. Arthritis Rheum 2003;48: Frediani B, Falsetti P, Storri L et al. Ultrasound and clinical evaluation of quadricipital tendon enthesitis in patients with psoriatic arthritis and rheumatoid arthritis. Clin Rheumatol 2002;21: Emad Y, Ragab Y, Shaarawy A et al. Can magnetic resonance imaging differentiate undifferentiated arthritis based on knee imaging? J Rheumatol 2009;36: Iagnocco A, Spadaro A, Marchesoni A et al. Power Doppler ultrasonographic evaluation of enthesitis in psoriatic arthritis. A multi-center study. Joint Bone Spine 2012;79: Narvaez J, Narvaez JA, de Albert M et al. Can magnetic resonance imaging of the hand and wrist differentiate between rheumatoid arthritis and psoriatic arthritis in the early stages of the disease? Semin Arthritis Rheum 2012;42: Falsetti P, Frediani B, Fioravanti A et al. Sonographic study of calcaneal entheses in erosive osteoarthritis, nodal osteoarthritis, rheumatoid arthritis and psoriatic arthritis. Scand J Rheumatol 2003;32: Genc H, Cakit BD, Tuncbilek I, Erdem HR. Ultrasonographic evaluation of tendons and enthesal sites in rheumatoid arthritis: comparison with ankylosing spondylitis and healthy subjects. Clin Rheumatol 2005;24: Ibrahim G, Groves C, Chandramohan M et al. Clinical and ultrasound examination of the Leeds enthesitis index in psoriatic arthritis and rheumatoid arthritis. ISRN Rheumatol 2011;2011: Marzo-Ortega H, Tanner SF, Rhodes LA et al. Magnetic resonance imaging in the assessment of metacarpophalangeal joint disease in early psoriatic and rheumatoid arthritis. Scand J Rheumatol 2009;38: Paramarta JE, van der Leij C, Gofita I et al. Peripheral joint inflammation in early onset spondyloarthritis is not specifically related to enthesitis. Ann Rheum Dis 2013, Advance Access published 25 April 2013, doi: /annrheumdis Naredo E, Moller I, de Miguel E et al. High prevalence of ultrasonographic synovitis and enthesopathy in patients with psoriasis without 22

24 psoriatic arthritis: a prospective case-control study. Rheumatology 2011;50: Emad Y, Ragab Y, Gheita T et al. Knee enthesitis and synovitis on magnetic resonance imaging in patients with psoriasis without arthritic symptoms. J Rheumatol 2012;39: Francois RJ, Gardner DL, Degrave EJ, Bywaters EG. Histopathologic evidence that sacroiliitis in ankylosing spondylitis is not merely enthesitis. Arthritis Rheum 2000;43: Muche B, Bollow M, Francois RJ et al. Anatomic structures involved in early- and late-stage sacroiliitis in spondylarthritis: a detailed analysis by contrast-enhanced magnetic resonance imaging. Arthritis Rheum 2003;48: Lories RJ, Matthys P, de Vlam K et al. Ankylosing enthesitis, dactylitis, and onychoperiostitis in male DBA/1 mice: a model of psoriatic arthritis. Ann Rheum Dis 2004; Van Duivenvoorde LM, Dorris ML, Satumtira N et al. Relationship between inflammation, bone destruction, and osteoproliferation in spondyloarthritis in the HLA-B27/human β2-microglobulin-transgenic rat model of spondyloarthritis. Arthritis Rheum 2012;64: Laloux L, Voisin MC, Allain J, et al. Immunohistological study of entheses in spondyloarthropathies: comparison in rheumatoid arthritis and osteoarthritis. Ann Rheum Dis 2001;60: McGonagle D, Marzo-Ortega H, O Connor P, et al. Histological assessment of the early enthesitis lesion in spondyloarthropathy. Ann Rheum Dis 2002;61: Armaka M, Apostolaki M, Jacques P, Kontoyiannis DL, Elewaut D, Kollias G. Mesenchymal cell targeting by TNF as a common pathogenic principle in chronic inflammatory joint and intestinal diseases. J Exp Med 2008;205: Sherlock JP, Joyce-Shaikh B, Turner SP, et al. IL-23 induces spondyloarthropathy by acting on ROR-gammat(+) CD3(+)CD4(-)CD8(-) entheseal resident T cells. Nat Med 2012;18: Adamopoulos IE, Tessmer M, Chao CC, et al. IL-23 is critical for induction of arthritis, osteoclast formation, and maintenance of bone mass. J Immunol 2011;187: Baeten D, Moller HJ, Delanghe J, Veys EM, Moestrup SK, De Keyser F. Association of CD163+ macrophages and local production of soluble CD163 with decreased lymphocyte activation in spondylarthropathy synovitis. Arthritis Rheum 2004;50: Baeten D, Kruithof E, De Rycke L, et al. Infiltration of the synovial membrane with macrophage subsets and polymorphonuclear cells reflects global disease activity in spondyloarthropathy. Arthritis Res Ther 2005;7:R359-R Canete JD, Santiago B, Cantaert T, et al. Ectopic lymphoid neogenesis in psoriatic arthritis. Ann Rheum Dis 2007;66: Cantaert T, Kolln J, Timmer T, et al. B lymphocyte autoimmunity in rheumatoid synovitis is independent of ectopic lymphoid neogenesis. J Immunol 2008;181: Kruithof E, Baeten D, De Rycke L, et al. Synovial histopathology of psoriatic arthritis, both oligoand polyarticular, resembles spondyloarthropathy more than it does rheumatoid arthritis. Arthritis Res Ther 2005;7:R569-R Baeten D, Demetter P, Cuvelier C, et al. Comparative study of the synovial histology in rheumatoid arthritis, spondyloarthropathy, and osteoarthritis: influence of disease duration and activity. Ann Rheum Dis 2000;59: Yeremenko N, Noordenbos T, Cantaert T, et al. Disease-specific and inflammation-independent stromal alterations in spondyloarthritis synovitis. Arthritis Rheum 2013;65: Noordenbos T, Yeremenko N, Gofita I, et al. Interleukin-17-positive mast cells contribute to synovial inflammation in spondylarthritis. Arthritis Rheum 2012;64: Dougados M, van der Linden S, Juhlin R, et al. The European Spondylarthropathy Study Group preliminary criteria for the classification of spondylarthropathy. Arthritis Rheum 1991;34: Baeten D, Breban M, Lories R, Schett G, Sieper J. Are spondylarthritides related but distinct conditions or a single disease with a heterogeneous phenotype? Arthritis Rheum 2013;65: Rahman P, Inman RD, Gladman DD, Reeve JP, Peddle L, Maksymowych WP. Association of interleukin-23 receptor variants with ankylosing spondylitis. Arthritis Rheum 2008;58: Rahman P, Inman RD, Maksymowych WP, Reeve JP, Peddle L, Gladman DD. Association of interleukin 23 receptor variants with psoriatic arthritis. J Rheumatol 2009;36: Said-Nahal R, Miceli-Richard C, Berthelot JM, et al. The familial form of spondylarthropathy: a clinical study of 115 multiplex families. Groupe Francais d Etude Genetique des Spondylarthropathies. Arthritis Rheum 2000;43: Said-Nahal R, Miceli-Richard C, d Agostino MA, et al. Phenotypic diversity is not determined 1 General introduction 23

25 by independent genetic factors in familial spondylarthropathy. Arthritis Rheum 2001;45: Porcher R, Said-Nahal R, d Agostino MA, Miceli-Richard C, Dougados M, Breban M. Two major spondylarthropathy phenotypes are distinguished by pattern analysis in multiplex families. Arthritis Rheum 2005;53: Vandooren B, Noordenbos T, Ambarus C, et al. Absence of a classically activated macrophage cytokine signature in peripheral spondylarthritis, including psoriatic arthritis. Arthritis Rheum 2009;60: Tran TM, Dorris ML, Satumtira N, et al. Additional human beta2-microglobulin curbs HLA-B27 misfolding and promotes arthritis and spondylitis without colitis in male HLA-B27-transgenic rats. Arthritis Rheum 2006;54: Mau W, Zeidler H, Mau R, et al. Clinical features and prognosis of patients with possible ankylosing spondylitis. Results of a 10-year followup. J Rheumatol 1988;15: Van der Linden S, Valkenburg HA, Cats A. Evaluation of diagnostic criteria for ankylosing spondylitis. A proposal for modification of the New York criteria. Arthritis Rheum 1984;27: Amor B, Dougados M, Mijiyawa M. [Criteria of the classification of spondylarthropathies]. Rev Rhum Mal Osteoartic 1990;57: Rudwaleit M, van der Heijde D, Landewe R, et al. The development of Assessment of SpondyloArthritis international Society classification criteria for axial spondyloarthritis (part II): validation and final selection. Ann Rheum Dis 2009;68: Rudwaleit M, van der Heijde D, Landewe R, et al. The Assessment of SpondyloArthritis International Society classification criteria for peripheral spondyloarthritis and for spondyloarthritis in general. Ann Rheum Dis 2011;70: Appel H, Maier R, Wu P, et al. Analysis of IL-17(+) cells in facet joints of patients with spondyloarthritis suggests that the innate immune pathway might be of greater relevance than the Th17-mediated adaptive immune response. Arthritis Res Ther 2011;13:R Clegg DO, Reda DJ, Abdellatif M. Comparison of sulfasalazine and placebo for the treatment of axial and peripheral articular manifestations of the seronegative spondylarthropathies: a Department of Veterans Affairs cooperative study. Arthritis Rheum 1999;42: Chen J, Liu C. Sulfasalazine for ankylosing spondylitis. Cochrane Database Syst Rev 2005;CD Chen J, Veras MM, Liu C, Lin J. Methotrexate for ankylosing spondylitis. Cochrane Database Syst Rev 2013;2:CD van Denderen JC, van der Paardt M, Nurmohamed MT, de Ryck YM, Dijkmans BA, van der Horst- Bruinsma IE. Double blind, randomised, placebo controlled study of leflunomide in the treatment of active ankylosing spondylitis. Ann Rheum Dis 2005;64: Strand V, Rao SA, Shillington AC, Cifaldi MA, McGuire M, Ruderman EM. Prevalence of axial SpA in US rheumatology practices: Assessment of ASAS criteria vs. rheumatology expert clinical diagnosis. Arthritis Care Res (Hoboken) 2013 Published on 22 February doi: /acr Rudwaleit M, van der Heijde D, Khan MA, Braun J, Sieper J. How to diagnose axial spondyloarthritis early. Ann Rheum Dis 2004;63: Paramarta JE, De Rycke L, Heijda TF, et al. Efficacy and safety of adalimumab for the treatment of peripheral arthritis in spondyloarthritis patients without ankylosing spondylitis or psoriatic arthritis. Ann Rheum Dis 2013;72: Garrett S, Jenkinson T, Kennedy LG, Whitelock H, Gaisford P, Calin A. A new approach to defining disease status in ankylosing spondylitis: the Bath Ankylosing Spondylitis Disease Activity Index. J Rheumatol 1994;21: Lukas C, Landewe R, Sieper J, et al. Development of an ASAS-endorsed disease activity score (ASDAS) in patients with ankylosing spondylitis. Ann Rheum Dis 2009;68: Machado P, Landewe R, Lie E, et al. Ankylosing Spondylitis Disease Activity Score (ASDAS): defining cut-off values for disease activity states and improvement scores. Ann Rheum Dis 2011;70: Van Den Bosch F, Kruithof E, Baeten D, De Keyser F, Mielants H, Veys EM. Effects of a loading dose regimen of three infusions of chimeric monoclonal antibody to tumour necrosis factor alpha (infliximab) in spondyloarthropathy: an open pilot study. Ann Rheum Dis 2000;59: Van Den Bosch F, Kruithof E, Baeten D, et al. Randomized double-blind comparison of chimeric monoclonal antibody to tumor necrosis factor alpha (infliximab) versus placebo in active spondylarthropathy. Arthritis Rheum 2002;46: Vandooren B, Kruithof E, Yu DT, et al. Involvement of matrix metalloproteinases and their inhibitors in peripheral synovitis and down-regulation by tumor necrosis factor alpha blockade in spondylarthropathy. Arthritis Rheum 2004;50:

26 69 Wanders AJ, Landewe RB, Spoorenberg A, et al. What is the most appropriate radiologic scoring method for ankylosing spondylitis? A comparison of the available methods based on the Outcome Measures in Rheumatology Clinical Trials filter. Arthritis Rheum 2004;50: Kavanaugh A, van der Heijde D, McInnes IB, et al. Golimumab in psoriatic arthritis: one-year clinical efficacy, radiographic, and safety results from a phase III, randomized, placebo-controlled trial. Arthritis Rheum 2012;64: Finzel S, Kraus S, Schmidt S, et al. Bone anabolic changes progress in psoriatic arthritis patients despite treatment with methotrexate or tumour necrosis factor inhibitors. Ann Rheum Dis 2013;72: Braun J, Brandt J, Listing J, et al. Treatment of active ankylosing spondylitis with infliximab: a randomised controlled multicentre trial. Lancet 2002;359: Brandt J, Haibel H, Cornely D, et al. Successful treatment of active ankylosing spondylitis with the anti-tumor necrosis factor alpha monoclonal antibody infliximab. Arthritis Rheum 2000;43: Mease PJ, Goffe BS, Metz J, VanderStoep A, Finck B, Burge DJ. Etanercept in the treatment of psoriatic arthritis and psoriasis: a randomised trial. Lancet 2000;356: De Vos M, Mielants H, Cuvelier C, Elewaut A, Veys E. Long-term evolution of gut inflammation in patients with spondyloarthropathy. Gastroenterology 1996;101: Baeten D, Demetter P, Culevelier CA, et al. Macrophages expressing the scavenger receptor CD163: a link between immune alterations of the gut and synovial inflammation in spondyloarthropathy. J Pathol 2002;196: Demetter P, De Vos M, Van Huysse JA, et al. Colon mucosa of patients both with spondyloarthritis and Crohn s disease is enriched with macrophages expressing the scavenger receptor CD163. Ann Rheum Dis 2005;64: Dougados M, Combe B, Braun J, et al. A randomised, multicentre, double-blind, placebocontrolled trial of etanercept in adults with refractory heel enthesitis in spondyloarthritis: the HEEL trial. Ann Rheum Dis 2010;69: Braun J, Baraliakos X, Listing J, Sieper J. Decreased incidence of anterior uveitis in patients with ankylosing spondylitis treated with the anti-tumor necrosis factor agents infliximab and etanercept. Arthritis Rheum 2005;52: Rutgeerts P, D Haens G, Targan S, et al. Efficacy and safety of retreatment with anti-tumor necrosis factor antibody (infliximab) to maintain remission in Crohn s disease. Gastroenterology 1999;117: Barkham N, Keen HI, Coates LC, et al. Clinical and imaging efficacy of infliximab in HLA-B27-Positive patients with magnetic resonance imagingdetermined early sacroiliitis. Arthritis Rheum 2009;60: Haibel H, Rudwaleit M, Listing J, et al. Efficacy of adalimumab in the treatment of axial spondylarthritis without radiographically defined sacroiliitis: results of a twelve-week randomized, double-blind, placebo-controlled trial followed by an open-label extension up to week fifty-two. Arthritis Rheum 2008;58: Sieper J, van der Heijde D, Dougados M, et al. Efficacy and safety of adalimumab in patients with non-radiographic axial spondyloarthritis: results of a randomised placebo-controlled trial (ABILITY-1). Ann Rheum Dis 2013;72: Song IH, Hermann K, Haibel H, et al. Effects of etanercept versus sulfasalazine in early axial spondyloarthritis on active inflammatory lesions as detected by whole-body MRI (ESTHER): a 48- week randomised controlled trial. Ann Rheum Dis 2011;70: Brandt J, Haibel H, Reddig J, Sieper J, Braun J. Successful short term treatment of severe undifferentiated spondyloarthropathy with the anti-tumor necrosis factor-alpha monoclonal antibody infliximab. J Rheumatol 2002;29: Brandt J, Khariouzov A, Listing J, et al. Successful short term treatment of patients with severe undifferentiated spondyloarthritis with the antitumor necrosis factor-alpha fusion receptor protein etanercept. J Rheumatol 2004;31: Kruithof E, De Rycke L, Roth J, et al. Immunomodulatory effects of etanercept on peripheral joint synovitis in the spondylarthropathies. Arthritis Rheum 2005;52: Marzo-Ortega H, McGonagle D, O Connor P, Emery P. Efficacy of etanercept in the treatment of the entheseal pathology in resistant spondylarthropathy: a clinical and magnetic resonance imaging study. Arthritis Rheum 2001;44: Marzo-Ortega H, McGonagle D, O Connor P, Emery P. Efficacy of etanercept for treatment of Crohn s related spondyloarthritis but not colitis. Ann Rheum Dis 2003;62: General introduction 25

27 90 Meyer A, Chatelus E, Wendling D, et al. Safety and efficacy of anti-tumor necrosis factor alpha therapy in ten patients with recent-onset refractory reactive arthritis. Arthritis Rheum 2011;63: Mease P, Sieper J, Van Den Bosch F, Rahman P, Obermeyer K, Pangan AL. Randomized controlled trial of adalimumab in patients with peripheral spondyloarthritis [abstract]. Arthritis Rheum 2012;64(Suppl):S248-S Kruithof E, Baeten D, Van Den Bosch F, Mielants H, Veys EM, De Keyser F. Histological evidence that infliximab treatment leads to downregulation of inflammation and tissue remodelling of the synovial membrane in spondyloarthropathy. Ann Rheum Dis 2005;64: Van der Laken CJ, Voskuyl AE, Roos JC, et al. Imaging and serum analysis of immune complex formation of radiolabelled infliximab and antiinfliximab in responders and non-responders to therapy for rheumatoid arthritis. Ann Rheum Dis 2007;66: Van Schouwenburg PA, van de Stadt LA, de Jong RN, et al. Adalimumab elicits a restricted antiidiotypic antibody response in autoimmune patients resulting in functional neutralisation. Ann Rheum Dis 2013;72: Glintborg B, Ostergaard M, Krogh NS, et al. Clinical response, drug survival and predictors thereof in 432 ankylosing spondylitis patients after switching tumour necrosis factor alpha inhibitor therapy: results from the Danish nationwide DANBIO registry. Ann Rheum Dis 2013;2: Glintborg B, Ostergaard M, Krogh NS, et al. Clinical response, drug survival and predictors thereof among 548 switchers of tumor necrosis factor alpha inhibitor therapy in psoriatic arthritis. Results from the Danish nationwide DANBIO registry [abstract]. Arthritis Rheum 2012;64(Suppl):S Baraliakos X, Listing J, Brandt J, et al. Clinical response to discontinuation of anti-tnf therapy in patients with ankylosing spondylitis after 3 years of continuous treatment with infliximab. Arthritis Res Ther 2005;7:R439-R Brandt J, Khariouzov A, Listing J, et al. Six-month results of a double-blind, placebo-controlled trial of etanercept treatment in patients with active ankylosing spondylitis. Arthritis Rheum 2003;48: Brandt J, Listing J, Haibel H, et al. Long-term efficacy and safety of etanercept after readministration in patients with active ankylosing spondylitis. Rheumatology (Oxford) 2005;44: Breban M, Vignon E, Claudepierre P, et al. Efficacy of infliximab in refractory ankylosing spondylitis: results of a six-month open-label study. Rheumatology (Oxford) 2002;41: Heldmann F, Brandt J, van der Horst-Bruinsma IE, et al. The European ankylosing spondylitis infliximab cohort (EASIC): a European multicentre study of long term outcomes in patients with ankylosing spondylitis treated with infliximab. Clin Exp Rheumatol 2011;29: Song IH, Althoff CE, Haibel H, et al. Frequency and duration of drug-free remission after 1 year of treatment with etanercept versus sulfasalazine in early axial spondyloarthritis: 2 year data of the ESTHER trial. Ann Rheum Dis 2012;71: Paramarta JE, Heijda TF, Baeten DL. Fast relapse upon discontinuation of tumor necrosis factor blocking therapy in patients with peripheral spondyloarthritis. Ann Rheum Dis : Van der Heijde D, Kavanaugh A, Gladman DD, et al. Infliximab inhibits progression of radiographic damage in patients with active psoriatic arthritis through one year of treatment: Results from the induction and maintenance psoriatic arthritis clinical trial 2. Arthritis Rheum 2007;56: Van der Heijde D, Landewe R, Baraliakos X, et al. Radiographic findings following two years of infliximab therapy in patients with ankylosing spondylitis. Arthritis Rheum 2008;58: Van der Heijde D, Landewe R, Einstein S, et al. Radiographic progression of ankylosing spondylitis after up to two years of treatment with etanercept. Arthritis Rheum 2008;58: Van der Heijde D, Salonen D, Weissman BN, et al. Assessment of radiographic progression in the spines of patients with ankylosing spondylitis treated with adalimumab for up to 2 years. Arthritis Res Ther 2009;11:R Sieper J, Appel H, Braun J, Rudwaleit M. Critical appraisal of assessment of structural damage in ankylosing spondylitis: implications for treatment outcomes. Arthritis Rheum 2008;58: Lories RJ, Luyten FP, de VK. Progress in spondylarthritis. Mechanisms of new bone formation in spondyloarthritis. Arthritis Res Ther 2009;11: Boersma JW. Retardation of ossification of the lumbar vertebral column in ankylosing spondylitis by means of phenylbutazone. Scand J Rheumatol 1976;5: Kroon F, Landewe R, Dougados M, van der Heijde D. Continuous NSAID use reverts the effects of inflammation on radiographic progression in 26

28 patients with ankylosing spondylitis. Ann Rheum Dis 2012;71: Poddubnyy D, Rudwaleit M, Haibel H, et al. Effect of non-steroidal anti-inflammatory drugs on radiographic spinal progression in patients with axial spondyloarthritis: results from the German Spondyloarthritis Inception Cohort. Ann Rheum Dis 2012;71: Wanders A, Heijde D, Landewe R, et al. Nonsteroidal antiinflammatory drugs reduce radiographic progression in patients with ankylosing spondylitis: a randomized clinical trial. Arthritis Rheum 2005;52: Landewe R, Dougados M, Mielants H, van der Tempel H, van der Heijde D. Physical function in ankylosing spondylitis is independently determined by both disease activity and radiographic damage of the spine. Ann Rheum Dis 2009;68: Mease P, Genovese MC, Gladstein G, et al. Abatacept in the treatment of patients with psoriatic arthritis: results of a six-month, multicenter, randomized, double-blind, placebocontrolled, phase II trial. Arthritis Rheum 2011;63: Song IH, Heldmann F, Rudwaleit M, et al. Treatment of active ankylosing spondylitis with abatacept: an open-label, 24-week pilot study. Ann Rheum Dis 2011;70: Lekpa FK, Farrenq V, Canoui-Poitrine F, et al. Lack of efficacy of abatacept in axial spondylarthropathies refractory to tumor-necrosis-factor inhibition. Joint Bone Spine 2012;79: Mease PJ, Gladman DD, Keystone EC. Alefacept in combination with methotrexate for the treatment of psoriatic arthritis: results of a randomized, double-blind, placebo-controlled study. Arthritis Rheum 2006;54: Mease PJ, Reich K. Alefacept with methotrexate for treatment of psoriatic arthritis: open-label extension of a randomized, double-blind, placebo-controlled study. J Am Acad Dermatol 2009;60: Papp KA, Caro I, Leung HM, Garovoy M, Mease PJ. Efalizumab for the treatment of psoriatic arthritis. J Cutan Med Surg 2007;11: Song IH, Heldmann F, Rudwaleit M, et al. Different response to rituximab in tumor necrosis factor blocker-naive patients with active ankylosing spondylitis and in patients in whom tumor necrosis factor blockers have failed: a twenty-four-week clinical trial. Arthritis Rheum 2010;62: Nocturne G, Dougados M, Constantin A, et al. Rituximab in the spondyloarthropathies: data of eight patients followed up in the French Autoimmunity and Rituximab (AIR) registry. Ann Rheum Dis 2010;69: Jimenez-Boj E, Stamm TA, Sadlonova M, et al. Rituximab in psoriatic arthritis: an exploratory evaluation. Ann Rheum Dis 2012;71: Mease P, Genovese M, Ritchlin C, Boyle D, Quistberg A, Kavanaugh A. Rituximab in psoriatic arthritis: results of an open label study [abstract]. Ann Rheum Dis 2010;69(Suppl3); Wendling D, Dougados M, Berenbaum F, et al. Rituximab treatment for spondyloarthritis. A nationwide series: data from the AIR registry of the French Society of Rheumatology. J Rheumatol 2012;39: Ambarus C, Yeremenko N, Tak PP, Baeten D. Pathogenesis of spondyloarthritis: autoimmune or autoinflammatory? Curr Opin Rheumatol 2012;24: Tan AL, Marzo-Ortega H, O Connor P, Fraser A, Emery P, McGonagle D. Efficacy of anakinra in active ankylosing spondylitis: a clinical and magnetic resonance imaging study. Ann Rheum Dis 2004;63: Haibel H, Rudwaleit M, Listing J, Sieper J. Open label trial of anakinra in active ankylosing spondylitis over 24 weeks. Ann Rheum Dis 2005;64: Jung N, Hellmann M, Hoheisel R, et al. An openlabel pilot study of the efficacy and safety of anakinra in patients with psoriatic arthritis refractory to or intolerant of methotrexate (MTX). Clin Rheumatol 2010;29: Sieper J, Porter-Brown B, Thompson L, Harari O, Dougados M. Tocilizumab (TCZ) is not effective for the treatment of ankylosing spondylitis (AS): results of a phase 2, international, multicentre, randomised, double-blind, placebo-controlled trial [abstract]. Ann Rheum Dis 2012;71(Suppl 3): Sieper J, Inman RD, Badalamenti S, Radin A, Braun J. Sarilumab for the treatment of ankylosing spondylitis: results of a phase 2, randomized, double-blind, placebo-controlled, international study (ALIGN) [abstract]. Ann Rheum Dis 2012;71(Suppl 3): Huang F, Gu J, Zhao W, Zhu J, Zhang J, Yu DT. Oneyear open-label trial of thalidomide in ankylosing spondylitis. Arthritis Rheum 2002;47: Wei JC, Chan TW, Lin HS, Huang F, Chou CT. Thalidomide for severe refractory ankylosing spondylitis: a 6-month open-label trial. J Rheumatol 2003;30: Maksymowych WP, Jhangri GS, Fitzgerald AA, et al. A six-month randomized, controlled, 1 General introduction 27

29 double-blind, dose-response comparison of intravenous pamidronate (60 mg versus 10 mg) in the treatment of nonsteroidal antiinflammatory drug-refractory ankylosing spondylitis. Arthritis Rheum 2002;46: Schett G, Wollenhaupt J, Papp K, et al. Oral apremilast in the treatment of active psoriatic arthritis: results of a multicenter, randomized, double-blind, placebo-controlled study. Arthritis Rheum 2012;64: Pathan E, Abraham S, Van-Rossen E, et al. Efficacy and safety of apremilast, an oral phosphodiesterase 4 inhibitor, in ankylosing spondylitis. Ann Rheum Dis 2013;72: Burton PR, Clayton DG, Cardon LR, et al. Association scan of 14,500 nonsynonymous SNPs in four diseases identifies autoimmunity variants. Nat Genet 2007;39: Duerr RH, Taylor KD, Brant SR, et al. A genomewide association study identifies IL23R as an inflammatory bowel disease gene. Science 2006;314: Cargill M, Schrodi SJ, Chang M, et al. A largescale genetic association study confirms IL12B and leads to the identification of IL23R as psoriasis-risk genes. Am J Hum Genet 2007;80: DeLay ML, Turner MJ, Klenk EI, Smith JA, Sowders DP, Colbert RA. HLA-B27 misfolding and the unfolded protein response augment interleukin-23 production and are associated with Th17 activation in transgenic rats. Arthritis Rheum 2009;60: Zeng L, Lindstrom MJ, Smith JA. Ankylosing spondylitis macrophage production of higher levels of interleukin-23 in response to lipopolysaccharide without induction of a significant unfolded protein response. Arthritis Rheum 2011;63: Jandus C, Bioley G, Rivals JP, Dudler J, Speiser D, Romero P. Increased numbers of circulating polyfunctional Th17 memory cells in patients with seronegative spondylarthritides. Arthritis Rheum 2008;58: Shen H, Goodall JC, Hill Gaston JS. Frequency and phenotype of peripheral blood Th17 cells in ankylosing spondylitis and rheumatoid arthritis. Arthritis Rheum 2009;60: Bowness P, Ridley A, Shaw J, et al. Th17 cells expressing KIR3DL2+ and responsive to HLA- B27 homodimers are increased in ankylosing spondylitis. J Immunol 2011;186: Kenna TJ, Davidson SI, Duan R, et al. Enrichment of circulating interleukin-17-secreting interleukin-23 receptor-positive gamma/delta T cells in patients with active ankylosing spondylitis. Arthritis Rheum 2012;64: Baeten D, Sieper J, Braun J, et al. Anti-interleukin- 17A monoclonal antibody secukinumab in treatment of ankylosing spondylitis: a randomised, double-blind, placebo-controlled trial. Lancet 2013;382: McInnes IB, Sieper J, Braun J, et al. Efficacy and safety of secukinumab, a fully human anti-interleukin- 17A monoclonal antibody, in patients with moderate-to-severe psoriatic arthritis: a 24-week, randomised, double-blind, placebo-controlled, phase II proof-of-concept trial. Ann Rheum Dis 2013, Advance Access published 29 January 2013, doi: /annrheumdis Gottlieb A, Menter A, Mendelsohn A, et al. Ustekinumab, a human interleukin 12/23 monoclonal antibody, for psoriatic arthritis: randomised, double-blind, placebo-controlled, crossover trial. Lancet 2009;373: McInnes IB, Kavanaugh A, Gottlieb AB, et al. Ustekinumab in patients with active psoriatic arthritis: results of the phase 3, multicenter, double-blind, placebo-controlled PSUMMIT I study [abstract]. Ann Rheum Dis 2012;71(Suppl3):S Ritchlin, CT, Gottlieb AB, McInnes IB, et al. Ustekinumab in active psoriatic arthritis including patients previously treated with anti-tnf agents: Results of a phase 3, multicenter, double-blind, placebo-controlled study [abstract]. Ann Rheum Dis 2012;71(Suppl3):S Eklund KK, Remitz A, Kautiainen H, Reitamo S, Leirisalo-Repo M. Three months treatment of active spondyloarthritis with imatinib mesylate: an open-label pilot study with six patients. Rheumatology (Oxford) 2006;45:

30

31

32 Outline of this thesis 2

33 AIM AND OUTLINE OF THIS THESIS In this thesis we aimed to study two major clinical challenges in spondyloarthritis (SpA) as described in the general introduction (Chapter 1). Part I of this thesis focuses on the disease phenotypes in SpA. Here we aimed to assess which tissue is primary affected in early SpA and describe and compare the different SpA subpopulations as classified by two types of criteria. Part II of this thesis focuses on novel treatment options in SpA where we investigated whether it is useful to extend the availability of tumor necrosis factor (TNF) inhibitors, which are known to be very effective in the treatment of the prototypic subtypes of SpA, to treat all the different subgroups described in Part I. Also we describe and evaluate novel therapeutic targets in the treatment of SpA. PART I: DISEASE PHENOTYPES IN SPONDYLOARTHRITIS Articular manifestations in SpA comprises involvement of the axial skeleton and of peripheral joints. The relative contribution of synovial versus entheseal versus bone inflammation in axial and peripheral SpA remains unclear. In Chapter 3 we revisited the hypothesis proposing that enthesitis is the primary peripheral joint lesion in SpA whereas synovitis is the primary lesion in rheumatoid arthritis (RA). In order to test the validity of this hypothesis we assessed enthesitis and synovitis in a combined magnetic resonance imaging (MRI) and histopathological study of early untreated SpA and RA. In the next two chapters we describe the clinical similarities and differences of various SpA subpopulations. In Chapter 4 we compared clinical features and disease activity of undifferentiated SpA with ankylosing spondylitis (AS) and psoriatic arthritis (PsA), using the classical phenotypic classification. In Chapter 5 we use the novel classification criteria developed by the Assessment of SpondyloArthritis international Society (ASAS) and assessed to what extent these criteria really reflect axial versus peripheral disease. PART II: NOVEL TREATMENTS IN SPONDYLOARTHRITIS The search for novel treatment options in SpA does not only consist of developing and testing new therapeutic compounds, but also concerns the optimal use of already existing therapies. In Chapter 6 we describe an investigator-initiated, randomized, placebo-controlled trial which aimed to investigate whether the TNF inhibitor adalimumab, which has been extensively studied in AS and PsA, is also effective in peripheral SpA patients who do not fulfill the criteria for AS and PsA. This group, which is described in Chapter 4, has not been included in most clinical trials and, accordingly, biologicals are not yet available in daily clinical practice for these patients. We also assessed whether TNF inhibition can lead to long-lasting remission in peripheral SpA, as sometimes observed in RA, or whether interruption of treatment leads to rapid disease relapse, as seen in AS (Chapter 7). Finally, we investigated in Chapter 8 whether anti-drug antibodies are associated with poor clinical response and/or rapid relapse in peripheral SpA treated with adalimumab. The final chapters of this thesis study novel therapeutic options in SpA. In Chapter 9 we discuss the rationale for targeting the interleukin (IL)-23/IL-17 axis in SpA. One of the compounds targeting this axis is secukinumab, a monoclonal antibody directed towards IL-17A. In Chapter 32

34 10 we report the proof-of-concept randomized, placebo-controlled trial with secukinumab in AS and indicate that this is the first biologic drug beyond TNF blockers that reached its primary endpoint in a phase II trial in AS. Finally, in Chapter 11 we address the mast cell, the main cell type expressing IL-17 in SpA synovitis, as possible therapeutic target in SpA by conducting an investigator-initiated, proof-of-concept study with nilotinib, a tyrosine kinase inhibitor. In this randomized placebo-controlled clinical trial we investigated both the immunopathological and clinical effects of nilotinib on peripheral and axial SpA. 2 Thesis outline 33

35

36 Part one Disease phenotypes in spondyloarthritis

37

38 Peripheral joint inflammation in early onset spondyloarthritis is not specifically related to enthesitis Jacqueline E. Paramarta 1*, Christiaan van der Leij 2*, Ioana Gofita 1,3, Nataliya Yeremenko 1,3, Marleen G. van de Sande 1, Maria J. de Hair 1, Paul P. Tak 1,4, Mario Maas 2, Dominique Baeten 1,3 1 Department of Clinical Immunology and Rheumatology, Academic Medical Center/University of Amsterdam, Amsterdam, The Netherlands, 2 Department of Radiology, Academic Medical Center/University of Amsterdam, Amsterdam, The Netherlands, 3 Laboratory of Experimental Immunology, Academic Medical Center/University of Amsterdam, Amsterdam, The Netherlands, 4 GlaxoSmithKline, Stevenage, UK 3 *Both authors contributed equally Ann Rheum Dis Apr;73(4):735-40

39 ABSTRACT Objectives A pivotal MRI study of knee arthritis indicated that enthesitis was more frequently observed in established spondyloartritis (SpA) than rheumatoid arthritis (RA). Subsequent MRI and ultrasound studies, however, failed to consistently demonstrate primary synovitis in RA versus primary enthesitis in SpA. Therefore, the current study aimed to reassess enthesitis versus synovitis in peripheral arthritis by a combined imaging and histopathological study in early untreated disease. Methods MRI and mini-arthroscopic synovial biopsy sampling were performed in 41 patients with early untreated knee or ankle arthritis, who were diagnosed with SpA (n=13), RA (n=20) or crystal arthropathy (n=8) at follow-up. MRI evaluation of enthesitis and synovitis, and immunohistochemical characterisation of synovitis were performed by two observers blinded to diagnosis. Results MRI showed similar prevalence of perientheseal fluid/oedema (67% vs 75%), perientheseal bone marrow oedema (0% vs 10%) and entheseal enhancement (46% vs 47%) in SpA versus RA, respectively. The number and distribution of affected entheseal sites were not different between both diseases. The MRI synovitis score was significantly higher in SpA (median 1.4; IQR ) compared with RA (median 0.5; IQR ) (p=0.028). Synovial histopathology showed a numerical increase in infiltrating cells in SpA versus RA synovitis which reached significance for CD163 macrophages in the synovial sublining (p=0.030). There were no differences compared with the crystal arthropathy control group. Conclusions Enthesitis on MRI is not a specific feature of peripheral arthritis in recent onset SpA versus RA. Synovitis is prominent in both diseases as evaluated by MRI and immunohistochemistry. 38

40 INTRODUCTION Differences in clinical pattern, synovial histopathology, structural damage and response to treatment indicate that peripheral arthritis in spondyloarthritis (SpA) is distinct from rheumatoid arthritis (RA). A pivotal hypothesis attempting to explain these differences proposed that in peripheral arthritis synovitis is the primary feature of RA, whereas enthesitis is the primary lesion in SpA. 1 This was based on a MRI study in 10 SpA and 10 patients with RA with recent onset knee effusion, which showed that entheseal abnormalities were prominent in SpA: all SpA but only four patients with RA had perientheseal fluid/oedema, and six SpA but no patients with RA had perientheseal bone marrow oedema. 2 The value of this hypothesis is that it proposed a conceptual framework potentially explaining the tissue-specificity of SpA lesions and that it emphasized the role of mechanical stress in SpA. 1,3 However, this hypothesis also has important weaknesses. First, the original study was cross-sectional in established disease and did not include histopathology. 2 Therefore, one cannot conclude that enthesitis is primary to synovitis in SpA. Second, other imaging studies attempting to reproduce these findings yielded conflicting results. Some studies confirmed the higher prevalence of enthesitis in SpA versus RA 4-8 but others did not find more enthesitis in SpA compared with other inflammatory joint diseases, 9-12 questioning the specificity of this finding for SpA. Third, studies of peripheral enthesitis in SpA-associated diseases such as psoriasis, inflammatory bowel disease and human leukocyte antigen (HLA)-B27 positive uveitis in patients without arthritic symptoms demonstrated more enthesitis in these patients compared with healthy or non-spa associated controls, but most studies only assessed subclinical enthesitis and not subclinical synovitis, making it impossible to conclude which relationship exists between enthesitis and synovitis in the state before onset of full-blown SpA. However, the studies which investigated subclinical enthesitis and subclinical synovitis, showed that both features were more often present in SpA-related diseases compared with controls. 13,14 Furthermore, since follow-up data is scarce, it is unknown whether these patients will really develop full-blown SpA over time. Fourth, animal models also yielded conflicting results. The ankylosing enthesitis model in DBA/1 mice demonstrates that entheseal stress can lead to osteoproliferation and mild inflammation, 21 reminiscent of heel enthesitis in human SpA. Enthesitis is, however, also observed in erosive polyarthritis of TNF-ΔARE mice. 22 Moreover, interleukin-23 overexpression can lead to primary synovitis 23 and enthesitis. 24 Of most relevance to human SpA, finally, spontaneous peripheral arthritis in human HLA-B27/β2 microglobulin transgenic rats is characterised by pronounced synovitis in the absence of enthesitis. 25 Testing the validity of the proposed hypothesis 1 is of relevance for pathophysiological and clinical research. If the hypothesis is correct, pathophysiological studies should focus on experimental models of enthesitis rather than synovitis and translational research should aim to obtain and study human entheseal rather than synovial biopsies. Similarly, clinical studies should focus on drugs and interventions targeting enthesitis rather than compounds with proven efficacy on synovitis. Therefore, this study aimed to further elucidate the share of enthesitis and synovitis in peripheral arthritis in early SpA by comparing the presence and extent of these features in early untreated SpA and RA by paired MRI and synovial histopathology. 3 Enthesitis is not specific for recent onset SpA 39

41 METHODS Study patients Forty-one patients from our early arthritis cohort with less than 12 months disease duration and active knee (n=28) or ankle (n=13) arthritis were included. 26 None of the patients had received disease modifying antirheumatic drugs, corticosteroids or biological treatment before inclusion. At 2 year follow-up, patients were classified as SpA (n=13) according to the European Spondyloarthropathy Study Group (ESSG) 27 and the Assessment of SpondyloArthritis International Society (ASAS) criteria 28 or RA (n=20) according to the 1987 ACR 29 and 2010 American College of Rheumatology (ACR)/European League Against Rheumatism (EULAR) criteria. 30,31 The remaining eight patients were diagnosed as crystal arthropathy (CA) based on crystals in their synovial fluid or chondrocalcinosis on x-ray. The characteristics of the patients are depicted in table 1. The SpA group consisted of seven patients with undifferentiated spondyloarthritis, four patients with psoriatic arthritis (PsA) and two patients with reactive arthritis. In the RA group 13/20 patients were anticyclic citrullinated protein antibodies and/ or IgM rheumatoid factor positive. Demographic and clinical data (patient s global assessment of disease activity, 66/68 swollen and tender joint count, C reactive protein and erythrocyte sedimentation rate) and MRI scans of an inflamed knee or ankle were obtained at baseline. Twenty-three patients (SpA=8, RA=10, CA=5) underwent paired mini-arthroscopic synovial biopsy sampling of the inflamed knee or ankle, as described previously. 32,33 All patients gave written informed consent to participate in the study as approved by the local ethics committee. MRI acquisitions Scans were made using a 1.5 T MRI Scanner (GE Signa Horizon Echospeed, LX9.0, General Electric Medical Systems, Milwaukee), in supine position with the knee or ankle placed in a Table 1. Clinical characteristics of study patients SpA (n=13) RA (n=20) CA (n=8) Male gender, n (%) 10 (77) 5 (25) 6 (75) Age (years) 46 (32-51) 53 (39-57) 53 (48-61) Disease duration (years) 0.2 ( ) 0.1 ( ) 0.2 ( ) BMI (kg/m2) 25 (22-27) 27 (23-32) 28 (23-34) Patient s global assessment (0-100 mm VAS) 61 (29-82) 62 (36-78) 47 (20-77) 66 Swollen joint count 1 (1-4) 9 (4-14) 3 (1-4) 68 Tender joint count 6 (1-10) 16 (9-29) 2 (1-6) CRP (mg/l) 14.9 ( ) 12.4 ( ) 3.0 ( ) ESR (mm/h) 30 (10-52) 27 (12-33) 10 (5-26) Except where indicated otherwise, values are the median (interquartile range). SpA = spondyloarthritis; RA = rheumatoid arthritis; CA = crystal arthropathy; BMI = body mass index; VAS = visual analogue scale; CRP = C-reactive protein; ESR = erythrocyte sedimentation rate. 40

42 dedicated extremity coil (Quadrature Detection) centrally in the magnetic field. A bolus of 0.1 mmol/kg Gd-DTPA contrast agent (Magnevist, Schering AG, Berlin) was given intravenously. The knee joint protocol consisted of a sagittal short τ inversion recovery (STIR) (repetition time/echo time/inversion time 4000/73/170, slice thickness 4 mm, field of view 240 mm, matrix ), sagittal T1-weighted spin echo (SE) (500/14, 4 mm, 240 mm, ), and axial T2- weighted SE with fat saturation (FS) (4000/97, 4 mm, 200 mm, ) before contrast, and sagittal T1-weighted SE and transversal T1-weighted SE FS (500/14, 4 mm, 200 mm, ) after contrast. The ankle joint protocol consisted of a sagittal STIR (7300/70/170, 3 mm, 200 mm, ), sagittal T1-weighted SE (600/14, 3 mm, 200 mm, ), and axial T2-weighted SE FS (3800/95, 3 mm, 200 mm, ) before contrast, and sagittal T1-weighted SE and axial T1-weighted SE FS (600/14, 3 mm, 200 mm, ) after contrast. MRI scoring Enthesitis was assessed by the presence or absence of perientheseal fluid/oedema, and perientheseal bone marrow oedema on the T2-weighted and T1-weighted images, as described by McGonagle et al, 2 and on STIR. Additionally, we assessed entheseal enhancement using precontrast and postcontrast T1-weighted images. The following entheseal regions of the knee were examined: quadriceps femoris tendon insertion, patellar ligament insertion on the patella and tibial tuberositas, iliotibial band insertion, lateral collateral ligament origo and insertion, lateral joint capsule origo and insertion, anterior cruciate ligament origo and insertion, posterior cruciate ligament origo and insertion, biceps femoris tendon insertion, semimembranosus tendon insertion, medial collateral ligament origo and insertion, posterior joint capsule origo and insertion. For the ankle the Achilles tendon insertion and plantar fascia insertion were assessed. Synovitis was scored semiquantitatively on postcontrast sagittal T1-weighted images in four compartments of the knee (ranging from 0 to 3: no or minimal enhancing synovium (compared with healthy individuals) to large volume of enhancing synovial tissue). The lateral compartment was the space lateral from the patella below the suprapatellar bursa, the medial compartment the space medial from the patella below the suprapatellar bursa, and the central compartment the space behind the patella (containing the Hoffa s fat pad) and around the cruciate ligaments below the suprapatellar bursa. The suprapatellar compartment consisted of the suprapatellar bursa. For the knee synovitis score the mean was taken from the four compartments. Concerning the ankles, synovitis was assessed similarly in the tibial/fibular/talar, talar/ calcaneal, talar/navicular and calcaneal/cuboidal joint. For the ankle synovitis score the mean was taken from the four joints. The scoring was done by two musculoskeletal radiologists (CvdL and MM), who were blinded to the patients diagnoses. 3 Enthesitis is not specific for recent onset SpA Synovial immunohistochemistry Synovial biopsy samples were snap-frozen in Tissue-Tek OCT (Miles, Elkhart, Indiana, USA) immediately after collection. Cryostat sections (5 μm) were cut and mounted on Star Frost adhesive glass slides (Knittelgläser, Braunschweig). Frozen sections were acetone fixed and stained with monoclonal antibodies directed towards T cells (CD3; UHT1, Dako, Glostrup, Denmark), B 41

43 cells (CD22; RFB4; Chemicon, Billerica, Massachusetts, USA), macrophages (CD68; EBM-11, Dako), alternatively activated macrophages (CD163; 5cFAT, BMA Biomedicals, Augst, Switzerland) and endothelial cells (von Willebrand factor; F8/86, Dako). After rinsing, sections were sequentially incubated with a biotinylated secondary antibody, a streptavidin-horseradish peroxidise link, aminoethylcarbazole substrate as chromogen (all Dako) and haematoxylin as counterstain. As negative control parallel sections were incubated with isotype and concentration-matched monoclonal antibodies. All samples were stained in a single run to minimise technical biases, and subsequently scored semiquantitatively for cellular infiltration by two independent observers (NY and DB) who were blinded to the patients diagnoses, as described previously Statistical analysis Data are presented as the median and IQR. Since our main question was to compare SpA to RA, Mann- Whitney U tests were used to compare these diseases. In case of differences, additional comparison to the control CA group was performed by Mann-Whitney U tests to assess whether SpA or RA was distinct form other types of arthritis. Categorical data were analysed using χ2 tests. Correlation between enthesitis and synovitis parameters was assessed using Spearman s correlation tests. All statistical tests were two-sided. P-values less than 0.05 were considered statistically significant. RESULTS Similar enthesitis frequency on MRI in SpA and RA peripheral arthritis We first assessed the frequency of perientheseal abnormalities on knee and ankle MRIs during peripheral arthritis. Perientheseal fluid/oedema and entheseal enhancement occurred frequently, but perientheseal bone marrow oedema was rarely observed (figure 1). In SpA 66.7% had perientheseal fluid/oedema, 0% perientheseal bone marrow oedema and 45.5% entheseal enhancement, compared with 75%, 10% and 47.1%, respectively in RA (table 2). Accordingly, the frequency of these lesions was not significantly different between early untreated SpA and RA. Also separate analysis of knee and ankle MR images failed to reveal differences (table 2). Similar extent and localisation of enthesitis on MRI in SpA and RA peripheral arthritis As perientheseal abnormalities were frequently observed in peripheral arthritis independently of diagnosis, we next investigated whether the number or localisation of the enthesitis sites differed between SpA and RA. The median number of sites with perientheseal fluid/oedema per patient was 1.0 (IQR ), perientheseal bone marrow oedema 0.0 (IQR ) and entheseal enhancement 0.0 (IQR ) in SpA, and respectively 1.0 (IQR ), 0.0 (IQR ) and 0.0 (IQR ) in RA (figure 2). Accordingly, the number of entheseal sites with MRI abnormalities was not different between early SpA and RA peripheral arthritis, in the combined analysis of knees and ankles and in the separate analysis of these joints. As to the localisation of the MRI abnormalities, perientheseal fluid/oedema and entheseal enhancement in the knee were most frequently seen at the patellar ligament insertion on the patella and tibial tuberositas, and the quadriceps femoris tendon insertion. In the ankle, these 42

44 3 Figure 1. Representative magnetic resonance (MR) images of spondyloarthritis (SpA) (A-E) and rheumatoid arthritis (RA) patients (F-J). The MR images are sagittal STIR images (A, F), sagittal T1 weighted images before (B, D, G, I) and after (C, E, H, J) contrast enhancement. Image A and F show perientheseal fluid/oedema at the patellar tendon insertion at the tibial tuberositas (closed arrows). Image C and H show subtle enhancement of the quadriceps tendon attachment (open arrows). Image E and J show extensive synovial hypertrophy and enhancement (*) in the lateral compartment of the knee. Table 2. Frequency of enthesitis on MRI by disease Peri-entheseal fluid/oedema Peri-entheseal bone marrow oedema Entheseal enhancement SpA (n=8) Knee (n=28) Ankle (n=13) Knee and ankle (n=41) RA (n=13) CA (n=7) SpA (n=5) RA (n=7) CA (n=1) SpA (n=13) RA (n=20) CA (n=8) 5 (62.5) 9 (69.2) 7 (100) 3 (75.0) 6 (85.7) 1 (100) 8 (66.7) 15 (75.0) 8 (100) 0 (0.0) 2 (15.4) 1 (16.7) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 2 (10.0) 1 (14.3) 3 (42.9) 4 (36.4) 5 (71.4) 2 (50.0) 4 (66.7) 0 (0.0) 5 (45.5) 8 (47.1) 5 (62.5) Enthesitis is not specific for recent onset SpA Values are the number (percentage) of patients with a positive finding on magnetic resonance imaging (MRI). SpA = spondyloarthritis; RA = rheumatoid arthritis; CA = crystal arthropathy. There were no significant differences in frequency of enthesitis between the SpA and RA, neither when the knee and ankle MR images were analyzed separately nor when the data of the knee and ankle joints were combined. were mostly observed at the Achilles tendon and plantar fascia insertion. These localisations were similar in SpA and RA (data not shown). Perientheseal bone marrow oedema was only seen at the quadriceps femoris tendon insertion (two patients with RA) and patellar ligament insertion on the patella (one patient with CA). Taken together, these data indicate that MRI signs of enthesitis were not more predominant or differently distributed in SpA and RA early peripheral arthritis. 43

45 More pronounced MRI synovitis in SpA and RA peripheral arthritis As synovitis has been proposed to be primary in RA but not SpA, 1 we additionally scored the degree of synovitis on MRI. The median synovitis scores were higher in SpA (1.4; IQR ) than RA (0.5; IQR ) (p=0.028) (figure 2). There were no significant differences compared with CA. Also when the knee joints were analysed separately, the synovitis score tended to be higher in SpA (1.5; IQR ) than RA (1.0; IQR ) (p=0.087). For the ankle joints there was a similar numerical difference (1.3; IQR in SpA vs 0.5; IQR in RA) (p=0.198). In both diseases there was no correlation between the synovitis score and the number of affected entheseal sites in the knee or ankle (data not shown). Similar degree of immunohistological synovitis in SpA and RA In order to confirm the MRI findings indicating that synovitis was not more pronounced in RA than SpA early peripheral arthritis, we additionally assessed the severity of synovial inflammation by immunohistochemistry (table 3). The number of infiltrating T cells (CD3), B cells (CD22) and Figure 2. Enthesitis and synovitis assessed by Gd-DTPA enhanced magnetic resonance imaging (MRI) in patients with early spondyloarthritis (SpA), rheumatoid arthritis (RA) and crystal arthorpathy (CA). The panel represents the number of entheseal sites which show signs of enthesitis, assessed by three different enthesitis parameters: peri-entheseal fluid/oedema (A), peri-entheseal bone marrow oedema (B) and entheseal enhancement (C). The mean synovitis score is shown in figure D. The closed symbols represent the results for the knee MR images and the open symbols the results for the ankle MR images. * P<0.05 determined by Mann-Whitney U tests. 44

46 Table 3. Synovitis assessed by immunohistochemistry Marker SpA (n=8) RA (n=10) CA (n=5) P-value SpA vs RA CD3 1.5 ( ) 0.5 ( ) 0.5 ( ) CD ( ) 0.0 ( ) 0.0 ( ) CD68 lining 2.0 ( ) 1.5 ( ) 2.0 ( ) CD68 sublining 2.0 ( ) 0.8 ( ) 1.5 ( ) CD163 lining 1.0 ( ) 1.0 ( ) 1.0 ( ) CD163 sublining 2.0 ( ) 1.0 ( ) 1.0 ( ) von Willebrand Factor 2.0 ( ) 1.5 ( ) 1.5 ( ) Values are the median (interquartile range). SpA = spondyloarthritis; RA = rheumatoid arthritis; CA = crystal arthropathy. Significance of the comparisons is determined by Mann-Whitney U tests. P<0.05 was considered statistically significant. There were no significant differences compared to CA. macrophages (CD68) in the synovial lining and sublining was not different between SpA and RA. In line with previous findings, the number of CD163 macrophages was significantly increased in the synovial sublining of SpA (2.0; IQR ) versus RA (1.0; IQR ) (p=0.030) The expression of von Willebrand factor (reflecting the degree of synovial vascularisation) was not different between SpA and RA. As we previously reported that the degree of synovial sublining infiltration with CD68 macrophages accurately reflected disease activity in peripheral arthritis, 35,37 we confirmed the validity of our synovitis assessments by demonstrating a good correlation between the both MRI synovitis score and the CD68 sublining score (R=0.686) (p=0.001). Taken together, both MRI and immunohistochemistry showed that peripheral arthritis is characterized by pronounced synovitis in early SpA and RA. 3 Enthesitis is not specific for recent onset SpA DISCUSSION This combined imaging and histopathological study assessed the presence and extent of enthesitis and synovitis in early untreated SpAversus RA peripheral arthritis. The major findings are that the frequency as well as the extent and localisation of enthesitis are similar in SpA and RA. Moreover, the degree of synovitis as assessed by MRI and immunohistology is similar or even slightly increased in SpA versus RA. These findings do not support the hypothesis that enthesitis would be primary and synovitis only secondary in SpA, compared with synovitis as primary lesion in RA. 1 Several studies previously attempted to confirm the findings of the original MRI study by McGonagle et al 2 with inconsistent results. Some reported increased frequency of enthesitis in SpA 4-8 whereas others did not find any differences between RA and SpA Interpretation of these data is complicated by the fact that these studies used different approaches (MRI, ultrasound, Power Doppler ultrasound) and enthesitis scores. This results in wide variability of findings with, for example, a reported frequency of enthesitis ranging from no enthesitis at all 6,8 to 60% in RA. 4 To avoid potential biases, we based our approach on the study by McGonagle et al, 2 focusing on the same patient groups and using globally the same MRI parameters. The 45

47 differences between the original report 2 and the present study are that (a) we included more patients, (b) all patients had early untreated disease, (c) diagnosis was made prospectively at follow-up, (d) ankle MRIs were included beside knee MRIs and (e) we assessed synovial histopathology in parallel. Our SpA and RA cohorts were well-matched for systemic disease activity (C reactive protein, erythrocyte sedimentation rate and patient s global assessment) and body mass index. The latter is crucial as patients with PsA often have high body mass index 38 which may lead to mechanical enthesopathy and thereby bias the comparison with RA. Besides demonstrating that there was no difference in enthesitis during peripheral joint inflammation in early SpA and RA, we additionally demonstrated that synovitis was equally present and, if anything, even more pronounced in early SpA than in RA. We and others have previously demonstrated manifest synovitis in established SpA peripheral arthritis ,39,40 We also demonstrated that the overall degree and specific features of synovitis, including mast cell infiltration and a disease-specific myogene signature, were similar in early and established SpA This is in agreement with the current study demonstrating, by MRI and immunohistology, pronounced synovitis in early SpA. Additionally, the synovitis and entheseal scores did not correlate in the present study, and previous studies indicated that not only subclinical enthesitis but also subclinical synovitis was frequently detected in SpA related diseases, even in the absence of clinical joint symptoms. 13,14 Taken together, these data plea against the hypothesis that enthesitis is the primary lesion in SpA and leads over time to secondary synovitis. 1 Although our study strongly questions the role of enthesitis as the unifying and primary feature in SpA, the data should be interpreted carefully. First, we do not question that clinical heel enthesitis is specific for SpA. Our study only assessed enthesitis in the context of peripheral arthritis and we did not study clinical enthesitis. Second, in line with the study of McGonagle et al 2 we only assessed large joints of the lower extremities. Therefore we cannot formally exclude that the presence and/or potential role of enthesitis may be different in small joints. It has been reported that distal interphalangeal (DIP) involvement in PsA may relate to the anatomical enthesis-like organisation of the nail bed. 44 However, similar observations were made in osteoarthritis. 44 Third, although we increased the patient number compared with the study by McGonagle et al 2, the study population remains relatively small. Despite this limitation, however, the current results show not even a trend towards more enthesitis in SpA versus RA. Therefore we do not think that increasing the patient number will change the current conclusions. Fourth, we did not assess axial involvement in SpA and can thus not exclude that enthesitis may play a more prominent role in axial than peripheral SpA. However, a study with sacroiliac joint MRIs found that the synovium and subchondral bone marrow were more frequently inflamed in early disease, while enthesitis was more common in advanced disease. 45 Also, a histopathological study of sacroiliac biopsies showed that synovitis and subchondral bone marrow changes were more predominant, while enthesitis was neither the earliest nor the principal pathological change. 46 This is in agreement with our observations in human HLA-B27/β2 microglobulin transgenic rats where spontaneous tail spondylitis was characterised by a pronounced inflammatory pannus in the absence of enthesitis. 25 Fifth, we only studied entheses by imaging. Ideally we would also include histopathology, however entheseal biopsies are difficult to obtain. Moreover, the studies which studied entheseal histopathology yielded conflicting results. 47,48 Finally, our data do not question that mechanical stress may lead to inflammation in SpA. In agreement with previous 46

48 studies, 49 we found that the most frequently affected sites are entheses exposed to mechanical stress. However, this was the case in SpA, and in RA and CA. Perientheseal abnormalities at these sites are also often found in osteoarthritis and even healthy individuals. 9,10.50 In line with animal data, 21 these data indicate that mechanical stress can induce entheseal changes and inflammation. To what extent this process in SpA is restricted to the entheses versus other stromal tissues such as synovium or bone, remains to be established in future studies. In conclusion, enthesitis and synovitis are important pathological features of early SpA peripheral arthritis. However, enthesitis is not more prevalent or pronounced, and synovitis not less frequent or severe in early SpA versus early RA. These data challenge the hypothesis that enthesitis is the primary immunopathological lesion in SpA. Acknowledgements DB was supported by a VIDI grant from The Netherlands Organisation for Scientific Research (NWO) and by a grant from the Dutch Arthritis Foundation (Reumafonds). REFERENCES 1 McGonagle D, Gibbon W, Emery P. Classification of inflammatory arthritis by enthesitis. Lancet 1998;352: McGonagle D, Gibbon W, O Connor P, et al. Characteristic magnetic resonance imaging entheseal changes of knee synovitis in spondylarthropathy. Arthritis Rheum 1998;41: McGonagle D, Lories RJ, Tan AL, et al. The concept of a synovio-entheseal complex and its implications for understanding joint inflammation and damage in psoriatic arthritis and beyond. Arthritis Rheum 2007;56: D Agostino MA, Said-Nahal R, Hacquard-Bouder C, et al. Assessment of peripheral enthesitis in the spondylarthropathies by ultrasonography combined with power Doppler: a cross-sectional study. Arthritis Rheum 2003;48: Frediani B, Falsetti P, Storri L, et al. Ultrasound and clinical evaluation of quadricipital tendon enthesitis in patients with psoriatic arthritis and rheumatoid arthritis. Clin Rheumatol 2002;21: Emad Y, Ragab Y, Shaarawy A, et al. Can magnetic resonance imaging differentiate undifferentiated arthritis based on knee imaging? J Rheumatol 2009;36: Iagnocco A, Spadaro A, Marchesoni A, et al. Power Doppler ultrasonographic evaluation of enthesitis in psoriatic arthritis. A multi-center study. Joint Bone Spine 2012;79: Narvaez J, Narvaez JA, de Albert M, et al. Can Magnetic Resonance Imaging of the Hand and Wrist Differentiate Between Rheumatoid Arthritis and Psoriatic Arthritis in the Early Stages of the Disease? Semin Arthritis Rheum 2012;42: Falsetti P, Frediani B, Fioravanti A, et al. Sonographic study of calcaneal entheses in erosive osteoarthritis, nodal osteoarthritis, rheumatoid arthritis and psoriatic arthritis. Scand J Rheumatol 2003;32: Genc H, Cakit BD, Tuncbilek I, et al. Ultrasonographic evaluation of tendons and enthesal sites in rheumatoid arthritis: comparison with ankylosing spondylitis and healthy subjects. Clin Rheumatol 2005;24: Marzo-Ortega H, Tanner SF, Rhodes LA, et al. Magnetic resonance imaging in the assessment of metacarpophalangeal joint disease in early psoriatic and rheumatoid arthritis. Scand J Rheumatol 2009;38: Ibrahim G, Groves C, Chandramohan M, et al. Clinical and ultrasound examination of the leeds enthesitis index in psoriatic arthritis and rheumatoid arthritis. ISRN Rheumatol 2011;2011: Naredo E, Moller I, de Miguel E, et al. High prevalence of ultrasonographic synovitis and enthesopathy in patients with psoriasis without psoriatic arthritis: a prospective case-control study. Rheumatology (Oxford) 2011;50: Emad Y, Ragab Y, Gheita T, et al. Knee Enthesitis and Synovitis on Magnetic Resonance Imaging in Patients with Psoriasis without Arthritic Symptoms. J Rheumatol 2012;39: Enthesitis is not specific for recent onset SpA 47

49 15 Erdem CZ, Tekin NS, Sarikaya S, et al. MR imaging features of foot involvement in patients with psoriasis. Eur J Radiol 2008;67: Gisondi P, Tinazzi I, El-Dalati G, et al. Lower limb enthesopathy in patients with psoriasis without clinical signs of arthropathy: a hospital-based casecontrol study. Ann Rheum Dis 2008;67: Munoz-Fernandez S, de Miguel E, Cobo-Ibanez T, et al. Enthesis inflammation in recurrent acute anterior uveitis without spondylarthritis. Arthritis Rheum 2009;60: Emad Y, Ragab Y, Bassyouni I, et al. Enthesitis and related changes in the knees in seronegative spondyloarthropathies and skin psoriasis: magnetic resonance imaging case-control study. J Rheumatol 2010;37: Gutierrez M, Filippucci E, De Angelis R, et al. Subclinical entheseal involvement in patients with psoriasis: an ultrasound study. Semin Arthritis Rheum 2011;40: Ash ZR, Tinazzi I, Gallego CC, et al. Psoriasis patients with nail disease have a greater magnitude of underlying systemic subclinical enthesopathy than those with normal nails. Ann Rheum Dis 2012;71: Lories RJ, Matthys P, de Vlam K, et al. Ankylosing enthesitis, dactylitis, and onychoperiostitis in male DBA/1 mice: a model of psoriatic arthritis. Ann Rheum Dis 2004;63: Armaka M, Apostolaki M, Jacques P, et al. Mesenchymal cell targeting by TNF as a common pathogenic principle in chronic inflammatory joint and intestinal diseases. J Exp Med 2008;205: Adamopoulos IE, Tessmer M, Chao CC, et al. IL-23 is critical for induction of arthritis, osteoclast formation, and maintenance of bone mass. J Immunol 2011;187: Sherlock JP, Joyce-Shaikh B, Turner SP, et al. IL-23 induces spondyloarthropathy by acting on ROR-gammat(+) CD3(+)CD4(-)CD8(-) entheseal resident T cells. Nat Med 2012;18: Van Duivenvoorde LM, Dorris ML, Satumtira N, et al. Relationship between inflammation, bone destruction, and osteoproliferation in spondyloarthritis in HLA-B27/Hubeta2m transgenic rats. Arthritis Rheum 2012;64: De Hair MJ, Harty LC, Gerlag DM, et al. Synovial tissue analysis for the discovery of diagnostic and prognostic biomarkers in patients with early arthritis. J Rheumatol 2011;38: Dougados M, van der Linden S, Juhlin R, et al. The European Spondylarthropathy Study Group preliminary criteria for the classification of spondylarthropathy. Arthritis Rheum 1991;34: Rudwaleit M, van der Heijde D, Landewe R, et al. The Assessment of SpondyloArthritis International Society classification criteria for peripheral spondyloarthritis and for spondyloarthritis in general. Ann Rheum Dis 2011;70: Arnett FC, Edworthy SM, Bloch DA, et al. The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 1988;31: Aletaha D, Neogi T, Silman AJ, et al Rheumatoid arthritis classification criteria: an American College of Rheumatology/European League Against Rheumatism collaborative initiative. Ann Rheum Dis 2010;69; Aletaha D, Neogi T, Silman AJ, et al Rheumatoid arthritis classification criteria: an American College of Rheumatology/European League Against Rheumatism collaborative initiative. Arthritis Rheum 2010;62: Baeten D, Van den Bosch F, Elewaut D, et al. Needle arthroscopy of the knee with synovial biopsy sampling: technical experience in 150 patients. Clin Rheumatol 1999;18: Gerlag DM, Tak PP. How to perform and analyse synovial biopsies. Best Pract Res Clin Rheumatol 2009;23: Baeten D, Moller HJ, Delanghe J, et al. Association of CD163+ macrophages and local production of soluble CD163 with decreased lymphocyte activation in spondylarthropathy synovitis. Arthritis Rheum 2004;50: Baeten D, Kruithof E, De Rycke L, et al. Infiltration of the synovial membrane with macrophage subsets and polymorphonuclear cells reflects global disease activity in spondyloarthropathy. Arthritis Res Ther 2005;7:R359-R Kruithof E, Baeten D, De Rycke L, et al. Synovial histopathology of psoriatic arthritis, both oligoand polyarticular, resembles spondyloarthropathy more than it does rheumatoid arthritis. Arthritis Res Ther 2005;7:R569-R Bresnihan B, Pontifex E, Thurlings RM, et al. Synovial tissue sublining CD68 expression is a biomarker of therapeutic response in rheumatoid arthritis clinical trials: consistency across centers. J Rheumatol 2009;36: Bhole VM, Choi HK, Burns LC, et al. Differences in body mass index among individuals with PsA, psoriasis, RA and the general population. Rheumatology (Oxford) 2012;51: Canete JD, Santiago B, Cantaert T, et al. Ectopic lymphoid neogenesis in psoriatic arthritis. Ann Rheum Dis 2007;66:

50 40 Cantaert T, Kolln J, Timmer T, et al. B lymphocyte autoimmunity in rheumatoid synovitis is independent of ectopic lymphoid neogenesis. J Immunol 2008;181: Baeten D, Demetter P, Cuvelier C, et al. Comparative study of the synovial histology in rheumatoid arthritis, spondyloarthropathy, and osteoarthritis: influence of disease duration and activity. Ann Rheum Dis 2000;59: Noordenbos T, Yeremenko N, Gofita I, et al. Interleukin-17-positive mast cells contribute to synovial inflammation in spondylarthritis. Arthritis Rheum 2012;64: Yeremenko N, Noordenbos T, Cantaert T, et al. Disease-specific and inflammation-independent stromal alterations in spondyloarthritis synovitis. Arthritis Rheum Published Online First: 12 Sep doi: /art Tan AL, Benjamin M, Toumi H, et al. The relationship between the extensor tendon enthesis and the nail in distal interphalangeal joint disease in psoriatic arthritis--a high-resolution MRI and histological study. Rheumatology (Oxford) 2007;46: Muche B, Bollow M, Francois RJ, et al. Anatomic structures involved in early- and late-stage sacroiliitis in spondylarthritis: a detailed analysis by contrast-enhanced magnetic resonance imaging. Arthritis Rheum 2003;48: Francois RJ, Gardner DL, Degrave EJ, et al. Histopathologic evidence that sacroiliitis in ankylosing spondylitis is not merely enthesitis. Arthritis Rheum 2000;43: Laloux L, Voisin MC, Allain J, et al. Immunohistological study of entheses in spondyloarthropathies: comparison in rheumatoid arthritis and osteoarthritis. Ann Rheum Dis 2001;60: McGonagle D, Marzo-Ortega H, O Connor P, et al. Histological assessment of the early enthesitis lesion in spondyloarthropathy. Ann Rheum Dis 2002;61; Freeston JE, Coates L, Helliwell P, et al. Is there subclinical enthesitis in early psoriatic arthritis? A clinical comparison with power Doppler ultrasound. Arthritis Care Res (Hoboken) 2012;64: Feydy A, Lavie-Brion MC, Gossec L, et al. Comparative study of MRI and power Doppler ultrasonography of the heel in patients with spondyloarthritis with and without heel pain and in controls. Ann Rheum Dis 2012;71: Enthesitis is not specific for recent onset SpA 49

51

52 Undifferentiated spondyloarthritis vs ankylosing spondylitis and psoriatic arthritis: a real-life prospective cohort study of clinical presentation and response to treatment Jacqueline E. Paramarta 1, Leen De Rycke 1, Carmen A. Ambarus 1, Paul P. Tak 1,2, Dominique Baeten 1 1 Department of Clinical Immunology and Rheumatology, Academic Medical Center/University of Amsterdam, Amsterdam, The Netherlands, 2 GlaxoSmithKline, Stevenage, UK 4 Rheumatology (Oxford) Oct;52(10):1873-8

53 ABSTRACT Objective SpA is a phenotypically heterogeneous disease, with AS and PsA as its best studied subtypes. This study aimed to investigate whether, despite a different phenotypic presentation, patients with undifferentiated SpA (USpA) have similar disease activity and response to treatment to those with AS and PsA. Methods 175 patients presenting at a dedicated SpA outpatient clinic were recruited in a real-life prospective cohort with follow-up every 3 months. Clinical characteristics, disease activity at presentation and response to treatment of USpA were compared with AS and PsA. Results Twenty-three per cent (n = 40) of the patients were classified as USpA. These patients were younger and tended to have a shorter disease duration than AS and PsA patients. USpA patients exhibited a mixed axial (inflammatory back pain in 87.5%) and peripheral (peripheral arthritis in 62.5%) phenotype, with almost half of the patients having low-grade sacroiliitis on conventional X-ray. The overall disease activity in USpA was similar to AS and higher than in PsA, also when analysing only anti-tnf naive patients. Initiation of TNF blockade significantly decreased disease activity in USpA, with a similar amplitude to that in AS and PsA. Conclusion USpA is a frequent, severe and anti-tnf-responsive phenotypic subtype of SpA. In agreement with the new ASAS classification criteria for axial and peripheral SpA and emerging data on TNF blockade in non-radiographic axial SpA and peripheral uspa, these data emphasize the need for early diagnosis and optimal treatment of not only AS and PsA but also other SpA subforms. 52

54 INTRODUCTION AS and PsA are the best studied SpA subtypes, which also includes IBD-related SpA (IBD-SpA), ReA and USpA. 1,2 The latter is a poorly described subgroup that fulfils the SpA criteria but cannot be classified in one of the other subtypes. However, increasing evidence indicates that these subtypes represent different presentations of one disorder rather than related but distinct entities. 2 The classification of SpA has implications for clinical practice. Registration trials for TNF blockers were conducted only in AS and PsA and not other SpA subtypes. While IBD-SpA may receive anti-tnf for the IBD, and ReA is not that prevalent, the issue arising here is the appropriate treatment of USpA, since several observations argue that USpA is not fundamentally different from other SpA subtypes. First, a substantial proportion of USpA will evolve to fullblown AS. Sacroiliitis on radiographs can take years to develop 3,4 but the disease burden is similar in axial SpA without and with radiographic sacroiliitis. 5,6 Also, USpA can progress to other SpA subtypes such as PsA. To classify patients in an early stage, the Association of SpondyloArthritis International Society (ASAS) developed classification criteria starting from the major clinical presentation, axial or peripheral joint inflammation, rather than from prototypical phenotypes, allowing the inclusion of early and less typical cases without key phenotypic features. 7 The diagnostic potential of these criteria, however, remains to be established. Second, increasing evidence demonstrates that TNF blockade is also effective in USpA. Pivotal phase II trials with infliximab were conducted simultaneously in AS and SpA. The latter trial included not only AS and PsA but also USpA. 8 Although the subgroups were too small for wellpowered sub-analyses, the treatment response appeared similar. Subsequent clinical trials suggested good response to anti-tnf in non-as, non-psa SpA, both in patients classified as USpA phenotypically 9 and in patients with non-radiographic axial SpA 10 and peripheral SpA according to ASAS. 11 These observations suggest that USpA is an important SpA subgroup at risk of late diagnosis and suboptimal treatment. This study aimed to test this hypothesis by comparing the clinical presentation, disease activity and response to anti-tnf treatment of USpA with AS and PsA in a prospective cohort study. 4 Comparison of USpA versus AS and PsA PATIENTS AND METHODS Patient cohort Patients presenting at the SpA outpatient clinic of the Department of Clinical Immunology and Rheumatology at the Academic Medical Center/University of Amsterdam between June 2007 and March 2010 were included in this real-life prospective cohort. In the Netherlands ethics approval is not required for this observational study as the patients were not subjected to any additional procedures. Patients were older than 17 years and fulfilled the ESSG criteria. 1 Demographic and disease characteristics, HLA-B27 and X-rays of the sacroiliac joints were collected at the first visit. Sacroiliitis grade 2 bilaterally or 3 unilaterally were designated high-grade sacroiliitis, and grade <2 bilaterally or <3 unilaterally low grade. Patients fulfilling the modified New York criteria were classified as AS, 3 those fulfilling the Classification Criteria for Psoriatic Arthritis (CASPAR) as 53

55 PsA, 12 those with arthritis and IBD as IBD-SpA and those with arthritis after proven genitourinary or gastrointestinal infection as ReA. Patients fulfilling the ESSG criteria but none of these other criteria were classified as USpA. Every 3 months the patient s and physician s global assessment of disease activity, BASDAI, 13 Ankylosing Spondylitis Disease Activity Score (ASDAS), 14 66/68 swollen joint count (SJC) and tender joint count (TJC), Schober, chest expansion, ESR, CRP and medication use were measured and recorded. Patients were treated according to standard clinical patient care, hence all therapies were allowed without restrictions. Statistical analysis Clinical features of USpA were compared with AS on the one hand and PsA on the other. AS was not compared with PsA to avoid false-positive findings due to multiple testing. Other subtypes were not analysed. Independent sample t-tests were used for normally distributed parameters, Mann- Whitney U test for non-normally distributed parameters, χ 2 tests for categorical parameters and Wilcoxon matched pairs tests for assessment of treatment effects in each subgroup. Statistical tests were two-sided. P values <0.05 were considered statistically significant. RESULTS Clinical characteristics of USpA compared with AS and PsA 175 SpA patients were included, consisting of 74 AS, 45 PsA, 29 IBD-SpA and 3 ReA patients. Ten fulfilled both the AS and PsA criteria, 6 both the AS and IBD-SpA criteria. The 40 remaining patients were classified as USpA (23% of the cohort). Compared with AS, USpA patients were younger (P = 0.002) and tended to have shorter disease duration (P = 0.060) (Table 1). Albeit a lower HLA-B27 prevalence in USpA (42.5%) than AS (68.1%) (P = 0.008), 50.0% had a positive SpA family history. USpA patients less frequently had inflammatory back pain (IBP) (P = 0.038) but more frequently had arthritis (P = 0.005) and enthesitis (P = 0.003) than AS patients. X-rays in USpA revealed sacroiliitis in 47.5% but, in contrast to AS, this was almost exclusively low grade (P<0.001). One patient had high-grade sacroiliitis without IBP or limitation in spinal or chest mobility, but with peripheral arthritis, and was diagnosed with USpA and not AS. Compared with PsA, USpA patients were younger (P<0.001), less frequently male (P = 0.025) and had a younger age at onset (P = 0.001). USpA displayed more frequently signs of axial disease such as IBP (P<0.001), buttock pain (P = 0.003) and low-grade sacroiliitis (P<0.001) than PsA. As 10 PsA patients also fulfilled the modified New York criteria, 3 high-grade sacroiliitis was more prevalent in PsA than USpA (P = 0.007). Peripheral arthritis (P = 0.041) and dactylitis (P = 0.063) were more common in PsA than USpA. We also assessed how USpA would be classified according to the ASAS criteria; 7 however, since MRI was not performed, post hoc application of these criteria was hampered. Twenty-five per cent fulfilled the axial SpA and another 25% the peripheral SpA criteria. This suggests that a majority of USpA patients would also fulfil the ASAS criteria, although the exact overlap remains unknown. 54

56 Table 1. Demographic and clinical characteristics, disease activity and medication use at inclusion USpA (n=40) AS (n=74) PsA (n=45) Age, mean (SD) years 38.9 (11.9) 46.8 (13.6)* 50.7 (10.0)* Age at disease onset, mean (SD) years 33.5 (11.7) 37.7 (15.4) 42.3 (11.7)* Disease duration, mean (SD) years 5.4 (8.4) 9.2 (10.9) 8.3 (9.1) Male gender, % * BMI, mean (SD) kg/m (4.0) 26.0 (5.7) 26.5 (5.9) HLA-B27 positive, % * 35.5 Positive SpA family history, % * 35.6 Inflammatory back pain (history/presence), % * 48.9* Buttock pain (history/presence), % * Sacroiliitis low grade, % * 11.1* Sacroiliitis high grade, % * 22.2* Peripheral arthritis (history/presence), % * 82.2* Enthesitis (history/presence), % * 44.4 Dactylitis (history/presence), % Uveitis (history/presence), % Urethritis/diarrhea (history/presence), % Patient s global assessment, median (IQR) mm VAS 63 (48-76) 50 (25-72)* 38 (15-63)* Physician s global assessment, median (IQR) mm VAS 51 (33-63) 45 (26-63) 30 (15-44)* BASDAI, median (IQR) 5.3 ( ) 5.0 ( ) 3.2 ( )* BASDAI 4, % * ASDAS, median (IQR) 2.7 ( ) 2.9 ( ) 1.9 ( )* ASDAS 2.1, % * Axial nocturnal pain, % * Schober, median (IQR) centimetres 4.0 ( ) 3.5 ( )* 4.0 ( ) Chest expansion, median (IQR) centimetres 4.3 ( ) 4.0 ( ) 5.0 ( ) Swollen joint count, median (IQR) 0-66 joints 0 (0-2) 0 (0-0)* 0 (0-2) Tender joint count, median (IQR) 0-68 joints 1 (0-4) 0 (0-2) 3 (0-6) Peripheral nocturnal pain, % CRP, median (IQR) mg/l 2.0 ( ) 4.0 ( ) 2.2 ( ) ESR, median (IQR) mm/h 7 (2-15) 11 (5-24) 5 (3-19) 4 Comparison of USpA versus AS and PsA NSAIDs, % * Corticosteroids, % Methotrexate, % * Sulphasalazine, % * 4.4* Other DMARDs, % Anti-TNF therapy, % * 40.0* BMI = body mass index; VAS = visual analogue scale; IQR = interquartile range; DMARDs = diseasemodifying anti-rheumatic drugs (the other abbreviations are explained in the text). Positive SpA family history includes the presence of 1 of the following: AS, psoriasis, IBD, ReA or uveitis, in first- or seconddegree relatives. Significance of the comparisons is determined by independent sample t-tests or Mann- Whitney U tests for continuous variables and chi-square tests for categorical variables. Trend (0.05 P<0.1) and *significant difference (P<0.05) versus USpA. 55

57 Disease activity of USpA compared with AS and PsA at inclusion Most disease activity parameters were similar between USpA and AS with the exception of a higher patient s global assessment (P = 0.026) and SJC66 (P = 0.001) in USpA (Table 1). ESR tended to be lower in USpA than AS (P = 0.052) and the Schober less limited (P = 0.009). Compared with PsA, USpA had a higher patient s global assessment (P = 0.002), physician s global assessment (P = 0.001), BASDAI (P = 0.013) and ASDAS (P = 0.013), and more frequent nocturnal axial pain (P = 0.002). There were no differences when axial and peripheral USpA were compared with each other (data not shown). Influence of treatment on disease activity at inclusion The use of SSZ was higher (P<0.001) and TNF blockers lower (P = 0.047) in USpA than AS. Compared with PsA, NSAID usage was higher (P = 0.020) and anti-tnf usage lower (P = 0.002) in USpA (Table 1). To assess whether the lower use of anti-tnf in USpA vs AS and PsA influenced the disease activity, we performed a sub-analysis in patients naive to anti-tnf (supplementary Table S1, available at Rheumatology Online). The disease activity was similar in anti-tnf naive USpA and AS with the exception of a higher Schober (P = 0.016) and SJC66 (P = 0.007) in USpA. Compared with anti-tnf naive PsA, patient s global assessment (P = 0.015) and axial nocturnal pain (P = 0.014) was higher in USpA, with a similar trend for physician s global assessment (P = 0.054) and ASDAS (P = 0.083). The TJC68 was lower in anti-tnf naive USpA compared with PsA (P = 0.045). Anti-TNF therapy in USpA The initiation and efficacy of TNF blockade in patients with a follow-up of 24 weeks, corresponding to 19 USpA, 53 AS and 38 PsA patients, was assessed prospectively. A first anti-tnf was initiated in 8 (42.1%) USpA, 22 (41.5%) AS and 10 (26.3%) PsA patients. The disease activity in USpA starting anti-tnf was high and comparable to AS and PsA (Fig. 1), with a similar CRP and ESR (data not shown). Initiation of anti-tnf seemed to reduce disease activity also in USpA as shown in Fig. 1. Additionally, the median SJC66 decreased from 2 (IQR 0-2) to 0 (IQR 0-2) and ESR from 7 (IQR 2-17) to 3 (IQR 2-12) mm/h. Although mostly not statistically significant because of the small number of patients included here, the response seemed comparable in USpA, AS and PsA. DISCUSSION This prospective cohort study assessed the similarities and differences in patient characteristics, disease activity and response to anti-tnf treatment between USpA and the better characterized SpA subgroups AS and PsA in a real-life outpatient setting. The major findings are that (I) USpA is a common SpA subtype that presents at younger age than AS and PsA, (II) USpA exhibits a mixed axial and peripheral phenotype, frequently with low-grade sacroiliitis, (III) USpA depicts a disease activity similar to AS and higher than PsA and (IV) USpA appears to respond as well to anti-tnf as AS and PsA. That USpA is a common SpA subtype (23%) is in agreement with previous reports. 1,4 The annual estimated incidence of 46.5 USpA patients per and incidence of SpA of 62.5 per in a Spanish registry, 15 together with observations that the prevalence of SpA 56

58 4 Figure 1. Changes in disease activity parameters during treatment with anti-tnf therapy. The panel represents the patient s and physician s global assessment of disease activity on a 100 mm visual analogue scale (VAS), the Bath Ankylosing Spondylitis Disease Activity Index (BASDAI), and Ankylosing Spondylitis Disease Activity Score (ASDAS) in undifferentiated spondyloarthritis (USpA), ankylosing spondylitis (AS) and psoriatic arthritis (PsA). Data are presented as the median (interquartile range). * P<0.05 by Wilcoxon matched pairs test, week 24 compared to week 0 by subtype. Comparison of USpA versus AS and PsA might be similar to that of RA, 16 indicate that USpA is a prevalent condition. Demographically USpA differed slightly from AS and PsA: patients were younger, had shorter disease duration and younger age of onset, and were more frequently female, which corresponds to previous studies. 4-6,17-20 HLA-B27 prevalence (42%) was slightly lower than previously described (57-86%). 4-6,17-20 However, in AS HLA-B27 positivity was also lower than commonly described, probably because of our medium-sized cohort, and would be higher if more patients were included. Interestingly, although HLA-B27 prevalence in USpA was not high, 50% had a positive SpA family history, consistent with observations that SpA subtypes segregate together in families. 2 USpA demonstrated a mixed axial and peripheral phenotype, intermediate between the pronounced axial AS and the marked peripheral PsA. In USpA almost 90% had IBP and almost 50% low-grade sacroiliitis on X-ray, which concords with other cohorts. 4,17 Since USpA had a young age and short disease duration and since low-grade sacroiliitis is a risk factor for development of AS, 17,19 this suggests that an important fraction of USpA may evolve to overt AS over time. USpA patients had a high disease activity, comparable to AS and higher than PsA, that was independent of phenotypical presentation and still valid when assessing exclusively 57

59 anti-tnf naive patients. Previous studies also reported similar disease activity between SpA subgroups. 4-6 Despite this high disease activity, a lower proportion of USpA received anti-tnf compared with AS and PsA at inclusion. While this reflects the anti-tnf approval and reimbursement regulations in different SpA subtypes, it also indicates that USpA patients are at risk of suboptimal treatment since TNF blockade seems equally effective in all SpA subtypes, 8-11 which is supported by the evaluation of anti-tnf initiation in USpA in the current cohort. This study gives a good reflection of the real-life situation but has limitations. First, the population in an academic hospital may be different from that of general rheumatology clinics. Second, the heterogeneity of USpA makes it difficult to define the most relevant disease activity parameter. AS parameters such as BASDAI and ASDAS may underestimate peripheral disease activity, whereas SJC and TJC as used in PsA will miss axial disease. Third, our observational open design and limited patient number preclude strong conclusions regarding response to anti-tnf. However, it indicates that randomized controlled trials with anti-tnf or novel therapies should be conducted not only in AS and PsA but also in other SpA subtypes. Finally, the ESSG criteria 1 were used and not the ASAS criteria 7 as these were not yet established when the study was started. Despite the limitations, however, the data from this real-life prospective cohort concord with the concept that non-as, non-psa patients represent a relevant part of SpA. Early recognition, adequate assessment and monitoring of disease activity, optimal use of existing therapies and studies of novel drugs should thus not be limited to AS and PsA but be extended to all SpA forms. RHEUMATOLOGY KEY MESSAGES -- USpA is a frequent subtype of SpA with a mixed peripheral and axial phenotype -- Disease activity in USpA is not lower than in AS and PsA -- Clinical trials are also needed in SpA subtypes other than AS and PsA Acknowledgements We would like to thank the following physicians who worked on the SpA outpatient clinic during our study period: Marc Kok, Radjesh Bisoendial, Ruth Klaassen, Lot Burgemeister, Liesbeth Hak, Frieda Koopman and Ivy Choi. Prof. Dr D. Baeten was supported by a VIDI grant from The Netherlands Organization for Scientific Research (NWO) and by a grant from the Dutch Arthritis Foundation (Reumafonds). REFERENCES 1 Dougados M, van der Linden S, Juhlin R et al. The European Spondylarthropathy Study Group preliminary criteria for the classification of spondylarthropathy. Arthritis Rheum 1991;34: Baeten D, Breban M, Lories R et al. Are spondylarthritides related but distinct conditions or a single disease with a heterogeneous phenotype? Arthritis Rheum 2013;65: Van der Linden S, Valkenburg HA, Cats A. Evaluation of diagnostic criteria for ankylosing spondylitis. A proposal for modification of the New York criteria. Arthritis Rheum 1984;27: Rojas-Vargas M, Munoz-Gomariz E, Escudero A et al. First signs and symptoms of spondyloarthritis data from an inception cohort with a disease course of two years or less (REGISPONSER-Early). Rheumatology 2009;48:

60 5 Rudwaleit M, Haibel H, Baraliakos X et al. The early disease stage in axial spondylarthritis: results from the German Spondyloarthritis Inception Cohort. Arthritis Rheum 2009;60: Kiltz U, Baraliakos X, Karakostas P et al. Patients with non-radiographic axial spondyloarthritis differ from patients with ankylosing spondylitis in several aspects. Arthritis Care Res 2012;64: Rudwaleit M, van der Heijde D, Landewe R et al. The Assessment of SpondyloArthritis International Society classification criteria for peripheral spondyloarthritis and for spondyloarthritis in general. Ann Rheum Dis 2011;70: Van Den Bosch F, Kruithof E, Baeten D et al. Randomized double-blind comparison of chimeric monoclonal antibody to tumor necrosis factor alpha (infliximab) versus placebo in active spondylarthropathy. Arthritis Rheum 2002;46: Paramarta JE, De Rycke L, Heijda TF et al. Efficacy and safety of adalimumab for the treatment of peripheral arthritis in spondyloarthritis patients without ankylosing spondylitis or psoriatic arthritis. Ann Rheum Dis 2012;doi: / annrheumdis Sieper J, van der Heijde D, Dougados M et al. Efficacy and safety of adalimumab in patients with non-radiographic axial spondyloarthritis: results of a randomized placebo-controlled trial (ABILITY-1). Ann Rheum Dis 2013;72: Mease P, Sieper J, Van Den Bosch F et al. Efficacy and safety of adalimumab in patients with peripheral spondyloarthritis: results from a phase 3 study [abstract]. Ann Rheum Dis 2012;71: Taylor WJ, Helliwell PS. Development of diagnostic criteria for psoriatic arthritis: methods and process. Curr Rheumatol Rep 2004;6: Garrett S, Jenkinson T, Kennedy LG et al. A new approach to defining disease status in ankylosing spondylitis: the Bath Ankylosing Spondylitis Disease Activity Index. J Rheumatol 1994;21: Machado P, Landewe R, Lie E et al. Ankylosing Spondylitis Disease Activity Score (ASDAS): defining cut-off values for disease activity states and improvement scores. Ann Rheum Dis 2011;70: Munoz-Fernandez S, de Miguel E, Cobo-Ibanez T et al. Early spondyloarthritis: results from the pilot registry ESPIDEP. Clin Exp Rheumatol 2010;28: Reveille JD, Witter JP, Weisman MH. Prevalence of axial spondylarthritis in the United States: estimates from a cross-sectional survey. Arthritis Care Res 2012;64: Huerta-Sil G, Casasola-Vargas JC, Londono JD et al. Low grade radiographic sacroiliitis as prognostic factor in patients with undifferentiated spondyloarthritis fulfilling diagnostic criteria for ankylosing spondylitis throughout follow up. Ann Rheum Dis 2006;65: Heuft-Dorenbosch L, Landewe R, Weijers R et al. Performance of various criteria sets in patients with inflammatory back pain of short duration; the Maastricht early spondyloarthritis clinic. Ann Rheum Dis 2007;66: Sampaio-Barros PD, Bortoluzzo AB, Conde RA et al. Undifferentiated spondyloarthritis: a longterm followup. J Rheumatol 2010;37: Dougados M, d Agostino MA, Benessiano J et al. The DESIR cohort: a 10-year follow-up of early inflammatory back pain in France: study design and baseline characteristics of the 708 recruited patients. Joint Bone Spine 2011;78: Comparison of USpA versus AS and PsA 59

61 SUPPLEMENTARY TABLE Supplementary Table 1. Disease activity at first visit in anti-tnf naive patients USpA (n=36) AS (n=55) PsA (n=27) Patient s global assessment, median (IQR) mm VAS 64 (48-78) 56 (26-75) 38 (18-65)* Physician s global assessment, median (IQR) mm VAS 50 (33-63) 53 (30-70) 36 (24-51) BASDAI, median (IQR) 5.4 ( ) 5.2 ( ) 3.8 ( ) BASDAI 4, % ASDAS, median (IQR) 2.8 ( ) 3.0 ( ) 2.2 ( ) ASDAS 2.1, % Axial nocturnal pain, % * Schober, median (IQR) centimetres 4.0 ( ) 3.5 ( )* 4.0 ( ) Chest expansion, median (IQR) centimetres 4.5 ( ) 4.0 ( ) 4.0 ( ) Swollen joint count, median (IQR) 0-66 joints 0 (0-2) 0 (0-0)* 0 (0-3) Tender joint count, median (IQR) 0-68 joints 1 (0-4) 0 (0-2) 4 (0-7)* Peripheral nocturnal pain, % CRP, median (IQR) mg/l 2.9 ( ) 4.0 ( ) 3.2 ( ) ESR, median (IQR) mm/h 9 (2-16) 11 (5-24) 9 (4-20) USpA = undifferentiated spondyloarthritis; AS = ankylosing spondylitis; PsA = psoriatic arthritis; VAS = visual analogue scale; BASDAI = Bath Ankylosing Spondylitis Disease Activity Index; ASDAS = Ankylosing Spondylitis Disease Activity Score; CRP = C-reactive protein; ESR = erythrocyte sedimentation rate; IQR = interquartile range. Significance of the comparisons is determined by Mann-Whitney U tests for continuous variables or chi-square tests for categorical variables. Trend (0.05 P<0.1) and * significant difference (P<0.05) versus USpA. 60

62

63

64 Prevalence and burden of peripheral disease in spondyloarthritis: proposal for modification of the Assessment of SpondyloArthritis international Society (ASAS) classification criteria Jacqueline E. Paramarta 1, Maureen C. Turina 1, Maud van de Schoot 2, Carmen A. Ambarus 1, Johannes W. Bijlsma 1,2, Leen De Rycke 1,2, Dominique Baeten Department of Clinical Immunology and Rheumatology, Academic Medical Center/University of Amsterdam, Amsterdam, The Netherlands, 2 Department of Rheumatology and Clinical Immunology, University Medical Center Utrecht, Utrecht, The Netherlands Submitted for publication

65 ABSTRACT Objectives To assess the presence and activity of peripheral disease in axial and peripheral spondyloarthritis (SpA) as defined by the Assessment of SpondyloArthritis international Society (ASAS) criteria. Methods 389 patients attending dedicated SpA outpatient clinics, having the clinical diagnosis of SpA and fulfilling the European Spondyloarthropathy Study Group (ESSG) criteria were included in a real life observational cohort. Clinical characteristics and disease activity were compared between the different subgroups defined by the ASAS criteria. Results 75 out of 389 patients did not fulfil the ASAS criteria. Although less typical in clinical presentation (shorter disease duration, more female, HLA-B27 negative, less sacroiliitis), they had similar disease activity as those fulfilling the ASAS criteria. Of the latter group, 230 fulfilled the axial and 84 the peripheral SpA criteria. Within axial SpA, however, 49% had pure axial disease whereas 51% had combined disease (back pain plus arthritis, enthesitis and/or dactylitis). The latter group had the highest disease activity. Of the 75 patients which initially did not fulfil the ASAS criteria 65% (with combined axial and peripheral manifestations) could be classified as SpA if they could enter the criteria via the peripheral arm. Conclusions Peripheral disease is a prevalent manifestation of SpA, even in axial SpA. Patients with combined axial and peripheral disease have the highest disease activity, but not all these patients fulfil the ASAS criteria. These data support a classification into axial, combined, or peripheral disease, where the combined group could enter either the axial or peripheral arm of the SpA criteria. 64

66 INTRODUCTION Spondyloarthritis (SpA) is a disease with a heterogeneous clinical presentation and has therefore traditionally been divided into different phenotypic subtypes: ankylosing spondylitis (AS), psoriatic arthritis (PsA), inflammatory bowel disease (IBD)-related spondyloarthritis, reactive arthritis and undifferentiated spondyloarthritis. 1,2 The Assessment of SpondyloArthritis international Society (ASAS) recently developed classification criteria which distinguish axial and peripheral SpA according to the disease manifestations of the patient. 3,4 This seems more accurate than the subdivision into the different phenotypic subtypes as data from genetic studies, 5-8 animal models, 9,10 and immunopathology suggest that these different subtypes belong to a single disease with a heterogeneous phenotype rather than distinct disease entities. 15 Moreover, the clinical classification into axial and peripheral disease is supported by clues that axial and peripheral SpA might be driven by slightly different mechanisms, including different cell types expressing interleukin (IL)-17, 13,16 the distinct value of matrix metalloproteinase 3 as biomarker, 17 and the differential response to sulfasalazine in peripheral versus axial disease. 18 One issue with the classification into axial and peripheral SpA, however, is that approximately 30% of the SpA patients has both axial and peripheral involvement. 4 If patients have peripheral disease (arthritis, enthesitis, and/or dactylitis) as well as active back pain, the ASAS criteria prespecify that the axial SpA criteria 3 should be applied and not the peripheral SpA criteria. 4 This decision, which is based on consensus rather than on evidence, 4 may potentially lead to an underestimation of the prevalence and disease burden of peripheral disease in SpA. Therefore, this study aimed to systematically assess the presence and activity of peripheral disease in axial and peripheral SpA as defined by the ASAS criteria in a large real life observational cohort. 5 Burden of peripheral disease METHODS Patient cohort Patients presenting on the specialized SpA outpatient clinic of the Department of Clinical Immunology and Rheumatology at the Academic Medical Center/University of Amsterdam between June 2007 and August 2012 (n=346) and of the University Medical Center Utrecht between January 2011 and August 2012 (n=43) were included in this real life prospective observational cohort, as approved by the local Medical Ethics Committees. Patients were older than 17 years, had the clinical diagnosis of SpA according to the expert opinion of the treating rheumatologist and fulfilled the European Spondyloarthropathy Study Group (ESSG) criteria. 1 Demographic and disease characteristics, HLA-B27 and X-rays of the sacroiliac joints were collected at the first visit. Sacroiliitis grade 2 bilaterally or 3 unilaterally were designated high grade sacroiliitis, and sacroiliitis grade <2 bilaterally or <3 unilaterally low grade. The patient s and physician s global assessment of disease activity, Bath Ankylosing Spondylitis Disease Activity (BASDAI), 19 Ankylosing Spondylitis Disease Activity Score (ASDAS) based on the C-reactive protein (CRP), 20,21 66/68 swollen joint count (SJC) and tender joint count (TJC), Schober, chest expansion, CRP, erythrocyte sedimentation rate (ESR), and medication use were measured and recorded every 3 months. Active inflammatory back pain (IBP) was defined as pain at night at least 1 time a week and/or an average morning stiffness of at least 30 65

67 minutes in the past week. Patients were treated according to standard clinical patient care, hence all therapies were allowed without restrictions for this observational study. For the current study only data of the patient s first visit to the SpA outpatient clinic were used. Statistical analysis This study aimed to compare the clinical characteristics and disease activity between SpA patients a) who do versus do not fulfil the ASAS criteria, b) who fulfil the ASAS axial versus the ASAS peripheral criteria, and c) who fulfil the ASAS criteria with combined versus pure axial or pure peripheral symptoms. The different groups were compared with Mann-Whitney U tests for continuous variables and chi-square tests for categorical variables. Data are presented as the median and interquartile range (IQR), or percentages of patients. Statistical tests were 2-sided and P values less than 0.05 were considered statistically significant. RESULTS Disease characteristics in SpA patients fulfilling the ESSG but not the ASAS classification criteria From our total population of 389 patients diagnosed with SpA by the rheumatologist s expert opinion and fulfilling the ESSG criteria, 314 (80.7%) patients also fulfilled either the ASAS axial or peripheral SpA criteria (Figure 1). As shown in Table 1, patients not fulfilling the ASAS criteria had a shorter disease duration (1.6 ( ) versus 4.0 ( ) years, p=0.005), later disease onset (38.2 ( ) versus 34.1 ( ) years, p=0.023), and were more frequently female (64% versus 38.2%, p<0.001) and HLA-B27 negative (97.1% versus 34.9%, p<0.001) than those who did fulfil the ASAS criteria. Although the occurrence of IBP was similar in both groups, the patients not fulfilling the ASAS criteria had less often high grade sacroiliitis (1.4% versus 53.4%, p<0.001) and more often low grade sacroiliitis on X-ray (32.9% versus 13.5%, p<0.001) than those who did fulfil the criteria. In contrast, peripheral arthritis (66.7% versus 50.0%, p=0.009) and enthesitis (50.7% versus 36.0%, p=0.019) occurred more frequently in the group not fulfilling the ASAS criteria. These patients had also more frequently associated IBD (26.7% versus 8.9%, p<0.001) and, in line with HLA-B27 positivity, less frequently uveitis (9.3% versus 20.4%, p=0.026). As to disease activity, most parameters were similar between the two groups or were even higher in the patients not fulfilling the ASAS criteria, as illustrated by BASDAI (5.3 ( ) versus 4.6 ( ), p=0.003), TJC68 (3.0 ( ) versus 0.0 ( ), p<0.001), and percentage of patients with a high or very high disease activity according to ASDAS-CRP (77.8% versus 66.8%, p=0.092) (Table 1). With the exception of methotrexate use, there was no difference in medication between both groups and the differences in BASDAI and TJC68 remained significant when excluding patients on anti-tnf therapy (data not shown). Comparison of axial versus peripheral SpA according to the ASAS criteria Within the group of patients fulfilling the ASAS criteria (n=314), 230 (73.2%) patients fulfilled the ASAS axial SpA criteria and 84 (26.8%) patients fulfilled the peripheral SpA criteria 4 (Figure 1). 66

68 Figure 1. Distribution of the different spondyloarthritis (SpA) subgroups by criteria. All the patients were diagnosed with SpA according to the rheumatologist s expert opinion and fulfilled the European Spondyloarthropathy Study Group (ESSG) criteria. The patients were then subdivided into whether they did or did not fulfil the Assessment of SpondyloArthritis International Society (ASAS) criteria, and subsequently into either the ASAS axial SpA or ASAS peripheral SpA criteria. Finally, the ASAS axial SpA group was subdivided into patients with either pure axial symptoms or combined axial and peripheral symptoms. The ASAS axial SpA group was characterized by earlier disease onset (32.3 ( ) versus 38.2 ( ) years, p=0.001) and higher HLA-B27 positivity (76.2% versus 27.3%, p<0.001) than ASAS peripheral SpA (Table 2). As expected, IBP and high grade sacroiliitis were more frequent in the axial SpA (p<0.001 for all comparisons) whereas arthritis (p<0.001), enthesitis (p<0.001) and dactylitis (p=0.009) were more frequent in peripheral SpA. In line with the frequency of HLA-B27, uveitis was also more prevalent in axial SpA (P<0.001). In contrast, psoriasis (p<0.001) and IBD (p=0.004) were more prevalent in peripheral SpA. A positive family history was observed in 1/3 to 1/4 of the patients for both axial and peripheral SpA (Table 2). As expected, disease activity measurements addressing specifically axial versus peripheral disease were higher in the ASAS axial SpA and ASAS peripheral SpA group, respectively, whereas CRP and ESR levels as markers of systemic inflammation were similar between both groups (Table 2). Interestingly, not only composite parameters originally developed for axial disease such as BASDAI and ASDAS-CRP, but also parameters of global disease activity (patient s and physician s global assessment) were significantly higher in axial than peripheral SpA (p<0.001 for all 4 comparisons). This was still the case upon analysis of patients naive to anti-tnf therapy only (data not shown). As a reflection of standard clinical care, non-steroidal anti-inflammatory drug (NSAID) usage was higher and corticosteroid and DMARD usage lower in the axial SpA group (all p<0.05). The prescription of TNF blocking therapies was equal between the two groups (Table 2). 5 Burden of peripheral disease Prevalence and disease burden of SpA patients with combined axial and peripheral disease manifestations In agreement with the fact that the ASAS criteria exclude patients with active axial symptoms from the peripheral SpA group but do not exclude patients with active peripheral symptoms 67

69 Table 1. Clinical characteristics, disease activity and treatment of SpA patients who do and do not fulfil the ASAS criteria ASAS (n=314) no ASAS (n=75) P-value Age, median (IQR) years 42.8 ( ) 43.4 ( ) Disease duration, median (IQR) years 4.0 ( ) 1.6 ( ) Age at disease onset, median (IQR) years 34.1 ( ) 38.2 ( ) Male gender, % <0.001 HLA-B27 positive, % <0.001 Inflammatory back pain (history/presence), % Buttock pain (history/presence), % Sacroiliitis low grade, % <0.001 Sacroiliitis high grade, % <0.001 Peripheral arthritis (history/presence), % Enthesitis (history/presence), % Dactylitis (history/presence), % Uveitis (history/presence), % Psoriasis (history/presence), % IBD (history/presence), % <0.001 Urethritis/diarrhea (history/presence), % Positive SpA family history, % Patient s global assessment, median (IQR) mm 52.0 ( ) 59.0 ( ) Physician s global assessment, median (IQR) mm 43.5 ( ) 44.0 ( ) BASDAI, median (IQR) 4.6 ( ) 5.3 ( ) BASDAI 4, % BASDAI #2 back pain, median (IQR) mm 52.0 ( ) 63.0 ( ) ASDAS-CRP, median (IQR) 2.7 ( ) 2.8 ( ) ASDAS-CRP 2.1, % Swollen joint count, median (IQR) 0-66 joints 0.0 ( ) 0.0 ( ) Tender joint count, median (IQR) 0-68 joints 0.0 ( ) 3.0 ( ) <0.001 Schober, median (IQR) centimetres 4.0 ( ) 4.0 ( ) Chest expansion, median (IQR) centimetres 5.0 ( ) 4.0 ( ) CRP, median (IQR) mg/l 3.7 ( ) 2.0 ( ) ESR, median (IQR) mm/h 9.0 ( ) 9.0 ( ) NSAIDs, % Corticosteroids, % Methotrexate, % Sulfasalazine, % Other DMARDs, % Anti-TNF therapy, % SpA = spondyloarthritis; ASAS = Assessment of SpondyloArthritis international Society; IQR = interquartile range; Sacroiliitis low grade = grade <2 bilateral or grade <3 unilateral, and high grade = grade 2 bilateral or grade 3 unilateral on X-ray; IBD = inflammatory bowel disease; Positive SpA family history = 1 first- or seconddegree relative with ankylosing spondylitis, psoriasis, IBD, reactive arthritis or uveitis; BASDAI = Bath Ankylosing Spondylitis Disease Activity Index; ASDAS = Ankylosing Spondylitis Disease Activity Score; CRP = C-reactive protein; ESR = erythrocyte sedimentation rate; NSAIDs = non-steroidal anti-inflammatory drugs; DMARDs = disease-modifying anti-rheumatic drugs; TNF = tumor necrosis factor. Significance of the comparisons is determined by Mann-Whitney U tests for continuous variables or chi-square tests for categorical variables.

70 Table 2. Clinical characteristics, disease activity and treatment of ASAS axial SpA versus ASAS peripheral SpA ASAS axial SpA (n=230) ASAS peripheral SpA (n=84) P-value Age, median (IQR) years 40.8 ( ) 47.9 ( ) Disease duration, median (IQR) years 3.9 ( ) 4.2 ( ) Age at disease onset, median (IQR) years 32.3 ( ) 38.2 ( ) Male gender, % HLA-B27 positive, % <0.001 Inflammatory back pain (history/presence), % <0.001 Buttock pain (history/presence), % <0.001 Sacroiliitis low grade, % Sacroiliitis high grade, % <0.001 Peripheral arthritis (history/presence), % <0.001 Enthesitis (history/presence), % Dactylitis (history/presence), % Uveitis (history/presence), % <0.001 Psoriasis (history/presence), % <0.001 IBD (history/presence), % Urethritis/diarrhea (history/presence), % Positive SpA family history, % Patient s global assessment, median (IQR) mm 55.0 ( ) 38.0 ( ) Physician s global assessment, median (IQR) mm 48.0 ( ) 30.5 ( ) <0.001 BASDAI, median (IQR) 4.9 ( ) 3.0 ( ) <0.001 BASDAI 4, % <0.001 BASDAI #2 back pain, median (IQR) mm 60.0 ( ) 11.5 ( ) <0.001 ASDAS-CRP, median (IQR) 2.8 ( ) 2.0 ( ) <0.001 ASDAS-CRP 2.1, % <0.001 Swollen joint count, median (IQR) 0-66 joints 0.0 ( ) 1.0 ( ) <0.001 Tender joint count, median (IQR) 0-68 joints 0.0 ( ) 1.0 ( ) <0.001 Schober, median (IQR) centimetres 3.5 ( ) 4.5 ( ) <0.001 Chest expansion, median (IQR) centimetres 5.0 ( ) 5.0 ( ) CRP, median (IQR) mg/l 3.6 ( ) 3.7 ( ) ESR, median (IQR) mm/h 9.0 ( ) 7.5 ( ) Burden of peripheral disease NSAIDs, % Corticosteroids, % Methotrexate, % <0.001 Sulfasalazine, % <0.001 Other DMARDs, % Anti-TNF therapy, % SpA = spondyloarthritis; ASAS = Assessment of SpondyloArthritis international Society; IQR = interquartile range; Sacroiliitis low grade = grade <2 bilateral or grade <3 unilateral, and high grade = grade 2 bilateral or grade 3 unilateral on X-ray; IBD = inflammatory bowel disease; Positive SpA family history = 1 first- or second-degree relative with ankylosing spondylitis, psoriasis, IBD, reactive arthritis or uveitis; BASDAI = Bath Ankylosing Spondylitis Disease Activity Index; ASDAS = Ankylosing Spondylitis Disease Activity Score; CRP = C-reactive protein; ESR = erythrocyte sedimentation rate; NSAIDs = non-steroidal anti-inflammatory drugs; DMARDs = disease-modifying anti-rheumatic drugs; TNF = tumor necrosis factor. Significance of the comparisons is determined by Mann- Whitney U tests for continuous variables or chi-square tests for categorical variables.

71 from the axial SpA group, we noted that peripheral arthritis and enthesitis each occurred in more than 30% of the patients with axial SpA. Therefore, we additionally subdivided patients who fulfilled the ASAS axial SpA criteria into pure axial (n=112 or 48.7%) and combined disease (back pain plus arthritis, enthesitis or dactylitis) (n=118 or 51.3%) (Figure 1). The clinical characteristics such as age at disease onset, gender, HLA-B27, axial and peripheral symptoms, and extra-articular manifestations tended to be intermediate for the combined SpA group in comparison with pure axial and pure peripheral disease (Table 3). Importantly, the combined group had the highest disease activity since the patient s and physician s global assessment as well as the BASDAI and ASDAS-CRP was higher than in the other two groups (all p<0.001) (Table 3). There was no significant difference in CRP and ESR levels. The group with combined axial and peripheral disease retained the highest disease activity after exclusion of patients treated with anti-tnf therapy (data not shown). Proposal for modification of the ASAS classification criteria The prevalence and contribution to increased disease burden of combined peripheral disease in axial SpA and the broad overlap between pure axial disease, combined disease and peripheral disease does not support the exclusive use of the axial SpA criteria in patients with both active axial and peripheral symptoms. We therefore reanalyzed our cohort using the ASAS classification criteria but allowing patients with combined disease to enter the classification through both arms rather than through exclusively the axial SpA arm (Figure 2). Of the 389 patients with SpA diagnosed by the rheumatologist and fulfilling the ESSG criteria, now 363 (93.3%) instead of 314 (80.7%) also fulfilled the ASAS criteria. This means that 49 of the 75 patients (65.3%) not fulfilling the strict ASAS criteria (meaning that patients with combined active axial and peripheral symptoms can only enter the axial arm) could now be classified as SpA by using the peripheral arm of the criteria. Now 133 (34.2%) patients fulfilled these modified ASAS peripheral Figure 2. Distribution of the different spondyloarthritis (SpA) subgroups by criteria when the Assessment of SpondyloArthritis International Society (ASAS) criteria are applied with a small modification: patients with both peripheral manifestations and active axial complaints were now also allowed to enter the peripheral SpA arm instead of the axial SpA arm only. 70

72 SpA criteria, which can be subdivided in 84 (63.2%) patients with pure peripheral symptoms and 49 (36.8%) patients with combined active axial and peripheral manifestations (Figure 2). DISCUSSION The clinical characteristics and disease activity of the different SpA subpopulations according to the ASAS criteria were investigated in SpA patients (who were both diagnosed by the rheumatologist s expert opinion and fulfilled the ESSG criteria) in a real life observational cohort study of nearly 400 patients, with special focus on the prevalence and disease burden of peripheral disease. The most important finding was that the cohort of patients classified as axial SpA according to the ASAS criteria in fact consisted of two separate groups of equal size: patients with exclusive axial disease and patients with combined axial and peripheral disease, defined as active IBP plus arthritis, enthesitis and/or dactylitis. Importantly, the latter subgroup showed higher disease activity. If applying the ASAS criteria strictly (this is: whenever a patient reports active axial symptoms, even if minor in comparison with pronounced peripheral disease, he should enter the axial arm of the criteria), our analysis would reveal that approximately 2/3 of patients has axial SpA versus 1/3 has peripheral SpA, and that those with axial SpA have higher disease activity. However, in depth analysis shows that 1/3 of patients has pure axial disease, 1/3 has pure peripheral disease, and 1/3 has combined axial and peripheral disease. The latter has the highest disease activity. Both prevalence and disease burden of peripheral disease appears thus to be higher than reflected by analysis according to the ASAS criteria. A second and partially related finding was that a substantial subgroup of patients fulfilling the ESSG criteria and fitting the clinical picture of SpA according to the treating rheumatologist failed to be classified by the ASAS criteria. It is obviously not surprising that criteria do not capture all patients as they always have to aim for a good balance between sensitivity and specificity and can thus not reach optimal sensitivity without jeopardizing specificity. Nevertheless, it was striking that many of these patients with predominantly peripheral disease had also active axial symptoms, and if these patients were allowed to enter the SpA classification through the peripheral arm of the criteria, 65% of these patients would also fulfil these modified ASAS criteria. Since the rheumatologist s expert opinion is still regarded as the golden standard this is of relevance This is also illustrated by a recent study which compared the prevalence of axial SpA in at-risk patients with chronic back pain according to the axial SpA criteria to the diagnosis of axial SpA by the rheumatologist s expert clinical diagnosis and concluded that, although the prevalence estimates were similar, the concordance yielded only a kappa of 0.41 (low to moderate agreement). This suggests that the ASAS axial SpA criteria and rheumatology experts captured different patient populations. According to the axial SpA criteria 24% of the patients were undiagnosed and 21% misdiagnosed in clinical practise, 23 which can also be interpreted as the ASAS classification criteria leading to false positive diagnosis in 24% of patients when applied as diagnostic tool (in comparison with the golden standard, the rheumatologist s expert opinion). This is probably related to the fact that MRI and HLA-B27 are major entry criteria in the ASAS classification but are diagnostically only useful in patients with a moderately high a priori chance of SpA. 27 It should be noted that in our study MRI of the sacroiliac joints was not systematically 5 Burden of peripheral disease 71

73 Table 3. Clinical characteristics, disease activity and treatment of patients with pure axial SpA, combined axial and peripheral SpA, versus ASAS peripheral SpA Pure axial SpA (n=112) Combined SpA (n=118) Peripheral SpA (n=84) P-value Pure axial vs combined SpA P-value Combined vs Peripheral SpA Age, median (IQR) years 38.8 ( ) 43.6 ( ) 47.9 ( ) Disease duration, median (IQR) years 3.1 ( ) 4.6 ( ) 4.2 ( ) Age at disease onset, median (IQR) years 31.8 ( ) 34.2 ( ) 38.2 ( ) Male gender, % HLA-B27 positive, % <0.001 Inflammatory back pain (history/presence), % NS <0.001 Buttock pain (history/presence), % <0.001 Sacroiliitis low grade, % Sacroiliitis high grade, % <0.001 Peripheral arthritis (history/presence), % <0.001 <0.001 Enthesitis (history/presence), % < Dactylitis (history/presence), % Uveitis (history/presence), % <0.001 Psoriasis (history/presence), % <0.001 IBD (history/presence), % Urethritis/diarrhea (history/presence), % Positive SpA family history, %

74 Patient s global assessment, median (IQR) mm 52.0 ( ) 61.0 ( ) 38.0 ( ) <0.001 Physician s global assessment, median (IQR) mm 44.0 ( ) 53.0 ( ) 30.5 ( ) <0.001 BASDAI, median (IQR) 4.4 ( ) 5.4 ( ) 3.0 ( ) <0.001 BASDAI 4, % <0.001 BASDAI #2 back pain, median (IQR) mm 57.0 ( ) 65.5 ( ) 11.5 ( ) <0.001 ASDAS-CRP, median (IQR) 2.6 ( ) 3.0 ( ) 2.0 ( ) <0.001 ASDAS-CRP 2.1, % <0.001 Swollen joint count, median (IQR) 0-66 joints 0.0 ( ) 0.0 ( ) 1.0 ( ) < Tender joint count, median (IQR) 0-68 joints 0.0 ( ) 1.0 ( ) 1.0 ( ) < Schober, median (IQR) centimetres 3.0 ( ) 4.0 ( ) 4.5 ( ) <0.001 Chest expansion, median (IQR) centimetres 5.0 ( ) 4.0 ( ) 5.0 ( ) CRP, median (IQR) mg/l 3.3 ( ) 4.0 ( ) 3.7 ( ) ESR, median (IQR) mm/h 8.0 ( ) 12.0 ( ) 7.5 ( ) NSAIDs, % Corticosteroids, % Methotrexate, % <0.001 Sulfasalazine, % Other DMARDs, % Anti-TNF therapy, % SpA = spondyloarthritis; ASAS = Assessment of SpondyloArthritis international Society; IQR = interquartile range; Sacroiliitis low grade = grade <2 bilateral or grade <3 unilateral, and high grade = grade 2 bilateral or grade 3 unilateral on X-ray; IBD = inflammatory bowel disease; Positive SpA family history = 1 first- or second-degree relative with ankylosing spondylitis, psoriasis, IBD, reactive arthritis or uveitis; BASDAI = Bath Ankylosing Spondylitis Disease Activity Index; ASDAS = Ankylosing Spondylitis Disease Activity Score; CRP = C-reactive protein; ESR = erythrocyte sedimentation rate; NSAIDs = non-steroidal anti-inflammatory drugs; DMARDs = disease-modifying anti-rheumatic drugs; TNF = tumor necrosis factor. Significance of the comparisons is determined by Mann-Whitney U tests for continuous variables or chi-square tests for categorical variables. 5 Burden of peripheral disease 73

75 performed in all patients as it was the treating rheumatologist who decided whether or not MRI was a useful additional diagnostic procedure or not in an individual patient. It is thus possible that systematically performing MRI would classify additional patients and that our data slightly overestimate the number of patients with SpA not fulfilling the ASAS criteria. However access to MRI is not universally available to patients due to costs and availability issues, and also other studies assessing the performance of the ASAS criteria did not include MRI in all cases. 22,28,29 The observations in the present study are in line with previous studies indicating that classifying patients with combined axial and peripheral disease exclusively according to the axial arm of the criteria may not be optimal. In our recent proof-of-concept clinical trial with adalimumab in non-as, non-psa peripheral SpA according to the ESSG criteria post-hoc application revealed that 38/40 patients would fulfil the ASAS criteria if not taking into account that 22/40 patients also had active IBP when actively questioned for axial complaints. 30 Moreover, the studies which assessed the performance of the ASAS criteria also did not strictly follow this artificial subdivision into either axial or peripheral disease. 22,28,29 In the Leiden early arthritis cohort the ASAS peripheral SpA criteria were applied on early arthritis patients with past and/or present IBP, 28 while strictly speaking these criteria were not allowed to be used since the ASAS axial SpA criteria should always be applied in case of active IBP and not the peripheral SpA criteria. 4 Another study in established SpA (the COSPA (Cochin SpondyloArthritis study) cohort) used either the ASAS axial or peripheral SpA criteria based on what the predominant manifestation was according to the treating physician, 29 while when stringently following the ASAS criteria it is predefined when which criteria should be applied. 4 And a study analyzing the performance of the ASAS criteria in early SpA units within the ESPERANZA program applied the axial SpA criteria only on patients with exclusively back pain and the peripheral SpA only on patients with peripheral arthritis, dactylitis or enthesitis in the absence of axial pain, 22 which is also not in line with what the ASAS classification criteria state. 4 Another study assessing the prevalence of SpA in southern Sweden concluded that 57% fulfilled the ASAS classification criteria for peripheral SpA, 91% the ASAS classification criteria for axial SpA, and that 45% of this latter group also fulfilled the ASAS criteria for peripheral SpA, 31 while when really following the ASAS criteria it is not possible to fulfill both criteria arms. 4 These studies show that there is a lot of confusion when to apply which criteria, which lead to wrong results for sensitivity and specificity of the ASAS criteria and make it impossible to reliably compare results for prevalence and patients characteristics from different studies with each other. However, it is of great importance that these criteria are investigated thoroughly since there are already reports which suggest to use the ASAS criteria not only as classification criteria but also as diagnostic criteria In conclusion, the data obtained in this cohort indicate that the ASAS criteria tend to underestimate the prevalence and disease burden of peripheral disease in SpA as a) a substantial subgroup of SpA patients having mainly peripheral disease and displaying high disease activity are not classified, and b) axial SpA as defined by ASAS consists of two separate groups with either exclusive axial disease or combined axial and peripheral disease, with the latter having the highest disease activity. If reproduced in independent cohorts, these data support a classification into axial, combined, or peripheral disease rather than just axial versus peripheral disease. We propose that the ASAS criteria should be modified in order to let the combined group enter the classification through either the axial or peripheral SpA criteria (Figure 3). 74

76 5 Figure 3. The Assessment of SpondyloArthritis International Society (ASAS) criteria and the modified ASAS criteria, where we propose a small modification concerning that patients with both active axial and peripheral symptoms can enter the SpA criteria either through the axial or peripheral arm instead of exclusively the axial arm. Acknowledgements We would like to thank the following physicians who worked on the SpA outpatient clinic during our study period: Marc Kok, Radjesh Bisoendial, Ruth Klaassen, Lot Burgemeister, Liesbeth Hak, Frieda Koopman, Ivy Choi, Els van Nood, Nadine Reyn-Gibson, Mirella Bes, and Anouk Vedder. Prof. Dr. D. Baeten was supported by a VIDI grant from The Netherlands Organization for Scientific Research (NWO) and by a grant from the Dutch Arthritis Foundation (Reumafonds). Burden of peripheral disease REFERENCES 1 Dougados M, van der Linden S, Juhlin R, Huitfeldt B, Amor B, Calin A, et al. The European Spondylarthropathy Study Group preliminary criteria for the classification of spondylarthropathy. Arthritis Rheum 1991;34: Dougados M, Baeten D. Spondyloarthritis. Lancet 2011;377: Rudwaleit M, van der Heijde D, Landewe R, Listing J, Akkoc N, Brandt J, et al. The development of Assessment of SpondyloArthritis international Society classification criteria for axial spondyloarthritis (part II): validation and final selection. Ann Rheum Dis 2009;68: Rudwaleit M, van der Heijde D, Landewe R, Akkoc N, Brandt J, Chou CT, et al. The Assessment of SpondyloArthritis International Society classification criteria for peripheral spondyloarthritis and for spondyloarthritis in general. Ann Rheum Dis 2011;70: Said-Nahal R, Miceli-Richard C, Berthelot JM, Duche A, Dernis-Labous E, Le Blevec G, et al. The familial form of spondylarthropathy: a clinical study of 115 multiplex families. Groupe Francais d Etude Genetique des Spondylarthropathies. Arthritis Rheum 2000;43: Said-Nahal R, Miceli-Richard C, D Agostino MA, Dernis-Labous E, Berthelot JM, Duche A, 75

77 et al. Phenotypic diversity is not determined by independent genetic factors in familial spondylarthropathy. Arthritis Rheum 2001;45: Rahman P, Inman RD, Gladman DD, Reeve JP, Peddle L, Maksymowych WP. Association of interleukin-23 receptor variants with ankylosing spondylitis. Arthritis Rheum 2008;58: Rahman P, Inman RD, Maksymowych WP, Reeve JP, Peddle L, Gladman DD. Association of interleukin 23 receptor variants with psoriatic arthritis. J Rheumatol 2009;36: Hammer RE, Maika SD, Richardson JA, Tang JP, Taurog JD. Spontaneous inflammatory disease in transgenic rats expressing HLA-B27 and human beta 2m: an animal model of HLA-B27-associated human disorders. Cell 1990;63: Tran TM, Dorris ML, Satumtira N, Richardson JA, Hammer RE, Shang J, et al. Additional human beta2-microglobulin curbs HLA-B27 misfolding and promotes arthritis and spondylitis without colitis in male HLA-B27-transgenic rats. Arthritis Rheum 2006;54: Kruithof E, Baeten D, De Rycke L, Vandooren B, Foell D, Roth J, et al. Synovial histopathology of psoriatic arthritis, both oligo- and polyarticular, resembles spondyloarthropathy more than it does rheumatoid arthritis. Arthritis Res Ther 2005;7:R569-R Vandooren B, Noordenbos T, Ambarus C, Krausz S, Cantaert T, Yeremenko N, et al. Absence of a classically activated macrophage cytokine signature in peripheral spondylarthritis, including psoriatic arthritis. Arthritis Rheum 2009;60: Noordenbos T, Yeremenko N, Gofita I, van de Sande M, Tak PP, Canete JD, et al. Interleukin- 17-positive mast cells contribute to synovial inflammation in spondylarthritis. Arthritis Rheum 2012;64: Yeremenko N, Noordenbos T, Cantaert T, van Tok M, van de Sande M, Canete JD, et al. Diseasespecific and inflammation-independent stromal alterations in spondylarthritis synovitis. Arthritis Rheum 2013;65: Baeten D, Breban M, Lories R, Schett G, Sieper J. Are spondylarthritides related but distinct conditions or a single disease with a heterogeneous phenotype? Arthritis Rheum 2013;65: Appel H, Maier R, Wu P, Scheer R, Hempfing A, Kayser R, et al. Analysis of IL-17(+) cells in facet joints of patients with spondyloarthritis suggests that the innate immune pathway might be of greater relevance than the Th17-mediated adaptive immune response. Arthritis Res Ther 2011;13:R Vandooren B, Kruithof E, Yu DT, Rihl M, Gu J, De Rycke L, et al. Involvement of matrix metalloproteinases and their inhibitors in peripheral synovitis and down-regulation by tumor necrosis factor alpha blockade in spondylarthropathy. Arthritis Rheum 2004;50: Clegg DO, Reda DJ, Mejias E, Cannon GW, Weisman MH, Taylor T, et al. Comparison of sulfasalazine and placebo in the treatment of psoriatic arthritis. A Department of Veterans Affairs Cooperative Study. Arthritis Rheum 1996;39: Garrett S, Jenkinson T, Kennedy LG, Whitelock H, Gaisford P, Calin A. A new approach to defining disease status in ankylosing spondylitis: the Bath Ankylosing Spondylitis Disease Activity Index. J Rheumatol 1994;21: Lukas C, Landewe R, Sieper J, Dougados M, Davis J, Braun J, et al. Development of an ASASendorsed disease activity score (ASDAS) in patients with ankylosing spondylitis. Ann Rheum Dis 2009;68: Machado P, Landewe R, Lie E, Kvien TK, Braun J, Baker D, et al. Ankylosing Spondylitis Disease Activity Score (ASDAS): defining cut-off values for disease activity states and improvement scores. Ann Rheum Dis 2011;70: Tomero E, Mulero J, de Miguel E, Fernandez- Espartero C, Gobbo M, Descalzo MA, et al. Performance of the Assessment of Spondyloarthritis International Society criteria for the classification of spondyloarthritis in early spondyloarthritis clinics participating in the ESPERANZA programme. Rheumatology (Oxford) 2014;53: Strand V, Rao SA, Shillington AC, Cifaldi MA, McGuire M, Ruderman EM. Prevalence of axial spondyloarthritis in United States rheumatology practices: Assessment of SpondyloArthritis International Society criteria versus rheumatology expert clinical diagnosis. Arthritis Care Res (Hoboken ) 2013;65: van den Berg R, de Hooge M, van Gaalen F, Reijnierse M, Huizinga T, van der Heijde D. Percentage of patients with spondyloarthritis in patients referred because of chronic back pain and performance of classification criteria: experience from the Spondyloarthritis Caught Early (SPACE) cohort. Rheumatology (Oxford) 2013;52: van den Berg R, de Hooge M, Rudwaleit M, Sieper J, van Gaalen F, Reijnierse M, et al. ASAS modification of the Berlin algorithm for diagnosing 76

78 axial spondyloarthritis: results from the SPondyloArthritis Caught Early (SPACE)-cohort and from the Assessment of SpondyloArthritis international Society (ASAS)-cohort. Ann Rheum Dis 2013;72: Molto A, Paternotte S, Comet D, Thibout E, Rudwaleit M, Claudepierre P, et al. Performances of the Assessment of SpondyloArthritis International Society axial spondyloarthritis criteria for diagnostic and classification purposes in patients visiting a rheumatologist because of chronic back pain: results from a multicenter, cross-sectional study. Arthritis Care Res (Hoboken) 2013;65: Rudwaleit M, van der Heijde D, Khan MA, Braun J, Sieper J. How to diagnose axial spondyloarthritis early. Ann Rheum Dis 2004;63: van den Berg R, van Gaalen F, van der Helm-van Mil A, Huizinga T, van der Heijde D. Performance of classification criteria for peripheral spondyloarthritis and psoriatic arthritis in the Leiden Early Arthritis cohort. Ann Rheum Dis 2012;71: Cheung PP, Paternotte S, Burki V, Durnez A, Elhai M, Fabreguet I, et al. Performance of the assessment in Spondyloarthritis International Society classification for axial and peripheral spondyloarthritis in an established clinical cohort: comparison with criteria sets of Amor and the European Spondylarthropathy Study Group. J Rheumatol 2012;39: Paramarta JE, De Rycke L, Heijda TF, Ambarus CA, Vos K, Dinant HJ, et al. Efficacy and safety of adalimumab for the treatment of peripheral arthritis in spondyloarthritis patients without ankylosing spondylitis or psoriatic arthritis. Ann Rheum Dis 2013;72: Haglund E, Bremander AB, Petersson IF, Strombeck B, Bergman S, Jacobsson LT, et al. Prevalence of spondyloarthritis and its subtypes in southern Sweden. Ann Rheum Dis 2011;70: Burden of peripheral disease 77

79

80 Part two Novel treatments in spondyloarthritis

81

82 Efficacy and safety of adalimumab for the treatment of peripheral arthritis in spondyloarthritis patients without ankylosing spondylitis or psoriatic arthritis Jacqueline E. Paramarta 1, Leen De Rycke 1, Tanja F. Heijda 1, Carmen A. Ambarus 1, Koen Vos 1, Huib J. Dinant 1, Paul P. Tak 1,2, Dominique L. Baeten 1 1 Department of Clinical Immunology and Rheumatology, Academic Medical Center/University of Amsterdam, Amsterdam, The Netherlands, 2 GlaxoSmithKline, Stevenage, UK 6 Ann Rheum Dis Nov 1;72(11):1793-9

83 ABSTRACT Objectives To evaluate the efficacy and safety of adalimumab in patients with peripheral spondyloarthritis (SpA) not fulfilling the criteria for ankylosing spondylitis (AS) or psoriatic arthritis (PsA). Methods 40 patients with active peripheral SpA fulfilling the European Spondyloarthropathy Study Group or Amor criteria but not the criteria for AS or PsA were included in a randomised, double-blind, placebo-controlled clinical trial. Patients were treated 1 : 1 with adalimumab or placebo for 12 weeks, followed by an open label extension up to week 24. Safety and efficacy measurements were performed every 6 weeks, with the patient s global assessment of disease activity at week 12 as the primary endpoint. Results At week 12, the patient s and physician s global assessment of disease activity, swollen joint count, Bath Ankylosing Spondylitis Disease Activity Index (BASDAI), Ankylosing Spondylitis Disease Activity Score (ASDAS) and erythrocyte sedimentation rate improved significantly in the adalimumab group compared with the baseline values and compared with placebo. A similar improvement was seen upon adalimumab treatment from weeks 12 to 24 in the patients originally randomised to placebo, whereas the clinical response was maintained or even augmented at week 24 in the patients who received adalimumab from the start. ASDAS inactive disease and BASDAI50 responses were met in 42% of the adalimumab group versus 0% 5% in the placebo group at week 12 (p=0.001 and p=0.008, respectively), and were further increased at week 24. The number of adverse events was not different between the adalimumab and placebo groups. Conclusions Adalimumab appears to be effective and well tolerated in SpA patients with peripheral arthritis, also in those patients not fulfilling the AS or PsA criteria. 82

84 INTRODUCTION The treatment of spondyloarthritis (SpA) has improved since the successful introduction of tumour necrosis factor (TNF)-blockade in this disease more than 10 years ago. 1-5 However, this therapy is only approved and reimbursed for ankylosing spondylitis (AS) and psoriatic arthritis (PsA), which are the best described and studied phenotypic subtypes of SpA. Approximately a third of the SpA population cannot be classified as AS or PsA and belong to one of the other SpA subtypes: inflammatory bowel disease related SpA, reactive arthritis and undifferentiated spondyloarthritis. 6,7 The last consists of patients with peripheral disease as well as patients with axial disease not fulfilling the modified New York criteria for AS, 8 now commonly called non-radiographic axial SpA (nr-axspa). Based on the similarities in symptoms and disease severity between nr-axspa and AS, 9 new classification criteria have been developed for axial SpA without phenotypic subclassifications. 10 Supporting the concept that nr-axspa is not fundamentally different from AS, several small placebo-controlled randomised clinical trials (RCTs) have demonstrated the efficacy of TNF-blockers in nr-axspa This has recently been confirmed in a large RCT with adalimumab. 14 As for axial SpA, familial aggregation studies, synovial immunopathology and experimental models 21 strongly suggest that also peripheral disease in SpA belongs to a single pathological entity. Accordingly, unifying criteria have recently been developed for peripheral SpA. 22 TNF-blockade has a proven beneficial effect on peripheral manifestations of AS and PsA. 3,4,23-26 In non-as and non-psa peripheral SpA, several small studies suggest a good efficacy of anti-tnf-therapy but RCTs are still lacking Therefore, the objective of this study was to assess the efficacy and safety of adalimumab in a randomised, double-blind, placebo-controlled clinical trial in patients with peripheral SpA not fulfilling the criteria for AS or PsA. METHODS Study design Overall, 40 patients were randomised (1 : 1) in a single centre, double-blind clinical trial to receive adalimumab 40 mg or placebo subcutaneously every other week for 12 weeks, followed by an open label extension with adalimumab for another 12 weeks. Written informed consent was obtained from each patient before study related procedures were performed. If eligible, patients entered the study within 3 weeks. The primary endpoint of the study was the improvement in the patient s global assessment of disease activity at week 12. This endpoint was chosen because most other disease activity measurements in SpA are validated in AS only and because it has been used previously as primary endpoint in anti-tnf trials in mixed SpA populations. 4 The study drug was provided in prefilled syringes containing adalimumab 40 mg or matching placebo (Abbott Laboratories, Abbott Park, Illinois, USA). The study was approved by the local Ethics Committee and is registered under the code NTR Adalimumab in peripheral SpA Patients Patients had to fulfil the European Spondyloarthropathy Study Group (ESSG) criteria 6 and/or the Amor criteria 33 for SpA for at least 3 months without fulfilling the criteria for AS 8 or PsA

85 The Assessment of SpondyloArthritis International Society (ASAS) criteria for peripheral SpA 22 were not used as they were not published yet when the study was started. Patients had to be between 18 and 70 years old and had to have an active arthritis (at least one swollen and tender joint) despite treatment with non-steroidal anti-inflammatory drugs (NSAIDs). For female subjects, a negative pregnancy test and adequate contraception during the study and for 150 days thereafter were criteria for entry. Exclusion criteria included serious infections in the previous 4 weeks, history of malignancy in the past 10 years, significant history of other severe diseases or uncontrolled concomitant disease. Chest radiography and a tuberculin skin test were performed at screening. In case of active tuberculosis, patients were excluded and in case of latent tuberculosis, patients had to receive at least 3 months of isoniazide before enrolment. Concurrent and prior medication Patients were allowed to continue NSAIDs, corticosteroids ( 10 mg/day prednisone or equivalent), methotrexate and sulphasalazine provided that the dosage was stable for at least 4 weeks prior to baseline and throughout the study and, for methotrexate and sulphasalazine, provided they had been started at least 3 months before inclusion. Intra-articular corticosteroids and other disease-modifying antirheumatic drugs (DMARDs) were not allowed and were discontinued at least 4 weeks prior to baseline. Prior anti-tnf-therapy or investigational drugs were permitted if stopped more than 2 months prior to baseline. Clinical response Patients were seen for clinical evaluation at baseline, weeks 6, 12, 18 and 24 where the following disease activity parameters were evaluated: patient s and physician s global assessment of disease activity (each a 100 mm visual analogue scale), 68/66 tender joint count (TJC) and swollen joint count (SJC), Bath Ankylosing Spondylitis Disease Activity Index (BASDAI), 35 Ankylosing Spondylitis Disease Activity Score (ASDAS), 36 modified Schober index, erythrocyte sedimentation rate (ESR) and C reactive protein (CRP). Improvement at weeks 12 and 24 was also measured by the different ASDAS improvement criteria and BASDAI50 response The Health Assessment Questionnaire Disability Index (HAQ-DI) 39 and Health Utility Index Mark-3 (HUI-3) 40,41 were measured as disability and health related quality of life (QoL) variables. At each visit patients were asked for side effects, and routine laboratory testing and physical examination were performed for safety evaluation. Statistical analysis The sample size calculation was based on previous anti-tnf trials which included peripheral SpA 1,4,27 with an estimated response in patient s global assessment at week 12 of 48±24 mm in the adalimumab group and 21±32 mm in the placebo group, a power of 80% and α level of The data were tested for normal distribution and presented as mean±sd or SEM. Analysis of covariance (ANCOVA) adjusting for the baseline score was used to compare change from baseline at week 12 between the adalimumab and placebo groups. Further comparison of the adalimumab and placebo groups was done by independent sample t test for continuous 84

86 variables and χ 2 test for categorical variables. Changes from baseline (intragroup analysis) were assessed by paired t tests. All statistical tests were two-sided and p values of <0.05 were considered statistically significant. RESULTS Clinical and demographic characteristics at baseline In all, 40 patients were enrolled in the study and subsequently randomised to treatment with adalimumab or placebo. A total of 31 patients fulfilled both the ESSG and Amor criteria, six only the ESSG criteria, and three only the Amor criteria. The demographic and clinical characteristics were similar across both treatment arms (table 1). Only human leukocyte antigen B27 (HLA-B27) positivity tended to be higher in the adalimumab group (55%) versus the placebo group (25%) (p=0.053). At baseline, both treatment groups were well matched for all disease activity parameters except for the physician s global assessment, which was somewhat higher in the placebo group (57.0±12.6 mm) compared with the adalimumab group (47.8±11.8 mm) (p=0.022) (table 1). The mean SJC66 tended to be slightly higher in the patients who were assigned to adalimumab (4.3±4.2) compared with placebo (2.5±1.9) (p=0.096). The number of patients with elevated CRP (n=7 and n=8 in the adalimumab and placebo groups, respectively) and ESR (n=3 and n=4, respectively) at baseline was similar in the two groups. Concomitant NSAID and DMARD use was not different between the treatment groups. One patient randomised to receive adalimumab and two patients randomised to receive placebo had previously been treated with TNF-blockers (infliximab or adalimumab) for inflammatory bowel disease, but in all cases TNF-blockade was stopped more than 1 year prior to baseline. Two patients tested positive for latent tuberculosis and were treated with isoniazide before entering the trial. 6 Adalimumab in peripheral SpA Improvement of disease activity at week 12 The primary endpoint of the study was met since the patient s global assessment showed a strong improvement at week 12 compared with baseline in the adalimumab group ( 31.0±23.3 mm), while the placebo group showed almost no improvement ( 5.9±21.4 mm) (p=0.001). Also the improvement in the physician s global assessment, BASDAI, ASDAS and ESR was significantly greater in the adalimumab treated patients compared with placebo (p<0.05 for all parameters) (table 2). These data were confirmed by direct comparison of the adalimumab and placebo groups: the patient s and physician s global assessment, BASDAI and ASDAS were significantly lower in adalimumab compared with placebo at week 6 as well as at week 12 (figure 1). Moreover, the intragroup analysis demonstrated a strong and significant improvement at week 12 compared with baseline in the adalimumab group for the patient s and physician s global assessment, SJC66, BASDAI and ASDAS (p<0.05 for all parameters) (table 3). In sharp contrast, none of these parameters showed any statistically significant change in the placebo group over this time period. The inflammatory parameters showed a numerical but non-significant decrease at week 12 of adalimumab treatment. Subanalysis of patients with increased CRP at baseline, however, showed that CRP levels decreased in seven out of seven patients treated with adalimumab. 85

87 60 Patient's global assessment 80 Placebo controlled Open label Physician's global assessment 80 Placebo controlled Open label 60 Adalimumab Placebo mm 40 mm 40 * 20 0 Week 0 Week 6 * * Week 12 Week 18 Week Week 0 * Week 6 * Week 12 Week 18 * Week 24 BASDAI 8 Placebo controlled Open label 4 Placebo controlled ASDAS Open label * * 1 * * 0 Week 0 Week 6 Week 12 Week 18 Week 24 0 Week 0 Week 6 Week 12 Week 18 Week 24 number of joints Swollen joint count Placebo controlled Open label Week 0 Week 6 Week 12 Week 18 * Week 24 Back pain 8 Placebo controlled Open label Week 0 Week 6 Week 12 Week 18 Week 24 Figure 1. Changes in the clinical disease activity parameters during treatment with adalimumab or placebo from week 0 until 12, and during the open label extension phase with adalimumab from week 12 until 24. The panel represents the patient s and physician s global assessment of disease activity on a 100 mm visual analogue scale (VAS), the Bath Ankylosing Spondylitis Disease Activity Index (BASDAI), Ankylosing Spondylitis Disease Activity Score (ASDAS), 66 swollen joint count, and back pain score (assessed by BASDAI question #2). Data are presented as mean (standard error of the mean). * P <0.05 compared to placebo by independent sample t-test. 86

88 Table 1. Baseline characteristics of the study population by treatment group Adalimumab (n=20) Placebo (n=20) Age, mean (SD) years 41.5 (12.8) 44.4 (11.1) Disease duration, mean (SD) years 7.9 (9.3) 6.7 (6.2) Number of men/women 9/11 12/8 BMI, mean (SD) kg/m (4.2) 26.8 (5.7) HLA-B27 positive, n (%) 11 (55) 5 (25) Subtype USpA, n (%) 15 (75) 17 (85) Subtype ReA, n (%) 4 (20) 0 (0) Subtype IBD-SpA, n (%) 1 (5) 3 (15) Inflammatory back pain (history/presence), n (%) 16 (80) 15 (75) Enthesitis (history/presence), n (%) 15 (75) 15 (75) Dactylitis (history/presence), n (%) 4 (20) 1 (5) Patient s global assessment, mm VAS 65.8 (18.1) 68.4 (17.4) Physician s global assessment, mm VAS 47.8 (11.8) 57.0 (12.6)* Swollen joint count, 0-66 joints 4.3 (4.2) 2.5 (1.9) Tender joint count, 0-68 joints 9.4 (8.2) 10.6 (5.9) BASDAI 5.5 (2.3) 6.0 (1.4) ASDAS 2.9 (1.0) 3.2 (0.9) CRP, mg/l 7.8 (13.3) 13.5 (26.4) ESR, mm/hour 11.6 (13.5) 15.7 (23.1) Concomitant NSAIDs, n (%) 13 (65) 14 (70) Concomitant corticosteroids, n (%) 0 (0) 2 (10) Concomitant MTX, n (%) 5 (25) 6 (30) Concomitant SSZ, n (%) 7 (35) 4 (20) Previous anti-tnf-treatment, n (%) 1 (5) 2 (10) 6 Adalimumab in peripheral SpA BMI = body mass index; USpA = undifferentiated spondyloarthritis; ReA = reactive arthritis; IBD-SpA = inflammatory bowel disease related spondyloarthritis; VAS = visual analogue scale; BASDAI = Bath Ankylosing Spondylitis Disease Activity Index; ASDAS = Ankylosing Spondylitis Disease Activity Score; CRP = C-reactive protein; ESR = erythrocyte sedimentation rate; NSAIDs = non-steroidal anti-inflammatory drugs; MTX = methotrexate; SSZ = sulphasalazine; TNF = tumour necrosis factor. Significance of the comparisons is determined by independent sample t-test for continuous variables and χ 2 test for categorical variables. * P <0.05 compared to the adalimumab group. Improvement of disease activity at week 24 After the 12-week placebo-controlled phase all patients were treated in an open label phase with adalimumab for an additional 12 weeks. In the original placebo group, adalimumab treatment induced a significant decrease of all clinical disease activity parameters between weeks 12 and 24 (p<0.05 for all parameters) (table 3 and figure 1). Additionally, ESR was significantly suppressed (p<0.05). Analysis of patients with increased CRP at baseline showed a decrease in four out of five patients during this treatment period. 87

89 Table 2. Mean changes in disease activity from baseline to week 12 by treatment group Baseline to week 12 Adalimumab (n=19) Placebo (n=19) P-value Patient s global assessment, mm VAS 31.0 (23.3) 5.9 (21.4) Physician s global assessment, mm VAS 19.8 (19.5) 4.1 (16.4) Swollen joint count, 0-66 joints 2.5 (4.0) 0.4 (1.8) Tender joint count, 0-68 joints 1.8 (9.2) 1.7 (6.5) BASDAI 1.8 (2.6) 0.3 (1.5) ASDAS 1.1 (1.2) 0.0 (0.9) CRP, mg/l 5.7 (12.4) 4.0 (22.9) ESR, mm/hour 6.0 (12.5) 1.7 (9.3) Values are the mean change (standard deviation) from baseline to week 12. VAS = visual analogue scale; BASDAI = Bath Ankylosing Spondylitis Disease Activity Index; ASDAS = Ankylosing Spondylitis Disease Activity Score; CRP = C-reactive protein; ESR = erythrocyte sedimentation rate. Significance of the comparisons is determined by analysis of covariance (ANCOVA) adjusted for baseline, comparing adalimumab with placebo treatment. In the patients who already received adalimumab from baseline, we observed a further improvement of all clinical disease activity parameters between weeks 12 and 24, reaching statistical significance for physician s global assessment, SJC66 and BASDAI (table 3 and figure 1). All clinical disease activity parameters as well as ESR were significantly improved at week 24 in comparison with baseline. ASDAS and BASDAI response criteria Although the trial was designed to assess the effect of adalimumab on peripheral SpA, 22 out of 40 patients had also axial complaints at baseline. As we observed a significant effect on BASDAI and ASDAS scores (tables 2, 3 and figure 1), we additionally assessed the ASDAS and BASDAI response criteria. At week 12, 53% of the patients treated with adalimumab met the criteria for ASDAS clinically important improvement, 42% the criteria for ASDAS inactive disease and 42% a BASDAI50 response (figure 2). All three assessments reached statistical significance compared with placebo, which showed an ASDAS improvement of 11%, ASDAS inactive disease of 0% and BASDAI50 response of 5%. By 24 weeks, the initial placebo treated group showed a similar response to 12 weeks of adalimumab treatment: ASDAS clinically important improvement, ASDAS inactive disease and BASDAI50 response were reached in 35% for all three parameters (figure 2). In the group receiving adalimumab from the beginning, treatment with adalimumab for an additional 12 weeks even further increased the percentage of patients who reached the ASDAS clinically important improvement, ASDAS inactive disease and BASDAI50 response to 63%, 58% and 53%, respectively (figure 2). Improvement in disability and QoL measurements To assess the impact of adalimumab treatment on disability and QoL, we measured the HAQ-DI and the HUI-3. As shown in table 3, an improvement in HAQ-DI was observed at week 12 in the 88

90 Table 3. Disease activity parameters at week 0, 12 and 24 by treatment group Adalimumab week 0-24 Placebo week 0-12 Adalimumab week week 24 (n=17) week 12 (n=19) week 0 (n=20) week 24 (n=19) week 12 (n=19) week 0 (n=20) Patient s global assessment, mm VAS 65.8 (18.1) 33.3 (26.6)* 22.5 (20.2)* 68.4 (17.4) 62.2 (21.5) 34.9 (25.7)* Physician s global assessment, mm VAS 47.8 (11.8) 28.4 (20.9)* 18.4 (14.8)* 57.0 (12.6) 52.1 (18.9) 33.8 (21.0)* Swollen joint count, 0-66 joints 4.3 (4.2) 1.9 (2.2)* 0.6 (1.0)* 2.5 (1.9) 2.2 (1.1) 1.3 (0.8)* Tender joint count, 0-68 joints 9.4 (8.2) 7.9 (14.0) 4.6 (6.9)* 10.6 (5.9) 12.6 (8.8) 7.8 (7.0) BASDAI 5.5 (2.3) 3.8 (3.2)* 2.5 (2.1)* 6.0 (1.4) 5.7 (1.7) 3.6 (2.4)* ASDAS 2.9 (1.0) 1.8 (1.2)* 1.4 (0.8)* 3.2 (0.9) 3.1 (0.9) 1.9 (0.9)* CRP, mg/l 7.8 (13.3) 2.5 (2.2) 2.1 (2.8) 13.5 (26.4) 12.8 (24.7) 7.9 (24.0) ESR, mm/hour 11.6 (13.5) 6.1 (7.2) 4.9 (2.7)* 15.7 (23.1) 13.1 (13.7) 4.1 (2.6) HAQ-DI 0.8 (0.6) 0.6 (0.7)* 0.4 (0.6)* 1.1 (0.5) 1.0 (0.4) 0.6 (0.4)* HUI (0.35) 0.60 (0.40) 0.67 (0.36)* 0.38 (0.28) 0.46 (0.34) 0.59 (0.31)* Values are the mean (standard deviation). VAS = visual analogue scale; BASDAI = Bath Ankylosing Spondylitis Disease Activity Index; ASDAS = Ankylosing Spondylitis Disease Activity Score; CRP = C-reactive protein; ESR = erythrocyte sedimentation rate; HAQ-DI = Health Assessment Questionnaire Disability Index; HUI-3 = Health Utility Index Mark 3. Significance of the comparisons is determined by a paired t-test. * P <0.05 compared to week 0 (by treatment group). P <0.05 week 24 compared to week 12 (by treatment group). 6 Adalimumab in peripheral SpA 89

91 Improvement at week 12 Improvement at week 24 % of patients * * * % of patients ASDAS improvement ASDAS inactive disease BASDAI50 response 0 Adalimumab (n=19) Placebo (n=19) 0 Adalimumab (n=19) Placebo/Adalimumab (n=17) Figure 2. Percentages of patients who achieved a clinically important improvement and inactive disease according to the Ankylosing Spondylitis Disease Activity Score (ASDAS) criteria, and a Bath Ankylosing Spondylitis Disease Activity Index (BASDAI) 50 response at week 12 and week 24. The first 12 weeks patients received either adalimumab or placebo, the last 12 weeks all the patients received adalimumab in the open label extension phase. * P<0.05 compared to placebo by χ 2 test. adalimumab group but not the placebo group. A similar trend was seen for the HUI-3. At week 24, both parameters were significantly improved in both treatment arms (p<0.05 for both parameters). Dropouts and safety analysis The safety analysis is summarised in table 4. In each treatment arm, 19 patients completed the 12-week placebo-controlled phase. Two patients discontinued the study because of a serious adverse event (SAE): one death due to suicide in the adalimumab group and one arthroscopy related septic arthritis in the placebo group. No other SAEs were observed. The number of nonserious adverse events (AEs) was similar (n=12 in each group). The number of infectious events was lower in the adalimumab group (n=4) than in the placebo group (n=8). Two adalimumab treated patients developed a diffuse skin rash, which resolved after topical corticosteroids and oral antihistaminic treatment, and did not result in treatment discontinuation. In the open label phase, all 19 patients from the group originally treated with adalimumab completed the full study. In the group initially randomised to receive placebo, 17 patients completed the open extension until week 24: one patient dropped out due to a SAE (hospital admission because of an acute psychosis) and one patient was lost to follow-up. There were no other SAEs. Non-serious AEs included 13 infections, consisting mainly of upper respiratory tract infections and cystitis, and one local injection reaction, which resolved spontaneously. Overall, there were no unexpected safety signals. DISCUSSION The present randomised double-blind, placebo-controlled clinical trial demonstrated the efficacy and safety of adalimumab in patients with active peripheral SpA not fulfilling the criteria for AS or PsA. The primary objective of the study was met as there was a rapid and significant decrease in the patient s global assessment upon adalimumab but not placebo treatment. Additionally, treatment 90

92 with adalimumab but not with placebo, significantly improved all other clinical parameters of disease activity except the TJC68, which showed a numerical but not significant decrease. The efficacy of adalimumab in this patient group was confirmed by three additional observations. First, ESR and CRP as objective measurements of inflammation did consistently decrease in those patients with elevated values at baseline. Second, the patients who were initially treated with placebo also showed good efficacy upon treatment with adalimumab in the open label phase. And third, the clinical benefit of TNF-blockade was sustained and even augmented on weeks 18 and 24 in the group continuously treated with adalimumab. Moreover, multiple regression analysis (data not shown) showed that features such as gender, HLA-B27 status, CRP, ESR, SJC66 and concomitant DMARD treatment did not act as confounders in this trial. These data are in agreement with the reported efficacy of adalimumab in large RCTs in AS 42 and PsA, 24 as well as in smaller trials in other SpA subtypes. 1,4,11-13,27-32 Taken together, these trials support the concept that anti-tnf-therapy is highly effective in SpA as a whole independent of the phenotypic manifestation. This concept is further strengthened by the significant improvement of measurements primarily developed for axial disease such as BASDAI and ASDAS despite the fact that the trial was designed for peripheral SpA. The efficacy of adalimumab on these parameters in our study is similar to that reported for AS in the ATLAS trial: a BASDAI50 response at week 12 was achieved in 42% of our study population compared with 45% reached in the Adalimumab Trial Evaluating Long-term Efficacy and Safety for Ankylosing Spondylitis (ATLAS) trial, 42 whereas at week 24, ASDAS inactive disease was reached in respectively 58% and 55% of the patients. 43 The results of our trial raise three additional issues with regard to peripheral SpA. First, we used the ESSG and Amor criteria as the ASAS criteria for peripheral SpA were not published yet when the study was started. Post hoc application of these criteria to our study population, however, indicated that 38 out of the 40 patients also fulfilled the ASAS criteria for peripheral SpA. Albeit presenting primarily with peripheral disease, 22 out of 40 patients also had inflammatory back pain when actively questioned for axial complaints. Whereas strict application of the ASAS criteria would thus imply that the criteria for axial SpA should be used rather than peripheral SpA, the overlap in signs and symptoms between different SpA phenotypes observed in the present study fits with the significant improvement in BASDAI and ASDAS in these patients with predominantly peripheral SpA, and indicates that the subclassification in axial versus peripheral SpA may not always completely cover the clinical picture. It remains to be determined whether global measurements such as patient s and physician s global assessment, 1,4 peripheral measurements such as SJC, or primarily axial measurements such as BASDAI and ASDAS are the most appropriate to monitor disease activity in SpA patients with a mixed phenotype. A second issue is whether the improvement in disease activity also translates in improved function and QoL in non-as, non-psa peripheral SpA. Functional outcome parameters have been developed and validated for AS and are also applied to axial SpA but similar parameters are not available for peripheral SpA. Interestingly, also at the functional level we noticed a significant improvement of axial outcome parameters such as the lumbar mobility measured by the modified Schober index in these patients presenting primarily with peripheral disease (data not shown). As to disability and QoL, both the HAQ-DI and the HUI-3 were significantly improved at week 24. This has medical and socio-economic implications as previous studies in AS have indicated that QoL scores are associated with working status and thus with economic and societal costs Adalimumab in peripheral SpA 91

93 Table 4. Adverse events through week 24 by treatment group Adalimumab Week 0-12 Placebo Number of patients with AEs Total number of AEs Infections Common cold 4 3 Gingivitis 0 1 Cystitis 0 2 Septic arthritis 0 1 Dermatomycosis 0 1 Skin Diffuse rash 2 0 Other 0 2 Gastrointestinal Nausea 1 0 Psychologic Depression 1 0 Other 4 2 Total number of SAEs 1 1 Death 1 0 Hospital admission 0 1 A third issue is that the peripheral SpA patients included in the study presented with arthritis rather than dactylitis or enthesitis (table 1). The low frequency of dactylitis, with only three patients having active dactylitis at baseline, may be related to the fact that we excluded patients fulfilling the ClASsification criteria for Psoriatic ARthritis (CASPAR) criteria for PsA. In contrast, enthesitis was noted in the history of 75% of the patients. However, only 25% of the patients had active enthesitis at inclusion, thereby precluding a detailed analysis of this manifestation in our trial. However, a previous study including specifically SpA patients with refractory heel enthesitis has indicated the efficacy of TNF-blockade by etanercept for this peripheral SpA manifestation. 45 The number of SAEs, AEs and infections was similar in the adalimumab treated patients compared with the placebo treated patients. The number of SAEs was quite high in our cohort (three out of 40 patients) but two out of three SAEs (suicide and acute psychosis) were not considered to be related to the study drug. The third SAE, a septic arthritis, was most probably triggered by needle arthroscopy and occurred in the placebo group and not the adalimumab group. Although safety data remain difficult to interpret correctly in relatively small proof-ofconcept trials, we did not observe unexpected safety signals. In conclusion, the results of this randomised double-blind, placebo-controlled clinical trial indicate that SpA patients with peripheral disease who do not fulfil the criteria for AS or PsA may 92

94 Table 4. Continued Adalimumab Week Number of patients with AEs 8 11 Total number of AEs 9 13 Infections Skin Common cold 3 3 Sinusitis 0 1 Otitis 0 1 Tooth abcess 1 1 Cystitis 0 3 Injection-site reaction 0 1 Other 0 1 Gastrointestinal Diarrhea 0 1 Bloating 1 0 Psychologic Depression 1 0 Psychosis 0 1 Other 3 0 Total number of SAEs 0 1 Hospital admission 0 1 Values are the number of events. Patients may be counted in more than 1 adverse event category and may be counted more than once within one sort adverse event. AE = adverse event; SAE = serious adverse event. 6 Adalimumab in peripheral SpA benefit from adalimumab treatment in terms of disease activity and QoL. This warrants further confirmation in larger studies. Acknowledgements We thank Abbott for the supply of the study medication for this investigator initiated and independent study. Also we would like to thank the following rheumatologists for referring patients: Dr ZN Jahangier, Dr G Collee, Dr MHW de Bois and Drs A van Sijl. We thank AN Scholten, A van Tillo, Dr DM Gerlag, R Heinrich, A Giesbers and L Gomes for their help in this study, and Professor Dr R Landewé for critical reading of the manuscript. Professor Dr DL Baeten is supported by a Vidi grant from The Netherlands Organization for Scientific Research (NWO). 93

95 REFERENCES 1 Van den Bosch F, Kruithof E, Baeten D, et al. Effects of a loading dose regimen of three infusions of chimeric monoclonal antibody to tumour necrosis factor alpha (infliximab) in spondyloarthropathy: an open pilot study. Ann Rheum Dis 2000;59: Brandt J, Haibel H, Cornely D, et al. Successful treatment of active ankylosing spondylitis with the anti-tumor necrosis factor alpha monoclonal antibody infliximab. Arthritis Rheum 2000;43: Mease PJ, Goffe BS, Metz J, et al. Etanercept in the treatment of psoriatic arthritis and psoriasis: a randomised trial. Lancet 2000;356: Van den Bosch F, Kruithof E, Baeten D, et al. Randomized double-blind comparison of chimeric monoclonal antibody to tumor necrosis factor alpha (infliximab) versus placebo in active spondylarthropathy. Arthritis Rheum 2002;46: Braun J, Brandt J, Listing J, et al. Treatment of active ankylosing spondylitis with infliximab: a randomised controlled multicentre trial. Lancet 2002;359: Dougados M, Van der Linden S, Juhlin R, et al. The European Spondylarthropathy Study Group preliminary criteria for the classification of spondylarthropathy. Arthritis Rheum 1991;34: Dougados M, Baeten D. Spondyloarthritis. Lancet 2011;377: Van der Linden S, Valkenburg HA, Cats A. Evaluation of diagnostic criteria for ankylosing spondylitis. A proposal for modification of the New York criteria. Arthritis Rheum 1984;27: Rudwaleit M, Haibel H, Baraliakos X, et al. The early disease stage in axial spondylarthritis: results from the German Spondyloarthritis Inception Cohort. Arthritis Rheum 2009;60: Rudwaleit M, van der Heijde D, Landewe R, et al. The development of Assessment of SpondyloArthritis international Society classification criteria for axial spondyloarthritis (part II): validation and final selection. Ann Rheum Dis 2009;68: Haibel H, Rudwaleit M, Listing J, et al. Efficacy of adalimumab in the treatment of axial spondylarthritis without radiographically defined sacroiliitis: results of a twelve-week randomized, double-blind, placebo-controlled trial followed by an open-label extension up to week fifty-two. Arthritis Rheum 2008;58: Barkham N, Keen HI, Coates LC, et al. Clinical and imaging efficacy of infliximab in HLA-B27-Positive patients with magnetic resonance imagingdetermined early sacroiliitis. Arthritis Rheum 2009;60: Song IH, Hermann K, Haibel H, et al. Effects of etanercept versus sulfasalazine in early axial spondyloarthritis on active inflammatory lesions as detected by whole-body MRI (ESTHER): a 48- week randomised controlled trial. Ann Rheum Dis 2011;70: Sieper J, Van der Heijde D, Dougados M, et al. Efficacy and safety of adalimumab in patients with non-radiographic axial spondyloarthritis: results o a randomized placebo-controlled trial (ABILITY-1). Ann Rheum Dis 2013;72: Said-Nahal R, Miceli-Richard C, D Agostino MA, et al. Phenotypic diversity is not determined by independent genetic factors in familial spondylarthropathy. Arthritis Rheum 2001;45: Said-Nahal R, Miceli-Richard C, Berthelot JM, et al. The familial form of spondylarthropathy: a clinical study of 115 multiplex families. Groupe Francais d Etude Genetique des Spondylarthropathies. Arthritis Rheum 2000;43: Porcher R, Said-Nahal R, D Agostino MA, et al. Two major spondylarthropathy phenotypes are distinguished by pattern analysis in multiplex families. Arthritis Rheum 2005;53: Kruithof E, Baeten D, De Rycke L, et al. Synovial histopathology of psoriatic arthritis, both oligo- and polyarticular, resembles spondyloarthropathy more than it does rheumatoid arthritis. Arthritis Res Ther 2005;7:R Vandooren B, Noordenbos T, Ambarus C, et al. Absence of a classically activated macrophage cytokine signature in peripheral spondylarthritis, including psoriatic arthritis. Arthritis Rheum 2009;60: Noordenbos T, Yeremenko N, Gofita I, et al. Interleukin-17-positive mast cells contribute to synovial inflammation in spondylarthritis. Arthritis Rheum 2012;64: Tran TM, Dorris ML, Satumtira N, et al. Additional human beta2-microglobulin curbs HLA-B27 misfolding and promotes arthritis and spondylitis without colitis in male HLA-B27-transgenic rats. Arthritis Rheum 2006;54: Rudwaleit M, Van der Heijde D, Landewe R, et al. The Assessment of SpondyloArthritis International Society classification criteria for peripheral spondyloarthritis and for spondyloarthritis in general. Ann Rheum Dis 2011;70:

96 23 Antoni C, Krueger GG, de Vlam K, et al. Infliximab improves signs and symptoms of psoriatic arthritis: results of the IMPACT 2 trial. Ann Rheum Dis 2005;64: Mease PJ, Gladman DD, Ritchlin CT, et al. Adalimumab for the treatment of patients with moderately to severely active psoriatic arthritis: results of a double-blind, randomized, placebocontrolled trial. Arthritis Rheum 2005;52: Goedkoop AY, Kraan MC, Picavet DI, et al. Deactivation of endothelium and reduction in angiogenesis in psoriatic skin and synovium by low dose infliximab therapy in combination with stable methotrexate therapy: a prospective singlecentre study. Arthritis Res Ther 2004;6:R Van Kuijk AW, Gerlag DM, Vos K, et al. A prospective, randomised, placebo-controlled study to identify biomarkers associated with active treatment in psoriatic arthritis: effects of adalimumab treatment on synovial tissue. Ann Rheum Dis 2009;68: Kruithof E, De Rycke L, Roth J, et al. Immunomodulatory effects of etanercept on peripheral joint synovitis in the spondylarthropathies. Arthritis Rheum 2005;52: Brandt J, Haibel H, Reddig J, et al. Successful short term treatment of severe undifferentiated spondyloarthropathy with the anti-tumor necrosis factor-alpha monoclonal antibody infliximab. J Rheumatol 2002;29: Brandt J, Khariouzov A, Listing J, et al. Successful short term treatment of patients with severe undifferentiated spondyloarthritis with the antitumor necrosis factor-alpha fusion receptor protein etanercept. J Rheumatol 2004;31: Marzo-Ortega H, McGonagle D, O Connor P, et al. Efficacy of etanercept in the treatment of the entheseal pathology in resistant spondylarthropathy: a clinical and magnetic resonance imaging study. Arthritis Rheum 2001;44: Marzo-Ortega H, McGonagle D, O Connor P, et al. Efficacy of etanercept for treatment of Crohn s related spondyloarthritis but not colitis. Ann Rheum Dis 2003;62: Meyer A, Chatelus E, Wendling D, et al. Safety and efficacy of anti-tumor necrosis factor alpha therapy in ten patients with recent-onset refractory reactive arthritis. Arthritis Rheum 2011;63: Amor B, Dougados M, Mijiyawa M. Criteria of the classification of spondylarthropathies. Rev Rhum Mal Osteoartic 1990;57: Taylor W, Gladman D, Helliwell P, et al. Classification criteria for psoriatic arthritis: development of new criteria from a large international study. Arthritis Rheum 2006;54: Garrett S, Jenkinson T, Kennedy LG, et al. A new approach to defining disease status in ankylosing spondylitis: the Bath Ankylosing Spondylitis Disease Activity Index. J Rheumatol 1994;21: Lukas C, Landewe R, Sieper J, et al. Development of an ASAS-endorsed disease activity score (ASDAS) in patients with ankylosing spondylitis. Ann Rheum Dis 2009;68: Machado P, Landewe R, Lie E, et al. Ankylosing Spondylitis Disease Activity Score (ASDAS): defining cut-off values for disease activity states and improvement scores. Ann Rheum Dis 2011;70: Braun J, Davis J, Dougados M, et al. First update of the international ASAS consensus statement for the use of anti-tnf agents in patients with ankylosing spondylitis. Ann Rheum Dis 2006;65: Bruce B, Fries JF. The Health Assessment Questionnaire (HAQ). Clin Exp Rheumatol 2005;23(Suppl 39):S Furlong WJ, Feeny DH, Torrance GW, et al. The Health Utilities Index (HUI) system for assessing health-related quality of life in clinical studies. Ann Med 2001;33: Gooch K, Feeny D, Wong RL, et al. Is the Health Utilities Index 3 valid for patients with ankylosing spondylitis? Value Health 2011;14: Van der Heijde D, Kivitz A, Schiff MH, et al. Efficacy and safety of adalimumab in patients with ankylosing spondylitis: results of a multicenter, randomized, double-blind, placebo-controlled trial. Arthritis Rheum 2006;54: Sieper J, Van der Heijde D, Dougados M, et al. Early response to adalimumab predicts longterm remission through 5 years of treatment in patients with ankylosing spondylitis. Ann Rheum Dis 2012;71: Maksymowych WP, Gooch KL, Wong RL, et al. Impact of age, sex, physical function, healthrelated quality of life, and treatment with adalimumab on work status and work productivity of patients with ankylosing spondylitis. J Rheumatol 2010;37: Dougados M, Combe B, Braun J, et al. A randomised, multicentre, double-blind, placebocontrolled trial of etanercept in adults with refractory heel enthesitis in spondyloarthritis: the HEEL trial. Ann Rheum Dis 2010;69: Adalimumab in peripheral SpA 95

97

98 Fast relapse upon discontinuation of tumour necrosis factor blocking therapy in patients with peripheral spondyloarthritis Jacqueline E. Paramarta, Tanja F. Heijda, Dominique L. Baeten Department of Clinical Immunology and Rheumatology, Academic Medical Center/University of Amsterdam, Amsterdam, The Netherlands 7 Ann Rheum Dis Sep 1;72(9):1581-2

99

100 Tumour necrosis factor (TNF) blockade is effective in axial spondyloarthritis (SpA), including both ankylosing spondylitis and non-radiographic axial SpA, 1 as well as peripheral SpA, which comprises psoriatic arthritis (PsA) but also other SpA subtypes. 2 It is well established that anti-tnf therapy discontinuation leads to fast relapse in almost all axial SpA patients. 3-8 This study aimed to investigate if similar relapses are seen after anti-tnf therapy discontinuation in peripheral SpA. Twenty-six patients from our randomised clinical trial with adalimumab in peripheral arthritis in non-as, non-psa SpA 2 were included. Patients had received either 12 (n=12) or 24 weeks (n=14) of adalimumab 2 before discontinuation of the anti-tnf therapy. After discontinuation, patients were followed for 16 weeks and seen for a relapse visit upon worsening of symptoms. Relapse was defined as increase of 1 swollen joint, or 2 points in patient s or physician s global assessment of disease activity or Bath Ankylosing Spondylitis Disease Activity Index. At the relapse visit, or in absence of relapse at the 16 weeks follow-up visit, disease activity parameters were measured. The study was approved by the local ethics committee. At the time point of adalimumab discontinuation, the disease activity was low (table 1). In all, 11 patients (42.3%) had reached a 66 swollen joint count (SJC66) of zero and 14 patients (53.8%) Ankylosing Spondylitis Disease Activity Score (ASDAS) inactive disease. After adalimumab discontinuation, 19 patients (73.1%) relapsed after a mean of 10.0±3.2 weeks. Only four patients (16.0%) maintained a SJC66 of zero or ASDAS inactive disease over 16 weeks. At the group level, there was a significant increase in all disease activity parameters after adalimumab interruption (table 1). The number of patients with relapse increased over time (figure 1A). Relapse was largely based on an increase in SJC66 (12/19 patients), reflecting the peripheral SpA phenotype of this population, but was also associated with worsening of systemic and axial parameters as most SpA patients depict a combination of different disease manifestations. 2 Univariate analysis Table 1. Disease activity upon anti-tnf therapy discontinuation (baseline) and at the follow-up visit (upon relapse or, in absence of relapse, at 16 weeks after discontinuation) Baseline Follow-up P-value Patient s global assessment, mm VAS 27.6 (23.4) 52.4 (25.8) <0.001 Physician s global assessment, mm VAS 24.0 (19.4) 39.3 (18.1) <0.001 Swollen joint count, 0-66 joints 0.8 (0.8) 2.1 (2.2) Tender joint count, 0-68 joints 5.3 (6.8) 10.7 (12.1) BASDAI 2.9 (2.3) 4.9 (2.5) <0.001 ASDAS 1.6 (0.8) 2.5 (1.0) <0.001 CRP, mg/l 2.2 (2.5) 4.4 (7.8) ESR, mm/hour 4.7 (2.8) 7.0 (8.3) Swollen joint count = 0, n (%) 11 (42.3) 4 (16.0) ASDAS inactive disease, n (%) 14 (53.8) 4 (16.0) Relapse after adalimumab discontinuation Except where indicated otherwise, values are the mean (standard deviation). VAS = visual analogue scale; BASDAI = Bath Ankylosing Spondylitis Disease Activity Index; ASDAS = Ankylosing Spondylitis Disease Activity Score; CRP = C-reactive protein; ESR = erythrocyte sedimentation rate. Significance of the comparisons is determined by paired t-tests. 99

101 did not identify parameters (including treatment duration, disease activity parameters and demographic or clinical characteristics) to be significantly associated with the occurrence of relapse. In particular, longer duration of adalimumab treatment, SJC66 of zero or ASDAS inactive disease at the time point of adalimumab discontinuation did not result in lower relapse rates (figure 1B D). However, time to relapse correlated with the duration of adalimumab treatment (R=0.722, p<0.001 assessed by Pearson correlation test) and SJC66 at the time point of adalimumab interruption (R= 0.585, p=0.002). In conclusion, our data show a rapid relapse in more than 70% of the peripheral SpA patients within 16 weeks after interruption of TNF blockade. Even patients with complete remission of arthritis or reaching ASDAS inactive disease did rapidly flare. This is in agreement with findings in axial SpA 3-8 and preliminary findings in combined axial and peripheral undifferentiated SpA. 9 We therefore hypothesise that rapid relapse upon anti-tnf therapy discontinuation is a general SpA feature. It needs to be further investigated whether the patients not flaring within 16 weeks can maintain a longer drug-free remission. Figure 1. Cumulative percentage of peripheral spondyloarthritis patients who experienced relapse upon follow-up after discontinuation of treatment with adalimumab. The panel represents the total group of patients (A), and the sub-analyses for the duration of adalimumab treatment before anti-tnf treatment discontinuation (B), the number of swollen joints at week 24 and the achievement of Ankylosing Spondylitis Disease Activity Score (ASDAS) inactive disease at week 24 (D). SJC = swollen joint count 66; Wk = week; ASDAS active = ASDAS 1.3; ASDAS inactive = ASDAS<

102 Acknowledgements We thank Abbott for the supply of the study medication for this investigator initiated and independent study. Also we would like to thank A van Tillo for her help in this study. DLB is supported by a Vidi grant from The Netherlands Organisation for Scientific Research (NWO). REFERENCES 1 Song IH, Weiss A, Hermann KG, et al. Similar response rates in patients with ankylosing spondylitis and non-radiographic axial spondyloarthritis after 1 year of treatment with etanercept: results from the ESTHER trial. Ann Rheum Dis 2013;72: Paramarta JE, De Rycke L, Heijda TF, et al. Efficacy and safety of adalimumab for the treatment of peripheral arthritis in spondyloarthritis patients without ankylosing spondylitis or psoriatic arthritis. Ann Rheum Dis. Published Online First 8 November doi: /annrheumdis Baraliakos X, Listing J, Brandt J, et al. Clinical response to discontinuation of anti-tnf therapy in patients with ankylosing spondylitis after 3 years of continuous treatment with infliximab. Arthritis Res Ther 2005;7:R Brandt J, Khariouzov A, Listing J, et al. Six-month results of a double-blind, placebo-controlled trial of etanercept treatment in patients with active ankylosing spondylitis. Arthritis Rheum 2003;48: Brandt J, Listing J, Haibel H, et al. Longterm efficacy and safety of etanercept after readministration in patients with active ankylosing spondylitis. Rheumatology (Oxford) 2005;44: Breban M, Vignon E, Claudepierre P, et al. Efficacy of infliximab in refractory ankylosing spondylitis: results of a six-month open-label study. Rheumatology (Oxford) 2002;41: Heldmann F, Brandt J, van der Horst-Bruinsma IE, et al. The European ankylosing spondylitis infliximab cohort (EASIC): a European multicentre study of long term outcomes in patients with ankylosing spondylitis treated with infliximab. Clin Exp Rheumatol 2011;29: Song IH, Althoff CE, Haibel H, et al. Frequency and duration of drug-free remission after 1 year of treatment with etanercept versus sulfasalazine in early axial spondyloarthritis: 2 year data of the ESTHER trial. Ann Rheum Dis 2012;71: Brandt J, Khariouzov A, Listing J, et al. Successful short term treatment of patients with severe undifferentiated spondyloarthritis with the antitumor necrosis factor-alpha fusion receptor protein etanercept. J Rheumatol 2004;31: Relapse after adalimumab discontinuation 101

103

104 Adalimumab serum levels and antidrug antibodies towards adalimumab in peripheral spondyloarthritis: No association with clinical response to treatment or with disease relapse upon treatment discontinuation Jacqueline E. Paramarta, Dominique L. Baeten 8 Department of Clinical Immunology and Rheumatology, Academic Medical Center/University of Amsterdam, Amsterdam, The Netherlands Arthritis Res Ther Jul 29;16(4):R160 [Epub ahead of print]

105 ABSTRACT Introduction To evaluate the clinical relevance of serum drug levels and anti-drug antibodies (ADAbs) in response to treatment as well as relapse upon treatment discontinuation in peripheral spondyloarthritis (pspa) patients treated with adalimumab. Methods The study included 26 peripheral SpA patients treated with adalimumab for either 12 (n=12) or 24 (n=14) weeks in a randomized controlled trial. Patients achieving ASDAS inactive disease at the end of the treatment period were classified as responders. Clinical characteristics, serum trough adalimumab levels and ADAbs were assessed at the end of the treatment period and at follow-up (upon relapse or, in absence of relapse, at 16 weeks after discontinuation). Results Serum adalimumab levels measured 2 weeks after the last adalimumab administration ranged from <0.002 to 23.0 ug/ml, with a median of 11.5 ug/ml. These levels were neither associated with response to treatment or disease activity measurements at end of treatment, nor with the occurrence of relapse and time to relapse after discontinuation of treatment. Anti-adalimumab ADAbs were present in 23% of the patients at end of treatment and in 35% at follow-up after treatment discontinuation, indicating that ADAbs have been masked by the presence of the drug in some patients. However, ADAbs at end of treatment as well as at follow-up were not different between responders and non-responders and were not associated with relapse upon discontinuation of treatment. Conclusions There is no clear association between adalimumab serum levels or anti-adalimumab ADAbs and clinical response to treatment or with relapse upon treatment discontinuation in pspa. 104

106 INTRODUCTION Tumor necrosis factor (TNF) inhibition is a highly effective treatment for axial and peripheral spondyloarthritis (SpA). 1-5 However, a significant proportion of patients fails to respond or does not tolerate the treatment because of side effects. Reasons for non-response or for intolerance are multiple, including potentially the development of anti-drug antibodies (ADAbs) directed towards the TNF blocker. It has been proposed that ADAbs may reduce therapeutic responses either by increasing the clearance of the TNF inhibitor 6 or by direct neutralisation of the functional part of the drug. 7 Accordingly, recent reviews suggest that monitoring of serum drug levels and ADAbs would be a promising tool for personalised cost-effective usage of biological therapies in immune-mediated inflammatory diseases (IMIDs). 8,9 Most studies on immunogenicity of TNF blockers have been performed in rheumatoid arthritis and Crohn s disease. In SpA, the available studies on immunogenicity yielded conflicting results. For infliximab and adalimumab some groups reported that ADAbs towards these TNF inhibitors are associated with decreased clinical response and increased risk of hypersensitivity reactions, while others do not find this association and even conclude that serum anti-tnf drug levels are not associated with response to treatment. 17,18 For etanercept these ADAbs have not been detected and it has been shown that the serum drug levels are similar in responders and non-responders. 19 Recently, also with golimumab ADAbs do not appear to have a major role in treatment success or failure. 20,21 Moreover, a recent meta-analysis on anti-tnf ADAbs in various IMIDs concluded that there was no relevant association of ADAbs with efficacy in SpA. 22 This corresponds with the clinical experience that the treatment failure to TNF blockade is similar among the various TNF blockers, 23,24 both the ones which do and do not cause ADAbs according to the before mentioned studies These conflicting results may be related to the diversity in methods and timing to evaluate the ADAbs as well as to the fact that the presence of detectable serum drug levels may mask the detection of ADAbs. 8,9 The latter issue can be avoided by the use of novel assay methods 25 and/or by assessing the ADAbs several weeks after stopping the TNF inhibitor. In this study, we assessed the potential clinical relevance of serum drug levels and ADAbs measured at the end of the treatment period as well as at a drug-free follow-up after a double-blind, placebocontrolled, randomized clinical trial (RCT) with adalimumab in peripheral SpA (pspa). 5 We correlated these serum levels both with clinical response at the end of treatment and with relapse upon discontinuation of the TNF inhibitor. METHODS Study patients Twenty six patients from our RCT with adalimumab in peripheral arthritis in SpA patients not fulfilling the criteria for ankylosing spondylitis (AS) or psoriatic arthritis (PsA) 5 were included in this study. The patients fulfilled the European Spondyloarthritis Study Group (ESSG) 26 and/or the Amor criteria 27 and had active disease at inclusion into the study. They were treated with either placebo (n=12) or adalimumab (n=14) for 12 weeks, followed by a 12 week open-label phase with adalimumab for all patients. 5 After this period adalimumab was discontinued. Patients were 8 ANTI-DRUG antibodies: no association with clinical response 105

107 allowed to continue non-steroidal anti-inflammatory drugs (NSAIDs), corticosteroids ( 10 mg/ day prednisone or equivalent), methotrexate and sulfasalazine on a stable dosage throughout the study. After discontinuation of the TNF inhibitor, patients were prospectively followed for 16 weeks and seen for a relapse visit upon worsening of symptoms or, in the absence of relapse, at the 16 weeks follow-up visit. The following disease activity parameters were measured: patient s and physician s global assessment of disease activity, 68/66 tender joint count (TJC) and swollen joint count (SJC), Bath Ankylosing Spondylitis Disease Activity Index (BASDAI), Ankylosing Spondylitis Disease Activity Score (ASDAS), erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP). Responders were defined as patients achieving ASDAS inactive disease at the end of treatment. 28 Relapse was defined as increase of 1 swollen joint, or 2 points in patient s or physician s global assessment of disease activity or BASDAI. 29 Fourteen patients (53.8%) reached ASDAS inactive disease at the end of the treatment period. Nineteen (73.1%) patients had a disease relapse within 16 weeks after discontinuation of adalimumab with a mean time to relapse of 10.0±3.2 weeks. The characteristics of the patients have been published previously. 29 Written informed consent was obtained from each patient before study-related procedures were performed and the study was approved by the Medical Ethics Committee of the Academic Medical Center/University of Amsterdam. Assessment of serum adalimumab levels Serum samples were obtained two weeks after the last adalimumab administration (n=26) and at follow-up (upon relapse or 16 weeks after discontinuation in the absence of relapse, n=25). Trough serum adalimumab concentrations were measured by enzyme-linked immunosorbent assay (ELISA) developed by Wolbink and co-workers as previously described and as accredited by the RvA/CCKL (Dutch Accreditation Council/Dutch Accreditation Board for Medical Laboratories) according to the International Standardization Organization (ISO) guideline ISO The detection limit of the assay is approximately ug/ml; serum adalimumab levels <5.0 ug/ml were designated low, as previously described. 15,30 Assessment of antibodies against adalimumab The same serum samples were analyzed by a radio immunoassay (RIA) (Sanquin) to detect the presence of anti-adalimumab antibodies as previously described. 30 After dilution of 1 μl of serum in phosphate-buffered saline/0.3% bovine serum albumin (pro analysi buffer), overnight incubation followed with 1 mg Sepharose-immobilized protein A (GE Health Care, Giles, England) in a final volume of 800 μl. Then, the samples were washed with phosphatebuffered saline 0.005% polysorbate. The anti-adalimumab binding was determined by overnight incubation with 20,000 disintegrations per minute (dpm [ 1 ng]) iodine 125-labeled F(ab)2 adalimumab diluted in Freeze buffer (Sanquin). Unbound label was removed by washing, and protein A-bound radioactivity was measured. Anti-adalimumab levels were expressed in arbitrary units (AU [1 AU 12 ng]) using a serum containing anti-adalimumab as standard. The mean cut-off value as derived from 100 healthy donors was set at 12 AU/mL. 106

108 Statistical analysis Data are presented as median and interquartile range (IQR). Mann-Whitney U tests were used to compare differences in serum levels in case of unpaired samples and Wilcoxon signed rank tests in case of paired samples. Chi square tests were used for categorical variables. Logistic regression analyses were conducted to examine associations between the ASDAS response, relapse status and trough serum adalimumab and anti-adalimumab ADAbs levels. Correlations between the serum measurements and the clinical disease activity measurements were assessed using Spearman s correlation tests. All statistical tests were 2-sided and P-values <0.05 were considered statistically significant. RESULTS Clinical response to treatment and relapse after anti-tnf treatment discontinuation are independent of trough serum adalimumab levels Trough serum adalimumab levels at the end of the treatment period (two weeks after the last injection) ranged from <0.002 to 23.0 ug/ml (median 11.5 ug/ml). Seven patients (26.9%) had serum adalimumab levels of <5.0 ug/ml. Levels were not different between patients treated with adalimumab for 12 or 24 weeks (p=0.292) (Figure 1A). There were no significant differences in trough serum adalimumab levels between responders (12.6 (IQR ) ug/ml) and nonresponders (9.3 (IQR ) ug/ml) as defined by the achievement of ASDAS inactive disease (p=0.237) (Figure 1B). Serum adalimumab levels did also not correlate with end-of-study disease activity parameters such as patient s global assessment (Figure 1C) and physician s global assessment of disease activity, TJC, SJC, BASDAI, ASDAS, ESR (data not shown) and CRP (Figure 1D). Moreover, adalimumab levels at the end of treatment were not different between patients with and without subsequent relapse upon discontinuation of therapy (p=0.931) (Figure 2A) and were not correlated with time to relapse (p=0.984) (Figure 2B). Taken together, these data indicate that the amplitude and/or duration of clinical response to adalimumab in pspa are not related to trough serum adalimumab levels. Clinical response to treatment and relapse after anti-tnf treatment discontinuation are independent of the presence of anti-adalimumab ADAbs At the end of the treatment period 6/26 patients (23.1%) tested positive for serum anti-adalimumab ADAbs: 4 were clearly positive with titres ranging from 89 to 2320 AU/mL and 2 were borderline positive both with a titre of 15 AU/mL. The presence of detectable anti-adalimumab ADAbs levels was similar between patients treated 12 (4/12 patients; 33.3%) or 24 weeks (2/14 patients; 14.3%) with adalimumab (p=0.250) (Figure 3A), and between responders (3/14 patients; 21.4%) and nonresponders (3/12 patients; 25.0%) (p=0.829) (Figure 3B). The anti-adalimumab ADAb titres did not correlate with the various disease activity measurements (data not shown). Finally, the number of patients who tested positive for anti-adalimumab ADAb was not different between those with and without subsequent relapse upon discontinuation of therapy (p=0.518) (Figure 4A) nor was there a difference in time to relapse (p=0.488) (Figure 4B). As for the trough serum adalimumab levels, 8 ANTI-DRUG antibodies: no association with clinical response 107

109 Figure 1. Clinical response to treatment is independent of trough serum adalimumab levels. Trough serum adalimumab levels at end of treatment (two weeks after the last injection) were not different between patients treated with 12 or 24 weeks of adalimumab (A) or between responders and nonresponders (B). Also there was no correlation between these drug levels and clinical disease activity parameters such as patient s global assessment of disease activity measured on a 100 mm visual analogue scale (C) or C-reactive protein (CRP) (D). our data thus do not provide any evidence that anti-adalimumab ADAbs have a significant impact on the amplitude and/or duration of clinical response to adalimumab in pspa. Anti-adalimumab ADAbs can be masked by the presence of adalimumab but also unmasked anti-adalimumab ADAbs do not correlate with clinical response to treatment As several factors can bias the measurement and/or interpretation of ADAbs, we conducted additional analyses to assess the potential clinical relevance of these antibodies. Firstly, there was a negative correlation between the anti-adalimumab ADAb titres and the serum trough adalimumab levels (R= 0.709, p<0.001). This may be explained by the fact that ADAbs contribute to the clearance of the drug or, alternatively, by the fact that serum adalimumab levels may interfere with the measurement of the ADAbs. 8,9 To investigate the latter possibility, we obtained additional serum samples at follow-up in 25/26 patients at 10.1 (IQR ) weeks after interruption of the adalimumab treatment. At this time point the median serum 108

110 Figure 2. Relapse after treatment discontinuation is independent of trough serum adalimumab levels. Trough serum adalimumab levels at end of treatment (two weeks after the last injection) were similar in patients who did or did not relapsed after discontinuation of the TNF inhibitor (A). Neither was there a correlation between the drug levels and time to relapse (B). 8 Figure 3. Clinical response to treatment is independent of the presence of anti-adalimumab ADAbs. Anti-adalimumab anti-drug antibodies (ADAbs) at the end of treatment, two weeks after the last injection, were not different between patients treated with 12 or 24 weeks of adalimumab (A) or between responders or non-responders (B). adalimumab level was 1.7 (IQR ) ug/ml, which was significantly lower than at end of treatment (p<0.0001) (Figure 5A). Measurement of anti-adalimumab ADAbs at the same timepoint, allowing to exclude interference of circulating adalimumab levels, showed that the 4 patients with high ADAbs at end of treatment maintained high ADAbs levels after interruption of treatment, whereas in the 3 patients with low ADAbs titres at end of treatment the titres tended to increase over time (Figure 5B). Additionally, 3 patients without detectable ADAbs at end of treatment depicted titres between 15 and 57 AU/mL at follow-up, confirming that low to intermediate levels of ADAbs may be partially masked by circulating adalimumab. However, also these unmasked anti-adalimumab ADAbs were not different between between responders ANTI-DRUG antibodies: no association with clinical response 109

111 Figure 4. Relapse after treatment discontinuation is independent of the presence of anti-adalimumab ADAbs. Anti-adalimumab anti-drug antibodies (ADAbs) at the end of treatment, two weeks after the last injection, were not different between patients who did or did not relapsed after discontinuation of adalimumab (A). There was no difference in time to relapse in patients who tested positive or negative for anti-adalimumab ADAbs (B). (33.3% positive for unmasked anti-adalimumab) and non-responders (40.0% positive for unmasked anti-adalimumab) (p=0.734), or between patients who did or did not experienced relapse after treatment discontinuation (respectively 42.1% and 16.7%, p=0.258). The titres of the unmasked ADAbs were neither correlated with the various disease activity measurements nor the time to relapse (data not shown). Secondly, we investigated the effect of concomitant treatment with disease-modifying antirheumatic drugs (DMARDs) on ADAbs since this has been described to reduce the frequency of ADAb formation. 8,9,14,31 Although numerically different, there was no statistical difference in antiadalimumab ADAbs positivity between patients without DMARDs (5/15 patients, 33.3%) versus patients with DMARDs (1/11 patients, 9.1%) (p=0.147) (Figure 6A). Strikingly, this difference was not present anymore when analysing the unmasked anti-adalimumab ADAbs: 5/14 patients (35.7%) positive in the without DMARDs group versus 4/11 patients (36.4%) in the DMARDs group (p=0.973) (Figure 6B). Adalimumab levels and anti-adalimumab ADAbs in patients with generalized drug reactions Two patients developed a generalized skin reaction upon adalimumab during the study. Both patients were able to continue adalimumab treatment after topical corticosteroids and oral 110

112 Figure 5. Anti-adalimumab ADAbs can be masked by the presence of adalimumab in the serum. Trough serum adalimumab levels (A) and anti-adalimumab anti-drug antibodies (ADAbs) titres (B) at the end of treatment (two weeks after the last injection) and at follow-up (upon relapse or, in absence of relapse, at 16 weeks after discontinuation). *p<0.05 assessed by Wilcoxon signed rank test. 8 Figure 6. The effect of DMARDs on anti-adalimumab ADAbs. At the end of treatment, two weeks after the last injection, anti-adalimumab anti-drug antibodies (ADAbs) were less often observed in patients who used concomitant disease-modifying anti-rheumatic drugs (DMARDs) compared to patients who did not use any DMARDs, although this was not statistically significant (A). However, this difference disappeared when assessing the unmasked anti-adalimumab ADAbs at follow-up (B). antihistaminic treatment. One of these patients had a low serum trough adalimumab level (<0.002 ug/ml) and tested positive for anti-adalimumab ADAbs with a titre of 2320 AU/mL. The second patient had a serum trough adalimumab level of 17.5 ug/ml and tested negative ANTI-DRUG antibodies: no association with clinical response 111

113 for anti-adalimumab ADAbs both at the end of the study and at follow-up after interruption of treatment. Neither of the patients used concomitant DMARDs. DISCUSSION In this study we assessed the potential relevance of serum drug levels and ADAbs on various aspects of clinical response to adalimumab treatment in pspa. The major findings were that a) trough serum adalimumab levels are heterogeneous but do not correlate with clinical response to treatment or relapse after anti-tnf treatment discontinuation, b) anti-adalimumab ADAbs are found in 1/4 patients but also did not correlate with clinical response to treatment or relapse after discontinuation of the TNF inhibitor, and c) low titer ADAbs can be masked by circulating adalimumab but also unmasked ADAbs show no clear relationship with clinical efficacy. More and more research is done addressing the immunogenicity of TNF inhibitors in various IMIDs including SpA, since the development of ADAbs towards these TNF inhibitors are assumed to play a major role in loss of response to treatment. The mechanism behind this is thought to be either an increased clearance of the drug or neutralization of the active component of the compound. 6,7 This hypothesis is supported by several studies with infliximab and adalimumab, which reported that treatment failure occurred more often in patients who test positive for ADAbs However, there is also evidence which is not in line with this concept, since other studies with infliximab did not find a relation between ADAbs and response to treatment. 17,18 Moreover, one of the studies which found serum trough infliximab levels to be significantly higher in responders compared to non-responders, showed that although statistical significance was reached, the difference between these groups was very low (8.2 ug/ml versus 6.3 ug/ml, respectively). 12 Whether such a small difference is really clinically relevant is questionable. ADAbs towards etanercept have not been found nor is there an association between serum drug levels and clinical effect. 19 Likewise, for golimumab there is no clear relation between ADAbs and clinical efficacy. 20,21 Furthermore, a recent meta-analysis on various TNF-inhibitors 22 and the clinical experience that there is no difference in efficacy or drug survival between the various TNF inhibitors 23,24 also question the clinical relevance of anti-tnf ADAbs. Several factors could be devised trying to explain these differences between the various studies. Firstly, not all TNF inhibitors are assumed to be equally immunogenic. E.g. the soluble dimeric fusion protein etanercept has a less immunogenic structure since only the fusion part of the molecule can contain immunogenic epitopes. Also it is administered more frequently than the other TNF inhibitors, thereby possibly creating more drug interference in ADAb detection. 19 However, this does not explain why different studies in the same TNF inhibitor (e.g. infliximab) come to different conclusions ,17,18 Secondly, there may be differences among the different SpA subtypes. However, even in studies with both the same disease as well as the same TNF inhibitor (e.g. infliximab in AS) contradicting conclusions are made ,17,18 Hence, this could also not explain why we and others 21 did not find a clinical association with anti-tnf drug levels or ADAbs in pspa, while another study did conclude that these had clinical relevance. 15 Thirdly, there is some variation in the size and duration of the studies but this did not influence the results whether ADAbs did or did not have clinical relevance Fourthly, the use of DMARDs, 112

114 especially methotrexate, is described to decrease immunogenicity of TNF inhibitors trough a mechanism which is not yet understood. 8,9,14,31 However, this is not in line with the finding that the addition of methotrexate in the management of SpA does not have effect on the efficacy of the treatment 23,24,32 or on drug survival of the TNF inhibitor. 23,24 In our study we indeed found less ADAbs in patients using DMARDs, however this difference was not present anymore when analyzing the unmasked ADAbs. Fifthly, the detection of ADAbs is also influenced by the assay used. However, the method used in the current study is the same which has been used in other studies 15,16 making it unlikely that this is the explanation for the differences in results. And finally, the timing of the samples also influences the measurement of ADAbs since the assays are sensitive for drug interference, even when measured before the next administration of the drug when the drug levels are the lowest. 8,9 Indeed we showed that when anti-adalimumab ADAbs were measured at follow-up after discontinuation of the TNF inhibitor more patients tested positive compared to when measured at end of treatment. Previously it has been described that anti-tnf ADAb titres can decrease and increase over time and vice versa, 10,15,33 causing a gradual increase in incidence over time when ADAbs status is presented cumulatively, but not when assessed at each time point independently. 33 This shows that the timing of the measurement can influence the interpretation of the ADAbs status, making it very difficult to make strong conclusions about the relationship with clinical response, and to apply ADAbs measurement in clinical practise. The limited number of patients is a limitation of our study. We can thus not exclude that clinical correlation with serum adalimumab levels and/or ADAbs would be found in larger patient cohorts. However, this would imply that this association is weak and thus anyway not relevant for treatment monitoring in individual patients. Also, one small study which investigated whether the serum trough infliximab levels modified therapeutic decisions in the management of AS did not find an improvement in the control of disease activity. 34 Similar efficacy and drug survival of TNF inhibitors which induce and which do not induce ADAbs 23,24 also question the relevance of testing the immunogenicity of these drugs. Although testing immunogenicity in clinical trials is standard practise and may yield interesting scientific insights, the real added value of the presence or absence of ADAbs in an individual patient in clinical practice remains thus to be demonstrated. CONCLUSIONS In conclusion, the link between either serum adalimumab levels or anti-adalimumab ADAbs with clinical response or relapse is not as strong as previously assumed. This argues against the use of these parameters in monitoring drug efficacy. And although the treatment with adalimumab and other TNF inhibitors is overall very successful in SpA, future research has to be done to unravel factors which can explain differences in clinical response to further improve disease management towards a personalized and more cost-effective approach. Acknowledgements We thank AbbVie for the supply of the study medication for this investigator initiated and independent study. Also we would like to thank T.F. Heijda and A. van Tillo for their help in this study. Prof. Dr. D.L. Baeten is supported by a Vidi grant from The Netherlands Organization for Scientific Research (NWO). 8 ANTI-DRUG antibodies: no association with clinical response 113

115 REFERENCES 1 Van den Bosch F, Kruithof E, Baeten D, Herssens A, De Keyser F, Mielants H, Veys EM: Randomized double-blind comparison of chimeric monoclonal antibody to tumor necrosis factor alpha (infliximab) versus placebo in active spondylarthropathy. Arthritis Rheum 2002, 46: Braun J, Brandt J, Listing J, Zink A, Alten R, Golder W, Gromnica-Ihle E, Kellner H, Krause A, Schneider M, Sörensen H, Zeidler H, Thriene W, Sieper J: Treatment of active ankylosing spondylitis with infliximab: a randomised controlled multicentre trial. Lancet 2002, 359: Antoni C, Krueger GG, de Vlam K, Birbara C, Beutler A, Guzzo C, Zhou B, Dooley LT, Kavanaugh A: Infliximab improves signs and symptoms of psoriatic arthritis: results of the IMPACT 2 trial. Ann Rheum Dis 2005, 64: Sieper J, van der Heijde D, Dougados M, Mease PJ, Maksymowych WP, Brown MA, Arora V, Pangan AL: Efficacy and safety of adalimumab in patients with non-radiographic axial spondyloarthritis: results of a randomised placebo-controlled trial (ABILITY-1). Ann Rheum Dis 2013, 72: Paramarta JE, De Rycke L, Heijda TF, Ambarus CA, Vos K, Dinant HJ, Tak PP, Baeten DL: Efficacy and safety of adalimumab for the treatment of peripheral arthritis in spondyloarthritis patients without ankylosing spondylitis or psoriatic arthritis. Ann Rheum Dis 2013, 72: Van der Laken CJ, Voskuyl AE, Roos JC, Stigter van Walsum M, de Groot ER, Wolbink G, Dijkmans BA, Aarden LA: Imaging and serum analysis of immune complex formation of radiolabelled infliximab and anti-infliximab in responders and non-responders to therapy for rheumatoid arthritis. Ann Rheum Dis 2007, 66: Van Schouwenburg PA, van de Stadt LA, de Jong RN, van Buren EE, Kruithof S, de GE, Hart M, van Ham SM, Rispens T, Aarden L, Wolbink GJ, Wouters D: Adalimumab elicits a restricted antiidiotypic antibody response in autoimmune patients resulting in functional neutralisation. Ann Rheum Dis 2013, 72: Vincent FB, Morand EF, Murphy K, Mackay F, Mariette X, Marcelli C: Antidrug antibodies (ADAb) to tumour necrosis factor (TNF)-specific neutralising agents in chronic inflammatory diseases: a real issue, a clinical perspective. Ann Rheum Dis 2013, 72: Garces S, Demengeot J, Benito-Garcia E: The immunogenicity of anti-tnf therapy in immunemediated inflammatory diseases: a systematic review of the literature with a meta-analysis. Ann Rheum Dis 2013, 72: Arends S, Lebbink HR, Spoorenberg A, Bungener LB, Roozendaal C, van der veer E, Houtman PM, Griep EN, Limburg PC, Kallenberg CG, Wolbink GJ, Brouwer E: The formation of autoantibodies and antibodies to TNF-alpha blocking agents in relation to clinical response in patients with ankylosing spondylitis. Clin Exp Rheumatol 2010, 28: De Vries MK, Wolbink GJ, Stapel SO, de Groot ER, Dijkmans BA, Aarden LA, van der Horst-Bruinsma IE: Inefficacy of infliximab in ankylosing spondylitis is correlated with antibody formation. Ann Rheum Dis 2007, 66: De Vries MK, Wolbink GJ, Stapel SO, de Vrieze H, van Denderen JC, Dijkmans BA, Aarden LA, van der Horst-Bruinsma IE: Decreased clinical response to infliximab in ankylosing spondylitis is correlated with anti-infliximab formation. Ann Rheum Dis 2007, 66: Ducourau E, Mulleman D, Paintaud G, Miow Lin DC, Lauferon F, Ternant D, Watier H, Goupille P: Antibodies toward infliximab are associated with low infliximab concentration at treatment initiation and poor infliximab maintenance in rheumatic diseases. Arthritis Res Ther 2011, 13:R Plasencia C, Pascual-Salcedo D, Nuno L, Bonilla G, Villalba A, Peiteado D, Diez J, Nagore D, del Agua AR, Moral R, Martin-Mola E, Balsa A: Influence of immunogenicity on the efficacy of longterm treatment of spondyloarthritis with infliximab. Ann Rheum Dis 2012, 71: Van Kuijk AW, de Groot M, Stapel SO, Dijkmans BA, Wolbink GJ, Tak PP: Relationship between the clinical response to adalimumab treatment and serum levels of adalimumab and anti-adalimumab antibodies in patients with psoriatic arthritis. Ann Rheum Dis 2010, 69: De Vries MK, Brouwer E, van der Horst-Bruinsma IE, Spoorenberg A, van Denderen JC, Jamnitski A, Nurmohamed MT, Dijkmans BA, Aarden LA, Wolbink GJ: Decreased clinical response to adalimumab in ankylosing spondylitis is associated with antibody formation. Ann Rheum Dis 2009, 68: Baraliakos X, Listing J, Rudwaleit M, Brandt J, Alten R, Burmester G, Gromnica-Ihle E, Haibel H, Schewe S, Schneider M, Sörensen H, Zeidler H, Visvanathan S, Sieper J, Braun J: Safety and efficacy of readministration of infliximab after longterm continuous therapy and withdrawal in patients with ankylosing spondylitis. J Rheumatol 2007, 34:

116 18 Krzysiek R, Breban M, Ravaud P, Prejean MV, Wijdenes J, Roy C, Henry YD, Barbey C, Trappe G, Dougados M, Emilie D, French Ankylosing Spondylitis Infliximab Network: Circulating concentration of infliximab and response to treatment in ankylosing spondylitis: results from a randomized control study. Arthritis Rheum 2009, 61: De Vries MK, van der Horst-Bruinsma IE, Nurmohamed MT, Aarden LA, Stapel SO, Peters MJ, van Denderen JC, Dijkmans BA, Wolbink GJ: Immunogenicity does not influence treatment with etanercept in patients with ankylosing spondylitis. Ann Rheum Dis 2009, 68: Inman RD, Davis JC, Jr., Heijde D, Diekman L, Sieper J, Kim SI, Mack M, Han J, Visvanathan S, Xu Z, Hsu B, Beutler A, Braun J: Efficacy and safety of golimumab in patients with ankylosing spondylitis: results of a randomized, double-blind, placebo-controlled, phase III trial. Arthritis Rheum 2008, 58: Kavanaugh A, McInnes I, Mease P, Krueger GG, Gladman D, Gomez-Reino J, Papp K, Zrubek J, Mudivarthy S, Mack M, Visvanathan S, Beutler A: Golimumab, a new human tumor necrosis factor alpha antibody, administered every four weeks as a subcutaneous injection in psoriatic arthritis: Twenty-four-week efficacy and safety results of a randomized, placebo-controlled study. Arthritis Rheum 2009, 60: Maneiro JR, Salgado E, Gomez-Reino JJ: Immunogenicity of Monoclonal Antibodies Against Tumor Necrosis Factor Used in Chronic Immune-Mediated Inflammatory Conditions: Systematic Review and Meta-analysis. JAMA Intern Med 2013, 173: Glintborg B, Ostergaard M, Krogh NS, Andersen MD, Tarp U, Loft AG, Lindegaard HM, Holland- Fischer M, Nordin H, Jensen DV, Olsen CH, Hetland ML: Clinical response, drug survival, and predictors thereof among 548 patients with psoriatic arthritis who switched tumor necrosis factor alpha inhibitor therapy: results from the Danish Nationwide DANBIO Registry. Arthritis Rheum 2013, 65: Glintborg B, Ostergaard M, Krogh NS, Dreyer L, Kristensen HL, Hetland ML: Predictors of treatment response and drug continuation in 842 patients with ankylosing spondylitis treated with anti-tumour necrosis factor: results from 8 years surveillance in the Danish nationwide DANBIO registry. Ann Rheum Dis 2010, 69: Van Schouwenburg PA, Bartelds GM, Hart MH, Aarden L, Wolbink GJ, Wouters D: A novel method for the detection of antibodies to adalimumab in the presence of drug reveals hidden immunogenicity in rheumatoid arthritis patients. J Immunol Methods 2010, 362: Dougados M, van der Linden S, Juhlin R, Huitfeldt B, Amor B, Calin A, Cats A, Dijkmans B, Olivieri I, Pasero G, Veys E, Zeidler H, The European Spondyloarthropathy Study Group: The European Spondylarthropathy Study Group preliminary criteria for the classification of spondylarthropathy. Arthritis Rheum 1991, 34: Amor B, Dougados M, Mijiyawa M: [Criteria of the classification of spondylarthropathies]. Rev Rhum Mal Osteoartic 1990, 57: Machado P, Landewe R, Lie E, Kvien TK, Braun J, Baker D, van der Heijde D: Ankylosing Spondylitis Disease Activity Score (ASDAS): defining cut-off values for disease activity states and improvement scores. Ann Rheum Dis 2011, 70: Paramarta JE, Heijda TF, Baeten DL: Fast relapse upon discontinuation of tumour necrosis factor blocking therapy in patients with peripheral spondyloarthritis. Ann Rheum Dis 2013, 72: Bartelds GM, Krieckaert CL, Nurmohamed MT, van Schouwenburg PA, Lems WF, Twisk JW, Dijkmans BA, Aarden L, Wolbink GJ: Development of antidrug antibodies against adalimumab and association with disease activity and treatment failure during long-term follow-up. JAMA 2011, 305: Jani M, Barton A, Warren RB, Griffiths CE, Chinoy H: The role of DMARDs in reducing the immunogenicity of TNF inhibitors in chronic inflammatory diseases. Rheumatology (Oxford) Published Online First: 14 August doi: /rheumatology/ket Fagerli KM, Lie E, van der Heijde D, Heiberg MS, Lexberg AS, Rodevand E, Kalstad S, Mikkelsen K, Kvien TK: The role of methotrexate comedication in TNF-inhibitor treatment in patients with psoriatic arthritis: results from 440 patients included in the NOR-DMARD study. Ann Rheum Dis Published Online First: 3 January doi: /annrheumdis Hanauer SB, Wagner CL, Bala M, Mayer L, Travers S, Diamond RH, Olson A, Bao W, Rutgeerts P: Incidence and importance of antibody responses to infliximab after maintenance or episodic treatment in Crohn s disease. Clin Gastroenterol Hepatol 2004, 2: Meric JC, Mulleman D, Ducourau E, Lauferon F, Miow Lin DC, Watier H, Goupille P, Paintaud G: Therapeutic drug monitoring of infliximab in spondyloarthritis: an observational open-label study. Ther Drug Monit 2011, 33: ANTI-DRUG antibodies: no association with clinical response 115

117

118 The il-23/il-17 immune axis as a promising new target in the treatment of spondyloarthritis Nataliya Yeremenko 1,2, Jacqueline E. Paramarta 2, Dominique Baeten 1,2 1 Department of Clinical Immunology and Rheumatology and 2 Laboratory of Experimental Immunology, Academic Medical Center/ University of Amsterdam, Amsterdam, The Netherlands 9 Curr Opin Rheumatol Jul;26(4):361-70

119 ABSTRACT Purpose of review Various novel therapies for spondyloarthritis (SpA) are currently under development. In this review, we discuss the scientific rational to target the interleukin (IL)-23/IL-17 axis in SpA and give an overview of the proof-of-concept trials with drugs directed towards this axis. Recent findings Cumulative evidence from genetics (e.g. the strong genetic association with the IL-23 receptor gene), in-vitro models (e.g. the increased IL-23 production upon HLA-B27 misfolding), human expression studies (e.g. the expansion of IL-17 producing innate cells in SpA) and animal models (e.g. the increased IL-17 production in HLA-B27 transgenic rats) strongly supports the involvement of the IL-23/IL-17 axis in the pathogenesis of SpA. Ustekinumab (a monoclonal antibody directed against the common p40 subunit of IL-23 and IL-12), secukinumab, ixekizumab (both monoclonal antibodies directed against IL-17A), and brodalumab a monoclonal antibody against the IL-17RA receptor) have been recently used in proof-of-concept and randomized trials in the ankylosing spondylitis and/or psoriatic arthritis subforms of SpA, with overall very promising clinical efficacy. Summary The first results for novel drugs blocking key cytokines in the IL-23/IL-17 axis are promising in SpA and more novel compounds are upcoming. This will teach us more on the role of the IL-23/ IL-17 axis in the pathophysiology of SpA. 118

120 INTRODUCTION Spondyloarthritis (SpA) affects approximately % of the Western population and is characterized by inflammation of axial and peripheral joints, and extra-articular manifestations such as skin psoriasis and inflammatory bowel disease (IBD). The disease presentation is heterogeneous and comprises the various phenotypic subtypes ankylosing spondylitis, nonradiographic axial SpA (nr-axspa), psoriatic arthritis (PsA), IBD-related SpA (IBD-SpA), reactive arthritis (ReA) and undifferentiated peripheral SpA (USpA). 1 Although the management of SpA patients has improved enormously since the introduction of tumor necrosis factor (TNF) inhibitors more than 10 years ago, 2,3 there is still a high unmet need for novel therapeutic alternatives. Up to 40% of patients do not respond sufficiently to TNF inhibitors, either due to intolerance or inefficacy. Moreover, in patients in whom TNF blockade is effective, remission is not long-lasting since a large majority of patients experience relapse of symptoms after treatment discontinuation. 4-6 Finally, even when clinical symptoms and inflammation are completely suppressed, TNF blockade does not appear to halt new bone formation. 1 Considering the medical need for other therapeutic agents in SpA, proof-of-concept (PoC) studies have been performed with biological agents directed towards interleukin (IL)-1, IL-6, B cells and costimulatory molecules. However, none of these strategies resulted in significant clinical improvement. More recently, several lines of evidence suggest that the IL-23/IL-17 immune axis may be a promising new target in the treatment of SpA. Our knowledge of this cytokine axis has expanded dramatically over the past decade. The potential importance of the IL-23/IL-17 axis for chronic tissue inflammation was originally described in experimental models of T-cell-driven autoimmune diseases, 7 but now starts to be confirmed by PoC studies with drugs targeting this axis in a series of human immune-mediated inflammatory diseases. 8 In this review, we give a short overview of the IL-23/IL-17 axis, discuss the lines of evidence linking IL-23/IL-17 with SpA, and summarize the emerging results of clinical trials with compounds targeting IL-23/IL-17 in SpA and related inflammatory diseases. THE INTERLEUKIN-23/INTERLEUKIN-17 AXIS Studies using IL-23-deficient mice have indicated a crucial role for this cytokine in the pathogenesis of autoimmune inflammation. In particular, it has been demonstrated that animals lacking functional IL-23 are resistant to experimental autoimmune encephalomyelitis (EAE) 7 and collagen-induced arthritis. 9 Later, it has been recognized that IL-23 mediates this effect by promoting the development of a novel subset of T cells that produce the signature cytokine IL-17 (Th17). 9,10 IL-23 is secreted by activated macrophages and dendritic cells as a heterodimer that comprises a unique p19 subunit and the p40 subunit shared with IL-12 (Fig. 1). IL-23 binds to IL-23R on IL-23R-expressing cells. Engagement of IL-23 with the IL-23R complex, composed of IL-23R and IL-12Rβ1, leads to activation of signal transducer and activator of transcription 3 (STAT-3), which subsequently up-regulates the expression of retinoic acid receptor-related orphan receptor C (RORC), a Th17-specific transcriptional regulator that is critical for the expression of two members of IL-17 family IL-17A and IL-17F. IL-17A and IL-17F are by far the best characterized cytokines from the six highly conserved IL-17 family 9 The IL-23/IL-17 axis as promising new target 119

121 members (IL-17A F). IL-17A and IL-17F are both covalent homodimers, but the single subunits can also form IL-17A IL-17F heterodimers. 11 IL-17A, IL-17F and IL-17A IL-17F all signal through a heteromeric receptor complex consisting of the IL-17RA and IL-17RC subunits (Fig. 1), mainly expressed on epithelial, endothelial and fibroblast cells resulting in activation of (nuclear factor kappa-light-chain-enhancer of activated B cells) pathway, which then leads to the production of pro-inflammatory cytokines, chemokines, adhesion molecules, leukocyte recruitment and activation of endothelium hereby playing an important role in directing development of subsequent immune responses. Remarkably, IL-17A is not only a potent inducer of TNF-α, but also synergizes with this cytokine to promote induction of nearly all its target genes, and in many cases synergy has also been observed with IFN-γ and IL-1β (reviewed in 12 ). Recent studies have indicated three new important concepts in Th17/IL-17 biology. Firstly, Th17 cells display a marked functional plasticity. Depending on cytokines or pathogens present in the milieu, Th17 cells can change into interferon (IFN)-γ-producing Th1 cells or IL-4-producing Th2 cells, 13 or even coexpress IL-10, 14,15* exerting tissue-protective and immunosuppressive effects. Secondly, IL-17A is not only produced by canonical Th17 cells but also by a variety of innate and acquired immune cells, including γδ T cells, myeloid cells and innate lymphoid cells (Fig. 1). 16 This implies that the IL-23/IL-17 axis can still play a pathogenic role in disorders which are not primarily driven by (autoreactive) T cells. And thirdly, the role of Th17 cells depends not only on IL-17A but also on other Th17-derived cytokines such as IL-21 and IL-22, as well as on the specific immunological and tissue context in which they are operating. 8 At mucosal surfaces, for example, IL-17 may have a dual role in promoting tissue inflammation and also protecting from fungal infections. Another good example is IL-22, which drives psoriasis-like dermal inflammation and protects from colitis. TARGETING THE INTERLEUKIN-23/INTERLEUKIN-17 AXIS IN SPONDYLOARTHRITIS There are multiple lines of indirect evidence for therapeutic targeting of the IL-23/IL-17 axis in SpA (Table 1) ,20**,21-30,31*,32*,33-37,38*,39,40* First, the disease shows a strong genetic association with a series of protective polymorphisms in the IL-23 receptor (IL23R) gene, including rs (Arg381Gln). This IL23R gene variant grants protection from IBD, 41 psoriasis 42 and ankylosing spondylitis 17 through selective impairment of IL-17A production 43 due to decreased STAT-3 phosphorylation. 44 Furthermore, identification of additional risk genes related to IL-17 signaling supported significance of the IL-23/IL-17 axis in the pathogenesis of SpA (Table 1). Second, experimental in-vitro and in-vivo models indicate that (Human Leukocyte Antigen (HLA) B27) (the major genetic risk factor for ankylosing spondylitis) heavy chain has a specific propensity to misfold during the assembly in the endoplasmic reticulum, which in specific conditions can lead to an unfolded protein response and increased IL-23 production. 37 HLA-B27 heavy chain can also form aberrant disulfide-linked homodimers on the cell surface, which directly trigger KIR3DL2-positive T cells and natural killer (NK) cells to produce IL Third, analysis of the peripheral blood compartment in humans shows that the number of circulating CD4 + IL-17 + cells (including KIR3DL2-expressing T cells 27 and IL-17-producing γδ T cells 31* ) is expanded in 120

122 9 ankylosing spondylitis.29 Expression studies in the inflamed target tissues, including axial skeleton, peripheral synovitis and subclinically affected gut tissue, indicate a SpA-specific increase in IL-23 and IL-17-producing innate immune cells (Table 1). Fourth, experimental SpA 121 The IL-23/IL-17 axis as promising new target Figure 1. The IL-23/IL-17 pathway. Heterodimeric IL-23 consists of p19 (IL-23A) and p40 (IL-12B) subunits and is secreted by activated macrophages and dendritic cells. IL-23 signals via IL-23R in a STAT3-dependent manner. This leads to typical Th17 effector cytokines production by the lineage expressing specific transcription factor RorC. IL-17A, IL-17F and IL-17A IL-17F-heterodimer signal through heteromeric receptor complex consisted of IL-17RA and IL-17RC subunits. DC, dendritic cell; IL, interleukin; mφ, macrophage; RorC, retinoic acid receptor-related orphan nuclear receptor; STAT, signal transducer and activator of transcription.

123 Table 1. Rationales for targeting IL-23/IL-17 axis in spondyloarthritis I. Genetics IL23R R381Q gene variant grants protection from AS 17 through selective impairment of IL-17A production. Other genes involved in SpA susceptibility: IL-12B, which encodes the 40-kd (p40) chain, a component of both IL-12 and IL-23; 18,19 STAT-3, JAK-2 and TYK-2, which encodes the major signalling molecules activated by the IL-23R. 18,20** Caspase recruitment domain 9 (CARD-9), which encodes a major signalling intermediate in the pathway whereby fungal products induce IL-23 production from dendritic cells (DCs); 21 Prostaglandin E receptor 2 subtype EP 2 (PTGER-4), which encodes a prostaglandin receptor that drives IL-23 secretion by DCs. 22 II. In vitro models Misfolding of HLA-B27 can generate ER stress that orchestrates the unfolded protein response leading to IL-23 production. There are evidences for UPR activation in cells from humans with spondyloarthritis: Gene expression profiling revealed enhanced expression of immunoglobulin heavy chain binding protein (BiP) in SpA PBMCs. 23 Increased expression of GRP78 in macrophages isolated from peripheral joints of AS compared with OA patients. 24 Ankylosing spondylitis macrophages produce increased levels of IL HLA-B27 is capable of forming H chain homodimers ( B27 2 ) that interact with the killer-cell Ig-like receptor (KIR) KIR3DL2. 26 TCR-stimulated peripheral blood KIR3DL2 + CD4 + T cell lines from SpA patients secrete increased levels of IL III. Human expression studies Peripheral blood Target tissues Increase in Th17 cell (SpA, AS ). Specific increase in killer immunoglobulin receptor (KIR) KIR3DL2 + CD4 + Th17 cells (AS 27 ). Increase in IL23R + γδ T cell (active AS 31 ). Monocyte-derived macrophages from the peripheral blood of AS patients produce more IL-23 in response to LPS than healthy controls. 25 Monocyte-derived DCs from AS patients and healthy controls produced more IL-23 than cells from rheumatoid arthritis (RA) patients. 32* CD15 + neutrophils and MPO + myeloid cells are the major cellular sources of IL-17 in the inflamed bone marrow of affected facet joints. 33 IL-23 is overexpressed in AS spinal facet joints comparing to osteoarthritis patients and localized in MPO + CD15 - myeloid precursors and to a lesser extent in CD68 + or CD163 + macrophages. 34 Specifically increased mast cells are the major source of IL-17 in the inflamed synovial joints of psoriatic and non-psoriatic SpA. 35 IL-23 is increased in the gut of SpA patients and has been localized to infiltrating monocytes and Paneth cells. 36 IV. Animal models Spondyloarthritis in HLA-B27 transgenic rat is associated with CD4 + Th17 T cell activation. 37,38* Experimental ankylosing enthesitis in (BXSB NZB) F1 mice is associated with an increase in IL-17 production by T cells. 39 IL-23 overexpression induces a spondyloarthritis-like disease in B10.RIII mice via RORγt(+)CD3(+) CD4( )CD8( ) T cells that produce IL-17 and IL * AS, ankylosing spondylitis; GRPT78, glucose-regulated protein, 78 kda; IL-23R, interleukin-23 receptor; JAK-2, Janus kinase 2; LPS, lipopolysaccharides; STAT-3, signal transducer and activator of transcription 3; TYK-2, tyrosine kinase

124 in HLA-B27 transgenic rats and experimental ankylosing enthesitis in (BXSB NZB) F1 mice are associated with an expansion of Th17 cells and an increase in IL-17 production. 39,45 Moreover, IL-23 overexpression induces SpA-like disease in B10.RIII mice via a previously unidentified population of IL-23R + resident cells expressing RoRγt (RAR-related orphan receptor gamma) that produce IL-17 and IL * Thus, cumulative evidence from genetics, in-vitro models, human expression studies and animal models strongly supports the involvement of the IL-23/IL-17 axis in the pathogenesis of SpA. It is important to note, however, that all this evidence is circumstantial and that the data emerging from these studies should be interpreted carefully. Indeed, genetic risk factors do not always predict pathogenic involvement of a pathway, as exemplified by the genetic association with IL-6R, 46 but the lack of effect of IL-6 blockade in ankylosing spondylitis. 47 Similarly, the UPR (unfolded protein response) induced by HLA-B27 misfolding has been elegantly shown in cell lines and HLA-B27 tg rats, but it remains to be demonstrated that this phenomenon is also occurring in vivo in our patients. 32*,48 Also, the exact cellular source of IL-23 and IL-17 in the target tissues remains to be fully defined. Interestingly, IL-23 serum levels are not specifically elevated and do not correlate with disease activity and/or treatment response in SpA. 49,50 And finally, animal models are key to study mechanistic aspects of the pathways of interest, but poorly mimic human diseases. Systemic IL-23 overexpression, for example, does not only lead to SpA-like enthesitis but can also induce a very severe and destructive myeloiddriven polyarthritis in the same mice. 51,52 These data indicate that preclinical studies should be interpreted carefully, and highlight the crucial importance of PoC studies in human patients. CLINICAL TRIALS TARGETING THE INTERLEUKIN-23/ INTERLEUKIN-17 AXIS Recognition of the importance of the IL-23/IL-17 axis resulted in a rapidly expanding repertoire of compounds targeting this axis. Ustekinumab 53 (a monoclonal antibody directed against the common p40 subunit of IL-23 and IL-12), secukinumab, 54 ixekizumab 55 (both monoclonal antibodies directed against IL-17A), and brodalumab 56 [a monoclonal antibody against the IL-17RA receptor, thereby blocking IL-17A, IL-17F and IL-17E (also known as IL-25)] have all successfully completed preclinical development programs and have subsequently been tested in randomized controlled trials (RCTs) in psoriasis as first-choice target disease (Fig. 2). All compounds showed very significant clinical efficacy in psoriasis, with 75% improvement in Psoriasis Area and Severity Index (PASI75) scores of 65 85% at week 12. As psoriasis, ankylosing spondylitis and Crohn s disease are clinically related conditions sharing the same polymorphisms in the IL-23R; these impressive clinical data in psoriasis provided further indirect support for testing these agents in SpA. However, PoC data in Crohn s disease showed a quite different picture. A phase II trial with ustekinumab did not reach its primary endpoint, but showed a positive trend in TNF inhibitor failers in a post-hoc analysis. 57 A phase III trial specifically designed to address this question showed a significant effect: 34 40% of the ustekinumab groups versus 24% of the placebo group reached the primary endpoint (clinical response at week 6). 58 In contrast with p40 blockade by ustekinumab, however, PoC 9 The IL-23/IL-17 axis as promising new target 123

125 Figure 2. The IL-23/IL-17 pathway as a target in the treatment of spondyloarthritis. The pharmacological mechanism of action is depicted schematically for inhibitors of IL-23 and IL-12 synthesis (apilimod), IL-23p40 (ustekinumab, briakinumab), IL-23p19 (tildrakizumab, guselkumab), IL-17A (secukinumab, ixekizumab), IL-22 (fezakinumab) and a blocker of the receptor IL-17RA (brodalumab). trials with IL-17 blockers did not show any benefit. A RCT with secukinumab in 60 patients with Crohn s disease did not reach its primary endpoint and, if anything, showed deterioration in the active treatment arm versus placebo (<0.1% probability that active treatment was better than placebo at the week 6 primary endpoint).59 A similar trial with brodalumab was terminated by the sponsor after interim analysis ( 124

126 With discrepant results in psoriasis versus Crohn s disease, the clinical efficacy of drugs targeting the IL-23/IL-17 axis was explored in PoC trials in SpA, in particular, in the ankylosing spondylitis and PsA subtypes (summarized in Table 2). 60,61*-65*,66 Blocking IL-17A with secukinumab was very efficacious in ankylosing spondylitis (Fig. 2). In a PoC phase II trial with 30 patients, the primary endpoint Assessment of SpondyloArthritis International Society 20% improvement criteria (ASAS20) at week 6 was 59% in secukinumab and 24% in placebo. Moreover, not only the clinical signs and symptoms but also the serum biomarkers of inflammation C-reactive protein (CRP) and S100A8 and S100A9, as well as MRI scores for inflammation, decreased upon active treatment. 65* Subsequently, a PoC open-label trial with ustekinumab in 20 ankylosing spondylitis patients showed clear signs of clinical improvement, with an ASAS40 of 65% at week 24, 63* but there was no significant effect on CRP levels in the whole group but only in the clinical responders and the clinical improvement should thus first be confirmed in a RCT before making firm conclusions. Results for ixekizumab and brodalumab in ankylosing spondylitis are not known yet (Fig. 2). In PsA, the superiority of ustekinumab above placebo has already been shown in several RCTs. 60,61* The phase II trial with 146 patients showed that 42% of the ustekinumab-treated PsA patients versus 14% of the placebo-treated patients reached the American College of Rheumatology 20% (ACR20) improvement criteria. 60 This was confirmed in two phase III trials (n=615 and n=312), where ACR20 at week 24 was reached in 42 50% of the ustekinumab group versus 20 23% of the placebo group. 61*,67 This response was still maintained at week 50 for the various ACR scores, and also for the PASI75 score and the Health Assessment Questionnaire- Disability Index (HAQ-DI). Moreover, ustekinumab was also efficacious in patients who previously failed on TNF blockers (week 24 ACR20 response: ustekinumab 36% versus placebo 15%). 67 Of the IL-17 blocking agents, only secukinumab was yet tested and reported in a RCT in PsA, with a trend towards clinical efficacy (week 6 ACR20: secukinumab 39% versus placebo 23%) in a phase II PoC trial with 42 patients. 64* The first results with brodalumab in PsA were reported in abstract form, but are not yet published: 66 in a phase II trial with 168 patients, 37 39% of the brodalumab group reached the ACR20 primary endpoint at week 12 compared to 18% of the placebo group. Importantly, these various compounds did not show unexpected safety concerns so far. Results for ixekizumab are still pending. Taken together, the first PoC trials with drugs blocking key cytokines of the IL-23/IL-17 axis showed promising results in SpA (Table 2). Both larger long-term trials with these compounds and PoC trials with new compounds are currently in progress. The latter include briakinumab (another IL-12/23 inhibitor targeting the common p40 subunit), tildrakizumab and guselkumab (both monoclonal antibodies targeting the p19 subunit of IL-23), apilimod (a small molecule that selectively suppresses synthesis of IL-12 and IL-23) and fezakinumab (a monoclonal antibody directed against compounds targeting IL-23R, Janus kinases (JAKs) and RORγt (RAR-related orphan receptor gamma), are also in preclinical or clinical development. 9 The IL-23/IL-17 axis as promising new target 125

127 Tabel 2. Summary of clinical trials targeting the IL-23/IL-17 axis in SpA Drug Study design SpA subtype Treatment arms Main results Monoclonal antibodies targeting the common p40 subunit of IL-23 and IL-12 Ustekinumab 60 Ustekinumab 61* Ustekinumab 62* Randomized, double blind, placebo-controlled, cross-over, phase II trial Randomized, double blind, placebo-controlled, phase III trial Randomized, double blind, placebo-controlled, phase III trial PsA PsA PsA Placebo (n=70) Ustekinumab (n=76) Placebo (n=206) Ustekinumab 45 mg (n=205) Ustekinumab 90 mg (n=204) Placebo (n=104) Ustekinumab 45 mg (n=103) Ustekinumab 90 mg (n=105) ACR20 week 12: - Placebo: 14% - Ustekinumab: 42% ACR20 week 24: - Placebo: 23% - Ustekinumab 45 mg: 42% - Ustekinumab 90 mg: 50% ACR20 week 24: - Placebo: 20% - Ustekinumab 45 mg: 44% - Ustekinumab 90 mg: 44% Ustekinumab 63* Open label single-arm trial AS Ustekinumab (n=20) ASAS40 week 24: - Ustekinumab: 65% Briakinumab Not yet tested in SpA Small molecules suppressing the synthesis of IL-23 and IL-12 Apilimod Not yet tested in SpA Monoclonal antibodies targeting the p19 subunit of IL-23 Tildrakizumab Guselkumab Not yet tested in SpA Not yet tested in SpA Monoclonal antibodies targeting IL-17A Secukimumab 64* Secukimumab 65* Ixekizumab Randomized, double blind, placebo-controlled, phase II trial Randomized, double blind, placebo-controlled, phase II trial Randomized, double blind, placebo-controlled trial Monoclonal antibodies targeting IL-17RA Brodalumab 66 Randomized, double blind, placebo-controlled, phase II trial Monoclonal antibodies targeting IL-22 Fezakinumab Not yet tested in SpA PsA AS PsA PsA Placebo (n=14) Secukinumab (n=28) Placebo (n=6) Secukinumab (n=24) Still recruiting Placebo (n=55) Brodalumab 140 mg (n=57) Brodalumab 280 mg (n=56) ACR20 week 6: - Placebo: 23% - Secukinumab: 39% ASAS20 week 6: - Placebo: 24% - Secukinumab: 59% ACR week 12: - Placebo: 18% - Brodalumab 140 mg: 37% - Brodalumab 280 mg: 39% ACR20 = American College of Rheumatology 20% improvement criteria; ASAS20 = Assessment of SpondyloArthritis international Society 20% improvement criteria; IL-17RA = IL-17-receptor A. 126

128 CONCLUSION The strong albeit circumstantial evidence that the IL-23/IL-17 axis contributes to SpA pathogenesis has now been confirmed by a series of PoC trials with ustekinumab, targeting the p40 subunit of IL-23 and IL-12, and secukinumab, a monoclonal anti-il17a antibody, in ankylosing spondylitis and PsA. These promising results open new avenues for treatment of patients with SpA, but many fundamental and clinical questions remain to be answered. Obviously, the long-term efficacy and safety of IL-23 and IL-17 blockers need to be demonstrated in large phase III and IV trials. Three additional questions, however, are of key translational and clinical relevance. Firstly, how does the IL-23/IL-17 axis relate to the TNF axis? Are both pathways synergistic, as suggested by in-vitro data, or are they redundant in vivo? Are patients failing TNF inhibitors responding well to IL-23/IL-17 blockade or, on the contrary, are nonresponders to inhibition of one pathway also nonresponders to blockade of the other axis? Or should we even consider combination treatment with blockade of both axes? Secondly, at what level should the IL-23/IL17 axis be targeted? Is it more effective to target an upstream component of the pathway, such as p19, as this will not only block IL-17A but also other IL-17 family members and IL-22? Or is it safer to target downstream effectors, such as one specific IL-17 isotype or even one specific cell type producing IL-17 in the target tissue of interest? Thirdly and finally, will inhibition of the IL-23/IL-17 axis not only suppress inflammation but also modify disease progression, in particular, with regard to new bone formation? And, if so, what is then the relative contribution of the different cytokines of this axis (e.g. IL-17A, which is known to costimulate osteoclasts, versus IL-22, which promotes stromal repair) to structural damage? Answering these key questions will require an integrated approach combining the ongoing clinical trials with translational research into the molecular and cellular pathways of disease. 9 KEY POINTS -- Cumulative evidence from genetics, in-vitro models, human expression studies and animal models strongly supports the involvement of the IL-23/IL-17 axis in the pathogenesis of SpA. -- Clinical trials with ustekinumab (anti-p40) and secukinumab (anti-il-17a) generally showed good clinical efficacy in ankylosing spondylitis and PsA; proof-of-concept trials with ixekizumab (anti-il-17a) and brodalumab (anti-il-17r) are ongoing. -- Targeting the IL-23/IL-17 axis is also effective in psoriasis, but this effect is less clear in Crohn s disease. -- More compounds targeting the IL-23/IL-17 axis at different levels are in preclinical and/or clinical development. -- Testing these compounds in the various subtypes of SpA is expected to lead to new therapeutic options of patients as well as to new insights in the pathophysiology of the disease. The IL-23/IL-17 axis as promising new target Acknowledgements Professor Dr Baeten was supported by a VICI grant from The Netherlands Organization for Scientific Research (NWO) and by a grant from the Dutch Arthritis Foundation (Reumafonds). 127

129 REFERENCES 1 Dougados M, Baeten D. Spondyloarthritis. Lancet 2011; 377: Brandt J, Haibel H, Cornely D, et al. Successful treatment of active ankylosing spondylitis with the antitumor necrosis factor alpha monoclonal antibody infliximab. Arthritis Rheum 2000; 43: Van den Bosch F, Kruithof E, Baeten D, et al. Effects of a loading dose regimen of three infusions of chimeric monoclonal antibody to tumour necrosis factor alpha (infliximab) in spondyloarthropathy: an open pilot study. Ann Rheum Dis 2000; 59: Baraliakos X, Listing J, Brandt J, et al. Clinical response to discontinuation of anti-tnf therapy in patients with ankylosing spondylitis after 3 years of continuous treatment with infliximab. Arthritis Res Ther 2005; 7:R439 R Paramarta JE, Heijda TF, Baeten DL. Fast relapse upon discontinuation of tumour necrosis factor blocking therapy in patients with peripheral spondyloarthritis. Ann Rheum Dis 2013; 72: Song IH, Althoff CE, Haibel H, et al. Frequency and duration of drug-free remission after 1 year of treatment with etanercept versus sulfasalazine in early axial spondyloarthritis: 2 year data of the ESTHER trial. Ann Rheum Dis 2012; 71: Cua DJ, Sherlock J, Chen Y, et al. Interleukin-23 rather than interleukin-12 is the critical cytokine for autoimmune inflammation of the brain. Nature 2003; 421: Baeten DL, Kuchroo VK. How Cytokine networks fuel inflammation: Interleukin-17 and a tale of two autoimmune diseases. Nat Med 2013; 19: Murphy CA, Langrish CL, Chen Y, et al. Divergent pro- and anti-inflammatory roles for IL-23 and IL-12 in joint autoimmune inflammation. J Exp Med 2003; 198: Aggarwal S, Ghilardi N, Xie MH, et al. Interleukin-23 promotes a distinct CD4 T cell activation state characterized by the production of interleukin-17. J Biol Chem 2003; 278: Wright JF, Guo Y, Quazi A, et al. Identification of an interleukin 17F/17A heterodimer in activated humancd4 + T cells. J Biol Chem2007; 282: Shen F, Gaffen SL. Structure-function relationships in the IL-17 receptor: implications for signal transduction and therapy. Cytokine 2008; 41: Lee YK, Turner H, Maynard CL, et al. Late developmental plasticity in the T helper 17 lineage. Immunity 2009; 30: Zielinski CE, Mele F, Aschenbrenner D, et al. Pathogen-induced human TH17 cells produce IFN-gamma or IL-10 and are regulated by IL-1beta. Nature 2012; 484: *Evans HG, Roostalu U, Walter GJ, et al. TNF-alpha blockade induces IL-10 expression in human CD4 + T cells. Nat Commun 2014; 5:3199. doi: / ncomms4199. This study demonstrates that IL-17 + CD4 + T cells can acquire regulatory activity through the expression of the anti-inflammatory cytokine IL Awasthi A, Riol-Blanco L, Jager A, et al. Cutting edge: IL-23 receptor gfp reporter mice reveal distinct populations of IL-17-producing cells. J Immunol 2009; 182: Burton PR, Clayton DG, Cardon LR, et al. Association scan of 14,500 nonsynonymous SNPs in four diseases identifies autoimmunity variants. Nat Genet 2007; 39: Danoy P, Pryce K, Hadler J, et al. Association of variants at 1q32 and STAT3 with ankylosing spondylitis suggests genetic overlap with Crohn s disease. PLoS Genet 2010; 6:e Evans DM, Spencer CC, Pointon JJ, et al. Interaction between ERAP1 and HLA-B27 in ankylosing spondylitis implicates peptide handling in the mechanism for HLA-B27 in disease susceptibility. Nat Genet 2011;43: **Cortes A, Hadler J, Pointon JP, et al. Identification of multiple risk variants for ankylosing spondylitis through high-density genotyping of immunerelated loci. Nat Genet 2013; 45: This important study supports mounting evidence that aberrant peptide processing before MHC class I presentation and alterations of the IL-23 pathway are key elements in the pathogenesis of ankylosing spondylitis. 21 Pointon JJ, Harvey D, Karaderi T, et al. Elucidating the chromosome 9 association with AS: CARD9 is a candidate gene. Genes Immun 2010;11: Chen Q, Muramoto K, Masaaki N, et al. A novel antagonist of the prostaglandin E(2) EP(4) receptor inhibits Th1 differentiation and Th17 expansion and is orally active in arthritis models. Br J Pharmacol 2010; 160: Gu J, Rihl M, Marker-Hermann E, et al. Clues to pathogenesis of spondyloarthropathy derived from synovial fluid mononuclear cell gene expression profiles. J Rheumatol 2002; 29: Dong W, Zhang Y, Yan M, et al. Upregulation of 78- kda glucose-regulated protein in macrophages in 128

130 peripheral joints of active ankylosing spondylitis. Scand J Rheumatol 2008; 37: Zeng L, Lindstrom MJ, Smith JA. Ankylosing spondylitis macrophage production of higher levels of interleukin-23 in response to lipopolysaccharide without induction of a significant unfolded protein response. Arthritis Rheum 2011; 63: Kollnberger S, Bird L, Sun MY, et al. Cell-surface expression and immune receptor recognition of HLA-B27 homodimers. Arthritis Rheum 2002; 46: Bowness P, Ridley A, Shaw J, et al. Th17 cells expressing KIR3DL2 + and responsive to HLA- B27 homodimers are increased in ankylosing spondylitis. J Immunol 2011; 186: Jandus C, Bioley G, Rivals JP, et al. Increased numbers of circulating polyfunctional Th17 memory cells in patients with seronegative spondylarthritides. Arthritis Rheum 2008; 58: Shen H, Goodall JC, Hill Gaston JS. Frequency and phenotype of peripheral blood Th17 cells in ankylosing spondylitis and rheumatoid arthritis. Arthritis Rheum 2009; 60: Zhang L, Li YG, Li YH, et al. Increased frequencies of Th22 cells as well as Th17 cells in the peripheral blood of patients with ankylosing spondylitis and rheumatoid arthritis. PLoS One 2012;7:e *Kenna TJ, Davidson SI, Duan R, et al. Enrichment of circulating interleukin-17-secreting interleukin-23 receptor-positive gamma/delta T cells in patients with active ankylosing spondylitis. Arthritis Rheum 2012;64: This study demonstrates that IL-23 drives IL-17 production by gamma/delta T cells, which are possibly involved in the pathogenesis of ankylosing spondylitis. 32 *Prevosto C, Goodall JC, Hill Gaston JS. Cytokine secretion by pathogen recognition receptorstimulated dendritic cells in rheumatoid arthritis and ankylosing spondylitis. J Rheumatol 2012;39: This study shows that monocyte-derived dendritic cells from rheumatoid arthritis patients produce less IL-23 than cells from ankylosing spondylitis and healthy controls. 33 Appel H, Maier R, Wu P, et al. Analysis of IL-17(+) cells in facet joints of patients with spondyloarthritis suggests that the innate immune pathway might be of greater relevance than the Th17-mediated adaptive immune response. Arthritis Res Ther 2011;13:R Appel H, Maier R, Bleil J, et al. In situ analysis of interleukin-23- and interleukin-12-positive cells in the spine of patients with ankylosing spondylitis. Arthritis Rheum 2013; 65: Noordenbos T, Yeremenko N, Gofita I, et al. Interleukin-17-positive mast cells contribute to synovial inflammation in spondylarthritis. Arthritis Rheum 2012;64: Ciccia F, Bombardieri M, Principato A, et al. Overexpression of interleukin-23, but not interleukin-17, as an immunologic signature of subclinical intestinal inflammation in ankylosing spondylitis. Arthritis Rheum 2009;60: DeLay ML, Turner MJ, Klenk EI, et al. HLA-B27 misfolding and the unfolded protein response augment interleukin-23 production and are associated with Th17 activation in transgenic rats. Arthritis Rheum 2009; 60: *Glatigny S, Fert I, Blaton MA, et al. Proinflammatory Th17 cells are expanded and induced by dendritic cells in spondylarthritis-prone HLA-B27-transgenic rats. Arthritis Rheum 2012; 64: This study demonstrates that defective function in HLA-B27-transgenic rat dendritic cells contributes to disease development by skewing CD4 + T cells toward Th Abe Y, Ohtsuji M, Ohtsuji N, et al. Ankylosing enthesitis associated with upregulated IFNgamma and IL-17 production in (BXSB x NZB) F(1) male mice: a new mouse model. Mod Rheumatol 2009; 19: *Sherlock JP, Joyce-Shaikh B, Turner SP, et al. IL-23 induces spondyloarthropathy by acting on RORgammat + CD3 + CD4 - CD8 - entheseal resident T cells. Nat Med 2012; 18: This study shows that IL-23 overexpression induces a spondyloarthritis-like disease in B10.RIII mice via (RAR-related orphan receptor gamma) ROR-γt(+) CD3(+)CD4(-)CD8(-) T cells that produce IL-17 and IL Duerr RH, Taylor KD, Brant SR, et al. A genomewide association study identifies IL23R as an inflammatory bowel disease gene. Science 2006; 314: Cargill M, Schrodi SJ, Chang M, et al. A largescale genetic association study confirms IL12B and leads to the identification of IL23R as psoriasis-risk genes. Am J Hum Genet 2007; 80: Di MP, Di CA, Laggner U, et al. The IL23R R381Q gene variant protects against immune-mediated diseases by impairing IL-23-induced Th17 effector response in humans. PLoS One 2011; 6:e Sarin R, Wu X, Abraham C. Inflammatory disease protective R381Q IL23 receptor polymorphism results in decreased primary CD4 + and CD8 + 9 The IL-23/IL-17 axis as promising new target 129

131 human T-cell functional responses. Proc Natl Acad Sci U S A 2011;108: Utriainen L, Firmin D, Wright P, et al. Expression of HLA-B27 causes loss of migratory dendritic cells in a rat model of spondylarthritis. Arthritis Rheum 2012;64: Parkes M, Cortes A, van Heel DA, Brown MA. Genetic insights into common pathways and complex relationships among immune-mediated diseases. Nat Rev Genet 2013; 14: Sieper J, Porter-Brown B, Thompson L, et al. Assessment of short-term symptomatic efficacy of tocilizumab in ankylosing spondylitis: results of randomised, placebo-controlled trials. Ann Rheum Dis 2014; 73: Neerinckx B, Carter S, Lories RJ. No evidence for a critical role of the unfolded protein response in synovium and blood of patients with ankylosing spondylitis. Ann Rheum Dis 2013; 73: Melis L, Vandooren B, Kruithof E, et al. Systemic levels of IL-23 are strongly associated with disease activity in rheumatoid arthritis but not spondyloarthritis. Ann Rheum Dis 2010; 69: Wendling D, Cedoz JP, Racadot E. Serum and synovial fluid levels of p40 IL12/23 in spondyloarthropathy patients. Clin Rheumatol 2009;28: Adamopoulos IE, Tessmer M, Chao CC, et al. IL-23 is critical for induction of arthritis, osteoclast formation, and maintenance of bone mass. J Immunol 2011;187: Molle C, Zhang T, Ysebrant de Lendonck L, et al. Tristetraprolin regulation of interleukin 23 mrna stability prevents a spontaneous inflammatory disease. J Exp Med 2013; 210: Leonardi CL, Kimball AB, Papp KA, et al. Efficacy and safety of ustekinumab, a human interleukin-12/23 monoclonal antibody, in patients with psoriasis: 76-week results from a randomised, double-blind, placebo-controlled trial (PHOENIX 1). Lancet 2008; 371: Papp KA, Langley RG, Sigurgeirsson B, et al. Efficacy and safety of secukinumab in the treatment of moderate-to-severe plaque psoriasis: a randomized, double-blind, placebo-controlled phase II doseranging study. Br J Dermatol 2013; 168: Leonardi C, Matheson R, Zachariae C, et al. Antiinterleukin-17 monoclonal antibody ixekizumab in chronic plaque psoriasis. N Engl J Med 2012;366: Papp KA, Leonardi C, Menter A, et al. Brodalumab, an antiinterleukin-17-receptor antibody for psoriasis. N Engl J Med 2012; 366: Sandborn WJ, Feagan BG, Fedorak RN, et al. A randomized trial of Ustekinumab, a human interleukin-12/23 monoclonal antibody, in patients with moderate-to-severe Crohn s disease. Gastroenterology 2008; 135: Sandborn WJ, Gasink C, Gao LL, et al. Ustekinumab induction and maintenance therapy in refractory Crohn s disease. N Engl J Med 2012;367: Hueber W, Sands BE, Lewitzky S, et al. Secukinumab, a human anti-il-17a monoclonal antibody, for moderate to severe Crohn s disease: unexpected results of a randomised, double-blind placebo-controlled trial. Gut 2012;61: Gottlieb A, Menter A, Mendelsohn A, et al. Ustekinumab, a human interleukin 12/23 monoclonal antibody, for psoriatic arthritis: randomised, double-blind, placebo-controlled, crossover trial. Lancet 2009;373: *McInnes IB, Kavanaugh A, Gottlieb AB, et al. Efficacy and safety of ustekinumab in patients with active psoriatic arthritis: 1 year results of the phase 3, multicentre, double-blind, placebo-controlled PSUMMIT 1 trial. Lancet 2013;382: These important clinical trials show good clinical efficacy of targeting of IL-23/IL-17 axis in PsA and ankylosing spondylitis. 62 *Ritchlin C, Rahman P, Kavanaugh A, et al. Efficacy and safety of the anti-il-12/23 p40 monoclonal antibody, ustekinumab, in patients with active psoriatic arthritis despite conventional nonbiological and biological antitumour necrosis factor therapy: 6-month and 1-year results of the phase 3, multicentre, double-blind, placebo-controlled, randomised PSUMMIT 2 trial. Ann Rheum Dis doi: /annrheumdis These important clinical trials show good clinical efficacy of targeting of IL-23/IL-17 axis in PsA and ankylosing spondylitis. 63 *Poddubnyy D, Hermann KG, Callhoff J, et al. Ustekinumab for the treatment of patients with active ankylosing spondylitis: results of a 28-week, prospective, open-label, proof-of-concept study (TOPAS). Ann Rheum Dis 2014; 73: These important clinical trials show good clinical efficacy of targeting of IL-23/IL-17 axis in PsA and ankylosing spondylitis. 64 *McInnes IB, Sieper J, Braun J, et al. Efficacy and safety of secukinumab, a fully human antiinterleukin-17a monoclonal antibody, in patients with moderate-tosevere psoriatic arthritis: a 24-week, randomised, double-blind, placebocontrolled, phase II proof-of-concept trial. Ann Rheum Dis 2014; 73:

132 These important clinical trials show good clinical efficacy of targeting of IL-23/IL-17 axis in PsA and ankylosing spondylitis. 65 *Baeten D, Baraliakos X, Braun J, et al. Antiinterleukin-17A monoclonal antibody secukinumab in treatment of ankylosing spondylitis: a randomised, double-blind, placebo-controlled trial. Lancet 2013; 382: These important clinical trials show good clinical efficacy of targeting of IL-23/IL-17 axis in PsA and ankylosing spondylitis. 66 Mease PJ, Genovese MC, Greenwald MW, et al. Efficacy of brodalumab, an anti-il-17r antibody, in subjects with psoriatic arthritis [abstract]. Ann Rheum Dis 2013; 72 (Suppl3): Ritchlin CT, Gottlieb AB, McInnes IB, et al. Ustekinumab in active psoriatic arthritis including patients previously treated with anti-tnf agents: results of a phase 3, multicenter, double-blind, placebo-controlled study [abstract]. Arthritis Rheum 2012; 64 (Suppl):S1080 S The IL-23/IL-17 axis as promising new target 131

133

134 Anti-interleukin-17a monoclonal antibody secukinumab in treatment of ankylosing spondylitis: a randomised, double-blind, placebo-controlled trial Dominique Baeten 1, Xenofon Baraliakos 2, Jürgen Braun 2, Joachim Sieper 3, Paul Emery 4, Désirée van der Heijde 5, Iain McInnes 6, Jacob M. van Laar 7, Robert Landewé 1,8, Paul Wordsworth 9, Jürgen Wollenhaupt 10, Herbert Kellner 11, Jacqueline Paramarta 1, Jiawei Wei 12, Arndt Brachat 13, Stephan Bek 13, Didier Laurent 13, Yali Li 14, Ying A Wang 14, Arthur P Bertolino 13, Sandro Gsteiger 15, Andrew M. Wright 15, Wolfgang Hueber 13 1 Department of Clinical Immunology and Rheumatology, Academic Medical Center/University of Amsterdam, Amsterdam, The Netherlands, 2 Rheumazentrum Ruhrgebiet, Herne, Germany, 3 Medicine/Rheumatology Department, Charité Campus Benjamin Franklin, Berlin, Germany, 4 Leeds MSK Biomedical Research Unit, University of Leeds, Leeds, United Kingdom, 5 Rheumatology Department, Leiden University Medical Center, Leiden, The Netherlands, 6 Glasgow Biomedical Research Centre, University of Glasgow, Glasgow, United Kingdom 7 Musculoskeletal Research Group, Institute of Cellular Medicine, Newcastle University, Newcastle upon Tyne, United Kingdom, 8 Atrium Medical Center, Heerlen, The Netherlands, 9 National Institute for Health Research, Oxford Musculoskeletal Biomedical Research Unit, and Nuffield Department of Orthopaedics, Rheumatology and Musculoskeletal Sciences, Nuffield Orthopaedic Centre, Oxford, United Kingdom, 10 Division of Rheumatology, Schön Klinik Hamburg Eilbeck, Hamburg, Germany, 11 Division of Rheumatology, Centre for Inflammatory Joint Diseases, Munich, Germany, 12 Novartis Pharma AG, Shanghai, China, 13 Novartis Institutes for BioMedical Research, Basel, Switzerland, 14 Novartis Institutes for BioMedical Research, Cambridge, MA, United States, 15 Novartis Pharma AG, Basel, Switzerland 10 Lancet Nov 23;382(9906):

135 SUMMARY Background Ankylosing spondylitis is a chronic immune-mediated inflammatory disease characterised by spinal inflammation, progressive spinal rigidity, and peripheral arthritis. Interleukin 17 (IL-17) is thought to be a key inflammatory cytokine in the development of ankylosing spondylitis, the prototypical form of spondyloarthritis. We assessed the efficacy and safety of the anti-il-17a monoclonal antibody secukinumab in treating patients with active ankylosing spondylitis. Methods We did a randomised double-blind proof-of-concept study at eight centres in Europe (four in Germany, two in the Netherlands, and two in the UK). Patients aged years were randomly assigned (in a 4:1 ratio) to either intravenous secukinumab (2 10 mg/kg) or placebo, given 3 weeks apart. Randomisation was done with a computer-generated block randomisation list without a stratification process. The primary efficacy endpoint was the percentage of patients with a 20% response according to the Assessment of SpondyloArthritis international Society criteria for improvement (ASAS20) at week 6 (Bayesian analysis). Safety was assessed up to week 28. This study is registered with ClinicalTrials.gov, number NCT Findings 37 patients with moderate-to-severe ankylosing spondylitis were screened, and 30 were randomly assigned to receive either intravenous secukinumab (n=24) or placebo (n=6). The final efficacy analysis included 23 patients receiving secukinumab and six patients receiving placebo, and the safety analysis included all 30 patients. At week 6, ASAS20 response estimates were 59% on secukinumab versus 24% on placebo (99.8% probability that secukinumab is superior to placebo). One serious adverse event (subcutaneous abscess caused by Staphylococcus aureus) occurred in the secukinumab-treated group. Interpretation Secukinumab rapidly reduced clinical or biological signs of active ankylosing spondylitis and was well tolerated. It is the first targeted therapy that we know of that is an alternative to tumour necrosis factor inhibition to reach its primary endpoint in a phase 2 trial. 134

136 INTRODUCTION Ankylosing spondylitis is a chronic immune-mediated inflammatory disease, with an estimated prevalence of % worldwide. It is characterised by spinal inflammation, progressive spinal ankylosis due to new bone formation, peripheral arthritis and enthesitis, extra-articular manifestations, familial clustering, and a strong genetic association with human leucocyte antigen (HLA)-B27. 1 Early onset of the disease in young adults, chronic spinal and extraspinal inflammation, and progressive irreversible structural damage can lead to important morbidity, functional deterioration, and socioeconomic burden. Non-steroidal anti-inflammatory drugs (NSAIDs) and physiotherapy are the cornerstones of treatment for ankylosing spondylitis. Conventional disease-modifying anti-rheumatic drugs (DMARDs) are not effective for the axial symptoms of the disease. For patients with inadequate response to NSAIDs, treatment with tumour necrosis factor (TNF)-blockers has been recommended. 2 However, no approved alternative therapies are available for the roughly 40% of patients who do not tolerate or respond to TNF-blockers. Thus, there is an unmet need for medicines with new modes of action. Interleukin 17A (IL-17A) has emerged as a novel therapeutic target for ankylosing spondylitis. First, the disease shows a strong genetic association with a series of protective polymorphisms in the IL-23 receptor (IL23R) gene, including rs (Arg381Gln). 3 Functional analyses have indicated that the Arg381Gln polymorphism impairs IL-23 dependent IL-17 production by Th17 cells, suggesting that genetically determined down-modulation of the IL-23/IL-17 axis could provide protection against ankylosing spondylitis. 4 Second, intracellular HLA-B27 misfolding leads to an unfolded protein response that potentiates abnormal IL-23 production, 5 and ankylosing spondylitis macrophages produce increased levels of IL Third, the number of circulating CD4 + IL-17 + cells is expanded in ankylosing spondylitis; 7 this includes KIR3DL2-expressing T cells that respond to cellsurface HLA-B27 homodimers 8 and IL-17 producing γ/δ T cells. 9 Fourth, inflamed target tissues indicate an ankylosing spondylitis-specific increase in IL-17 producing innate immune cells. 10,11 Fifth, spondyloarthritis in HLA-B27 transgenic rats and experimental ankylosing enthesitis in (BXSB NZB) F1 mice were associated with an expansion of Th17 cells and with up-regulated IL-17 expression, respectively. 12 Moreover, IL-23 overexpression induces a spondyloarthritis-like disease in B10.RIII mice via RAR-related orphan receptor γt(+)cd3(+)cd4( )CD8( ) T cells that produce IL-17 and IL Finally, trials with anti-il-12/il-23 and anti-il-17 agents have shown significant clinical efficacy in psoriasis, a disease closely related to ankylosing spondylitis We therefore undertook a proof-of-concept study to assess the efficacy and safety of secukinumab, a fully human anti-il-17a monoclonal antibody, in patients with active ankylosing spondylitis. METHODS 10 Secukinumab anti-il-17a therapy in AS Study design We did a 28-week multicentre, randomised, double-blind, placebo-controlled study between March, 2009, and May, 2011, at eight centres in Europe (four in Germany, two in the Netherlands, 135

137 and two in the UK; appendix). After a 4-week screening period, patients were randomly assigned (in a 4:1 ratio) to receive secukinumab or placebo at days 1 and 22. The randomisation plan was generated by Novartis Drug Supply Management using a validated system. It was reviewed and approved by the Biostatistics Quality Assurance group of Novartis and locked after approval. Randomisation was done by using a computer-generated block randomisation list without a stratification process. Data were accessible only to authorised personnel until database lock. The unblinded pharmacist or designated person (eg, study nurse) at each site received treatment allocation cards. When an eligible patient at a participating study site was to be entered into the study, the unblended pharmacist or designated person completed a randomization request by to Novartis. The assigned randomisation number was returned to the pharmacist within 24 h of the request. After inclusion of each new patient, the pharmacist returned the to confirm the patient s randomisation number and the date the patient received the first administration of study medication. Efficacy and safety assessments were done up to week 28, with week 6 being the primary endpoint. All week-6 analyses were designated as interim analyses. All centres received approval from independent ethics committees or institutional review boards, and the study was done in accordance with the principles of the Declaration of Helsinki. Signed informed consent for the main and pharmacogenetic studies was obtained from each patient before any study-related procedures were undertaken. The full study protocol is available from the sponsor. Patients The study included 30 patients, years of age, with definite ankylosing spondylitis as defined by the modified New York criteria, 20 a score of at least 4 on the Bath ankylosing spondylitis disease activity index (BASDAI), and a total back pain and nocturnal pain score of 40 or more on the visual analogue scale (0 100 mm), irrespective of the maximum tolerated doses of NSAIDs. Major exclusion criteria included total spinal ankylosis, concomitant fulfilment of criteria for psoriatic arthritis, and presence of severe acute or subacute anterior uveitis. Patients with a history of or current stable psoriasis or inflammatory bowel disease were allowed to participate. Patients with active tuberculosis, hepatitis B or C, HIV, or any active systemic infection within the 2 weeks before baseline were excluded. Patients with latent tuberculosis had to complete their treatment and be considered cured before study entry. Patients with a history of malignancy (except basal cell carcinoma or adequately treated carcinoma in situ of the cervix) or significant cardiac, renal, neurological, psychiatric, endocrinologic, metabolic, or hepatic disease were also excluded. Previous use of DMARDs, cyclosporine, azathioprine, or TNF-blockers was allowed. However, a maximum of ten patients with previous TNF-blocking therapy were allowed, to account for the possibility that response rates for secukinumab in such patients could be lower, thereby biasing the mean efficcacy readout. Washout periods of 3 months for adalimumab and certolizumab and 2 months for etanercept and infliximab were required before baseline. Patients were allowed to continue any of the following medications if the dosage remained stable for at least 4 weeks before the baseline visit and during the study: sulfasalazine (up to 3 g a day), methotrexate (up to 25 mg a week), prednisone or prednisone equivalent (up to 10 mg a day), and NSAIDs. 136

138 Study medication and dosage The intravenous secukinumab dose of 10 mg/kg was based on previous pharmacokinetic and pharmacodynamic experience with secukinumab in rheumatoid arthritis. 15 A treatment regimen of two infusions of 10 mg/kg given intravenously 3 weeks apart was used to achieve high systemic exposure for induction of response, and because the 4-week toxicology coverage available at study start did not allow a longer treatment period. Endpoints The primary efficacy endpoint was the percentage of patients with a 20% response according to the Assessment of SpondyloArthritis international Society criteria for improvement (ASAS20) at week Secondary efficacy endpoints included ASAS40 (40% response according to ASAS criteria for improvement) and ASAS5/6 responses (improvement in five of six domains: pain, patient global assessment, function, inflammation, spinal mobility, C-reactive protein [acutephase reactant] without deterioration in the 6th domain), BASDAI, 22 and assessment with the ankylosing spondylitis quality-of-life (ASQoL) measure. 23 All of these assessments were done at screening, baseline, days 8, 15, and 29, and weeks 6, 8, 10, 12, 16, 20, 24, and 28. Biological endpoints, which were measured in secukinumab-treated patients, included erythrocyte sedimentation rate and serum C-reactive protein (CRP). Additionally, analyses of S100-proteins A8, A9, and A12 in secukinumab patients were done by multiplexed liquid chromatography mass spectrometry multiple reaction monitoring. Sagittal MRI of the entire spine in secukinumab-treated and placebo-treated patients was done using short tau inversion recovery and T1-weighted sequencing to assess bone marrow oedema at baseline, week 6, and week 28. MRI studies were analysed by an independent reader, who was blinded to treatment allocation and chronology of images, using the Berlin score. 24 To further assess the relevance of targeting IL-17A, 13 genetic polymorphisms that are either associated with ankylosing spondylitis or directly related to IL-17 were tested for association with response to secukinumab (appendix) All genotyping was done by Novartis: single-nucleotide polymorphism (SNP) using TaqMan, and HLA genotyping using sequencespecific oligonucleotide probes. ERAP1 transcript levels in whole blood were determined using Affymetrix DNA microarrays (Santa Clara, CA, USA). Adverse events and vital signs were assessed at every visit for safety evaluations. Laboratory measurements were done at screening; baseline; days 2, 8, 15, 23, and 29; and weeks 6, 8, 10, 12, 16, 20, 24, and 28. Randomisation and blinding details are provided in the appendix. Statistical analysis The primary endpoint was analysed using Bayesian methods, as described previously. 28 For the placebo ASAS20 response rate, data from eight previous trials of ankylosing spondylitis (N=533) were included. 29 The information contained in these trials was transformed into an informative prior distribution by means of the meta-analytic-predictive methodology described by Neuenschwander and colleagues. 30 This approach accounts for between-trial heterogeneity, since the data from the previous trials are downweighted compared with the data collected in this trial. The resulting prior distribution for the placebo response rate was equivalent to 43 patients (appendix). For the secukinumab response rate, an uninformative prior was used. 10 Secukinumab anti-il-17a therapy in AS 137

139 The predefined criterion to declare a positive proof-of-concept required a posterior probability of at least 95% that the ASAS20 response rate on secukinumab is larger than that on placebo. With data from 20 patients on secukinumab and five patients on placebo, the study showed a 90% probability of achieving proof-of-concept, assuming true response rates of 25% on placebo and 60% on secukinumab (appendix). Secondary and exploratory endpoint analyses were non-bayesian and are summarised using descriptive statistics. Exploratory (post-hoc) analyses of MRI total scores and protein biomarkers were done to compare the change from baseline and post-treatment measurements (Wilcoxon signed-rank test). Patients who dropped out of the study were considered non-responders in all planned ASAS assessments, irrespective of the reason for discontinuation. Comparisons between baseline and post-treatment measurements, as well as between ASAS responders and non-responders, applied Mann-Whitney tests. An exploratory analysis of genetic data (13 polymorphisms shown to be associated with ankylosing spondylitis or in the target IL-17A gene region) was done with a subgroup of consenting patients taking secukinumab (N=22; appendix). The polymorphisms were tested individually for association with ASAS20 and ASAS40 response at week 6 using logistic regression models and with BASDAI using a linear regression model, with number of copies of the minor (less common) allele carried by the individual as the predictor and baseline BASDAI score as a covariate. A permutation test was done to account for small sample size and multiple comparisons of polymorphisms, with genotypes randomly permuted times. Role of the funding source An academic advisory committee was involved in study design and data interpretation, together with authors from Novartis Institutes for Biomedical Research. The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication. RESULTS Of 37 patients with active ankylosing spondylitis who were screened, 30 were enrolled in the study and randomly assigned to receive intravenous 2 10 mg/kg secukinumab (n=24) or placebo (n=6) on day 1 and day 22 (figure 1). Baseline demographics were similar between the two groups (table 1). 22 of the 24 (92%) secukinumab-treated patients and three of the six (50%) placebo-treated patients reached week 6. Of the 25 patients still in the study at week 6, 15 in the secukinumab group and all three in the placebo group completed the study up to week 28. Patients who discontinued the study were considered to be non-responders for the efficacy analysis. Reasons for discontinuation were mainly unsatisfactory therapeutic effect, withdrawal of consent, and adverse events. Additionally, a dosing error led to one patient in the secukinumab group being excluded from the efficacy analysis but not from the safety analysis (as predefined by the protocol). Thus the efficacy analysis included 23 secukinumab patients and six placebo patients. The primary efficacy outcome, ASAS20 response at week 6, was met by 14 (61%) of the 23 patients in the secukinumab group and one (17%) of the six patients in the placebo group. Calculated response rates were 59% for the secukinumab group versus 24% for the placebo 138

140 Figure 1. Trial profile group. The observed placebo response rate was similar to the results obtained from the metaanalysis of eight comparable trials (26.5%). 29 Bayesian analysis (table 2) indicated that the probability of secukinumab inducing greater response rates than placebo was 99.8%. The secondary efficacy endpoints, ASAS20, ASAS40, and ASAS5/6 response rates, and BASDAI scores up to week 28 are summarised in the appendix and shown for the secukinumabtreated patients in figure 2. Response rates for all ASAS measures and BASDAI were seen at week 6. Consistent with the loading dose regimen, there was a decreasing response rate by week 28. Clinically relevant improvement in quality of life, defined as a change of two or more points in ASQoL, was seen in 11 of 21 (52%) patients in the secukinumab group versus two of six (33%) patients in the placebo group at day 29. Additional ASQoL data are shown in the appendix. Mean serum CRP levels decreased from 14.4 mg/l (SD 17.60) at baseline to 6.1 mg/l (9.23) at week 6 (p=0.03; figure 3A). Similarly, mean erythrocyte sedimentation rate decreased from 35.5 mm/h (SD 24.81) at baseline to 22.4 mm/h (16.27) at week 6 after secukinumab treatment (p=0.11; appendix). Mean levels of the inflammatory biomarker protein S100A8 decreased from ng/ml (SD ) at baseline to ng/ml (478.37) at week 6 (p=0.01; figure 3B). Mean levels of the inflammatory biomarker protein S100A9 decreased from ng/ml (SD ) at baseline to ng/ml (445.28) at week 6 (p=0.01; figure 3C). Additionally, ASAS20 response at week 6 was significantly correlated with both baseline levels of S100A8 (R=0.55) and A9 (0.39) and changes between baseline and week 6 levels of S100A8 (0.43) and A9 (0.28). Even stronger correlations were seen with the ASAS40 response at week 6 (R=0.40, 0.59, 0.52, 10 Secukinumab anti-il-17a therapy in AS 139

141 Table 1. Baseline characteristics Safety analysis dataset Secukinumab (N=24) Placebo (N=6) Total (N=30) Age (years) 41.1± ± ±10.03 Men no. (%) 14 (58) 5 (83) 19 (63) Race no. (%) White 20 (83) 6 (100) 26 (87) Black 1 (4) - 1 (3) Asian 1 (4) - 1 (3) Other 2 (8) - 2 (7) Height (cm) 170.6± ± ±7.92 Weight (kg) 78.9± ± ±15.14 Body-mass index 27.1± ± ±4.83 Efficacy analysis dataset Disease duration (years) 10.1± ± ±11.95 HLA-B27 no. (%) 16 (70) 5 (83) 21 (72) History of extra-axial involvement no. (%) Uveitis 7 (30) 2 (33) 9 (31) Psoriasis 3 (13) 1 (17) 4 (14) Inflammatory bowel disease 3 (13) 1 (17) 4 (14) Dactylitis 1 (4) 1 (17) 2 (7) Peripheral arthritis 12 (52) 4 (67) 16 (55) Concomitant DMARD no. (%) 8 (35) 3 (50) 11 (38) Methotrexate 4 (17) - 4 (14) Leflunomide 1 (4) - 1 (3) Sulfasalazine 5 (22) 3 (50) 8 (28) Other immunosuppressant 2 (9) - 2 (7) NSAID no. (%) 22 (96) 6 (100) 28 (97) Steroid no. (%) 3 (13) - 3 (10) Oral 2 (9) - 2 (7) Topical 1 (4) - 1 (3) Prior TNF-blocker no. (%) 10 (43) 3 (50) 13 (45) ASQoL 12.1± ± ±3.72 BASDAI 7.1± ± ±1.45 BASMI 4.3± ± ±1.73 BASFI 6.4± ± ±2.18 MASES 4.3± ± ±3.40 LEI 1.2± ± ±1.55 CRP (mg/l) 13.3± ± ±16.65 Values are means ±SD. DMARD=disease-modifying anti-rheumatic drug, NSAID=non-steroidal anti-inflammatory drug, TNF=tumor necrosis factor, ASQoL=Ankylosing Spondylitis Quality of Life, BASDAI=Bath Ankylosing Spondylitis Disease Activity Index, BASMI=Bath Ankylosing Spondylitis Metrology Index, BASFI=Bath Ankylosing Spondylitis Functional Index, MASES=Maastricht Ankylosing Spondylitis Enthesis Score, LEI=Leeds Enthesis Index, CRP=C-reactive protein. Secukinumab 2x10 mg/ kg. Efficacy analysis dataset included only 29 participants, since one patient on secukinumab was excluded due to a dosing error. Azathioprine or cyclosporine.

142 Figure 2. Primary and secondary endpoints over time in patients receiving secukinumab 2 10 mg/kg. Arrows below x-axis denote drug administration. Error bars in ASAS graphs represent mean (SE). Error bars in BASDAI graph represent mean (SD). ASAS=Assessment of SpondyloArthritis international Society criteria. BASDAI=Bath ankylosing spondylitis disease activity index. 10 and 0.64, respectively), indicating that clinical improvement was paralleled by a decrease in inflammatory parameters (appendix). Findings from direct visualisation of the axial inflammatory lesions by MRI (figure 4) were consistent with the clinical effects and results seen for the protein measures. The mean MRI scores decreased from 9.2 (SD 8.87) at baseline to 6.6 (6.56) at week 6 (p=0.10) and 5.7 (6.20) at week 28 (p=0.16) in the secukinumab group (appendix). The mean MRI scores were unchanged in the placebo group: 20.6 (SD 20.18) at baseline, 21.0 (24.56) at week 6, and 19.0 (19.33) at week 28. For the genetic biomarkers, a non-synonymous SNP of the ERAP1 gene, rs30187, which leads to significantly reduced endopeptidase activity in vitro and is associated with protection against ankylosing spondylitis, 27,31 had a highly significant association with ASAS40 at week 6 in secukinumab-treated patients (nominal p= ⁵ and empirical p=0.004 in permutation test after correction for multiple comparisons). The association of SNP rs30187 with the ASAS40 response was consistent over time (figure 5A). The analysis also yielded a significant p value when restricted to the HLA-B27 positive subset of patients (n=17, p= for association between rs30187 and ASAS40 at week 6). A similar, but weaker association with ASAS20 was seen for Secukinumab anti-il-17a therapy in AS 141

143 Table 2. Primary endpoint Bayesian analysis of ASAS20 responders at week 6 Treatment Responders, n (%) Response rate* Difference vs Placebo) 95% credibility interval Probability Secukinumab 14 (60.9) 59.2% 34.7% 11.5%-56.4% 99.8% Placebo 1 (167) 24.5% ASAS=Assessment of SpondyloArthritis international Society criteria. *Means from the posterior beta (0.5 + x, 1 + n x) distribution for secukinumab and beta (11 + x, 32 + n x) distribution for placebo, where x represents the number of responders and n x represents the number of non-responders in the corresponding treatment group. Difference in response rates simulated from the posterior probability distributions of secukinumab and placebo. Secukinumab 2 10 mg/kg. The efficacy dataset included only 23 of 24 patients in the secukinumab group, since one patient was excluded due to a dosing error. Figure 3. Biological markers of response. C-reactive protein (A), S100A8 (B), and S100A9 (C) levels at week 6, after secukinumab (2 10 mg/kg) administration at weeks 0 and

144 Figure 4. Reduction of inf ammation in thoracic (T2 T3) and (T7 T8) units at baseline and at week 6 in a patient treated with secukinumab. 10 Secukinumab anti-il-17a therapy in AS 143