The Incidence of Uveitis in Ankylosing Spondylitis Patients Undergoing Tumor Necrosis Factor Inhibiting Therapy in Korea

Journal of Rheumatic Diseases Vol. 22, No. 5, October, 2015 http://dx.doi.org/10.4078/jrd.2015.22.5.288 Original Article The Incidence of Uveitis in Ankylosing Spondylitis Patients Undergoing Tumor Necrosis Factor Inhibiting Therapy in Korea Bon San Koo 1 *, Seokchan Hong 2 *, You Jae Kim 2, Chang-Keun Lee 2, Bin Yoo 2, Yong-Gil Kim 2 1 Division of Rheumatology, Department of Internal Medicine, Konkuk University Chungju Hospital, Konkuk University College of Medicine, Chungju, 2 Division of Rheumatology, Department of Internal Medicine, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea Objective. The purpose of this study is to evaluate the outcome of uveitis in ankylosing spondylitis (AS) during tumor necrosis factor (TNF)-inhibiting therapy and to compare the incidence rate of uveitis in infliximab, adalimumab, and etanercept. Methods. A retrospective evaluation was performed in AS patients who had started TNF-inhibiting therapy from June 2003 to June 2011. The clinical characteristics of patients with documented uveitis were evaluated. Results. Among 316 patients treated with TNF inhibitor, 26 patients (8%) had experienced uveitis during TNF-inhibiting therapy. Among them, 15 patients were treated with etanercept, eight with adalimumab, and three with infliximab. The overall incidence rate of uveitis flare during therapy with TNF inhibitor was 46 per 1,000 person-years (pys) (95% confidence interval [CI], 32 to 64). The incidence rate did not differ between TNF inhibitors, with 54/1,000 pys (95% CI, 34 to 81) for etanercept, 46/1,000 pys (95% CI, 21 to 87) for adalimumab, and 22/1,000 pys (95% CI, 5 to 64) for infliximab. Fourteen patients experienced a first episode of uveitis. The overall incidence rate of new onset-uveitis after therapy with TNF inhibitor was 19 per 1,000 pys (95% CI, 10 to 31). The incidence rate for etanercept was 24/1,000 pys (95% CI, 12 to 45); adalimumab, 15/1,000 pys (95% CI, 3 to 45); and infliximab, 7/1,000 pys (95% CI, 0 to 40). There was no statistical difference in the incidence of uveitis flare or the cumulative uveitis-free rate among the three TNF inhibitors. Conclusion. The relative rate of uveitis, including the first episode, was determined using the TNF inhibitor. However, there was no difference in the incidence rate of uveitis among the three TNF inhibitors. (J Rheum Dis 2015;22:288-292) Key Words. Tumor necrosis factor, Ankylosing spondylitis, Uveitis INTRODUCTION Uveitis is the most common extra-articular manifestation occurring in patients with ankylosing spondylitis (AS) [1,2]. Patients with AS have a 20% to 30% chance of developing uveitis during the course of their disease, and 85% of the AS patients who experience uveitis are diagnosed with acute anterior uveitis [1]. It generally affects men and is unilateral, painful, and self-limiting. Tumor necrosis factor (TNF) inhibitors have been known to be effective agents in treating articular as well as extra-articular manifestations of AS. However, differing from monoclonal antibodies, concerns have been raised that etanercept may have a less protective effect on uveitis flare [3-7]. Although the pathogenic mechanisms have not yet been fully identified, a recombinant soluble receptor to TNF might be related to lack of its effect on uveitis flare or its paradoxically adverse effect of ocular inflammation. In our study, we investigated the incidence rate of uveitis in AS patients treated with TNF inhibitors in Korea. We also compared the incidence rate of uveitis in patients treated with infliximab, adalimumab, and etanercept. Received:March 12, 2015, Revised:June 4, 2015, Accepted:June 5, 2015 Corresponding to:yong-gil Kim, Division of Rheumatology, Department of Internal Medicine, Asan Medical Center, University of Ulsan College of Medicine, 88 Olympic-ro 43-gil, Songpa-gu, Seoul 05505, Korea. E-mail : bestmd2000@amc.seoul.kr *The first two authors contributed equally to this work. pissn: 2093-940X, eissn: 2233-4718 Copyright c 2015 by The Korean College of Rheumatology. All rights reserved. This is a Free Access article, which permits unrestricted non-commerical use, distribution, and reproduction in any medium, provided the original work is properly cited. 288

Uveitis during TNF Inhibitor MATERIALS AND METHODS We retrospectively reviewed the electronic medical reports (EMR) of 316 AS patients treated with TNF-inhibiting therapy in a tertiary care center in Korea from June 2003 to June 2011. All of the patients had been previously diagnosed as having AS which satisfied the modified New York criteria [8] and had received at least one TNF inhibitor, including etanercept, adalimumab or infliximab, due to the lack of efficacy or the adverse effect of non-steroidal, anti-inflammatory drugs and disease modifying, anti-rheumatic drugs. This study was approved by the Asan Medical Center Institutional Review Board (IRB 2013-0882). A previous history of uveitis was evaluated at the first medical examination and it was recorded in the EMR. We defined uveitis patients as those with at least one episode of uveitis diagnosed by an ophthalmologist. Clinical data, including patient age, sex, disease duration, human leukocyte antigen (HLA)-B27 allele, body-mass index, and duration of TNF inhibitor exposure, were collected. When comparing the three TNF inhibitors, the chi-square test or Kruskal-the Wallis test was performed. The incidence rates of uveitis were calculated as the number of events or patients per 1,000 person-years (pys) of follow-up with a 95% confidential interval (CI) in the infliximab, adalimumab, and etanercept groups. To estimate the difference in the uveitis risk in the three groups, a survival analysis was performed using the Kaplan-Meier method and Cox proportional hazards model. The log-rank test was used to compare survival curves. A p<0.05 was considered statistically significant. All statistical analyses were performed using Predictive Analytics Software SPSS Statistics ver. 17.0 (SPSS Inc., Chicago, IL, USA). RESULTS Among the 316 AS patients treated with TNF inhibitor, 42 patients (13%) experienced uveitis before TNF-inhibiting therapy. Table 1 shows a comparison of the infliximab, adalimumab, and etanercept groups. Infliximab was used for 85 patients, adalimumab for 148 patients, and etanercept for 142 patients. Male predominance and high HLA-B27 positivity were similar in all of the groups. As each TNF inhibitor was approved by Korean National Health Insurance at points in time, the exposure duration differed in each group. In patients treated with infliximab and adalimumab, the number of patients who experienced uveitis was decreased following their treatment, although not in those treated with etancercept. The overall incidence rate of uveitis flare in all TNF inhibitors was 46 per 1,000 pys (95% CI, 32 to 64). The incidence rate for etanercept was Table 1. Comparison of the clinical characteristics of AS patients treated with infliximab, adalimumab, or etanercept Characteristic Total (n=316) Infliximab (n=85) Adalimumab (n=148) Etanercept (n=142) p-value Age at the start of TNFi (yr) 36.1±11.8 36.4±11.6 36.3±11.6 36.4±12.4 0.987 Male (%) 81 77 74 83 0.189 HLA-B27 positivity (%) 95 95 94 95 0.901 AS disease duration (mo) 75.7±52.9 86.1±54.8 58.63±50.1 86.8±51.1 0.000 Duration of TNFi exposure (mo) 23.9±19.4 19.5±15.0 16.1±12.5 34.6±22.6 0.001 Body mass index (kg/m 2 ) 24.2±3.9 23.9±3.5 24.5±4.3 24.1±4.0 0.598 Previous history of uveitis 42 (13) 6 (7) 21 (14) 15 (11) 0.252 Uveitis during TNFi therapy 26 (8) 3 (4) 8 (5) 15 (11) 0.000 Uveitis during TNFi therapy (episode) 34 (11) 3 (4) 9 (6) 22 (15) 0.004 Incidence rate of uveitis during TNFi 46 (32 64) 22 (5 64) 46 (21 87) 54 (34 81) 0.257* therapy (per 1,000 pys) New onset of uveitis during TNFi 14 (4) 1 (1) 3 (2) 10 (7) 0.029 therapy Incidence rate of new onset of uveitis during TNFi therapy (per 1,000 pys) 19 (10 31) 7 (0 40) 15 (3 45) 24 (12 45) 0.357* Values are presented as mean±standard deviation, percent only, number (%), or median (95% CI). AS: ankylosing spondylitis, HLA: human leukocyte antigen, TNFi: tumor necrosis factor inhibitor, CI: confidence interval, pys: person-years. *Using Poisson regression; including 59 patients who were treated with second or third TNFi. www.jrd.or.kr 289

Bon San Koo et al. 54/1,000 pys (95% CI, 34 to 81) for adalimumab, 46/1,000 pys (95% CI, 21 to 87), and for infliximab, 22/1,000 pys (95% CI, 5 to 64). The incidence rate of uveitis flare did not different among the three groups even in cases of new onset. Of the 42 patients who experienced uveitis, nine patients were changed to another TNF inhibitor, i.e. five due to uveitis, three due to the lack of efficacy of axial symptoms, and one due to an injection site reaction. Interestingly, among the five patients who switched the TNF inhibitor due to uveitis, four patients changed etancercept to adalimumab and one changed etanercept to infliximab. And after alternative TNF-inhibiting therapy, they never suffered from uveitis again. Among the three patients who changed their TNF inhibitor due to its lack of efficacy, one suffered from uveitis flare after changing adalimumab to etancercept. Fourteen patients experienced their first episode of uveitis after TNF-inhibiting therapy and which was observed more often in the etanercept group (10 patients) and also in the adalimumab group (three patients) and the infliximab group (one patient). The overall incidence rate of new onset uveitis after TNF-inhibiting therapy was 19 per 1,000 pys (95% CI, 10 to 31). The incidence rate for etanercept was 24/1,000 pys (95% CI, 12 to 45), that for adalimumab was 15/1,000 pys (95% CI, 3 to 45), and that for infliximab was 7/1,000 pys (95% CI, 0 to 40), respectively. The incidence rate of patients who used etanercept was 1.6 to 3.4 times higher than that of infliximab and adalimumab; however, there was statistically no difference of incidence rate in three TNF inhibitors. The duration of TNF-inhibiting therapy before the new onset of uveitis was not statistically different. The cumulative, uveitis-free rate was analyzed in the three TNF inhibitors and there was no difference in the development of uveitis (Figure 1A). Moreover, the rate of developing uveitis in patients who had not previously experienced it before starting the TNF inhibitor also did not differ between them (Figure 1B). The risk of developing uveitis and new onset of uveitis during TNF inhibiting therapy increased with treatment with adalimumab and etanercept without significant difference (Table 2). The hazard ratio adjusted for disease duration and duration of TNF inhibitor exposure was also analyzed (hazard ratio 0.9 to 3.1) and there was no significant difference. DISCUSSION In our AS cohort, 26 patients (8%) developed uveitis and 14 patients (4%) experienced new onset of uveitis during TNF-inhibiting therapy. Although there have been previous reports regarding uveitis flare during etanercept therapy, in our study there was no difference in the rate of developing uveitis in the three TNF inhibitors. The TNF inhibitor is a highly effective agent for controlling the activity of AS, including extra-articular manifestations (EAM). The TNF inhibitor can also decrease the incidence of uveitis flare which is a major EAM of AS [9-11]. Although our study did not compare numerous episodes of uveitis flare before and after each TNF inhibitor exposure, the number of patients who developed uveitis during TNF-inhibitor therapy was low (8%) com- Figure 1. Cumulative uveitis-free survival rate after tumor necrosis fator (TNF) inhibitors (A) in all patients and (B) in patients who had never experienced uveitis before TNF-inhibiting therapy. 290 J Rheum Dis Vol. 22, No. 5, October, 2015

Uveitis during TNF Inhibitor Table 2. Cox proportional hazard analysis for uveitis of ankylosing spondylitis patients treated with infliximab, adalimumab, or etanercept Characteristic Infliximab (n=85) Adalimumab (n=148) Etanercept (n=142) 95% CI p 95% CI p HR developing uveitis during TNFi therapy Reference 2.0 (0.5 7.8) 0.295 1.6 (0.5 5.8) 0.461 Adjusted HR developing uveitis during TNFi Reference 1.8 (0.4 7.9) 0.429 1.7 (0.4 6.5) 0.443 therapy* HR of new onset of uveitis Reference 2.8 (0.3 28.2) 0.37 2.8 (0.3 22.7) 0.336 Adjusted HR of new onset of uveitis* Reference 0.9 (0.1 12.3) 0.98 3.1 (0.3 27.0) 0.312 CI: confidence interval, HR: hazard ratio, TNFi: tumor necrosis factor inhibitor. *Adjusted for ankylosing spondylitis disease duration and duration of TNFi exposure. pared to AS patients in general, who usually have a 20% to 30% chance of developing uveitis [1]. However, some studies have reported that etanercept did not reduce the incidence of uveitis differently from that of TNF monoclonal antibodies. Guignard et al. [11] reported a difference in the efficacies of the three TNF inhibitors in 46 spondyloarthropathy patients who had at least one uveitis flare. They suggested that etanercept did not reduce the number of flares, whereas infliximab and adalimumab greatly reduced the number of flares. However, it has also been seen in AS clinical trials that etanercept is effective for lowering the uveitis flare rate in AS patients [12]. Our study showed that developing a uveitis flare and the new onset of uveitis did not statistically differ in the three TNF inhibitors. However, five of our patients who were previously treated with etanercept were switched to infliximab or adalimumab alternatively after their uveitis flare, and after which they experienced no subsequent flare. Considering the results of previous reports related to TNF inhibitor and uveitis, these clinical experiences support the supposition that etanercept might be less effective for treating uveitis compared to the other TNF inhibitors. Etanercept is a soluble TNF receptor molecule, whereas infliximab and adalimumab are monoclonal antibodies. Because their mechanisms of antagonizing TNF-α differ, there are pharmacodynamic and pharmacokinetic differences between the TNF inhibitors that may potentially contribute to their action in different disease states [4,13,14]. This evidence suggests that the monoclonal antibodies may be more appropriate than etanercept for treating EAM, including uveitis [15]. Although our study showed that there was no difference in the uveitis flare among the three TNF inhibitors, changing from etanercept to a monoclonal antibody, including infliximab and adalimumab, could be effective for preventing uveitis flare in patients experienced uveitis before. Even if the cohort was small, our patients characteristics showed a male predominance and a high percentage of HLA-B27 positivity similar to that seen in other reports of AS patients who developed uveitis during TNF-inhibiting therapy [4,5,12,16]. In addition, most studies reporting uveitis onset stated that the duration of TNF-inhibitor exposure was from one to 96 months [16]. In our study, with regard to the new onset of uveitis, the mean duration of TNF-inhibiting therapy was 30.5 months and the range was four to 71 months. Our study has several limitations. First, because of retrospective study in single center, small number of patients and loss of follow-up might be influence the incidence rate of uveitis. Second, we could not determine the cause for developing uveitis during TNF-inhibiting therapy and whether it was an adverse effect or the lack of efficacy of the TNF inhibitor. Third, the physician s preference for patients, based on their age and comorbidity, could have affected the choice of TNF inhibitor. CONCLUSION In conclusion, a considerable proportion of the AS patients experienced uveitis flare and approximately half of them developed new uveitis onset even during their treatment with the TNF inhibitor. Although there was no difference in the uveitis flare in the three TNF inhibitors in our study, it might be helpful for patients developing uveitis during etanercept therapy to change to monoclonal antibodies. www.jrd.or.kr 291

Bon San Koo et al. CONFLICT OF INTEREST No potential conflict of interest relevant to this article was reported. REFERENCES 1. Rosenbaum JT. Acute anterior uveitis and spondyloarthropathies. Rheum Dis Clin North Am 1992;18:143-51. 2. Zeboulon N, Dougados M, Gossec L. Prevalence and characteristics of uveitis in the spondyloarthropathies: a systematic literature review. Ann Rheum Dis 2008;67:955-9. 3. Coates LC, McGonagle DG, Bennett AN, Emery P, Marzo- Ortega H. Uveitis and tumour necrosis factor blockade in ankylosing spondylitis. Ann Rheum Dis 2008;67:729-30. 4. Kakkassery V, Mergler S, Pleyer U. Anti-TNF-alpha treatment: a possible promoter in endogenous uveitis? Observational report on six patients: occurrence of uveitis following etanercept treatment. Curr Eye Res 2010;35:751-6. 5. Lim LL, Fraunfelder FW, Rosenbaum JT. Do tumor necrosis factor inhibitors cause uveitis? A registry-based study. Arthritis Rheum 2007;56:3248-52. 6. Cobo-Ibáñez T, del Carmen Ordóñez M, Muñoz-Fernández S, Madero-Prado R, Martín-Mola E. Do TNF-blockers reduce or induce uveitis? Rheumatology (Oxford) 2008;47: 731-2. 7. Gaujoux-Viala C, Giampietro C, Gaujoux T, Ea HK, Prati C, Orcel P, et al. Scleritis: a paradoxical effect of etanercept? Etanercept-associated inflammatory eye disease. J Rheumatol 2012;39:233-9. 8. 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;27:361-8. 9. Rosenbaum JT. Effect of etanercept on iritis in patients with ankylosing spondylitis. Arthritis Rheum 2004;50:3736-7. 10. 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;52:2447-51. 11. Guignard S, Gossec L, Salliot C, Ruyssen-Witrand A, Luc M, Duclos M, et al. Efficacy of tumour necrosis factor blockers in reducing uveitis flares in patients with spondylarthropathy: a retrospective study. Ann Rheum Dis 2006;65: 1631-4. 12. Sieper J, Koenig A, Baumgartner S, Wishneski C, Foehl J, Vlahos B, et al. Analysis of uveitis rates across all etanercept ankylosing spondylitis clinical trials. Ann Rheum Dis 2010;69:226-9. 13. Ranganathan P. Pharmacogenomics of tumor necrosis factor antagonists in rheumatoid arthritis. Pharmacogenomics 2005;6:481-90. 14. Van den Brande JM, Braat H, van den Brink GR, Versteeg HH, Bauer CA, Hoedemaeker I, et al. Infliximab but not etanercept induces apoptosis in lamina propria T-lymphocytes from patients with Crohn's disease. Gastroenterology 2003;124:1774-85. 15. Elewaut D, Matucci-Cerinic M. Treatment of ankylosing spondylitis and extra-articular manifestations in everyday rheumatology practice. Rheumatology (Oxford) 2009;48: 1029-35. 16. Wendling D, Paccou J, Berthelot JM, Flipo RM, Guillaume- Czitrom S, Prati C, et al; CRI. New onset of uveitis during anti-tumor necrosis factor treatment for rheumatic diseases. Semin Arthritis Rheum 2011;41:503-10. 292 J Rheum Dis Vol. 22, No. 5, October, 2015