Dr. Cacoub has received consultant fees, speaking fees, and/or honoraria (less than $10,000 each) from AstraZeneca, BMS, Sanofi-Aventis, Gilead, and Schering-Plough, and (more than $10,000 each) from Roche and Servier.
Azathioprine in severe uveitis of Behçet's disease
Article first published online: 27 JUL 2010
Copyright © 2010 by the American College of Rheumatology
Arthritis Care & Research
Volume 62, Issue 12, pages 1733–1738, December 2010
How to Cite
Saadoun, D., Wechsler, B., Terrada, C., Hajage, D., Le Thi Huong, D., Resche-Rigon, M., Cassoux, N., Le Hoang, P., Amoura, Z., Bodaghi, B. and Cacoub, P. (2010), Azathioprine in severe uveitis of Behçet's disease. Arthritis Care Res, 62: 1733–1738. doi: 10.1002/acr.20308
- Issue published online: 30 NOV 2010
- Article first published online: 27 JUL 2010
- Accepted manuscript online: 27 JUL 2010 12:00AM EST
- Manuscript Accepted: 20 JUL 2010
- Manuscript Received: 5 FEB 2010
To investigate the efficacy and tolerance of azathioprine in severe uveitis related to Behçet's disease (BD).
We reported 157 consecutive patients with severe uveitis (active posterior uveitis or panuveitis) fulfilling the international criteria for BD and treated with corticosteroids (0.5–1 mg/kg/day) and azathioprine (2.5 mg/kg/day). Long-term outcome and factors associated with complete remission were assessed.
Mean ± SD age at diagnosis was 29.9 ± 10.1 years, with 71.3% men. At baseline, 59 patients (37.6%) had loss of useful vision, 54 (34.4%) had retinal vasculitis, 66 (42.0%) had panuveitis, and 132 (84.1%) had bilateral uveitis. Following azathioprine therapy, 81 patients (51.6%) were complete responders, 65 (41.4%) were partial responders, and 11 (7%) were nonresponders. The visual acuity significantly improved (P < 0.001), and a significant decrease in the mean oral prednisone dosage (55.3 to 10.5 mg/day; P < 0.001) was observed after therapy. Patients with retinal vasculitis (odds ratio [OR] 0.45 [95% confidence interval (95% CI) 0.2–0.9], P = 0.02) and severe visual loss (OR 0.28 [95% CI 0.2–0.7], P < 0.0001) at diagnosis were less likely to be complete responders. Azathioprine was well tolerated, with only 3 withdrawals due to hepatotoxic effects (n = 2) and bacterial septicemia (n = 1).
Azathioprine represents an effective and safe therapy in patients with severe uveitis of BD.
Behçet's disease (BD) is a vasculitis of unknown etiology characterized by mucocutaneous, ocular, arthritic, vascular, and central nervous system manifestations (1, 2). Uveitis is one of the most severe complications of the disease. As many as 50% of patients with ocular involvement became legally blind within 5 years after the onset of symptoms in the 1980s (3). The prognosis has improved in recent years with the increasing use of steroids and immunosuppressive agents (3, 4). Eye involvement in BD follows a remitting–relapsing course and the recurrent inflammatory attacks result in irreversible damage and visual loss. Suppression of inflammation and the prevention of recurrences of ocular attacks should be the goals. Corticosteroids rapidly suppress the inflammation, but potential side effects, including cataracts and glaucoma, cause concern. The European League Against Rheumatism recommends azathioprine as the initial agent for ocular involvement of BD (5). However, data regarding its efficacy in large cohorts of BD patients with severe uveitis are lacking.
In this study, we reported 157 consecutive patients with severe uveitis (active posterior uveitis or panuveitis) fulfilling the international criteria for BD, treated with corticosteroids and azathioprine, and seen in a single university hospital (Pitié-Salpétrière, Paris, France) between 1970 and 2006. Long-term outcome and factors associated with clinical remission were assessed.
PATIENTS AND METHODS
Clinical records of 157 consecutive patients treated with corticosteroids and azathioprine for severe uveitis and fulfilling the international criteria for the classification of BD (6) were analyzed. All of the patients were referred to the internal medicine and ophthalmology departments of the Pitié-Salpétrière University Hospital, Paris, France, between 1970 and 2006. To be included, patients had to present active posterior uveitis or panuveitis, had to receive corticosteroids and azathioprine as the sole immunosuppressive drug, and had to have a minimum followup of 12 months after initiation of azathioprine. Uveitis associated with a decrease in visual acuity below 20/100, an increase in the vitreous haze of ≥2, a relapse of cystoid macular edema, or serous retinal detachment was defined as severe. Clinical and laboratory findings of all of the patients were analyzed retrospectively by both internists and ophthalmologists before inclusion and during followup. Patients with evidence of active infection, blood abnormalities, liver function abnormalities, and/or concern about treatment compliance were excluded.
Patients were regularly followed in both the internal medicine and ophthalmologic departments. For each patient, the following data were collected: age at diagnosis of uveitis; sex; date of criteria for BD; main features of BD, including cutaneous, ocular, rheumatologic, neurologic, and/or vascular involvement; clinical features; and outcome and treatment of uveitis. Lumbar puncture and brain magnetic resonance imaging were performed when clinically indicated. Laboratory tests, including complete blood cell counts, liver transaminases, and creatininemia, were performed at regular intervals.
Ophthalmologic examination included measurement of the best-corrected visual acuity, tonometry, and slit lamp examination. Funduscopy was performed to assess vitritis, retinal hemorrhages, vasculitis, papillitis, and retinal ischemia. Fluorescein angiography was performed in all cases. Improvement of uveitis was based on international criteria and was defined as a 2-step decrease in the level of inflammation or a decrease to grade 0 (7). Main outcome measures included relapses, best-corrected visual acuity, and dependence on steroids during azathioprine therapy.
All of the patients received oral azathioprine 2.5 mg/kg/day. Azathioprine was initiated in association with oral prednisone (0.5–1 mg/kg/day). Fifty-eight (36.9%) of the 157 BD patients were considered steroid dependent (≥15 mg/day of prednisone) and received azathioprine as a steroid-sparing agent. Intravenous pulses of methylprednisolone (1 gm per day over 3 days) preceded oral prednisone in 57 patients. Thirty-one patients (19.6%) had been previously treated by another regimen (i.e., cyclophosphamide, cyclosporin, chlorambucil, and interferon [IFN] alfa-2a). Doses of steroids were tapered based on clinical improvement of uveitis starting 1 month after initiation. Steroids were then progressively tapered to obtain an average daily dosage of 0.25–0.5 mg/kg/day at 3 months and of 10 mg at 6 months. No other therapy was allowed except colchicine and topical steroids. No patients received intraocular or periocular steroids. The corticosteroid threshold was evaluated according to the dose, which, if tapered, led to relapse. In our tertiary care center for autoimmune diseases, the objective of corticosteroid tapering ranged between 7 and 10 mg/day. Patients remained receiving approximately 10 mg/day of prednisone for a long duration to avoid relapses (8). The duration of azathioprine was left to the discretion of the clinician in charge of the patient. The median duration of azathioprine therapy was 3.4 years (range 1–5 years). Based on the criteria recently reported by the Standardization of Uveitis Nomenclature Working Group, improved activity was defined as a 2-step decrease in the level of inflammation or a decrease to grade 0 (7). Evolution of visual acuity, control of intraocular inflammation (regression of retinal vasculitis), and levels of steroid dependence were evaluated as major criteria for drug efficacy. Loss or gain of visual acuity was evaluated across the 20/50 or worse (visual impairment) and the 20/200 or worse (legal blindness) thresholds according to the Standardization of Uveitis Nomenclature Working Group (7). A complete response consisted of the complete resolution of inflammatory ocular lesions and the absence of relapse during corticosteroid tapering. A partial response was considered in cases of a significant but incomplete decrease of ocular inflammation. If no improvement occurred within 2 months, we concluded that the treatment was a failure (i.e., nonresponse), and a new immunosuppressive strategy was proposed. Likewise, if severe inflammatory relapse occurred during azathioprine treatment or if steroids could not be tapered, we concluded that azathioprine therapy was a failure, and a new immunosuppressive strategy was proposed.
Data are summarized as frequencies and percentages for categorical variables. Quantitative variables are presented as the median (interquartile range) or the mean ± SD. Factors associated with complete ocular remission were assessed using the Fisher's exact test or logistic regression model. Odds ratios (ORs) with their 95% confidence intervals (95% CIs) are shown as a measure of association. Log linearity of the visual acuity effect was graphically assessed using restricted cubic splines. A categorized variable was constructed using cutoffs obtained by restricted cubic splines, and clinically relevant log linearity of the new categorized variable was assessed using Wald's test.
Uveitis relapse rates with their 95% CIs were estimated by using the Kaplan-Meier method. Cox proportional hazards regression model was used to estimate hazard ratios with their 95% CIs. All tests were 2-sided, with a significance level of 0.05. Analyses were performed using the R statistical package (online at http://www.R-project.org).
Characteristics of BD.
The main features of the 157 BD patients are shown in Table 1. The mean ± SD age at diagnosis was 29.9 ± 10.1 years, with 71% men. Most BD patients come from Europe (49%) and North Africa (44.6%). Clinical signs of BD included recurrent oral (100%) and genital (52.9%) ulcerations, skin lesions (61.8%; mainly pseudofolliculitis), articular involvement (51.3%), central nervous system involvement (33.3%), venous involvement (24.8%), and arterial involvement (8.3%). HLA–B5 was detected in 60.8% of BD patients. This was higher than the prevalence for HLA–B5 in our cohort of BD patients (259 [39.7%] of 652).
|Age at diagnosis, mean ± SD years||29.9 ± 10.1|
|Male sex||112 (71.3)|
|North Africa||70 (44.6)|
|HLA–B5 (n = 120)||73 (60.8)|
|Characteristics of uveitis|
|Retinal vasculitis||54 (34.4)|
|Clinical features of BD|
|Oral ulcerations||157 (100)|
|Genital ulcerations||83 (52.9)|
|Skin lesions||97 (61.8)|
|Arterial involvement||13 (8.3)|
|Venous thrombosis||39 (24.8)|
|CNS involvement||52 (33.3)|
All of the patients had active posterior uveitis or panuveitis. Ocular disease was bilateral in 132 (84.1%) of 157 patients, with panuveitis in 42% and retinal vasculitis in 34.4% of the patients. Mean visual acuity before initiation of therapy was 4.2/10 (mean ± SD 4.5 ± 0.3 in the right eye and 4.1 ± 0.3 in the left eye), with a loss of useful vision at baseline in 59 patients (37.6%). Other immunosuppressants were used before azathioprine initiation in 31 cases (19.7%) and included cyclophosphamide (n = 18), cyclosporine (n = 6), chlorambucil (n = 4), and IFN alfa-2a (n = 3).
Efficacy and tolerance.
Based on clinical and imaging findings and on corticosteroid tapering, we concluded that 146 patients (92.9%) had an improvement (i.e., partial or complete response) of ocular lesions with azathioprine (Table 2). After a mean ± SD followup of 71.5 ± 68.6 months, 81 (51.6%) were complete responders, 65 (41.4%) were partial responders, and 11 (7%) were nonresponders. Visual acuity in the better eye progressed from 4.49 to 6.8/10 (P < 0.0001) after azathioprine therapy. Visual acuity in the worse eye progressed from 4.18 to 6.45/10 (P < 0.0001) after azathioprine therapy. Loss of useful vision was noted in 59 cases (37.6%) at baseline and 31 cases (19.6%) at the end of followup (P < 0.01) (Table 2). The mean ± SD oral prednisone threshold decreased significantly from 55.3 ± 13.8 mg/day (range 25–80) to 10.5 ± 6.5 mg/day (range 5–25; P < 0.001). Fourteen patients with persistent ocular inflammation despite high doses of steroids and azathioprine were considered nonresponders. They were switched to cyclophosphamide (n = 6), IFN alfa-2a (n = 5), or infliximab (n = 3) to control inflammation, with complete remission in 12 patients and partial response in the 2 remaining patients. Among the 146 responders, the cumulative relapse rates of uveitis were 11% at 1 year (receiving azathioprine therapy) and 32.6% at 5 years (after azathioprine discontinuation) (Figure 1). The median duration of azathioprine therapy was 3.4 years (range 1–5 years). Side effects of azathioprine were noted in 67 patients (42.6%) and mainly included gastrointestinal events (19.1%), cytopenia (18.4%), and infections (17.8%). There were 3 withdrawals due to toxicity during azathioprine therapy, 2 for hepatotoxicity and 1 for septicemia.
|Baseline||End of followup|
|Followup, mean ± SD months||71.5 ± 68.6|
|Complete response, no. (%)||81 (51.6)|
|Partial response, no. (%)||65 (41.4)|
|Nonresponse, no. (%)||11 (7)|
|Loss of useful vision, no. (%)||59 (37.6)||31 (19.6)*|
|Daily prednisone dose, mean ± SD mg||55.3 ± 13.8||10.5 ± 6.5†|
|Relapse after azathioprine discontinuation (n = 144), no. (%)||47 (32.6)|
Incidence of vision loss/gain.
The incidence rates for loss of visual acuity among patients with severe posterior uveitis related to BD and treated with azathioprine are shown in Table 3. The incidence rates of loss of visual acuity to 20/50 or worse and to 20/200 or worse among affected eyes were 0.35/eye-year and 0.18/eye-year, respectively, among eyes with visual acuity better than these thresholds at presentation. The incidence of gaining 3 lines of visual acuity during followup was 0.63/eye-year.
|Event||Eye*||Eye-years||Rate/eye-year (95% confidence interval)|
|Visual acuity change|
|To 20/50 or worse||57/103||76.24||0.35 (0.25–0.47)|
|To 20/200 or worse||29/53||65.34||0.18 (0.12–0.26)|
|Loss of 3 lines||19/270||61.21||0.16 (0.11–0.28)|
|Gain of 3 lines||50/270||78.44||0.63 (0.51–0.71)|
Predictive factors of complete ocular remission.
Comparison between complete responders of uveitis (n = 81) and partial/nonresponders (n = 76) is detailed in Table 4. No significant differences were noted between the complete responders and the partial/nonresponders in terms of age, sex, HLA–B5 status, previous immunosuppressant use, and extraocular symptoms of BD (i.e., oral and genital ulceration and articular, venous, central nervous system, or arterial involvement). Patients with retinal vasculitis (OR 0.45 [95% CI 0.2–0.9], P = 0.02) and severe visual loss (OR 0.28 [95% CI 0.2–0.7], P < 0.0001) at diagnosis were less likely to be complete responders (Table 4 and Figure 2). Among the complete responders, 43 (53%) of 81 had a visual acuity lower than 20/50 at diagnosis compared with 69 (91%) of 76 partial/nonresponders. Thirty-three partial/nonresponders (43.4%) had retinal vasculitis at diagnosis compared with 21 complete responders (25.9%). The likelihood of a complete response of uveitis was 3.4 in patients with a visual acuity between 1 and 4/10, 8.6 in those with a visual acuity between 5 and 6/10, and 33.1 in those with a visual acuity between 7 and 10/10 (Figure 2).
|All (n = 157)||Complete responders (n = 81)||Partial/nonresponders (n = 76)||OR (95% CI)||P|
|Age, mean ± SD years||29.9 ± 10.1||29.3 ± 9.3||30.7 ± 10.9||0.98 (0.9–1.1)||0.37|
|Male sex||112 (71.3)||56 (69.1)||56 (73.7)||0.80 (0.4–1.7)||0.59|
|HLA–B5||72 (60)†||36 (60)||36 (60)||0.95 (0.4–2.1)||1|
|Previous immunosuppressant||31 (19.7)||14 (17.3)||17 (22.4)||0.71 (0.5–3.1)||0.55|
|Retinal vasculitis||54 (34.4)||21 (25.9)||33 (43.4)||0.45 (0.2–0.9)||0.02|
|Bilateral uveitis||132 (84.1)||66 (81.5)||66 (86.8)||0.66 (0.2–1.7)||0.39|
|Visual acuity, mean ± SD||4.3 ± 3.6||5.4 ± 3.2||3.1 ± 3.4||0.28 (0.2–0.7)||< 0.0001|
|Oral ulceration||157 (100)||81 (100)||76 (100)||–||–|
|Genital ulceration||83 (52.9)||40 (49.4)||43 (56.6)||0.75 (0.4–1.5)||0.42|
|Articular involvement||80 (51.3)||42 (51.9)||38 (50)||1.04 (0.5–2.1)||1|
|Venous involvement||39 (24.8)||23 (28.4)||16 (21.1)||1.48 (0.6–3.3)||0.35|
|Arterial involvement||13 (8.3)||10 (12.5)||3 (3.9)||1.47 (1.2–1.7)||0.08|
|CNS involvement||52 (33.3)||32 (39.5)||20 (26.7)||1.78 (0.8–3.7)||0.12|
In the present study, we report on the long-term efficacy and tolerance of 157 patients with BD treated with steroids and azathioprine for severe uveitis. We showed that following this regimen, posterior uveitis was controlled in 92.9% of patients, of whom 51.6% were complete responders and 41.4% were partial responders. We observed a significant improvement of visual acuity under azathioprine, especially in the complete responders. The proportion of patients with a loss of useful vision dropped from 37.6% to 19.7% after therapy. Visual prognosis has improved in recent years with the increasing use of immunosuppressive agents (9–11). Nevertheless, in few cases, uveitis remains refractory to conventional therapy and despite these aggressive strategies, blindness occurs in 16–25% of patients after 5 years (3, 12). Azathioprine has been proven effective in BD in a controlled study, showing a decrease in hypopyon uveitis attacks, preservation of visual acuity, and decreased development of new eye disease (13). However, this study was not designed to address the ocular outcome, and the number of patients with eye disease (n = 25) who received azathioprine was rather low (13). Azathioprine was also effective in decreasing the development of new genital ulcers, arthritis, and thrombophlebitis (13). The 7 years of followup of these BD patients showed less blindness in the original azathioprine group and less development of new eye disease (14). Moreover, the beneficial effect of azathioprine was especially pronounced among patients who had recent eye involvement (13, 14). Pacheco et al also observed, in 27 patients with autoimmune uveitis (of whom 3 had BD) treated with azathioprine, a complete response in 92%, with only minor side effects (15). CD8bright/CD56+ T cells were shown to function as strong cytotoxic effectors upon ocular infiltration, resulting in visual loss of BD uveitis (16). These cells are polarized to produce a large amount of IFNγ upon stimulation and exert cytolytic functions through both FasL- and perforin-dependent pathways (16).
In our study, azathioprine had a corticosteroid-sparing effect, allowing a significant diminution of the daily prednisone dose. This sparing effect is likely to be clinically meaningful because potential side effects of corticosteroids, including cataracts and glaucoma, cause concern. Some authors advocated a rapid corticosteroid tapering after initiation of immunosuppressants (i.e., IFN alfa-2a), starting during the first month (11). Corticosteroid tapering varies widely among reported studies. Our approach was based on the potential severity of relapsing uveitis and, in our study, corticosteroid tapering was initiated 1 month after initiation (i.e., based on clinical improvement). The cumulative relapse rate of uveitis was 11% at 1 year and increased to 32% after azathioprine discontinuation (i.e., at 5 years). This was approximately the same relapse rate reported in BD uveitis after discontinuation of IFN alfa-2a (10), but lower than that reported with a tumor necrosis factor α antagonist (i.e., infliximab) (17).
Tolerance of azathioprine was excellent, with only 3 discontinuations for hepatotoxicity or septicemia. There was no withdrawal due to toxicity during the controlled study with azathioprine (13). It was observed in an open study that the concomitant use of azathioprine with IFN alfa-2a could cause leukopenia (18).
This large cohort population of BD patients with severe uveitis enables us to assess predictive factors of complete ocular remission. BD patients with severe visual loss (i.e., visual acuity <20/50) and retinal vasculitis at diagnosis were 3.5 and 2 times less likely to be complete clinical responders, respectively. Among the complete responders, 53% had severe visual loss at diagnosis compared with 91% of the partial/nonresponders. The likelihood of complete response of uveitis was 10 times higher in BD patients with good visual acuity (i.e., visual acuity >20/50) at diagnosis compared with those with severe visual loss. Only 25% of the complete responders presented with retinal vasculitis at diagnosis compared with 44% of the partial/nonresponders. Other immunosuppressive agents proved effective in Behçet's uveitis and may represent interesting alternative therapeutic options. Recent studies with a tumor necrosis factor α antagonist emphasized that infliximab may be a rapidly acting agent in patients with relapsing posterior uveitis inadequately controlled with other immunosuppressants (19–21). However, repeated infusions are needed due to its transient efficacy (17). IFN alfa-2a is an interesting alternative (22), but side effects are frequent, with flu-like symptoms reported in more than 70% of patients (23). In conclusion, azathioprine represents an effective and well-tolerated therapy in patients with BD and severe uveitis. Azathioprine should be recommended in BD patients without retinal vasculitis or severe vision loss.
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Saadoun had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design. Saadoun, Wechsler, Le Thi Huong, Resche-Rigon, Le Hoang, Amoura, Bodaghi, Cacoub.
Acquisition of data. Saadoun, Wechsler, Terrada, Hajage, Le Thi Huong, Resche-Rigon, Cassoux, Le Hoang, Amoura, Bodaghi.
Analysis and interpretation of data. Saadoun, Wechsler, Terrada, Hajage, Resche-Rigon, Le Hoang, Bodaghi, Cacoub.
- 6International Study Group for Behçet's Disease. Criteria for diagnosis of Behçet's disease. Lancet 1990; 335: 1078–80.
- 9Efficacy of infliximab in the treatment of uveitis that is resistant to treatment with the combination of azathioprine, cyclosporine, and corticosteroids in Behçet's disease: an open-label trial. Arthritis Rheum 2005; 52: 2478–84., , , , , , et al.