PURPOSE
To evaluate the effectiveness and reasons for discontinuation including the side effect profiles of adalimumab in a real-world setting.
DESIGN
Retrospective clinical cohort study.
METHODS
A medical chart review of clinical practice in 2 tertiary eye care services in Rotterdam, the Netherlands, was performed Data were collected from May 1, 2004, through September 1, 2020. Patients with noninfectious uveitis treated with adalimumab (n = 341; 633 affected eyes) were included. The primary outcome was the effectiveness of adalimumab, measured by the number of patients achieving inactive disease, remission, and relapse-free survival. The secondary outcomes were the reasons for discontinuation, including side effects, and the number of patients who developed antibodies.
RESULTS
In total, 341 patients were treated with adalimumab between May 2004 and September 2020. The uveitis recurrence-free survival interval was 3.4 years (range, 0-13 years). Adalimumab had an acceptable side effect profile. A total of 178 patients achieved inactive disease while continuing adalimumab, and 51 patients maintained remission after discontinuing adalimumab. Reasons for discontinuation of adalimumab were no response, relapse, or reasons unrelated to the effectiveness of treatment. Adalimumab antibodies were present in 40 of 115 patients (35%). Antibodies were associated with lower adalimumab levels, and antibodies were observed more often in patients on adalimumab monotherapy ( P < .01).
CONCLUSIONS
Adalimumab is effective for patients with noninfectious uveitis, with an acceptable side effect profile. Although relapses can occur, the majority of the patients achieved inactive disease or remission after cessation of adalimumab, without other systemic immunosuppressive medication.
S ystemic and topical glucocorticosteroids have been the first choice of treatment for noninfectious uveitis in the past 50 years. Long-term use of systemic glucocorticosteroids is not advisable, because of severe side effects such as steroid-induced diabetes mellitus, osteoporosis, and cataracts. , To reduce glucocorticosteroid dosage and to prevent side effects, “second-line drugs” such as methotrexate, azathioprine, mycophenolate, and cyclosporine have been used. , , Nowadays, treatment with biologics, such as anti−tumor necrosis factor (anti-TNF) is initiated as third-line therapy, if the inflammation cannot be controlled with second-line therapy or if unacceptable side effects occur.
TNF-α is a pro-inflammatory cytokine that is synthesized by T-helper cells (Th cells), activated macrophages, and neutrophils. It activates the cytokine-driven inflammatory processes. High TNF-α levels are present in the aqueous humor of patients with active uveitis. From the early 2000s, TNF-α blocking therapy, using the chimeric TNF-α monoclonal antibody infliximab, has demonstrated efficacy in the treatment of refractory uveitis. ,
Later on, several studies demonstrated the efficacy of adalimumab (Humira), a humanized recombinant monoclonal anti-TNF antibody, in infliximab-intolerant patients or patients with refractory uveitis. An advantage of adalimumab over infliximab and other chimeric monoclonal antibodies is its lower immunogenicity, as shown in rheumatoid arthritis and Crohn disease because of the humanized recombinant compound. Moreover, superiority of adalimumab over infliximab was reported in a study of chronic childhood uveitis by maintaining remission for more than 3 years. In large case-control studies, adalimumab was well tolerated, decreased inflammatory activity, and enabled the reduction of steroid requirement. Recently, studies with long-term follow-up, including 5 to 6 years of treatment duration, were published. , , However, in these studies, concerning long-term follow-up, the reason for loss of effectiveness, how often antidrug antibodies were formed, and the relapse rate after adalimumab cessation were not addressed. Concerning the drug retention rate, there was a single observation study, showing a drug retention rate of 54% after 60 months. Concerning adalimumab antibody formation, antibodies have been reported to develop in 2.7% to 35% of patients. , Data on relapse after cessation of adalimumab are scarce. In a small cohort study of 28 patients with juvenile idiopathic arthritis, 11 of 12 patients (92%) treated with adalimumab and methotrexate had a relapse within 1 to 2 years after discontinuation of treatment.
The aim of the present study is to evaluate the long-term effectiveness of adalimumab in noninfectious uveitis and to determine the response rate, causes of lack of effectiveness, and treatment failure as well as long-term side effects.
METHODS
DATA COLLECTION
All patients with uveitis treated with adalimumab for noninfectious uveitis who visited the Rotterdam Eye Hospital (Rotterdam, the Netherlands), or the Department of Ophthalmology or Department of Immunology of the Erasmus University Medical Center (Rotterdam, the Netherlands) were eligible for inclusion in this observational clinical cohort study. Patients from the Erasmus University Medical Center were retrieved through a database search, including all patients visiting the Department of Ophthalmology and/or the Department of Immunology, who were prescribed adalimumab at any point in time. Manually, all patients with uveitis were selected from this search. Patients from the Rotterdam Eye Hospital were retrieved through a manual database search. All patients who were prescribed adalimumab were included. Data were collected through retrospective medical chart review. Approval from the Medical Ethic Committee for analysis of patient data was requested and obtained from the Rotterdam Eye Hospital and the Erasmus University Medical Center (MEC-2012-016). Additional informed consent was not deemed necessary, because of the retrospective observational character of this study.
PATIENTS AND DEFINITIONS
Data included both adult patients and pediatric patients; “pediatric” was defined as uveitis onset before the age of 18 years. The anatomic classification of uveitis used in this study was defined by the Standardization of Uveitis Nomenclature Working Group criteria.
In clinical practice, patients are screened for infectious diseases (eg, hepatitis B and C) before initiation of adalimumab therapy. Patients were screened for active tuberculosis by means of a Quantiferon test and/or Mantoux test. After initiation of adalimumab, patients were evaluated for treatment response within 3 months of treatment. Topical steroid therapy, disease-modifying anti-rheumatic drugs (methotrexate, azathioprine, mycofenolate/myfortic, cyclosporin, hydroxychloroquine) (DMARDs), and systemic therapy were continued, the dosage was adjusted, or the medication was discontinued according to the discretion of the treating physician.
Patients started with subcutaneous adalimumab with a dose of least 40 mg biweekly, although some patients started with a dose of 80 mg at baseline and 40 mg after 1 week in case of severe inflammation. If deemed necessary, the dosage and therapy interval were adjusted by the treating physician based on the response to therapy. Immunosuppressive drugs that were added or still used in addition to adalimumab 3 months after the start of adalimumab were included in the study analysis. Dosage of topical steroid drops was not taken into account in this study.
OUTCOME MEASUREMENT
The primary outcome was effectiveness of adalimumab. Effectiveness was measured by the number of patients achieving inactive disease or remission and by the relapse-free survival.
The secondary outcomes were the percentage of patients with inactive disease at 6 months, 1 year, 3 years, 5 years and 10 years, and the number of patients developing antibodies, the number of patients experiencing side effects, and the reasons for discontinuation.
Remission was defined as no sign of inflammation for 3 months after cessation of systemic immunomodulating therapy. Inactive disease was defined as no sign of inflammation for 3 months during treatment with systemic immunomodulating therapy, including systemic corticosteroids. Because one of the treatment goals in clinical practice is tapering of systemic corticosteroids, the number of patients still using corticosteroids after inactive disease was reached was reported. A cut-off of 10 mg or less of prednisone per day was chosen as succesful tapering of systemic corticosteroid therapy, in accordance to the Standardization of Uveitis Nomenclature for Reporting Clinical Data. Relapse of uveitis was defined as active inflammation after inactivity for at least 3 months. Cystoid macular edema was not included in this definition.
The worst eye analysis method was used. In the worst eye method, treatment is considered successful if improvement or quiescence is observed in both eyes of a single patient. A therapeutic switch is considered if 1 eye shows activity or does not improve despite therapy. Activity, inactivity while on therapy, and remission reflect the response of the worst eye, not of the individual eyes (eg, if both eyes respond to therapy, the remission or time to relapse will be calculated for the worst eye, or as a single response and not counted as 2 responses).
In case a patient discontinued therapy, defined as interruption of more than 4 weeks, the cause of discontinuation was reported. Reasons for discontinuation were classified as remission, no primary response, relapse of disease, complications or adverse events, side effects, patient request, and pregnancy or plan for pregnancy. The primary response to therapy was defined as decrease of uveitis activity, as reported by an ophthalmologist. A diagnosis of no response was made in patients who did not exhibit a change in the severity of uveitis activity.
In case of a relapse, the treating physician decided whether adalimumab therapy was continued with additional immunosuppressive therapy or switched to other therapy.
STATISTICAL ANALYSIS
Patient characteristics were summarized using descriptive statistics, including median, range, and percentage. The nonparametric test was used to test for significant differences. Differences in groups of unpaired categorical data (eg, antibody formation, prescribed therapy, or uveitis localization) were analyzed using the Fisher exact test. Survival analysis for uveitis relapse-free survival was performed by plotting a Kaplan−Meier curve. For the survival analysis, an event was defined as the first uveitis relapse during adalimumab treatment. Patients who stopped using adalimumab or patients who reached end of follow-up without a relapse were considered as censored at that time point. Differences in time to first relapse based on uveitis localization, uveitis etiology, and systemic vs solitary ocular involvement were tested using the log-rank (Mantel−Cox) test. Time to second relapse or time after restart of adalimumab was not included in this analysis.
All statistical analyses were performed using SPSS 25.0.0 for Windows (SPSS Inc). A P value of <.05 was considered statistically significant.
RESULTS
PATIENT CHARACTERISTICS
The database search revealed 342 patients, with uveitis in 1 or both eyes, resulting in 633 affected eyes treated with adalimumab. One patient was prescribed adalimumab, but did not start the treatment (due to noncompliance) and was excluded from the analysis ( Table 1 ).
| Characteristic | N (%) |
|---|---|
| Total Number of Patients | 341 (100%) |
| Female Sex | 209 (61%) |
| Bilateral Eye Involvement | 292 (86%) |
| Median Age at Start of Adalimumab Treatment, years | 44 (IQR = 33-56) |
| Anatomical Localization of Uveitis | |
| Anterior | 66 (19%) |
| Intermediate | 22 (7%) |
| Posterior | 92 (27%) |
| Panuveitis | 160 (47%) |
| Not specified | 1 (0.3 %) |
| Diseases Limited to the Eye | |
| Idiopathic | 70 (21%) |
| Birdshot chorioretinopathy | 56 (16%) |
| Postinfectious | 13 (3.8%) |
| Sympathetic ophthalmia | 8 (2.3%) |
| Multifocal choroiditis with panuveitis | 5 (1.5%) |
| APMPPE | 3 (0.9%) |
| Serpiginous/ampiginous choroiditis | 2 (0.6%) |
| IRVAN syndrome | 1 (0.3%) |
| AZOOR | 1 (0.3%) |
| POHS | 1 (0.3%) |
| Systemic diseases | |
| Sarcoidosis, including ocular sarcoidosis (n = 7) | 75 (22%) |
| Behçet disease | 26 (7.6%) |
| Juvenile rheumatoid arthritis | 16 (4.7%) |
| Spondylarthropathy | 16 (4.7%) |
| Psoriasis | 15 (4.4%) |
| Inflammatory bowel disease | 8 (2.3%) |
| Rheumatoid arthritis | 7 (2.1%) |
| Vogt−Koyanagi−Harada disease | 7 (2.1%) |
| HLA-B27−associated uveitis | 6 (1.8%) |
| TINU | 2 (0.6%) |
| Systemic sclerosis | 1 (0.3%) |
| Morbus Sjögren | 1 (0.3%) |
| ROSAH syndrome (ALPK-1 mutation) | 1(0.3%) |
| Medication Before Starting Adalimumab a | |
| None | 22 (6.5%) |
| Topical steroids only | 39 (11%) |
| Steroids intravitreal and/or systemic b Monotherapy/with DMARDS/with anti-TNF/with interferon | 152 (45%) 44 /102/5 /1 |
| DMARDs b Monotherapy/with anti-TNF/with interferon | 107 (31%) 103/3/1 |
| Anti-TNF (etanercept, infliximab) b Monotherapy | 12 (3.5%) |
| Interferon b Monotherapy | 2 (0.6%) |
a In 7 patients, the exact medication prior to the start of adalimumab was unknown.
Uveitis was active in 244 of the 341 patients (72%) when starting adalimumab. Nearly half of the patients (160 of 341; 47%) had inflammation limited to the eye(s) (eg, birdshot retinopathy) or remained idiopathic after screening. The other patients (181 of 341; 53%) had a systemic noninfectious disease. The most prevalent systemic diseases were sarcoidosis (22%) and Behçet disease (8%) ( Table 1 ). Most of these systemic diseases are associated with uveitis, with the exception of systemic sclerosis.
The median age at the start of adalimumab treatment was 44 years (range, 33-56 years). The cohort of 341 patients included 20 pediatric patients (5.9%). There were no differences in sex, bilateral involvement, localization of uveitis, number of relapses, and median time to relapse between pediatric and adult patients. Therefore, data from both adult and pediatric patients were combined.
All patients received immunosuppressive therapies before initiation of adalimumab ( Table 1 ).
TREATMENT OUTCOME
The median follow-up period after initiation of adalimumab was 4.9 years (range, 0.11-14.7 years). The median treatment duration of adalimumab was 3.3 years (range, 0.03-14.7 years). Of the patients, 18% started adalimumab monotherapy with or without topical steroids. All other patients had combination immunosuppressive therapy consisting of adalimumab with steroids and/or DMARDs. During the follow-up, 148 of the 341 patients (43%) continued treatment with adalimumab until end of follow-up (ie, time of data collection).
Inactive disease/remission was achieved in 298 of the 341 patients (87%). Inactive disease, which enabled the cessation of adalimumab, was achieved in 57 patients, with a median treatment duration until discontinuation of 3.2 years (range, 0.4-11 years). In all, 178 patients reached inactive disease and continued adalimumab treatment ( Table 2 ). Adalimumab was effective in 235 patients ([57 + 178]/341 patients; 69%) of patients. In the remaining 63 of the 298 patients, inactive disease was achieved when patients switched to other medication. Of the 103 patients using adalimumab and systemic corticosteroids who reached inactive disease, systemic corticosteroids could be tapered to zero in 86 patients. None of the 103 patients had received intravitreal steroids in the prior 3 months. A minority of patients (33 of 86; 38%) still used systemic corticosteroids such as prednisolone. In total, 31 of 33 patients (94%) used prednisolone 10 mg or less. The remaining 2 patients used 12.5 and 15 mg, respectively. In those patients still using prednisone despite inactive disease, prednisone was not tapered further because of previous visual loss and because the risk of a relapse was considered too high.
