Factors associated with cause-specific discontinuation of long-term anti-tumor necrosis factor agent use in patients with ankylosing spondylitis: a retrospective cohort study

Abstract

Object: To investigate the factors associated with cause-specific discontinuation of long-term anti-tumor necrosis factor (TNF) agent use in patients with ankylosing spondylitis (AS). Methods: AS patients who initiated first-line anti-TNF treatment between 2004 and 2018 and continued treatment for at least two years were enrolled in the study. Enrolled patients were observed until the last visit, discontinuation of treatment, or September 2022. Reasons for discontinuation of the first-line anti-TNF agent were categorized into the following: (1) clinical remission, (2) loss of efficacy, (3) adverse events, and (4) other reasons including loss to follow-up, cost, or reimbursement issues. A cumulative incidence function curve was used to visualize the cumulative failure rates over time for each specific reason. Univariable and multivariable cause-specific hazard models were utilized to identify factors associated with cause-specific discontinuation of the first-line anti-TNF agent. Results: A total of 429 AS patients was included in the study, with 121 treated with adalimumab (ADA), 176 with etanercept (ETN), 89 with infliximab (INF), and 43 with golimumab (GLM). The median overall survival on the first-line anti-TNF agent was 10.6 (7.9-14.5) years. Among the patients, 103 (24.0%) discontinued treatment, with 36 (34.9%) due to inefficacy, 31 (30.1%) due to clinical remission, 15 (14.6%) due to adverse events, and 21 (20.4%) due to other reasons. Patients treated with ETN had a lower risk of discontinuation due to clinical remission compared to those receiving ADA (hazard ratio [HR] 0.45 [0.21-0.99], P = 0.048). Higher baseline Bath Ankylosing Spondylitis Disease Activity Index (BASDAI; HR 1.31 [1.04-1.65], P = 0.023) and INF use were linked to a higher risk of treatment discontinuation for inefficacy compared to ADA use (HR 4.53 [1.45-14.16], P = 0.009). Older age was related to an increased risk of discontinuation due to infection-related adverse events (HR 1.07 [1.02-1.12], P = 0.005), and current smoking was a risk factor for discontinuation due to other reasons (HR 6.22 [1.82-21.28], P = 0.004). Conclusion: AS patients on their first anti-TNF treatment for at least two years demonstrated a favorable long-term treatment retention rate, with a 24.0% discontinuation rate over a 10.6-year overall survival period. The predictors for discontinuation varied by causes, underscoring the complexity of treatment response and the importance of personalized approaches to treatment management.