Background/Purpose: To describe in a cohort of rheumatoid arthritis (RA) patients followed up to 12 years, in standard clinical practice, the rate of biological agents (BAs) discontinuation due to adverse drug reactions (ADRs). Also, to analyze which factors influence the rate of severe ADRs (defined as those that required hospital admission or resulted in patient death), paying special attention on the comparison between BAs.

Methods: Observational retrospective longitudinal study, that included RA patients attending the rheumatology outpatient clinic of our center, whom started treatment with any BA [including etanercept (ETN), golimumab, certolizumab, infliximab (IFX), adalimumab (ADA), rituximab (RTX), abatacept (ABA), or tozilizumab], between January 1^st, 2000, and January 1^st, 2012. Clinical records were examined until withdrawal of the drug, loss of follow up, or December 18^th, 2012. Survival techniques were used to estimate the incidence ratio (IR) of BA withdrawal due to ADRs, expressed per 100 patient-years with 95% confidence interval (95%CI). Cox bivariate and multivariate regression models (adjusted by age, gender and calendar time) were used to examine risk factors for BAs discontinuation due to severe.

Results: 405 patients were included, whom began 744 different courses of BA treatments, with a total follow up of 1,612 patient-years. 81% were women, with a mean [standard deviation (SD)]  age at diagnosis of 52.5 (13) years and a mean (SD) elapsed time to the first BA of 5 (4.8) years. The most frequently used BAs were ADA (32%), ETN (25%), IFX (20.7%), and RTX (13.5%). There were 198 discontinuations due to ADRs, mostly due to infections (46%), mucocutaneous (14%), and infusion reactions (10%). 18% of the BAs therapies were discontinued during the first year due to ADRs, with a mean survival time of 5 years (95% CI 4.3-7.6). The IR was estimated in 12.8 (95%CI 10.7-14.1). 77 (39%) were severe ADRs, with an IR of 4.8 (95% CI 3.8-6.0). The most frequent types of severe ADRs were infections (51%) [IFX (55%), ADA (30%), RTX (7.5%) and ETN (2.5%)], cancer (13%) [IFX (30%), ADA (30%), and ETN (30%)] , and congestive heart failure (9%) [IFX (58%), and RTX (42%)]. 50% were lower respiratory tract infections and 11% tuberculosis. 20% were lung and 20% were pancreatic cancers.  6 ADRs resulted in death [IFX (n=3), ADA (n=1), RTX (n=1), ABA (n=1)] with an IR of 0.4 (95% CI 0.2-0.8).

Both in the bivariate and multivariate regression analysis, age (HR: 1.04, 95%CI: 1.02-1.06), maximum rheumatoid factor (HR: 1.0001, 95%CI: 1.00006- 1.0002, p=0.002), and, compared with ETN, treatment with IFX or RTX (HR:5.7, 95%CI: 2.2-14.8, p=0.001; HR:3.3 95% CI: 1.2-9.8, p=0.02, respectively) were significantly associated to a higher risk of BAs discontinuation due to severe ADRs. Treatment with ADA did not achieve statistical signification (HR:2.4, 95%CI: 0.96-6.3, p=0.06).

Conclusion: This study describes the incidence of ADRs (50% related to infections) in RA patients taking BA in real life conditions. 39% resulted in hospital admissions and even in death, thus close monitoring is required in these patients. Within the BA more widely used in our setting, it seems that ETN generates less hospital admissions due to ADRs.

ACR Meeting Abstracts - https://acrabstracts.org/abstract/discontinuation-of-biological-therapy-due-to-adverse-drug-reactions-in-rheumatoid-arthritis-patients-12-years-follow-up/