Predictors of drug retention and treatment response in axial spondyloarthritis patients treated with certolizumab: real-life results from the HURBIO registry

Objectives To assess the real-life retention rate of certolizumab and factors related to retention of certolizumab. Methods We analysed all patients who received at least 1 dose of certolizumab and were registered in the HURBIO database. Patients with at least 1 control visit were included in efficacy analysis. Drug retention rates were calculated using the Kaplan-Meier method and predictors of drug retention was determined by Cox proportional hazard model. Factors predicting BASDAI50 response at first visit were analysed by the logistic regression analysis. Reasons of switching and discontinuation were also determined. Results A total of 325 (AS (76%), female 55%) patients were recruited. Median follow-up while receiving certolizumab was 13 (4.7-22.7) months. At 1 year, overall certolizumab retention rate was 72.5%. Predictors of poor certolizumab retention were: Current or ex-smoker [HR 1.11 (0.70-1.76), p=0.65], high CRP levels [HR 0.72 (0.45-1.16), p=0.18], biologic-naive [HR 0.81 (0.49-1.32), p=0.39] and good BASDAI50 response at first control visit [HR 0.54 (0.30-0.96), p=0.04]. Mean duration from starting certolizumab to the first control visit was 3 (3-6) months. Predictors of poor BASDAI50 response: Presence of nr-axSpa [RR 2.12 (1.01-4.51), p=0.05], female gender [RR 2.14 (1.20-3.82), p=0.01] and history of biologic therapy [RR 3.52 (1.95-6.33), p<0.001]. The most common causes of drug switch were primary failure and drug side-effects. Conclusions In this study, good BASDAI50 response at first visit seems to be a strong predictor of higher retention of certolizumab in patients with axial spondyloarthritis.