OBJECTIVE: To evaluate drug adherence, clinical response and predictors thereof for tocilizumab in patients with RA in routine care based on prospectively collected data from the Swedish biologics register, Anti-Rheumatic Therapies in Sweden.
METHODS: RA patients who had started with tocilizumab from September 2008 until March 2012 were identified. Cox regression and logistic regression models were used.
RESULTS: A total of 530 RA patients were included, of whom 80.6% were female, 64.7% were on concomitant DMARDs, of which 300 were on MTX and 12% were biologic naive. The overall 6 month, 1 and 2 year estimated drug continuations were 79%, 64% and 50%, respectively. In the multivariate analyses, a low initial level of CRP [hazard ratio (HR) 0.76/1s.d. (95% CI 0.63, 0.91)], high HAQ score [HR 1.23/1s.d. (95% CI 1.06, 1.44)] and prior exposure to different biologics [HR 1.43 (95% CI 1.12, 1.83)] were predictors for drug termination, whereas concomitant DMARD therapy was not. European League Against Rheumatism (EULAR) good, moderate, and no response were achieved by 184 (46.7%), 133 (33.8%) and 77 (19.5%) patients, respectively. Predictors for EULAR good response vs no response (at 2.5-8 months) were low HAQ [odds ratio (OR) 0.56/1s.d. (95% CI 0.40, 0.78)], high 28-joint DAS [OR 2.0/1s.d. (95% CI 1.44, 2.78)] and not being on prednisolone [OR 0.47 (95% CI 0.25, 0.88)] at baseline.
CONCLUSION: In this RA cohort treated with tocilizumab, the estimated 1 year drug continuation was 64% and 80% of the patients achieved a EULAR response. Drug discontinuation was not predicted by no concomitant DMARD, but by low CRP, high HAQ and prior exposure to biologics.