POS0861 EFFECTIVENESS AND SAFETY OF TOCILIZUMAB IN PATIENTS WITH SYSTEMIC SCLEROSIS: A PROPENSITY SCORE CONTROL MATCHED OBSERVATIONAL STUDY OF THE EUSTAR COHORT

Abstract

Background:Tocilizumab (TCZ) showed trends for improving skin fibrosis and prevented progression of lung fibrosis in patients with systemic sclerosis (SSc) in placebo-controlled randomised clinical trials (RCTs). However, safety and effectiveness of TCZ beyond these selected and enriched clinical trial populations in SSc is still unknown.Objectives:To assess safety and effectiveness of TCZ treatment compared to standard of care in SSc patients from the large, multicentre, observational, real-life EUSTAR network/database using propensity score matching.Methods:SSc patients from the EUSTAR network/database, who fulfilled the ACR/EULAR 2013 classification criteria, with a baseline and a follow-up visit at 12±3 months, receiving TCZ or standard of care (controls), were selected. The following variables were used for the propensity score matching (1:1): age at diagnosis, gender, disease subtype, baseline modified Rodnan skin score (mRSS), forced vital capacity (FVC), and diffusing capacity for carbon monoxide (DLCO), co-therapy with immunosuppressives, disease duration, and year of treatment. Primary endpoints were mRSS and FVC at 12±3 months follow-up compared between the groups, using paired t-tests. Secondary endpoints were the percentage of progressive/regressive patients for skin and lung at 12±3 months follow-up according to standard definitions (1,2). Sensitivity analyses assessed pre-processing decisions (selection of most recent vs. random observation for control patients with multiple suitable time intervals), as well as the matching method (optimal vs. exact matching). Missing values were addressed with 100-fold multiple imputation using chained equations. Safety data were analysed in all patients. The study including the statistical analysis plan was pre-registered at www.drks.de (DRKS-ID: DRKS00015537).Results:We identified 93 SSc patients treated with TCZ and 2370 SSc patients with standard of care who fulfilled the inclusion criteria. Forty nine (57.7%) of the TCZ treated patients were diffuse, eight patients were not classified, disease duration was (mean±SD) 6.35±5.40 years, their baseline mRSS was 15.05±10.85, and 76 (81.7%) received immunosuppressive therapy in addition to TCZ.Through multiple imputation and propensity score matching, 100 imputed sets of 93 pairs of TCZ/controls were generated. Comparison between groups showed consistent effects of TCZ across all pre-defined primary and secondary endpoints: mRSS was lower in the TCZ group (mean difference (95% confidence interval (CI)) -1.8 (-4.79 to 1.19), p=0.24, Figure 1A). Similarly, FVC % predicted was higher in the TCZ group mean difference (2.25, 95% CI -4.57 to 9.06), p=0.51, Figure 1B). Considering secondary endpoints, the percentage of skin progressors as well as lung progressors at follow up was lower in the TCZ group (odds ratio OR 0.67 (95% CI 0.07 to 6.41), p=0.74 and OR 0.53 (95% CI 0.16 to 1.7); p=0.2, respectively. Consistently, the percentage of regressors for skin (OR 1.6 (95% CI 0.56 to 4.54), p=0.38) and for lung (OR 1.74 (95% CI 0.66 to 4.58), p=0.26) was higher in TCZ. These results were robust regarding the sensitivity analyses. Safety analysis confirmed previously reported adverse event profiles.Conclusion:In this large, observational, controlled, real-life EUSTAR study, effectiveness of TCZ did not reach statistical significance compared to standard of care treatment but showed consistent positive effects of TCZ on skin and lung fibrosis across all pre-defined primary and secondary endpoints confirming data from recent RCTs.