Individual therapeutic DAS28-d_crit responses differentiate between effectiveness of rheumatoid arthritis therapies and reflect patient-reported outcomes: retrospective analysis of DAS28 responses in comparative tocilizumab studies

Assessment of individual therapeutic responses provides valuable information concerning treatment benefits in individual patients. We evaluated individual therapeutic responses as determined by the Disease Activity Score-28 joints critical difference for improvement (DAS28-d_crit) in rheumatoid arthritis (RA) patients treated with intravenous tocilizumab or comparator anti-tumor necrosis factor (TNF) agents. The previously published DAS28-d_crit value [DAS28 decrease (improvement) ≥ 1.8] was retrospectively applied to data from two studies of tocilizumab in RA, the 52-week ACT-iON observational study and the 24-week ADACTA randomized study. Data were compared within (not between) studies. DAS28 was calculated with erythrocyte sedimentation rate as the inflammatory marker. Stability of DAS28-d_crit responses and European League Against Rheumatism (EULAR) good responses was determined by evaluating repeated responses at subsequent timepoints. A logistic regression model was used to calculate p values for differences in response rates between active agents. Patient-reported outcomes (PROs; pain, global health, function, and fatigue) in DAS28-d_crit responder versus non-responder groups were compared with an ANCOVA model. DAS28-d_crit individual response rates were 78.2% in tocilizumab-treated patients and 58.2% in anti-TNF-treated patients at week 52 in the ACT-ion study (p = 0.0001) and 90.1% versus 59.1% at week 24 in the ADACTA study (p < 0.0001). DAS28-d_crit responses showed greater stability over time (up to 52 weeks) than EULAR good responses. For both active treatments, DAS28-d_crit responses were associated with statistically significant improvements in mean PRO values compared with non-responders. The DAS28-d_crit response criterion provides robust assessments of individual responses to RA therapy and may be useful for discriminating between active agents in clinical studies and guiding treat-to-target decisions in daily practice.