Since their development in 1995, the American College of Rheumatology (ACR) responses have been used as primary and secondary endpoints in almost all important drug approval trials in rheumatoid arthritis [1]. However, which response (ACR20, 50, or 70) would represent the optimal balance between statistical power and clinical relevance in the therapeutic landscape of modern trials of biologics and small molecules was unclear for many years. This current work therefore systematically compares the three ACR responses at different timepoints and in different patient populations in randomized controlled approval trials in RA, based on the assumptions that a) the abovementioned balance between statistical power and clinical relevance differs at different timepoints and in different patient groups, that b) a statistically significant outcome marker is essential for limiting the extent and duration of placebo exposure in clinical trials, and that c) a clinically relevant marker should be preferred if the statistical power is the same. To summarize, the current work validates use of the ACR20 threshold as the primary endpoint, particularly at early timepoints in clinical trials, as its discriminative capacity is significantly higher than that of ACR50 or 70. At later timepoints (e.g. around week 24), the ACR70 response can also be applied with comparable statistical power. This longitudinal comparison of the ACR metrics over the entire duration of placebo-controlled trials provides new insights into the dynamics and relevance of these parameters so frequently used in RA trials, and is therefore relevant for planning and performance of clinical trials as well as for the interpretation of their results.
