In this issue of EJC, Perez-Gracia and colleagues question the need for confirmation of response reported in cancer clinical trials. To illustrate this problem, the authors have selected several phase II trials testing the activity of a new anticancer agent in different tumour types. The authors have compared the response rates obtained after the first response evaluation with those obtained after confirmation. The correlation between response rates (assessed with Cohen’s Kappa coefficient) was excellent for the pooled analysis of the trials and also within each study.
Therasse, P., & Eisenhauer, E. (2005). Confirmation of response in cancer clinical trials: A meaningless exercise? European Journal of Cancer, 41(11), 1501-1502. https://doi.org/10.1016/j.ejca.2005.04.021 (Original work published 2005)