Pooled analysis of progression-free survival and overall survival in phase II clinical trials of breast cancer

Abstract

Metastatic breast cancer requires more investigation because of its high incidence. Efforts to improve the accessibility, evaluative simplicity, and predictive accuracy of clinical trial endpoints are ongoing. In this paper, we analyzed 410 Phase II clinical trials’ endpoints to assess the efficacy and impact of anticancer agents. We estimated the treatment effects for overall survival and progression-free survival using appropriate statistical methodologies. We found a significant moderate positive correlation between OS and PFS (r = 0.63, 95% CI 0.55–0.71, P < 0.001). Moreover, PFS predicted the OS with a 40% accuracy (R-sq= 0.40). Interestingly, PFS showed significant differences in 1-2-and-3-agent trials, however, this was not reflected in the OS for 2-and-3-agent trials, indicating the need for further validation of the surrogacy of PFS. Nonetheless, PFS is somewhat reliable surrogate of OS in phase II trials of breast cancer and we recommend finding a better surrogate than PFS.