Overall survival (OS) in contemporary randomized clinical trials (RCT) in advanced breast cancer (ABC)

Abstract

1002 Background: OS has been considered an elusive endpoint, while progression-free survival (PFS), time to tumor progression (TTP), and time to treatment failure (TTF) have been used frequently and often interchangeably as primary endpoints (PE) in ABC (Saad, Ann Oncol 2009). Methods: We searched PubMed using the medical subject headings “breast neoplasms” and “drug therapy,” limiting the search to phase III trials on systemic antineoplastic therapies published between 1/2000 and 12/2007 in 8 leading medical journals (Ann Oncol, BCRT, BJC, Cancer, EJC, JCO, Lancet Oncol, and NEJM). PE was the one stated explicitly, used for N calculation, or listed first. Significant PE (SigPE) was considered as P <0.05 for superiority trials (N=47), or proven non-inferiority/equivalence otherwise. Results: We retrieved 58 RTCs, with a median sample size of 329 evaluable patients. The table shows the breakdown of these trials by size and PE, and the number of trials that reached statistical significance on their PE and on OS. Overall, a statistically significant gain in OS was reported in 11 of 58 RCTs (19%). In 30 RCTs with gain in PE, 26 favored the experimental arm. Conclusions: While publication and journal selection biases may not be excluded, our review suggests that gain in OS, although seldom a primary study objective, is not infrequent in contemporary RCTs in ABC. However, the majority of RCTs are underpowered for OS, which remains a secondary or exploratory endpoint in most cases. [Table: see text] No significant financial relationships to disclose.