PURPOSE

We investigated whether tumor response rate (TRR), disease control rate (DCR), or progression-free survival (PFS) was a valid surrogate for overall survival (OS) in phase II trials of second-line therapies for patients with platinum-resistant ovarian carcinoma (PROC).

METHODS

We retrospectively evaluated data from 11 second-line phase II trials conducted for PROC by the Gynecologic Oncology Group (GOG). TRR included complete response and partial response (CR/PR) and DCR was defined as either tumor response or stable disease (CR/PR+SD). Survival by tumor response was analyzed using a landmark approach. Correlations of OS with TRR, DCR, and PFS were estimated.

RESULTS

Among 407 patients analyzed the TRR was 13.8% (56/407) and DCR was 38.8% (158/407). Median OS was 10.2 months while median PFS was only 2.4 months. Median OS among patients with a best response of CR/PR, SD, and progressive disease (PD) was 13.3, 12.1 and 5.7 months, respectively, showing no difference between CR/PR and SD. From a protocol level, DCR correlated better with OS (Pearson r=0.748; Tau-b r=0.514) compared to TRR (Pearson r=0.564; Tau-b r=0.404). PFS rate at 6 months (Pearson r=0.661; Tau-b r=0.514) also correlated strongly with OS.

CONCLUSIONS

This study demonstrates the limitations of the use of response rate alone in PROC. Clinical benefit, as defined by OS, appeared similar for patients with an objective response and those with SD. The DCR, by including tumor response and SD may have utility as a surrogate endpoint for survival in phase II therapeutic trials in PROC.