PALETTE: Final overall survival (OS) data and predictive factors for OS of EORTC 62072/GSK VEG110727, a randomized double-blind phase III trial of pazopanib versus placebo in advanced soft tissue sarcoma (STS) patients.

Abstract

10009 Background: At ASCO 2011 we presented an increase in median progression-free survival (PFS) of pazopanib of 13 weeks (median 7 to 20 weeks; HR 0.31) in a randomized double-blind, placebo-controlled phase III trial in advanced non-adipocytic STS patients pretreated with chemotherapy. Now we present final OS data and predictive factors for OS, such as ethnicity, ancestry and well-known prognostic and predictive factors in STS patients treated with chemotherapy. Methods: At clinical cut-off median follow-up was 25 months. Comparison of both treatments for OS was performed using a two-sided log rank test stratified for number of prior lines of systemic therapy for advanced disease and WHO performance status (PS). HR was calculated with a 2-sided CI. A Kaplan-Meier OS curve was used. Baseline characteristics as potential predictive factors for OS were: PS (0 vs 1), number of lines of prior therapy for advanced disease (0-1 vs 2+), age (≤55 yrs vs > 55 yrs), gender, ethnicity (Hispano or Latino vs other), geographic ancestry (White Caucasian, Asian, other), tumour grade (1-2 vs 3), histology (leiomyosarcoma vs synovial sarcoma vs other). A Cox model was fitted with the investigated factor, treatment arm and associated interaction factor. Data of all 369 randomized patients (123 placebo, 246 pazopanib) were used for the analyses. Results: Median OS (95% CI) was 10.7 (8.7-12.8) in the placebo versus 12.5 (10.6-14.8) months in the pazopanib group, HR: 0.86 (0.67-1.1). Post study protocol systemic treatment was administered in 63% of placebo, and in 49% of pazopanib treated patients. None of the investigated factors showed any statistical significant predictive value for OS. Conclusions: Although pazopanib demonstrated a statically significant increase in PFS compared to placebo, final OS analysis, which showed a trend in favor of pazopanib in patients with metastatic non-adipocytic STS, did not reach statistical significance. The factors studied were not predictive for OS.