Phase III trial of induction gemcitabine (G) or paclitaxel (T) plus carboplatin (C) followed by elective T consolidation in advanced ovarian cancer (OC): Final safety and efficacy report.

Abstract

LBA5008 Background: Safety and efficacy of GC or TC induction followed by elective T consolidation (Tcon) were evaluated. Methods: Patients (pts) with stage IC-IV epithelial OC were randomized to GC: G 1,000 mg/m2 on days 1, 8 plus C AUC=5 on day 1; or TC: T 175 mg/m2 plus C AUC=6 on day 1 for a total of six 21-day cycles. Pts with a complete response (CR) could receive Tcon 135 mg/m2every 28 days for 12 cycles. Non-CR pts received single-agent crossover (CO) therapy (CO-T 175 mg/m2 on day 1 or CO-G 1,000 mg/m2 on days 1, 8) every 21 days until CR or progression of disease (PD). PD or death in 636 pts was required to compare GC and TC with 80% power for progression-free survival (PFS), the primary endpoint. Efficacy results were compared by log-rank test. Results: The trial was stopped in 8/2009 after an ad hoc futility analysis showed low probability of a positive PFS result. Of 919 pts enrolled, 88 pts were excluded (clerical errors); 831 pts were entered; 820 pts had induction; 352 pts had Tcon (GC-Tcon=169, TC-Tcon=183); 155 pts had crossover (CO-T=77, CO-G=78); 313 pts discontinued after induction (GC=165, TC=148). Baseline pt characteristics were balanced across arms. Overall response and adverse events for induction regimens were similar to interim results (Gordon, Clin Ovar Cancer, 2009). For GC and TC, median PFS were 20.0 and 22.2 months; median overall survival (OS) were 43.8 and 57.3 months, respectively. There was no significant difference in PFS (p=0.199) comparing GC and TC. Despite high censorship (GC: 52.8%, TC: 61.4%), OS was greater for TC (p=0.013) compared to GC, but there was no statistical difference after adjusting for significant covariates. For pts with CR, median OS was 65.6 months with Tcon versus 51.4 months without Tcon (p=0.041). Median OS was not reached for TC-Tcon and was 56.1 months for GC-Tcon (p=0.035). For pts not receiving Tcon or receiving either CO-T or CO-G crossover, there was no difference in OS. Conclusions: PFS was similar for GC and TC. Tcon improved OS. However, OS analysis was limited by study design and high censorship. GC does not offer an advantage over standard of care TC for first-line chemotherapy in advanced OC. No significant financial relationships to disclose.