Molecular analysis-directed, international, phase III trial in patients with advanced non-small-cell lung cancer.

Abstract

8001 Background: ERCC1 (E1) and RRM1 (R1) are predictive markers for platinum agents and gemcitabine (G) respectively. In a phase II trial that utilized E1 and R1 mRNA expression levels for therapeutic decision-making, a response rate of 44%, PFS of 6.6 m, and OS of 13.3 m was achieved. Methods: Pts with advanced, chemo-naive NSCLC, PS 0-1, measurable disease, and a FFPE histological or cell block tumor specimen were eligible. E1 and R1 were analyzed by AQUA and categorized as high or low. Pts were randomized 2:1 to arm A, which received G and carboplatin (Cb) for R1/E1 low, docetaxel (D) and Cb for R1 high and E1 low, G and D for R1 low and E1 high, and D and vinorelbine (V) for R1/E1 high, or arm B, which received GC. Pts received up to 6 cycles. Efficacy was assessed every 6-8 weeks until 1 year from treatment initiation. The primary goal was to improve the 6-m PFS from 38% in arm B (median PFS 4.3 m) to 50% in arm A (median PFS 6.0 m). Secondary goals were improvements in OS (8.0 to 12.0 m) and RR (25% to 50%). Results: Of 275 eligible pts, 183 randomized to arm A. 56 were assigned to GCb, 26 to DCb, 37 to GD, and 64 to DV, which was not significantly different (p=.2) from the expected assignment (30%, 20%, 20%, 30%). In arm B, all 92 pts received GCb. Protein analysis was successful in 91% of pts. The median time from consent to completed gene analysis was 11 days. A tumor rebiopsy for the specific purpose of gene analysis was required in 17% of pts. The 6-m PFS rate and median PFS were 52.0% and 6.1 m in arm A and 56.5% and 6.9 m in arm B (p = 0.18). PFS was not significantly different among the treatment groups in arm A (p = 0.1). A significant survival advantage was found for pts with low E1 and low R1 in arm B compared to the same group in arm A (p = 0.02) although both received GCb therapy. OS and RR were not significantly different between both groups. A comparison between protein and mRNA levels for both genes revealed no significant correlation. Conclusions: This demonstrates that gene expression analysis for therapeutic decision making is feasible in newly diagnosed advanced-stage NSCLC pts. A tumor rebiopsy is safe, required in 17%, and acceptable to 89% pts. The survival results are false negative based on the internal control included in the trial. Clinical trial information: NCT00499109.