Molecular and clinical predictors of outcome for cetuximab in non-small cell lung cancer (NSCLC): Data from the FLEX study

Abstract

8007 Background: The multinational, randomized, phase III FLEX study compared cisplatin/vinorelbine (CT) plus the EGFR- antibody, cetuximab (Erbitux), with CT alone in the 1st-line treatment of patients (pts) with advanced EGFR-expressing NSCLC and demonstrated a statistically significant OS benefit for the cetuximab combination. We hypothesized that KRAS mutation status is predictive for cetuximab efficacy and enables optimal use of cetuximab. The relationship between early-onset acne-like rash (ie rash that developed ≤21 days of treatment initiation) and OS time of pts treated with CT and cetuximab was also evaluated. Methods: Archived tumor samples from 554/1125 pts were available. Genomic DNA derived from formalin-fixed paraffin embedded tumor tissue was analyzed for KRAS using an LNA-mediated qPCR clamping assay capable of detecting oncogenic mutations at codons 12 and 13. The Kaplan-Meier method was used to estimate OS time and PFS time in pts with KRAS wild-type (wt) and mutant (mt) tumors for each treatment arm. All pts treated with cisplatin/vinorelbine plus cetuximab who were alive at 21 days were included in a landmark analysis evaluating the relationship between early-onset acne-like rash and OS time. Results: KRAS results were obtained from 379 pts. A KRAS mutation was detected in 72 (19%) pts. The comparison of the cetuximab treatment effects in pts with KRAS wt tumors and pts with KRAS mt tumors showed no marked differences with regard to OS or PFS. A total of 518 pts were included in the landmark analysis. Pts treated with cetuximab who developed early acne-like rash of any grade (grade 1–3; 56%, n=290) had a longer median OS than those without acne-like rash (n=228) (median [95% CI]: 15.0 months [12.8–16.4] vs 8.8 months [7.6–11.1]; HR [95% CI]: 0.63 [0.52–0.77]; p<0.001). Analysis of EGFR FISH is ongoing and results will be presented. Conclusions: Clinical data from the FLEX study do not support the hypothesis that KRAS mutation status is predictive for cetuximab efficacy when combined with 1st- line chemotherapy in advanced NSCLC, whereas early acne-like rash of any grade appears to be associated with better outcome in pts treated with platinum-based chemotherapy plus cetuximab in this setting. [Table: see text]