KRAS and EGFR mutations in the molecular epidemiology of NSCLC: Interim analysis of S0424.

Abstract

7013 Background: The Intergroup molecular epidemiology study S0424 was designed to investigate relationships between tumor biology with tobacco carcinogen exposure and gender in a cohort of early stage NSCLC patients enriched for never-smokers. Each patient completed a questionnaire on smoking history and submitted tumor and blood specimens. Here we report associations between KRAS and EGFR mutational status with histology, gender, and smoking status. Methods: KRAS mutations were assessed in 549 path-confirmed tumor tissue using a cold-PCR single-strand conformational polymorphism (SSCP) approach established to distinguish at least 10 different base substitutions in codons 12 and 13. DNA sequencing was used for confirmation. Preliminary EGFR mutation analysis was conducted on 195 patients using similar methodology. Researchers are blinded to patient-level data. Results: KRAS mutations were identified in 106/549 specimens (19.3%), and were more frequent in smokers (101/495, 20%) compared to never-smokers (5/54, 9%)(p=0.047). Among female smokers, 22% (60 of 274) had KRAS mutations, nonsignificantly higher than the 16.7% rate observed in male smokers (46/275, p=0.13). KRAS mutations were more frequent in adenocarcinomas compared to other histologies (p<0.0001). Logistic regression showed that histology and smoking status, but not gender, were independent variables for KRAS mutations. 47% of KRAS mutations were codon 12 GGT→TGT (Gly→Cys), and 72% were tobacco carcinogenesis-associated G→T transversions. G→T transversions were more frequent in smokers compared to never-smokers (p=0.03). EGFR mutations were observed in 12.8% (25/195; E19del: 9; L858R: 16) and were significantly more common in never-smokers (45%) than in smokers (9%)(p<0.0001) and in patients with adenocarcinoma (p=0.04). EGFR and KRAS mutations were exclusive with the exception of 1 patient. Conclusions: S0424 is the largest prospectively designed study of gender and smoking in NSCLC to date. Smoking history and adenocarcinoma histology, but not gender, were independent variables for KRAS and EGFR mutations. Additional studies exploring tobacco-associated DNA adducts and estrogen receptor expression are ongoing. No significant financial relationships to disclose.