Abstract
BACKGROUND: Circulating cell-free DNA (cfDNA) is emerging as a surrogate sample type for mutation analyses. To improve the clinical utility of cfDNA, we developed a sensitive peptide nucleic acid (PNA)-based method for analyzing EGFR and KRAS mutations in the plasma cfDNA of patients with advanced non-small cell lung cancer (NSCLC). METHODS: Baseline tissue and plasma samples were collected from treatment-naïve advanced NSCLC patients participated in a randomized phase II study, which was registered with ClinicalTrials.gov at Feb. 2009 (NCT01003964). EGFR and KRAS mutations in the plasma cfDNA were analyzed retrospectively using a PNA clamping-assisted fluorescence melting curve analysis. The results were compared with those obtained from tissue analysis performed using the direct sequencing. Exploratory analyses were performed to determine survival predicted by the plasma and tissue mutation status. RESULTS: Mutation analyses in matched tissue and plasma samples were available for 194 patients for EGFR and 135 patients for KRAS. The mutation concordance rates were 82.0 % (95 % confidence interval [CI], 76.5-87.4) for EGFR and 85.9 % (95 % CI, 80.1-91.8) for KRAS. The plasma EGFR mutation test sensitivity and specificity were 66.7 % (95 % CI, 60.0-73.3) and 87.4 % (95 % CI, 82.7-92.1), respectively, and the plasma KRAS mutation test sensitivity and specificity were 50.0 % (95 % CI, 41.6-58.4) and 89.4 % (95 % CI, 84.2-94.6), respectively. The predictive value of the plasma EGFR and KRAS mutation status with respect to survival was comparable with that of the tissue mutation status. CONCLUSIONS: These data suggest that plasma EGFR and KRAS mutations can be analyzed using PNA-based real-time PCR methods and used as an alternative to tumor genotyping for NSCLC patients when tumor tissue is not available.