The detection of EGFR mutation status in plasma is reproducible and can dynamically predict the efficacy of EGFR-TKI.

Abstract

The validity of epidermal growth factor receptor (EGFR) mutation in serum and plasma DNA as a surrogate of tumor tissue has been comprehensively explored. However, the concordance between peripheral blood and tumor tissue samples in EGFR mutation detection remains variable. The question as to whether real-time samples for EGFR mutation analysis are required before epidermal growth factor receptor tyrosine kinase inhibitor (EGFR-TKI) therapy remains unanswered. This study included two cohorts:(i) 822 non-small cell lung cancer (NSCLC) patients with primary tumor tissue and matched plasma samples at initial diagnosis; and (ii) 207 patients with advanced NSCLC who had plasma samples taken immediately before EGFR-TKI therapy, in which 157 cases had matched tumor tissues. Denaturing High-Performance Liquid Chromatography (DHPLC) determined EGFR mutation status. Among a total of 822 patients with matched samples, the EGFR mutation rates were 36.3% and 32.1% in tissue and plasma samples, respectively. Concordance of EGFR mutation between two kinds of samples was 77.0% (631/822),with 63.5% (188/296) of accuracy of EGFR mutation in plasma DNA. In 207 advanced NSCLC patients who had plasma samples taken immediately before EGFR-TKI therapy, the objective response rate (ORR) after EGFR-TKI therapy was significantly higher in EGFR mutant patients than those in EGFR wild-type patients (51.4% vs. 22.6%, P < 0.001), regardless of the treatment lines of EGFR-TKI. In patients with two or more lines of EGFR-TKI therapy, EGFR mutation status in plasma samples, but not in tissues, was a predictor for progression-free survival (PFS) after EGFR-TKI therapy (mutant vs. wild-type: 10.1 months vs. 3.7 months, P = 0.038). An EGFR mutation test using plasma DNA samples was validated and reproducible. Obtaining real-time samples for EGFR mutation detection is critical in order to predict the outcomes of EGFR-TKI.