Specific Gene Co-variation Acts Better Than Number of Concomitant Altered Genes in Predicting EGFR-TKI Efficacy in Non–small-cell Lung Cancer

Background There occurs huge heterogeneity in clinical outcomes for patients with epidermal growth factor receptor (EGFR)-mutated non–small-cell lung cancer (NSCLC) treated with EGFR tyrosine kinase inhibitors (EGFR-TKIs). The purpose of this study was to indicate genetic biomarkers predicting primary resistance of EGFR-TKIs in these patients. Patients and Methods Using a next-generation sequencing panel with 168 cancer-related genes, matched tumor biopsy and plasma samples before treatments from patients with NSCLC were analyzed. Patients taking EGFR-TKIs were followed-up with imaging examination. Correlation of co-alterative genes with progression-free survival (PFS) was analyzed. Results Of the 48 patients treated with EGFR-TKIs, 46 (95.83%) had at least 1 genetic co-variant beyond EGFR mutation. Multivariate analysis indicated that RB1, PIK3CA, and ERBB2 co-alterations, rather than number of co-alterative genes, were independently associated with poorer PFS. Grouping patients by specific gene status showed best likelihood ratio χ2, Akaike information criterion, and Harrell concordance index. The median PFS for patients in group A (less genetic co-variations and wild specific genes), group B (more genetic co-variations and wild specific genes), group C (less genetic co-variations and altered specific genes), and group D (more genetic co-variations and altered specific genes) were 10.4, 9.13 (vs. group A; P = .3112), 6.33 (vs. group B; P = .0465), and 3.90 (vs. group C; P = .0309) months, respectively. Conclusions This study revealed a high concomitant genetic alteration rate in patients with EGFR-mutated NSCLC. Specific gene variants were more important than number of altered genes in predicting poor PFS, and may help select patients needing new treatment strategies.