Clinical application of targeted deep sequencing for predicting responsiveness to epidermal growth factor receptor (EGFR)-tyrosine kinase inhibitors (TKIs) in never-smoking lung adenocarcinomas (NSLA).

Abstract

e19019 Background: Most patients with EGFR-mutant lung adenocarcinoma response well to EGFR-TKIs. However, a subset of patients showed primary resistance to EGFR-TKIs. The routine conventional methods have limitation in predicting primary resistance among EGFR mutant lung cancers. We performed targeted deep sequencing to explore the accuracy and additional values in predicting responsiveness to EGFR-TKIs in NSLA. Methods: We performed direct sequencing to identify EGFR mutations in 48 NSLA received gefitinib or erlotinib for their recurrent advanced disease after surgery. The peptide nucleic acid (PNA) clamp polymerase chain reaction (PCR) was used to confirm the EGFR mutation status. To exclude well known causes of resistance, we analyzed KRAS, BRAF and PIK3CA mutations by PNA clamp PCR, ALK rearrangement by FISH, and RET and ROS-1 rearrangements by reverse transcriptase (RT)-PCR and quantitative real-time PCR. After screening, Ion Torrent Personal Genome Machine (PGM) was performed in 31 patients harboring EGFR exon 19 deletions (19DEL) or L858R mutations or none of above molecular changes. Results: The conventional screening found 25 EGFR mutations (10 19DEL; 9 L858R, 1 G719X, 4 double EGFR mutations, 1 concomitant L858R and PIK3CA mutations), 4 KRAS mutations, 3 ALK, 2 RET and 2 ROS-1 rearrangements. No mutation was found in 12 patients. After screening, samples were divided into 4 groups; 1) good responders to EGFR-TKIs with only 19DEL or L858R (n=15); 2) non-responder to EGFR-TKI with only 19DEL or L858R (n=4); 3) non-responders to EGFR-TKI without any mutation (n=8); 4) stable disease to EGFR-TKI without any mutations (n=4). With PMG, all conventionally detected mutations were confirmed, and some additional mutations were detected in a subset of samples; two KRAS and one PIK3CA mutations in group 3, one concomitant PIK3CA and L858R mutations in group 3, one PIK3CA mutation in group 2, and one 19DEL mutation in group 4. Conclusions: Targeted deep sequencing may improve the detection of clinically actionable genomic alterations and enable mutation-driven personalized approach in lung cancer patients.