A multi-institutional study to evaluate the feasibility of next-generation sequencing and genomic analysis using formalin-fixed, paraffin-embedded biopsies of gastric cancer.

Abstract

Formalin-fixed, paraffin-embedded (FFPE) samples acquired and preserved adequately are expected to faithfully maintain tumor characteristics. Endoscopic biopsy tissues represent an attractive resource for identifying predictive biomarkers to evaluate pretreatment responses of patients with advanced gastric cancer (GC). However, whether genomic profiles obtained through next-generation sequencing (NGS) using biopsy samples match well with those gained from surgical FFPE samples remains a concern. We collected 50 FFPE samples (26 biopsies and 24 surgical samples) from patients with GC who participated in phase III clinical trial JCOG1509. The quality and quantity of FFPE samples were determined for deep sequencing using NGS. We queried a 435-gene panel CANCERPLEX-JP to generate comprehensive genomic profiling data including the tumor mutation burden (TMB). The median DNA yields and NGS success rates of biopsy samples compared with surgical samples were 879ng and 80.8% vs 8523ng and 100%, respectively. Epstein-Barr virus and microsatellite instability-high were detected in 9.5% of biopsy samples. Comparing the genomic profiles of 18 paired samples for which NGS data were available, we detected identical somatic mutations in paired biopsy and surgical samples (kappa coefficient, 0.8692). TMB positively correlated between paired biopsy and surgical samples (correlation coefficient, 0.6911). NGS is applicable to the analysis of FFPE samples of GC acquired by the endoscopic biopsy, and the data were highly concordant with those obtained from surgical specimens of the same patients.