Validation of a targeted sequencing panel with automatic analysis system for clinical decision support in cancer therapy

Abstract

Abstract Purpose Precision cancer therapy relies on the identification of tumor-specific genomic alterations, which can be achieved through next-generation sequencing (NGS). In the clinic, personalized treatment for patients with advanced treatment-refractory solid tumors often requires rapid and comprehensive multi-dimensional molecular signature analysis using tumor-only samples because paired normal specimens are unavailable in most cases. To address this issue, we developed a CancerMaster panel, targeted NGS panel with 524 key genes specifically designed for multi-dimensional molecular signature analysis of solid tumors. Methods Its asynchronous and parallel one-stop automated analysis pipeline with a reporting system provides a comprehensive solution to shorten the turnaround time from analysis to reporting. The panel can detect common genomic alteration types, including SNVs/Indels and CNVs, fusions, Epstein-Barr virus (EBV)/Human papillomavirus (HPV) infection, microsatellite instability (MSI), tumor mutational burden (TMB) status and human leukocyte antigen (HLA) typing. Results We confirmed its reproducibility (100%) and analytical sensitivity (99%) using reference materials and performed clinical validation of the panel, which demonstrated a high accuracy (94%). Using the CancerMaster panel, we identified actionable mutations (TP53, KRAS, and PIK3CA) and CNV (ERBB2 amplification) mainly in gastric and colorectal cancer. We also found a high correlation between MSI and TMB in our patient samples (n = 668, r = 0.75, p < ), especially for gastric cancer (n = 412, r = 0.75, p < ) and colorectal cancer (n = 66, r = 0.87, p < ). Conclusion The CancerMaster panel demonstrated the potential for clinical decision support in personalized cancer treatment.