Comprehensive Analysis of Gene Mutation Profile in Chinese NSCLC Patients by Next-generation Sequencing

Overview

In recent years, the development of lung cancer has been improved from pathological level to the molecular level. Research showed that there are many gene mutations in non-small cell lung cancer (NSCLC), and some activating mutations have become the hotspots in target therapy area. With the development of targeted drug research, the molecular classification of NSCLC will be more and more important. But a large number of clinical data showed that gene mutation in Chinese NSCLC patients is significantly different from Caucasian population, which suggesting that it is necessary to identify gene mutation profile in Chinese patients with NSCLC. Six hundred NSCLC paraffin tissue samples was collected during operation from Tianjin Cancer hospital in 2009-2012, which including lung squamous cell carcinoma and adenocarcinoma. The target area of 295 genes, including lung cancer drive genes, important signal pathway genes, drug resistance genes will be detected by next-generation sequencing deep (average 1000X). We will identify gene mutation profile for Chinese lung squamous cell carcinoma and adenocarcinoma patients. The aim is to find related predictor and prognostic factors by analysing the relationship between these gene mutations and clinical characteristics and follow-up treatment.

The objective of this study is to build NSCLC gene mutation profile in China and find related correlation between gene mutation panel and clinical outcome. Approximately 600 surgical tissue samples will be collected during operation in Tianjin Cancer hospital from 2009-2012, including lung squamous cell carcinoma and adenocarcinoma. The target area of 295 genes, including lung cancer drive genes, important signal pathway genes, drug resistance genes will be detected by New generation Sequencing (NGS) deep (average 1000X). This genes was selected from Mutations can guide treatment or as prognosis factors in NCCN/FDA/CFDA guideline, related mutations in phase II/III studies and NCCN/FDA/CFDA approved in other type tumors and related mutations in phase I or pre-clinical studies and can not guide treatment or as prognosis factors. All mutations detected in 600 samples are summarized for statistics, calculating the mutation proportion in overall population. Clustering analysis is performed according to the main drive genes related biological pathways, correlation between gene mutation data and categorical clinical variables is performed by Fisher's Exact Test. The Log-rank test will be used to explore the relationship between the clinical outcomes (DFS and OS, respectively) and gene mutations (present or absent) or each of the clinical features (gender, age, smoking status, TNM staging, histology, tumor location, recurrence, number of lymph node metastasis, tumor size, postoperative adjuvant treatment, DFS and OS). Then a Cox Proportional Hazards model will be constructed to evaluate the effect of multiple variables (genomic and clinical features) on DFS, and OS, respectively. Benjamini-Hochberg false discovery rate (FDR) method is used to adjust the p-value and calculate the statistically differences. All p values were two-sided, and P\<0.05 was assumed to be significant.