Confirmation of Gene Expression–Based Prediction of Survival in Non–Small Cell Lung Cancer

Open Access
Publisher Website
Google Scholar
Abstract
Purpose: It is a critical challenge to determine the risk of recurrence in early stage non–small cell lung cancer (NSCLC) patients. Accurate gene expression signatures are needed to classify patients into high- and low-risk groups to improve the selection of patients for adjuvant therapy. Experimental Design: Multiple published microarray data sets were used to evaluate our previously identified lung cancer prognostic gene signature. Expression of the signature genes was further validated with real-time reverse transcription-PCR and Western blot assays of snap-frozen lung cancer tumor tissues. Results: Our previously identified 35-gene signature stratified 264 patients with NSCLC into high- and low-risk groups with distinct overall survival rates (P < 0.05, Kaplan-Meier analysis, log-rank tests). The 35-gene signature further stratified patients with clinical stage 1A diseases into poor prognostic and good prognostic subgroups (P = 0.0007, Kaplan-Meier analysis, log-rank tests). This signature is independent of other prognostic factors for NSCLC, including age, sex, tumor differentiation, tumor grade, and tumor stage. The expression of the signature genes was validated with real-time reverse transcription-PCR analysis of lung cancer tumor specimens. Protein expression of two signature genes, TAL2 and ILF3, was confirmed in lung adenocarcinoma tumors by using Western blot analysis. These two biomarkers showed correlated mRNA and protein overexpression in lung cancer development and progression. Conclusions: The results indicate that the identified 35-gene signature is an accurate predictor of survival in NSCLC. It provides independent prognostic information in addition to traditional clinicopathologic criteria.