Messenger-RNA expression of five gemcitabine sensitivity-related genes predicting outcome in advanced-stage non-small cell lung cancer

Abstract

Background/Aim: Tumoural transcriptional levels of RRM1, RRM2, CDA, dCK and hENT1 genes are potential biomarkers for gemcitabine's efficacy in non-small cell lung cancer (NSCLC). Patients and Methods: We retrospectively analysed each gene's relative mRNA expression by quantitative, real-time polymerase chain reaction in microdissected, formalinfixed, paraffin-embedded primary-tumour specimens from 219 chemonaïve patients with advanced-stage NSCLC, treated with gemcitabine-based regimens within clinical trials. The five genes' transcriptional patterns were integrated into an ordinal, five-level gemcitabine-susceptibility classifier (5L-GSC). Results: Treatment efficacy increased progressively across the five susceptibility levels, with the very-high chemosensitivity cases obtaining the most clinical benefit. 5L-GSC emerged as an independent prognosticator for overall response and disease control rates, time to progression and overall survival at pvalues of 0.03, 0.004, <0.001 and <0.001, respectively, with results remaining significant after bootstrapping. Penalised, optimally-scaled, categorical-regression modelling of overall response identified 5L-GSC as the most stable predictor. Conclusion: The proposed composite biomarker is promising for customising front-line chemotherapy in NSCLC.