Development and validation of a polygenic hazard score to predict prognosis and adjuvant chemotherapy benefit in early-stage non-small cell lung cancer

Background: It remains controversial who would benefit from adjuvant chemotherapy (ACT) in patients with early-stage non-small cell lung cancer (NSCLC). We aim to construct a polygenic hazard score (PHS) to predict prognosis and ACT benefit among NSCLC patients.Methods: We conducted a retrospective study including 1,395 stage I-II NSCLC patients. We performed a genome-wide association study (GWAS) on overall survival (OS) in patients treated with ACT (SYSUCC ACT set, n=404), and then developed a PHS using LASSO Cox regression in a random subset (training, n=202) and tested it in the remaining set (test, n=202). The PHS was further validated in two independent datasets (SYSUCC surgery set, n=624; PLCO cohort, n=367).Results: The GWAS-derived PHS consisting of 37 single-nucleotide polymorphisms (SNPs) was constructed to classify patients into high and low PHS groups. For patients treated with ACT, those with low PHS had better clinical outcomes than high PHS (test set: HR =0.21, P<0.001; PLCO ACT set: HR =0.33, P=0.260). Similar results were found in the extended validation cohorts including patients with or without ACT (SYSUCC: HR =0.48, P<0.001; PLCO: HR =0.60, P=0.033). Within subgroup analysis by treatment or clinical factors, we further observed consistent results for the prognostic value of the PHS. Notably, ACT significantly improved OS in stage II patients with low PHS (HR =0.26, P<0.001), while there was no ACT survival benefit among patients with high PHS (HR =0.97, P=0.860).Conclusions: The PHS improved prognostic stratification and could help identify patients who were most likely to benefit from ACT in early-stage NSCLC.