Lung Function–Associated Genetic Variants and Overall Survival in Non–Small Cell Lung Cancer: A Multi-Platform Analysis

Abstract Background: Genome-wide association studies (GWAS) have identified numerous variants associated with spirometric lung function, and Mendelian randomization studies suggest that impaired pulmonary function contributes to lung cancer susceptibility. Whether these variants influence survival after non–small cell lung cancer (NSCLC) diagnosis remains unclear. Methods: We evaluated 77 lung function GWAS–identified single-nucleotide polymorphisms (SNPs) in a European-ancestry NSCLC cohort genotyped on three platforms (HSPH, n=979; OncoArray, n=2,322; MGH, n=1,088; total n=4,389). Within each subgroup, Cox debiased lasso regression jointly estimated conditional associations between SNPs and overall survival (OS), adjusting for age, sex, smoking, stage, treatment, and ancestry principal components. Debiased log-hazard ratios were combined by inverse-variance–weighted fixed-effect meta-analysis, with false discovery rate (FDR) correction across the 65 SNPs available in ≥2 subgroups. Results: Among 4,389 patients, 3,630 deaths (82.7%) occurred. Eleven SNPs were nominally associated with OS (P<0.05) and two reached FDR q<0.05. Only rs11022690 (11p15.4) was available in all three subgroups and met prespecified Tier 1 criteria; it was associated with reduced mortality (hazard ratio [HR], 0.927; 95% confidence interval [CI], 0.887–0.969; P=7.79×10⁻⁴; q=0.050), with consistent direction across subgroups. Heterogeneity was low (I²=0% for 86% of SNPs). In leave-one-cohort-out analyses, a weighted polygenic risk score was associated with survival in held-out subgroups (pooled HR per SD, 1.06; 95% CI, 1.02–1.09). Conclusions: Selected lung function–associated variants are modestly associated with NSCLC survival; these findings warrant replication in independent populations. Impact: This study provides a reproducible framework for evaluating moderate-dimensional germline prognostic associations across genotyping platforms, extending standard candidate-SNP survival analysis.