Feed

Immune checkpoint inhibitors are used to treat head and neck cancer (HNC), but clinical responses vary, and systematic genetic evaluation is needed to prioritize candidate susceptibility genes. We applied summary-data-based Mendelian randomization (SMR) and HEIDI testing to screen 624 immune checkpoint-related genes, integrating GWAS data from the Million Veteran Program with blood-derived multi-omics QTLs, and explored replication in FinnGen. Two-sample MR sensitivity analyses (MR-Egger, weighted median, MR-PRESSO) assessed pleiotropy for prioritized candidates, which were further contextualized using tumor expression and survival outcomes (overall survival [OS], progression-free interval [PFI]) in The Cancer Genome Atlas (TCGA-HNSC) cohort. Several SMR-prioritized genes showed directional concordance between genetically predicted effects and tumor expression, including the protective candidate HGF (OR_SMR = 0.37, P = 0.010) and the risk-associated AKT1 (OR_SMR = 1.19, P = 0.024). Both were further supported by two-sample MR sensitivity analyses (HGF: OR_IVW = 0.72, P = 0.007; AKT1: OR_IVW = 1.17, P = 0.009), with no evidence of horizontal pleiotropy (Egger intercept P > 0.05). Higher tumor HGF expression correlated with better OS (HR_Cox = 0.70, P = 0.008) and PFI (HR_Cox = 0.72, P = 0.020), whereas higher AKT1 expression correlated with poorer OS (HR_Cox = 1.42, P = 0.010). None of the eQTL or pQTL associations replicated in FinnGen. By integrating genetic signals with tumor transcriptomic and clinical data, this study prioritized immune checkpoint-related genes potentially involved in HNC susceptibility. HGF and AKT1 emerged as the most consistently supported candidates, showing convergent tumor transcriptomic and survival support despite lacking genetic replication in FinnGen, providing a foundation for further biological investigation.