Mendelian randomization evidence for a causal link between Epstein–Barr virus antibody levels and head and neck cancer risk

Epstein-Barr virus (EBV) has been consistently detected in head and neck cancer (HNC) tissues, suggesting its potential involvement in disease pathogenesis. These observations underscore the need to further investigate the plausible causal relationship between genetically predicted EBV antibody levels and the risk of developing HNC. We performed a 2-sample Mendelian randomization (MR) analysis. Genetic instruments for 5 EBV antibody levels including immunoglobin G, EBV nuclear antigen 1 (EBNA-1), EBV early antigen diffuse (EA-D), viral capsid antigen, and BamHI Z EBV replication activator were derived from a publicly available genome-wide association study dataset. Summary-level genome-wide association study data for HNC and oral and oropharyngeal cancer were obtained independently. Causal inference was performed using inverse-variance weighted (IVW), weighted median, and MR-Egger regression. Benjamini-Hochberg false discovery rate (FDR) correction was applied for multiple testing. Reverse MR analysis, Cochran Q test, MR-Egger intercept test, leave-one-out analysis, and funnel plot were subsequently conducted to assess the robustness of the results. The results of MR analysis revealed that higher genetically predicted EA-D antibody levels were significantly associated with oral and oropharyngeal cancer (IVW odds ratio [OR] = 1.00146, 95% confidence interval [CI] = 1.00066-1.00227, P = .00035, FDR-adjusted P = .004). The association between EBV EA-D antibody levels and overall HNC was of borderline significance (OR = 1.00132, 95% CI = 1.00012-1.00253, P = .032, FDR-adjusted P = .079). Conversely, elevated levels of EBV EBNA-1 antibody were inversely associated with the risk of HNC (IVW OR = 0.99886, 95% CI = 0.99809-0.99963, P = .00373, FDR-adjusted P = .012,) and oral and oropharyngeal cancer (IVW OR = 0.99897, 95% CI = 0.99836-0.99958, P = .00099, FDR-adjusted P = .005). No significant causal associations were observed for other EBV antibody levels. Reverse MR analysis did not provide evidence for a causal effect of HNC on genetically predicted EBV antibody levels. Sensitivity analyses indicated no major outliers. This study provides suggestive evidence of associations between genetically predicted EBV antibody levels and HNC risk. The inverse associations of EBNA-1 with both HNC subtypes and the positive association of EA-D with oral/oropharyngeal cancer remained significant after FDR correction. The observed effect sizes were modest, and the biological significance of these associations remains uncertain. Further validations in independent populations and mechanistic studies are warranted.