Mendelian randomization analysis of inflammatory proteins and lung adenocarcinoma risk

Abstract

To investigate the potential causal effects of six circulating inflammatory proteins on lung adenocarcinoma risk using Mendelian randomization. Genetic instrumental variables were selected from genome-wide association study (GWAS) summary statistics for six inflammatory proteins (beta-nerve growth factor, C-C motif chemokine 25, cystatin D, interleukin-18, tumor necrosis factor, and VEGF-A). Three Mendelian randomization methods, including inverse-variance weighted (IVW) analysis as the primary method, MR-Egger regression, and weighted median analysis, were applied with a relaxed instrument selection threshold of P < 1 × 10^-5. Sensitivity analyses were conducted to assess heterogeneity, horizontal pleiotropy, and the stability of the findings. Among the six inflammatory proteins examined, C-C motif chemokine 25 (CCL25) showed a significant causal association with increased lung adenocarcinoma risk (IVW OR = 1.055, 95% CI: 1.021–1.090, P = 0.0013), which survived Bonferroni correction for multiple testing (threshold P < 0.0083). Beta-nerve growth factor, cystatin D, interleukin-18, tumor necrosis factor, and VEGF-A did not show significant associations after correction. Sensitivity analyses did not indicate substantial horizontal pleiotropy, though interleukin-18 showed evidence of heterogeneity (Cochran Q P = 0.002). Among six circulating inflammatory proteins, C-C motif chemokine 25 was significantly associated with increased lung adenocarcinoma risk, whereas the remaining five proteins showed null associations after correction for multiple testing. These findings suggest a specific rather than generalized role of inflammatory mediators in lung adenocarcinoma risk. Results should be interpreted cautiously because of the relaxed instrument selection threshold and the need for external validation in independent populations.