Disentangling the genetic relationship between malignant skin neoplasms and sepsis: Evidence from bidirectional Mendelian randomization.

Sepsis, a potentially fatal condition marked by an abnormal immune response, is likewise noted during the development of malignant skin tumors. However, it remains uncertain whether there is a causal link between malignant skin tumors and sepsis. A bidirectional 2-sample Mendelian randomization (MR) analysis was performed utilizing data from the most comprehensive genome-wide association studies on malignant skin neoplasms, involving 218,792 participants, along with genome-wide association studies data from the UK Biobank cohort (N = 486,484; comprising 1896 sepsis cases and 484,558 controls). The inverse-variance weighted method served as the primary analytical approach, complemented by 4 additional MR techniques (MR-Egger, weighted median, and weighted mode) for sensitivity assessments. Heterogeneity was evaluated using Cochran Q test. To verify the robustness of the MR findings, a leave-one-out analysis was conducted. The Bonferroni correction was applied to assess the strength of causal associations. Finally, data from the FinnGen study and the UK Biobank cohort were synthesized through meta-analysis. In the primary inverse-variance weighted analysis, genetically predicted malignant skin neoplasms were significantly associated with sepsis (odds ratio = 1.14, 1.12, 1.14, 1.12; 95% confidence interval: 1.04-1.26, 1.01-1.23, 1.04-1.25, 1.01-1.23; P < .001 or P = .03). These results were consistent across sensitivity analyses, and no evidence of directional pleiotropy was observed (MR-Egger intercept: P = .15). Moreover, genetically predicted sepsis showed no causal effect on the development of malignant skin neoplasms. The results of this study suggest that malignant skin neoplasms may be a risk factor for sepsis and could contribute to the progression of the condition.