Objective: We aimed to identify potential causal immune cells and their phenotypes in benign and malignant thymic tumors from the FinnGen database using Mendelian randomization (MR). Methods: A two-sample MR approach was employed to evaluate the causal associations between 731 immune cell traits and the risk of benign and malignant thymomas using publicly available genetic datasets. Immune traits analyzed included median fluorescence intensity (MFI), relative cell count (RC), absolute cell count (AC), and morphological parameters (MP). Robustness of findings was assessed through sensitivity analyses, ensuring minimal heterogeneity and horizontal pleiotropy. Results: Four immune cell phenotypes, including CCR2 on CD14+ CD16+ monocytes (95% CI: 1.163–1.775, P = 0.0007) and transitional AC (95% CI: 1.176–3.232, P = 0.0095), were causally associated with benign thymomas and identified as risk factors. Conversely, naive-mature B cell AC (95% CI: 0.438–0.887, P = 0.0087) and natural killer (NK) AC (95% CI: 0.237–0.812, P = 0.0087) displayed protective effects. For malignant thymomas, seven immune phenotypes, including Memory B cell %B cell (95% CI: 1.160–2.242, P = 0.0045), were linked as risk factors, while three, such as IgD+ CD24− %B cells (95% CI: 0.438–0.868, P = 0.0055), exhibited protective associations. Sensitivity analyses confirmed the robustness of the results. Conclusion: Using genetic methodologies, we identified a robust link between immune cells and both benign and malignant thymic tumors. This study highlights the distinct immune phenotypes between benign and malignant thymic tumors, shedding light on their immunological mechanisms and suggesting new avenues for clinical immunotherapy.

Lung cancer is a leading cause of morbidity and mortality worldwide, with smoking as a major risk factor. This study investigated the causal relationship between smoking intensity, plasma lipid phenotypes, and lung cancer risk using Mendelian randomization (MR) analysis. Bidirectional MR analysis examined daily cigarette consumption’s impact on non-small cell carcinoma, not otherwise specified (NOS), lung adenocarcinoma (LUAD), lung squamous cell carcinoma (LUSC), and small cell lung cancer (SCLC). Associations between smoking and 179 plasma lipid phenotypes were explored, followed by analysis of lipid groups’ causal links to lung cancer subtypes. Mediation MR estimated the mediation proportion. Smoking intensity showed a significant positive correlation with all lung cancer subtypes (non-small cell carcinoma, not otherwise specified (NOS) [NSCLC-NOS]: odds ratio [OR] = 2.31, 95% confidence interval [CI] = 1.30–4.10; LUAD: OR = 2.99, 95% CI = 1.47–6.10; LUSC: OR = 3.69, 95% CI = 1.39–9.80; SCLC: OR = 3.66, 95% CI = 1.06–12.56). Increased smoking was linked to lower levels of 3 phosphatidylcholine (PC) and higher levels of 2 sphingomyelin (SM) molecules. Bayesian weighted MR revealed 41 plasma lipid phenotypes significantly affected lung cancer risk. PC (16:0_0:0) and SM (d36:2) mediated 1.79% and 3.27% of the risk for NSCLC-NOS, respectively, while SM (d42:2) mediated 2.72% of the risk for SCLC. The study established a causal relationship between daily smoking intensity and the risk of both NSCLC and SCLC, mediated by changes in specific plasma lipid phenotypes. This highlights the critical role of lipid metabolism in smoking-related lung cancer. Understanding the lipid metabolic pathways involved in smoking-induced lung cancer could aid in the development of targeted interventions, potentially reducing the elevated cancer risk associated with increased smoking intensity.

The rising incidence of lung cancer in never-smokers (LCINS) warrants investigation into novel etiological factors. Gut microbiota-associated metabolic pathways (e.g., glycolysis, glyoxylate degradation) and immune traits, such as the proportion of memory B cells, have been implicated in various diseases, but their causal roles in LCINS remain unclear. Using summary-level data from IEU OpenGWAS, we performed 2-sample Mendelian randomization (MR) to assess causal relationships of 412 gut microbiota features and 731 immune cell traits with LCINS. Mediation MR quantified immune cells’ role in linking microbiota to LCINS. Twelve gut microbial taxa and 15 metabolic pathways (including glycolysis and glyoxylate degradation) demonstrated causal associations with LCINS ( P _IVW < .05). Separately, 36 peripheral immune cell traits showed causal links to LCINS. Mediation analysis further revealed that memory B cell ratio attenuated the effect of microbiota-derived metabolic pathways on LCINS by 7.53% (95% confidence interval: −2.63% to 17.70%; β _mediation = −0.017, 95% confidence interval: −0.041 to 0.006). The ratio of memory B cells partially mediates the effect of gut microbiota-driven glycolysis/glyoxylate degradation on LCINS pathogenesis. These findings suggest the gut-immune-lung axis may inform preventive strategies for LCINS.