Background Lung cancer is the leading cause of cancer-related deaths globally, with non-small cell lung cancer (NSCLC) and small cell lung cancer (SCLC) being its two primary subtypes. While inflammation plays a crucial role in cancer development, its specific functions in different lung cancer subtypes remain unclear. Methods This study employed a bidirectional two-sample Mendelian randomization (MR) approach, utilizing Genome-Wide Association Studies (GWAS) data for NSCLC (5,315 cases) and SCLC (717 cases) from the FinnGen biobank, along with 314,193 controls and GWAS data for 91 circulating inflammatory factors. We investigated the causal relationships between inflammatory factors and lung cancer risk. Instrumental variables were selected based on p-value < 1 × 10− 5 and F-statistic > 10, with linkage disequilibrium (LD) clumping (R2 <0.001, physical distance threshold 10,000 kb) applied. The primary analysis used the inverse variance weighted (IVW) method, supplemented by MR-Egger, weighted median, simple mode, and weighted mode methods for sensitivity analyses. Heterogeneity was assessed using Cochran’s Q test, horizontal pleiotropy was evaluated using the intercept term from MR-Egger regression, and bias was assessed using the MR-PRESSO global test. Additionally, we conducted sensitivity analyses to evaluate the impact of individual Single Nucleotide Polymorphisms (SNPs) on the overall causal effect estimates. Results For NSCLC, Eotaxin (OR = 1.1150, 95%CI 1.0031–1.2394, P = 0.0436) and MCP-1 (OR = 1.0982, 95%CI 1.0078–1.1967, P = 0.0326) levels were positively associated with risk, while IL-12β (OR = 0.9214, 95%CI 0.8571–0.9906, P = 0.0267) and TRAIL (OR = 0.9188, 95%CI 0.8479–0.9956, P = 0.0388) levels were negatively associated with risk. For SCLC, IL-13 (OR = 1.4195, 95%CI 1.0736–1.8768, P = 0.0139), IL-17 A (OR = 1.4288, 95%CI 1.0058–2.0297, P = 0.0463), and MCP-1 (OR = 1.2898, 95%CI 1.0253–1.6224, P = 0.0297) levels were positively associated with risk, while CXCL1 (OR = 0.7448, 95%CI 0.5594–0.9915, P = 0.0435), DNER (OR = 0.7210, 95%CI 0.5265–0.9875, P = 0.0415), IL-15Rα (OR = 0.8035, 95%CI 0.6582–0.9809, P = 0.0316), and IL-18R1 (OR = 0.8090, 95%CI 0.6710–0.9753, P = 0.0262) levels were negatively associated with risk. Heterogeneity and pleiotropy tests showed no significant variables (P > 0.05). Reverse analyses did not reveal significant effects of lung cancer on these inflammatory factor levels (all p-values > 0.05). Conclusion This study elucidated potential causal relationships between specific inflammatory factors and the risk of NSCLC and SCLC, highlighting the complex roles of different inflammatory factors in the pathogenesis of these two lung cancer subtypes. These findings provide novel insights for molecular subtyping and personalized treatment strategies in lung cancer.

Bidirectional Mendelian randomization identifies differential causal associations between inflammatory cytokines and lung cancer subtypes

期刊论文
出版模式：

开放获取
发表日期：

2026年2月10日

查看全文

Menglu Sun,
Xue Gao

摘要

Background

Lung cancer is the leading cause of cancer-related deaths globally, with non-small cell lung cancer (NSCLC) and small cell lung cancer (SCLC) being its two primary subtypes. While inflammation plays a crucial role in cancer development, its specific functions in different lung cancer subtypes remain unclear.

Methods

This study employed a bidirectional two-sample Mendelian randomization (MR) approach, utilizing Genome-Wide Association Studies (GWAS) data for NSCLC (5,315 cases) and SCLC (717 cases) from the FinnGen biobank, along with 314,193 controls and GWAS data for 91 circulating inflammatory factors. We investigated the causal relationships between inflammatory factors and lung cancer risk. Instrumental variables were selected based on p-value < 1 × 10^− 5 and F-statistic > 10, with linkage disequilibrium (LD) clumping (R² <0.001, physical distance threshold 10,000 kb) applied. The primary analysis used the inverse variance weighted (IVW) method, supplemented by MR-Egger, weighted median, simple mode, and weighted mode methods for sensitivity analyses. Heterogeneity was assessed using Cochran’s Q test, horizontal pleiotropy was evaluated using the intercept term from MR-Egger regression, and bias was assessed using the MR-PRESSO global test. Additionally, we conducted sensitivity analyses to evaluate the impact of individual Single Nucleotide Polymorphisms (SNPs) on the overall causal effect estimates.

Results

For NSCLC, Eotaxin (OR = 1.1150, 95%CI 1.0031–1.2394, P = 0.0436) and MCP-1 (OR = 1.0982, 95%CI 1.0078–1.1967, P = 0.0326) levels were positively associated with risk, while IL-12β (OR = 0.9214, 95%CI 0.8571–0.9906, P = 0.0267) and TRAIL (OR = 0.9188, 95%CI 0.8479–0.9956, P = 0.0388) levels were negatively associated with risk. For SCLC, IL-13 (OR = 1.4195, 95%CI 1.0736–1.8768, P = 0.0139), IL-17 A (OR = 1.4288, 95%CI 1.0058–2.0297, P = 0.0463), and MCP-1 (OR = 1.2898, 95%CI 1.0253–1.6224, P = 0.0297) levels were positively associated with risk, while CXCL1 (OR = 0.7448, 95%CI 0.5594–0.9915, P = 0.0435), DNER (OR = 0.7210, 95%CI 0.5265–0.9875, P = 0.0415), IL-15Rα (OR = 0.8035, 95%CI 0.6582–0.9809, P = 0.0316), and IL-18R1 (OR = 0.8090, 95%CI 0.6710–0.9753, P = 0.0262) levels were negatively associated with risk. Heterogeneity and pleiotropy tests showed no significant variables (P > 0.05). Reverse analyses did not reveal significant effects of lung cancer on these inflammatory factor levels (all p-values > 0.05).

Conclusion

This study elucidated potential causal relationships between specific inflammatory factors and the risk of NSCLC and SCLC, highlighting the complex roles of different inflammatory factors in the pathogenesis of these two lung cancer subtypes. These findings provide novel insights for molecular subtyping and personalized treatment strategies in lung cancer.