No genetic causal association between human papillomavirus and lung cancer risk: a bidirectional two-sample Mendelian randomization analysis

Abstract

Introduction: Several observational or retrospective studies have previously been conducted to explore the possible association between lung cancer and human papillomavirus (HPV) infection. However, there may be inconsistencies in the data and conclusions due to differences in study design and HPV testing methods. There are currently no studies that provide conclusive evidence to support the involvement of HPV in the occurrence and development of lung cancer. Therefore, the relationship between HPV and lung cancer remains controversial and uncertain. This study aimed to explore whether HPV infection is causally related to lung cancer risk by systematically performing a two-way Two-Sample Mendelian Randomization (TSMR) analysis.

Methods: In the International Lung Cancer Consortium (ILCCO) genome-wide association study dataset, we included 11,348 lung cancer (LUCA) cases, including 3275 squamous cell carcinoma (LUSC) cases, 3442 adenocarcinoma (LUAD) cases, and 15,861 cases of control. Using genetic variants associated with the HPV E7 protein as instrumental variables, we summarized statistics associated with HPV infection in the MRC IEU OpenGWAS database, which included the HPV-16 E7 protein and the HPV-18 E7 protein. Two-sample Mendelian randomization (MR) results are expressed as odds ratios (OR) and 95% confidence intervals (CI).

Results: Based on a comprehensive analysis of genome-wide association study (GWAS) data from public databases, we mainly used inverse-variance weighted (IVW) to estimate causal relationships, while using MR-Egger, weighted median, simple mode, and weighted mode, and other four methods as supplements. Two-sample MR Analysis revealed no causal relationship between exposure factors (HPV-16 E7 protein and HPV-18 E7 protein) and outcome factors (lung cancer (LUCA) and its subtypes squamous cell carcinoma (LUSC) and adenocarcinoma (LUAD)) in forward MR Analysis using the IVW approach.HPV-16 E7 protein and LUCA and its subtypes LUSC and LUAD by IVW method results: [OR] = 1.002; 95% [CI]: 0.961 - 1.045; p = 0.920; [OR] = 1.023; 95% [CI]: 0.966 - 1.084; p = 0.438; [OR] = 0.994; 95% [CI]: 0.927 - 1.066; p = 0.872); HPV-18 E7 protein and LUCA and its subtypes LUSC and LUAD by IVW method results: [OR] = 0.965; 95% [CI]: 0.914 - 1.019; p = 0.197; [OR] = 0.933; 95% [CI]: 0.834 - 1.043; p = 0.222; [OR] = 1.028; 95% [CI]: 0.945 - 1.118; p = 0.524. It was observed through reverse MR that LUCA and its subtypes LUSC and LUAD were used as exposure factors, and HPV infection (HPV-16 E7 protein and HPV-18 E7 protein) was used as the outcome factors, the results of the IVW method are also invalid.LUCA and HPV-16 E7 protein and HPV-18 E7 protein by IVW method results: [OR] = 1.036; 95% [CI]: 0.761 - 1.411; p = 0.82; [OR] = 1.318; 95% [CI]: 0.949 - 1.830; p = 0.099; LUSC and HPV-16 E7 protein and HPV-18 E7 protein by IVW method results: [OR] = 1.123; 95% [CI]0.847 - 1.489; p = 0.421; [OR] = 0.931; 95% [CI]: 0.660 - 1.313; p = 0.682; LUAD and HPV-16 E7 protein and HPV-18 E7 protein by IVW method results: [OR] = 1.182; 95% [CI] 0.983 - 1.421; p = 0.075; [OR] = 1.017; 95% [CI]: 0.817 - 1.267; p = 0.877.Our results indicate that there is no causal relationship between genetically predicted HPV infection and LUCA and its subtypes LUSC and LUAD. In addition, in the reverse MR analysis, we did not observe a significant causal relationship between LUCA and its subtypes LUSC and LUAD on HPV infection.

Conclusions: Our findings do not support a genetic association between HPV infection and lung cancer.

Keywords: Causal relationship; HPV infection; Human papillomavirus; Lung adenocarcinoma; Lung cancer; Lung squamous cell carcinoma; Mendelian randomization.

Fig. 1

Schematic representation of the bidirectional bidirectional Mendelian randomization (MR) study design. In the figure, blue indicates forward MR analysis with HPV as the exposure and lung cancer as the outcome, while red represents reverse MR analysis with lung cancer as the exposure and HPV as the outcome. Dashed lines indicate unrelated relationships, while solid lines denote associations between variables and confounding factors or outcomes that cannot be overcome

Fig. 2

Causal estimates of the relationship between HPV (HPV-16 E7 protein, HPV-18 E7 protein) and lung cancer (LUCA), including its subtypes squamous cell carcinoma (LUSC) and adenocarcinoma (LUAD), represented by odds ratios (OR) and 95% confidence intervals (CI). MR: Mendelian randomization; IVW: Inverse variance weighting; LUCA: Lung cancer; LUSC: Squamous cell carcinoma; LUAD: Adenocarcinoma

Fig. 3

Scatter plots assessing the causal relationship between HPV (HPV-16 E7 protein, HPV-18 E7 protein) and lung cancer, as well as its subtypes. Specifically, (A) causal estimates of HPV-16 E7 protein on lung cancer, (B) causal estimates of HPV-16 E7 protein on squamous cell carcinoma, (C) causal estimates of HPV-16 E7 protein on adenocarcinoma, (D) causal estimates of HPV-18 E7 protein on lung cancer, (E) causal estimates of HPV-18 E7 protein on squamous cell carcinoma, and (F) causal estimates of HPV-18 E7 protein on adenocarcinoma. The slope of each line corresponds to the causal estimate of each method. Individual SNP effects on both the outcome and exposure are depicted by vertical and horizontal lines, respectively. LUCA: Lung cancer; LUSC: Squamous cell carcinoma; LUAD: Adenocarcinoma

Fig. 4

Funnel plots depicting overall heterogeneity in MR estimates of the impact of HPV (HPV-16 E7 protein, HPV-18 E7 protein) on lung cancer and its subtypes. A Funnel plot for the causal effect of HPV-16 E7 protein on lung cancer. B Funnel plot for the causal effect of HPV-16 E7 protein on squamous cell carcinoma. C Funnel plot for the causal effect of HPV-16 E7 protein on adenocarcinoma. D Funnel plot for the causal effect of HPV-18 E7 protein on lung cancer. E Funnel plot for the causal effect of HPV-18 E7 protein on squamous cell carcinoma. F Funnel plot for the causal effect of HPV-18 E7 protein on adenocarcinoma. IVW, Inverse Variance Weighting

Fig. 5

Leave-one-out analysis of the causal relationship between HPV (HPV-16 E7 protein, HPV-18 E7 protein) and lung cancer and its subtypes. The red line represents the result of the random-effects IVW analysis. A HPV-16 E7 protein and lung cancer. B HPV-16 E7 protein and squamous cell carcinoma. C HPV-16 E7 protein and adenocarcinoma. D HPV-18 E7 protein and lung cancer. E HPV-18 E7 protein and squamous cell carcinoma. F HPV-18 E7 protein and adenocarcinoma. LUCA: Lung cancer; LUSC: Squamous cell carcinoma; LUAD: Adenocarcinoma. IVW: Inverse Variance Weighting

Fig. 6

Forest plots of the causal effects of single nucleotide polymorphisms (SNPs) associated with HPV (HPV-16 E7 protein, HPV-18 E7 protein) on lung cancer and its subtypes. A Forest plot for the causal effect of HPV-16 E7 protein on lung cancer. B Forest plot for the causal effect of HPV-16 E7 protein on squamous cell carcinoma. C Forest plot for the causal effect of HPV-16 E7 protein on adenocarcinoma. D Forest plot for the causal effect of HPV-18 E7 protein on lung cancer. E Forest plot for the causal effect of HPV-18 E7 protein on squamous cell carcinoma. F Forest plot for the causal effect of HPV-18 E7 protein on adenocarcinoma. LUCA: Lung cancer; LUSC: Squamous cell carcinoma; LUAD: Adenocarcinoma