Mendelian Randomization Using a Japanese GWAS Identifies an HLA-Linked Causal Effect of Chronic Hepatitis B on Cholangiocarcinoma Risk

Elgeshy, Khaled; Nagaoka, Katsuya; Yamazaki, Hajime; Honda, Munenori; Nagaoka, Maiko; Mizuta, Takahiro; Toyota, Toshinori; Maeda, Daiki; Kurano, Sotaro; Tanaka, Kentaro; Narahara, Satoshi; Inada, Hiroki; Tokunaga, Takayuki; Iio, Etsuko; Watanabe, Takehisa; Setoyama, Hiroko; Uojima, Haruki; Tanaka, Yasuhito

doi:10.31557/APJCP.2026.27.8.2923

Mendelian Randomization Using a Japanese GWAS Identifies an HLA-Linked Causal Effect of Chronic Hepatitis B on Cholangiocarcinoma Risk

Document Type : Research Articles

Authors

¹ Department of Gastroenterology and Hepatology, Graduate School of Medical Sciences, Kumamoto University, Kumamoto, Japan.

² Section of Clinical Epidemiology, Department of Community Medicine, Graduate School of Medicine, Kyoto University, Kyoto, Japan.

³ Health Care Center, Kumamoto University, Kumamoto, Japan.

10.31557/APJCP.2026.27.8.2923

Abstract

Background: Cholangiocarcinoma (CCA) is a highly malignant cancer that develops in the bile ducts. Its incidence is particularly high in East Asian populations, but the underlying genetic factors remain unclear. To investigate potential risk factors for CCA, we conducted a Mendelian randomization study to infer causality. Methods: Using large-scale genome-wide association study data from the BioBank Japan resource, we systematically investigated the causal effects of genetic predisposition to seven conditions, chronic hepatitis B (CHB), chronic hepatitis C, autoimmune hepatitis, type 1 diabetes, type 2 diabetes, chronic gastritis, and chronic pancreatitis, on CCA risk. Results: Our analysis reveals a significant association between genetic susceptibility to CHB with a 24% higher likelihood of developing CCA than non-susceptible individuals (Inverse-Variance Weighted Odds Ratio = 1.24, 95% Confidence Interval: 1.08–1.42; p = 0.002). This genetic association is significantly driven by instrumental variables enriched in the immune-regulatory HLA class II region (6p21), suggesting a plausible biological mechanism. For the primary outcome (CCA), statistical significance was assessed across seven exposures at a Bonferroni-corrected threshold (two-sided p<0.0071). Notably, the CHB–CCA association remains significant after correction. This primary finding is strongly supported by comprehensive sensitivity analyses that showed no evidence of confounding by horizontal pleiotropy or heterogeneity. Conversely, no significant causal effects on CCA were identified for the other six conditions. Conclusions: Our MR analysis supports a causal role of HBV infection in CCA development, highlighting the importance of targeted HBV screening and surveillance.

Keywords