Causal Effect of Autoimmune Liver Diseases on Cancer

Meta-analyses of Cohort Studies and Mendelian Randomization Study

Dongqing Gu; Min Zhang; Yutong Wang; Ye Bai; Xin Wang; Guohong Deng


Liver International. 2022;42(10):2216-2226. 

In This Article

Abstract and Introduction


Background and Aims: Prior studies suggested that patients with autoimmune liver diseases (AiLDs) had an increased risk of cancer, whereas the causal effect remained unclear.

Methods: Meta-analyses concerning the relationship between AiLD and cancer risk were performed to calculate the pooled relative risk (RR) and corresponding 95% confidence intervals (CIs). Then, the associations with a p value of <.05 were further validated by two-sample Mendelian randomization studies.

Results: A total of 37 cohort studies covering more than 34 558 patients were included, and we observed an increased risk of overall cancers (pooled RR = 3.64, 95% CI: 2.64–5.03, p < .001) and cancer-related death (pooled RR = 2.48, 95% CI: 1.73–3.53, p < .001) for patients with AiLD. Besides, overall and several site-specific cancers risk were found in patients with primary biliary cholangitis (PBC), autoimmune hepatitis (AIH), and primary sclerosing cholangitis (PSC) (p < .05). However, associations between genetically predisposed AIH, PBC, and PSC and the risk of specific cancers did not reach a significant level, except for PBC and gastric cancer (OR = 0.96, 95% CI: 0.93–0.99; p = .02).

Conclusions: In addition to hepatobiliary cancer, results from the meta-analyses suggest that patients with AiLD might have an increased risk of several extrahepatobiliary cancers. However, the causal role of AiLD in cancer development needs to be further investigated.


Autoimmune liver diseases (AiLD) are a group of immune-mediated liver inflammatory diseases with three major subtypes including primary biliary cholangitis (PBC), autoimmune hepatitis (AIH) and primary sclerosing cholangitis (PSC).[1] These AiLD can lead to cirrhosis, hepatic and extrahepatic malignancies, liver failure and even death.[2–6] However, the quantitative data on the associations between AiLD and cancer and mortality risk remains scarce, and conflicting results also exist. Besides, the observational studies are susceptible to methodological biases, particularly residual confounding and reverse causation, these causal associations remain unestablished.

A Mendelian randomization study is a powerful tool that can be used to validate the causal effect of an exposure on the outcome in the presence of unobserved confounding factors using genetic variants as instrumental variables for the exposure.[7] Since the genotypes are presumed to be randomly allocated in the process of gamete formation, Mendelian randomization analysis is not affected by reverse causation and is less susceptible to confounding factors.[8] This approach is analogous to a randomized controlled trial and has become a powerful approach to assess the potential causal associations in recent years.[9,10]

To comprehensively evaluate the associations between AiLD (including AIH, PBC and PSC) and the risk of cancer development and cancer-related death, we firstly conducted a systematic review and meta-analysis of cohort studies. Additionally, the two-sample Mendelian randomization analyses were conducted to confirm the associations with a p value of <.05.