A polygenic risk score for alcohol-related cirrhosis among heavy drinkers with European ancestry

Background & Aims: Polygenic Risk Scores (PRS) based on results from genome-wide association studies offer the prospect of risk stratification for many common and complex diseases. We developed a PRS for alcohol-related cirrhosis by comparing SNPs among patients with alcohol-related cirrhosis (ALC) vs. drinkers who did not have evidence of liver fibrosis/cirrhosis.Approach & Results: Using a data driven approach, a PRS for ALC was generated using a meta-GWAS of ALC (N=4,305) and an independent cohort of heavy drinkers with ALC and without significant liver disease (N=3,037). It was validated in two additional independent cohorts from the UK Biobank with diagnosed ALC (N=467) and high-risk-drinking controls (N=8,981) and participants in the Indiana Biobank Liver cohort with alcohol-related liver disease (N=121) and controls without liver disease (N=3,239). A 20-SNPs PRS for ALC (PRSALC) was generated that stratified risk for ALC comparing the top and bottom deciles of PRS in the two validation cohorts (odds ratios 2.83 [95% CI 1.82-4.39] in UK Biobank; 4.40 [1.56-12.44] in Indiana Biobank Liver cohort). Furthermore, PRSALC improved the prediction of ALC risk when added to the models of clinically known predictors of ALC risk. It also stratified risk for MASLD-cirrhosis (3.94 [2.23-6.95]) in the Indiana Biobank Liver cohort based exploratory analysis.Conclusions: PRSALC incorporates 20 SNPs and predicts increased risk for ALC and improves risk stratification for ALC compared to the models that only include clinical risk factors. This new score has potential for early detection of heavy-drinking patients who are at high risk for ALC