METHODS: We performed a GWAS of 35,839 participants in the UKB with high intake of alcohol against markers of hepatic fibrosis (FIB-4, APRI and Forns index scores) and hepatocellular injury (levels of aminotransferases). Loci identified in the discovery analysis were tested for their association with alcohol-related cirrhosis in 3 separate European cohorts (phase 1 validation cohort; n=2545). Variants associated with alcohol-related cirrhosis in the validation at a false-discovery rate of less than 20% were then directly genotyped in 2 additional European validation cohorts (phase 2 validation, n=2068).
RESULTS: In the GWAS of the discovery cohort, we identified 50 independent risk loci with genome-wide significance (P<5 x 10-8). Nine of these loci were significantly associated with alcohol-related cirrhosis in the phase 1 validation cohort; 6 of these 9 loci were significantly associated with alcohol-related cirrhosis in phase 2 validation cohort, at a false discovery rate below 5%. The loci included variants in the mitochondrial amidoxime reducing component 1 gene (MARC1) and the heterogeneous nuclear ribonucleoprotein U like 1 gene (HNRNPUL1). After we adjusted for age, sex, body mass index, and type-2 diabetes in the phase 2 validation cohort, the minor A allele of MARC1:rs2642438 was associated with reduced risk of alcohol-related cirrhosis (adjusted odds ratio, 0.76; P=.0027); conversely the minor C allele of HNRNPUL1:rs15052 was associated with an increased risk of alcohol-related cirrhosis (adjusted odds ratio, 1.30; P=.020).
CONCLUSIONS: In a GWAS of samples from the UKB, we identified and validated (in 5 European cohorts) single-nucleotide polymorphisms that affect risk of alcohol-related cirrhosis in opposite directions: the minor A allele in MARC1:rs2642438 decreases risk whereas the minor C allele in HNRNPUL1:rs15052 increases risk.
- prognostic factor
- hepatic fibrogenesis