P: 81 Previously Identified Candidate Gene Associations in Hepatic Encephalopathy Do Not Replicate in the STOPAH Cohort

Abstract

BACKGROUND: Hepatic encephalopathy (HE) is a common complication of cirrhosis and negatively affects quality of life and prognosis. However, the development of HE is not inevitable, even in patients severe hepatic decompensation. Recently, there has been interesting the possibility that the susceptibility to develop HE is, at least in part, genetically determined. In 2017, Gil-Gómez et al (J Hepatol 2017;66(Suppl 1):S144) reported significant associations between variants in nine candidate genes and the presence of overt HE in 400 patients included in the CLIF consortium cross-sectional study. In this study overt HE was diagnosed using the West Haven criteria, while acute on chronic liver failure (ACLF) was diagnosed using the CLIF-OF score. The aim of the present study was to replicate these findings in participants in the STOPAH alcoholic hepatitis treatment trial. METHODS: Genomic DNA was available in 731 participants in the STOPAH trial. Overt HE was diagnosed using the West Haven criteria; ACLF was diagnosed using the CLIF-OF score. Data on six of the nine SNP associated with overt HE and/or ACLF in the Spanish study were identified in the STOPAH GWAS database. Genotypic association analyses were undertaken for each of the genetic variants in the entire population and in subgroups defined by the presence of overt HE and/or ACLF. RESULTS: A total of 199 (27.2%) of the 731 STOPAH participants had overt HE on admission, while 193 (26.4%) had ACLF; patients were classified as follows: HE and ACLF (n = 61), HE but no ACLF (138); ACLF but no HE (n = 132); no ACLF and no HE (n = 400). Only one significant association was identified viz. rs752949 in SLC1A2 in participants with overt HE but without ACLF (Table 1); however, the significance of this association was lost when the data were corrected for multiple testing. CONCLUSIONS: The previously reported genetic associations with overt HE were not replicated in the present study. The number of participants was almost double that in the Spanish study but the proportions of patients with HE and ACLF were similar. Not all of the reported SNPs were available but at least two SNPs were tested for association in each patient subgroup. Further large case-control studies and meta-analyses of existing data are needed.