Alcohol-related cognitive impairment confounds covert hepatic encephalopathy screening by PHES and the simplified...

Diagnosing Covert Hepatic Encephalopathy (CHE) in alcohol-related cirrhosis is complicated by potential overlap with alcohol-related cognitive impairment (ARCI), whose effect on recommended CHE tests remains poorly characterized. We aimed to assess the performance of PHES, MoCA and the Simplified Animal Naming Test (sANT) in patients with chronic alcohol exposure, with and without cirrhosis.

In this prospective single-center cohort, patients admitted for alcohol withdrawal or pre-liver-transplant workup were assessed with PHES (CHE if <-4), MoCA (impairment if <26) and sANT. Correlations and factors associated with CHE were analyzed by univariate and multivariate analysis; sANT diagnostic performance for CHE detection was evaluated by ROC analysis against PHES as the reference standard.

236 patients were included, 137(58%) with cirrhosis (age 61 [54-64], MELD 14 [9-19]; 13% clinical HE) and 99 without (age 48 [42-55]). PHES<-4 occurred in 34(33%) non-cirrhotic versus 52(38%) cirrhotic patients; sANT was below the CHE cut-off in 27% and 27% respectively. sANT AUROC for CHE in cirrhosis was 0.77(0.69-0.85). sANT correlated significantly with PHES and MoCA in both groups (rho 0.44-0.56). In multivariate analysis, PHES was associated with sANT, MoCA, Creatinine and albumin; MoCA with sANT, PHES and education.

A comparable proportion of patients crossed the PHES and sANT cut-offs whether or not they had cirrhosis, suggesting failure to distinguish ARCI from CHE in alcohol-related cirrhosis. Clinicians should interpret abnormal CHE tests with caution in this population. Further research is required to better disentangle ARCI from CHE. Magnetic resonance imaging and spectroscopy, therapeutic tests, or more accurate biomarkers, are promising opportunities.