Hepatic Encephalopathy Expands the Predictivity of Meld in Liver Transplant Setting: Evidence by Means of two Independent Cohorts

Despite its documented prognostic relevance, hepatic encephalopathy (HE) is not considered in liver transplantation (LT) due to its possible poor objectivity. To override this problem, we aimed to analyze if an objective diagnosis of HE may confer additional mortality risk beyond MELD. A study and a validation cohorts of cirrhotic patients were considered in Italy and Canada respectively. Patients were considered as "HE+" if an episode of overt HE was documented in a hospitalization. Of the 486 patients enrolled in Italy, 184(38%) were HE+. During the 6-month follow-up, 77 patients died, 50 were transplanted. The 6-month mortality of HE+ vs HE- patients was significantly higher (P<0.001). MELD (sHR 1.16; IC:1.1-1.22; P<0.001), "HE+"(sHR 3.6; IC:1.8-7.1; P<0.001) and sodium (sHR 0.9; IC:0.8-0.9; P<0.001) were independent predictors of 6-month mortality. In HE+ patients, short-term mortality was increased across the entire MELD spectrum (range 6-40). The results were unchanged by including or excluding patients with hepatocellular carcinoma or stratifying patients according to HE characteristics. The higher 6-month mortality of HE+ vs HE- patients was confirmed also in the Canadian cohort (p<0.001) (n=300, 33% HE+; 33 died, 104 transplanted). A similar and statistically significant c-index increase deriving by the incorporation of HE in MELD was observed both in the Italian (from 0.67 to 0.75) and Canadian (from 0.69 to 0.74) cohorts. A score based on MELD+7 points (95% CI: 4-10) for HE+ patients optimally predicted 6-month mortality in the 2 cohorts. According to the net-reclassification-index, by not considering HE, 29% of overall patients were misclassified by MELD.