Validation of the Model for End-Stage Liver Disease 3.0 in Korean Patients on the Liver Transplant Waiting List

Abstract

Background & Aims: Model for End-stage Liver Disease (MELD) 3.0 has recently been introduced with an improved discriminative ability over MELD-Na by adding serum albumin and sex. This study aims to validate the performance of MELD3.0 in Korean patients awaiting liver transplantation. Methods: Between June 2016 and December 2021, 2,132 Korean patients on the liver transplant waiting list with end-stage liver disease were analyzed. The 90-day mortality predictive performance was assessed based on the MELD, MELD-Na, MELD3.0, and MELD3.0 without albumin. Results: The mean age of the patients was 55.9 years, and 1,556 (73.0%) were male. Chronic hepatitis B (64.5%) was the most common liver disease etiology. The median Child?Pugh score was 9 (IQR, 7?11). The 90-day cumulative death rate was 8.3%. Using Harrel’s method, the C-statistic for 90-day mortality of MELD, MELD-Na, MELD3.0, and MELD3.0 without albumin was 0.853, 0.859, 0.870, and 0.871, respectively. C-statistics according to Uno’s method were 0.745, 0.769, 0.782, and 0.782, respectively. No significant difference in performance was observed between MELD 3.0 and MELD3.0 without albumin. In contrast to the MELD3.0 original cohort, female sex was not associated with 90-day mortality in our study. With MELD3.0, more female (14.1%) than male (5.6%) patients were up-categorized compared with MELD-Na in the entire cohort as well as the deceased 150 patients (male:6.5% and female: 27.9%). Conclusions: MELD3.0 and MELD3.0 without albumin showed improved predictive performance for 90-day mortality in Korean patients on the liver transplant waiting list compared with MELD-Na.