Proposal of Muscle-MELD Score, Including Muscularity, for Prediction of Mortality After Living Donor Liver Transplantation

Abstract

BACKGROUND: The Model for End-stage Liver Diseases (MELD) is currently the most widely accepted scoring system for predicting prognosis and for allocation of liver grafts in patients awaiting liver transplantation (LT). However, this system is independent of the severity of recipient nutritional and functional status. METHODS: The present study retrospectively evaluated the preoperative quality and quantity of skeletal muscle in 231 adult patients undergoing living donor LT (LDLT) between January 2008 and December 2014. Using plain computed tomography imaging at the third lumbar vertebral level, the preoperative muscularity (quantity and quality of skeletal muscle) was evaluated. We included muscularity with the MELD score (Muscle-MELD) and investigated its ability to predict posttransplant mortality. RESULTS: Cox regression analysis provided the following equation for Muscle-MELD: Muscle-MELD = MELD score + 27.0 × low muscle quality + 25.2 × low muscle mass. The overall survival rate after LDLT was significantly lower in patients with high Muscle-MELD than in patients with low Muscle-MELD scores (P < 0.001). Muscle-MELD scores more accurately predicted posttransplant mortality than conventional MELD scores (P = 0.038 for 3 months, P = 0.002 for 6 months, P < 0.001 for 12 months, and P = 0.001 for overall mortality). With multivariate analysis, high Muscle-MELD score was an independent predictor of 6-month mortality after LDLT (odds ratio, 6.715; 95% confidence interval, 3.305-14.653; P < 0.001). CONCLUSIONS: Muscle-MELD score closely correlates with posttransplant mortality. Our results suggest that perioperative nutritional intervention and rehabilitation would affect posttransplant outcomes with improving pretransplant low muscularity.