Predicting Morbidity and Mortality After Hepatic Resection in Patients with Hepatocellular Carcinoma: The Role of Model for End-Stage Liver Disease Score

Background The Model for End-Stage Liver Disease (MELD) score is currently used as a disease severity index of cirrhotic patients awaiting liver transplantation. This study evaluated the usefulness of the MELD score in predicting mortality and morbidity of patients with hepatocellular carcinoma (HCC) undergoing hepatic resection. Methods The study cohort consisted of 1,017 patients who underwent hepatic resection for HCC between 1991 and 2005. Patient variables were examined by univariate and multivariate analyses to identify risk factors for morbidity and mortality. Accuracy in predicting mortality was assessed with the area under the receiver operator characteristic curve (AUC) analysis. Results The morbidity and mortality rates were 30.7% and 1.9%, respectively. Age, liver cirrhosis, operation time, and MELD score were risk factors for mortality, whereas indocyanine green retention rate at 15-min value, operation time, blood loss, and Child-Turcotte-Pugh score were risk factors for morbidity. Patients with MELD score > 8 had higher mortality (4.0% vs. 0.6%, p = 0.004) and higher liver-related morbidities (16.1% vs. 4.3%, p < 0.001), including massive ascites, intra-abdominal hemorrhage, and hepatic failure, compared with patients with MELD score < 6. High MELD score also was related to longer postoperative hospital stay (score > 8, 14.5 days vs. score < 6, 12.6 days, p = 0.015). The AUC for MELD score as a predictor of mortality was 0.718, indicating high clinical usefulness. Conclusions The MELD score relates with mortality and liver-related morbidities in HCC patients who undergo hepatic resection. A MELD sore > 8 represents the trigger for intensive treatment to improve patient outcome.