Huo T-I, Lin H-C, Hsia C-Y, Huang Y-H, Wu J-C, Chiang J-H, Chiou Y-Y, Lui W-Y, Lee P-C, Lee S-D
Department of Medicine, Taipei Veterans General Hospital, National Yang-Ming University School of Medicine, Taipei, Taiwan.
Dig Liver Dis. 2008 Nov;40(11):882-9. doi: 10.1016/j.dld.2008.01.015. Epub 2008 Mar 12.
Serum sodium has been suggested to incorporate into the model for end-stage liver disease to enhance its prognostic ability for cirrhosis. A mathematical equation based on model for end-stage liver disease and sodium, known as "MELD-Na", was developed for outcome prediction for cirrhosis. The severity of liver cirrhosis is a key component to predict survival in patients with hepatocellular carcinoma. This study investigated the prognostic role of MELD-Na for hepatocellular carcinoma.
A total of 535 unselected hepatocellular carcinoma patients were prospectively enrolled to evaluate the performance of MELD-Na.
The MELD-Na was better than model for end-stage liver disease in predicting 6-month mortality by comparing the area under receiver operating characteristic curve (0.782 vs. 0.761, p=0.101). MELD-Na, but not model for end-stage liver disease, was an independent predictor associated with 6-month mortality in multivariate logistic regression analysis (odds ratio: 1.14, p=0.001). In the survival analysis, MELD-Na also independently predicted mortality, with an additional risk of 4.3% per unit increment of the score (p<0.001). Patients with MELD-Na scores between 10 and 20 and scores >20 had 2.1-fold (p<0.001) and 7.5-fold (p<0.001) risk of mortality, respectively, compared to patients with a score <10 in the Cox proportional hazard model.
The MELD-Na score is a feasible and independent prognostic predictor for both short- and long-term outcome predictions in patients with hepatocellular carcinoma.
血清钠已被建议纳入终末期肝病模型，以增强其对肝硬化的预后预测能力。基于终末期肝病模型和钠的数学方程，即“MELD-Na”，被开发用于肝硬化的预后预测。肝硬化的严重程度是预测肝细胞癌患者生存的关键因素。本研究调查了MELD-Na对肝细胞癌的预后作用。
前瞻性纳入535例未经选择的肝细胞癌患者，以评估MELD-Na的性能。
通过比较受试者工作特征曲线下面积，MELD-Na在预测6个月死亡率方面优于终末期肝病模型（0.782对0.761，p=0.101）。在多因素逻辑回归分析中，MELD-Na而非终末期肝病模型是与6个月死亡率相关的独立预测因子（比值比：1.14，p=0.001）。在生存分析中，MELD-Na也独立预测死亡率，评分每增加一个单位，额外死亡风险为4.3%（p<0.001）。在Cox比例风险模型中，与MELD-Na评分<10的患者相比，评分在10至20之间以及评分>20的患者死亡风险分别为2.1倍（p<0.001）和7.5倍（p<0.001）。
MELD-Na评分是肝细胞癌患者短期和长期预后预测的可行且独立的预后预测指标。
