【摘要】 目的 旨在结合肌少症及其他临床指标，构建并验证一个慢加急性肝衰竭（ACLF）患者死亡风险的新预测模型，以提高对ACLF患者预后评估的准确性。方法 选取2019年1月—2022年1月于首都医科大学附属北京佑安医院住院的ACLF患者380例，采用分层随机抽样法按照6∶4的比例将其分为训练组（n=228）和测试组（n=152）。在训练组中，通过CT图像测量第三腰椎骨骼肌面积，计算第三腰椎骨骼肌指数（L3-SMI）。肌少症的诊断依据前期多中心研究建立的中国北方正常成年人L3-SMI参考值。采用单因素和多因素Cox回归分析，构建结合肌少症及临床风险因素的“肌少症-ACLF模型”，并通过列线图展示。采用受试者操作特征曲线下面积（AUC）评估模型的预测效能，使用校准曲线评估模型的校准度，使用决策曲线分析（DCA）评估其临床应用价值。计量资料两组间比较采用成组t检验或Mann-Whitney U检验。计数资料两组间比较采用χ~2检验。采用Kaplan-Meier方法绘制生存曲线，组间比较使用Log-rank检验。不同模型间AUC的差异比较采用De Long检验。结果 根据多因素Cox回归分析结果，将肌少症（HR=1.962,95%CI:1.185～3.250,P=0.009）、总胆红素（HR=1.003,95%CI:1.002～1.005,P<0.001）、国际标准化比值（HR=1.997,95%CI:1.674～2.382,P<0.001）和乳酸（HR=1.382,95%CI:1.170～1.632,P<0.001）纳入肌少症-ACLF模型。训练队列中，肌少症-ACLF模型预测ACLF患者90天死亡风险的AUC为0.80，较MELD-Na评分的AUC(0.73)有所提高（Z=1.97,P=0.049）。测试队列中，肌少症-ACLF模型的AUC为0.79，显著高于MELD评分（AUC=0.69）(Z=2.70,P=0.007)和MELD-Na评分（AUC=0.68）(Z=2.92,P=0.004)。校准曲线显示该模型具有良好的校准能力，预测的死亡风险与实际观察结果之间一致性较好。DCA结果显示，在一定的阈值概率范围内，训练队列和测试队列中的肌少症-ACLF模型均表现出较MELD评分和MELD-Na评分更高的净收益。结论 本研究开发的肌少症-ACLF模型为预测ACLF患者90天死亡风险提供了更准确的工具，可支持临床决策和优化治疗策略。更多还原

【Abstract】 Objective To establish and validate a new prediction model for the risk of death in patients with acute-on-chronic liver failure(ACLF) based on sarcopenia and other clinical indicators, and to improve the accuracy of prognostic assessment for ACLF patients. Methods A total of 380 patients with ACLF who were admitted to Beijing YouAn Hospital, Capital Medical University, from January 2019 to January 2022 were enrolled, and they were divided into training group with 228 patients and testing group with 152 patients in a ratio of 6∶4 using the stratified random sampling method. For the training group, CT images were used to measure the cross-sectional area of the skeletal muscle at the third lumbar vertebra(L3), and L3 skeletal muscle index(L3-SMI) was calculated. Sarcopenia was diagnosed based on the previously established L3-SMI reference values for healthy adults in northern China. Univariate and multivariable Cox regression analyses were used to establish a sarcopenia-ACLF model which integrated sarcopenia and clinical risk factors, and a nomogram was developed for presentation. The area under the ROC curve(AUC) was used to assess the predictive performance of the model, the calibration curve was used to assess the degree of calibration, and a decision curve analysis was used to investigate the clinical application value of the model. The independentsamples t test or the Mann-Whitney U test was used for comparison of continuous data between two groups, and the chi-square test was used for comparison of categorical data between two groups. The Kaplan-Meier method was used to plot survival curves, and the Log-rank test was used for comparison between groups. The DeLong test was used for comparison of AUC between different models.Results The multivariate Cox regression analysis showed that sarcopenia(hazard ratio [HR]=1.962, 95% confidence interval [CI]: 1.185 — 3.250, P=0.009), total bilirubin(HR=1.003, 95%CI: 1.002 — 1.005, P<0.001), international normalized ratio(HR=1.997, 95%CI: 1.674 — 2.382, P<0.001), and lactic acid(HR=1.382, 95%CI: 1.170 — 1.632, P<0.001) were included in the sarcopenia-ACLF model. In the training cohort, the sarcopenia-ACLF model had a larger AUC than MELD-Na score in predicting 90-day mortality in patients with ACLF(0.80 vs 0.73, Z=1.97, P=0.049). In the test cohort, the sarcopenia-ACLF model had a significantly larger AUC than MELD score(0.79 vs 0.69, Z=2.70, P=0.007) and MELD-Na score(0.79 vs 0.68, Z=2.92, P=0.004). The calibration curve showed that the model had good calibration ability, with a relatively good consistency between the predicted risk of mortality and the observed results. The DCA results showed that within a reasonable range of threshold probabilities, the sarcopenia-ACLF model showed a greater net benefit than MELD and MELD-Na scores in both the training cohort and the test cohort. Conclusion The sarcopenia-ACLF model developed in this study provides a more accurate tool for predicting the risk of 90-day mortality in ACLF patients, which provides support for clinical decision-making and helps to optimize treatment strategies.更多还原