术前肝功能储备预测肝癌手术患者预后价值的倾向性匹配研究

目的:运用倾向性评分匹配法(PSM)探讨术前肝功能储备对肝癌术后生存情况的预测价值。方法:绘制术前肝功能储备预测肝癌患者术后生存价值的ROC曲线，获得术前肝功能储备的最佳诊断截点,根据术前肝功能储备的最佳诊断截点将所有患者分为两组,运用1∶1倾向性匹配评分(PSM)法将Logistic模型估计的倾向性评分相近的两组患者进行配对,检验匹配前后两组间各临床病理指标的均衡性,Kaplan-Meier生存分析比较匹配后两组患者术后无瘤生存率和总生存率,Cox回归模型进行敏感性分析,验证匹配后术前肝功能储备对肝癌术后生存情况的预测价值。结果:研究共纳入207例肝癌患者,平均年龄(54.1±15.4)岁,其中男性132例(63.8%),女性75例(36.2%),术后1年、3年、5年无瘤生存率和总生存率分别为75.3%、45.9%、37.1%和84.7%、57.6%、48.2%,术前肝功能储备预测肝癌患者术后生存价值的AUC为0.821(95%CI:0.763～0.894),最佳诊断截点为14.13,相应的灵敏度和特异度分别为84.5%和81.2%,根据术前肝储备功能的最佳诊断截点,将患者分别分为肝储备功能良好组134例(64.7%)和肝储备功能不全组73例(35.3%),匹配后肝储备功能不全组患者术后出血和肝衰竭的发生率明显高于肝功能储备良好组,Kaplan-Meier生存分析显示,肝储备功能良好组术后1年、3年、5年总生存率明显高于肝储备功能不全组(91.4%、82.5%、74.7% vs 83.2%、52.7%、43.6%,P＜0.05),Cox多因素分析显示,肝癌患者术前肝储备功能每降低1%,患者术后5年内死亡的风险增加1.124倍(OR:2.124,95%CI:1.278～2.875,P=0.028)。结论:术前肝储备功能可用于准确评估肝癌患者术后肝衰风险以及远期总生存率。

Propensity matching study of preoperative liver function reserve in predicting prognostic value of hepatocellular carcinoma patients underwent surgery

Objective:To explore the predictive value of preoperative liver function reserve for postoperative survival of hepatocellular carcinoma (HCC) patients by propensity score matching (PSM).Methods:The ROC curve of preoperative liver function reserve to predict the postoperative survival value of HCC patients was drawn,and the best cut-off point of preoperative liver function reserve was obtained.According to the best cut-off point of preoperative liver function reserve,all patients were divided into two groups.The patients between the two groups with similar propensity score estimated by Logistic model were matched by 1∶1 propensity matching score (PSM) method.To test the balance of clinical and pathological indexes between the two groups before and after matching.Kaplan Meier survival analysis was used to compare the postoperative disease-free survival rate and overall survival rate of postoperative HCC patients between the two groups.Cox regression model was used for sensitivity analysis to verify the predictive value of preoperative liver function reserve on survival of postoperative HCC patients.Results:A total of 207 patients with HCC were included in the study.The average age was (54.1±15.4) years,including 132 males (63.8%) and 75 females (36.2%).The 1-year,3-year,5-year disease-free survival rates and overall survival rates were 75.3%,45.9%,37.1% and 84.7%,57.6%,48.2% respectively.The AUC of preoperative liver function reserve for predicting postoperative survival value of HCC patients was 0.821 (95%CI:0.763~0.894),the optimal cut-off point was 14.13,and the corresponding sensitivity and specificity were 84.5% and 81.2% respectively.According to the optimal cut-off point of preoperative liver reserve function,the patients were divided into two groups:134 cases (64.7%) with good liver reserve function and 73 cases (35.3%) with liver reserve function insufficiency,after matching,the incidence of postoperative bleeding and liver failure of patients in liver reserve insufficiency group was significantly higher than that in good liver reserve group.Kaplan Meier survival analysis showed that the 1-year,3-year and 5-year overall survival rates of postoperative HCC patients in good liver reserve were significantly higher than those in poor liver reserve (91.4%,82.5%,74.7% vs 83.2%,52.7%,43.6%,P＜0.05).Cox multivariate analysis showed that for every 1% decrease in liver reserve function before operation,the risk of death of HCC patients within 5 years after operation increased by 1.124 times (OR:2.124,95%CI:1.278~2.875,P=0.028).Conclusion:Preoperative liver reserve function can be used to accurately assess the risk of postoperative liver failure and long-term overall survival rate of HCC patients.