论文部分内容阅读
目的探讨肝细胞癌(HCC)肝切除术后预后预测模型的建立与评价。方法回顾性分析128例手术切除并经病理证实的HCC患者病例资料及随访资料,以单因素分析和多因素Cox比例风险模型进行筛选与评估。将筛选出的独立危险因素纳入二分类Logistic回归分析,建立预后预测模型,受试者工作特征(ROC)曲线对模型进行评价。结果肿瘤分化程度、侵犯血管数目、甲胎蛋白(AFP)水平是影响预后的独立因素,3年生存时间的二分类Logistic回归模型:Ln(P/l-P)=-2.547+1.601×肿瘤分化程度+0.798×侵犯血管数目+0.002×AFP。似然比检验:χ2=42.336,P=0.000。ROC曲线下面积0.924,95%CI:0.872-0.976;最佳截断值0.538,灵敏度90.77%,特异度90.48%。结论通过对HCC肝切除术后影响预后的危险因素分析,可建立高效的预后预测模型。
Objective To investigate the prognosis prediction model of hepatocellular carcinoma (HCC) after hepatectomy. Methods A retrospective analysis of 128 cases of HCC patients with surgical resection and confirmed by pathology and follow-up data, univariate analysis and multivariate Cox proportional hazards model for screening and evaluation. The independent risk factors screened were included in the dichotomous Logistic regression analysis to establish the prognosis prediction model and the receiver operating characteristic (ROC) curve to evaluate the model. Results The degree of tumor differentiation, blood vessel invasion and AFP level were independent prognostic factors. The Logistic regression model of three-year survival time: Ln (P / lP) = -2.547 + 1.601 × tumor differentiation degree + 0.798 × Invasion of blood vessels + 0.002 × AFP. Likelihood ratio test: χ2 = 42.336, P = 0.000. The area under the ROC curve was 0.924, 95% CI: 0.872-0.976; the best cutoff was 0.538, the sensitivity was 90.77% and the specificity was 90.48%. Conclusion Through the analysis of risk factors influencing prognosis after hepatic resection of HCC, we can establish an efficient prognostic model.