肝细胞癌术后复发风险因素及预测模型分析

Analysis of Risk Factors and Prediction Models for Postoperative Recurrence of Hepatocellular Carcinoma

摘要: 目的：探讨肝细胞癌(HCC)患者术后复发的临床及影像学风险因素，并构建预测模型。方法：回顾性分析143例经病理确诊并接受根治性切除的HCC患者，根据术后3个月CT复查结果分为复发组(84例)和未复发组(59例)。通过组间差异分析及Logistic回归筛选术后复发的独立风险因素，并构建多因素预测模型，采用ROC曲线评估模型效能。结果：微血管浸润、病灶数目及浸润程度为HCC术后复发的独立预测因素(P < 0.05)。ROC分析显示，去除病灶数目的对比模型相比全因素模型，AUC值具有最大的下降，由0.796下降至0.765，表明病灶数目在预测HCC术后复发中具有重要贡献。结论：微血管浸润、病灶数目及浸润程度是HCC术后复发的关键独立预测因素，其中病灶数目与浸润程度对模型预测效能的提升作用更为显著。基于三项独立变量建立的多因素模型可较准确地预测HCC术后复发风险，为术后个体化管理提供重要参考。

Abstract: Objective: To investigate the clinical and imaging risk factors for postoperative recurrence in patients with hepatocellular carcinoma (HCC) and to construct a predictive model. Methods: A retrospective analysis was conducted on 143 patients with pathologically confirmed HCC who underwent curative resection. Based on CT re-examination at 3 months post-surgery, patients were classified into a recurrence group (n = 84) and a non-recurrence group (n = 59). Intergroup comparisons and logistic regression analyses were performed to identify independent risk factors for postoperative recurrence, and a multivariate predictive model was established. The model’s performance was evaluated using receiver operating characteristic (ROC) curve analysis. Results: Microvascular invasion, number of lesions, and degree of invasion were identified as independent predictors of postoperative recurrence (P < 0.05). ROC analysis showed that among the comparison models, the model excluding the number of lesions exhibited the largest decrease in AUC, declining from 0.796 in the full-factor model to 0.765, indicating that the number of lesions makes an important contribution to predicting postoperative recurrence in HCC. Conclusion: Microvascular invasion, number of lesions, and degree of invasion are key independent predictors of postoperative recurrence in HCC, with the number of lesions and degree of invasion contributing most significantly to the predictive performance of the model. A multivariate model based on these three factors can accurately predict postoperative recurrence risk and provide valuable guidance for individualized postoperative management.