术前临床指标对肝细胞癌患者微血管侵犯分级的预测价值

目的 通过对根治性肝切除患者术前临床指标综合分析构建术前预测模型，预测肝细胞癌（HCC）患者是否合并微血管侵犯（MVI），并验证其预测效能。方法 对2017年3月至2022年6月在南京鼓楼医院肝胆外科收治的579例肝切除HCC患者的临床资料进行回顾性研究，根据手术时间顺序分为模型组279例和验证组300例。采用单因素与多因素Logistic回归分析术前临床指标影响MVI分级的独立危险因素，并建立预测评分模型，通过ROC曲线判断MVI的诊断价值，并在验证组中进行独立验证。结果多因素Logistic回归分析显示，肿瘤最大径>5 cm(OR=8.356,95%CI 3.950～17.675,P<0.001)、肿瘤数目为多个（OR=8.652,95%CI 3.213～23.302,P<0.001）、肿瘤包膜强化（OR=4.636,95%CI 2.266～9.483,P<0.001）及AFP>400μg/L(OR=8.938,95%CI 4.182～19.105,P<0.001)为MVI分级的独立危险因素。根据Logistic回归分析结果构建预测模型，ROC曲...

Predictive value of preoperative clinical indicators for microvascular invasion in hepatocellular carcinoma

Objective To establish a preoperative prediction model for microvascular invasion (MVI) in patients with hepatocellular carcinoma (HCC) by comprehensive analysis of preoperative clinical indicators in patients with HCC after radical hepatectomy, and to verify its predictive efficacy. Methods Clinical data of 579 patients with HCC who underwent hepatectomy in Nanjing Drum Tower Hospital from Mar. 2017 to Jun. 2022 were retrospectively analyzed. According to the order of operation, they were divided into model group (279 cases) and validation group (300 cases). Univariate and multivariate Logistic regression were used to analyze the independent risk factors of preoperative clinical indicators affecting MVI grade, and the predictive score model was established. The diagnostic value of MVI was judged by receiver operating characteristic (ROC) curve, and independent verification was performed in the validation group. Results Multivariate Logistic regression analysis showed that tumor diameter > 5 cm (OR=8.356, 95%CI 3.950-17.675), multiple tumor number (OR=8.652, 95%CI 3.213-23.302), capsule enhancement (OR=4.636, 95%CI 2.266-9.483) and AFP > 400 μg/L (OR=8.938, 95%CI 4.182-19.105) were independent risk factors for MVI classification (all P<0.001). The prediction model was constructed according to the results of Logistic regression analysis. The ROC curve analysis showed that the area under the curve (AUC) of the prediction model for predicting MVI in the model group was 0.866, and the sensitivity and specificity were 81.1% and 79.9% when the cut-off value was 4. In the validation group, the AUC of the prediction model to predict MVI was 0. 815, the sensitivity was 84.0%, and the specificity was 79.0%.Conclusion The preoperative prediction model based on Logistic regression analysis has important reference value for preoperative evaluation of MVI in patients with HCC, further guiding surgical plan, and improving the prognosis.