小肝癌微血管侵犯风险术前列线图预测模型的构建与验证

Construction and validation of a preoperative nomogram prediction model for microvascular invasion in small hepatocellular carcinoma

目的分析小肝癌(SHCC)发生微血管侵犯(MVI)的危险因素,构建术前列线图预测模型并进行验证。方法回顾性分析2018年8月至2023年8月因SHCC于华中科技大学同济医学院附属协和医院行根治性肝切除术的288例患者的临床资料,根据术后病理结果是否存在MVI分为MVI阳性组(n=96)和MVI阴性组(n=192)。收集患者一般资料、血清学指标、炎症指标、术前影像学资料、病理指标等,采用Lasso回归分析结合单因素、多因素Logistic回归分析探究SHCC患者发生MVI的主要危险因素,并构建术前列线图预测模型。结果288例单发SHCC患者中有96例发生MVI,发生率为33.3%。多因素Logistic回归分析结果显示,肿瘤最大径、包膜强化、AFP≥200 ng/mL、中性粒细胞/淋巴细胞比值(NLR)≥1.63、全身免疫炎症指数(SII)≥170.86、血小板计数(PLT)≥183.0×10^(9)/L为SHCC患者发生MVI的独立危险因素(P<0.05)。列线图预测模型的曲线下面积(AUC)为0.823,通过Youden指数计算出列线图的最佳截断值为181.3分,截断值下的灵敏度、特异度、阳性预测值和阴性预测值分别为65.6%、84.9%、68.5%和83.2%,具备较好的区分度和一致性。结论基于肿瘤最大径、包膜强化、AFP≥200 ng/mL、NLR≥1.63、SII≥170.86、PLT≥183.0×10^(9)/L所建立的术前列线图预测模型对SHCC发生MVI风险有较高的预测价值,对指导临床决策,改善不良预后具有重要意义。

Objective To investigate the risk factors for microvascular invasion(MVI)in small hepatocellular carcinoma(SHCC)and establish a preoperative nomogram prediction model,and assess the accuracy of it.Methods A retrospective analysis was conducted on the clinical data of 288 patients who underwent hepatectomy due to SHCC at the Affiliated Union Hospital of Tongji Medical College,Huazhong University of Science and Technology from Aug.2018 to Aug.2023.Based on the postoperative pathological results,they were divided into MVI positive group(n=96)and MVI negative group(n=192).The general data,serological indicators,inflammatory factors,preoperative imaging data and pathological indicators were collected.Lasso regression analysis combined with univariate and multivariate logistic regression analysis were used to explore the main risk factors for MVI in SHCC patients,and establish a preoperative nomogram prediction model.Results Out of 288 SHCC patients,MVI was confirmed in 96(33.3%).The results of multivariate logistic regression analysis showed that the maximum diameter of the tumor,capsule enhancement,AFP≥200 ng/mL,neutrophil/lymphocyte ratio(NLR)≥1.63,and systemic immune inflammation(SII)≥170.86,platelet count(PLT)≥183.0×10^(9)/L were independent risk factors for MVI in SHCC patients(P<0.05).The area under the curve(AUC)of the preoperative nomogram prediction model was 0.823.The optimal cutoff value of the column chart calculated by the Youden index was 181.3 points.The sensitivity,specificity,positive predictive value,and negative predictive value under the cutoff value were 65.6%,84.9%,68.5%,and 83.2%,which showed good discrimination and consistency.Conclusion The preoperative nomogram prediction model based on tumor maximum diameter,capsule enhancement,AFP≥200 ng/mL,NLR≥1.63,SII≥170.86 and PLT≥183.0×10^(9)/L has high predictive value for the risk of MVI in SHCC,and is of great significance for guiding clinical decision-making and improving poor prognosis.