放射学实践2024,Vol.39Issue(10) :1367-1373.DOI:10.13609/j.cnki.1000-0313.2024.10.014

钆塞酸二钠增强MRI联合血管包绕肿瘤细胞簇预测肝细胞癌术后早期复发的价值

Preoperative gadoxetic acid-enhanced MRI combine with vessels encapsulating tumor clusters in predicting early recurrence of hepatocellular carcinoma

张继云 张涛 张学琴 刘茂童 鲁梦恬 瞿琦
放射学实践2024,Vol.39Issue(10) :1367-1373.DOI:10.13609/j.cnki.1000-0313.2024.10.014

钆塞酸二钠增强MRI联合血管包绕肿瘤细胞簇预测肝细胞癌术后早期复发的价值

Preoperative gadoxetic acid-enhanced MRI combine with vessels encapsulating tumor clusters in predicting early recurrence of hepatocellular carcinoma

张继云 1张涛 1张学琴 1刘茂童 1鲁梦恬 1瞿琦1
扫码查看

作者信息

  • 1. 226000 江苏,南通大学附属南通第三医院、南通市第三人民医院影像科
  • 折叠

摘要

目的:本研究主要是探讨钆塞酸二钠增强MRI联合血管包绕肿瘤细胞簇(VETC)预测肝细胞癌(HCC)术后早期复发的价值.方法:回顾性分析2013年1月~2022年1月间在本院行肝部分切除术并经病理和免疫组织化学检测诊断为HCC患者234例.其中术后早期复发患者90例,非早期复发患者144例,记录所有入组患者的术前实验室相关检查指标、VETC状态以及影像学特征.通过COX单因素及多因素回归分析分别确定术前影像学征象联合VETC(with VETC)及仅术前影像学征象(without VETC)预测HCC术后早期复发的危险因素,根据每个危险因素的回归系数β建立风险评分表,依据风险评分模型的早期复发预测值与实际早期复发的观测值一致性采用Hosmer-Lemeshow拟合优度检验.通过绘制受试者工作特征(ROC)曲线检验模型的诊断效能.结果:术前钆塞酸二钠增强MRI联合VETC状态预测HCC术后早期复发的多因素分析结果显示VETC(HR:2.385;95%CI:1.518~3.746)、肿瘤直径(HR:1.089;95%CI:1.003~1.183)、动脉期非环形高强化(HR:1.978;95%CI:1.265~3.093)是预测HCC术后早期复发的独立危险因素.仅术前钆塞酸二钠增强MRI预测HCC术后早期复发的多因素分析结果显示肿瘤直径(HR:1.012;95%CI:1.001~1.284)、动脉期非环形高强化(HR:2.032;95%CI:1.292~3.198)、坏死或缺血严重(HR:1.530;95%CI:1.339~1.828)是预测HCC术后早期复发的独立危险因素.建立两个风险评分模型:风险评分模型(with VETC)和风险评分模型(without VETC),Hosmer-Lemeshow拟合优度检验结果显示:风险评分模型(with VETC)x2=4.964,P=0.761,风险评分模型(without VETC)x2=4.776,P=0.092,表明两个模型预测 HCC术后早期复发一致性较好.两个模型的ROC曲线下面积分别为AUCwith VETC=0.801,AUCwithout VETC=0.747,Z=2.783,P=0.025.两个模型的最佳截断值分别为 cutoffwith VETC=7,cutoffwithout VETC=5,患者的风险评分≥cutoff值为术后早期复发高风险.结论:影像学特征肿瘤直径、动脉期非环形高强化及VETC状态是HCC术后早期复发的独立危险因素,以此所构建的HCC术后早期复发风险评分模型诊断效能优于仅通过术前影像学特征所构建的风险评分模型.

Abstract

Objective:The aim of this study was to assess the effectiveness of combining gadox-etic acid-enhanced magnetic resonance imaging with vessels encapsulating tumor clusters(VETC)in predicting early recurrence of hepatocellular carcinoma(HCC)after surgery.Methods:We conducted a retrospective analysis from January 2013 to January 2022 at the Third People's Hospital of Nantong,wherein 234 patients with hepatocellular carcinoma(HCC)diagnosed by pathology and immunohisto-chemistry after hepatectomy were identified.We recorded preoperative laboratory-related examination indicators,pathological findings,and imaging features of all enrolled patients.Of these,90 patients ex-perienced early recurrence,and the remaining 144 patients did not exhibit early recurrence.We recor-ded the preoperative laboratory-related examination indicators,VETC status,and imaging features of all enrolled patients.We analyzed the data using univariate and multivariate COX regression analysis to identify risk factors for predicting early postoperative recurrence of HCC with preoperative imaging factors combined with VETC and only preoperative imaging factors.We established a risk index based on the regression coefficient(β)for each risk factor.We used the Hosmer Lemeshow goodness of fit test to assess the consistency between the risk-scoring model predicted early recurrence status and the true early recurrence status.Furthermore,we plotted a receiver operating characteristic(ROC)curve to verify the diagnostic efficiency of the models.Results:The multivariate COX regression analysis of preoperative MRI combined with VETC status to predict early recurrence of HCC revealed that VETC(HR:2.385;95%CI:1.518-3.746),tumor diameter(HR:1.089;95%CI:1.003~1.183),and nonrim arterial phase hyperenhancement(HR:1.978;95%CI:1.265~3.093)were significant independent risk factors.In comparison,the multivariate COX regression analysis of only preoperative MRI high-lighted that tumor diameter(HR:1.012;95%CI:1.001~1.284),nonrim arterial phase hyperenhance-ment(HR:2.032;95%CI:1.2 92~3.198),and necrosis or severe ischemia(HR:1.530;95%CI:1.339~1.828)were independent risk factors for early recurrence of HCC.We developed two risk scoring models:the risk scoring model(with VETC)and the risk scoring model(without VETC).The Hos-mer-Lemeshow goodness of fit test revealed that both models predicted early recurrence of HCC with good consistency(risk score model(with VETC)x2=4.964,P=0.761;risk score model(without VETC)x2=4.776,P=0.092).The areas under curves for the two models were AUC with VETC=0.801 and AUC without VETC=0.747(Z=2.783,P=0.025).The optimal cutoff value for the risk score model with VETC was cutoff with VETC=7 while that for the risk score model without VETC was cutoff without VETC=5.Patients with risk scores ≥cutoff are at a high risk of early recurrence after surgery.Conclusion:The imaging features such as tumor diameter,nonrim arterial phase hyperen-hancement,and VETC status are significant independent risk factors for early recurrence after hepato-cellular carcinoma(HCC)surgery.The risk score model for predicting early recurrence after HCC sur-gery,constructed based on these factors,has better diagnostic efficacy than the risk score model built solely on preoperative imaging features.

关键词

癌,肝细胞/复发/磁共振成像

Key words

Carcinoma,hepatocellular/Recurrence/Magnetic resonance imaging

引用本文复制引用

基金项目

江苏省卫生健康委2023年医学科研立项项目(Z2023091)

江苏省研究型医院学会感染影像科研专项科技项目(GY202203)

南通市科技计划项目(MS22022056)

出版年

2024
放射学实践
华中科技大学同济医学院

放射学实践

CSTPCDCSCD北大核心
影响因子:1.08
ISSN:1000-0313
参考文献量25
段落导航相关论文