基于CT测定的细胞外体积分数对结肠腺癌术前病理分级的预测价值

Predictive value of the extracellular volume fraction based on CT in the preoperative pathological grading of colon adenocarcinoma

目的:探讨CT测定的细胞外体积分数(ECV)预测结肠腺癌病理分级可行性。方法:选取经手术病理证实的结肠腺癌患者380例,术前均行腹部多期动态CT增强扫描。结肠癌病灶的ECV通过平扫和平衡期图像在ROI测量计算。根据病理分级将患者分为高级别组100例和低级别组280例,比较2组影像特征及临床特征。ECV与结肠腺癌病理分级的相关性行二元logistic回归分析。采用ROC曲线评价ECV对结肠腺癌病理分级的诊断效能。结果:多因素logistic回归分析显示,ECV是结肠腺癌病理分级的独立预测因子,OR值为1.380(95%CI1.289~1.477)。ROC曲线显示:ECV对结肠腺癌病理分级的诊断效能较好,AUC为0.898(95%CI0.851~0.947),敏感度为90.0%,特异度为91.4%。结论:基于CT测定的ECV有助于预测结肠腺癌的病理分级,且对个体化治疗具有一定意义。

Objective:To investigate the feasibility of the extracellular volume fraction(ECV)determined using CT to predict the pathologic grade of colon adenocarcinoma.Methods:A total of 380 patients with pathologically proven colon adenocarcinoma were selected.Pre-operative multi-phase dynamic enhanced CT scans were conducted for all patients.The ECVs of colon cancer lesions were calculated using ROI measurements on unenhanced and equilibrium-phase CT images.According to pathological grade,the patients were divided into a high-grade group(100 patients)and a low-grade group(280 patients).The imaging features and clinical data between the two groups were compared.The correlation between ECV and the pathological grade of colon adenocarcinoma was analyzed with binary logistic regression.ROC curve was used to evaluate the diagnostic efficiency.Results:The multi-factor logistic regression analysis showed that ECV was an independent predictor for the pathological grade of colon adenocarcinoma,with OR of 1.380(95%CI 1.289~1.477).The ROC curve demonstrated that ECV had better diagnostic accuracy in predicting the pathologic grade of colon adenocarcinoma,with the AUC of 0.898(95%CI 0.851~0.947),the sensitivity of 90.0%and the specificity of 91.4%.Conclusion:The ECV determined using enhanced CT is helpful for predicting the pathological grade of colon adenocarcinoma and may contribute to personalized treatment.