【摘要】 目的 探讨基于术前腔内超声(TRUS)及肿瘤标志物特征的列线图模型预测T3期直肠癌壁外血管侵犯(EMVI)的价值。方法 回顾性分析225例于术前行TRUS检查的直肠癌患者声像图及肿瘤标志物特征，采用单因素及多因素Logistic回归，筛选术前预测EMVI的独立影响因素，构建列线图模型并评价其效能。结果 肿瘤大小、壁外血管内径、壁外血管血流状态、有无淋巴结肿大及血清癌胚抗原(CEA)水平是影响EMVI的独立危险因素。根据多因素分析结果构建列线图模型，受试者工作特征(ROC)曲线下面积为0.885(95%CI:0.835～0.934),校准曲线显示该模型预测直肠癌EMVI的结果与实际结果有较好的一致性，决策曲线显示该模型具有临床价值。结论 基于TRUS检查及肿瘤标志物特征的列线图模型可在术前预测T3期直肠癌患者发生EMVI的个体风险。更多还原

【Abstract】 Objective To investigate the value of a nomogram model based on transrectal ultrasound(TRUS) and tumour marker profiles for the diagnosis of extramural vascular invasion(EMVI) in T3 rectal cancer. Methods The ultrasonographic and tumour marker characteristics of 225 patients with rectal cancer who underwent preoperative TRUS were retrospectively analyzed. Univariate and multivariate Logistic regression was used to screen for independent influences on preoperative prediction of EMVI as a means of constructing a nomogram model and evaluating its efficacy. Results The tumour size, intramural vessel diameter, extramural vessel flow status, presence of lymph node enlargement and carcinoembryonic antigen(CEA) level were independent risk factors for EMVI. The area under the receiver operating characteristic(ROC) curve was 0.885(95%CI:0.835-0.934), based on a nomogram model constructed from the multifactor results. Calibration curves showed good agreement between the model predicting EMVI in rectal cancer and the actual results. The decision curve showed that the model had a significant positive net benefit. Conclusions A nomogram model based on TRUS examination and tumour marker profiles can predict the individual risk of EMVI in T3 rectal cancer patients before surgery.更多还原