基于双能量CT增强图像的列线图预测早期声门型喉癌术后复发

目的探讨基于双能量CT(DECT)增强图像的列线图预测早期声门型喉癌(EGC)术后无复发生存期的价值。方法回顾性分析2015年1月至2018年7月天津市第一中心医院经手术病理证实的EGC患者的临床病理和DECT资料。共入组178例患者,男162例、女16例,年龄44~86(62±9)岁。对所有患者进行随访,根据随访资料,分为复发组(n=32)和无复发组(n=146)。2组间临床病理资料和碘图定量参数虚拟平扫CT值、碘浓度及标准化碘浓度(SIC)的比较采用χ²检验、独立样本t检验或Mann-Whitney U检验。用X-tile软件进行2组间定量资料生存相关截断值的筛选。采用Kaplan-Meier法绘制无复发生存曲线,使用log-rank法检验生存率差异。将有差异的变量纳入Cox比例风险回归模型筛选术后无复发生存期的独立影响因素,基于多因素Cox分析,绘制预测1、2和5年无复发生存率的列线图。用C指数、校准曲线和决策曲线评估联合DECT参数及T分期的列线图模型的预测效能及临床有效性。结果患者随访时间2~63个月,中位随访时间为24.3个月。复发组与无复发组间肿瘤T分期差异有统计学意义(χ²=9.21,P=0.002)。X-tile软件获得相关定量参数预后截断值为SIC动脉期=0.28、SIC静脉期=0.87。log-rank检验结果显示不同T分期、SIC动脉期、SIC静脉期患者的无复发生存期差异有统计学意义(χ²=10.74、15.50、17.97,P=0.001、<0.001、<0.001)。Cox多因素分析显示,T分期、SIC动脉期、SIC静脉期为术后无复发生存期的独立影响因素(风险比为2.271、3.552、3.266,P=0.026、<0.001、0.003)。列线图模型的C指数为0.785,高于单独使用T分期模型的C指数(0.622)。校准曲线显示样本的实际和预测概率之间具有较好的一致性。决策曲线分析表明,列线图模型比T分期模型的临床获益高。结论基于术前T分期及DECT碘图定量参数SIC动脉期和SIC静脉期构建的列线图可以有效预测EGC患者术后无复发生存期。

A nomogram based on dual-energy CT enhanced imaging for prediction of postoperative recurrence of early-stage glottic carcinoma

Objective To explore the value of nomogram based on dual-energy CT(DECT)enhanced imaging in predicting postoperative recurrence-free survival(RFS)of early-stage glottic carcinoma(EGC).Methods The clinicopathological and DECT data of patients with EGC confirmed by pathology in the Tianjin First Central Hospital from January 2015 to July 2018 were analyzed retrospectively.A total of 178 patients were enrolled,including 162 males and 16 females,with the age from 44 to 86(62±9)years old.According to the follow-up data,the patients were divided into recurrent group(n=32)and non-recurrent group(n=146).The differences of clinicopathological data and DECT iodine maps parameters between the two groups were analyzed usingχ² test,independent-sample t test and Mann-Whitney U test.The survival related cut-off values of the quantitative data between the two groups were selected by X-tile software.The survival curve was drawn using Kaplan-Meier method,and the difference of survival rate was tested with log-rank analysis.The variables with statistical differences were included in the Cox proportional hazard model for multivariate analysis to select the independent predictors of postoperative RFS.Based on the multivariate Cox analysis,the nomogram was drawn to predict the RFS at 1,2 and 5 years.The prediction efficiency and clinical benefit of the nomogram were evaluated by C-index,calibration curve and decision curve analysis.Results The median follow-up time was 24.3 months,ranging from 2 to 63 months.There was a significant difference in T-stage between recurrent and non-recurrent groups(χ²=9.21,P=0.002).The prognostic cutoff values obtained by X-tile software were arterial phase standardized iodine concentration(SICAP)=0.28 and venous phase standardized iodine concentration(SICVP)=0.87.The results of log-rank test showed that there were significant differences in RFS among patients with different T-stage,SICAP and SICVP(χ²=10.74,15.50,17.97,P=0.001,<0.001,<0.001).T-stage,SICAP and SICVP were identified as independent predictors of postoperative RFS(hazard ratio=2.271,3.552,3.266,P=0.026,<0.001,0.003).The C-index of the nomogram combined with DECT parameter and T-stage was 0.785,which was higher than that of T-stage alone(0.622).The calibration curve showed that there was good consistency between the actual and predicted probability of the sample.The decision curve analysis showed that the clinical benefit of the nomogram was higher than that of the T-stage alone.Conclusion The nomogram based on preoperative clinical factors(T-stage)and DECT iodine map factors(SICAP and SICvp)can predict the postoperative RFS of patients with EGC.