腹腔镜肝切除治疗肝内胆管癌的远期预后因素分析及生存列线图的建立

The analysis of long-term prognostic factors after laparoscopic liver resection for intrahepatic cholangiocarcinoma and establishment of survival Nomogram model

目的基于肝内胆管癌(ICC)患者腹腔镜肝切除术后的独立预后因素建立生存预测模型。方法回顾性收集2010年8月至2021年5月国内13家医疗中心连续收治的接受腹腔镜根治性肝切除治疗的351例ICC患者的临床和病理学资料,男性190例,女性161例,年龄[M(IQR)]61(14)岁(范围:23~93岁)。将总队列随机分为训练集(264例)和验证集(87例)。通过门诊或电话对患者进行随访,随访截止时间为2021年10月。基于训练集数据,利用多因素Cox比例风险回归模型筛选ICC患者腹腔镜肝切除术后的独立预后因素,构建列线图模型,绘制受试者工作特征(ROC)曲线,计算曲线下面积(AUC),并通过内部和外部验证评估列线图模型的区分度、校准度和临床获益。结果训练集和验证集患者的临床和病理学特征及远期生存结果的差异均无统计学意义(P值均>0.05)。多因素分析结果提示,CA19-9、CA125、腹腔镜术中中转开腹和淋巴结转移状态是训练集ICC患者腹腔镜肝切除术后的独立预后因素(P值均<0.05)。基于上述筛选得到的独立预后因素建立生存列线图。ROC曲线提示,训练集患者术后的1、3、5年总体生存率的AUC分别为0.794(95%CI:0.721~0.867)、0.728(95%CI:0.618~0.839)、0.799(95%CI:0.670~0.928),验证集患者术后的1、3、5年AUC分别为0.787(95%CI:0.660~0.915)、0.831(95%CI:0.678~0.983)、0.810(95%CI:0.639~0.982)。内部和外部验证结果证明,该模型具有较好的区分度、校准度和临床实用性。结论基于CA19-9、CA125、腹腔镜术中中转开腹和淋巴结转移状态等ICC患者腹腔镜肝切除术后远期预后的独立危险因素构建的生存列线图模型具有较好的区分度、校准度和临床实用性。

Objective To establish a survival prediction model based on the independent prognostic factors of long-term prognosis after laparoscopic liver resection(LLR)for intrahepatic cholangiocarcinoma(ICC).Methods The clinical and pathological data of 351 consecutive patients with ICC who received radical LLR in 13 Chinese medical centers from August 2010 to May 2021 were collected retrospectively.There were 190 males and 161 females,aged(M(IQR))61(14)years(range:23 to 93 years).The total cohort was randomly divided into a training dataset(264 cases)and a validation dataset(87 cases).The patients were followed up by outpatient service or telephone,and the deadline for follow-up was October 2021.Based on the training dataset,the multivariate Cox proportional hazards regression model was used to screen the independent influencing factors of long-term prognosis to construct a Nomogram model.The Nomogram model′s discrimination,calibration,and clinical benefit were evaluated through internal and external validation,and an assessment of the overall value of two groups was made through the use of a receiver operating characteristic(ROC)curve.Results There was no significant difference in clinical and pathological characteristics and long-term survival results between the training and validation datasets(all P>0.05).The multivariate Cox analysis showed that CA19-9,CA125,conversion to laparotomy during laparoscopic surgery,and lymph node metastasis were independent prognostic factors for ICC patients after LLR(all P<0.05).The survival Nomogram was established based on the independent prognostic factors obtained from the above screening.The ROC curve showed that the area under the curve of 1,3 and 5-year overall survival rates of patients in the training dataset were 0.794(95%CI:0.721 to 0.867),0.728(95%CI:0.618 to 0.839)and 0.799(95%CI:0.670 to 0.928),and those in the validation dataset were 0.787(95%CI:0.660 to 0.915),0.831(95%CI:0.678 to 0.983)and 0.810(95%CI:0.639 to 0.982).Internal and external validation proved that the model exhibited a certain discrimination,calibration,and clinical applicability.Conclusion The survival Nomogram model based on the independent influencing factors of long-term prognosis after LLR for ICC(including CA19-9,CA125,conversion to laparotomy during laparoscopic surgery,and lymph node metastasis)exhibites a certain differentiation,calibration,and clinical practicability.