Wang Xiang, Liu Li, Liu Zhi-Peng, Wang Jiao-Yang, Dai Hai-Su, Ou Xia, Zhang Cheng-Cheng, Yu Ting, Liu Xing-Chao, Pang Shu-Jie, Fan Hai-Ning, Bai Jie, Jiang Yan, Zhang Yan-Qi, Wang Zi-Ran, Chen Zhi-Yu, Li Ai-Guo
Department of Hepatobiliary Surgery, Southwest Hospital, Third Military Medical University (Army Medical University), Chongqing, China.
Department of Digital Medicine, School of Biomedical Engineering and Medical Imaging, Third Military Medical University (Army Medical University), Chongqing, China.
J Gastrointest Surg. 2024 Dec;28(12):2039-2047. doi: 10.1016/j.gassur.2024.09.027. Epub 2024 Oct 3.
Early recurrence is the leading cause of death for patients with perihilar cholangiocarcinoma (pCCA) after surgery. Identifying high-risk patients preoperatively is important. This study aimed to construct a preoperative prediction model for the early recurrence of patients with pCCA to facilitate planned treatment with curative resection.
This study ultimately enrolled 400 patients with pCCA after curative resection in 5 hospitals between 2013 and 2019. They were randomly divided into training (n = 300) and testing groups (n = 100) at a ratio of 3:1. Associated variables were identified via least absolute shrinkage and selection operator (LASSO) regression. Four machine learning models were constructed: support vector machine, random forest (RF), logistic regression, and K-nearest neighbors. The predictive ability of the models was evaluated via receiving operating characteristic (ROC) curves, precision-recall curve (PRC) curves, and decision curve analysis. Kaplan-Meier (K-M) survival curves were drawn for the high-/low-risk population.
Five factors: carbohydrate antigen 19-9, tumor size, total bilirubin, hepatic artery invasion, and portal vein invasion, were selected by LASSO regression. In both the training and testing groups, the ROC curve (area under the curve: 0.983 vs 0.952) and the PRC (0.981 vs 0.939) showed that RF was the best. The cutoff value for distinguishing high- and low-risk patients was 0.51. K-M survival curves revealed that in both groups, there was a significant difference in RFS between high- and low-risk patients (P < .001).
This study used preoperative variables from a large, multicenter database to construct a machine learning model that could effectively predict the early recurrence of pCCA in patients to facilitate planned treatment with curative resection and help clinicians make better treatment decisions.
早期复发是肝门部胆管癌（pCCA）患者术后死亡的主要原因。术前识别高危患者很重要。本研究旨在构建pCCA患者早期复发的术前预测模型，以促进根治性切除的计划性治疗。
本研究最终纳入了2013年至2019年间在5家医院接受根治性切除的400例pCCA患者。他们以3:1的比例随机分为训练组（n = 300）和测试组（n = 100）。通过最小绝对收缩和选择算子（LASSO）回归确定相关变量。构建了四种机器学习模型：支持向量机、随机森林（RF）、逻辑回归和K近邻。通过接受操作特征（ROC）曲线、精确召回率曲线（PRC）和决策曲线分析评估模型的预测能力。为高/低风险人群绘制了Kaplan-Meier（K-M）生存曲线。
通过LASSO回归选择了五个因素：糖类抗原19-9、肿瘤大小、总胆红素、肝动脉侵犯和门静脉侵犯。在训练组和测试组中，ROC曲线（曲线下面积：0.983对0.952）和PRC（0.981对0.939）均显示RF是最佳模型。区分高风险和低风险患者的临界值为0.51。K-M生存曲线显示，在两组中，高风险和低风险患者之间无复发生存期（RFS）存在显著差异（P <.001）。
本研究使用来自大型多中心数据库的术前变量构建了一个机器学习模型，该模型可以有效预测pCCA患者的早期复发，以促进根治性切除的计划性治疗，并帮助临床医生做出更好的治疗决策。
