基于多因素分析的直肠癌术后低位前切除 综合征(LARS)预测列线图模型的构建与验证

Development and Validation of a Nomogram for Predicting Low Anterior Resection Syndrome (LARS) after Rectal Cancer Surgery Based on Multivariable Analysis

摘要: 目的：探讨直肠癌术后发生低位前切除综合征(LARS)的独立影响因素，并构建个体化预测列线图模型。方法：回顾性收集2019年10月至2024年9月某院506例行直肠癌低位前切除术患者的临床资料。根据LARS评分分为LARS组(243例)和非LARS组(263例)。通过单因素及多因素Logistic回归分析筛选LARS的独立影响因素。将数据按7:3随机分为训练集(n = 354)与验证集(n = 152)，基于独立影响因素构建列线图预测模型，并采用ROC曲线、校准曲线及决策曲线分析(DCA)对模型的区分度、校准度和临床实用性进行内部验证。结果：多因素分析显示，吻合口瘘(OR = 9.025)、造口还纳时间延长(OR = 1.260)、合并糖尿病(OR = 2.281)是LARS发生的独立危险因素，而肿瘤下缘距肛缘距离(OR = 0.446)及肿瘤下切缘距肛缘距离(OR = 0.439)增大是保护因素。基于此构建的列线图模型在训练集中的AUC为0.83，验证集中为0.78。校准曲线显示预测概率与实际概率具有良好的一致性。决策曲线分析表明该模型在较宽的阈值概率范围内具有临床净获益。结论：本研究构建的列线图模型能有效预测直肠癌患者术后发生LARS的风险，可为临床早期识别高危患者及制定个体化干预策略提供可视化工具。

Abstract: Objective: To investigate the independent influencing factors for low anterior resection syndrome (LARS) after rectal cancer surgery and to develop an individualized predictive nomogram model. Methods: Clinical data of 506 patients who underwent low anterior resection for rectal cancer at a single center between October 2019 and September 2024 were retrospectively collected. Patients were categorized into a LARS group (n = 243) and a non-LARS group (n = 263) based on their LARS score. Univariate and multivariate logistic regression analyses were performed to identify independent factors associated with LARS. The data were randomly split into a training set (n = 354) and a validation set (n = 152) in a 7:3 ratio. A nomogram prediction model was constructed based on the identified independent factors. The model’s discriminative ability, calibration, and clinical utility were internally validated using the receiver operating characteristic (ROC) curve, calibration curve, and decision curve analysis (DCA). Results: Multivariate analysis revealed that anastomotic leakage (odds ratio [OR] = 9.025), prolonged time to stoma reversal (OR = 1.260 per month), and comorbid diabetes (OR = 2.281) were independent risk factors for LARS. Conversely, a greater distance from the tumor’s lower edge to the anal verge (OR = 0.446) and a greater distance from the tumor’s distal resection margin to the anal verge (OR = 0.439) were protective factors. The nomogram model constructed based on these factors demonstrated an area under the curve (AUC) of 0.83 in the training set and 0.78 in the validation set. The calibration curve indicated good agreement between predicted and observed probabilities. Decision curve analysis showed that the model provided a clinical net benefit across a wide range of threshold probabilities. Conclusion: The developed nomogram model effectively predicts the risk of LARS in patients after rectal cancer surgery, offering a visual tool for the early identification of high-risk patients and for formulating individualized intervention strategies.