肾部分切除术后近期术侧肾功能损失的危险因素及预测模型构建

Risk factors and predictive model for short-term ipsilateral renal function loss after partial nephrectomy

摘要:

目的 分析肾部分切除术（PN）后近期术侧肾功能损失的危险因素，并以此构建预测模型进行验证。方法 回顾性分析2015年1月—2023年12月江南大学附属医院收治的107例行PN治疗的患者的临床资料，根据患者术后近期术侧肾功能损失发生情况将其分为损失组（27例）和无损失组（80例）。采用多因素逐步Logistic回归模型分析影响患者术后术侧肾功能损失的危险因素，并以此构建Nomogram列线图模型预测患者术后肾功能损失的发生风险；绘制受试者工作特征（ROC）曲线，分析该模型对患者术后肾功能损失的预测效能。结果 损失组肿瘤最大径、R.E.N.A.L评分、热缺血占比、缺血时间> 45 min占比均高于无损失组，肾体积保留率低于无损失组（P <0.05）。多因素逐步Logistic回归分析结果显示，R.E.N.A.L评分［O^R=5.609（95% CI：2.710，11.606）］、缺血类型［O^R=4.462（95% CI：1.978，10.064）］是PN患者术后近期术侧肾功能损失的危险因素（P <0.05）；肾体积保留率［O^R=0.285（95% CI：0.098，0.826）］是保护因素（P <0.05）。基于上述影响因素构建的列线图预测模型经Bootstrap法内部验证，结果显示，C-index指数为0.852（95% CI：0.783，0.964），预测患者肾功能损失的校正曲线趋近于理想曲线（P >0.05）。列线图模型预测患者肾功能损失的敏感性为88.90%（95% CI：0.791，0.984）、特异性为91.20%（95% CI：0.841，0.994）。结论 基于影响因素构建的列线图预测模型可较好地评估患者PN术后近期术侧肾功能损失的发生风险。

关键词:肾部分切除术;肾功能损失;影响因素;风险预测模型;列线图

Abstract:

Objective To analyze the risk factors of short-term renal function loss of the operated side after partial nephrectomy (PN), and to construct and verify a prediction model based on this, in order to provide some guidance for improving postoperative renal function of patients.Methods The clinical data of 107 patients who underwent PN in our hospital from January 2015 to December 2023 were retrospectively analyzed. The patients were divided into renal function loss group (27 cases) and non-renal function loss group (80 cases) according to the occurrence of renal function loss on the operated side after operation. The risk factors affecting postoperative renal function loss of the surgical side of patients were analyzed, and a Nomogram model was constructed to predict the risk of postoperative renal function loss of patients. The receiver operating characteristic curve (ROC) was drawn, and the area under the curve (AUC) was used to analyze the predictive efficiency of the prediction model for postoperative renal function loss of patients.Results The maximum tumor diameter, R.E.N.A.L score, the proportion of warm ischemia type and the proportion of ischemic time > 45 min in the loss group were higher than those in the non-loss group, and the renal volume preservation rate was lower than that in the non-loss group (P <0.05). Multivariate stepwise Logistic regression analysis showed that R.E.N.A.L score [O^R = 5.609 (95% CI: 2.710, 11.606) ], ischemic type [O^R = 4.462 (95% CI: 1.978, 10.064) were the risk factors of postoperative short-term renal function loss in PN patients (P < 0.05). Renal volume preservation rate [O^R = 0.285 (95% CI: 0.098, 0.826) ] was a protective factor (P < 0.05). The nomogram prediction model based on the above influencing factors was internally verified by the Bootstrap method, and the C-index index was 0.852 (95% CI: 0.783, 0.964), and the calibration curve for predicting renal function loss of patients was close to the ideal curve (P > 0.05). ROC showed that the sensitivity and specificity of the nomogram model in predicting renal function loss were 88.90% (95% CI: 0.791, 0.984) and 91.20% (95% CI: 0.841, 0.994), respectively.Conclusion R.E.N.A.L score and ischemia type are independent risk factors for postoperative short-term renal function loss in PN patients, and renal volume preservation rate is an independent protective factor. The nomogram prediction model based on this model can better evaluate the risk of short-term renal function loss on the operated side after PN.

Key words:partial nephrectomy;renal function loss;risk factors;risk prediction model;nomogram