术前控制营养状态评分对结直肠癌根治术患者并发术后肠梗阻的预测价值

The predictive value of preoperative controlling nutritional status score for concurrent postoperative ileus in patients undergo⁃ing radical resection of colorectal cancer

目的探讨术前控制营养状态(CONUT)评分对接受结直肠癌根治术患者并发术后肠梗阻(POI)的预测价值。方法选择2019年1月至2022年9月在海南医学院第二附属医院择期全麻下行结直肠癌根治术患者640例,收集患者的一般情况、实验室检查和组织病理学检查结果等临床资料。根据术前3 d内的血常规结果、血清清蛋白和总胆固醇水平计算中性粒细胞与淋巴细胞比值(NLR)、全身免疫炎症指数(SII)和CONUT评分。应用受试者工作特征(ROC)曲线分析比较上述3种指标预测接受结直肠癌根治术患者并发POI的效能。根据ROC曲线确定的CONUT评分最佳截断值,将患者分成高CONUT评分组和低CONUT评分组,并比较两组患者临床病理特征的差异。应用单因素分析和多因素logistic回归分析确定接受结直肠癌根治术患者并发POI的独立危险因素,并以此构建预测接受结直肠癌根治术患者并发POI的列线图。分别通过C-index、校正曲线和决策曲线来确定列线图模型的区分度、一致性和临床实用性。结果NLR、SII、CONUT评分预测接受结直肠癌根治术患者并发POI的曲线下面积(AUC)分别为0.616(95%CI:0.577~0.654)、0.631(95%CI:0.592~0.668)、0.728(95%CI:0.692~0.726)。CONUT评分预测接受结直肠癌根治术患者并发POI的效能高于NLR和SII(Z分别为3.427、2.897,均P<0.05)。CONUT评分的最佳阈值为4分,以此将640例患者分为低CONUT评分组(CONUT评分<4分,484例)和高CONUT评分组(CONUT评分≥4分,156例)。与低CONUT评分组相比,高CONUT评分组年龄≥65岁比例和POI发生率更高(均P<0.05)。多因素logistic回归分析结果显示,CONUT评分≥4分(OR=6.016,95%CI:3.771~9.597),CCI评分≥3分(OR=1.702,95%CI:1.015~2.854),开腹手术(OR=1.742,95%CI:1.029~2.951),造口(OR=3.096,95%CI:1.594~6.015)是影响接受结直肠癌根治术患者并发POI的独立危险因素(均P<0.05)。列线图模型内部验证的C-index为0.771(95%CI:0.722~0.820)。校准曲线显示列线图模型的预测概率与实际概率均接近,具有较好的一致性。决策曲线显示,当阈值0.16~0.75时,列线图模型具有较好的临床实用性。结论术前CONUT评分可作为预测接受结直肠癌根治术患者并发POI的指标,并且预测效能高于NLR和SII。以术前CONUT评分≥4分等独立危险因素构建的列线图模型具有较高的预测价值和临床实用性。

Objective To investigate the predictive value of concurrent preoperative controlling nutritional status(CONUT)score for postoperative ileus(POI)in patients undergoing radical resection of colorectal cancer.Methods From January 2019 to September 2022,a to⁃tal of 640 patients undergoing radical resection of colorectal cancer under general anesthesia in the Second Affiliated Hospital of Hainan Medi⁃cal University were recruited,and the clinical data such as general conditions,laboratory examination and histopathological examination results were collected.Neutrophil-to-lymphocyte ratio(NLR),systemic immune-inflammatory index(SII)and CONUT score were calculated based on the blood routine results,serum albumin and total cholesterol levels within three days before operation.The receiver operating characteristic(ROC)curves were used to compare the efficiency of the above three indicators in predicting concurrent POI in patients undergoing radical re⁃section of colorectal cancer.According to the optimal cut-off value of CONUT score determined by the ROC curve,the patients were divided into the high CONUT score and the low CONUT score groups,and the differences in the clinical pathological characteristics of the patients were compared between the two groups.Univariate analysis and multivariate logstic regression analysis were used to identify the independent risk factors of concurrent POI in patients undergoing radical resection of colorectal cancer,so as to establish a nomogram model based on the re⁃sults to predict concurrent POI in patients.The differentiation,consistency and clinical practicability of the nomogram model were determined by C-index,the calibration curve and the decision curve,respectively.Results The areas under the curve(AUC)of NLR,SII,and CONUT scores for predicting concurrent POI in patients undergoing radical resection of colorectal cancer was 0.616(95%CI:0.577~0.654),0.631(95%CI:0.592~0.668),and 0.728(95%CI:0.692~0.726),respectively.There was higher efficiency of CONUT score in predicting concurrent POI in patients undergoing radical resection of colorectal cancer than that of NLR and SII(Z=3.427,2.897,respectively,both P<0.05).The op⁃timal cut-off value of CONUT score was 4 points,and 640 patients were divided into the low CONUT score group(CONUT score<4 points,n=484)and the high CONUT score group(CONUT score≥4 points,n=156).Compared with the low CONUT score group,the proportion of age≥65 years and the incidence of POI were higher in the high CONUT score group(all P<0.05).Multivariate logstic regression analysis showed that CONUT score≥4 points(OR=6.016,95%CI:3.771~9.597),CCI score≥3 points(OR=1.702,95%CI:1.015~2.854),open surgery(OR=1.742,95%CI:1.029~2.951),and stoma(OR=3.096,95%CI:1.594~6.015)were independent risk factors for concurrent POI in patients undergoing radical resection of colorectal cancer(all P<0.05).The C-index for internal verification of the nomogram model was 0.771(95%CI:0.722~0.820).The calibration curves showed that the predicted probability of nomogram model was close to the actual probability with displayed good agreement.The decision curve showed that when the threshold value was between 0.16 and 0.75,the nomograph model had better clinical prac⁃ticability.Conclusion Preoperative CONUT score can be used as an indicator to predict concurrent POI in patients undergoing radical resec⁃tion of colorectal cancer.The predictive efficiency of CONUT score is higher than that of NLR and SII.The nomogram according to independent risk factors such as preoperative CONUT score≥4 points has high predictive value and clinical practicability.