Risk Factors And A New Nomogram Of Anastomotic Leak After Laparoscopic Anterior Resection For Rectal Cancer

OBJECTIVE: This retrospective study intends to analyze the risk factors associated with anastomotic leak(AL)after laparoscopic anterior resection(AR),and then to develop a model to assess the risk of AL in patients with rectal cancer.METHODS: A total of 1016 patients with rectal cancer who underwent laparoscopic AR at our hospital between August 2019 and February 2022 were retrospectively collected.Clinical factors collected included: age,BMI;history of smoking,alcohol consumption,underlying comorbidities(diabetes,hypertension,COPD),history of abdominal surgery and neoadjuvant chemoradiotherapy;preoperative hemoglobin,serum albumin,PLR(Platelet-lymphocyte ratio),NLR(Neutrophil-lymphocyte ratio);ASA grade,operation time,intraoperative blood loss,splenic flexure mobilization,the IMA high-ligation,distance from the anal verge,tumor diameter,and clinical TNM stage.The training set(n=777)and the validation set(n=239)were divided according to the time-validation principle.In the training set,the difference test of clinical factors between AL and non-AL groups was performed,and the factors with statistical difference were included in the stepwise multivariate Logistic regression analysis to obtain a model for predicting the probability of AL after laparoscopic AR,and the Nomogram was plotted according to the model.We tested the performance of the model by ROC curves,calibration plots,DCA curves and the Hosmer-Lemeshow test.RESULTS: The model obtained by stepwise multivariate logistic regression analysis included: age,clinical N stage,distance from the anal verge,PLR,operation time,splenic flexure mobilization and specimen length.Age between 60 and 75,c N2,PLR>142,splenic flexure mobilization and specimen length>12cm were independent risk factors(p<0.05).The AUC was 0.809(95% CI: 0.742-0.877)in training set,0.821(95% CI: 0.728-0.915)in validation set,and was 0.813(95% CI: 0.758-0.868)in the full set.The sensitivity and the specificity of the model in the training set were 79.8% and 72.5%,and were 77.8% and 72.2% in the validation set,and were 79.0% and 72.5% in the full set.The Hosmer-Lemeshow test shown that X2=7.47 with a p=0.588 in the training set and X2=3.83 with a p= 0.922 in the validation set.The calibration plots of the model in both groups were close to the standard curve,suggesting that the model has a high accuracy of prediction.The DCA curves analysis suggest that the model has good clinical applicability in both the training set and the validation set.CONCLUSION: In this study,age between 60 to 75,c N2,PLR>142,splenic flexure mobilization and specimen length>12cm were found to be independent risk factors for AL after laparoscopic AR.A predicting model of AL after laparoscopic AR for rectal cancer was constructed,which can improve the accuracy of predicting AL after surgery in suspicious patients.It can provide a reference for perioperative interventions and the selection of surgical methods to promote individualized treatment.