Preoperative Evaluation Of Hepatectomy-associated Acute Kidney Injury-a New Nomogram Prediction Model

Objective: : Hepatectomy-associated acute kidney injury(HEA-AKI)is a common complication in patients with liver resection,is associated with considerable morbidity,mortality and high costs thus placing a heavy burden to society.At present,the risk factors for HEA-AKI are still unclear.This study was aimed to observe the incidence of HEA-AKI,analyze the preoperative related risk factors of HEA-AKI,to build a predictive model based on preoperative features in order to early recognize and intervene for patients with high risk of HEA-AKI.Methods: This is a retrospective cohort study enrolling 2769 inpatients who underwent hepatectomy from the Affiliated Hospital of Qingdao University from October 2012 to July 2018,data was split into training(2050)and test(719)cohorts.The post-hepatectomy AKI was diagnosed according to the 2012 KDIGO AKI guideline.Univariate regression analysis was used to evaluate the potential risk factors for HEA-AKI,and the univariate analysis result p <0.05 was included in the COX regression.COX regression was used to analyze the independent risk factors of HEA-AKI.Identified independent predictive factors associated with HEA-AKI were incorporated into the nomogram,the discriminative ability and predictive accuracy of the model was assessed by the area under the receiver operating characteristic curve(AUC).Result : In the end,the incidence of HEA-AKI after hepatectomy was 4.80%(133/2769).COX regression results showed hypertriglyceridemia(HR=1.594,95%CI=1.252-2.031,P<0.001),hypoalbuminemia(HR=0.652,95%CI=0.464-0.915,P=0.013),alkaline phosphatase(HR=1.900,95%CI=1.228-2.939,P=0.004),aspartate aminotransferase(HR=1.711,95%CI=1.182-2.476,P=0.004),open surgery(HR=2.678,95%CI=1.164-6.161,P=0.020),lower preoperative glomerular filtration rate(HR=0.595,95%CI=0.425-0.832,P=0.002),aspirin(HR=1.748,95%CI=1.182-2.584,P=0.005)and duration of surgery(HR=1.042,95%CI=1.015-1.070,P=0.002)were independent risk factors for HEA-AKI.We incorporated these indicators into the nomogram to construct a predictive model for HEA-AKI,the area under the ROC curve is 0.764.The area under the ROC curve of the text group is 0.781.Moreover,calibration plots showed an optimal consistency with the actual presence of HEA-AKI.Conclusion: This study shows that hypertriglyceridemia,hypoalbuminemia,hyperalkaline phosphatase,high aspartate transaminase,laparotomy,low glomerular filtration rate,preoperative aspirin,operation duration is independent risk factors for HEA-AKI.This study constructed a nomogram prediction model for HEA-AKI.The nomogram had excellent predictive ability and might have significant clinical implications for the early detection of HEA-AKI.