Derivation And Validation Of A Novel Risk Nomogram For Predicting Contrast-Induced Acute Kidney Injury In Patients With STEMI Undergoing Primary PCI

Objective:To establish a novel risk nomogram,the CI-AKI risk nomogram to evaluate the risk of contrast-induced acute kidney injury(CI-AKI)in Patients with ST-Segment Elevation Myocardial Infarction(STEMI)Undergoing Primary Percutaneous Coronary Intervention.Methods:(1)The present study was a retrospective,single-center study.We selected 1167 patients admitted with the diagnosis of acute ST-elevation myocardial infarction who underwent primary PCI at the Heart Center,First Hospital of Lanzhou University from October 2017 to May 2018.524 patients with missing key data or compliance with exclusion criteria were excluded,a total of 643 patients were finally included in this study,These patients were randomly assigned into two cohorts:one was for the derivation cohort(n=448)and another for validation(n=195).The following information with general demographics,laboratory indicator,angiographic and procedural characteristics,use of the drug in hospital,and hemodynamic parameters was collected within 24 hours after admission to PCI.The primary endpoint was all-cause death during in-hospital stay.(2)CI-AKI is defined based on KDIGO criteria:In absence of other causes,a 0.3 mg/dl(26.5?mol/L)increase in serum creatinine within 48 hours after the intravascular injection of contrast agents or urine output less than 0.5 ml/kg/h for more than 6 hours within 7 days after the use of contrast.(3)The development of cohort was divided into 2 groups according to CI-AKI criteria,and both groups were compared according to clinical factors.SPSS25.0(SPSS Inc,Chicago,IL,USA)was used for statistical analysis.The comparison was done using?2–test or Fisher?s exact test for categorical variables and Student T-test for continuous variables.Non-normally distributed data were log-transformed for use with parametric statistics,including NT-ProBNP,admission glucose and monocyte count/lymphocyte count ratio,Multivariate logistic regression analysis was performed to determine the risk factors associated with CI-AKI(P values less than0.05 are considered statistically significant);Multivariate Cox proportional hazards regression analysis with backward,stepwise selection was used to evaluate independent prognostic factors associated with endpoint.Using receiver operating characteristic(ROC)curves to determine the diagnostic efficacy of predictive models.(4)Model construction:Combined the results of multivariable logistic regression with previous reports of risk factors independently associated with CI-AKI and formulated the CI-AKI nomogram risk score model by R software.Discrimination and calibration of the predicting model were assessed using the C-index and calibration plot analysis.Internal validation using 1000 bootstrap in development cohort by R software,External validation:calculate individual scores and take risk stratification of the validation cohort(n=195),use the area under the ROC curve to evaluate the predictive value of the nomogram,and using Hosmer-Lemeshow test to evaluate the goodness-of-fit.Results:(1)The basic clinical characteristics:There were 383 males and 65females included in our study,60.5±11.4 years old,a total of 35 patients(7.81%)were diagnosed with CI-AKI,Two patients required renal replacement therapy for CIAKI.CI-AKI group was prone to older age,female sex,the proportion of smoking history,hypertension,diabetes,insulin-treated diabetes mellitus,previous myocardial infarction,Killip classification?2,basic renal insufficiency,in-hospital death,hemodialysis,the use of IABP,diuretics,dopamine,and epinephrine,cardiogenic shock,cardiac arrest,atrial fibrillation were significantly higher than non-CI-AKI Group.Regarding biochemical and procedural characteristics,the level of baseline NT-ProBNP,peak NT-ProBNP,admission creatinine,serum osmolarity,serum glucose and the dosage of contrast media were higher in CI-AKI group;LVEF,LMR,hemoglobin content and hematocrit level were significantly lower than the non-CI-AKI group.(2)The following factors were independently associated with increased risk for AKI:hypertension,diabetes requires insulin treatment,cardiogenic shock,and peak ln NT-ProBNP,the odds ratio was 2.302,4.856,1.819,and 8.271,respectively.all P values less than 0.05.Then we combined these factors and developed predictive model 1,compared with Mehran score and ln peak NT-ProBNP.according to our analysis,predictive model 1 had better prediction abilities.AUROC for the scoring systems ranged from 0.737(Mehran score)to 0.804(prediction model1)for the prediction of CI-AKI risk.(3)Derivation and validation of a novel nomogram(prediction model 2):We combined age,use of IABP,basic renal insufficiency,hypertension,diabetes,cardiogenic shock,and peak NT-ProBNP within48h to form a new scoring system,and the model displayed good discrimination with a C-index of 0.815(95%confidence interval:0.728-0.902)and good calibration.High C-index value of 0.774 could still be reached in the interval validation.External validation based on the validation cohort(n=195),the area under the AUC curve is0.8622,95%CI:0.806-0.907,P<0.0001,with 58.8%sensitivity and 96.9%specificity.Hosmer-Lemeshow?~2=2.669,P=0.953,and shows a good calibration.(4)Survival analysis:8 patients(1.8%)with all-cause death in the hospital,5 in the CI-AKI group(14.3%),3 in the non-CI-AKI group(0.7%),and the cumulative survival rate between the two groups had significant differences(P<0.001);multivariate Cox regression analyses showed that perioperative use of epinephrine and CI-AKI was independently associated with the risk of all-cause death during hospitalization,the hazard ratio(HR)and 95%confidence interval were 17.46(3.51-97.14),17.43(2.89-105.24),respectively;P-values were 0.001 and 0.002,respectively.Conclusions:In this study,the main findings are as follows:(1)STEMI patients undergoing primary PCI had a high incidence of postoperative CI-AKI(7.8%),which was independently associated with the short-term adverse outcomes.(2)insulin-treated DM,hypertension,cardiogenic shock(within 48 hours after admission to hospital)and peak NT-ProBNP were independently associated with CI-AKI in STEMI patients undergoing PPCI.(3)The present study developed a simple CI-AKI risk score based on common clinical factors and laboratory indicators.This tool can accurately identify the risk of CI-AKI after primary PCI in STEMI patients,but This risk model still needs validation in an external cohort.