Contrast induced nephropathy: a new predictive model based on pre procedural glycemia and glomerular filtration rate

Aims: The risk of contrast induced nephropathy (CIN) is predicted by the already proposed formula of the ratio of contrast volume to glomerular filtration rate (GFR). Recent data from literature undescore that the incidence of CIN is significantly influenced by admission glycemia. Therefore, our aim was to identify a predictive model of CIN based on the quantity of contrast used during pPCI, known independent predictors of CIN, and pre procedural glycemia. Methods and results: 679 STEMI patients treated with primary PCI (pPCI) were enrolled in our prospective study. CIN was defined as an absolute serum creatinine increase ≥0.3 mg/dl after procedure. Admission hyperglycemia was defined as glucose levels >198 mg/dl. Medium volume of contrast (CV) we used ranged from 30 to 700 ml. We observed a significant increase in the incidence of CIN with the increase of the CV/GFR ratio which was in turn influenced by admission hyperglycemia. We therefore created a model of CIN prediction based on CV, GFR and admission glycemia. This model results from the product of admission glycemia and CV/GFR × 100. We then confronted our model with previously proposed models for prediction of CIN in pPCI patients (model 1: CV/GFR >3.7; model 2: CV/{[5 × weigth (kg)]/serum creatinin (mg/dl)} >1) obtaining better AUC with ROC analisys (AUC 0.72 vs 0.63, p<0.001 with model 1, AUC 0.72 vs 0,59, p<0.001 with model 2). At multivariate analisys a value of 4.2 was the best indipendent predictor of CIN. Conclusions: Our CIN risk model based on CV and admission glycemia was more accurate than the existing models. This model can be useful before pPCI to identify the safe quantity of contrast to use, and after pPCi to identify the subgroup of patients at higher risk of CIN.