Glasgow prognostic score and its derived scores predicts contrast-associated acute kidney injury in patients undergoing coronary angiography

Background: Glasgow prognostic score (GPS) is a reliable scoring system reflecting both nutritional and inflammatory factors. The association of inflammation and nutrition with contrast-associated acute kidney injury (CA-AKI) has been validated. This study set out to determine the impact of GPS and its derived scores on CA-AKI incidence.Methods: Populations treated with coronary angiography with/without percutaneous coronary intervention were screened retrospectively. According to C-reactive protein and albumin, three kinds of GPSs were involved: GPS, modified GPS (mGPS), and the cutoff-based GPS (cGPS) which was derived by calculating the optimal cutoff values of two parameters. Primary endpoint was CA-AKI. Pearson' r correlation, linear/logistic regression, receiver operating characteristic curve as well as subgroup analyses were conducted.Results: Totally, 3150 patients were valid for analysis, and the mean age was 67.5 years old, with 66.4 % male. Of these, 610 patients suffered CA-AKI. All three kinds of GPSs were independently associated with the SCr elevation proportion (GPS: beta = 4.850, 95%CI [3.700 to 8.722], P < 0.001; mGPS: beta = 3.450, 95%CI [1.896 to 6.888], P = 0.001; cGPS: beta = 3.992, 95%CI [2.368 to 6.940], P < 0.001). GPS, mGPS and cGPS were proved to be the independent risk factors for CA-AKI risk (all P for trend <0.05). Compared with GPS and mGPS, cGPS was of greater prognostic value for predicting CA-AKI incidence (cGPS: AUC = 0.633; mGPS: AUC = 0.567; GPS: AUC = 0.611). Main findings were also consistent in all subgroup analysis.Conclusion: Preprocedural GPS and its derived scores (mGPS and cGPS), especially cGPS, were correlated with the incidence of CA-AKI, which might assist in clinical decision making in treating CA-AKI.