#4677 EARLY MARKERS OF POSTOPERATIVE ACUTE KIDNEY INJURY

Abstract

Abstract Background and Aims Acute kidney injury (AKI) is a common clinical condition among patients who undergo cardiac surgery, significantly affecting morbidity and mortality[1]. To date, preoperative biomarkers and solid predictive models for AKI are not yet currently available in clinical practice. Recent studies reported endogenous ouabain (EO), a stress hormone secreted by the adrenal glands, as associated to worse kidney outcomes after cardiac surgery[2]. The aim was to validate the use of EO as biomarker of individual susceptibility for AKI after cardiac surgery and to create a new powerful score for postoperative AKI risk. Method 1174 patients undergoing elective cardiac surgery were enrolled in the study. Preoperative biological samples were collected and analysed. The primary outcome was AKI development, according to KDIGO 2012 guidelines. Results 21.6% of patients developed AKI (9% developed severe AKI, stage≥2). AKI confirmed strong correlation to postoperative death: all patients dead within 90 days for postsurgical complication developed AKI and each AKI stage was associated to a 10 times higher mortality risk (p < 0.001). Different preoperative clinical variables were analyzed, identifying five independent risk factors significantly correlated to AKI and severe AKI: age, FE, NYHA class, reoperation and complex surgical intervention (p < 0.001 for all of them). Preoperative EO levels turned out significantly associated to the incidence of AKI (p < 0.001) and clinical complications. Specifically, patients with higher EO levels had greater incidence of AKI (p < 0.001) and worse cardiac and kidney basal function (p = 0.005 and p = 0.003, respectively). Finally, a simple score was developed on the base of those preoperative clinical values significantly associated to AKI (AUC for severe AKI = 0.79, p <0,001). Adding preoperative EO to the model significantly improved the predictive capability (AUC for severe AKI = 0.82, p < 0.001; Δ-AUC +0.0229, p = 0.026). Conclusion We confirmed in a large population the role of preoperative plasma level of EO as an important early predictor for post-surgical AKI and we built a powerful clinical model for postoperative AKI risk, exclusively using few simple preoperative clinical factors and a single biohumoral marker.