Urinary π-glutathione S-transferase Predicts Advanced Acute Kidney Injury Following Cardiovascular Surgery

Urinary biomarkers augment the diagnosis of acute kidney injury (AKI), with AKI after cardiovascular surgeries being a prototype of prognosis scenario. Glutathione S-transferases (GST) were evaluated as biomarkers of AKI. Urine samples were collected in 141 cardiovascular surgical patients and analyzed for urinary alpha-(alpha-) and pi-(pi-) GSTs. The outcomes of advanced AKI (KDIGO stage 2, 3) and all-cause in-patient mortality, as composite outcome, were recorded. Areas under the receiver operator characteristic (ROC) curves and multivariate generalized additive model (GAM) were applied to predict outcomes. Thirty-eight (26.9%) patients had AKI, while 12 (8.5%) were with advanced AKI. Urinary p-GST differentiated patients with/without advanced AKI or composite outcome after surgery (p < 0.05 by generalized estimating equation). Urinary p-GST predicted advanced AKI at 3 hrs post-surgery (p = 0.033) and composite outcome (p = 0.009), while the corresponding ROC curve had AUC of 0.784 and 0.783. Using GAM, the cutoff value of 14.7 mu g/L for p-GST showed the best performance to predict composite outcome. The addition of p-GST to the SOFA score improved risk stratification (total net reclassification index = 0.47). Thus, urinary pi-GST levels predict advanced AKI or hospital mortality after cardiovascular surgery and improve in SOFA outcome assessment specific to AKI.