FC 050THE PREDICTIVE ABILITY OF URINARY BIOMARKERS FOR PROGRESSION OF ACUTE KIDNEY INJURY IN CRITICAL ILLNESS

Abstract

Abstract Background and Aims Acute Kidney Injury (AKI) is a common hospital complication associated with high morbidity, mortality and the risk of progression to CKD. Despite globally increasing rates of AKI and a shortage of critical care beds, clinicians currently lack the tools to predict progression of early AKI. The Dublin Acute Biomarker Group Evaluation (DAMAGE) Study is a prospective multi-center observational study with a heterogenous cohort of critically ill patients (n= 717). We hypothesised that a panel of 14 novel urinary biomarkers would predict progression of AKI. Method Biomarker values were measured daily for 7 days from ICU admission, including on the day of Stage 1-2 AKI diagnoses occurring within 48 hours of ICU admission. The primary outcome was progression to the composite outcome of increased AKI KDIGO Stage, Renal Replacement Therapy (RRT) or Death within 7 days of ICU admission. AUC-ROC values and adjusted Odds Ratios were calculated using multivariate logistic regression, and the additional value of biomarkers over the clinical model (serum creatinine and urine output at time of diagnosis) was assessed. In a secondary analysis,all AKI events within 7 days (including those diagnosed at Stage 3, and/or on days 3-7) were included, with the endpoint of progression to RRT/Death within 30 days of ICU admission. Results AKI developed in 268 patients (37%) within 7 days of ICU admission. 95 patients were diagnosed with AKI at Stage 1or2 AKI within 48 hours, of whom 32 (33.7%) progressed to the primary outcome. Adjusted odds ratios for the prediction of AKI progression were significantly higher for the highest tertile, versus the lowest, of eight biomarkers including uCystatin C (OR 5.2, 95% CI; 1.3-23.6), uAlbumin (OR 4.9, 95% CI; 1.5-18.3), uNGAL (OR 4.6, 95% CI; 1.4-17.9) and uIGFBP-7 (OR 4.7, 95% CI; 1.2-21.5). LFABP-1, KIM-1, IL-18 and Alpha-1-M also significantly improved prediction (Table 1). In ROC analysis of the primary outcome, uCystatin C (AUC 0.75, 95% CI; 0.63-0.87), uNGAL (AUC 0.72, 95% CI;0.61-0.84), uAlbumin (AUC 0.70, 95% CI; 0.59-0.82) and IL-18 (AUC 0.70, 95% CI; 0.58-0.82) had moderate accuracy; the others did not. Conclusion In this cohort study of AKI progression within 7 days of ICU admission, eight novel urinary biomarkers improved the prediction of AKI progression after multivariate adjustment. Similarly, ROC analysis found that several biomarkers had moderate accuracy for the prediction of AKI progression in the ICU.