Abstract
Acute kidney injury (AKI) is a major complication with high mortality in patients admitted to the intensive care unit (ICU). Early diagnosis of AKI is not achievable with plasma creatinine alone.
The authors report a prospective observational study of six urinary biomarkers (γ-glutamyltranspeptidase [GGT], alkaline phosphatase [AP], neutrophil-gelatinase-associated lipocalin [NGAL], cystatin C [CysC], kidney injury molecule-1 [KIM-1] and interleukin-18 [IL-18]) as both diagnostic and predictive markers of AKI in a heterogeneous high-risk population.
Comparisons were made using the area under the receiver operating characteristic curve (AUC) for the diagnosis or prediction of AKI, need for dialysis and prediction of mortality at seven days. Data were reassessed after patient stratification by baseline renal function (estimated glomerular filtration rate) and time after renal insult.
Analysis included 528 patients with a mean age of 60 years. The primary diagnoses were variable, including sepsis (19%), cardiac surgery (18%), cardiac arrest (12%) and neurological causes (14%). Over 27% were diagnosed with AKI upon entry to the ICU (plasma creatinine >26.4 μmol/L or >50% above the baseline plasma creatinine concentration).
All biomarkers except AP were diagnostic of AKI, but with low sensitivity (AUC 0.59–0.67, sensitivity 0.27–0.40). CysC, IL-18 and NGAL were the strongest predictors of dialysis (AUC > 0.70) with high negative predictive values. All biomarkers except KIM-1 were predicting mortality within seven days (AUC > 0.60), especially IL-18 (AUC = 0.68).
The performance of all urinary biomarkers was improved by stratification for time after renal insult and baseline renal function before injury. GGT was the only early predictor of AKI within 6–12 h of insult, whereas CysC and IL-18 showed high predictive values beyond 36 h. In patients with pre-existing renal dysfunction, biomarkers showed limited utility and a delayed detection window (12–36 h after insult), attributed to impaired excretion.
The authors concluded that the high negative predictive value observed for all urinary biomarkers can guide management, but the absence of a high positive predictive value impedes triaging to intervention. The duration of injury and baseline renal function should also be considered when evaluating biomarker performance to diagnose AKI.