Abstract
Kidney International advance online publication: 22 May 2013
Acute kidney injury (AKI) is common in the critically ill and is associated with a high hospital mortality rate. In this patient group, the use of biomarkers to detect AKI is hindered by the heterogeneity of injury and the effects of prevalent AKI. A single biomarker is unlikely to reflect the multiple pathways involved.
This study attempted to improve diagnostic performance by using three distinct urine markers: neutrophil gelatinase-associated lipocalin (NGAL), L-type fatty acid-binding protein (L-FABP) and cystatin C. The authors measured these markers in a large nested case–control study of critically ill patients to assess whether they improved prediction of the development of incident AKI, and subsequent dialysis and death, compared with a single marker alone.
Urine NGAL, L-FABP and cystatin C and serum creatinine were measured at study enrolment and after 48 h. AKI cases were identified (n = 130), and compared to controls (n = 250) without AKI, by measurement of serum creatinine on patients admitted to one of four intensive care units. Patients were followed up prospectively until hospital discharge.
Urine cystatin C did not show any difference between those with and without AKI and was not included in subsequent analysis. NGAL and L-FABP did not reliably discriminate between patients who did and did not develop AKI. Discrimination was improved when comparing those with more severe AKI with those with no injury. Using a multivariate regression model, both NGAL and L-FABP independently predicted AKI and subsequent need for dialysis. However, the sample size of patients who received dialysis was limited.
The study benefited from using a large and diverse critically ill population but the three urine biomarkers failed to reliably detect AKI in this patient group.