Rapid detection of acute kidney injury by plasma and urinary neutrophil gelatinase-associated lipocalin after cardiopulmonary bypass

Abstract

Background: Cardiopulmonary bypass (CPB) is associated with a significant risk of postoperative renal dysfunction. We studied the utility of a novel biomarker in predicting acute kidney injury (AKI) in adult patients undergoing cardiac surgery.

Methods and results: Blood and urine were obtained from 50 patients undergoing CPB-requiring surgery. Patients were divided into group A (n=41) with normal creatinine pre-bypass and post-bypass and group B (n=9) who developed an increase in serum creatinine of >0.5 mg/dL within the first 48 hours post CPB. Plasma and urinary neutrophil gelatinase-associated lipocalin (NGAL) was determined at baseline and 2 hours after CPB.Plasma levels of NGAL were higher in patients who developed AKI [214+/-16.7 ng/mL (95% CI 176.9-252.9)] compared with those who did not [149.5+/-13.5 ng/mL (95% CI 122.1-175.7); P=0.035]. Two hours after CPB, there was a significant increase (P=0.0003) in NGAL levels, greater in those patients who developed AKI [476.1+/-41.1 ng/mL (95% CI 380.6-571.6); P=0.0003] compared with those who did not [278.4+/-22 ng/mL (95% CI 233.9-323.0)]. In the AKI group, urinary NGAL increased from 7.13+/-2.30 ng/mL (95% CI 2.5-11.8) to 2924+/-786 ng/mL (95% CI 1110-4739). In the non-AKI group, there was an increase from 1.6+/-0.6 (95% CI 0.3-3.0) to 749+/-179 ng/mL (95% CI 386-1113). The post-CPB levels of urinary NGAL were significantly different in the AKI group (P<0.0001) such that a suitable threshold for use as a diagnostic test could be determined. Receiver operating characteristics were determined for plasma and urinary NGAL with area under the curve (AUC) of 0.80 and 0.96, respectively. For a threshold of 433 ng/mmol creatinine, the test had 90% sensitivity and 78% specificity for the detection of post-CPB renal dysfunction.

Conclusions: Measurement of this novel biomarker in the urine or plasma of patients in the first hours after CPB is predictive of subsequent renal injury. Although the AUC for plasma NGAL seemed inferior to urine, even an AUC of 0.8 as reported compares very favorably to that for other "outstanding" biomarkers (eg, AUCs in the 0.7 range for troponin).