Urinary versus plasma neutrophil gelatinase-associated lipocalin (NGAL) as a predictor of mortality for acute kidney injury in intensive care unit patients.

OBJECTIVE

To examine urinary and plasma neutrophil gelatinase-associated lipocalin (NGAL) levels in predicting ICU mortality.

DESIGN

Prospective observational.

SETTING

University Critical Care setting.

PARTICIPANTS

50 patients with acute kidney injury (AKI).

INTERVENTIONS

None.

MEASUREMENTS AND MAIN RESULTS

Serial urinary and plasma concentrations of NGAL were measured. Twenty-five patients had early progression (EP) and 25 patients had early improvement (EI) of AKI. Plasma concentrations of NGAL in the EP group (N=25) were significantly higher than those in the EI group (129 [IQR; 20] vs. 111 [IQR; 32] ng/mL; P=0.009), while urine NGAL levels on admission were similar in both groups (61 [IQR; 20] vs. 65 [IQR; 20] ng/mL; P=0.767). Plasma NGAL concentrations rapidly decreased to 87 [32] ng/mL in the EI group (P<0.001) and while it remained elevated in the EP group (138 [21] ng/mL). Within 28-days, 50% of the patients died in the EP group, whereas no patient died in the EI group (P<0.001). Plasma NGAL was a fair predictor for progression of AKI (AUC; 0.719±0.063; P=0.006). 48-hour changes in plasma NGAL levels predicted death within 28-days of ICU admission (AUC; 0.874±0.048; P<0.001).

CONCLUSION

Early progression of AKI was associated with more death within 28 and 90days. While one time measurement of plasma NGAL levels at the time ICU admission may represent the kidney health status in critical care settings, it does not reliably predict mortality. On the other hand, changes in plasma NGAL within 48h of admission improve the value of this biomarker in predicting ICU mortality.