Plasma Neutrophil Gelatinase-Associated Lipocalin Is Useful for Predicting Mortality in Critically Ill Patients

Elevated neutrophil gelatinase-associated lipocalin (NGAL) occurs in a wide range of systemic diseases. This study examined the clinical utility of plasma NGAL to predict intensive care unit (ICU) and in-hospital mortality in critically ill patients. A total of 62 patients hospitalized in a mixed ICU were included; pNGAL, creatinine, and C-reactive protein (CRP) were assayed on four consecutive days (D1-D4) following ICU admission. APACHE II score (Acute Physiology and Chronic Health Evaluation) was calculated 24 h post-admission. ICU mortality reached 35% and in-hospital mortality was 39%. The median pNGAL at admission was 142.5 (65.6-298.3) ng/mL. pNGAL was significantly higher in non-survivors compared to survivors. The highest accuracy for ICU mortality prediction was achieved at the pNGAL cutoff of 93.91 ng/mL on D4 area under the curve (AUC) = 0.89; 95%CI 0.69-0.98 and for in-hospital mortality prediction was achieved at the pNGAL cutoff of 176.64 ng/mL on D3 (AUC = 0.86; 95%CI 0.69-0.96). The APACHE II score on ICU admission predicted ICU mortality with AUC = 0.89 (95%CI 0.79-0.96) and in-hospital mortality with AUC = 0.86 (95%CI 0.75-0.94). Although pNGAL on D1 poorly correlated with APACHE II (R = 0.3; p = 0.01), the combination of APACHE II and pNGAL on D1 predicted ICU mortality with AUC = 0.90 (95%CI 0.79-0.96) and in-hospital mortality with AUC = 0.95 (95%CI 0.78-0.99). Maximal CRP during study observation failed to predict ICU mortality (AUC = 0.62; 95%CI 0.49-0.74), but helped to predict in-hospital mortality (AUC = 0.67; 95%CI 0.54-0.79). Plasma NGAL with combination with the indices of critical illness is a useful biomarker for predicting mortality in heterogeneous population of ICU patients.

Automated summary

The study demonstrated that plasma NGAL, combined with indices of critical illness, is a useful biomarker for predicting mortality in heterogeneous population of ICU patients. pNGAL on D3 was found to be the most suitable for predicting in-hospital mortality.