Urinary biomarkers of cell cycle arrest are delayed predictors of acute kidney injury after pediatric cardiopulmonary bypass

Liqun Dong,Qing Ma,Michael Bennett,Prasad Devarajan
DOI: https://doi.org/10.1007/s00467-017-3748-7
Abstract:Background: Several novel biomarkers that predict acute kidney injury (AKI) have recently been proposed. We have evaluated the sequential patterns of biomarker elevation after pediatric cardiopulmonary bypass (CPB) and determined their diagnostic accuracy. Methods: We measured the ability of neutrophil gelatinase-associated lipocalin (NGAL), interleukin-18 (IL-18), liver type fatty-acid binding protein (L-FABP), kidney injury molecule-1 (KIM-1), tissue inhibitor of metalloproteinase-2 (TIMP-2), and insulin-like growth factor binding protein 7 (IGFBP7), to predict AKI (≥50% increase in serum creatinine from baseline). Areas under the receiver-operator characteristic curves (AUCs) were calculated for each biomarker and for various biomarker combinations at multiple time points after CPB. Results: Of 150 patients examined, AKI had developed in 50 patients by 24 h after CPB, with an elevated NGAL concentration first noted at 2 h post-CPB, increases in IL-18, L-FABP, and the product of TIMP-2 and IGFBP7 first noted at 6 h, and an elevated KIM-1 level noted at 12 h. At each time point, urine NGAL remained the marker with the highest predictive ability (AUC > 0.9). The addition of any other biomarker did not increase the predictive accuracy of NGAL alone at 2 and 6 h. At 12 h, when compared to NGAL alone, the combination of NGAL, IL-18, and TIMP2 improved the AUC for AKI prediction (from 0.938 to 0.973). Conclusions: While urine NGAL remains a superior stand-alone test at the 2 and 6 h time points after pediatric CPB, a panel of carefully selected biomarkers may prove optimal at later time points.
What problem does this paper attempt to address?