Background: Donation after cardiac death (DCD) increases the number of donor kidneys but is associated with more primary nonfunction (PNF) and delayed graft function (DGF). It has been suggested that biomarkers in the preservation solution of machine perfused kidneys may predict PNF, although evidence is lacking.
Methods: We analyzed the diagnostic accuracy of the perfusate biomarkers glutathione S-transferase, lactate dehydrogenase (LDH), heart-type fatty acid binding protein, redox-active iron, interleukin (IL)-18, and neutrophil gelatinase-associated lipocalin to predict PNF and DGF in 335 DCD kidneys preserved by hypothermic machine perfusion at our center between 1 January 1997 and 1 January 2008. The diagnostic accuracy of these biomarkers to predict PNF was evaluated with the area under the receiver operating characteristics curves. Additionally, the risk of DGF and graft failure was assessed.
Results: LDH and IL-18 concentrations were associated with PNF (odds ratio [95% confidence interval], 1.001 [1.000-1.002]; P=0.005 and 1.001 [1.000-1.002]; P=0.003, respectively) in a multivariate analysis; the diagnostic accuracy for PNF was "poor" for all biomarkers but increased to "fair" for redox-active iron and IL-18 in a multivariate analysis (area under the receiver operating characteristics curves, 0.701 and 0.700, respectively). LDH and IL-18 concentrations were associated with DGF; biomarker concentration was not associated with 1-year graft survival.
Conclusions: The diagnostic accuracy of the perfusate biomarkers glutathione S-transferase, LDH, heart-type fatty acid binding protein, redox-active iron, IL-18, and neutrophil gelatinase-associated lipocalin to predict viability of DCD kidneys varies from "poor" to "fair". Therefore, DCD kidneys should not be discarded because of high biomarker perfusate concentration.