Background: Machine perfusion use has been reported to promote graft outcome in case of donation after cardiac death. Our objective was to evaluate the potential for nuclear magnetic resonance (NMR) to predict graft outcome by analyzing perfusates during machine perfusion time.
Method: We used a renal autotransplantation model mimicking deceased after cardiac death donors with pigs. Organs were subjected to 60 min of warm ischemia before the hypothermic machine preservation during 22 hr. We studied the correlation between creatinemia after transplantation and the NMR data from perfusates.
Results: A metabonomic analysis allowed us to highlight the evolution of several metabolites during perfusion: the concentration of lactate, choline, or amino acids such as valine, glycine, or glutamate increased with time, whereas there was a diminution of total glutathione during this period. The changes in these biomarkers were less severe in the group with the better outcome. Statistical analysis revealed a strong association between the level of those metabolites during machine perfusion and function recovery (Spearman rank ≥0.89; P<0.05).
Conclusion: Multivariate analysis of lesion biomarkers during kidney perfusion using NMR data could be an interesting tool to assess graft quality, particularly because analyses times (2 hr total) are compatible with clinical application.