Acellular Perfusate is an Adequate Alternative to Packed Red Blood Cells During Normothermic Human Kidney Perfusion

Alban Longchamp,Fermin M Fontan,Mohamed M Aburawi,Corey Eymard,Negin Karimian,Danielle Detelich,Casie Pendexter,Stephanie Cronin,Thomas Agius,Sonal Nagpal,Peony Dutta Banik,Shannon N Tessier,Sinan Ozer,Francis L Delmonico,Korkut Uygun,Heidi Yeh,James F Markmann
DOI: https://doi.org/10.1097/TXD.0000000000001609
2024-03-12
Abstract:Background: Brief normothermic machine perfusion is increasingly used to assess and recondition grafts before transplant. During normothermic machine perfusion, metabolic activity is typically maintained using red blood cell (RBC)-based solutions. However, the utilization of RBCs creates important logistical constraints. This study explored the feasibility of human kidney normothermic perfusion using William's E-based perfusate with no additional oxygen carrier. Methods: Sixteen human kidneys declined for transplant were perfused with a perfusion solution containing packed RBCs or William's E medium only for 6 h using a pressure-controlled system. The temperature was set at 37 °C. Renal artery resistance, oxygen extraction, metabolic activity, energy metabolism, and histological features were evaluated. Results: Baseline donor demographics were similar in both groups. Throughout perfusion, kidneys perfused with William's E exhibited improved renal flow (P = 0.041) but similar arterial resistance. Lactic acid levels remained higher in kidneys perfused with RBCs during the first 3 h of perfusion but were similar thereafter (P = 0.95 at 6 h). Throughout perfusion, kidneys from both groups exhibited comparable behavior regarding oxygen consumption (P = 0.41) and reconstitution of ATP tissue concentration (P = 0.55). Similarly, nicotinamide adenine dinucleotide levels were preserved during perfusion. There was no evidence of histological damage caused by either perfusate. Conclusions: In human kidneys, William's E medium provides a logistically convenient, off-the-shelf alternative to packed RBCs for up to 6 h of normothermic machine perfusion.
What problem does this paper attempt to address?