Does hypothermic machine perfusion of human donor livers affect risks of sinusoidal endothelial injury and microbial infection? A feasibility study assessing flow parameters, sterility, and sinusoidal endothelial ultrastructure

Abstract

Background: Hypothermic machine perfusion (HMP) is better than conventional cold storage in kidney transplantation. Large animal models suggest that HMP may be beneficial for the liver as well, but questions remain about perfusion mode (dual portal/arterial flow versus single flow) and hepatic vascular injury including endothelial dysfunction or potential microbial infectivity during HMP.

Methods: Sixteen human livers rejected for transplantation by all UK centers with appropriate consent for research were randomized into 4 groups (n = 4 each): group 1: ≥7 hours cold storage (CS) and 1 hour HMP through hepatic artery (HA) alone; group 2: ≥7 hours CS and 1 hour HMP through HA and portal vein (PV); group 3: ≥7 hours CS and 1 hour HMP through PV alone; and group 4: ≥8 hours CS. A pressure-controlled prototype based on Lifeport Kidney Transporter (Organ Recovery Systems) was used. Livers were perfused at 4-8°C under sterile conditions with Belzer MPS KPS-1. Perfusion parameters (pressure, flow, resistance, and temperature) were recorded every 15 minutes. Perfusate for microbial culture and sensitivity were taken before and after HMP. Electron microscopy of 3 liver biopsy samples taken before perfusion, were compared with 3 samples from adjacent areas after perfusion.

Results: Preset HA pressure of 30 mm Hg and PV pressure of 7 mm Hg were maintained throughout the perfusion. HA and PV flow ranged, respectively, from 11 to 107 mL/min (mean 59.5) and 39 to 199 mL/min (mean 96.2), with no differences between groups. The same was true for resistance: HA and PV resistance ranged, respectively from 0.17 to 1.99 mm Hg/mL/min (mean 0.71) and 0.07 to 0.17 mm Hg/mL/min (mean 0.08). Temperature was maintained at 4-8°C with the use of an external heat exchanger. No difference in sinusoidal endothelial ultrastructure was seen before and after machine perfusion or between any of the groups. Sterility was maintained throughout the HMP.

Conclusions: HMP of human livers did not produce evidence of sinusoidal endothelial injury or breach of sterility. Single or dual perfusion modes did not affect vascular resistance or flow. The results suggest that further studies of HMP with human livers are warranted.