In vivo evaluation of an implantable portal pump system for augmenting liver perfusion

Background: Increasing portal inflow in cirrhosis using a mechanical pump r educes portal venous pressure and improves liver function. A pump has been developed for portal vein implantation in human cirrhosis. This study descr ibes the initial in vivo evaluation in a porcine model. Methods: Five Large White pigs underwent laparotomy and exposure of the liv er. Flow in the hepatic artery, portal vein and hepatic microcirculation wa s monitored continuously. Hepatic tissue oxygenation was measured by near-i nfrared spectroscopy. After baseline measurements the pump was inserted int o the portal vein. Pump flow rate was then increased stepwise to 50 per cen t over the baseline value for a period of 2 h. The pump was then stopped fo r 20 min and left in sift! while continuing to collect systemic and hepatic haemodynamic data. The animal was killed and biopsies for histological exa mination were taken from the liver, small intestine and spleen. Results: The baseline total hepatic blood flow was 626(39) ml/min; the hepa tic artery supplied 18.4(2.1) per cent and the portal vein 81.6(2.1) per ce nt. The pump was inserted successfully in all animals without surgical comp lications. During surgical insertion of the pump, the temporary portal vein occlusion resulted in a significant rise in hepatic artery blood flow (22( 3) per cent; P < 0.01 versus baseline). Portal vein flow was augmented by p umping; there was a significant correlation between the pump motor speed an d portal vein flow (P < 0.0001). This inflow correlated directly with flow in the hepatic microcirculation and hepatic tissue oxygenation (P < 0.001). The pump ran satisfactorily throughout the study. Histological examination revealed no evidence of structural damage to the liver or ischaemic change s in the small intestine or spleen. Conclusion: It is technically possible and safe to insert an implantable pu mp in the portal vein. Portal venous blood flow can be increased up to 50 p er cent with a resultant increase in flow in the hepatic microcirculation a nd hepatic oxygenation and without adverse effects on either hepatic or sys temic haemodynamics.