Real-world implementation of normothermic machine perfusion: A detailed analysis of intraoperative and early postoperative impact.

Abstract

Outcome data for the great majority of liver normothermic machine perfusion (NMP) cases derive from the strict confines of clinical trials. Detailed specifics regarding the intraoperative and early postoperative impact of NMP on reperfusion injury and its sequelae during real-world use of this emerging technology remain largely unavailable. We analyzed transplants performed in a 3-month pilot period during which surgeons invoked commercial NMP at their discretion. Living donor, multi-organ, and hypothermic machine perfusion transplants were excluded. Intraoperatively, NMP (n=24) compared to static cold storage (n=25) recipients required less peri-reperfusion bolus epinephrine (0vs. 60μg; p<.001) and post-reperfusion fresh frozen plasma (2.5vs. 7.0 units; p=.0069), platelets (.0vs. 2.0 units; p=.042), and hemostatic agents (0%vs. 24%; p=.010). Time from incision to venous reperfusion did not differ (3.6vs. 3.1; p=.095) but time from venous reperfusion to surgery end was shorter for NMP recipients (2.3vs. 2.8h; p=.0045). Postoperatively, NMP recipients required fewer red blood cell (1.0vs. 4.0 units; p=.0083) and fresh frozen plasma (4.0vs. 7.0 units; p=.046) transfusions, had shorter intensive care unit stays (33.5vs. 58.4h; p=.012), and experienced less early allograft dysfunction according to both the Model for Early Allograft Function Score (3.4vs. 5.0; p=.0047) and peak AST within 10 days of transplant (619vs. 1,181U/L; p=.036). Liver acceptance for the corresponding recipient was conditional on NMP use for 63% (15/24) of cases. Real-world NMP use was associated with significantly lower intensity of reperfusion injury and intraoperative and postoperative care that may translate into patient benefit.