Donor-Related Risk Factors for Normothermic Machine Perfusion in Liver Transplantation: A Meta-Analysis

Abstract

Background and aims: During normothermic machine perfusion (NMP), a variety of criteria are used to gauge the suitability of an organ for transplantation. However, the relations between donor factors and these criteria are poorly understood. The aim of this meta-analysis was to investigate the association between donor-related risk factors and the decision to transplant a liver subjected to NMP.

Methods: A comprehensive literature search was performed for articles published up to March 2025 in four databases, reporting livers subjected to NMP for viability assessment prior to transplantation. Effect size (ES) was calculated using Cohen's D and log odds ratio.

Results: Out of 806 unique articles, 18 were included in this meta-analysis, encompassing 690 liver grafts that underwent NMP. Following viability assessment during NMP, utilisation rate was 82% from donors after brain death and 68% from donors after circulatory death (ES: 0.08, p = 0.88). Transplanted livers had shorter cold ischemia time (ES: -0.34, p = 0.003) and lower liver weight (ES: -0.53, p < 0.001). Donor age, BMI and donor warm ischemia time did not differentiate between transplanted and unused groups. Differences were observed in viability assessment for lactate clearance (ES: 2.0, p = 0.005), glucose metabolism (ES: 2.2, p < 0.001), bile production (ES: 1.0, p = 0.003) and pH (ES: 1.9, p < 0.001). Excellent outcomes, including 10% non-anastomotic strictures, 89% graft survival and 93% patient survival, were achieved in a large cohort of high-risk livers.

Conclusion: Cold ischemia time and liver weight were identified as donor-related risk factors, whereas donor type, age and donor warm ischemia time appear not to impact the decision to transplant during NMP.

Keywords: diagnosis; metabolism; methods; mortality; therapy; transplantation.

FIGURE 1

Forest plots of donor and donation parameters between livers used for and not used for transplantation. (A) DBD grafts had similar likelihood compared to DCD grafts to be transplanted after being subjected to NMP (ES: 0.08, I ² = 0.68, p = 0.88). (B) Donor age (ES: −0.10, I ² = 0.00, p = 0.29) and (C) donor BMI (ES: 0.08, I ² = 0.79, p = 0.88) were similar between transplanted and unused grafts. (D) CIT was shorter in transplanted livers (ES: −0.33, I ² = 0.00 p = 0.003), whereas, (E) dWIT was comparable between transplanted and unused grafts (ES: 0.02, I ² = 0.00, p = 0.90). (F) Liver weight of transplanted grafts was lower than unused grafts (ES: −0.53, I ² = 0.00, p < 0.001).

FIGURE 2

Successfully meeting viability assessment criteria between transplanted and unused grafts. Meeting arterial (ES: 0.78, I ² = 0.00, p = 0.40) and portal flow (ES: 0.78, I ² = 0.00, p = 0.40) criteria did not differ between transplanted and unused grafts. Evidence of glucose metabolism (ES: 2.2, I ² = 0.00, p < 0.001), lactate clearance (ES: 2.0, I ² = 0.39, p = 0.005), perfusate pH (ES: 2.5, I ² = 0.00, p < 0.001), bile pH (ES: 1.9, I ² = 0.00, p < 0.001) and bile production (ES: 1.0, I ² = 0.00, p = 0.003) were all worse in unused grafts. Forest plots are included in Data S7. pH, potential of hydrogen. ***p < 0.001, **p < 0.01.