Screening for mitochondrial function before use-routine liver assessment during hypothermic oxygenated perfusion impacts liver utilization

Abstract

Background: To report on a concept of liver assessment during ex situ hypothermic oxygenated perfusion (HOPE) and its significant impact on liver utilization.

Methods: An analysis of prospectively collected data on donation after circulatory death (DCD) livers, treated by HOPE at our institution, during a 11-year period between January 2012 and December 2022.

Findings: Four hundred and fifteen DCD Maastricht III livers were offered during the study period in Switzerland, resulting in 249 liver transplants. Of those, we performed 158 DCD III liver transplants at our institution, with 1-year patient survival and death censored graft survival (death with functioning graft) of 87 and 89%, respectively, thus comparable to benchmark graft survivals of ideal DBD and DCD liver transplants (89% and 86%). Correspondingly, graft loss for primary non-function or cholangiopathy was overall low, i.e., 7/158 (4.4%) and 11/158 (6.9%), despite more than 82% of DCD liver grafts ranked high (6-10 points) or futile risk (>10 points) according to the UK-DCD score. Consistently, death censored graft survival was not different between low-, high-risk or futile DCD III livers. The key behind these achievements was the careful development and implementation of a routine perfusate assessment of mitochondrial biomarkers for injury and function, i.e., release of flavin mononucleotide from complex I, perfusate NADH, and mitochondrial CO₂ production during HOPE, allowing a more objective interpretation of liver quality on a subcellular level, compared to donor derived data.

Interpretation: HOPE after cold storage is a highly suitable and easy to perform perfusion approach, which allows reliable liver graft assessment, enabling surgeons to make a fact based decision on whether or not to implant the organ. HOPE-treatment should be combined with viability assessment particularly when used for high-risk organs, including DCD livers or organs with relevant steatosis.

Funding: This study was supported by the Swiss National Foundation (SNF) grant 320030_189055/1 to PD.

Keywords: Donation after circulatory death; Flavin mononucleotide; Hypothermic oxygenated perfusion; Liver utilization.

Fig. 1

Number of DCD III liver offers and number of implanted or discarded DCD III livers in Switzerland 2012–2022. The number of offered DCD III livers in Switzerland increased continuously in the study period (a). While until 2017, all liver were offered only to one center (Zurich), all three centers participate in the DCD III liver transplant program since 2018 (b). The DCD III utilization decreased during the study period due to introduction of ex situ assessment before transplantation (c).

Fig. 2

Risk factors in implanted and discarded DCD III livers in Switzerland. Donor age, donor BMI, and donor warm ischemia times were not different between implanted and discarded DCD III livers (a). The percentage of implanted livers outside clinical cutoffs was high (94/158, 59.5%), and not different from the percentage of discarded livers within clinical cutoffs (29/59, 50%) (b). Eighty percent of implanted DCD III livers qualified as high risk or futile according to the UK DCD risk score (c). P values refer to the Mann–Whitney U test.

Fig. 3

Outcome in DCD III livers implanted in Zurich 2012–2022. DCD III livers showed overall low liver enzyme release and excellent liver graft function (a). Graft loss per year was low during the first 5 years and increased from 2017 to 2020, due to use of DCDIII livers also in sick and retransplant recipients (b). Overall death censored graft survival was 89 and 85% at 1 and 5 years, respectively (c).

Fig. 4

Examples of DCD III livers assessed during HOPE. Sixteen examples of assessed livers during HOPE are shown, with twelve livers, exceeding clinical thresholds for donor age or fDWIT (>65 y, >30 min). Five of these livers (42%) were implanted based on low mitochondrial injury within the first hour of HOPE (red encircled, FMN ≤ 6000 A.U., i.e., 0.04 μg/g liver). In contrast, four livers demonstrated no clinical signs for risk, but the perfusate assessment resulted in unexpected high values for mitochondrial injury (FMN > 6000 A.U.), leading to discard of these livers (blue encircled).

Fig. 5

Excitation and emittance wavelengths of NADH and FMN, as well as perfusate fluorescence during HOPE. The excitation wavelengths for NADH were between 310 and 390 and for FMN between 400 and 470 nm. The emitted light was detected at 450–490 for NADH and at 505–570 nm for FMN (a). Perfusate fluorescence was detected by plate reader analysis or real time during HOPE (b).

Fig. 6

Mitochondrial and perfusate metabolites, correlation of perfusate FMN and liver histology. Mitochondrial FMN content was significantly lower in discarded (n = 5) compared to implanted livers (n = 7), and inversely correlated with high (n = 7) and low perfusate FMN (n = 110) (a). Mitochondrial FMN highly correlated with mitochondrial ATP (b) (Pearson correlation coefficents). Perfusate FMN and purine metabolites were significantly different between implanted and discarded livers (n = 33 each) (c). With increasing perfusate FMN, the degree of inflammation increased, as quantified by necrosis (HE staining), oxidative DNA injury (OHdG staining), and Kupffer cell activation (CD68 staining) (d,e). Presented p values refer to the Mann–Whitney U test.

Fig. 7

Example of a discarded human liver based on perfusate fluorescence and subsequent ex situ normothermic reperfusion. Example of a discarded human DCD liver exceeding the FMN threshold (a), and subsequent cannulation (b) and normothermic ex situ reperfusion for 3 days, with high liver enzyme release and bilirubin increase (c) and large amounts of liver necrosis (circle) (d).

Fig. 8

Mechanism of mitochondrial injury and assessment of mitochondrial function during HOPE. Mechanism of mitochondrial injury during ischemia reperfusion induced by HOPE. Accumulated succinate and NADH during oxygenated reperfusion induce electron overflow at complex I and release of FMN at complex I (a). The level of complex I derived FMN or NADH can be detected by means of perfusate fluorescence (a). Detection is possible because of the natural fluorescence properties of FMN and NADH with excitation wavelengths between 310–390 nm and 400–470 nm, and emission wavelengths at 450–490 nm and 505–570 nm, respectively (b). Perfusate measurements of HOPE treated DCD III livers (n = 37) are shown with levels below (open square) or above (red circle) the arbitrary cutoff of 6000 A.U. (arbitrary units). An increased CO₂ activity in the first 30 min (c) was detected in cases with increased perfusate FMN (d), with correlation of produced CO₂ and perfusate FMN (e); CO₂ produced during HOPE was detected at the oxygenator outlet with laser spectrometry (C¹³ORlab, Argos GmbH). Presented p values refer to the Mann Withney U test.

Fig. 9

Correlation of complex I function and loss of FMN, correlation of perfusate FMN with perfusate NADH. Complex I function decreased with increasing FMN loss (a). High perfusate FMN correlated with high perfusate NADH (b) (Pearson correlation coefficients).