Metabolic profiling during ex vivo machine perfusion of the human liver

Abstract

As donor organ shortages persist, functional machine perfusion is under investigation to improve preservation of the donor liver. The transplantation of donation after circulatory death (DCD) livers is limited by poor outcomes, but its application may be expanded by ex vivo repair and assessment of the organ before transplantation. Here we employed subnormothermic (21 °C) machine perfusion of discarded human livers combined with metabolomics to gain insight into metabolic recovery during machine perfusion. Improvements in energetic cofactors and redox shifts were observed, as well as reversal of ischemia-induced alterations in selected pathways, including lactate metabolism and increased TCA cycle intermediates. We next evaluated whether DCD livers with steatotic and severe ischemic injury could be discriminated from 'transplantable' DCD livers. Metabolomic profiling was able to cluster livers with similar metabolic patterns based on the degree of injury. Moreover, perfusion parameters combined with differences in metabolic factors suggest variable mechanisms that result in poor energy recovery in injured livers. We conclude that machine perfusion combined with metabolomics has significant potential as a clinical instrument for the assessment of preserved livers.

Figure 1. Schematic representation of the SNMP system.

Two parallel circulations provide continuous flow of oxygenated perfusate to the liver through the portal vein (left) and hepatic artery (right). Flow and pressure on the vessels are monitored continuously. Oxygenation of the perfusate was provided by membrane oxygenators using carbogen gas(95% O₂/5% CO₂) mixture. The temperature is ambiently maintained at 21 ± 0.3 °C. The pressure of the system is controlled by flow adjustments to achieve target pressures of 3–7 mmHg and 30–80 mmHg on the PV and HA, respectively.

Figure 2. Characterization of function and injury.

(A) Cumulative production of bile. Bile flow was measured hourly and was sustained for 3 hours of SNMP (n = 21, mean ± sem) (B) Oxygen uptake from the perfusate. The oxygen uptake rate calculated from in- and outflow partial oxygen (pO₂) increased substantially during the first two hours of SNMP preservation (n = 21, mean ± sem (shaded)) (C) Clearance of indocyanine green (ICG). The postperfusion clearance of ICG was reduced from reference values (blue shaded) under ex vivo subnormothermic conditions (n = 7, mean ± sem (shaded)). (D) Release of alanine aminotransferase. Significant ALT release was observed in the first 30 minutes, after which release of ALT was reduced to a minimum for the remainder of the SNMP (n = 21).

Figure 3. Analysis of adenosine and redox cofactors in human liver tissue.

(A) ATP concentration in biopsies from SNMP- preserved liver grafts. Absolute ATP content increased significantly during SNMP preservation (n = 9, P = 0.03) (B) NADH:NAD ratios in human liver biopsies before (n = 9) and after (n = 9) SNMP compared to fresh samples (n = 12). The ratio NADH:NAD was unchanged during SNMP (n = 9) and did not significantly differ from fresh samples (n = 12) at either time point (P ≥ 0.05) (C) Estimated FAD were increased after ischemia, relative to fresh samples (P = 0.03) and remained unchanged during perfusion (n = 9, P ≥ 0.05) (D) NADPH:NADP ratios increased significantly during SNMP (n = 9), and were restored to fresh liver levels (n = 12; P = 0.0012). Bars represent mean ± sem.

Figure 4. Metabolite changes of key metabolic pathways.

(A) Relative values of detected glycolytic intermediates; Glu-6-P, glucose-6-phosphate, Fru-6-P, fructose-6-phosphate; 3-PG, 3-phospoglycerate. Glu-6-P and Fru-6-P decreased significantly during perfusion (P = 0.03, P = 0.015, respectively), while 3-PG showed an insignificant increase (P = 0.09) (B) Relative values of lactic acid in tissue and perfusate concentration of lactate. A large release of lactate was observed in the first hour, after which both tissue lactic acid and perfusate lactate levels decreased significantly (P = 0.004). (C) Relative values of intermediates of the TCA cycle. Significant increases are observed in levels of citrate, α-ketoglutarate and succinate (P < 0.05). Changes in these metabolic pathways represent a reversal of ischemic metabolic shifts and suggest an increase in TCA cycle activity. Bars represent mean ± sem.

Figure 5. Injury markers between control DCD, steatotic, severely warm ischemic livers.

(A) Levels of alanine transaminase (ALT), (B) potassium, and (C) lactate in the perfusate throughout 3 hours of SNMP. Dashed lines reflect normal blood reference values for potassium. Data shown as mean ± sem.

Figure 6. Metabolomic profile of ischemic and steatotic injury.

(A) Heatmap visualization of untargeted metabolomic assessment of preperfusion and postperfusion liver grafts. DCD livers with a WIT <30 min (DCD <30 min) are compared to DCD livers with severe warm ischemia (DCD >30 min). The heatmap on the left shows the profile preperfusion and the right shows the profile postperfusion. (B) DCD <30 min livers are compared to steatotic DCD livers (right). From left to right, the macrosteatosis scores are 0, 0, 0, 2, 3, 3. The heatmap on the left shows the profile preperfusion and the right shows the profile postperfusion. Individual livers are arranged in columns with increasing warm ischemia from left to right, metabolites are ordered by slope correlating relative metabolite levels to ischemia or steatosis. The color gradient used to color the entries ranges from green to red, corresponding to relatively low and high abundance of the metabolite respectively. Both DCD >30 min and steatotic DCD livers could be discriminated from DCD <30 min livers based on their visual metabolomic pattern.

Figure 7. Multivariate analysis of pre- and postperfusion metabolomic profile.

Principal component analysis of DCD <30 min (green), DCD >30 min (black), and steatotic DCD livers (grey) pre- (open circles) and postperfusion (filled circles), providing a 2-dimensional (PC1/PC2) representation of the untargeted metabolomic profile. Clustering of control livers is seen, both pre- and postperfusion. All but one of the steatotic DCD and DCD >30 min livers show separation from the DCD <30 min livers, lying outside the 95% confidence interval of the pre- and postperfusion controls (black ellipses).

Figure 8. Mitochondrial respiratory function and ATP recovery.

(A) Absolute ATP content at the end of perfusion in the different groups (n = 3). Final ATP levels were significantly lower in the DCD >30 min and steatotic DCD groups (P < 0.02). (B) Oxygen uptake rates between groups throughout perfusion and total oxygen consumption, determined by the area under the curve. Total oxygen uptake was significantly lower in steatotic DCD livers, compared to the non-steatotic livers (DCD >30 min and DCD <30 min) (P < 0.03). (C) Change in ratio of oxidized to reduced glutathione (GSSG:GSH) during perfusion (t3/t0) in the different groups as an indicator of oxidative stress. DCD >30 min livers exhibited the largest increase in GSSG:GSH, with a significantly higher ratio change than the DCD <30 min group (P = 0.007). (D) Change in ratio of NADH to NAD⁺ during perfusion in different groups. DCD <30 min and DCD >30 min groups show minimal change in NADH:NAD, while steatotic DCD livers exhibit a significant increase (P = 0.046). (E) Arterial resistance at the end of perfusion and flow through the portal vein (PV) and hepatic artery (HA). Arterial resistance was significantly higher in the steatotic DCD livers (P < 0.05), but the difference in flow through the arterial or portal system did not reach statistical significance (shown as mean ± sem). (F) Structural and ultrastructural preservation. Transmission electron microscopy images of representative DCD <30 min, DCD >30 min, and steatotic DCD livers (top). Corresponding H&E photomicrographs are shown in the bottom row. m, mitochondria; Nucl, nucleus; lipid; intracellular lipid droplet. (G) Mitochondrial injury scoring. Injury is scored based on the appearance of cristae, swelling and intramitochondrial deposition of amorphous material. Mitochondrial injury was increased in both DCD >30 min livers and steatotic DCD livers compared to DCD <30 min livers (n = 2 × 20 cells, mean ± sem) (P < 0.01).