Anticoagulant Management and Synthesis of Hemostatic Proteins during Machine Preservation of Livers for Transplantation

Normothermic Machine-perfused Human Donor Livers Produce Functional Hemostatic Proteins

BACKGROUND

Normothermic machine perfusion (NMP) is used for the viability assessment of high-risk donor livers before transplantation. The production of hemostatic proteins is one of the major synthetic functions of the liver. The objective of this study was to measure the concentration and functionality of hemostatic proteins concentration in the NMP perfusate of human donor livers.

METHODS

Thirty-six livers that underwent NMP for viability assessment were included in this study. Perfusate samples taken during NMP (start, 150 min, and 300 min) were used for the measurement of antigen and activity levels of hemostatic proteins (factors II, VII, and X; fibrinogen; plasminogen; antithrombin; tissue plasminogen activator; von Willebrand factor; and proteins induced by vitamin K absence). The antigen levels were correlated with hepatocellular function according to previously proposed individual hepatocellular viability criteria: lactate clearance and perfusate pH.

RESULTS

Antigen levels of hemostatic proteins reached subphysiological levels in the NMP perfusate. Hemostatic proteins that were produced during NMP were at least partially active. All livers produced all hemostatic proteins tested within 150 min of NMP. Hemostatic protein concentrations did not significantly correlate with perfusate lactate and perfusate pH after 150 min of NMP.

CONCLUSIONS

All livers produce functional hemostatic proteins during NMP. The generation of a functional hemostatic system in NMP perfusate confirms the need for adequate anticoagulation of the perfusate to avoid generation of (micro)thrombi that may harm the graft.

The international normalised ratio to monitor coagulation factor production during normothermic machine perfusion of human donor livers

BACKGROUND

Normothermic machine perfusion (NMP) of donor livers allows for new diagnostic and therapeutic strategies. As the liver produces most of the haemostatic proteins, coagulation assays such as the International Normalised Ratio (INR) performed in perfusate may be useful to assess hepatocellular function of donor livers undergoing NMP. However, high concentrations of heparin and low levels of fibrinogen may affect coagulation assays.

METHODS

Thirty donor livers that underwent NMP were retrospectively included in this study, of which 18 were subsequently transplanted. We measured INRs in perfusate in presence or absence of exogenously added fibrinogen and/or polybrene. Additionally, we prospectively included 14 donor livers that underwent NMP (of which 11 were transplanted) and measured INR using both a laboratory coagulation analyser and a point-of-care device.

RESULTS

In untreated perfusate samples, the INR was above the detection limit in all donor livers. Addition of both fibrinogen and polybrene was required for adequate INR assessment. INRs decreased over time and detectable perfusate INR values were found in 17/18 donor livers at the end of NMP. INR results were similar between the coagulation analyser and the point-of-care device, but did not correlate with established hepatocellular viability criteria.

CONCLUSIONS

Most of the donor livers that were transplanted showed a detectable perfusate INR at the end of NMP, but samples require processing to allow for INR measurements using laboratory coagulation analysers. Point-of-care devices bypass this need for processing. The INR does not correlate with established viability criteria and might therefore have additional predictive value.

Coagulation Factors Accumulate During Normothermic Liver Machine Perfusion Regardless of Donor Type and Severity of Ischemic Injury

BACKGROUND

Coagulation factors may inform on liver function during normothermic machine perfusion (NMP). We investigated whether graft ischemic injury impairs the accumulation of anticoagulation factors during NMP of porcine and human livers.

METHODS

Dynamics of FV, FVII, FVIII, FIX, and FX during NMP and their correlation with graft injury was investigated in porcine livers with minimal (no warm ischemia, n = 5) or severe injury (60 min warm ischemia, n = 5). Next, FV, FVIII, FIX, fibrinogen, and antithrombin were measured in 35 matched human liver NMPs from the COPE trial. Correlation of these factors with outcomes was explored. Livers were categorized in to 4 groups depending on donor type and posttransplant peak aspartate aminotransferase (AST) as surrogate of minimal (peak < 500 IU/L) or moderate injury (peak > 1000 IU/L).

RESULTS

Factor concentrations increased significantly during NMP regardless of severity of injury. In porcine livers, factor concentrations were 2- to 6-fold lower in severely injured grafts (all P < 0.05). All factors negatively correlated with AST (coefficient range: from -0.50 to -0.93; all P < 0.05) and lactate (range: from -0.51 to -0.67; all P < 0.05). In human livers, no difference in factor accumulation rates and no correlation with other markers were observed. One graft with primary nonfunction had low rate of factor accumulation.

CONCLUSIONS

Anticoagulation factors accumulate during NMP regardless of donor type and severity of injury. In pigs, severe ischemic injury resulted in significantly lower factor concentrations. In human livers with life-sustaining function, they do not correlate with hepatic injury. Whether low concentrations predict nonfunction in high-risk livers with severe injury requires further investigation.

Long-term normothermic machine preservation of human livers: what is needed to succeed?

Although short-term machine perfusion (≤24 h) allows for resuscitation and viability assessment of high-risk donor livers, the donor organ shortage might be further remedied by long-term perfusion machines. Extended preservation of injured donor livers may allow reconditioning, repairing, and regeneration. This review summarizes the necessary requirements and challenges for long-term liver machine preservation, which requires integrating multiple core physiological functions to mimic the physiological environment inside the body. A pump simulates the heart in the perfusion system, including automatically controlled adjustment of flow and pressure settings. Oxygenation and ventilation are required to account for the absence of the lungs combined with continuous blood gas analysis. To avoid pressure necrosis and achieve heterogenic tissue perfusion during preservation, diaphragm movement should be simulated. An artificial kidney is required to remove waste products and control the perfusion solution's composition. The perfusate requires an oxygen carrier, but will also be challenged by coagulation and activation of the immune system. The role of the pancreas can be mimicked through closed-loop control of glucose concentrations by automatic injection of insulin or glucagon. Nutrients and bile salts, generally transported from the intestine to the liver, have to be supplemented when preserving livers long term. Especially for long-term perfusion, the container should allow maintenance of sterility. In summary, the main challenge to develop a long-term perfusion machine is to maintain the liver's homeostasis in a sterile, carefully controlled environment. Long-term machine preservation of human livers may allow organ regeneration and repair, thereby ultimately solving the shortage of donor livers.

Synthesis of coagulation factors during long-term ex situ liver perfusion

Robust viability assessment of grafts during normothermic liver perfusion is a prerequisite for organ use. Coagulation parameters are used commonly for liver assessment in patients. However, they are not yet included in viability assessment during ex situ perfusion. In this study, we analysed coagulation parameters during one week ex situ perfusion at 34℃. Eight discarded human livers were perfused with blood-based, heparinised perfusate for one week; perfusions in a further four livers were terminated on day 4 due to massive ongoing cell death. Coagulation parameters were well below the physiologic range at perfusion start. Physiologic levels were achieved within the first two perfusion days for factor V (68.5 ± 35.5%), factor VII (83.5 ± 26.2%), fibrinogen (2.1 ± 0.4 g/L) and antithrombin (107 ± 26.5%) in the livers perfused for one week. Despite the increased production of coagulation factors, INR was detectable only at 24h of perfusion (2.1 ± 0.3) and prolonged thereafter (INR > 9). The prolongation of INR was related to the high heparin level in the perfusate (anti-FXa > 3 U/mL). Intriguingly, livers with ongoing massive cell death also disclosed synthesis of factor V and improved INR. In summary, perfused livers were able to produce coagulation factors at a physiological level ex situ. We propose that single coagulation factor analysis is more reliable for assessing the synthetic function of perfused livers as compared to INR when using a heparinised perfusate.