Machine perfusion keeps a donor liver alive and testable outside the body
The preservation clock starts the instant a donor organ loses its blood supply. Rapid in situ flushing with an ice-cold preservation solution is the single most important step in classical organ banking — it arrests warm ischemic injury, clears the vasculature of blood and clotting factors, and buys the hours needed to transport, match and transplant the organ.
Once circulation stops, cells are deprived of oxygen and switch to anaerobic glycolysis, rapidly depleting ATP and accumulating lactate and hydrogen ions. This "warm ischemia time" (WIT) is the most damaging phase of organ injury — every additional minute at body temperature accelerates cell death, and for donation-after-circulatory-death (DCD) donors, WIT begins even before formal death is declared.
Surgeons race to cannulate the aorta (and portal vein for liver) and infuse cold flush solution, which does two things simultaneously: it displaces warm blood with an ice-cold, non-clotting fluid, and it reduces core temperature by roughly 30°C in a matter of minutes. The abrupt drop in temperature is itself protective — it is the fastest, cheapest intervention available to slow the biochemical clock before any machine can be connected.
Cold alone is not enough. Hypothermia inhibits the Na⁺/K⁺-ATPase pump that normally keeps cells from swelling, so a plain saline flush would leave cells to absorb water and rupture. University of Wisconsin (UW) solution — the historic gold standard — is formulated to mimic the intracellular ionic environment and counteract this failure mode:
• Lactobionate & raffinose: large, impermeant sugars that create osmotic pressure to prevent cell swelling during hypothermia • Hydroxyethyl starch (HES): a colloid that limits interstitial and endothelial cell edema • Adenosine: substrate reserve for rapid ATP resynthesis on reperfusion • Allopurinol: xanthine oxidase inhibitor, limiting reactive oxygen species generation • Glutathione: antioxidant buffer against the reperfusion oxidative burst • High potassium, low sodium: matches intracellular electrolyte composition to blunt ionic gradients
Histidine-tryptophan-ketoglutarate (HTK) solution is a lower-viscosity alternative favored for faster, more homogeneous flush, particularly in DCD and pediatric procurement.
UW solution's colloid and impermeant-sugar combination was engineered specifically to prevent the cold-induced cell swelling that plain electrolyte flushes cause — a design insight from the 1980s that still underlies nearly every modern preservation and perfusate formulation.
For liver procurement, cannulae are placed in the aorta (and often the portal vein for a dual flush) while the abdomen is packed in ice. Flush is delivered by gravity or light pressure to avoid barotrauma to the endothelium. The inferior vena cava is vented to allow outflow, and the organ visibly blanches as blood is displaced by the pale, cold solution.
On the back table, the surgical team trims vascular pedicles, inspects for anatomical variants, and performs a second flush before packaging the organ in sterile bags surrounded by ice slurry for transport — the state that defines classical static cold storage, the baseline against which all machine perfusion technologies are measured.
Simple, cheap, and effective for decades: place the flushed organ in an ice-filled cooler and transport it. Static cold storage (SCS) revolutionized transplantation by slowing metabolism, but hypothermia is a blunt instrument — it decelerates injury without stopping it, and it offers no way to know how the organ is actually doing until it is warmed up inside the recipient.
The Q10 rule of thumb states that biochemical reaction rates fall by a factor of roughly 2–3 for every 10°C drop in temperature. Cooling from 37°C to 4°C therefore reduces oxygen consumption to somewhere near 5–10% of baseline — dramatic, but not zero.
Residual anaerobic metabolism continues to consume the last available ATP, generating lactate and hydrogen ions that acidify the cytoplasm. As ATP falls, the Na⁺/K⁺-ATPase pump fails, sodium and water flood into cells, and mitochondria slowly lose membrane potential. None of this is visible from outside an ice cooler — cold storage is metabolically a black box that is quietly getting worse the entire time it sits on ice.
Paradoxically, much of the injury attributed to "cold storage" actually manifests at reperfusion, when the organ is rewarmed and reconnected to the recipient's circulation:
• Mitochondrial permeability transition pore (MPTP) opens on reoxygenation, collapsing the proton gradient and triggering cell death pathways • A burst of reactive oxygen species (ROS) is generated as electron transport chain components, primed during ischemia, suddenly meet abundant oxygen • Kupffer cells (liver-resident macrophages) and complement activation drive a sterile inflammatory cascade • Microvascular "no-reflow" — swollen endothelial cells and sinusoidal narrowing — can block blood from reaching regions of the organ even after the main vessels are reconnected
The longer the cold ischemic time, the more primed the tissue is for this reperfusion injury cascade — meaning cold storage duration and reperfusion severity are directly linked.
Marginal organs — steatotic ("fatty") livers, older donors, and donation-after-circulatory-death (DCD) organs that already sustained a warm ischemic insult — tolerate cold ischemia far worse than standard organs, which is why they are disproportionately discarded under a cold-storage-only paradigm.
The defining limitation of static cold storage is informational, not just biological: a hypothermic, non-perfused organ performs essentially none of its physiological functions, so clinicians have no direct signal of how well it will work after transplant.
Decisions are instead made from indirect proxies — donor age, cause of death, warm ischemia time, biopsy-based steatosis grading, and historical statistics about similar organs. These proxies correlate imperfectly with actual post-transplant function, and when the risk looks too high, usable organs are conservatively declined. This uncertainty is a central driver of the persistent gap between the number of organs procured and the number ultimately transplanted — a gap that ex vivo machine perfusion was specifically developed to close.
Instead of sitting inert on ice, the organ is cannulated onto a closed-loop perfusion circuit: a pump drives oxygenated perfusate through a membrane oxygenator, a heat exchanger brings it to the target temperature, and it flows through the organ's own vasculature exactly as blood would in the body — before draining to a reservoir to recirculate.
A modern liver perfusion device typically includes:
• A centrifugal or roller pump generating physiological flow and pressure, often with separate circuits for the low-pressure portal vein and higher-pressure hepatic artery • A hollow-fiber membrane oxygenator that diffuses oxygen (and removes CO₂) across a gas-permeable membrane, mimicking the lung • A heat exchanger to precisely control perfusate temperature • A leukocyte filter to reduce inflammatory cell load in blood-based perfusate • A reservoir that acts as a venous return buffer and access point for sampling and drug/nutrient additions
Dual perfusion — separately cannulating both the portal vein and hepatic artery, as the body does — better reproduces the liver's unique dual blood supply (roughly 70% portal, 30% arterial) than single-vessel perfusion.
Unlike static storage, which is binary (cold or not), machine perfusion spans a temperature spectrum, each tier serving a different purpose:
• Normothermic machine perfusion (NMP), ~37°C: restores near-full metabolic activity, enabling genuine functional testing — bile production, lactate clearance, and hemodynamics can be measured exactly as they would occur in vivo • Subnormothermic perfusion, ~20–25°C: partially active metabolism at reduced oxygen demand, used as an intermediate resuscitation step • Hypothermic oxygenated perfusion (HOPE), ~8–12°C: metabolism stays low, but continuous oxygen delivery lets mitochondria replenish ATP and clear succinate accumulated during ischemia — "priming" the organ to blunt the reperfusion ROS burst rather than testing function
Some protocols combine modes — e.g., a short period of HOPE followed by NMP viability testing — to capture both the mitochondrial-protective benefit of cold oxygenation and the functional read-out of warm perfusion.
The concept now extends before procurement too: normothermic regional perfusion (NRP) restores circulation inside the donor's body immediately after circulatory death, and "back-to-base" transport perfusion keeps the organ on a portable pump throughout transit rather than only at the receiving center.
Connecting a cold organ directly to a warm circulation is itself a mechanism of injury — the abrupt thermal and oxidative shock is a major contributor to early graft dysfunction. Machine perfusion allows gradual, controlled rewarming over 20–60 minutes before the organ ever reaches full physiological temperature, giving mitochondria and antioxidant systems time to re-equilibrate rather than being shocked all at once.
This controlled transition — cannulation, gradual warming, gas and pressure titration — is itself a clinically meaningful intervention, independent of any of the diagnostic information perfusion goes on to provide.
This is the defining advantage of normothermic perfusion: instead of inferring organ quality from donor history, clinicians watch the liver actually perform its physiological jobs in real time. Bile secretion, lactate metabolism, acid-base handling and vascular hemodynamics become live, continuously sampled data streams rather than static risk factors.
Healthy hepatocytes clear lactate from the perfusate via gluconeogenesis (the Cori cycle), a highly ATP-dependent process that only functions if mitochondria are intact. A liver that fails to clear lactate below roughly 2.5 mmol/L within the first hours of normothermic perfusion has a very high likelihood of primary non-function after transplant.
This single metric — trending lactate downward over successive perfusate samples — has become the anchor criterion in essentially every published liver NMP viability protocol, including the widely used Birmingham/UK and Toronto criteria, precisely because it is fast, cheap to measure, and mechanistically tied to core hepatocellular energetics.
Bile production requires coordinated function of both hepatocytes (bile synthesis and secretion) and cholangiocytes (bile duct lining cells that modify its composition) — making it a uniquely informative combined biomarker:
• Production rate: sustained output above roughly 10 mL/h suggests adequate hepatocellular synthetic function • Bile pH and bicarbonate: cholangiocytes actively secrete bicarbonate into bile; a bile pH above perfusate pH (a positive "bile pH gradient") indicates intact ductal function • Bile glucose: healthy cholangiocytes reabsorb glucose from bile, so low bile glucose relative to perfusate glucose is reassuring, while a high concentration suggests ductal injury
Biliary viability testing during NMP is particularly important because ischemic cholangiopathy — bile duct strictures from inadequate peribiliary vascular perfusion — is a leading cause of graft loss after transplantation of DCD livers, and bile chemistry during perfusion is a much earlier predictor than anything cold storage can offer.
Continuous flow and pressure monitoring reveals microvascular health directly: rising portal and arterial flow at stable, physiological pressures (i.e., falling vascular resistance) over the course of perfusion indicates a relaxing, functional sinusoidal bed, while persistently high resistance suggests endothelial swelling or microthrombosis.
Simultaneously, enzymes leaking into the perfusate — aspartate and alanine aminotransferase (AST/ALT) and lactate dehydrogenase (LDH) — quantify the extent of hepatocellular necrosis, while glucose metabolism (net consumption versus release) and potassium trends add further texture to the picture. No single number is decisive; modern protocols combine several trending biomarkers into a composite viability score rather than relying on any one threshold.
Because these are continuously trending signals rather than one-off snapshots, clinicians can watch an organ improve over the course of perfusion — a marginal liver that looks poor at 1 hour but shows falling lactate and rising bile output by hour 4 may still be entirely transplantable.
| Product | Indication | Trial Design | Key Result |
|---|---|---|---|
| Perfusate lactate | Clearance to <2.5 mmol/L by ~2 h | Mitochondrial gluconeogenic capacity (Cori cycle) | Strongest single predictor of post-transplant function |
| Bile production & pH | >10 mL/h, bile pH >7.45 | Combined hepatocyte synthesis + cholangiocyte bicarbonate secretion | Predicts ischemic cholangiopathy risk early |
| Perfusate pH / bicarbonate | 7.35–7.45, stable without correction | Buffering capacity reflects overall metabolic acid handling | Flags global metabolic failure quickly |
| Vascular flow & resistance | Rising flow, falling resistance | Sinusoidal endothelial and microvascular integrity | Detects microthrombosis / no-reflow in real time |
| AST / ALT / LDH leakage | Low and plateauing, not rising | Hepatocellular membrane integrity / necrosis extent | Quantifies cumulative cell injury directly |
Once functional data accumulates, the perfusion team faces a decision point unavailable to cold storage: transplant the organ now, continue perfusing to gather more data or let borderline function improve, or — in emerging protocols — actively treat and recondition a marginal organ while it is still on the pump.
Normothermic machine perfusion has moved rapidly from experimental technique to clinical adoption over the past decade. The VITTAL trial (Birmingham, UK) applied normothermic perfusion specifically to livers that transplant centers had already declined as too high-risk for cold-storage transplantation. Using a composite viability protocol — lactate clearance below 2.5 mmol/L by 4 hours plus at least two of: perfusate pH >7.30, bile production, glucose metabolism, and adequate hepatic artery flow — the trial transplanted a majority of these previously-discarded organs with outcomes comparable to standard-criteria transplants.
Randomized trials of both liver and kidney NMP/hypothermic perfusion (including large multi-center European and North American programs) have consistently shown reduced rates of early allograft dysfunction and delayed graft function versus static cold storage, alongside meaningful increases in the proportion of procured organs that are ultimately usable.
Beyond viability testing, simply extending safe preservation time from single-digit hours to a day or more reshapes how transplantation is scheduled and delivered:
• Long-distance procurement becomes feasible without racing the cold-ischemia clock • Surgery can be scheduled for daytime, fully-staffed operating rooms rather than emergency overnight cases • Procurement and transplantation can be decoupled in time, allowing better recipient preparation and cross-matching • Perfusion time itself becomes usable — for continued monitoring, treatment, or simply logistics buffer — rather than being pure risk accrual as it is with cold storage
Perhaps the most forward-looking application of machine perfusion is active therapeutic intervention on the organ itself while it remains outside the body:
• Defatting protocols: steatotic ("fatty") livers — a common reason for donor organ decline — have been treated during NMP with lipid-lowering perfusate additives (e.g., forskolin, L-carnitine, and defatting cocktails) that measurably reduce intrahepatocellular triglyceride content over hours of perfusion • Pharmacological and gene therapies: perfusion allows organ-specific drug or vector delivery at concentrations that would be systemically toxic if given to the recipient • Infection clearance: donor organ-specific infections can potentially be treated with antimicrobial-supplemented perfusate before transplantation, isolated entirely from the recipient • Immunomodulation: depleting donor leukocytes or modulating antigenicity during perfusion, aiming to reduce rejection risk before the organ ever meets the recipient's immune system
Every year, a substantial share of procured organs are never transplanted — discarded due to uncertainty about their quality under a cold-storage-only paradigm. Machine perfusion's combination of objective functional testing and active reconditioning directly targets this utilization gap, a major lever against transplant waitlist mortality.
Active research areas building on ex vivo perfusion include:
• Xenoperfusion: using genetically modified pig blood, or perfusion techniques developed for organ preservation, to bridge and condition xenotransplant organs before implantation • Portable and wearable perfusion devices: miniaturized pumps and oxygenators enabling perfusion during transport with minimal footprint • Multi-day subnormothermic storage: research into stable intermediate-temperature perfusion for preservation windows measured in days rather than hours • AI-driven predictive viability modeling: integrating continuous multi-parameter perfusion data streams to predict post-transplant outcomes earlier and more accurately than any single threshold
Taken together, these directions point toward organ preservation evolving from a passive shipping problem into an active extension of clinical care — treating, testing, and even improving organs before they ever reach the recipient.