For decades, the decision to accept or reject a donor lung has depended on a narrow set of measurements taken under intense time pressure. Oxygenation, airway pressure, lung compliance and the appearance of the organ provide important clues, but they do not always reveal whether a marginal lung can recover. A review republished from the Japanese Journal of Artificial Organs examines how cellular ex vivo lung perfusion, or EVLP, is expanding that diagnostic window. By maintaining donated lungs outside the body under controlled physiological conditions, EVLP turns an organ that would otherwise be judged quickly into a living experimental system—one in which pulmonary function can be tested, monitored and potentially improved before transplantation.
The principle behind EVLP is deceptively simple. After retrieval, the lungs are connected to a circuit that circulates a warm, oxygenated perfusate through the pulmonary artery while the lungs are ventilated through the trachea. The system reproduces selected features of the human cardiopulmonary environment, generally at near-body temperature, without exposing the organ to the metabolic demands of a recipient. Unlike static cold storage, which suppresses cellular activity to slow injury, normothermic perfusion allows the lung to resume metabolism. This makes it possible to observe how the tissue handles oxygen, carbon dioxide, fluid and pressure, and to distinguish potentially reversible injury from irreversible structural damage.
The most established assessment remains gas exchange. During ventilation, oxygen transfer from the alveoli into the perfusate and carbon dioxide removal from the perfusate into the exhaled air can be quantified. The partial pressure of oxygen in the perfusate, often considered alongside the inspired oxygen concentration, provides a practical measure of pulmonary efficiency. Yet oxygenation alone can be misleading. A lung may produce an acceptable oxygen value while developing interstitial edema, high vascular resistance or localized collapse. Conversely, a lung with initially poor gas exchange may improve during perfusion as inflammatory debris is cleared, alveolar units reopen and endothelial function recovers. The review therefore places gas exchange within a broader panel of physiological measurements rather than treating it as a solitary pass-or-fail test.
Mechanical behavior offers another layer of information. Airway pressure, tidal volume and dynamic compliance reveal how easily the lung expands and recoils during ventilation. Rising airway pressures at a fixed tidal volume can indicate worsening edema, airway obstruction or loss of compliance, while improving compliance may signal recruitment of previously collapsed alveoli. Peak and plateau pressures must be interpreted carefully because the isolated lung is affected by ventilator settings, airway resistance and the properties of the perfusion circuit. Pressure–volume relationships can help separate restrictive changes in the parenchyma from problems in the airways. Continuous monitoring is particularly valuable because a single measurement may conceal deterioration that becomes evident only over time.
The pulmonary circulation can also be interrogated directly. Pulmonary vascular resistance is estimated from pulmonary artery pressure, left atrial or venous pressure and perfusion flow. Excessive resistance may reflect vasoconstriction, microvascular obstruction, endothelial injury or thrombotic material. Tracking the relationship between flow and pressure can show whether the vascular bed is becoming more receptive to perfusion or progressively overloaded. At the same time, changes in perfusate composition can expose hidden cellular distress. Lactate accumulation, glucose consumption, electrolyte shifts and acid–base changes provide indirect evidence of metabolic activity. Oxygen consumption, measured by comparing oxygen content before and after passage through the lung, may offer a more direct indication of mitochondrial and cellular function than gas exchange alone.
A major development highlighted by the review is the movement from organ-level physiology toward cellular and molecular analysis. Samples of perfusate can be tested for inflammatory mediators, endothelial injury markers, epithelial damage proteins and signals associated with cell death. Cytokines such as interleukin-6 and interleukin-8 may reflect activation of innate immune pathways, although their interpretation depends on the perfusate, the duration of perfusion and the initial cause of donor injury. Measurements of cell-free DNA, extracellular vesicles and other circulating biomarkers could eventually help identify damage that is not yet visible through pressure or oxygenation changes. These approaches are especially relevant because transplantation can trigger both immediate ischemia–reperfusion injury and delayed immune responses.
The composition of the perfusate itself has become an important technical variable. Acellular solutions are designed to carry oxygen while limiting cellular immune reactions, whereas cellular perfusates may contain red blood cells or other components that more closely reproduce blood’s oxygen-carrying capacity and rheological behavior. Cellular EVLP can improve the interpretation of oxygen transport and may support tissue repair by providing nutrients and physiological shear stress. It also introduces additional complexity: leukocytes can amplify inflammation, blood components can affect laboratory measurements and viscosity influences the distribution of flow through the microcirculation. Comparing acellular and cellular systems is therefore not simply a matter of choosing one fluid over another; it is a question of which biological signals clinicians need to preserve, suppress or measure.
Imaging technologies are helping researchers see beyond global averages. Bronchoscopy can identify airway secretions, mucosal injury and obstruction, while radiography and computed tomography can reveal consolidation, edema, atelectasis and uneven recruitment. Ultrasound may provide a rapid, radiation-free assessment of pleural and parenchymal changes. Electrical impedance tomography, which maps changes in thoracic electrical conductivity, has attracted interest as a way to follow regional ventilation and perfusion in real time. Such information could show that an apparently satisfactory organ is relying on a small number of overworked regions while other areas remain poorly ventilated. The combination of imaging with localized sampling may eventually allow transplant teams to determine whether a defect is diffuse, reversible or confined to tissue that can be safely excised.
The review also points toward emerging analytical tools, including metabolomics, transcriptomics and machine-learning models. Metabolomic profiling can capture shifts in energy production, oxidative stress and membrane turnover, while gene-expression analysis may reveal pathways activated by hypoxia, inflammation or repair. These datasets are potentially powerful but difficult to standardize. Temperature, perfusion duration, ventilator settings, donor history and the formulation of the perfusate can all alter the molecular readout. Artificial intelligence could help integrate physiological curves, biomarker concentrations and imaging data, but a predictive model is only as reliable as the multicenter data used to train and validate it. Before such systems can guide clinical acceptance decisions, investigators will need common protocols, defined thresholds and outcome studies linking EVLP measurements to post-transplant survival and graft function.
EVLP is also evolving from a diagnostic platform into a therapeutic one. During perfusion, clinicians can adjust ventilation, control fluid balance, administer vasodilators or antibiotics and remove inflammatory mediators through filtration. Experimental strategies have explored targeted delivery of drugs, gene-based interventions, anti-inflammatory agents and cell-derived products directly to the isolated lung. Mesenchymal stromal cells and their secreted factors, for example, are being investigated for their potential to modulate inflammation and restore endothelial or epithelial integrity. These therapies remain under active study, and the ability of an ex vivo circuit to repair a severely injured organ has not eliminated the need for rigorous safety testing. Nevertheless, the platform offers a rare advantage: treatment can be evaluated outside the recipient, while changes in oxygenation, compliance, vascular resistance and molecular injury markers are observed in real time.
The central message of the review is that no single test can capture the biological condition of a donor lung. Effective assessment will likely depend on a composite profile combining gas exchange, mechanics, hemodynamics, metabolism, inflammatory biomarkers and regional imaging. Such a profile could reduce unnecessary organ discard, make the selection of extended-criteria lungs more transparent and create a common language for centers using different perfusion systems. The challenge is to translate a technically rich stream of data into decisions that are reproducible, clinically meaningful and safe. As cellular EVLP matures, its greatest contribution may be more than rescuing individual organs: it may transform lung transplantation from a brief inspection of a damaged graft into a measured period of observation, intervention and biological recovery.
