Donor Lung Oxygen Numbers Predict Early Complication but Not Survival After Transplant

Every year, thousands of donor lungs are turned away…

The distinction between early and late primary graft dysfunction deserves closer attention, because the two phenomena may not share the same underlying biology. In the Leuven–Prague cohort, the association between lower donor PaO2/FiO2 ratio and grade 3 PGD was confined to the 24-hour assessment window and disappeared by 72 hours. This temporal pattern suggests that a low donor PFR may act as a marker of transient perioperative vulnerability rather than of durable structural lung injury. Immediately after implantation, the transplanted lung must cope with ischemia–reperfusion injury, surgical trauma, and the abrupt resumption of blood flow, and a graft that already entered the operating room with suboptimal oxygenation may have less physiological reserve to absorb these insults. By 72 hours, however, much of the early capillary leak and atelectatic burden has typically begun to resolve in survivors, and the influence of the donor’s pre-procurement oxygenation appears to fade, giving way to other determinants of outcome such as recipient comorbidity, immunological events, and perioperative management.

The physiological ambiguity of the PFR measurement itself is central to interpreting these findings. The ratio of arterial oxygen tension to inspired oxygen fraction is routinely treated as a straightforward index of lung quality, yet it is sensitive to conditions far removed from the alveolus. Mixed venous oxygen saturation, cardiac output, hemoglobin concentration, and the overall balance between oxygen delivery and consumption all modulate the measured value. A hemodynamically unstable donor with low cardiac output, for example, may display a depressed PFR even when the lung parenchyma is essentially healthy, because venous blood arriving at the pulmonary circulation is already profoundly desaturated. Conversely, aggressive donor resuscitation with high FiO2 can transiently inflate the ratio. This means that a single arterial blood gas drawn at an arbitrary moment before procurement captures not only the state of the lung but also the hemodynamic circumstances of the donor at that instant. The authors’ observation that low PFR predicts only the earliest form of PGD is consistent with this interpretation: what is being measured may partly reflect a reversible physiological state rather than fixed parenchymal damage.

The composition of the donor pool in this study also merits consideration when generalizing the results. With a median donor age of 51 years and 18% of grafts coming from donation after circulatory death, the cohort reflects the contemporary reality of lung transplantation, in which programs increasingly rely on older donors and on DCD pathways that were once considered extended criteria. The authors note that DCD, although still formally classified as an extended criterion in many frameworks, has become so widely adopted that it now functions as standard practice in numerous centers. This gradual redefinition of what counts as acceptable illustrates how thresholds in transplantation medicine are often historical artifacts rather than evidence-derived boundaries. The PFR cutoff of 300, which leads to frequent declination of otherwise acceptable lungs, is explicitly described in the source as arbitrarily set, and the present findings add weight to the argument that such fixed numeric gates deserve re-examination.

The scale of the unused donor pool gives this question practical urgency. A significant proportion of offered lungs are declined, and this contributes directly to waiting list mortality, since candidates who deteriorate while waiting have no alternative therapy with comparable efficacy. Any donor parameter that could be shown to be over-restrictive represents an opportunity to expand utilization without compromising outcomes. If a lung with a PFR below 300 carries essentially the same two-year survival prospects as one above the threshold, as the regression and Kaplan–Meier analyses in this cohort suggest, then declining such organs on the basis of the number alone may be discarding transplantable grafts. The caveat, of course, is that the increased risk of early grade 3 PGD at 24 hours is not clinically trivial. PGD severe enough to reach grade 3 in the first day after surgery is associated with prolonged mechanical ventilation, longer intensive care stays, and greater early morbidity, so transplant teams using low-PFR lungs must be prepared for a higher likelihood of demanding early postoperative courses even if the ultimate survival trajectory is unaffected.

The two-center design strengthens the analysis in ways that single-institution studies cannot. Leuven and Prague operate within different healthcare systems, with different donor coordination structures, preservation logistics, and recipient populations, yet their combined dataset of 984 recipients spans more than a decade at the Belgian center and several years at the Czech center. The inclusion windows differed between the sites, reflecting the availability of consistently recorded donor PFR, PGD grading, and outcome data rather than any deliberate clinical divergence, and the median follow-up of four years provides a reasonably mature view of two-year survival. The choice of a two-year survival endpoint was deliberate: it captures early and intermediate-term outcomes that are plausibly influenced by donor and perioperative factors while limiting confounding from the long-term comorbidities that accumulate in transplant recipients regardless of graft quality. Chronic lung allograft dysfunction, by contrast, typically emerges over longer horizons, and the CLAD-free survival analysis in this study should be read with that timeframe in mind.

The secondary endpoints add a further layer of reassurance regarding the safety of lower-PFR grafts. Acute cellular rejection within the first year, diagnosed histologically through transbronchial forceps or cryobiopsy samples graded according to ISHLT guidelines, showed no association with donor PFR. This is noteworthy because one might plausibly hypothesize that a graft injured at the time of procurement, or subjected to more aggressive resuscitation, could be more immunogenic or more susceptible to alloimmune attack. The absence of such a signal in this large cohort argues against a meaningful link between pre-procurement oxygenation and early cellular immune events. Similarly, CLAD-free survival was not adversely related to donor PFR, suggesting that the early functional perturbation signaled by a low ratio does not translate into accelerated chronic rejection pathways within the observed follow-up.

Several variables that are known to influence PGD could not be incorporated into the models. Donor smoking history and heavy alcohol use have both been associated with increased PGD risk in prior work, but the volume of missing data for these characteristics in the two registries precluded their inclusion. This is a reminder that donor PFR does not exist in isolation: the same donor whose oxygenation is marginal may also carry a smoking history, and disentangling the independent contribution of the blood gas from these co-occurring features requires datasets with more complete documentation. Ex vivo lung perfusion, used at both centers for marginal or uncertain lungs, including those with impaired oxygenation, DCD or extended criteria features, and concerning radiographic findings, represents one practical avenue through which programs can gain additional information about a questionable graft before committing to implantation. EVLP assessment may in effect substitute a dynamic, controlled evaluation for the static snapshot provided by a single arterial blood gas, and its growing role in donor evaluation interacts directly with the question of how much weight the PFR number should carry.

The methodological choice to define early PGD as grade 3 at 24 hours and late PGD as grade 3 at 72 hours follows the ISHLT consensus framework, in which the diagnosis rests on pulmonary edema visible on chest imaging combined with an impaired PaO2/FiO2 ratio. It is worth noting the inherent irony that PGD grading itself relies on a PFR-type measurement in the recipient, so the donor and recipient oxygenation metrics are conceptually linked even though they reflect different physiological states at different times. Recipient-side factors, including the need for postoperative extracorporeal membrane oxygenation, intensive care and hospital length of stay, and the surgical approach, were captured in the analysis, allowing the models to adjust for important perioperative determinants of early graft function.

Taken together, the findings support a nuanced approach to donor lung acceptance rather than a wholesale abandonment of oxygenation assessment. A low donor PFR should prompt scrutiny and, where available, additional evaluation such as EVLP or careful review of donor imaging and bronchoscopy, but it should not function as an automatic exclusion criterion. The observation that the prognostic signal is confined to the first 24 hours after transplantation, with no measurable penalty at 72 hours, in rejection incidence, in CLAD-free survival, or in two-year overall survival, aligns with the position the Leuven group articulated as early as 2008, when they argued that a low PFR alone should not constitute an absolute contraindication to utilization. In an era of persistent donor shortage and growing waiting list mortality, the practical implication is that transplant programs may be able to safely widen their acceptance criteria for lungs with marginal oxygenation, provided they are equipped to manage an elevated incidence of early postoperative graft dysfunction. Confirmatory work in additional populations, ideally with prospective collection of donor hemodynamic variables alongside the blood gas, would help clarify how much of the early PGD signal reflects true graft injury and how much reflects the physiological context in which the donor number was generated.

Subject of Research: Effect of donor PaO2/FiO2 ratio on primary graft dysfunction and survival after lung transplantation

Article Title: Lower lung donor PaO2/FiO2 ratio correlates with early primary graft dysfunction, but does not impact overall survival after lung transplantation: a dual center study

Article References: Novysedlak, R., Vanluyten, C., Van Slambrouck, J., Barbarossa, A., Beeckmans, H., Van Raemdonck, D., De Leyn, P., Van Veer, H., Depypere, L., Jansen, Y., Vanaudenaerde, B. M., Vos, R., Belmans, A., Zajacova, A., Svorcova, M., Vajter, J., Tavandzis, J., Pozniak, J., Simonek, J., … Ceulemans, L. J. (2026). Lower lung donor PaO2/FiO2 ratio correlates with early primary graft dysfunction, but does not impact overall survival after lung transplantation: a dual center study. Transplant International, 39, Article 16414. https://doi.org/10.3389/ti.2026.16414

Image Credits: AI Generated

DOI: 10.3389/ti.2026.16414

Keywords: lung transplantation, donor lungs, PaO2/FiO2 ratio, primary graft dysfunction, extended criteria donors, ex vivo lung perfusion, organ shortage, transplant outcomes, DCD donors, acute cellular rejection, CLAD, survival