Heart transplantation does not end when the graft starts beating. In many respects, that moment marks the beginning of one of the most vulnerable phases of the entire process, a period that carries a substantial share of early prognosis: the first 72 hours.
For years, the 2014 ISHLT consensus served as the main framework for interpreting events during this interval. However, the present-day setting differs markedly from that earlier era. We now transplant more complex recipients, many of them bridged with durable mechanical support, and we use preservation strategies that have meaningfully altered early graft physiology. The 2026 update reflects that new reality.
Rather than offering a purely terminological revision, this document proposes a broader conceptual shift: a more accurate definition of what truly constitutes graft dysfunction and, above all, a clearer separation of conditions that are physiopathologically distinct and should not be grouped together.
- Farewell to PGD, welcome to EGD
Abandoning the term primary graft dysfunction (PGD) is probably the most visible change. It is replaced by early graft dysfunction (EGD), a concept that explicitly incorporates the temporal course of the process.
The central idea is to stop viewing dysfunction as a static event confined to the first postoperative hours. The new framework recognizes that graft deterioration may evolve over time and therefore distinguishes between dysfunction arising within the first 24 hours and dysfunction developing between 24 and 72 hours, termed subsequent early graft dysfunction (sEGD).
From a clinical standpoint, this makes sense. A heart that cannot be weaned from cardiopulmonary bypass is not the same as a graft that initially performs adequately and then deteriorates later in the ICU. The timeline, mechanism, and prognosis are different, and the consensus now attempts to reflect that distinction.
- Severity: fewer nuances, greater reproducibility
Another major change is the simplification of severity grading. Scales based on inotrope dose have been removed because, in everyday practice, they were difficult to interpret and lacked specificity.
The classic example is vasoplegia: patients receiving high-dose vasopressors despite preserved ventricular function who nonetheless ended up being labeled as having graft dysfunction. The new consensus seeks to avoid that kind of misclassification.
The current criterion is intentionally straightforward: dysfunction is considered severe when it requires implantation of mechanical circulatory support after transplantation. The need for VA-ECMO, some form of ventricular assist support, or even a newly placed intra-aortic balloon pump defines that threshold.
This is a less granular approach, but probably a more practical one in day-to-day care and, above all, one that is easier to compare across centers.
- The “footprint of the journey”: from concept to quantification
One of the most compelling ideas in the document is the recognition that the graft arrives shaped by everything that happened before implantation. Donor age, cause of death, ischemic times, preservation strategy, recipient inflammatory status—each leaves a trace.
What is new is that this concept is beginning to translate into concrete tools. Models such as PGD-AI integrate multiple variables, approximately 18, and have shown greater predictive performance than traditional scores such as RADIAL. In some studies, the area under the curve exceeds 0.80, clearly outperforming earlier models.
Beyond the number itself, what matters is the shift in mindset: we are moving from a relatively intuitive assessment to a more structured risk estimate. This has direct implications for decision-making, particularly when selecting preservation strategies, such as cold ischemia versus normothermic perfusion, or when anticipating the need for early support. In fact, this is the first document to give a leading role to transport systems such as SherpaPak or the OCS, which can reduce the incidence of severe dysfunction by as much as 50%.
We are still far from universal implementation, but the message is clear: prediction of graft dysfunction risk will become increasingly quantitative.
- What has really changed since 2014
Compared with the previous document, the update is not merely incremental; it is structural.
The inotrope score, long used as a reference despite its obvious limitations, is no longer central. Severity is now defined by the need for intervention rather than by the intensity of pharmacologic support.
The timing of presentation is also given much greater weight. Patients who cannot be weaned from bypass remain at the most severe end of the spectrum and continue to face very high mortality. By contrast, delayed deterioration in the ICU likely reflects more heterogeneous mechanisms and, in some cases, potentially reversible ones.
- Right ventricular dysfunction: less subjectivity
Historically, the right ventricle has been one of the weakest points in this field. Its assessment has depended heavily on clinical judgment and indirect parameters.
The 2026 consensus attempts to make the diagnosis more objective by combining echocardiographic and hemodynamic criteria. In practical terms, clinically relevant dysfunction is considered present when right ventricular ejection fraction is ≤35% or TAPSE is below 1.6 cm in an appropriate hemodynamic setting.
That hemodynamic context matters: elevated right atrial pressure, usually >15 mm Hg, with normal or low pulmonary capillary wedge pressure (<15 mm Hg) and reduced cardiac index (<2.0 L/min/m²). In other words, a relatively “pure” right-sided failure pattern without significant left-sided congestion.
As in the rest of the document, severity is again defined by the need for support. The requirement for RV support or VA-ECMO marks the threshold for severe disease. If treatment remains pharmacologic only, with inotropes or pulmonary vasodilators, it is not classified as severe.
Although these criteria add clarity, real-world diagnosis will still require integration of multiple variables, especially in patients with preexisting pulmonary hypertension or complex ventricular interdependence.
- Refining what truly constitutes graft dysfunction
Another key aspect is the emphasis on distinguishing primary dysfunction from other causes of clinical deterioration. For EGD to be diagnosed, graft failure must be essentially idiopathic. If an identifiable cause is present, such as acute rejection, significant pulmonary hypertension, or a technical complication, the syndrome is classified as secondary graft dysfunction.
This may appear to be a theoretical distinction, but it has important implications. Grouping different entities under the same label makes both clinical interpretation and research much harder.
- The uncomfortable finding: the “disappearance” of >90% of PGD cases
When the new criteria are applied, most patients who would previously have been diagnosed with PGD no longer meet the definition. Some series report this in more than 90% of cases.
Rather than undermining the concept, this likely suggests that we were previously including situations that did not truly reflect intrinsic graft failure. Vasoplegia, transient hemodynamic instability, and mixed syndromes were often subsumed under the same label.
The new approach is more restrictive and leaves a smaller but more homogeneous group, the one that truly concentrates risk and that is most likely to benefit from aggressive strategies such as early mechanical support.
Even so, the consensus itself acknowledges an important paradox: some patients who no longer fit the strict definition still carry substantial mortality. In other words, they may change category, but they do not cease to be clinically high-risk patients.
COMMENTARY:
The first 72 hours after heart transplantation remain an uncertain territory, but we probably now have a more useful framework for interpreting what happens during that period. The 2026 ISHLT consensus does not simplify reality, but it does force us to be more precise. It reduces background noise, defines the problem more accurately, and shifts the focus toward the patients who truly have graft dysfunction in the strict sense. For us, the take-home message is highly practical: make better distinctions, anticipate earlier, and escalate support sooner when needed. Above all, we must accept that not everything that destabilizes a transplant recipient is graft failure, even if that is what we called it for years.
REFERENCE:
Kobashigawa J, Zuckermann A, Potena L, Ardehali A, Berman M, Chang PP, et al; Consensus Conference participants. Summary of the International Society for Heart and Lung Transplantation (ISHLT) Consensus Conference on Graft Dysfunction within the First 72 hours after Heart Transplantation: A 10-year Update. J Heart Lung Transplant. 2026 Feb 18:S1053-2498(25)02486-6. doi: 10.1016/j.healun.2025.12.029. Epub ahead of print. PMID: 41823895.
