All consecutive high-risk patients who underwent HTx with grafts from marginal donors and preserved with EVP at the University Hospital of Udine and Hannover Medical School, from 2016 to 2021, were retrospectively analyzed. Indication for EVP was the same in the two centers and involved in all cases employment of a marginal graft for a high-risk recipient, as subsequently defined. The study was approved by the Institutional Review Board (Clinical Registration Number: 18386, 1 August 2016) and informed consent was waived due to its retrospective nature. The major end-point of the study was the assessment of early and mid-term clinical outcome in terms of survival and major complications after HTx.

In the present series EVP was achieved in both participating centers employing the Organ Care System (OCS) (Transmedics Inc., Boston, MA, United States). The OCS perfusion technique has been described in detail previously [8]; briefly, it is instituted by cannulating the aorta and pulmonary artery of the graft and connecting them to a perfusion circuit with an oxygenator and a pulsatile pump. The beating heart is perfused with warm, oxygenated, nutrient-enriched donor blood mixed with priming solution. Graft function is assessed by continuous monitoring of aortic pressure, coronary flow and lactate trend. Final arterial lactate value (<5 mmol/L), lactate trend, difference between arterial and venous lactate concentration, visual contractility and stability of the perfusion data on OCS are all considered for HTx acceptance of donor grafts. The OCS device was transported either by car, plane or helicopter based on the distance of the donor hospital.

Criteria for recipient selection and donor-recipient matching were based on standard guidelines [9–12]. HTx was performed using the bicaval anastomosis technique and immunosuppression therapy was standardized with steroids, calcineurin inhibitors and mycophenolate mofetil. Postoperative and long-term follow-up protocols at the University Hospital of Udine have been previously published and have remained unchanged during the study period [13–17]. In Hannover, all HTx recipients underwent triple immunosuppressive therapy with tacrolimus, prednisolone and mycophenolate mofetil. All patients underwent induction with anti-thymocite globulin on day 3, 4 and 5 after HTx. Endomyocardial biopsies were performed every 2 weeks during the first 3 months of follow-up. Seven patients who showed pre-transplant allosensitization and a positive virtual crossmatch were treated with a perioperative combination of therapeutic plasmapheresis, followed by infusions of Tocilizumab before allograft reperfusion, a single infusion of human immunoglobulins with or without a single infusion of Rituximab.

High-risk recipients were defined as those on pre-HTx dependence from inotropic support, pre-HTx implantation of an intra-aortic balloon pump, those bridged to HTx with extracorporeal life support systems (ECLS) or a ventricular assist device (VAD). For ECLS support a Quadrox-i oxygenator and a Cardiohelp centrifugal pump (Getinge, Göteborg, Sweden), were employed. ECLS was used in 26% of patients in one center and in 5% in the other depending on different treatment policies of moderate primary graft dysfunction (PGD). Donors selected according to extended criteria, defined as “marginal,” were considered those ≥55 years of age, with a history of drug abuse, cardiac resuscitation or severe prolonged hypotension (>20 min), coronary artery disease, with at least a >30% stenosis of a major coronary branch, expected graft ischemia time ≥4 h, left ventricular ejection fraction <50%, interventricular septum thickness >14 mm, or body surface area difference between donor and recipient >20%. Hemodynamic support of the donor grafts was generally obtained with infusion of ≤0.1 y/kg/min of norepinephrine.

Chronic renal failure was defined by a glomerular filtration rate <50 mL/min/m² or a persistent 50% increase of serum creatinine level. Impaired liver function was considered as at least a twofold increase of bilirubin and/or liver enzymes.

PGD was defined according to the International Society for Heart and Lung Transplantation (ISHLT) consensus statement, considering it relevant when grade ≥ moderate [18, 19].

Acute renal failure was defined following the Risk of renal dysfunction, Injury to the kidney, Failure of kidney function, Loss of kidney function, End-stage kidney disease (RIFLE) criteria [20]. The indication for continuous renal replacement therapy was based on persistent oliguria (<0.5 mL/Kg/h) for at least 6 h, a double increase of serum creatinine levels or at least 50% reduction of the glomerular filtration rate within 24 h.

Infections were defined as any episode requiring antibiotic, antiviral or antifungal treatment. Acute rejection was diagnosed, scored and treated following the ISHLT guidelines [21]. Grade ≥ 2 acute rejection or any type of rejection, cellular and/or antibody mediated with hemodynamic impairment was considered as a post-HTx complication and treated accordingly. Coronary allograft vasculopathy was diagnosed by yearly angiographies and defined according to ISHLT classification [22, 23].

For prediction of PGD, the RADIAL score has been used which is calculated on six factors with similar influence, four of which are related to the recipient: Right atrial pressure ≥10 mmHg, Age ≥60 years, Diabetes and Inotropic support dependence while two are related to the donor: Age ≥30 years and Length of total graft ischemic time ≥240 min. The presence of each of these factors in an individual patient adds 1 point to the final PGD predictive score [6].

Continuous variables were expressed as mean ± standard deviation or median and range (min-max) according to the data distribution, after performing the Kolmogorov-Smirnov test for normality. Categorical variables were presented as absolute numbers and percentages. Cumulative overall survival was defined as freedom form all-cause mortality and was determined by the Kaplan-Meier method. Binary logistic regression was used to assess factors for PGD ≥ moderate and in-hospital mortality, while Cox-regression model was used for long-term mortality after HTx. De Long’s nonparametric receiver operating characteristic analysis of the area under the curve (AUC) was performed to estimate the accuracy of risk factors that were identified at the univariate analysis and to determine a cut-off value.

Analyses were performed with IBM SPSS Statistics 22 for Microsoft Windows (IBM Corp., Armonk, NY).

During the study period, out of 88 marginal grafts 80 were employed for HTx in high-risk recipients; 37% were transplanted at the University Hospital of Udine and 71% at the Hannover medical School; 8 (10%) were discarded after being considered unsuitable for HTx through OCS graft assessment because of pathological increase of lactates despite adequate coronary perfusion in most cases; in particular, right ventricle dysfunction, severe aortic valve regurgitation, coronary anomaly and coronary dissection were detected in four grafts.

Recipient data are shown in Table 1. Median age was 57 years (range, 13–75); 80% were males, with a mean body mass index of 26 ± 4 and 12 patients had a weight ≥100 kg and a body mass index ≥30 with a significant size mismatch occurring in 30% of them. Chronic renal failure was present in 61% patients, three of whom where in pre-HTx dialysis and underwent combined heart-kidney transplantation; 11% had impaired liver function and 90% had a previous cardiac operation; 66% were bridged to HTx on long-term VAD with a median support time of 22 months (range, 1–133). In 81% urgent HTx was performed: of these 14% were on short-term ECLS (mean support time of 12 ± 12 days), 12% were dependent from inotropic support and 55% were patients with VAD-related complications. Median RADIAL score was 3 (range, 0–6); a score ≥4 was present in 35% of recipients.

Donor data are summarized in Table 2. Median donor age was 48 years (range, 17–69); 28% were ≥55 years of age, 65% were males, 46% smokers, 29% had a history of alcohol abuse and 39% suffered a previous episode of cardiac arrest. On transthoracic 2D- echo, 6% of grafts had a left ventricular ejection fraction ≤50%, while 21% showed left ventricular hypertrophy; in 4% coronary artery disease was disclosed at angiography revealing 40% stenosis of the left anterior descending in two cases and stenoses of 45% and 50% of left anterior descending and right coronary artery in another, respectively.

Mean graft ischemic time was 118 ± 25 min, mean EVP time 289 ± 62 min, mean total “out-of-body” time 420 ± 66 min and median cardiopulmonary bypass (CPB) time 228 min (range 126–416 min). Three patients underwent combined heart- kidney transplantation (Table 3).

Post-operative data are shown in Table 4. In-hospital mortality was 11%. Causes of early deaths were septic (n = 2) or haemorrhagic shock (n = 2), pancreatitis (n = 2), multi-organ failure (n = 2), and ECLS-related complications (n = 1). Postoperatively, 15 patients (19%) needed de novo ECLS support, because of ≥ moderate PGD in 13 patients (16%), vasoplegia syndrome in 1 and respiratory insufficiency in 1. Moreover, two patients needed veno-venous ECLS for pulmonary complications while an intra-aortic balloon was implanted in three patients. Two patients were assisted with ECLS for <24 h and 8 for <72 h.

Median intensive care unit stay was 6 days (range, 1–123); median mechanical assisted ventilation time was 29 h (range, 3 h to 110 days) with 26% of patients requiring >72 h of ventilation and 18% a tracheostomy. Median hospital stay was 36 days (range, 3–236), during which 64% of patients needed dialysis for acute renal failure, 24% had new onset of atrial fibrillation and 11% acute rejection grade ≥2 (combined with antibody-mediated rejection 2 in two cases); 19% needed sternal re-entry for bleeding and 1% had a stroke.

At univariable analysis, CPB time resulted a risk factor for both ≥moderate PGD (p = 0.001) and in-hospital mortality (p = 0.031).

In addition, high pre-HTx central venous pressure (CVP) was also a risk factor for hospital mortality (p = 0.050); PGD ≥ moderate and in-hospital mortality predictions of the CPB time showed AUC of 0.82 and 0.73 with cutoff values of 246 and 272 min, respectively. The predictive role of CVP regarding in-hospital mortality showed an AUC of 0.69, with a cutoff value of 15 mmHg (Table 5).

During a median follow-up of 16 months (range, 2–43 months), eight patients died. The causes of late mortality are shown in Table 6. Actuarial survival at 1 and 3 years HTx was 83% ± 4% and 72% ± 7% (Figure 1). The rate of grade ≥2R acute rejection episodes and coronary allograft vasculopathy during follow-up was 6% and 10%, respectively.

At univariable analysis, recipient age (p = 0.021), pre-HTx ECLS (p = 0.010), donor age (p = 0.031) were reported to be risk-factors for late mortality (Table 5).

This study has certainly some limitations mostly represented by its retrospective nature and those pertaining to any multicenter collaboration. In this specific study only two centers were involved minimizing potential biases in patient selection and treatment; however, some differences in each center policy, especially in postoperative immunosuppressive treatment and follow-up protocols, have emerged, which might have had an impact on patient outcomes but with negligible influence on PGD and perioperative complications. Moreover, the number of patients enrolled in this study was limited due to the specific patient characteristics, both of donor and recipients, thus precluding to select a control group for comparison.

Our study, dealing with a very complex setting represented by marginal donors, very high-risk recipients and expected long ischemic time, indicates that EVP appears to be an effective tool in reducing overall donor graft ischemic time and allowing continuous evaluation of graft function and viability during transportation. This should provide adequate grafts for high-risk recipients who would be otherwise excluded from the possibility of a HTx. Nevertheless, based on the gratifying results observed in the present study, we advocate the employment of EVP using OCS technology as a promising and valid tool to further extend the donor pool, to successfully perform HTx even in high-risk settings.