It is evident that DCD heart transplantation has successfully expanded the donor pool, increasing heart transplantation activity globally [1, 14, 15]. In the United States (US), United Network for Organ Sharing (UNOS) Registry data analysis demonstrates that DCD heart transplantation has significantly contributed to a reduction in wait-list time, particularly in candidates who would typically have longer expected wait-list times; these are candidates who are wait-list status 4 (stable with a VAD) and greater [16, 17]. In our experience, since 2021, the majority of heart transplants performed on patients who would qualify as status 1 (critically ill on mechanical support), were from DCD donors [18].

Described in our previous review [5], two commonly used methods of DCD heart procurement include: (I) DPP-NMP: Direct procurement pathway (DPP) followed by NMP and (II) taNRP-CS: Thoracoabdominal normothermic regional perfusion followed by cold storage (CS), most commonly referring to static cold storage (SCS).

Direct Procurement Protocol Followed by Normothermic Machine Perfusion

Our unit in Australia has the longest experience with DCD heart transplantation, with our program now having over a decade of experience. DCD hearts are procured using a DPP followed by NMP. Our published data reporting on DCD heart transplants from 2014 to 2022 demonstrates that there has been no significant difference in the 8-year survival outcomes of DCD heart transplant recipients compared to BDD receipients [1]. Since 2021, our unit has had access to HMP for BDD with a predicted prolonged donor ischaemic time (DIT). In comparing our heart transplantation outcomes from recipients of DCD hearts procured using NMP, BDD hearts preserved with either HMP or SCS, there appears to be no significant difference in survival during this new modern era of machine perfusion [18].

In the initial Sydney experience reported by Chew et al. [19], there was a high rate (35%) of severe primary graft dysfunction (sPGD) requiring mechanical circulatory support (MCS) - this is often cited in the literature as a concern with DCD heart transplantation. However, as our experience and familiarity with NMP and DCD donation grew, there was a statistically significant improvement in sPGD [1]. When directly compared to the initial cohort, a more contemporary cohort had a significantly reduced rate of sPGD at 8%; this improvement has continued and since 2021, our DCD program now has a 7% rate of sPGD [1, 18].

The ongoing reduction in Australian DCD heart transplant sPGD rates can be attributed to the efforts being made by all parties involved in the organ retrieval process and reflects growing experience. Particular attention has been paid to reducing the warm ischaemic time. An asystolic warm ischaemic time (aWIT) of > 15 min has been consistently identified as a vital, and statistically significant predictor of sPGD [1, 18, 19]. Where possible, and when local policy permits, there is an increased effort being made to advocate for withdrawal of life support (WLS) to occur in the anaesthetic bay and on an operating table as our experience has shown withdrawal location to play a significant role in influencing aWIT [20]. Retrieval techniques have also been refined in order to minimise aWIT. Furthermore, where once it was a novel phenomenon, our exposure to NMP over the last decade has resulted in a greater understanding of the nuances in the ex-situ management of the DCD heart, as experience accumulated with each retrieval and subsequent transplant, including any organ rejections following NMP assessment.

The Papworth group in the United Kingdom (UK) have also reported there to be no significant difference in survival when comparing recipients of BDD to DCD donor hearts [15]. In 2023, results from an 18 month observational national pilot study in the UK were published, this reported on outcomes after the DCD heart retrieval service was nationalised (as opposed to DCD heart transplants occurring at a single centre) [2]. DCD hearts were procured utilising DPP-NMP with results from this multi-centre study once again demonstrating there to be no significant difference in survival between recipients of BDD and DCD donor hearts [2]. Furthermore, there were no survival differences when compared to pre-pilot era outcomes [2].

Though it did not seem to influence survival, there was a significantly increased rate of extracorporeal membrane oxygenation (ECMO) in recipients of DCD hearts during the pilot study (40%) [2]. This was thought to be likely secondary to higher ECMO rates in centres that were not experienced with DCD heart transplantation [2]. The results and the authors highlight the importance of experience, and advocate for centres that are new to DCD heart transplantation to learn from established DCD heart transplant units [2]. Encouragingly, this study demonstrates that with appropriate training and expertise in DCD retrievals, the national expansion of DCD programs is feasible without compromising survival outcomes, resulting in improved access to an expanding donor pool.

Importantly, 2023 also marked the publication of the first randomised control trial (RCT) involving DCD heart transplantation, further validating DCD heart transplantation as a standard of care for heart transplant units [4]. This was a multi-centre trial based in the United States [4]. Patients assigned to the DCD heart transplantation group received a DCD heart procured via DPP followed by NMP using the OCS Heart and compared to recipients of BDD retrieved hearts following SCS; DCD heart transplantation was found to be non-inferior when considering the primary end-point of 6 month survival [4]. Overall, the sPGD rate amongst DCD heart transplant recipients in the trial [4] was 15%, compared to 5% in the BDD group – this is likely a reflection of the early experience with DCD retrievals and NMP (which has been observed in early outcomes internationally [2, 19]), as well as the multi-centre nature of the trial. Whilst it does not seem to have impacted survival, it is likely that the rate of sPGD will continue to improve with broadening use of DCD donors.

Thoracoabdominal Normothermic Regional Perfusion Followed by Cold Storage

Considering taNRP specifically, results from an international multicentre retrospective observational study of 157 DCD taNRP-CS hearts conducted across the United States, Belgium, Spain and the UK have been published [14]. Across the 15 centres enrolled in this study, there was a 23% increase in heart transplantation activity with no differences in survival up to 5yrs between recipients of DCD and BDD heart transplants [14].

The rapid uptake of DCD heart transplantation has resulted in a large amount of registry data from the United States now being available for analysis [21, 22]. A recent analysis of the United States Organ Procurement and Transplantation Network (OPTN) heart transplantation data between December 2019-September 2023 demonstrated there to be no significant difference in 3 year survival or rates of primary graft dysfunction (PGD) between recipients of DCD and BDD hearts [22]. This was regardless of DCD procurement method with taNRP and DPP accounting for 249 and 543 DCD heart transplants respectively (compared to 10,833 transplants from BDD) [22].

A large single centre study from the Vanderbilt University Medical Centre serves as an example of a centre being able to facilitate both taNRP as well as DPP approaches for DCD retrievals [23]. Out of 122 DCD heart transplants performed in the study, taNRP-CS accounted for 101 (83%) with 21 (17%) of DCD transplants occurring via DPP-NMP. This was compared to transplant outcomes from BDD hearts procured using either NMP (10/263, 4%) or SCS (253/263, 96%) [23]. Consistent with recent international literature, there were no significant differences in 1 year survival between the two groups [23].

The international analysis of taNRP outcomes by Louca et al. [14] reported no significant differences in the incidence of post-transplant MCS between the DCD and BDD groups (12.8% and 12.7% respectively). The Vanderbilt group also reported no differences in sPGD rates between the two groups (6% across both) [23]. The Vanderbilt study however was not powered to report differences between DCD procurement techniques [23]. This is where rapidly growing registry data and analysis can help provide further answers with the most recent US OPTN analysis showing no significant difference in survival or PGD rates when comparing DCD DPP-NMP and DCD taNRP-CS outcomes [22].

Cold Ischaemia in Directly Procured DCD Hearts Preserved with NMP

In DCD Hearts procured via DPP, donor hearts are also subject to short cold ischaemic periods (see Fig. 1). The first exposure is following the administration of cardioplegia during the initial organ procurement. Following this the heart is then placed in a cold slurry and instrumented prior to reperfusion on the OCS Heart NMP device. In our experience this can be referred to as the back table cold ischaemic time (CIT). Following this, the heart is then re-perfused and assessment on the NMP device ensues. The donor heart is then subject to a second cold ischaemic period as cardioplegia is once again administered in order to facilitate decannulation from the OCS Heart once the implanting surgeon is ready. In our experience, prior to this second administration of cardioplegia, the OCS Heart is connected to a water-heater cooler and the heart is cooled to 16 degrees Celsius [1]. During the implant, cold blood cardioplegia is then administered at regular intervals and a “hot shot” dose of dilute warm blood cardioplegia is then administered prior to the removal of the cross clamp at the discretion of the surgeon.

Fig. 1

Retrieval timeline for directly procured donation after circulatory death donor hearts

Exposure of the DCD heart to two cold ischaemic times results in two potential subsequent periods of ischaemia reperfusion injury (IRI). Of late, questions have arisen as to whether this IRI exposure, along with the duration of CIT, may potentially contribute to primary graft dysfunction [2, 24]. This issue is discussed in the UK national pilot study where the time to reperfusion on the OCS Heart following procurement was influenced by the method retrieval teams would use to mount the cardiac allograft onto the NMP device [2]. “Method A” (previously described by Messer et al. [15]) involved the inferior and superior vena cavae (IVC and SVC) being left open, along with the pulmonary artery (PA) cannula being disconnected; this allowed the right ventricle (RV) to remain in an unloaded state. In “Method B” (similar to the approach in Sydney [1] and other centres [4, 25]), the cavae are oversewn/ligated and the PA cannula attached to a return connector on the OCS Heart allowing for coronary flow to be measured.

There were no significant differences in survival or sPGD outcomes reported between the two methods in the pilot study [2], however Method B was found to have a significantly longer time from a systolic blood pressure (SBP) < 50mmHg up to reperfusion on the OCS Heart (which was defined in this particular study as the functional warm ischaemic time [fWIT]). This significant increase in time (median of 25 min for Method A vs. 33 min for Method B) is due to the extra steps taken to instrument the allograft for the OCS Heart. Messer et al. [2] suggested that a non-significant trend for lower 30 day survival with Method B may reach significance with longer follow up, however, longer term data from the Sydney experience suggests that this is not the case [1]. In our experience, increasing CIT and increasing CIT + fWIT was not associated with an increased risk of sPGD (fWIT defined in our unit as time from SBP < 90mmHg to administration of cardioplegia) [1, 18]. Furthermore, our more contemporary cohort, which has experienced a significantly lower rate of sPGD, had a significantly higher CIT exposure compared to our initial cohort [1].

The heart transplant team at Stanford University (Stanford, CA, USA) described an innovative operative approach to avoid a second period of cold ischaemia [24, 26]. Krishnan et al. [26] describe an original set up that allows for concurrent cardiopulmonary bypass (CPB) to be initiated in the recipient, alongside ongoing normothermic perfusion on the OCS Heart via a shared CPB circuit. Briefly, upon return to the recipient operating theatre, an aortic root needle is introduced to the donor heart and a cross clamp applied - this allows for antegrade perfusion of the donor heart from the CPB circuit [26]. The donor heart is then removed from the OCS device and implanted with ongoing antegrade perfusion [26]. Once the left atrial and aortic anastomoses are complete, donor heart and recipient cross clamps are removed and the heart is re-perfused via the aortic cannula while the remainder of the anastomoses are completed [26]. This novel approach describes the first instance of a “beating-heart” transplant and results in the DCD donor heart being exposed to just one period of CIT (at the donor hospital site). Outcomes appear to be promising with a case series of 10 DCD heart transplants utilising this method being described with no instance of primary graft dysfunction or early mortality [24].

The impact of CIT in DCD hearts preserved with NMP is an area for future potential research. Whilst Krishnan et al. [24] suggest that the second period of CIT may potentially contribute to graft dysfunction, data from our contemporary experiences with DCD heart transplantation demonstrate that exposure to CIT does not adversely impact outcomes, with excellent survival and sPGD rates despite two periods of CIT [1, 18]. We believe this suggests: the DCD heart is likely able to tolerate periods of controlled cold ischaemia; that ischaemic preconditioning may confer some cardioprotection following a second CIT [27]; and, that the asystolic warm ischaemic time likely plays a more important role in predicting sPGD [1, 18]. Novel techniques in “beating-heart” transplantation that limit a second CIT however, do appear to be feasible and safe [24, 26].

Ethics of taNRP and Impact on Concurrent Retrieval of Other Organs

The ethics surrounding taNRP remains a contentious topic, with concerns and discussion focussing on whether the in-situ re-animation of donor hearts – and the potential for any blood flow (including through collateral vessels) to the cerebral circulation – violates the “dead donor rule.” [28,29,30,31].

One potential argument supporting the use of taNRP is the increased utilisation of donor hearts compared to a DPP approach. Bakhtiyar et al. [22] reported a 1.64% rate of non-use in the DCD taNRP group compared to 10.83% in the DCD-DPP group. Arguments favouring normothermic regional perfusion (NRP) suggest that it is a technique that maximises organ utilisation of the heart as well as abdominal organs [32]. DCD livers procured through NRP are known to have significantly reduced rates of 30-day graft loss and significantly lower rates of ischaemic cholangiopathy [33]. Furthermore, UNOS registry data shows, compared to DPP-NMP procurement, when taNRP is used for the retrieval of DCD hearts, there is a significantly increased utilisation rate of concurrently procured DCD livers, as well as a reduced rate of delayed graft function in DCD kidneys [34].

Whilst there appears to be a benefit to abdominal organs via, the impact of taNRP on the retrieval of DCD lungs is less well known. This is due to concerns surrounding the potential risk of pulmonary oedema from fluid used to prime CPB or ECMO circuits. Cain et al. [35] describes a small case series from the University of Colorado Denver where lung transplants were performed successfully following a taNRP procurement technique with a specific focus on reducing pulmonary oedema. Strategies utilised in the Colorado case series included: early and aggressive donor diuresis pre-procurement; early venous cannulation and donor decongestion; early venting of the donor PA, and, early diuresis post-transplant [35]. Overall however, UNOS data suggests concurrent DCD lung utilisation rates are low regardless of procurement technique (14.9% and 13.8% for taNRP and DPP approaches respectively) [36]. With no significant differences in 6-month survival in taNRP retrieved lung transplants compared to DPP [36], procurement rates from both techniques have the potential to improve. Emerging interest in strategies to minimise pulmonary oedema during taNRP is promising and highlights the need to consider the lungs (and not just the heart and abdominal organs) in future decisions and ethical discussions surrounding which retrieval technique maximises organ procurement.

From an alternative viewpoint, arguments can be made that taNRP and the subsequent implications of in-situ re-perfusion of organs, challenge the definition, as well as our understanding of death [