Patients with advanced heart failure who are listed for heart transplantation may face significant waiting times before a graft becomes available.1 During this period, a non-negligible risk of hemodynamic decompensation exists, ranging from indolent progression based on the Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS) classification to overt cardiogenic shock. Beyond standard inotropic and vasopressor drugs, short- or medium-term mechanical circulatory support (MCS) devices may be needed to provide hemodynamic stability and halt progression to irreversible end-organ dysfunction, bridging the patient to the desired outcome. Although the mechanistic rationale for these devices is sound, solid evidence to guide the exact timing of insertion or preference for one device over another is scarce at best. Accordingly, European Guidelines assign a class 2b recommendation (level of evidence C) to their use.2

In the context of device use and recommendations, it is important to be aware of the variety of MCS options currently available for the failing heart.3 These devices can be broadly divided by phenotype (left, right, or biventricular), mode of insertion (percutaneous vs. surgical), and level of support (liters per minute), which may include gas exchange. In practice, this range of options includes the intra-aortic balloon pump (IABP), microaxial transaortic pumps (the Impella® family), surgically placed ventricular assist devices (VAD; e.g., Centrimag®), surgical or percutaneously placed right VAD (RVAD; e.g., Centrimag® with Protek Duo® cannula), and veno-arterial extracorporeal membrane oxygenation (VA-ECMO). As these systems are all associated with variable degrees of invasiveness and require systemic anticoagulation, they are inevitably linked to complications such as bleeding, hemolysis, infection, stroke, limb ischemia, and death.3

To complicate matters further, the initial bridge-to-transplant (BTT) strategy may need to be revised to a bridge-to-bridge (BTB) approach. This pivot is often necessary because short-term devices are limited in their duration of support due to inherent shortcomings. A typical scenario, for example, is the “crash and burn” patient, involving ongoing or recovered cardiac arrest. In such urgent situations, due to perceived time and logistical constraints, ECMO is the only device that can be rapidly deployed outside the operating room or catheterization laboratory, promptly providing circulatory support and shock reversal.4 Once ECMO is initiated, the timeframe for definitive intervention begins - unless the patient develops progressive multiorgan dysfunction or receives a graft within a few days, in which case escalation to a more durable device becomes inevitable.

In this issue of the Portuguese Journal of Cardiology, Domínguez-Massa et al. provide an analysis of a cohort of patients listed for heart transplantation who received MCS. They sought to compare the outcomes of two approaches: ECMO as BTB (ECMO-to-VAD) versus direct VAD placement as BTT. VADs included Centrimag and axillary Impella (CP, 5.0, and 5.5); patients bridged with a durable LVAD were excluded. In a single-center eight-year retrospective analysis, data from 73 patients were collected. Of these, the majority (n=51) were placed on VAD support as BTT. The remaining 22 patients initially received ECMO and later progressed to VAD support. Isolated left-sided VAD support was the most common phenotype (n=58), followed by biventricular (n=14) and isolated RVAD (n=1).

As expected, patients placed on ECMO were in a more serious clinical condition, as reflected by the INTERMACS classification, peri-implant lactate levels, and markers of end-organ dysfunction. When clinical outcomes were compared, patients in the ECMO-to-VAD group had a more than three-fold higher in-hospital mortality (86.4% vs. 27.5%). In a multivariate analysis model adjusted for clinical severity, ECMO-to-VAD was independently associated with increased in-hospital mortality. The model demonstrated optimal discriminative ability for predicting in-hospital mortality. Among VAD types, Centrimag (but not Impella) was also independently associated with worse survival.

Several factors can be expected to mediate this impact on mortality. After accounting for patient-related variables, which can be roughly grouped into higher clinical severity (e.g., shock stage, acute kidney injury, invasive mechanical ventilation), the addition of device-related factors (e.g., invasiveness, bleeding, blood transfusions, limb ischemia) will further increase both morbidity and mortality. A third factor, intrinsically related to the former, concerns transplant eligibility. Of the 22 patients on ECMO, only six were ultimately transplanted, suggesting that multiorgan dysfunction in the remaining 16 patients rendered them ineligible for transplantation. In the six patients who proceeded to heart transplantation, mortality was still 50%. Conversely, in the VAD group, four out of five patients were transplanted, and in-hospital post-transplant mortality was 9.8%.

Does this mean that a BTB strategy or specific support devices, such as ECMO or Centrimag, are necessarily worse options for these patients? Multicenter observational studies have yielded conflicting results. In an Organ Procurement and Transplantation Network (OPTN) database study analyzing 1820 patients from 2005 to 2020 (IABP=1385, VA-ECMO=201, VAD=234), Zhou et al. found no difference in post-transplant mortality up to 10 years between BTB and BTT strategies, when considering each device individually.5 Post-transplant survival rates for patients undergoing ECMO in a BTB strategy were 91.8%, 85.5%, and 75.8% at one, five, and 10 years, respectively. Although not formally compared, post-transplant survival rates in the VAD BTT strategy were 89.0%, 73.9%, and 68.0% at one, five, and 10 years, respectively. In another, more recent study that also examined the OPTN database, bridging from ECMO to different MCS devices (including Centrimag and Impella) before heart transplantation was found not to affect outcomes compared with direct transplantation from ECMO support.6

The findings from Domínguez-Massa et al. contrast with these large registry data but are nonetheless valuable for reporting novel data from an Iberian center that may be more closely aligned with Portuguese reality. Interestingly, in a previous work by the same group, the use of ECMO as BTT was associated with a 90.9% rate of transplanted patients, of whom 75% survived the index admission.7 To explain the differences in outcomes between studies, some key considerations should be discussed. The main explanation may lie in the interaction between patient severity and the timing of support implantation, which will ultimately determine shock reversibility.8 In cardiogenic shock, Ton et al. demonstrated that the Society of Cardiovascular Angiography and Interventions classification swifts rapidly within the first hours after diagnosis, altering prognosis in a significant proportion of patients.9 Deciding on the optimal timing is critical and among the most difficult tasks when implanting a temporary MCS device. While “too late” may mean irreversible organ dysfunction, “too early” may bring unnecessary device-related complications.8 This is further compounded by unmeasured factors not captured by multivariate analysis, which can tilt the balance toward worse outcomes in the BTB group. These are inexorably tied to patient selection and the legitimate desire to achieve the best outcome for the “fitter” patient, which adds complexity and variation to clinical decisions across studies.

What are the practical implications of the present analysis? Ideally, inpatients awaiting heart transplantation who are in INTERMACS 2 would be bridged with minimally invasive short-to-medium-term devices, allowing ambulation. For INTERMACS 3, the decision may not be straightforward, as prolonged inotropic support may suffice for some time. For patients in INTERMACS 1 in whom clinical deterioration was not foreseeable, ECMO will inevitably remain the only option for many, with responsibility for outcomes split between the patient's clinical status and the device's hazards. As the authors acknowledge, a preventive approach that focuses on early detection of hemodynamic decline and avoids the need for ECMO would be preferable. Additionally, when considering specific VAD types, the more favorable profile of less invasive devices should be framed in the context of each center's experience and device availability.

Given the perceived constraints on setting up randomized trials to test all these approaches, the adequate response will likely rest on the shoulders of well-trained, dedicated, multidisciplinary teams. Even though there is still a lack of consensus10 regarding this protocolized approach, based on locally established criteria and defined support pathways, it may pave the way toward better results. In the daunting task of achieving this goal, analyses such as those by Domínguez-Massa et al. will help shape the prognostic horizon for patients with advanced heart failure.