Progressive elimination of adverse events: the key to success for left ventricular assist devices

This article refers to ‘Primary results of long-term outcomes in the MOMENTUM 3 pivotal trial and continued access protocol study phase: a study of 2200 HeartMate 3 left ventricular assist device implants’ by M.R. Mehra et al., published in this issue on pages 1392–1400. Implantable left ventricular assist devices (LVAD) have undergone significant technological improvements over the past decades. From the bulky, noisy, and short-lived pulsatile devices, such as the HeartMate XVE in the 1990s, to the silent and much more durable continuous-flow devices with first clinical results reported in 20051 and now further developed with the full magnetically levitated centrifugal pumps with implants beginning a decade later.2, 3 Given the improved durability of the devices achieved, patients may now live more than 10 years with the same LVAD. Consequently, two main challenges need to be addressed to improve outcomes in this population further: (i) optimal patient selection and timing, (ii) elimination of remaining inherent adverse events. It has been repeatedly documented that heart failure (HF) patients are often referred late for evaluation for LVAD implantation and that this compromises outcomes for patients.4 Many patients undergo implantation in INTERMACS profiles 1 and 2, despite well-documented inferior survival compared to patients in INTERMACS 3–4. Moreover, many patients are never referred for evaluation. In a recent study, patients followed in eight European electrophysiology clinics with either an implantable cardioverter-defibrillator or cardiac resynchronization therapy were screened for the presence of advanced HF and indication for advanced therapies (LVAD or heart transplantation).5 Of patients with New York Heart Association III–IV symptoms and ejection fraction <40%, 26% either had an indication for transplantation or LVAD, without any contraindications, yet strikingly none of the patients had been referred for evaluation for these interventions. This is remarkable, especially as most of the participating electrophysiology clinics were located in hospitals with a fully operational advanced HF programme, including the option to implant LVADs. So clearly, despite recommendations from professional societies, under-referral to advanced HF centres remains a considerable problem.6 One reason for the late referral of patients for consideration for LVAD implantation may be the absence of a contemporary randomized trial comparing LVAD implantation to optimal medical care in elective patients.7 However, fear of adverse events in patients and referring physicians likely also plays a role. LVAD therapy is associated with a risk of bleeding, pump thrombosis, driveline or pump infection, stroke, and recurrence of HF symptoms. While patients in INTERMACS 1–3 (and, importantly, their physicians) often feel they have no choice but to opt for LVAD implantation, less advanced patients may weigh the risk of adverse events higher than the likelihood of longer survival, better exercise capacity and heightened quality of life. The ROADMAP study8 compared outcomes in patients with advanced HF in INTERMACS profile 4–7 depending on whether they opted for an LVAD (HeartMate II) implantation or not. Albeit non-randomized, the study documented that prognosis in electively LVAD implanted patients was associated with a higher chance of survival free from urgent LVAD implantation or transplantation. This difference was particularly prominent in patients in INTERMACS profile 4. Functional status and quality of life were better in the group treated electively with an LVAD than in the group treated with optimal medical therapy, despite the occurrence of adverse events in the LVAD group. This, again, highlights that timing of referral is essential as patients in INTERMACS 1–2 have a higher risk of adverse events than more stable patients.9 In this manner, adverse events play into the outcome of LVAD recipients, not just by their direct effect on the LVAD patient when they occur, but also indirectly as concerns about adverse events are likely an impactful physician decision-making factor resulting in late referral of relevant patients and thus, in turn, further compromise actual vs. desirable outcomes. To underscore this poignant dissociation, 1-year survival of advanced HF patients in INTERMACS profile 4 (mean age 57 years) treated medically is as low as 60%.10 Consequently, mitigation of adverse events represents a quintessential effort for the entire field of mechanical circulatory support. Fortunately, newer devices are helping the field in that direction. In an analysis published in this Journal, Mehra and colleagues report on long-term outcomes in 2200 patients implanted with a HeartMate 3 (HM3) either as part of the randomized MOMENTUM 3 study (comparing HeartMate II to HM3) or in the continued access protocol (CAP) where all patients received a HM3.11 This is the largest reported cohort of HM3 patients and as such the study provides essential data to inform health care professionals working in the field of advanced HF. The study has several important findings. First, it confirms in a large population, in considerable proportion not specifically selected for a randomized clinical trial (RCT), that survival after LVAD implantation is excellent at 81.2% after 2 years. Survival was numerically higher in the CAP than in the RCT cohort, even though the CAP population had several characteristics suggesting even more advanced disease at baseline compared with the RCT population. In fact, more patients in the CAP were in INTERMACS profile 1, and more patients were implanted with intra-aortic balloon pumps prior to LVAD surgery. This would have been expected to translate into worse survival but did not. Further, it validates that the need for exchange of the HM3 pump is very low, with less than 2% at 2 years. As some pump exchanges were likely performed because of the – now assumed to be historic – HM3-specific complication of outflow graft twist, pump exchange could be expected to become an even rarer event given design refinement now preventing this specific issue. Finally, it was demonstrated that the overall burden of adverse events and the rehospitalization rate were lower in the CAP cohort than in the RCT cohort. A paramount advantage of the study is that all endpoints were adjudicated. This is not the case for registries like INTERMACS or EUROMACS or even industry-owned registries such as ELEVATE. The absence of endpoint adjudication is associated with an inherent risk of under- or misreporting of important clinical events such as bleeding episodes. Hence, the results of the extended MOMENTUM 3 analysis provide both an inclusive (in terms of included patients) and comprehensive (in terms of adverse events) analysis of contemporary LVAD therapy. What do these findings tell us about LVADs as a therapeutic option for advanced HF patients in the current era? First, more than one-third of the patients in the CAP cohort were in INTERMACS 1–2. This proportion has been largely unchanged since the emergence of continuous-flow devices, including the European registries of HM3 implantations.12 So, while LVADs have been demonstrated to be indeed lifesaving in cardiogenic shock, are we still expanding the use in the more stable patients with advanced HF, where the devices are clearly associated with the best outcome? Second, the results demonstrate that survival with LVADs has steadily improved and, maybe even more importantly, that the incidence of adverse events is declining with improved technology and expanding knowledge of LVAD recipient management. Indeed, this is highly encouraging news that should be collectively embraced by the HF community when considering LVADs in the continuum of care for patients with advanced HF. Given the robustness of the analysis of Mehra et al., we can now with confidence quote lower risks of serious adverse events when we discuss pros and cons regarding elective LVAD implantation with patients and relatives as compared to previous magnitude-limited ROADMAP study data. Despite the now well-documented improvements in LVAD outcomes, there is still work to be done. Some adverse events must and can be minimized further. Bleeding appears to be a straightforward target given the better haemocompatibility of the HM3 pump compared with previous continuous-flow devices. A large randomized trial is currently testing whether the removal of aspirin from the antithrombotic regimen is safe and leads to lower bleeding risk.13 Exploratory trials of warfarin removal in selected patients14 have also been performed, and more anticoagulation modifying studies are underway. Driveline infections may be a thing of the past when systems for internalization of the power supply are proven effective and safe.15 Smart pumps regulating speed operation according to left ventricular filling pressures and patient activity may improve the exercise capacity of LVAD patients in the future. In summary, there is a tremendous need for ongoing development of the technology, emphasizing joint responsibility of the advanced HF academic milieu and the device industry, both current and hopefully soon incoming players. The recent withdrawal of a major device – the HVAD – from the market has raised concern that the development process may cool off for a period. For the sake of our patients with advanced HF, we need to make sure that this will not translate into a gloomy reality. Conflict of interest: F.G. is and advisor to Abbott. I.N. has nothing to disclose.