Contemporary left ventricular assist devices (LVADs) have improved survival and quality of life in patients with advanced heart failure. However, to reap these benefits our patients accept the important tradeoffs associated with this therapy; namely, bleeding, infection, neurologic events, and pump malfunction or failure.1 Although it has become customary to consider these adverse events or other negative clinical outcomes in the context of rigorously conducted clinical trials, it is equally important to acknowledge the key milestones that have decorated the evolution of this therapy. Mechanically assisted circulation has evolved dramatically over the past half-century. From the original large nondischargeable extracorporeal systems used for temporary circulatory support to the contemporary miniaturized, implantable, and durable pumps, we have witnessed a truly remarkable feat of iterative design and engineering that has afforded incremental improvement in our patients’ lives.2 In the last 2 decades, we have seen the transition from pulsatile devices to continuous flow pumps and more recently the design shift from the axial flow HeartMate II (Abbott, Abbott Park, IL) to centrifugal pumps such as the HeartMate 3 (Abbott, Abbott Park, IL) and the HeartWare HVAD (Medtronic, Framingham, MA). What’s more, the recent advances in magnetic levitation and artificial pulsatility have added another layer of improvements geared towards addressing adverse events and device reliability. At each step, newer devices have been tested meticulously against older generations in randomized controlled trials providing a strong evidence base for the clinical benefits of adopting newer technologies.
In this issue of the ASAIO Journal, Mahr et al.3 present a cross-trial comparison of neurologic outcomes between the HeartWare HVAD and the HeartMate 3. Although these two contemporary centrifugal pumps have never been tested head to head in a randomized controlled setting, each has been tested against the HeartMate II. In the ENDURANCE trial, HVAD was found to be noninferior to HeartMate II in terms of survival free of disabling stroke or reoperation at 2 years in a destination therapy population but was associated with significantly more stroke (29.7% vs. 12.1%; p < 0.001).4 A retrospective analysis revealed that elevated blood pressure was an independent risk factor for stroke in the HVAD population.5 As a result, the ENDURANCE Supplemental study was conducted to assess the effect of blood pressure control on the incidence of stroke among HVAD recipients. Despite successful lowering of mean arterial pressure, the difference in neurologic events which included strokes with a modified Rankin score > 0 and transient ischemic attacks failed to reach prespecified noninferiority margins compared with a HeartMate II control group.6 However, the ENDURANCE Supplemental cohort experienced less overall stroke (16.9% vs. 22.3%; p = 0.10) and less hemorrhagic stroke (5.2% vs. 10.5%; p = 0.02) than subjects in the primary trial. The MOMENTUM 3 trial that included patients who were considered for both bridge to transplant and destination therapy indications demonstrated that HeartMate 3 was associated with improved survival free of disabling stroke or reoperation at 2 years compared with the HeartMate II (79.5% vs. 60.2%; hazard ratio (HR), 0.46; p < 0.001) and importantly, also with a lower stroke risk (10.1% vs. 19.2%; HR, 0.47; p = 0.02).7
Taken together, these data are provocative and beg the question of how neurologic outcomes differ between the HeartMate 3 and HVAD devices. However, direct comparison between these two trials is hindered by key differences in study population, endpoint definition, reporting methodology, and possible differences in medical management during the trials. In an attempt to address these gaps, the current study assembled a new study cohort (HYBRID-HVAD) by combining clinical data sets from the ENDURANCE Supplemental trial and the continued access protocol (CAP) cohort of a previously performed HVAD study in bridge to transplant patients.6,8 The HYBRID-HVAD and MOMENTUM 3 cohorts were similar in terms of major baseline characteristics, the ratio of destination therapy to bridge to transplant patients, and the proportion of patients transplanted in 2 years. The authors compared “total neurologic events” given that the ENDURANCE and MOMENTUM trials used different specific definitions of adverse neurologic outcomes in an attempt to capture a broader, unbiased, and more directly comparable endpoint. The main finding of this study was that there was no significant difference in the incidence of overall neurologic events at 6 months or 2 years between the HYBRID-HVAD group and the MOMENTUM 3 populations.
This study seeks to lend insight into the important question of device selection between two competing but similar technologies and suggests that the potential difference in neurologic adverse events is negligible. Certainly, these findings challenge the more intuitive interpretation of the discordant neurologic outcomes data between the ENDURANCE and MOMENTUM clinical trials programs. Although the results of this analysis are provocative, several methodological concerns should prompt cautious interpretation. First, the finding of similar neurologic outcomes was based on incidence of total neurologic events that included strokes, transient ischemic attack (TIA), and other nonstroke neurologic events. This methodology was required because of differing definitions of neurologic outcomes between the trials.8 Unfortunately, despecifying the primary outcome variable detracts from our ability to distinguish between strokes and neurologic events less likely to cause long-term disability such as confusion, encephalopathy, or seizures as the authors note. Second, our ability to truly understand differences in stroke rates and the clinical implications of these events in this analysis is impossibly confounded by variability in definitions, observation periods, and accounting of clinical impact. Third, although combining the ENDURANCE Supplemental cohort with the ADVANCE cohort led to similar mean arterial pressures compared with MOMENTUM, exclusion of the larger ENDURANCE dataset biased the primary finding towards similarity in total neurologic events. Although it is contemporary to recognize that treatment of blood pressure is an important component of LVAD management, it would be interesting to know how the findings from the current study would change if the aggregated ENDURANCE cohort was included rather than just the ENDURANCE Supplemental cohort. Finally, the 2 year HeartMate II stroke rate in these contemporaneous trials was surprisingly discordant (12.1% vs. 19%), potentially related to variations in definitions or other undescribed confounders. Together, these issues raise the question: are we comparing different strokes for different folks?
It is clear that randomization and carefully conducted clinical trials are paramount to detect small signals that differentiate real clinical outcomes such as stroke between various LVAD technologies. What is not entirely clear is if this hybridization of two separate cohorts spanning different eras across two trials disambiguates or confounds the stroke risk between the HVAD and the HeartMate 3 devices. Furthermore, although this study intended to overcome the gaps that challenge cross-trial comparisons, we need more data either in the form of randomized controlled trials or longer contemporary registries that facilitate a closer comparison. Until then, clinical judgment and knowledge of the reported literature will guide decision-making. It is the responsibility of the LVAD team to have a clear and working understanding of these complex data to properly counsel patients and their families in a shared decision-making model. The crucial lessons we have learned over the last several decades that have enabled iterative improvement must continue to inform the design of the next generation of this technology and advance patient management strategies.
1. Slaughter MS, Rogers JG, Milano CA, et al.; HeartMate II Investigators: Advanced heart failure
treated with continuous-flow left ventricular assist device. N Engl J Med 2009.361: 2241–2251.
2. Drazner MH. A new left ventricular assist device - Better, but still not ideal. N Engl J Med 2018.378: 1442–1443.
3. Mahr C, Thinh Pham D, Mokadam NA, et al. Interpreting neurological outcomes in a changing trial design landscape: An analysis of aHVAD utilizing a hybrid intention to treat population. ASAIO 2019. doi: 10.1097/MAT.0000000000000931.
4. Rogers JG, Pagani FD, Tatooles AJ, et al. Intrapericardial left ventricular assist device for advanced heart failure
. N Engl J Med 2017.376: 451–460.
5. Teuteberg JJ, Slaughter MS, Rogers JG, et al.; ADVANCE Trial Investigators: The HVAD left ventricular assist device: Risk factors for neurological events and risk mitigation strategies. JACC Heart Fail 2015.3: 818–828.
6. Milano CA, Rogers JG, Tatooles AJ, et al.; ENDURANCE Investigators: HVAD: The ENDURANCE Supplemental Trial. JACC Heart Fail 2018.6: 792–802.
7. Mehra MR, Goldstein DJ, Uriel N, et al. Two-year outcomes with a magnetically levitated cardiac pump in heart failure
. N Engl J Med 2018.378: 1386–1395.
8. Slaughter MS, Pagani FD, McGee EC, et al.; HeartWare Bridge to Transplant ADVANCE Trial Investigators: HeartWare ventricular assist system for bridge to transplant: Combined results of the bridge to transplant and continued access protocol trial. J Heart Lung Transplant 2013.32: 675–683.