Heart transplantation (HT) remains the standard treatment for selected patients with advanced heart failure (HF). However, only a small proportion of patients with advanced HF ultimately undergo transplantation. The main limitation to expanding HT is the shortage of donor organs. In addition, HT requires strict adherence to immunosuppressive therapy, self-care measures, and prophylactic strategies to prolong graft survival and minimise complications.
Conversely, durable left ventricular assist devices (LVADs) have emerged as a transformative treatment that has dramatically changed the prognosis of selected patients with advanced HF who are ineligible for HT or cannot safely wait for a donor organ. Technological advances, together with a better understanding of the physiology of these patients and their clinical management, have substantially improved outcomes, with survival exceeding 90% at 1 year and 70% at 5 years.
Limited evidence is available regarding which of these two strategies offers better outcomes in middle-aged patients aged 50-64 years and in patients older than 65 years. This is the question addressed by Uriel et al. in a study recently published in JACC: Heart Failure. Importantly, the authors assessed not only survival after treatment but also the time spent on the transplant waiting list as an integral component of the survival analysis.
The investigators combined data from the MOMENTUM 3 trial, which included 1763 patients implanted with the HeartMate 3® (HM3) at 69 centres in the United States between 2014 and 2018, with data from the United Network for Organ Sharing (UNOS) registry. The latter included 5336 adults listed for HT during the same period. After applying the relevant exclusion criteria, the analysable population comprised 811 HM3 recipients and 3435 HT candidates in the middle-aged group, and 952 HM3 recipients and 1505 HT candidates in the group aged >65 years. Both age groups were evaluated using propensity score matching based on 12 clinically relevant covariates. Follow-up was 2 years for survival and 1 year for adverse events.
When survival was analysed from the time definitive treatment was received, HT was associated with significantly better 2-year survival than HM3 in both age groups. In the matched cohort aged 50-64 years, survival was 90.7% after HT compared with 83.8% after HM3 implantation (HR: 1.76; 95% CI: 1.29-2.42; p = .0004). Among patients aged >65 years, the difference was even greater: 87.5% versus 77.7% (HR: 1.89; 95% CI: 1.39-2.56; p < .0001).
However, the findings were reversed when the analysis started at the time of waitlist registration and the primary endpoint was defined as the composite of death or delisting due to clinical deterioration. Among middle-aged patients, 2-year survival with HM3 was 83%, compared with 75% freedom from death or delisting due to deterioration among matched HT candidates (HR: 0.62; 95% CI: 0.49-0.78; p < .0001). This represented an absolute difference of 8 percentage points in favour of LVAD therapy. In patients aged >65 years, the numerical difference also favoured HM3, at 76.6% versus 72.2%, although statistical significance was not reached (HR: 0.81; 95% CI: 0.65-1.01; p = .064).
Differences were also observed in adverse events during follow-up. Infection-related hospitalisation was more frequent after HT: 32.8% versus 22.4% in patients aged 50-64 years and 36.0% versus 21.6% in those aged >65 years (p < .0001). Stroke rates were similar between strategies in the middle-aged group. However, among patients aged >65 years, HM3 recipients had a significantly higher rate of disabling stroke than HT recipients (5.1% vs 2.1%; p = .011). Rates of renal dysfunction requiring dialysis were similar between strategies in both age groups.
COMMENTARY:
The study by Uriel et al. makes a relevant contribution to the debate surrounding the treatment of advanced HF and directly complements their previous publication comparing the same two strategies, HT and HM3, in patients aged <50 years. Their earlier findings suggested that survival after definitive treatment was comparable in younger patients, at 88.7% in the HM3 group and 90.2% in the HT group (p = .53). Nevertheless, freedom from the composite endpoint of death or delisting due to clinical deterioration was significantly higher with HM3 than with HT listing (90.1% vs 76.7%; HR: 0.38; p < .0001).
Taken together, the two studies suggest that survival after HM3 implantation and HT is comparable in younger patients, whereas HT as the initial strategy is associated with better outcomes in patients aged 50 years or older. However, when the probability of the composite outcome of death or removal from the waiting list is considered, HM3 appears to provide a clear advantage. These findings nevertheless require careful interpretation to avoid misleading conclusions.
Probably the most important limitation is the study period, from 2014 to 2018, which preceded the reform of the US heart allocation system. In 2018, the previous three-tier urgency system was replaced by a six-status framework. This change markedly reduced the use of durable LVADs as a bridge to transplantation and shortened both waiting times and waitlist mortality, profoundly altering the dynamics of HT in the United States. The waiting times reported in this study therefore reflect a clinical reality that no longer applies to the current US system and also differs from European transplantation systems, including the Spanish model. Direct extrapolation of these findings to contemporary practice should consequently be undertaken with considerable caution.
The study also combines heterogeneous data sources: a clinical trial characterised by strict selection criteria, prospective follow-up, and rigorous event definitions, and a national administrative registry with less standardised definitions. This asymmetry is particularly problematic when complications are compared. In MOMENTUM 3, infection required specific microbiological and clinical criteria, whereas in UNOS any infection-related hospitalisation was sufficient. Similarly, stroke in MOMENTUM 3 was defined as a disabling event with a modified Rankin Scale score >3, whereas UNOS recorded any stroke. Direct comparison of adverse events therefore partly involves comparing different definitions, and the observed differences should be interpreted cautiously.
The study’s time horizon also raises important questions. A follow-up of 2 years for survival and 1 year for adverse events is insufficient to fully answer the question posed by the authors. The benefits of HT are more clearly expressed over the medium and long term. Moreover, the post-transplant complication profile is strongly influenced by the intensity of immunosuppression, which is necessarily greatest during the initial months. This may bias the short-term adverse-event analysis in favour of HM3.
Long-term LVAD outcomes, quality of life, the burden of daily care—including anticoagulation, driveline management, and infection risk—and the psychosocial impact on patients and their families were not assessed. These dimensions are nevertheless crucial when treatment decisions are individualised.
Finally, although this study represents an important methodological advance by comparing these two strategies using propensity score matching, its objective of estimating net survival benefit was only partially achieved. The clinical course of patients implanted with an HM3 as a bridge to transplantation who subsequently underwent HT was not captured. This is a clinically and ethically relevant issue because transplantation after LVAD support is associated with a higher risk of perioperative mortality and primary graft dysfunction, factors that must necessarily be incorporated into therapeutic decision-making.
In conclusion, the central message of the study is robust: LVAD therapy should not be regarded as an inferior alternative to transplantation, but rather as a treatment option with its own outcomes that should be discussed with patients while taking the actual waiting-list period into account. Nevertheless, the limitations described require a balanced interpretation, and the optimal strategy must remain, and will probably continue to remain, individualised.
REFERENCE:
Uriel N, Sayer GT, Colombo PC, Yuzefpolskaya M, Richter I, Goldstein DJ, et al. Survival Outcomes in Middle-Aged and Older Patients With Advanced Heart Failure: A Propensity-Matched Analysis of HeartMate 3 LVAD and Heart Transplant Using MOMENTUM 3 and UNOS Registry. JACC Heart Fail. 2026 May 10:103159. doi: 10.1016/j.jchf.2026.103159. Epub ahead of print. PMID: 42262301.
