Abstract
Aims
Left ventricular assist devices (LVADs) are a critical intervention for advanced heart failure (HF), serving as destination therapy or bridge to transplantation. Obesity and diabetes impact outcomes in patients with LVADs. Glucagon‐like peptide‐1 receptor agonists (GLP1‐RAs) demonstrate cardiovascular benefits; however, their role in patients with LVADs remains underexplored. We evaluated the association of GLP1‐RA therapy with cardiovascular outcomes in patients with LVADs.
Methods
This retrospective cohort study used the TriNetX database, a research network database from 98 healthcare organizations. We queried for all adult LVAD recipients (≥18 years) and were stratified into GLP1‐RA users and non‐users. Propensity score matching (PSM) (1:1) balanced demographics, comorbidities, medication use and laboratory data. Outcomes included heart transplantation rates, heart failure hospitalizations, all‐cause mortality, all‐cause hospitalizations and cardiovascular events. Logistic regression models were used to estimate adjusted odds ratios (aOR).
Results
After PSM, we included a total of 1036 adult LVAD recipients (518 GLP1‐RA users, 518 matched non‐users) with a mean follow‐up time of 311.6 ± 98.4 days for the GLP1‐RA cohort and 304.0 ± 111.5 days for the non‐GLP1‐RAs cohort. Mean age was 56.7 ± 12.2 years in the GLP1‐RA cohort and 58.0 ± 12.5 years in the non‐GLP1‐RA cohort. Females comprised 28.0% of both cohorts while White patients represented 52.1% of the GLP1‐RA group and 53.1% of the non‐GLP1‐RA group. GLP1‐RA users had higher heart transplantation rates [n = 98 (18.9%) vs. n = 44 (8.5%); aOR 2.514 (95% CI: 1.720–3.673)]. Acute HF events and all‐cause hospitalizations were lower among GLP1‐RA users compared with non‐users [n = 288 (55.6%) vs. n = 357 (68.9%); aOR 0.565 (95% CI: 0.438–728) and n = 324 (62.5%) vs. n = 390 (75.3%); aOR 0.548 (95% CI: 0.420–0.716)]. No differences were observed when comparing the GLP1‐RA cohort with the non‐GLP1‐RA cohort in regard to all‐cause mortality [n = 32 (6.2%) vs. n = 44 (8.5%); aOR 0.709 (95% CI: 0.442–1.138)], stroke [n = 42 (8.1%) vs. n = 58 (11.2%); aOR 0.700 (95% CI: 0.461–1.062)] or cardiac arrest [n = 18 (3.5%) vs. n = 17 (3.3%); aOR 1.061 (95% CI: 0.541–2.082)].
Conclusions
GLP1‐RA therapy in patients with advanced HF and LVADs is potentially associated with improved heart transplantation rates while decreasing hospitalization and acute HF event rates.
Key question: does GLP1‐RA therapy improve heart transplantation rates and clinical outcomes in patients with LVADs? Key findings: in this propensity score‐matched cohort study of 1068 LVAD recipients, GLP1‐RA use was associated with higher heart transplantation rates (18.9% vs. 8.5%) and lower risks of acute HF events and all‐cause hospitalizations.
Introduction
Cardiovascular disease (CVD) remains a leading cause of morbidity and mortality globally. 1 Among its manifestations, advanced heart failure (HF) is associated with particularly high mortality rates. Left ventricular assist devices (LVADs) play an important role in this context, serving either as destination therapy or as a bridge to heart transplantation. 2 LVADs significantly improve survival and quality of life for patients awaiting transplantation or as a long‐term treatment option. Importantly, the duration of LVAD support can also provide an opportunity to optimize comorbid conditions such as obesity and diabetes, which are known to impact both short‐ and long‐term outcomes after heart transplantation. 3 Obesity has been shown to lead to higher LVAD complications along with worse outcomes post‐heart transplant. 2 , 4 , 5 , 6 One such therapy for obesity management includes bariatric surgical procedures, which has been shown to lead to increased successful heart transplantation rates in observational studies. 7
Glucagon‐like peptide‐1 receptor agonists (GLP1‐RAs), a therapeutic class initially developed for diabetes management, have demonstrated substantial cardiovascular benefits, particularly in reducing HF exacerbations, atherosclerotic events and systemic inflammation. 8 , 9 , 10 These agents are also associated with significant weight loss, offering an alternative to other obesity treatments such as bariatric surgery. However, the use of GLP1‐RAs in patients with LVADs remains underexplored. This study aimed to investigate the association between GLP1‐RA therapy and key outcomes, including heart transplantation rates and cardiovascular events, in LVAD‐supported patients.
Methods
The data for this study were independently queried, analysed and interpreted by the authors. As the analysis relied on aggregated, de‐identified data from a publicly available federated research network database, institutional review board (IRB) approval was not required. This study is a secondary analysis of already existing data, which involved no interaction or intervention with human subjects. De‐identification measures were completed in compliance with the HIPAA Privacy Rule. The study adheres to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines for cohort studies.
We utilized the TriNetX Network platform, an international, collaborative, multi‐centre database with extensive representation primarily in the United States. At the time of this study, the platform contained approximately 135 000 000 de‐identified patient records sourced from electronic health records across 104 participating healthcare organizations. This study used a retrospective cohort design with propensity score matching (PSM) to examine adult patients (≥18 years) with a history of LVAD implantation between 1 January 2017 and 31 December 2024. Patients were divided into two cohorts based on the use of GLP1‐RAs: those receiving GLP1‐RA therapy and a matched cohort of patients with LVADs not receiving GLP1‐RAs. Use of GLP1‐RA therapy was defined as being on any of the pre‐specified types of GLP1‐RA therapy (semaglutide, liraglutide or dulaglutide) starting prior to date of LVAD implantation or up to 1 month post‐LVAD implantation. The control group had no use of GLP1‐RA therapy either before or after LVAD implantation. Time zero was defined as 1 month post‐LVAD implantation in both cohorts. Matching was performed to minimize baseline differences and ensure comparability between the groups. We matched both cohorts based on demographic characteristics (i.e., age, sex, race and ethnicity), comorbidity burden (i.e., hypertension, dyslipidaemia, obesity, cerebral infarction, chronic kidney disease and diabetes mellitus), medication use [i.e., beta‐blockers, antilipemic agents, angiotensin‐converting enzyme inhibitors (ACE‐I), angiotensin receptor blockers (ARBs), sodium‐glucose co‐transporter 2 inhibitors (SGLT2‐I), loop diuretics, mineralocorticoid receptor antagonists, and platelet aggregation inhibitors] and laboratory data (i.e., glomerular filtration rate, body mass index and haemoglobin A1C). We also adjusted for potential health hazards related to psychosocial and socioeconomic circumstances through query of ICD10 codes Z55–Z65. Outcomes of interest included the rate of heart transplantation, cardiovascular events such as cardiac arrest, acute myocardial infarction (AMI) and acute HF events, as well as non‐cardiovascular outcomes including all‐cause mortality and all‐cause hospitalizations (Supporting information S1). Outcomes were assessed from 1 month following the time of LVAD implantation through 3 years post‐implantation.
Statistical analysis
Patients were stratified into two cohorts based on GLP1‐RA use. Continuous variables were reported as mean ± standard deviation (SD) and compared using independent‐sample t‐tests while categorical variables were expressed as frequencies (n, %) and analysed using χ 2 tests. PSM was conducted in a 1:1 ratio using a greedy nearest‐neighbour algorithm with a calliper width of 0.1 pooled SDs. Adjusted odds ratios (aORs) with 95% confidence intervals (CIs) were calculated through logistic regression models for all pre‐specified outcomes. Kaplan–Meier survival curves were constructed to evaluate time‐to‐event outcomes, and log‐rank tests were used to assess differences between cohorts. Cox proportional hazard models were used to estimate hazard ratios (HRs). A two‐tailed P value of <0.05 was considered statistically significant. Statistical analyses were performed using the TriNetX platform and R for computational analysis.
Results
A total of 534 patients were included in the LVAD with GLP1‐RAs cohort and 3674 in the LVAD without GLP1‐RAs cohort prior to PSM (Table 1 ). Following matching, both cohorts included 518 patients. Baseline differences were successfully minimized, achieving balance in key demographic, clinical and laboratory variables (standardized differences <0.1) (Table 2 and Figure 1 ). The mean follow‐up time after matching was 311.6 (SD 98.4) days for the GLP1‐RA cohort and 304.0 (SD 111.5) days for the non‐GLP1‐RA cohort.
Table 1.
Baseline characteristics before PSM. Baseline characteristics before the PSM, including demographic factors, comorbidity burden and medication use
| Characteristics | LVAD with GLP1‐RAs | LVAD without GLP1‐RAs | Standardized difference |
|---|---|---|---|
| Age (mean ± SD) | 56.2 ± 12.4 | 60.6 ± 14.4 | 0.326 |
| White (%) | 52.40% | 56.80% | 0.088 |
| Female (%) | 29.20% | 23.10% | 0.14 |
| Hispanic (%) | 8.20% | 6.30% | 0.073 |
| Black (%) | 33.30% | 26.60% | 0.148 |
| Hypertension (%) | 98.30% | 80.90% | 0.594 |
| Lipid disorders (%) | 85.40% | 64.10% | 0.506 |
| Obesity (%) | 79.40% | 41.60% | 0.837 |
| Cerebral infarction (%) | 20.80% | 13.40% | 0.197 |
| Chronic kidney disease (%) | 74.70% | 60.80% | 0.302 |
| Diabetes mellitus (%) | 81.60% | 47.80% | 0.758 |
| Sleep apnoea | 62.90% | 36.80% | 0.541 |
| Social influences (Z coding) (%) | 16.30% | 9.10% | 0.217 |
| Beta‐blockers (%) | 94.40% | 78.90% | 0.467 |
| Antilipemic agents (%) | 87.30% | 67.10% | 0.494 |
| ACE inhibitors (%) | 65.90% | 49.40% | 0.34 |
| Angiotensin II inhibitors (%) | 78.10% | 52.60% | 0.557 |
| SGLT2 inhibitors (%) | 60.10% | 23.50% | 0.8 |
| Loop diuretics (%) | 99.40% | 92.10% | 0.371 |
| Spironolactone (%) | 89.50% | 67.20% | 0.563 |
| Platelet aggregation inhibitors (%) | 94.80% | 74.80% | 0.579 |
| Anticoagulants | 98.90% | 90.90% | 0.370 |
| Estimated glomerular filtration rate (mean ± SD) | 60.8 ± 25.8 | 61.1 ± 25.8 | 0.012 |
| Body mass index (mean ± SD) | 35.8 ± 7.6 kg/m2 | 29.4 ± 7.1 kg/m2 | 0.872 |
| Haemoglobin A1C (mean ± SD) | 7.1 ± 1.7 | 6.5 ± 1.4 | 0.432 |
Abbreviations: GLP1‐RAs, glucagon‐like peptide‐1 receptor agonists; LVAD, left ventricular assist device; PSM, propensity score matching; SD, standard deviation.
Table 2.
Baseline characteristics after PSM. Baseline characteristics after the PSM, including demographic factors, comorbidity burden and medication use
| Characteristics | LVAD with GLP1‐RAs | LVAD without GLP1‐RAs | Standardized Difference |
|---|---|---|---|
| Age (mean ± SD) | 56.7 ± 12.2 | 58.0 ± 12.5 | 0.105 |
| White (%) | 52.10% | 53.10% | 0.019 |
| Female (%) | 28.00% | 28.00% | <0.001 |
| Hispanic (%) | 8.30% | 9.80% | 0.054 |
| Black (%) | 33.40% | 32.20% | 0.025 |
| Hypertension (%) | 98.30% | 98.50% | 0.015 |
| Lipid disorders (%) | 85.30% | 86.50% | 0.033 |
| Obesity (%) | 78.80% | 77.80% | 0.023 |
| Cerebral infarction (%) | 20.70% | 19.30% | 0.034 |
| Chronic kidney disease (%) | 74.70% | 75.30% | 0.013 |
| Diabetes mellitus (%) | 81.30% | 86.30% | 0.136 |
| Sleep apnoea | 62.40% | 62.50% | 0.004 |
| Social influences (Z coding) (%) | 16.20% | 16.00% | 0.005 |
| Beta‐blockers (%) | 94.40% | 94.80% | 0.017 |
| Antilipemic agents (%) | 87.30% | 87.80% | 0.018 |
| ACE inhibitors (%) | 65.30% | 61.20% | 0.084 |
| Angiotensin II inhibitors (%) | 77.80% | 77.80% | <0.001 |
| SGLT2 inhibitors (%) | 58.90% | 62.20% | 0.067 |
| Loop diuretics (%) | 99.40% | 99.40% | <0.001 |
| Spironolactone (%) | 89.20% | 88.20% | 0.03 |
| Platelet aggregation inhibitors (%) | 94.60% | 93.10% | 0.064 |
| Anticoagulants | 98.8% | 98.3% | 0.048 |
| Estimated glomerular filtration rate (mean ± SD) | 60.7 ± 25.8 | 58.9 ± 22.2 | 0.074 |
| Body mass index (mean ± SD) | 35.6 ± 7.3 kg/m2 | 33.8 ± 7.4 kg/m2 | 0.245 |
| Haemoglobin A1C (mean ± SD) | 7.1 ± 1.7 | 7.1 ± 1.6 | 0.051 |
Abbreviations: GLP1‐RAs, glucagon‐like peptide‐1 receptor agonists; LVAD, left ventricular assist device; PSM, propensity score matching; SD, standard deviation.
Figure 1.
Propensity score density function. Pre‐ and post‐propensity score matching. *Purple = LVAD with GLP1‐RAs; green = LVAD without GLP1‐RAs. GLP1‐RAs, glucagon‐like peptide‐1 receptor agonists; LVAD, left ventricular assist device.
Patients in the GLP1‐RA cohort had a significantly higher rate of heart transplantation compared with the non‐GLP1‐RA cohort [18.9% vs. 8.5%; aOR: 2.514 (95% CI: 1.720–3.673)] (Table 3 ). Kaplan–Meier analysis showed improved transplant‐free survival in the GLP1‐RA group [HR: 2.365 (95% CI: 1.657–3.376), P < 0.001] (Figure 2 ). Patients receiving GLP1‐RAs had a lower risk of acute HF events [risk: 55.6% vs. 68.9%; aOR: 0.565 (95% CI: 0.438–0.728)]. Kaplan–Meier analysis demonstrated a higher HF hospitalization‐free survival probability at the end of the study window [40.3% vs. 23.9%; HR: 0.582 (95% CI: 0.498–0.681)]. No significant differences were observed for AMI rates among the GLP1‐RA cohort versus the non‐GLP1‐RA cohort [risk: 9.5% vs. 10.8%, respectively; aOR: 0.862 (95% CI: 0.575–1.291)] (Figure 3 ).
Table 3.
Outcomes between cohorts. Comparison between the two cohorts (LVAD with and without GLP1‐RAs) in cardiovascular, heart transplant and non‐cardiovascular (all‐cause mortality and hospitalization) outcomes
| Outcome | LVAD with GLP1‐RAs (N = 503) events | LVAD with GLP1‐RAs (%) | LVAD without GLP1‐RAs (N = 503) events | LVAD without GLP1‐RAs (%) | Adjusted odds ratio (95% CI) | P value | Hazard ratio (95% CI) |
|---|---|---|---|---|---|---|---|
| Heart Transplant | 98 | 18.9 | 44 | 8.5 | 2.514 (1.720–3.673) | <0.001 | 2.365 (1.657–3.376) |
| All‐cause hospitalization | 324 | 62.5 | 390 | 75.3 | 0.548 (0.420–0.716) | <0.001 | 0.589 (0.508–0.683) |
| Acute HF Events | 288 | 55.6 | 357 | 68.9 | 0.565 (0.438–0.728) | <0.001 | 0.582 (0.498–0.681) |
| All‐cause mortality | 32 | 6.2 | 44 | 8.5 | 0.709 (0.442–1.138) | 0.153 | 0.707 (0.449–1.116) |
| Cardiac arrest | 18 | 3.5 | 17 | 3.3 | 1.061 (0.541–2.082) | 0.863 | 1.026 (0.529–1.991) |
| Stroke | 42 | 8.1 | 58 | 11.2 | 0.700 (0.461–1.062) | 0.092 | 0.684 (0.460–1.017) |
| AMI | 49 | 9.5 | 56 | 10.8 | 0.862 (0.575–1.291) | 0.471 | 0.833 (0.568–1.223) |
Abbreviations: AMI, acute myocardial infarction; CI, confidence interval; GLP1‐RAs, glucagon‐like peptide‐1 receptor agonists; HF, heart failure; LVAD, left ventricular assist device.
Figure 2.
Kaplan–Meier survival probabilities. Heart transplant‐free survival probabilities among the two cohorts (LVAD with and without GLP1‐RAs). *Purple = LVAD with GLP1‐RAs; green = LVAD without GLP1‐RAs. *Patients in each cohort: 518. Patients with outcome: 98 in GLP1‐RA cohort (survival probability = 78.92%) and 44 in non‐GLP1‐RA cohort (survival probability = 89.56%). GLP1‐RAs, glucagon‐like peptide‐1 receptor agonists; LVAD, left ventricular assist device. *Index date of both cohorts: one month post‐LVAD implantation.
Figure 3.
Kaplan–Meier survival probabilities. Event‐free survival probabilities among the two cohorts (LVAD with and without GLP1‐RAs). *Purple = LVAD with GLP1‐RAs; green = LVAD without GLP1‐RAs. *Acute HF = patients in each cohort: 518. Patients with outcome: 288 in GLP1‐RA cohort (survival probability = 40.33%) and 357 in non‐GLP1‐RA cohort (survival probability = 23.86%). *AMI = patients in each cohort: 518. Patients with outcome: 49 in GLP1‐RA cohort (survival probability = 89.58%) and 56 in non‐GLP1‐RA cohort (survival probability = 87.67%). AMI, acute myocardial infarction; GLP1‐RAs, glucagon‐like peptide‐1 receptor agonists; HF, heart failure; LVAD, left ventricular assist device. *Index date of both cohorts: one month post‐LVAD implantation.
All‐cause hospitalization was also lower in the GLP1‐RA group [62.5% vs. 75.3%; aOR: 0.548 (95% CI: 0.420–0.716)], with improved median hospitalization‐free days (164 vs. 50 days) and lower hazard for hospitalization [HR: 0.589 (95% CI: 0.508–0.683)]. All‐cause mortality occurred in 6.2% of the GLP1‐RA group versus 8.5% in the control group [aOR: 0.709 (95% CI: 0.442–1.138)]. While the survival difference was not statistically significant (P = 0.134), the Kaplan–Meier analysis suggested a trend towards improved survival [92.9% vs. 90.3% at 1 year; HR: 0.707 (95% CI: 0.449–1.116)] (Figure 4 ).
Figure 4.
Kaplan–Meier survival probabilities. Event‐free survival probabilities among the two cohorts (LVAD with and without GLP1‐RAs). *Purple = LVAD with GLP1‐RAs; green = LVAD without GLP1‐RAs. *All‐cause mortality = patients in each cohort: 518. Patients with outcome: 32 in GLP1‐RA cohort (survival probability = 92.91%) and 44 in non‐GLP1‐RA cohort (survival probability = 90.28%). *All‐cause hospitalizations = patients in each cohort: 518. Patients with outcome: 324 in GLP1‐RA cohort (survival probability = 32.77%) and 390 in non‐GLP1‐RA cohort (survival probability = 18.78%). GLP1‐RAs, glucagon‐like peptide‐1 receptor agonists; LVAD, left ventricular assist device. *Index date of both cohorts: one month post‐LVAD implantation.
No statistically significant differences were found for stroke [risk: 8.1% vs. 11.2%; aOR: 0.700 (95% CI: 0.461–1.062)] or cardiac arrest [risk: 3.5% vs. 3.3%; aOR: 1.061 (95% CI: 0.541–2.082)] (Figure 5 ).
Figure 5.
Kaplan–Meier survival probabilities. Event‐free survival probabilities among the two cohorts (LVAD with and without GLP1‐RAs). *Purple = LVAD with GLP1‐RAs; green = LVAD without GLP1‐RAs. *Stroke = patients in each cohort: 518. Patients with outcome: 42 in GLP1‐RA cohort (survival probability = 90.87%) and 58 in non‐GLP1‐RA cohort (survival probability = 86.97%). *Cardiac arrest = patients in each cohort: 518. Patients with outcome: 18 in GLP1‐RA cohort (survival probability = 96.05%) and 17 in non‐GLP1‐RA cohort (survival probability = 96.16%). GLP1‐RAs, glucagon‐like peptide‐1 receptor agonists; LVAD, left ventricular assist device. *Index date of both cohorts: one month post‐LVAD implantation.
Discussion
Obesity is a growing public health concern and remains an independent risk factor for both HF and poor outcomes after heart transplantation. Despite the substantial body of research on obesity's impact on outcomes following LVAD implantation, many studies are limited by small sample sizes. 11 Furthermore, randomized controlled trials (RCTs) in this area are scarce, constrained by ethical and practical challenges, as well as difficulties in patient recruitment. Our study addresses this gap by investigating the role of GLP1‐RAs in patients with LVADs. To the best of our knowledge, this is the first real‐world retrospective cohort of this size evaluating GLP1‐RAs in this patient population. Our findings reveal that GLP1‐RAs use is potentially associated with higher rates of heart transplantation listing and reduced adverse cardiovascular events, which may reflect improved diabetic control and weight loss, as all of our included patients were at least consider overweight with a body mass index (BMI) >25 and majority had diabetes (>85%). These results highlight important clinical insights into the net benefits for this vulnerable population.
Although several medications have been approved for obesity management, the use of anti‐obesity medications in the context of LVAD implantation remains underexplored. 3 , 12 Our findings suggest that patients with LVADs who were on GLP1‐RAs were more likely to be listed for heart transplantation and less likely to experience adverse cardiovascular outcomes. This aligns with prior research on anti‐obesity medications in populations with HF, although not specifically those with mechanical circulatory support. 13 , 14 For example, earlier trials investigated the impact of orlistat in patients with HF with reduced ejection fraction (HFrEF), demonstrating improvements in NYHA functional classification, 6 min walk distance and significant weight loss compared with placebo. 15 However, results with GLP1‐RAs have been mixed. The LIVE trial (2016), one of the first to assess GLP1‐RAs in chronic, stable HFrEF, showed no significant improvement in cardiac function compared with placebo and noted an increase in serious cardiac events. 16 Similarly, the FIGHT trial (2016), which evaluated liraglutide in patients post‐acute HFrEF exacerbation, reported no reductions in rehospitalization rates or mortality. 17 In contrast, newer studies such as the STEP‐HFpEF, STEP‐HFpEF DM and a subgroup analyses of the SELECT trials demonstrated significant benefits of semaglutide in patients with HF both reduced and preserved ejection fraction, 8 , 9 , 10 , 18 which included substantial weight loss, improved quality of life and favourable secondary outcomes like reduced hospitalizations and mortality. The discrepancy between these studies that investigated liraglutide and semaglutide highlights the heterogeneity in GLP1‐RAs effects across HF phenotypes and emphasizes the need for more nuanced investigations, particularly in populations with advanced HF and mechanical circulatory support.
Our inclusion of all GLP1‐RA agents in the analysis acknowledges these distinctions in the literature. For example, the FIGHT trial did not evaluate GLP1‐RAs as obesity medications, limiting its relevance to weight‐loss outcomes. Furthermore, the endpoints of these trials differed significantly: while FIGHT used a global rank score encompassing mortality, hospitalization and N terminal pro brain natriuretic peptide levels, STEP‐HFpEF focused on patient‐reported outcomes like symptom burden and changes in body weight, with hospitalization and mortality as secondary outcomes. The baseline characteristics of the patients in these studies also varied; for example, the higher prevalence of diabetes in FIGHT (59%) contrasts with the STEP‐HFpEF trials, which primarily differentiated cohorts based on comorbid diabetes status. Given that our analysis sought to investigate GLP1‐RAs as a whole in the setting of advanced HF with LVAD implantation, inclusion of all GLP1‐RAs should prompt future investigation to identify differences within this class of medications.
The prevalence of advanced HF has necessitated the use of mechanical circulatory support, such as LVADs, for patients awaiting heart transplantation. 4 , 19 The complex relationship between obesity and heart transplant outcomes is highlighted by the International Society for Heart and Lung Transplantation (ISHLT) guidelines, which have evolved over time. The 2016 guidelines identified a BMI ≥ 35 kg/m2 as a relative contraindication to heart transplantation, whereas the 2022 guidelines acknowledged the limitations of BMI as a measure of health. 2 Patients with obesity and LVADs face unique challenges, with studies showing higher rates of adverse events such as neurological complications, device thrombosis and increased mortality. 4 , 5 For example, a 2020 analysis demonstrated higher rates of LVAD thrombosis among obese patients, 4 and a meta‐analysis of 26 842 patients confirmed these findings. 5 Similarly, a 2023 retrospective study of 11 216 LVAD recipients found higher mortality rates in patients with comorbid obesity compared with those without. 6 Bariatric surgery has been investigated as a potential solution for weight management in patients with LVADs. A 2023 meta‐analysis of 271 patients who underwent bariatric surgery after LVAD implantation found that 67.4% were subsequently listed for heart transplantation, and 32.5% underwent successful transplantation. 7 However, no studies to date have evaluated the impact of GLP1‐RAs in patients with LVADs, making our findings particularly novel.
Our study's retrospective design introduces inherent limitations, including the potential for misclassification of diagnoses due to reliance on medical record data and lack of granular data availability. While we used PSM to analyse two distinct cohorts based on GLP1‐RA usage, the adequacy of appropriate matching based on severity of illness is limited, which can impact outcomes. For example, frailty status remains an important prognostic factor in outcomes in this vulnerable population; however, appropriate adjustments for frailty are not possible. The TriNetX platform used for this analysis is also an international, multi‐centre repository and the difference in underlying practices may also impact the treatment and related outcomes in these vulnerable populations with advanced HF, including time to transplant. Furthermore, although we included all patients with LVAD implantation, the ultimate goal of therapy, whether destination therapy, bridge to transplantation or eventual LVAD explantation, was unable to be assessed. This also includes potential bridge to recovery or bridge to transplant ability, which we are unable to ascertain. We are also unable to query for heart transplant listing status, which may impact outcomes, particularly heart transplantation rates. Lastly, we are unable to account for advancements in LVAD therapies during our study period, as LVAD technology and clinical indications have changed over time. Nonetheless, this analysis is strengthened by the cohort sizes and the first study to evaluate the impact of this class of anti‐obesity medications in patients with LVADs.
Conclusions
Although our findings are hypothesis generating, it highlights the potential use of GLP1‐RAs in patients with LVADs as a therapeutic option to improve transplant eligibility and reduce adverse cardiovascular outcomes. This study emphasizes the importance of future clinical trials to validate these findings.
Conflict of interest statement
Authors have no conflict of interest or disclosures.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial or not‐for‐profit sectors.
Supporting information
Table S1. Outcome Definitions.
Acknowledgements
Graphical abstract was created with Biorender.com
Ibrahim, R. , Nhat, H. , Abdelnabi, M. , Forst, B. , Allam, M. , Mee, X. C. , Lim, G. K. , Bcharah, G. , Barry, T. , Farina, J. , Ayoub, C. , Arsanjani, R. , and Lee, K. (2025) Use of glucagon‐like peptide‐1 receptor agonists in patients with left ventricular assist devices. ESC Heart Failure, 12: 4326–4335. 10.1002/ehf2.15379.
References
- 1. Joynt Maddox KE, Elkind MSV, Aparicio HJ, Commodore‐Mensah Y, de Ferranti SD, Dowd WN, et al. Forecasting the burden of cardiovascular disease and stroke in the United States through 2050‐prevalence of risk factors and disease: a presidential advisory from the American Heart Association. Circulation 2024;150:e65‐e88. doi: 10.1161/CIR.0000000000001256 [DOI] [PubMed] [Google Scholar]
- 2. Mehra MR, Canter CE, Hannan MM, Semigran MJ, Uber PA, Baran DA, et al. The 2016 International Society for Heart Lung Transplantation listing criteria for heart transplantation: a 10‐year update. J Heart Lung Transplant 2016;35:1‐23. doi: 10.1016/j.healun.2015.10.023 [DOI] [PubMed] [Google Scholar]
- 3. Ibrahim R, Pham HN, Vest AR, William P. Managing obesity in heart failure: latest evidence and knowledge gaps. Curr Treat Options Cardiovasc Med 2024;26:355‐367. doi: 10.1007/s11936-024-01056-0 [DOI] [Google Scholar]
- 4. Zhigalov KSM, Arjomandi Rad A, Vardanyan R, Goerdt L, Chrosch T, Zhigalov K, et al. The impact of obesity on left ventricular assist device outcomes. Medicina (Kaunas) 2020;56:556. doi: 10.3390/medicina56110556 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Khan MS, Yuzefpolskaya M, Memon MM, Usman MS, Yamani N, Garan AR, et al. Outcomes associated with obesity in patients undergoing left ventricular assist device implantation: a systematic review and meta‐analysis. ASAIO J 2020;66:401‐408. doi: 10.1097/MAT.0000000000001019 [DOI] [PubMed] [Google Scholar]
- 6. Alba AC, Kirklin JK, Cantor RS, Deng L, Ross HJ, Jacobs JP, et al. The impact of obesity and LVAD‐bridging on heart transplant candidate outcomes: a linked STS INTERMACS–OPTN/UNOS data analysis. J Heart Lung Transplant 2023;42:1587‐1596. doi: 10.1016/j.healun.2023.06.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Sharma I, Nakanishi H, Hage K, Marrero K, Diwan TS, daSilva‐deAbreu A, et al. Bariatric surgery and left ventricular assist device in patients with heart failure: a systematic review and meta‐analysis. Am J Surg 2023;226:340‐349. doi: 10.1016/j.amjsurg.2023.06.014 [DOI] [PubMed] [Google Scholar]
- 8. Kosiborod MN, Abildstrom SZ, Borlaug BA, Butler J, Rasmussen S, Davies M, et al. Semaglutide in patients with heart failure with preserved ejection fraction and obesity. N Engl J Med 2023;389:1069‐1084. doi: 10.1056/NEJMoa2306963 [DOI] [PubMed] [Google Scholar]
- 9. Kosiborod MN, Petrie MC, Borlaug BA, Butler J, Davies MJ, Hovingh GK, et al. Semaglutide in patients with obesity‐related heart failure and type 2 diabetes. N Engl J Med 2024;390:1394‐1407. doi: 10.1056/NEJMoa2313917 [DOI] [PubMed] [Google Scholar]
- 10. Lincoff AM, Brown‐Frandsen K, Colhoun HM, Deanfield J, Emerson SS, Esbjerg S, et al. Semaglutide and cardiovascular outcomes in obesity without diabetes. N Engl J Med 2023;389:2221‐2232. doi: 10.1056/NEJMoa2307563 [DOI] [PubMed] [Google Scholar]
- 11. Jaiswal A, Truby LK, Chichra A, Jain R, Myers L, Patel N, et al. Impact of obesity on ventricular assist device outcomes. J Card Fail 2020;26:287‐297. doi: 10.1016/j.cardfail.2019.10.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. daSilva‐deAbreu A, Rodgers JE, Seltz J, Mandras SA, Lavie CJ, Loro‐Ferrer JF, et al. Obesity, challenges, and weight‐loss strategies for patients with ventricular assist devices. JACC Heart Fail 2024;12:1661‐1676. [DOI] [PubMed] [Google Scholar]
- 13. Medina CK, Barnes SG, Felker GM, Mentz RJ, Pagidipati NJ, Seymour KA, et al. Severe obesity among patients with left ventricular assist devices. Am Heart J 2024;274:130‐133. doi: 10.1016/j.ahj.2024.05.006 [DOI] [PubMed] [Google Scholar]
- 14. Yost G, Coyle L, Gallagher C, Graney N, Siemeck R, Tatooles A, et al. The impact of extreme obesity on outcomes after left ventricular assist device implantation. J Thorac Dis 2017;9:4441‐4446. doi: 10.21037/jtd.2017.10.52 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Beck‐da‐Silva L, Higginson L, Fraser M, Williams K, Haddad H. Effect of orlistat in obese patients with heart failure: a pilot study. Congest Heart Fail 2005;11:118‐123. doi: 10.1111/j.1527-5299.2005.03827.x [DOI] [PubMed] [Google Scholar]
- 16. Jorsal A, Kistorp C, Holmager P, Tougaard RS, Nielsen R, Hanselmann A, et al. Effect of liraglutide, a glucagon‐like peptide‐1 analogue, on left ventricular function in stable chronic heart failure patients with and without diabetes (LIVE)‐a multicentre, double‐blind, randomised, placebo‐controlled trial. Eur J Heart Fail 2017;19:69‐77. doi: 10.1002/ejhf.657 [DOI] [PubMed] [Google Scholar]
- 17. Margulies KB, Hernandez AF, Redfield MM, Givertz MM, Oliveira GH, Cole R, et al. Effects of liraglutide on clinical stability among patients with advanced heart failure and reduced ejection fraction: a randomized clinical trial. JAMA 2016;316:500‐508. doi: 10.1001/jama.2016.10260 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Deanfield J, Verma S, Scirica BM, Kahn SE, Emerson SS, Ryan D, et al. Semaglutide and cardiovascular outcomes in patients with obesity and prevalent heart failure: a prespecified analysis of the SELECT trial. Lancet 2024;404:773‐786. doi: 10.1016/S0140-6736(24)01498-3 [DOI] [PubMed] [Google Scholar]
- 19. Clerkin KJ, Naka Y, Mancini DM, Colombo PC, Topkara VK. The impact of obesity on patients bridged to transplantation with continuous‐flow left ventricular assist devices. JACC Heart Fail 2016;4:761‐768. doi: 10.1016/j.jchf.2016.05.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1. Outcome Definitions.