What this is
- Liver transplant recipients often face complications like weight gain, diabetes, and hypertension post-surgery.
- This study examines the effects of glucagon-like peptide-1 receptor agonists (GLP1RAs) on health outcomes in these patients.
- Using data from the TriNetX Research Network, outcomes were compared between those using GLP1RAs and nonusers.
- Findings indicate that early GLP1RA use is linked to reduced mortality and hospitalizations without harming graft safety.
Essence
- GLP1RA therapy initiated within one month post-liver transplant is associated with lower all-cause mortality and fewer hospitalizations, as well as reduced risks of acute heart failure and renal failure.
Key takeaways
- GLP1RA use led to a 43% lower all-cause mortality (7.0% vs. 12.9%) compared to nonusers, indicating a significant survival benefit.
- Patients on GLP1RAs experienced 39% fewer hospitalizations (60.4% vs. 74.5%), suggesting improved overall health status.
- The risk of acute heart failure was reduced by 61% (10.9% vs. 26.2%) in the GLP1RA group, highlighting potential cardiovascular benefits.
Caveats
- The study's retrospective design limits causal inferences and may introduce residual confounding despite propensity score matching.
- Data limitations from the TriNetX database may affect the accuracy of patient outcomes and comorbidity assessments.
- Findings may not be generalizable to all liver transplant recipients due to the specific characteristics of the studied cohort.
Simplified
INTRODUCTION
Liver transplantation remains the definitive and often life-saving treatment for patients with end-stage liver disease; however, long-term survival in this population is complicated by increased incidence of nonhepatic complications.1 After transplantation, recipients frequently face an elevated risk of cardiometabolic disease, driven by the combined effects of weight gain, insulin resistance, and hypertension.2 In addition, the use of immunosuppressive therapy such as calcineurin inhibitors, although essential for maintaining graft integrity, exert nephrotoxic effects as well as direct adverse effects on glucose metabolism and lipid regulation, thereby accelerating the development of adverse cardiovascular outcomes in an already high-risk cardiovascular population.3–5 Furthermore, obesity and diabetes often develop or significantly worsen after transplantation, creating a self-reinforcing cycle of metabolic dysfunction that threatens both graft longevity and overall patient survival.2
Glucagon-like peptide-1 receptor agonists (GLP1RAs), initially developed for glycemic control, have emerged as effective therapies for metabolic and cardiovascular disease.2 Beyond stimulating insulin secretion and suppressing glucagon, these agents promote weight loss through delayed gastric emptying and appetite suppression, enhance myocardial glucose utilization, reduce oxidative stress, and mitigate maladaptive cardiac remodeling.6 They also improve endothelial function by stimulating nitric oxide release and modulating blood pressure through natriuresis and vasodilation.7
Despite these advances, the role of GLP1RAs in liver transplant recipients remains poorly studied. While newer studies suggest that GLP1RAs are well tolerated in transplant recipients and result in weight loss and better glycemic control, cardiovascular and renal outcomes have not been systematically evaluated in this population.8
In this study, we evaluated the effect of GLP1RA therapy on all-cause mortality, all-cause hospitalizations, cardiovascular events, renal dysfunction, and graft function in patients after liver transplantation.
MATERIALS AND METHODS
Study Design and Population
This retrospective cohort study utilized de-identified electronic health records (EHRs) from the TriNetX Research Network, a global database spanning >100 healthcare organizations. Adults aged ≥18 years who underwent liver transplantation between January 1, 2010, and December 31, 2023, were included. Patients were required to be on guideline-directed dual immunosuppressive therapy, defined as concurrent use of a calcineurin inhibitor (tacrolimus or cyclosporine) and an antiproliferative agent (mycophenolate, azathioprine, or sirolimus).9
The exposed cohort comprised patients who initiated GLP1RA therapy (semaglutide, dulaglutide, or liraglutide) within 1 mo posttransplant to minimize immortal time bias. The unexposed cohort included patients without any recorded GLP1RA use during the study period. Individuals with prior GLP1RA use before transplantation were excluded. A look-back period was also completed to ensure all patients had appropriate documentation of baseline characteristics, and those with no documented healthcare visits were excluded.
Baseline Characteristics and Risk Factors
Demographic variables, comorbidities, laboratory values, and medication use were extracted from the EHRs. Demographic variables included age, sex, race/ethnicity (White, Black or African American, Hispanic or Latino), and socioeconomic status. Comorbidities encompassed diabetes mellitus, hypertensive diseases, disorders of lipoprotein metabolism, obesity, chronic kidney disease (CKD), ischemic heart diseases, nicotine dependence, sleep apnea, heart failure (HF), alcohol-related disorders, atrial fibrillation/flutter, and prior cerebral infarction.
Pharmacotherapy use included insulin, metformin, glipizides, antiarrhythmics, anticoagulants, beta-blockers, loop diuretics, platelet aggregation inhibitors, antilipemic agents, calcium channel blockers, spironolactone, thiazide diuretics, angiotensin-converting enzyme inhibitors, angiotensin II receptor blockers, angiotensin receptor-neprilysin inhibitors, and sodium-glucose cotransporter-2 inhibitors.
Laboratory and clinical variables comprised serum creatinine, body mass index (BMI), low-density lipoprotein (LDL) cholesterol, hemoglobin A1c (HbA1c), N-terminal pro-B-type natriuretic peptide, and left ventricular ejection fraction. We also adjusted for Z-coding, which is an International Classification of Diseases, Tenth Revision (ICD-10) coded system (Z55-65) to account for social hazards to health.
Clinical Outcomes
The event index date was 1 mo after liver transplantation. The primary study outcomes included all-cause mortality and all-cause hospitalizations, as well as secondary clinical outcomes including acute HF events (ICD-10 I50.21, I50.23, I50.31, I50.33, I50.41, or I50.43), stroke (ICD-10 I63), acute myocardial infarction (MI) (I21.0-I21.4, I21.9), cardiac arrest (ICD-10 I46), new atrial fibrillation/flutter (AF) (ICD-10 I48), ventricular tachycardia (ICD-10 I47.2), acute renal failure or the need for dialysis (ICD-10 N17, N18.6, CPT codes: 90945, 1012752, 90947, 90940), liver transplant failure (ICD-10: T86.42), liver transplant rejection (ICD-10: T86.41), and respiratory failure (ICD-10: J96). We also included ischemic optic neuropathy (ICD-10: H46) as an outcome given recent observational data suggesting nonarteritic anterior optic neuropathy as an adverse event in patients on GLP1RA therapy.10 We also included follow-up LDL, HbA1C, and BMI. Follow-up was censored at the occurrence of an event, death, or the last recorded encounter within the study period. Groups were compared according to GLP1RA use.
Statistical Analysis
Nominal variables are presented as frequencies and percentages. Continuous variables are reported as mean ± SD for normally distributed data or median (interquartile range) for nonnormally distributed data. Baseline characteristics were compared between the GLP1RA and non-GLP1RA cohorts using chi-square tests for categorical variables and independent-sample t-tests or Mann-Whitney U tests for continuous variables, as appropriate.
A 1:1 propensity score matching (PSM) was performed using a greedy nearest-neighbor algorithm with a caliper width of 0.1. The PSM model included baseline demographics, comorbidities, and baseline medications. Time-to-event outcomes were analyzed using Kaplan-Meier survival curves with log-rank testing. Cox proportional hazards regression models were used to calculate hazard ratios (HRs) and 95% confidence intervals (CIs). A two-tailed P value <0.05 was deemed statistically significant. All analyses were conducted using R software (v4.3.2, R Foundation for Statistical Computing, Vienna, Austria) within the TriNetX platform. The Mayo Clinic Institutional Review Board deemed this study exempt from approval given the nature of the publicly available and anonymized data (Institutional Review Board #25-008530). Accordingly, individual participant consent was not required.
RESULTS
Before PSM, 546 liver transplant recipients prescribed GLP1RAs within 1 mo posttransplant were compared with 37 153 non-GLP1RA users. The GLP1RA cohort was older (mean age 62.9 versus 60.1 y, P < 0.001) and had a higher prevalence of diabetes mellitus (92.1% versus 37.2%, P < 0.001), hypertensive diseases (89.9% versus 57.8%, P < 0.001), obesity (70.0% versus 23.4%, P < 0.001), and CKD (62.5% versus 39.5%, P < 0.001). The GLP1RA group also included higher baseline use of insulin (93.2% versus 59.1%, P < 0.001), metformin (45.8% versus 7.6%, P < 0.001), and cardiovascular therapies such as beta-blockers (84.8% versus 61.0%, P < 0.001) and anticoagulants (85.2% versus 64.9%, P < 0.001). However, over the follow-up period, only 18% and 22% of the GLP1RA and non-GLP1RA cohorts were on anticoagulation, respectively. Laboratory and clinical parameters revealed higher HbA1c (7.0% versus 5.7%, P < 0.001), BMI (33.2 versus 28.2 kg/m2, P < 0.001), and lower serum creatinine (1.4 versus 1.6 mg/dL, P = 0.010) in the GLP1RA cohort. After 1:1 PSM, 541 patients were included in each group, with well-balanced baseline characteristics. Baseline characteristics of the study cohort are reported in Table 1.
At follow-up, LDL levels were lower in the GLP-1RA cohort compared with the non-GLP-1RA cohort (75.62 ± 35.59 mg/dL versus 83.09 ± 37.75 mg/dL; t = –2.83; P < 0.004). BMI was also lower among patients receiving GLP-1RA therapy (28.12 ± 6.30 kg/m2 versus 30.80 ± 6.29 kg/m2; t = 3.144; P = 0.002). HbA1c levels were similar between groups, with no statistically significant difference (6.85% ± 1.63% versus 6.84% ± 1.65%; t = 0.104; P = 0.918).
Over a mean follow-up of 838.5 d (SD 291.9) in the GLP1RA cohort and 884.3 d (SD 313.7) in the non-GLP1RA cohort, patients treated with GLP1RAs demonstrated significantly lower risks of all-cause mortality and hospitalization (Table 2). The endpoint of all-cause mortality occurred in 38 (7.0%) patients in the GLP1RA group compared with 70 (12.9%) patients in the non-GLP1RA group, corresponding to a 43% risk reduction (HR, 0.566; 95% CI, 0.381-0.841). Similarly, all-cause hospitalizations occurred in 327 patients in the GLP1RA group compared with 403 patients in the non-GLP1RA group (60.4% versus 74.5%; HR, 0.613; 95% CI, 0.530-0.710) (Figure 1).
GLP1RA use was associated with a 61% lower risk of acute HF (59 [10.9%] versus 142 [26.2%]; HR, 0.386; 95% CI, 0.285-0.524) and a 51% reduction in acute renal failure or dialysis requirement (223 [41.2%] versus 346 [64.0%]; HR, 0.489; 95% CI, 0.413-0.579) (Figure 2). Respiratory failure events were also significantly lower in the GLP1RA group (59 [10.9%] versus 116 [21.4%]; HR, 0.484; 95% CI, 0.354-0.662).
No significant differences were observed in acute MI (23 [4.3%] versus 35 [6.5%]; HR, 0.671; 95% CI, 0.397-1.137), stroke (26 [4.8%] versus 39 [7.2%]; HR, 0.675; 95% CI, 0.411-1.110), cardiac arrest (<10 [1.8%] versus 13 [2.4%]; HR, 0.626; 95% CI, 0.259-1.511), or ventricular tachycardia (<10 [1.8%] versus 11 [2.0%]; HR, 0.942; 95% CI, 0.400-2.220). Liver transplant failure (24 [4.4%] versus 37 [6.8%]; HR, 0.651; 95% CI, 0.389-1.089) and rejection rates (80 [14.8%] versus 77 [14.2%]; HR, 1.055; 95% CI, 0.772-1.443) also did not differ between groups (Figure 3). Similarly, the risk of ischemic optic neuropathy was insignificant between the 2 cohorts (HR, 0.986; 95% CI, 0.062-15.760).
Kaplan-Meier curve. Event-free survival analyses for all-cause mortality and hospitalization outcomes.
Kaplan-Meier curve. Event-free survival analyses for acute heart failure events and acute renal failure/dialysis outcomes.
Kaplan-Meier curve. Event-free survival analyses for liver transplant failure and rejection outcomes.
| Pre-PSM | Post-PSM | |||||
|---|---|---|---|---|---|---|
| On GLP1RAs, n (%) | Not on GLP1RAs, n (%) | P | On GLP1RAs, n (%) | Not on GLP1RAs, n (%) | P | |
| Demographics | ||||||
| Current age, y (mean ± SD) | 62.9 ± 11.2 | 60.1 ± 16.3 | <0.001 | 62.9 ± 11.3 | 63.0 ± 12.0 | 0.99 |
| Female | 230 (42.1) | 13 793 (37.1) | 0.02 | 228 (42.1) | 238 (44.0) | 0.54 |
| White | 385 (70.5) | 25 299 (68.1) | 0.23 | 380 (70.2) | 382 (70.6) | 0.89 |
| Hispanic or Latino | 116 (21.2) | 4407 (11.9) | <0.001 | 115 (21.3) | 98 (18.1) | 0.19 |
| Black or African American | 57 (10.4) | 4353 (11.7) | 0.36 | 57 (10.5) | 62 (11.5) | 0.63 |
| Comorbidities | ||||||
| Diabetes mellitus | 503 (92.1) | 13 814 (37.2) | <0.001 | 498 (92.0) | 495 (91.5) | 0.74 |
| Hypertensive diseases | 491 (89.9) | 21 473 (57.8) | <0.001 | 486 (89.8) | 476 (88.0) | 0.33 |
| Disorders of lipoprotein metabolism | 406 (74.4) | 11 286 (30.4) | <0.001 | 401 (74.1) | 398 (73.6) | 0.84 |
| Overweight and obesity | 382 (70.0) | 8711 (23.4) | <0.001 | 377 (69.7) | 391 (72.3) | 0.35 |
| Chronic kidney disease | 341 (62.5) | 14 658 (39.5) | <0.001 | 338 (62.5) | 333 (61.5) | 0.75 |
| Ischemic heart diseases | 269 (49.3) | 9074 (24.4) | <0.001 | 267 (49.4) | 272 (50.3) | 0.76 |
| Nicotine dependence | 192 (35.2) | 7790 (21.0) | <0.001 | 191 (35.3) | 196 (36.2) | 0.75 |
| Sleep apnea | 183 (33.5) | 4040 (10.9) | <0.001 | 179 (33.1) | 187 (34.6) | 0.61 |
| Heart failure | 142 (26.0) | 4575 (12.3) | <0.001 | 139 (25.7) | 152 (28.1) | 0.37 |
| Alcohol-related disorders | 132 (24.2) | 8046 (21.7) | 0.16 | 130 (24.0) | 140 (25.9) | 0.48 |
| Atrial fibrillation and flutter | 85 (15.6) | 3785 (10.2) | <0.001 | 85 (15.7) | 87 (16.1) | 0.87 |
| Cerebral infarction | 47 (8.6) | 1084 (2.9) | <0.001 | 45 (8.3) | 46 (8.5) | 0.91 |
| Persons with potential health hazards related to socioeconomic circumstances (Z-Codes) | 45 (8.2) | 1282 (3.5) | <0.001 | 44 (8.1) | 44 (8.1) | 1 |
| Pharmacotherapies | ||||||
| Insulin | 509 (93.2) | 21 937 (59.1) | <0.001 | 504 (93.2) | 511 (94.5) | 0.38 |
| Metformin | 250 (45.8) | 2824 (7.6) | <0.001 | 245 (45.3) | 177 (32.7) | <0.001 |
| Glipizides | 92 (16.9) | 1400 (3.8) | <0.001 | 89 (16.4) | 91 (16.8) | 0.87 |
| Antiarrhythmics | 473 (86.6) | 24 706 (66.5) | <0.001 | 468 (86.5) | 470 (86.9) | 0.86 |
| Anticoagulants | 465 (85.2) | 24 115 (64.9) | <0.001 | 460 (85.0) | 466 (86.1) | 0.6 |
| Beta-blockers | 463 (84.8) | 22 664 (61.0) | <0.001 | 459 (84.8) | 464 (85.8) | 0.67 |
| Loop diuretics | 443 (81.1) | 23 952 (64.5) | <0.001 | 438 (81.0) | 442 (81.7) | 0.75 |
| Platelet aggregation inhibitors | 401 (73.4) | 13 535 (36.4) | <0.001 | 396 (73.2) | 392 (72.5) | 0.78 |
| Antilipemic agents | 391 (71.6) | 10 526 (28.3) | <0.001 | 386 (71.3) | 391 (72.3) | 0.74 |
| Calcium channel blockers | 374 (68.5) | 13 437 (36.2) | <0.001 | 369 (68.2) | 350 (64.7) | 0.22 |
| Spironolactone | 280 (51.3) | 14 301 (38.5) | <0.001 | 275 (50.8) | 291 (53.8) | 0.33 |
| Thiazides | 181 (33.1) | 6114 (16.5) | <0.001 | 179 (33.1) | 183 (33.8) | 0.8 |
| ACE inhibitors | 205 (37.5) | 5856 (15.8) | <0.001 | 200 (37.0) | 198 (36.6) | 0.9 |
| Angiotensin II blockers | 165 (30.2) | 3929 (10.6) | <0.001 | 163 (30.1) | 176 (32.5) | 0.39 |
| ARNI | 10 (1.8) | 101 (0.3) | <0.001 | 10 (1.9) | 10 (1.9%) | 1 |
| Sodium-glucose cotransporter 2 inhibitors | 120 (22.0) | 572 (1.5) | <0.001 | 115 (21.3) | 109 (20.1) | 0.65 |
| Labs and clinical variables | ||||||
| Creatinine in serum (mean ± SD) | 1.4 ± 1.0 | 1.6 ± 1.6 | 0.01 | 1.4 ± 1.0 | 1.6 ± 1.3 | <0.001 |
| BMI (mean ± SD) | 33.2 ± 6.2 | 28.2 ± 6.4 | <0.001 | 33.2 ± 6.2 | 31.5 ± 6.3 | <0.001 |
| Cholesterol in LDL (mean ± SD) | 84.4 ± 40.7 | 83.6 ± 56.0 | 0.76 | 84.2 ± 40.3 | 78.9 ± 42.6 | 0.06 |
| Hemoglobin A1c (mean ± SD) | 7.0 ± 1.8 | 5.7 ± 1.5 | <0.001 | 7.0 ± 1.8 | 6.5 ± 1.6 | <0.001 |
| Natriuretic peptide.B prohormone N-terminal (mean ± SD) | 1529.9 ± 3205.1 | 3607.6 ± 8713.1 | 0.04 | 1412.3 ± 3058.2 | 2989.5 ± 5732.6 | 0.04 |
| Left ventricular ejection fraction (%) (mean ± SD) | 60.0 ± 10.3 | 61.8 ± 9.2 | 0.21 | 60.0 ± 10.3 | 62.0 ± 8.0 | 0.3 |
| Outcome | With GLP1RAs events, n (%) | Without GLP1RAs events, n (%) | Hazard ratio (95% CI) | Event-free survival probabilities (GLP1RA vs without) |
|---|---|---|---|---|
| All-cause mortality | 38 (7.0) | 70 (12.9) | 0.566 (0.381, 0.841) | 90.8% vs 85.1% |
| All-cause hospitalizations | 327 (60.4) | 403 (74.5) | 0.613 (0.530, 0.710) | 33.1% vs 20.3% |
| Acute heart failure events | 59 (10.9) | 142 (26.2) | 0.386 (0.285, 0.524) | 87.2% vs 71.0% |
| Acute myocardial infarction | 23 (4.3) | 35 (6.5) | 0.671 (0.397, 1.137) | 94.9% vs 92.4% |
| Stroke | 26 (4.8) | 39 (7.2) | 0.675 (0.411, 1.110) | 94.2% vs 91.9% |
| Cardiac arrest | <10 (1.8) | 13 (2.4) | 0.626 (0.259, 1.511) | 98.1% vs 97.4% |
| Ventricular tachycardia | <10 (1.8) | 11 (2.0) | 0.942 (0.400-2.220) | 97.8% vs 97.6% |
| New atrial fibrillation/flutter | 83 (15.3) | 92 (17.0) | 0.907 (0.674, 1.220) | 83.0% vs 81.4% |
| Acute renal failure or need for dialysis | 223 (41.2) | 346 (64.0) | 0.489 (0.413, 0.579) | 54.9% vs 31.4% |
| Liver transplant failure | 24 (4.4) | 37 (6.8) | 0.651 (0.389, 1.089) | 94.5% vs 92.3% |
| Liver transplant rejection | 80 (14.8) | 77 (14.2) | 1.055 (0.772, 1.443) | 83.0% vs 84.1% |
| Ischemic optic neuropathy | <10 (1.8) | <10 (1.8) | 0.986 (0.062, 15.760) | 99.8% vs 99.8% |
| Respiratory failure | 59 (10.9) | 116 (21.4) | 0.484 (0.354, 0.662) | 87.1% vs 75.9% |
| Follow-up labs | ||||
| Outcomes | GLP1RA cohort | Non-GLP1RA cohort | T statistics | P |
| LDL | 75.62 (SD 35.59) | 83.09 (SD 37.75) | −2.83 | <0.004 |
| BMI | 28.12 (SD 6.30) | 30.80 (SD 6.29) | 3.144 | 0.002 |
| HbA1C | 6.85 (SD 1.63) | 6.84 (SD 1.65) | 0.104 | 0.918 |
DISCUSSION
In our cohort of liver transplant recipients, the initiation of GLP1RA therapy was associated with significantly improved outcomes. Patients receiving GLP1RA had lower all-cause mortality and fewer all-cause hospitalizations compared with nonusers. Moreover, the GLP1RA group experienced fewer acute HF events and acute renal failure/dialysis events. These findings align with emerging data in other transplant populations.11,12
Our study showed similar mortality benefits seen in other GLP1RA studies involving transplant patients. For example, Dotan et al reported a significant mortality reduction with GLP1RA in a mixed transplant cohort,11 and Lin et al similarly found a significant mortality benefit in kidney transplant recipients on GLP1RA.12 The reduced hospitalization rate in our study likely stems from fewer acute events (HF, renal failure) and perhaps better overall metabolic health status. GLP1RA-treated patients in our study also had significantly lower risk of acute HF events. A recent meta-analysis of HF with preserved ejection fraction trials found that GLP1RAs significantly reduced the composite of cardiovascular death or HF hospitalization and lowered worsening HF events.13 This may be explained by GLP1RAs’ ability to improve multiple cardiovascular risk factors that contribute to HF since they promote significant weight loss, improved blood pressure and lipid profiles.7 GLP1RAs also improve endothelial function and reduce oxidative stress and inflammation, contributing to their antiatherogenic effects.7 In the myocardium, GLP1RAs can enhance myocardial glucose utilization and reduce maladaptive remodeling.6
Moreover, the significant decrease in risk of acute renal failure or need for dialysis is consistent with known renoprotective effects of GLP1RA. A recent meta-analysis showed that GLP1RAs reduce the risk of a composite kidney outcome consisting of kidney failure (kidney replacement therapy or a persistent estimated glomerular filtration rate (eGFR) <15 mL/min per 1·73 m2, a sustained reduction in eGFR by at least 50% or the nearest equivalent, or death from kidney failure,14 likely through improving glycemic control, reducing hyperfiltration, and attenuating inflammation.15 In addition, GLP1RAs reduce albuminuria and slow decline in eGFR in diabetic kidney disease.15 In the context of calcineurin inhibitor nephrotoxicity, these benefits may be especially valuable.
The lack of differences in graft failure or rejection may suggest that adding GLP1RA does not adversely affect the immune function or graft viability. Similarly, studies that evaluated GLP1RA use in patients with known liver transplantation tolerated immunosuppresive therapies well without necessitating changes in immunosuppressive therapy.16,17 For instance, Grancini et al found no need to adjust calcineurin inhibitor dosing when adding GLP1RA, and adverse events were minimal.16 Similarly, heart transplant recipients on GLP1RA rarely required any immunosuppression dose changes.17 There is also no evidence of direct immunostimulation by GLP-1Ras. In fact, no immune-related adverse events or rejection signals have emerged in the available posttransplant clinical data.18 Preclinical findings even suggest GLP-1R signaling may have a tolerogenic effect on T cells, helping to mitigate alloimmune responses.19 Beyond their glycemic effects, GLP-1RAs may indirectly improve graft health by alleviating metabolic stress, fatty infiltration, and inflammation.16 Importantly, liver stiffness (a marker of fibrosis) decreased with GLP1-RA therapy, suggesting less graft steatosis/fibrosis with GLP-1RA treatment.16 Thus, GLP1RAs appear safe in transplant patients, without evidence of immunologic interactions.
Our study provides evidence that GLP1RA therapy may offer multifactorial benefits in liver transplant recipients, potentially improving survival and promoting cardiorenal stability without compromising graft safety. By mimicking endogenous GLP-1, these drugs increase glucose-dependent insulin secretion and suppress glucagon, improving glycemic control, among multiple other beneficial effects.6 This positions GLP1RAs as a promising adjunctive therapy in post-liver transplant patients. However, given the observational nature of our study, these findings require cautious interpretation and warrant further investigation through prospective, controlled trials.
Several limitations in our study exist. Our study is retrospective and observational which precludes causal inference. Furthermore, residual confounding cannot be excluded despite propensity matching. Moreover, the TriNetX database lacks granular patient data, and there is a potential of misclassification or under-coding of data with ICD diagnosis coding. To that point, we are unable to identify specific causes of death or the exact indications for liver transplantation and causes of the end-stage liver disease, although the majority of our patients had high burden of metabolic comorbidities. Acuity of the perioperative state during liver transplantation, blood type, and underlying severity of systemic illness are not available. In addition, our cohort comprised liver transplant recipients with a substantial burden of cardiometabolic comorbidities, including diabetes, obesity, hypertension, and CKD, which may lead to selection bias. As such, the findings may reflect outcomes in metabolically high-risk recipients and may not be generalizable to the broader liver transplant population. We also lacked data on medication adherence and do not have data on longitudinal medication prescriptions, as it cannot be verified in EHR databases through the TriNetX network. Similarly, we do not have the initial indication for GLP1RA use; although the majority of the GLP1RA cohort had a diabetic range mean HbA1C and a BMI >30 kg/m2. Finally, follow-up time was limited to approximately three years given skewed follow-up data among the 2 cohorts if extended any further.
CONCLUSIONS
Our findings suggest that GLP1RA therapy after liver transplantation is associated with better long-term outcomes, particularly in reducing mortality, HF, hospitalizations, renal failure, and respiratory failure events. Prospective studies and randomized trials are needed to confirm these benefits and to assess long-term cardiovascular and renal outcomes in transplant recipients.
Footnotes
Contributor Information
Hesham Sheashaa, Email: sheashaa.hesham@mayo.edu;sheashaahesham@outlook.com.
Amani Elshaer, Email: Elshaer.Amany@mayo.edu.
Hoang Nhat Pham, Email: npham917@arizona.edu.
Rama Mouhaffel, Email: rmouhaffel@arizona.edu.
Eiad Habib, Email: habib.eiad@mayo.edu.
Mahmoud Abdelnabi, Email: abdelnabi.mahmoud@mayo.edu.
Juan M. Farina, Email: Farina.JuanMaria@mayo.edu.
Steven J. Lester, Email: lester.steven@mayo.edu.
David Simper, Email: simper.david@mayo.edu.
Said Alsidawi, Email: Alsidawi.Said@mayo.edu.
Eric D. Steidley, Email: Steidley.D@mayo.edu.
Bashar A. Aqel, Email: Aqel.Bashar@mayo.edu.
Michele Barnhill, Email: Barnhill.Michele@mayo.edu.
W. Ray Kim, Email: ray.thekims@gmail.com.
Chadi Ayoub, Email: ayoub.chadi@mayo.edu.
Reza Arsanjani, Email: arsanjani.reza@mayo.edu.
GLP-1 Therapies