Highlights
-
•
Medical therapy after TAVR is largely understudied.
-
•
GLP-1 RA monotherapy after TAVR showed decreased mortality.
-
•
SGLT2i monotherapy after TAVR showed decreased rates of arrhythmias.
-
•
Combination therapy showed decreased mortality, acute HF, MI, and arrhythmia.
-
•
GLP-1 RA and SGLT2i therapy may be underutilized in post-TAVR patients.
Aortic valve disease is the most common valvular disease and is often seen in the elderly population. Transcatheter aortic valve replacement (TAVR) is a favorable and rapidly evolving intervention. Despite highly effective procedural outcomes, the TAVR population remains at high risk for heart failure and death, and medical therapy after TAVR is understudied. Diabetic medications including GLP-1 receptor agonists (GLP-1 RA) and SGLT2 inhibitors (SGLT2i) have had emerging data suggesting cardioprotective effects. In this study, we aim to evaluate cardiovascular outcomes in post-TAVR patients who are treated with GLP-1 RA, SGLT2i, or both compared to those who are not treated. Using TriNetX, we identified a cohort of patients who underwent a TAVR procedure and then classified them as treated with GLP-1 RA, SGLT2i, both, or none. After propensity matching for demographics, comorbidities, and medications, patient outcomes for all-cause mortality (ACM) and cardiovascular disorders were evaluated using a Kaplan–Meier analysis. Our results showed a decrease in ACM in those who were on GLP-1 RA after TAVR compared to those who were not. There was also a statistically significant decrease in arrhythmia in patients on SGLT2i after TAVR compared to those who were not. When GLP-1 RA and SGLT2i were combined, there was a decrease in ACM, myocardial infarction, acute heart failure, and arrhythmia after TAVR compared to those who were not. In conclusion, these findings further suggest cardioprotective effects of these drugs in patients treated with TAVR. Future trials should further investigate the role of these medications in patients with aortic valve stenosis.
Calcific aortic stenosis is one of the most common valvular diseases in the western world and high-income countries with large disease burden in the growing elderly population. , In recent years, TAVR has emerged as the most common procedural treatment for severe aortic valve stenosis surpassing surgical aortic valve replacement in frequency of use. While TAVR has emerged as a less invasive and comparable alternative to surgical aortic valve replacement, its implementation into clinical medicine has outpaced the development of comprehensive post-TAVR medical management strategies. , Patients receiving TAVR are at higher risk of adverse events, such as heart failure, cerebrovascular events, and conduction disorders, yet optimal pharmacological regimens remain largely understudied in the post-TAVR population. Concurrently, use of sodium-glucose cotransporter-2 inhibitors (SGLT2i) and glucagon-like peptide-1 receptor agonists (GLP-1 RA) has been increasing in recent years due to favorable outcomes in patients treated with these medications. In the LEADER (Liraglutide Effect and Action in Diabetes: Evaluation of Cardiovascular Outcome Results) trial, liraglutide showed a lower rate of death from cardiovascular causes when compared to the placebo group. Moreover, Lincoff et al showed that patients on semaglutide with preexisting cardiovascular disease and obesity, but without diabetes mellitus, had a lower risk of death from cardiovascular causes in the SELECT (Semaglutide Effects on Cardiovascular Outcomes in People with Overweight or Obesity) trial. However, there are limited data on the frequency of use and outcomes associated with GLP-1 RA or SGLT2i in the post-TAVR population. This gap in literature and lack of robust post-TAVR medical management strategies provides a rationale for evaluating whether these therapies may contribute additional cardiovascular protection in the post-TAVR population. In this study, we aim to evaluate cardiovascular outcomes among post-TAVR patients who are on SGLT2 inhibitors, GLP-1 receptor agonists, or combined therapy compared to post-TAVR patients not receiving any therapy.
Methods
Ethical approval
This retrospective study is exempt from informed consent. The data reviewed are a secondary analysis of existing data, do not involve intervention or interaction with human subjects, and are de-identified in accordance with the de-identification standard defined in Section §164.514(a) of the Health Insurance Portability and Accountability Act Privacy Rule. The process by which the data are de-identified is attested to through a formal determination by a qualified expert as defined in Section §164.514(b)(1) of the Health Insurance Portability and Accountability Act Privacy Rule. This formal determination by a qualified expert was refreshed in December 2020.
Study design
This retrospective cohort study utilized the TriNetX Research Network in August 2025. This database includes information on medications, procedures, diagnoses, laboratory values, and genomic information on approximately 160 million patients from 106 healthcare organizations (HCO). Codes for diagnosis were based on the International Classification of Diseases (ICD-10-CM), laboratory tests used Logical Observation Identifiers Names and Codes, TriNetX curated codes (TNX), Anatomical Therapeutic Chemical codes, and procedures were identified using Current Procedural Terminology codes. Our study was conducted in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement guidelines for cohort studies, which are provided in the supplementary information .
Cohorts
Our cohorts included patients older than 18 years of age who underwent TAVR. Three exposure groups were constructed. The first group included SGLT2i use within 1 month before the procedure or within 1 year after, with no documented history of GLP-1 RA use. The second group was composed of patients who were prescribed a GLP-1 RA within 1 month before the procedure or up to 1 year after, and no history of SGLT2i use. The third group included the use of both GLP-1 RA and SGLT2i. For each exposure group, a control group was created consisting of patients with no documented use of SGLT2i or GLP-1 RA at any time. For the SGLT2i cohort, 50 HCOs contributed patient data. While in the GLP-1 RA cohort, 46 HCOs contributed patient data. The combined therapy cohort had 48 HCOs that contributed patient data. The cohorts were propensity-matched in a 1:1 ratio with their respective control groups based on patient characteristics such as demographics, preexisting conditions, and medication use.
Outcomes
Patients were excluded from the analysis if they had any previous history of an outcome of interest for each outcome-specific analysis to ensure capture of incident events only. All outcomes were assessed 1 year after the index event. The primary outcome was all-cause mortality (death status in medical records). Secondary outcomes included acute myocardial infarction (ICD-10-CM I21), acute heart failure (ICD-10-CM I50.21, I50.23, I50.31, I50.33, I50.41, I50.43, I50.813), ischemic stroke (ICD-10-CM I63), hemorrhagic stroke (ICD-10-CM I60, I61, I62), acute kidney injury (ICD-10-CM N17), and clinically significant arrhythmias including atrial fibrillation or flutter and ventricular arrhythmias (ICD-10-CM I48, I49, I49.0).
Statistical analysis
Our propensity matching utilized TriNetX’s greedy nearest neighbor matching algorithm to control for any differences between the cohorts. Propensity scores were generated using demographic characteristics, baseline comorbidities, and medication use listed in Table 1 . An acceptable match for covariate balancing was indicated by a standardized mean difference of <0.10.
Table 1
Patient demographics and characteristics after propensity matching in the GLP-1 RA, SGLT2i, and combination therapy cohorts
| Baseline characteristics after matching | GLP-1 RA | Control | SGLT2i | Control | Combined | Control |
|---|---|---|---|---|---|---|
| n = 854 | n = 854 | n = 4,077 | n = 4,077 | n = 584 | n = 584 | |
| Demographics | ||||||
| Age at index | 73.9 ± 7.5 | 73.9 ± 9.7 | 77.0 ± 8.7 | 77.1 ± 11.0 | 73.0 ± 7.7 | 72.6 ± 10.8 |
| White | 720 (84.3%) | 721 (84.4%) | 3,410 (83.6%) | 3,419 (83.9%) | 483 (82.7%) | 480 (82.2%) |
| Black or African American | 53 (6.2%) | 50 (5.9%) | 254 (6.2%) | 254 (6.2%) | 38 (6.5%) | 48 (8.2%) |
| Hispanic or Latino | 35 (4.1%) | 42 (4.9%) | 162 (4.0%) | 174 (4.3%) | 35 (6.0%) | 33 (5.7%) |
| Asian | 13 (1.5%) | 16 (1.9%) | 149 (3.7%) | 154 (3.8%) | 17 (2.9%) | 13 (2.2%) |
| Diagnosis | ||||||
| Heart failure | 585 (68.5%) | 569 (66.6%) | 3,476 (85.3%) | 3,454 (84.7%) | 452 (77.4%) | 456 (78.1%) |
| Chronic kidney disease (CKD) | 356 (41.7%) | 328 (38.4%) | 1,912 (46.9%) | 1,900 (46.6%) | 286 (49.0%) | 295 (50.5%) |
| Atrial fibrillation and flutter | 313 (36.7%) | 293 (34.3%) | 2,094 (51.4%) | 2,093 (51.3%) | 254 (43.5%) | 236 (40.4%) |
| Overweight and obesity | 588 (68.9%) | 591 (69.2%) | 1,726 (42.3%) | 1,708 (41.9%) | 400 (68.5%) | 409 (70.0%) |
| Essential hypertension | 739 (86.5%) | 730 (85.5%) | 3,298 (80.9%) | 3,321 (81.5%) | 535 (91.6%) | 542 (92.8%) |
| Type 1 diabetes mellitus | 112 (13.1%) | 97 (11.4%) | 194 (4.8%) | 196 (4.8%) | 82 (14.0%) | 92 (15.8%) |
| Type 2 diabetes mellitus | 707 (82.8%) | 705 (82.6%) | 2,382 (58.4%) | 2,408 (59.1%) | 548 (93.8%) | 548 (93.8%) |
| Other peripheral vascular disease | 144 (16.9%) | 158 (18.5%) | 861 (21.1%) | 856 (21.0%) | 135 (23.1%) | 137 (23.5%) |
| Other chronic obstructive pulmonary disease | 197 (23.1%) | 200 (23.4%) | 1,059 (26.0%) | 1,034 (25.4%) | 140 (24.0%) | 140 (24.0%) |
| Acute myocardial infarction | 158 (18.5%) | 146 (17.1%) | 1,232 (30.2%) | 1,228 (30.1%) | 148 (25.3%) | 150 (25.7%) |
| Tobacco use | 25 (2.9%) | 19 (2.2%) | 254 (6.2%) | 244 (6.0%) | 29 (5.0%) | 35 (6.0%) |
| Medications | ||||||
| Aspirin | 770 (90.2%) | 769 (90.0%) | 3,659 (89.7%) | 3,655 (89.6%) | 544 (93.2%) | 552 (94.5%) |
| ACE Inhibitors | 456 (53.4%) | 441 (51.6%) | 1,865 (45.7%) | 1,827 (44.8%) | 348 (59.6%) | 343 (58.7%) |
| Angiotensin II receptor blockers (ARBs) | 381 (44.6%) | 380 (44.5%) | 2,139 (52.5%) | 2,109 (51.7%) | 316 (54.1%) | 306 (52.4%) |
| Beta blockers | 686 (80.3%) | 691 (80.9%) | 3,523 (86.4%) | 3,522 (86.4%) | 516 (88.4%) | 514 (88.0%) |
| Clopidogrel | 467 (54.7%) | 491 (57.5%) | 1,981 (48.6%) | 1,953 (47.9%) | 335 (57.4%) | 325 (55.7%) |
| Angiotensin II inhibitor | 381 (44.6%) | 380 (44.5%) | 2,139 (52.5%) | 2,109 (51.7%) | 316 (54.1%) | 306 (52.4%) |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
