The use of bioprosthetic aortic valves (bAVs) has increased due to their low thrombogenicity and the emergence of new generation valves with longer durability.1 However, structural valve degeneration (SVD) remains one of the most concerning complications of bAVs, often leading to valve dysfunction and reintervention. The pathophysiology of SVD is not fully understood; however, it may share similar mechanisms with native aortic valve (AV) degeneration, including atherosclerotic, endothelial dysfunction, and inflammatory-oxidative processes.2 In vitro studies have proposed a role of sodium-glucose cotransporter 2 (SGLT2) pro-oxidant pathways in native AV degeneration and a potential protective effect of SGLT2 inhibitors (SGLT2i) for mitigating valve dysfunction.3 However, the impact of SGLT2i on SVD has not been previously investigated.
What is the clinical question being addressed?
Do SGLT2 inhibitors offer benefits in preventing the degeneration of bioprosthetic aortic valves?
What is the main finding?
SGLT2 inhibitors were associated with decreased risk of bioprosthetic aortic valve degeneration.
Methods
This retrospective cohort study evaluated all adult patients who underwent aortic valve replacement (AVR)—either transcatheter AVR or surgical AVR—using bAVs between 2010 and 2020 at 3 Mayo Clinic Centers in the United States. The study was approved by the Mayo Clinic Institutional Review Board. To be included, patients required at least one baseline transthoracic echocardiogram (TTE) within the first year after the procedure, with the closest one to the procedure being selected when patients had multiple available TTEs. Additionally, each patient required to have one follow-up TTE at least 3 months after the baseline with the most recent TTE being selected for the analysis. This interval was determined according to the updated guidelines of the American Society of Echocardiography.4
The diagnosis of SVD was determined according to the same guidelines, as follows: new occurrence or increase of at least one grade of intraprosthetic aortic regurgitation resulting in moderate or greater aortic regurgitation compared to baseline TTE; or/and increase in AV mean gradient ≥10 mm Hg resulting in a mean gradient ≥20 mm Hg with concomitant decrease in effective aortic valve area (EOA) ≥0.3 cm2 or ≥25% and/or decrease in Doppler velocity index ≥0.1 or ≥20% compared to baseline TTE. Patient prosthesis mismatch was defined as EOA index (EOA/body surface area) < 0.65.5
Exposure to SGLT2i was defined as patients on chronic treatment prior to AVR who continued the treatment post-AVR, or those who started the medication after AVR. All data were extracted from electronic medical records. To control for potential confounders between patients who were on SGLT2i and patients who were not, 4:1 propensity score matching was performed using optimal pair matching algorithm and adjusting for the following variables: age, sex, smoking status, baseline body surface area, patient prosthesis mismatch, comorbidities (type 2 diabetes mellitus, hypertension, atherosclerotic cardiovascular disease, congestive heart failure), medications history at the time of the intervention (beta-blockers, statins, clopidogrel, warfarin, direct oral anticoagulants, calcium-channel blockers, angiotensin-converting enzyme inhibitors/angiotensin receptors blockers), and laboratory measurements at the time of the intervention (low-density lipoprotein cholesterol, creatinine).
Adequacy of matching was verified by testing the standardized differences between the groups. Cumulative incidence of SVD was compared between the groups using Kaplan-Meier curves and the log-rank test. Additionally, univariate Cox regression analysis was performed to evaluate the association between SGLT2i exposure and SVD. Time zero in the Cox models was set at the time of AVR and patients were censored at the time of the last follow-up TTE up to 10 years. Statistical analyses were conducted using R (v 4.3, R Foundation for Statistical Computing).
Results
A total of 1838 patients were identified: among them, 83 (4.16%) were on SGLT2i treatment. After propensity matching, 332 controls who were not on SGLT2i treatment were identified and included for comparison. Adequacy of matching was confirmed with standardized differences between groups being <0.10 for all variables used for matching. The median age of the overall matched cohort was 74 (68,79) years with 71.00% of them male, 52.00% had type 2 diabetes mellitus, 54.00% were smokers, and 71.00% had congestive heart failure.
Over a median follow-up time of 4.85 (IQR: 2.79-6.70) years, the overall 10-year cumulative incidence of SVD was 44.70%. The overall 10-year cumulative incidence of SVD was significantly lower among patients who were taking SGLT2i compared to patients who were not according to Kaplan-Meier analysis (29.00% vs 54.10%; log rank P = 0.006) (Figure 1) and Cox regression analysis (HR: 0.37; 95% CI: 0.18-0.78; P = 0.008).
Figure 1.
Kaplan-Meier Survival Curve Illustrates the Comparison of Survival Free From Structural Valve Degeneration Between Patients Who Were Receiving SGLT2i and Patients Who Were Not
SGLT2i = sodium-glucose cotransporter 2 inhibitors.
Discussion
In our analysis, we demonstrated a potential benefit of SGLT2i in preventing SVD, with 63% less risk of SVD in patients who were on SGLT2i compared to patients who were not. To our knowledge, this is the first study examining the association between SGLT2i and SVD.
The mechanism behind these potential benefits of SGLT2 inhibitors is not yet fully understood. Hmadeh et al3 revealed that degenerative AVs extracted from patients post-AVR showed increased levels of oxidative stress and immunofluorescent staining of SGLT2 in the calcified areas of the valve. Interestingly, the level of oxidative stress was significantly reduced by empagliflozin. Furthermore, it was shown that angiotensin II and extracellular vesicles, which are possible inducers of endothelial damage, may stimulate the expression of SGLT2 in valvular endothelial cells. SGLT2 activation may subsequently induce endothelial valvular dysfunction and inflammation.3 Given the comparable mechanism of SVD to native AV degeneration,1 similar effects of SGLT2 on SVD progression may be involved in the pathophysiology and could be attenuated by SGLT2i.
This study has several limitations. The detection of SVD was achieved solely through TTE. Employing a comprehensive multimodality imaging approach may enhance precision of diagnosis and identify earlier stages of SVD. The magnitude of the sample was small and only represented patients from 3 academic centers in the United States, introducing potential selection bias and limiting the generalizability of the findings. The time between baseline and follow-up TTE was variable among patients. Additionally, the retrospective nature of the study makes controlling for all confounders challenging; however, propensity matching was used to address potential confounders. The use of SGLT2i was extracted from electronic medical records according to patients’ medication history and the specific date of initiation was not available, limiting the possibility of treating the exposure as time-dependent covariate. Therefore, these results should be considered as hypothesis-generating to pave the way for prospective studies to confirm the potential effects of SGLT2i in preventing SVD and increasing the durability of bAV.
Funding support and author disclosures
This publication was supported by the Mayo Clinic Arizona Cardiovascular Clinical Research Center (MCA CV CRC).
Acknowledgments
The authors thank Amro Badr MD, Hana Mousa MD, Milagros Pereyra MD, Isabel G. Scalia MD, Nima Baba Ali MD, Ahmed K. Mahmoud MD, Niloofar Javadi MD, Nadera Naquib Bismee MD, Sogol Attaripour Esfahani MD, Hesham Sheashaa MD, Omar H. Ibrahim MD, and Fatmaelzahraa E. Abdelfattah MD for their invaluable assistance with data collection and curation.
Footnotes
The contents of this publication are solely the responsibility of the authors and do not necessarily represent the official views of the MCA CV CRC.
The authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’ institutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information, visit the Author Center.
References
- 1.Rodriguez-Gabella T., Voisine P., Puri R., Pibarot P., Rodes-Cabau J. Aortic bioprosthetic valve durability: incidence, mechanisms, predictors, and management of surgical and transcatheter valve degeneration. J Am Coll Cardiol. 2017;70:1013–1028. doi: 10.1016/j.jacc.2017.07.715. [DOI] [PubMed] [Google Scholar]
- 2.Kostyunin A.E., Yuzhalin A.E., Rezvova M.A., et al. Degeneration of bioprosthetic heart valves: update 2020. J Am Heart Assoc. 2020;9 doi: 10.1161/JAHA.120.018506. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Hmadeh S., Trimaille A., Matsushita K., et al. Human aortic stenotic valve-derived extracellular vesicles induce endothelial dysfunction and thrombogenicity through AT1R/NADPH oxidases/SGLT2 pro-oxidant pathway. JACC Basic Transl Sci. 2024;9:845–864. doi: 10.1016/j.jacbts.2024.02.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Zoghbi W.A., Jone P.N., Chamsi-Pasha M.A., et al. Guidelines for the evaluation of prosthetic valve function with cardiovascular imaging: a report from the American Society of Echocardiography developed in collaboration with the Society for Cardiovascular Magnetic Resonance and the Society of Cardiovascular Computed Tomography. J Am Soc Echocardiogr. 2024;37:2–63. doi: 10.1016/j.echo.2023.10.004. [DOI] [PubMed] [Google Scholar]
- 5.Lancellotti P., Pibarot P., Chambers J., et al. Recommendations for the imaging assessment of prosthetic heart valves: a report from the European Association of Cardiovascular Imaging endorsed by the Chinese Society of Echocardiography, the Inter-American Society of Echocardiography, and the Brazilian Department of Cardiovascular Imaging. Eur Heart J Cardiovasc Imaging. 2016;17:589–590. doi: 10.1093/ehjci/jew025. [DOI] [PubMed] [Google Scholar]