Abstract Body (Do not enter title and authors here): Background:
Sodium-glucose co-transporter-2 inhibitors (SGLT2i) are guideline-directed therapy for heart failure with reduced ejection fraction (HFrEF), yet their utility in patients supported with left ventricular assist devices (LVADs) remains understudied. This study examined the association between SGLT2i use and clinical outcomes in LVAD recipients over a standardized 1-year follow-up.

Methods:
We queried the TriNetX Global Collaborative Network to identify adult patients with an LVAD (ICD-10 Z95.811 and procedure codes for insertion/revision). Patients were stratified based on exposure to SGLT2 inhibitors (canagliflozin, dapagliflozin, empagliflozin, ertugliflozin) post-implant. The primary outcome was all-cause mortality at 1 year. Secondary outcomes included hospitalizations, acute kidney injury (AKI), bloodstream infections (MSSA/MRSA), and device-related complications. Propensity score matching (1:1) was performed on demographics, comorbidities, and background medical therapy. Kaplan-Meier survival analysis was used to compare time-to-event outcomes.

Results:
After matching, 711 patients were included in each cohort. At 1 year, mortality was significantly lower in the SGLT2i group (8.0% vs. 24.2%; HR: 0.28 [95% CI: 0.21–0.38]; p<0.001). SGLT2i use was also associated with reduced hospitalization (62.3% vs. 70.7%; HR: 0.59; p<0.001), AKI (43.6% vs. 58.2%; HR: 0.56; p<0.001), and bloodstream infections (10.7% vs. 15.7%; HR: 0.59; p=0.002). The SGLT2i group had a higher rate of device-related infection (26.7% vs. 20.3%; HR: 1.20; p=0.099), but this did not reach statistical significance in survival analysis.

Conclusion:
In this large real-world analysis, SGLT2 inhibitor use in LVAD patients was associated with markedly reduced 1-year mortality and improved cardiorenal and infectious outcomes. These findings suggest a potential role for SGLT2i in optimizing medical therapy in LVAD recipients, a population not included in pivotal HF trials. This retrospective analysis may be affected by residual confounding despite rigorous matching. Medication timing, dose, and adherence were not available, and cause-specific mortality could not be assessed. While a 1-year follow-up standardized time at risk, long-term effects remain unknown. Prospective validation is needed.