The risk of cardiovascular (CV) events after initiation of SGLT2i, GLP-1RA, DPP4i, or SUs in patients with T2D and low-to-moderate CV risk

Wednesday, August 28, 2024

9:00 AM – 9:15 AM CEST

Location: Room II

Presenting Author(s)

Helen Tesfaye, PharmD, MSc

Senior Research Specialist
Division of Pharmacoepidemiology, Dept. of Medicine, Brigham & Women’s Hospital, Harvard Medical School, United States

Co-Author(s)

JP

Julie M. Paik, MD, ScD

Brigham and Women's Hospital
Boston, United States
CA

Caroline Alix, BA

Brigham and Women's Hospital, United States
Elyse C. DiCesare, BA

Research Assistant
Division of Pharmacoepidemiology and Pharmacoeconomics, Brigham and Women's Hospital and Harvard Medical School, United States
SC

Sara J. Cromer, MD

Massachusetts General Hospital, United States
BE

Brendan M. Everett, MD, MPH

Brigham and Women’s Hospital, United States
RG

Robert J. Glynn, PhD, ScD

Professor
Brigham and Women's Hospital
Boston, United States
DW

Deborah J. Wexler, MD, MSc

Massachusetts General Hospital, United States
Elisabetta Patorno, MD, DrPH (she/her/hers)

Associate Professor of Medicine
Brigham and Women's Hospital and, United States

Background: Patients with type 2 diabetes (T2D) are at increased risk of major CV events (MACE) and heart failure. There is limited guidance on which T2D medications clinicians should prescribe in patients with low-to-moderate CV risk.

Objectives: To compare the risk of CV events in initiators of SGLT2 inhibitors (SGLT2i), GLP1 receptor agonists (GLP-1RA), DPP-4 inhibitors (DPP4i), or sulfonylureas (SU) with T2D and low-to-moderate CV risk.

Methods: Using Medicare and 2 U.S. private health plans (2013-2023), we identified six pairwise cohorts of 1:1 propensity score-matched (PSM) patients aged ≥18 years (≥66 years in Medicare) with T2D and no established atherosclerotic CV disease or heart failure, who initiated SGLT2i, GLP-1RA, DPP4i, or SU. Primary outcomes were MACE (i.e., myocardial infarction, stroke, or mortality), and a composite of hospitalization for heart failure (HHF, discharge diagnosis in the primary position) or all-cause mortality (HHF-ACM). We estimated pooled per protocol hazard ratios (HRs) with 95% confidence intervals (CI) in Cox proportional hazard models, adjusting for 156 covariates via PSM.

Results: We identified 342,849 PSM pairs in the SGLT2i vs. GLP-1RA cohort, 289,977 pairs in the SGLT2i vs. DPP4i cohort, 232,138 pairs in the SGLT2i vs. SU cohort, 264,880 pairs in the GLP-1RA vs. DPP4i cohort, 194,957 pairs in the GLP-1RA vs. SU cohort, and 278,245 pairs in the DPP4i vs. SU cohort. Mean age ranged between 60-63 years, and 49-56% were females. Compared to DPP-4i or SU, SGLT2i and GLP-1RA were associated with 15-28% reductions in the risk of MACE [HR: 0.85 (95% CI, 0.75-0.96) for SGLT2i vs. DPP4i; HR: 0.72 (0.68-0.77) for SGLT2i vs. SU; HR: 0.80 (0.74-0.87) for GLP-1RA vs. DPP4i; HR: 0.73 (0.67-0.79) for GLP-1RA vs. SU] and 32-41% reductions in the risk of HHF-ACM [HR: 0.60 (0.51-0.71) for SGLT2i vs. DPP4i; HR: 0.59 (0.54-0.64) for SGLT2i vs. SU; HR: 0.68 (0.63-0.74) for GLP-1RA vs. DPP4i; HR: 0.66 (0.60-0.73) for GLP-1RA vs. SU], over a mean follow-up of ~11 months. Compared with GLP-1RA, SGLT2i were associated with a small increase in the risk of MACE [HR:1.06 (1.01-1.11)] and a trend towards a small reduction in the risk of HHF-ACM [HR: 0.93 (0.85-1.01)]. In the DPP4i vs. SU cohort, the HR for MACE was 0.90 (0.86-0.93) and the HR for HHF-ACM was 0.93 (0.87-1.00).

Conclusions: This study suggests robust reductions in rates of MACE and HHF-ACM with SGLT2i and GLP-1RA vs. DPP4i or SU in patients with T2D and low-to-moderate CV risk. There were no large differences in risk of CV events when comparing GLP-1RA vs. SGLT2i and DPP-4i vs. SU.