Title: Comparing the effectiveness of glucagon-like peptide 1 receptor agonists and sodium glucose co-transporter 2 inhibitors in a nationwide observational cohort study on the rates of cardiovascular outcomes.
Authors: Madison Jones, PharmD; Tanvi Patil, PharmD, BCPS, DPLA; John Minchak, PharmD, MBA, BCPS, BCGP; Alamdeep Kaur, PharmD, BCPS
Learning Objective: At the conclusion of my presentation, the participant will be able to explain the comparative effectiveness of glucagon-like peptide 1 receptor agonists (GLP1-RA) versus sodium glucose co-transporter 2 inhibitors (SGLT2i) on cardiovascular outcomes in a nationwide observational cohort study.
Background/Objective: The purpose of this study is to compare cardiovascular composite outcomes in patients newly initiation on GLP1-RA or SGLT2i.
Methods: This retrospective active comparator new-user cohort study included veteran patients who newly initiated on either GLP1-RA or SGLT2i. We collected electronic health record data from nationwide Veterans Health Administration (VHA) database from 1/1/2018 through 1/1/2024. Data was collected and combined from individual electronic medical records to the VA Corporate Data Warehouse (CDW) where it was modeled and prepared for use by The VA Informatics and Computing Infrastructure (VINCI) and extracted for study using sequel query language. We included patients who were newly started on either an SGLT2i or GLP1-RA. Patients taking a combination of a GLP1-RA and an SGLT2i any time before, during, or after the study period were excluded. Patients with type I diabetes, renal or liver transplants prior to or during the duration of the study were excluded. Patients were excluded if they did not have any encounters with VA healthcare system within the past 2 years of study index date or have history of prior use of any combination of GLP1RA or SGLT2i in the previous 1-year lookback period. Confounding was accounted via nearest-neighbor pairwise propensity score (PS) matching informed by expert-identified variables meeting the disjunctive cause criterion.
The primary outcome was a composite rate of 4-point major adverse cardiovascular events (MACE): ischemic stroke (IS), myocardial infarction (MI), coronary revascularization (CV) and hospitalization for heart failure (HHF). Secondary outcomes included comparing the rates of individual components of the primary outcomes, peripheral arterial disease (PAD) as well as chronic pulmonary disease (CPD) hospitalizations and emergency room visits and atrial fibrillation between the two groups.
Results: Matched cohorts included 51,919 patients in each of the GLP1RA and SGLT2i exposure groups. Greater than 86.8% were males with mean age of 62.5 years. The rate of four-point MACE was increased in the matched cohort when compared to the GLP1 cohort (incidence rate per 100 person-years (IR)= 22.5 vs 20.4; adjusted Hazard Ratio (aHR)= 0.88 [0.85-0.92]; p= <0.001
[MJ1] ). The rate of 3-pt MACE was lower in the GLP1RA group compared to the SGLT2i. (IR= 19.6 vs 17.9; aHR= 0.91 [0.87-0.95]; p= <0.001).
The secondary outcomes rates were lower in the GLP1RA group compared to SGLT2i for : MI (IR= 12.7 vs 11.5; aHR= 0.81[0.76-0.86]; p= <0.001), IS (IR= 16.2 vs 13.8; aHR= 0.84[0.8-0.89]; p=<0.0010), Any stroke (IR=27.4 vs 24.9; aHR=0.88[0.82-0.94]; p=<0.001), and coronary vascularization (IR= 2.06 vs 1.77; aHR=0.8[0.68-0.95]; p=0.01). The rates of HHF were not significantly different between the cohorts (IR= 6.5 vs 4.44; aHR=0.82[0.64-1.04]; p=<0.097). No statistically significant difference was noted in the rates of PAD(IR=25.9 vs 22; aHR= 0.99[0.93-1.07]; p=0.953) and atrial fibrillation (IR=41.2 vs 37.8; aHR=1.03[0.98-1.08; p=0.259), however atrial fibrillation did occur less frequently in the GLP1-RA group when limiting timeline to 90 days after therapy initiation (IR=27.5 vs 23.9; aHR=0.9[0.86-0.94]; p=<0.001) while the rates were similar at 180 days. The rates of CPD hospitalization (IR: 1.76 vs 1.42 SGLT2i ; aHR= 0.68[0.56-0.81]; p= <0.001) and emergency room visits (IR: 7.74 vs 6.21 SGLT2i; aHR= 0.85[0.78-0.92]; p=<0.001) were significantly lower in the GLP1RA group compared to SGLT2i.
Conclusions(s): Overall, the rate of 4-point MACE and 3-point MACE was significantly lower in GLP1RA group compared to SGLT2i, however, the rates of HHF were similar with numerically higher incidence rates in GLP1RA as compared to SGLT2i. No difference in the rates of Atrial fibrillation was found however the risk was lower at 90 days in the GLP1RA group however no significant difference in rates were noted between the cohorts at 180 days, indicating time varying confounders and warrants further research. Our study also showed that the rates of ER and Hospitalization for CPD were significantly lower in the GLP1RA group compared to SGLT2i in agreement. Future prospective studies should focus on ascertaining the difference between COPD and ASTHMA rates individually, differences in the rates of Atrial fibrillation with long term GLP1RA utilization as well as differences based on previous history of diabetes.
Self Assessment Question: MACE (MI, ischemic stroke, HF, CV) occurred more frequently in patients receiving an SGLT2i than patients using GLP1-RAs. (True/False)
