Impact of GLP-1 agonists on cardiovascular outcomes in patients with sickle cell anemia: A propensity score matched retrospective cohort study

Abstract

Abstract Background: Sickle Cell Disease (SCD) is associated with worse cardiovascular (CV) outcomes. The Glucagon-like peptide-1 receptor agonists (GLP-1 agonists) semaglutide, liraglutide, and dulaglutide have been shown to reduce the risk of major adverse CV events in large-scale prospective trials of patients with type two diabetes mellitus and high CV risk. There is limited data, however, assessing the impact of GLP-1 agonists on CV outcomes in patients with SCD. Methods: We utilized data from the Global Collaborative Network-TriNetX to assess the impact of GLP-1 agonists on CV outcomes in patients with SCD. A literature search was conducted and nine CV endpoints reported to occur at higher rates in SCD patients were selected. Analysis was then split into two parts. The first assessed the prevalence of these endpoints in patients with SCD compared to the general population. Patients who received a GLP1 agonist were excluded. The prevalence of each outcome was assessed using corresponding ICD-10 codes. The second phase assessed the impact of the three GLP-1 agonists listed above on 10-year overall survival in SCD patients with these endpoints. For each endpoint, patients aged 0 to 90 were divided into two cohorts: those who received a GLP1 agonist and had both SCD plus the CV outcome of interest, and those who had not received a GLP1 agonist and had both SCD and the CV outcome of interest. Using ICD-10 codes, the following endpoints were selected: pulmonary hypertension (unspecified), HFpEF, HFrEF, combined heart failure, high output heart failure (HOHF), arrhythmias (unspecified), cardiac arrest, acute myocardial ischemia, chronic myocardial ischemia, and a composite endpoint including patients with any one of the nine outcomes. Prior to analysis, cohorts underwent propensity matching based on age, sex, and ethnicity. Generalized linear models were used to measure the association and estimates were presented as risk ratios and 95% confidence intervals. Kaplan-Meier analysis was used to assess survival. Estimates were presented as survival probability, hazard ratios, and 95% confidence intervals. Results: In the first phase of analysis, eight of nine outcomes were more prevalent in the SCD cohort compared to the general population (Table 1). Rates of HOHF were found to be similar in both cohorts. Due to the size of the cohorts and software limitations, measures of association could not be calculated. During the second phase, propensity score matching was performed for nine of ten analyses. Matching could not be performed for the HOHF cohort due to insufficient sample size. For example, in the composite endpoint analysis, after matching, the SCD cohort consisted of 848 patients, had a mean age 58.6 +/- 14 years, 68.75% were female, Caucasians accounted for 20.05% of patients and black or African American 70.87%. The risk of death at 10 years was found to be significantly lower in SCD patients receiving GLP-1 agonists in the following CV outcome cohorts: HFpEF - RR 0.496 (95% CI 0.344-0.716, p=0.0001), HFrEF - RR 0.544 (0.358-0.825, p=0.0033), arrhythmia – RR 0.453 (0.259-0.793, p=0.0041), cardiac arrest RR 0.626 (0.425-0.924, p=0.0138), acute myocardial ischemia – RR 0.585 (0.368-0.931, p=0.0213), and the composite CV outcome – RR 0.624 (0.475-0.818, p=0.0006). The survival probability at 10 years was found to be significantly higher in SCD patients receiving GLP-1 agonist in the following CV outcome cohorts: HFpEF – 73.354% vs. 41.189% (p=<0.0001), HFrEF – 66.002% vs. 50.215% (p=<0.0001), arrhythmia – 83.34% vs. 70.277% (p=0.0012), cardiac arrest – 41.08% vs. 19.087% (p=0.0032), acute myocardial ischemia – 72.324% vs. 57.788% (p=0.0030), and the composite CV endpoint – 61.846% vs. 75.471% (p=<0.0001). All hazard ratios were found to be statistically insignificant (Table 2). Conclusions: Our study confirms that SCD is associated with worse CV outcomes and that the use of GLP1 agonists significantly reduces the risk of these endpoints in patients with SCD. These results are comparable to other studies showing worse CV outcomes in SCD patients and add to the growing field of data demonstrating improved CV outcomes with the use of GLP1 agonists. More prospective studies are needed to assess the impact of GLP1 agonist therapy on rates of adverse CV outcomes in various patient populations, including those with hemoglobinopathies.