6488 Factors associated with the response to SGLT2 inhibitors and GLP-1 Receptor Agonists in Veterans with Type 2 Diabetes

• Published in: Journal of the Endocrine Society 2024-10, Vol.8 (Supplement_1)
• Main Authors: Hashmi, Sulman, Samson, Kaeli, Arora, Gunjan, Azad, Sarah Khozema Hussain, Awad, Dana, Desouza, Cyrus V
• Format: Article
• Language: English
• Subjects: Abstract
• ISSN: 2472-1972; 2472-1972
• DOI: 10.1210/jendso/bvae163.772

Abstract Disclosure: S. Hashmi: None. K. Samson: None. G. Arora: None. S. Azad: None. D. Awad: None. C.V. Desouza: None. Background: SGLT2 inhibitors and GLP1 agonists are recommended as first line pharmacological therapy (along with metformin) for type 2 DM, particularly in patients with CKD, CVD, or HF. Despite their effectiveness, there is limited evidence available to predict which patient subsets will respond favorably to SGLT2 inhibitors or GLP1 agonist. Purpose of study: To determine if baseline characteristics can predict response to therapy and to compare the differences in characteristics between SGLT2 inhibitor and GLP1 agonist groups. Methodology: Retrospective study of veterans ages 19-89 yr with T2DM initiated on treatment from Jan 1, 2015, to Dec 31, 2022, with either SGLT2 inhibitor or GLP 1 agonist for at least 6 months. The cohort comprised 866 patients, of which 544 individuals met the inclusion criteria (332 in the GLP-1 agonist group and 212 in the SGLT2 inhibitor group). Baseline characteristics assessed included age, sex, race, comorbidities, weight, BMI, HbA1c, eGFR, BP and lipids. Treatment response thresholds were defined as an HbA1c reduction of more than 0.5 percentage points and more than 3% of pretreatment weight loss, evaluated 6-12 months after initiating treatment. A complete response was defined as both treatment thresholds being met. Statistical analysis involved Chi-Square tests and independent samples t-test between variables of interest. Logistic regression was conducted with the dichotomous complete response outcome, for variables with bivariate p-values less than 0.10, in addition to variables preselected for inclusion based on the clinical relevance. Results: Three variables were statistically significant in the final logistic regression with complete response outcome. Patients on GLP1 had an increased adjusted odds of having a complete outcome (AOR: 2.45 (95% CI: 1.45, 4.15) relative to patients on SGLT2. In addition, older patients (AOR: 1.34 (95% CI: 1.02, 1.74)) and patients with a higher baseline HbA1c (AOR: 1.18 (95% CI: 1.02, 1.36)) had increased adjusted odds of having a complete response relative to patients 10 years younger and patients with an A1c 1 unit lower, respectively. 72.6% of patients on GLP1 agonists achieved the A1c response goal, in contrast to 48.6% on SGLT2 inhibitors (p < 0.001). Also, 47.6 % of individuals achieved weight loss goals on GLP1 agonist as compared to 41 % on SGLT2 inhibitor. A