Pharmacometrician Pfizer Fenton, Michigan, United States
Disclosure(s):
Sima Ahadieh, Master of Science: No financial relationships to disclose
Objective: This MBMA aimed to evaluate weight loss efficacy of glucagon-like peptide-1 (GLP1) agonists including liraglutide, semaglutide (SC and oral), orforglipron, GLP-1/glucose-dependent insulinotropic peptide (GIP) receptor agonist tirzepatide, and GLP-1/glucagon receptor agonist efinopegdutide in obesity trials.
Methods: A comprehensive literature search of electronic databases (Medline, Embase, Biosis, and clinicaltrials.gov) was conducted for relevant randomized controlled trials (RCTs) published up until January 2024 for the drugs of interest listed above and the primary endpoint of body weight percent change from baseline in obesity patients. Each publication returned from the literature search was reviewed for relevancy and digitized into a MBMA data file. NONMEM 7.5.0 (ICON plc) was used for analyses and R 4.2.1 was used for data cleaning, modification, and plotting. The total effect consisted of drug effect and placebo effect. Emax models, differentiated by class and formulation, were used to characterize dose response relationships, and an exponential model plateaued with respect to time was used to assess placebo responses as well as reaching the maximum drug effect. The covariates of interest included existence of diabetes type 2 (T2DM) comorbidity, average baseline body weight, and intensive behavioral therapies added to pharmacological interventions. A loss of response over time was also evaluated via a durability function. The final model was used to predict percent changes from baseline in body weight, and the mean and confidence intervals (CI) for each treatment. A shiny app was developed for interactive assessments of analysis results.
Results: The final MBMA dataset included 70 citations out of which 33 studies with 89 study arms were included in the quantitative analysis compromising 18023 subjects. The MBMA demonstrated significant body weight reductions across GLP1, GLP1/GIP, GLP1/Glucagon receptor compared to placebo. The effects of diabetes comorbidity and intensive behavioral therapy were significant on the drug effect. Durability function in placebo effect improved the model. The predicted percent changes in body weight [95% CI] for liraglutide 3 mg, semaglutide SC 2.4 mg, semaglutide oral 50 mg, orforglipron 45 mg, efinopegdutide 10 mg, and tirzepatide 10 and 15 mg, and placebo after 72 weeks for non-diabetic obese patients were -8.08 [-7.39, -8.77], -14.3 [-13.1, -15.4], -16.4 [-13.6, -18.7], -17.8 [-16.0, -19.6], -12.0 [-9.14, -14.5], -20.0 [-17.9, -22.2], -21.9 [-19.7, -24.4], and -1.79 [-1.35 , -2.26] respectively.
Conclusion: Longitudinal changes in body weight by GLP1 classes were characterized via MBMA. This analysis provided a benchmark and information about the dose-response relationship as well as temporal dynamics in body weight changes for future trial designs.
