Impact of changes in conventional risk factors induced by once-weekly GLP-1 receptor agonist exenatide on cardiovascular outcomes: an EXSCEL post hoc analysis.

BACKGROUND: The objective of this study was to examine the degree to which conventional cardiovascular (CV) risk factor changes induced by once-weekly exenatide (EQW) might explain the placebo-controlled differences in CV outcomes observed in the Exenatide Study of Cardiovascular Event Lowering (EXSCEL). METHODS: We entered participant-level risk factor values over time into a validated type 2 diabetes-specific clinical outcomes model to estimate event rates, and compared simulated with observed relative risk changes in EXSCEL. We performed simulations for each participant to minimize uncertainty and to optimize confidence interval precision around risk point estimates. Six outcomes were examined: major adverse CV event (MACE), all-cause mortality (ACM), CV death, fatal or nonfatal myocardial infarction (MI), fatal or nonfatal stroke, and hospitalization for heart failure (hHF). We also performed a mediation analysis using Cox regression models to evaluate potential key mediators for ACM. RESULTS: Model simulations explained only modest proportions of the observed relative risk reductions for MACE (29%), ACM (15%), CV death (18%), and stroke (29%), but greater proportions for hHF (67%) and MI (200%). Mediation analysis suggested that baseline-to-6 or 12-month changes in HbA1c, blood pressure, heart rate, low-density lipoprotein cholesterol, triglycerides, and weight did not mediate the EQW effect on ACM. CONCLUSIONS: These model simulations explain only a modest proportion of the impact of observed EQW-induced changes in conventional CV risk factors on EXSCEL outcomes, apart from hHF and MI. Up to 1-year changes in conventional risk factors did not mediate the observed ACM risk reduction.