Exploratory Cardiovascular Risk Modeling of Orforglipron Using the PREVENT Equations Compared with the Pooled Cohort Equations: A Post-Hoc Analysis from a Systematic Review and Meta-analysis.

Abstract

Introduction: Orforglipron, an oral non-peptide GLP-1 receptor agonist, improves multiple cardiometabolic risk factors. However, prior modeling used only the 2013 Pooled Cohort Equations (PCE), which ignore body mass index (BMI), estimated glomerular filtration rate (eGFR), and hemoglobin A1c (HbA1c). We re-evaluated its cardiovascular benefit using the 2023 AHA PREVENT equations, which account for more nuanced variables.

Aim: To compare the projected cardiovascular risk reduction from orforglipron when modeled using the 2013 PCE versus the 2023 AHA PREVENT equations, and to quantify the additional exploratory cardiometabolic benefit captured by PREVENT's equations.

Methods: In this post-hoc analysis, we applied pooled effect estimates from our systematic review and meta-analysis of five orforglipron RCTs to representative hypothetical patient profiles. We calculated 10-year and 30-year Cardiovascular disease (CVD) risk using the 2013 PCE and the 2023 PREVENT equations (base and HbA1c-enhanced models) at baseline and after modeled treatment effects at each dose (12, 24, 36, and 45 mg). Absolute and relative risk reductions were compared across tools. A sensitivity analysis was performed using the 95% confidence interval of the pooled treatment effects.

Results: In these modeled scenarios, the PREVENT HbA1c-enhanced model projected model-dependent relative risk reductions approximately 75-90% larger than those generated by the PCE across all patient profiles and dose tiers. For a representative 50-year-old man with T2D BMI 33 kg/m², HbA1c 8.0%, SBP 130 mmHg, TC 200 mg/dL, HDL 42 mg/dL, eGFR 85 mL/min/1.73 m²), the PCE estimated a 10-year RRR of 16.8% at the 36 mg dose, whereas PREVENT HbA1c-enhanced projected an RRR of 31.4% at the same dose-driven by additional capture of BMI reduction (- 2.6 kg/m²), HbA1c lowering (- 1.40%). Approximately one-third of the projected total risk reduction by PREVENT was attributable to HbA1c alone, a dimension the PCE structurally cannot assess. PREVENT also projected a 22.4% reduction in 10-year heart failure risk, a dimension absent from the PCE. PREVENT yielded lower baseline risk estimates than the PCE, consistent with known PCE overestimation. Notably, the PREVENT ASCVD-specific model projected reclassification of Profile A from borderline to low risk (< 5.0%) at all doses, whereas the PCE showed no reclassification from intermediate risk at any dose.

Conclusions: The AHA PREVENT equations may more comprehensively capture the projected cardiovascular benefit of orforglipron than the legacy PCE. In these modeled scenarios, exclusive reliance on the PCE may provide a less complete estimate of the projected cardiometabolic risk-factor benefit of orforglipron. For therapies targeting the cardiovascular-kidney-metabolic axis, PREVENT may offer a more comprehensive and clinically informative framework for estimating the cardiovascular benefit of orforglipron. Nevertheless, these projections remain model-derived and should be validated with individual patient data from future cardiovascular outcomes trials.