Duke Health Referring Physicians

Article

Insights From EXSCEL: A Risk-Modeling Analysis of the Effect of Once-Weekly Exenatide on Cardiovascular Outcomes

The multinational Exenatide Study of Cardiovascular Event Lowering (EXSCEL) trial demonstrated that glucagon-like peptide-1 receptor agonist exenatide taken once weekly does not increase the risk of a cardiovascular (CV) event. Full results of the effect of once-weekly exenatide in patients with CV disease and associated treatment response are pending.

Robert J. Mentz, MD, of Duke, presented a risk-modeling analysis of the EXSCEL trial at the American Heart Association’s Scientific Sessions held November 11 to 15, 2017, in Anaheim, CA. Other investigators involved in the analysis included Vivian P. Thompson, Yuliya Lokhnygina, PhD, MS, Neha J. Pagidipati, MD, and Adrian F. Hernandez, MD, of Duke.

EXSCEL was a large-scale, double-blind, event-driven pragmatic trial. Researchers randomized 14,752 individuals with type 2 diabetes mellitus to receive either once-weekly exenatide or placebo. Patients enrolled in EXSCEL had glycated hemoglobin (HbA1C) levels of 6.5% to 10%. There were no criteria specifying degree of CV risk; however, about 70% of enrolled participants had a previous CV event.

The primary endpoint was a composite of CV mortality, nonfatal myocardial infarction, or nonfatal stroke (MACE). Mentz and colleagues hypothesized that patients with a higher risk of all-cause mortality and MACE would see a greater treatment benefit from once-weekly exenatide than those with lower risk. At a median follow-up of 3.2 years, once-weekly exenatide was noninferior to placebo on the composite endpoint (11.4% and 12.2%, respectively; P < .001). Although exenatide was associated with fewer events, it did not demonstrate superiority regarding efficacy (P = .061). In addition, exenatide had a 14% relative reduction in all-cause mortality, but this was not statistically significant based on a prespecified hierarchy. Whether a patient’s baseline CV risk affected treatment response was unclear based on the top-line results.

Researchers used a predictive model that included a stepwise approach to select baseline characteristics. The model was validated using an optimism-corrected c-index statistic. A hazard risk (HR) score for all-cause mortality and MACE was created for each patient. A time-to-event model was then developed, which included HR score, treatment assignment, and interaction between risk score and treatment.

Mentz commented, “The model was well calibrated. The predictive model for all-cause mortality matched what was observed in the EXSCEL patient population. The multivariable model for MACE tended to overestimate risk, particularly in patients with higher risk.”

The multivariable model included demographic information, such as age and sex; past medical history (comorbidities); and key findings from examination and laboratory tests. In the analysis, patients were segmented into 2 groups: those who died from any cause and those who experienced MACE. Patients who died during the study had a greater CV comorbidity burden than those who experienced MACE. In particular, patients in the all-cause mortality group were more likely to have hypertension, cerebrovascular disease, heart failure, atrial fibrillation, and chronic respiratory disease than patients who experienced MACE. The HR for the overall cohort was 0.86 for all-cause mortality, favoring exenatide. There was consistent benefit across patient groups stratified by CV risk on both all-cause mortality and MACE.

Baseline characteristics, including age, prior CV events, comorbidity burden, laboratory values (eg, HbA1C, body mass index), and examination findings (eg, diastolic blood pressure), had a prognostic value for all-cause mortality and MACE, according to Mentz.

Mentz concluded: “Exenatide is associated with modest reductions in all-cause mortality and MACE and has a broad spectrum of effectiveness across various levels of CV risk. We, as clinicians, have to assess the potential benefit and risk of each drug class—and drugs within a class—for each patient.”

Source: Mentz RJ, Bethel MA, Thompson VP, et al. Effect of exenatide once-weekly on clinical outcomes in patients with type 2 diabetes mellitus and cardiovascular disease: insights from the EXSCEL trial. Presented at: American Heart Association Scientific Sessions 2017; November 11-15, 2017; Anaheim, CA. Abstract LBS.04.