Abstract 15282: Individual Participant Meta Analysis of the Cardiovascular Safety of Taspoglutide Among Individuals with Diabetes

Abstract

Background In 2008, the FDA recommended that novel agents for the treatment of diabetes be tested for cardiovascular (CV) safety. Taspoglutide, a long-acting once-weekly sc GLP-1 agonist reduced HbA1c but development of the compound was stopped by Roche due to gastrointestinal side effects and serious hypersensitivity reactions. It remained unclear whether treatment with a once-weekly GLP-1 agonist is associated with an acceptable CV safety profile.

Methods The T-EMERGE program included 7056 patients in 9 randomized controlled trials (See Figure) evaluating safety and efficacy of Taspoglutide vs. placebo or active comparator. The primary aim of this analysis was to examine the effect of Taspoglutide on CV outcomes (CV death, acute MI, stroke or hospitalization for unstable angina) assessing whether the upper boundary of the confidence interval was below 1.8, in line with FDA guidelines. Individual patient data were obtained from each trial and ORs were pooled across all studies using the Peto method.

Results During 7478 person-years of follow-up there were 58 patients with CV events (CV deaths, acute MI or stroke) and 9 patients hospitalized for unstable angina. There were 23 deaths. For the primary analysis the pooled OR (See Figure) was 0.94 (95% CI 0.57, 1.56), among those treated with Taspoglutide vs. the comparison groups. The OR was 0.97 (95% CI 0.57, 1.66) for CV death, acute MI or stroke and 0.89 (95% CI 0.38, 2.07) for all-cause mortality. Taspoglutide resulted in greater reduction in HbA1c than all comparators pooled (mean difference -0.43 95% CI -0.19, -0.67). Sensitivity analyses retaining studies with comparable exposure and populations produced materially similar results.

Conclusion Despite tolerability issues, exposure to Taspoglutide over 7478 person years of follow-up was within FDA CV safety margins for new anti-diabetics. Further development of GLP-1 agonists which are better tolerated may offer improvement in HbA1c along with a reduction in CV risk.