Use of Time-to-event analysis to qualify activity of Ivabradine durign exercise tolerance tests

Ivabradine was developed for the treatment of angina pectoris. Its action combined to that of S 18982 (its active metabolite) results in a decrease of heart rate.

Objective: Assess efficacy of ivabradine and S 18982 during exercise tolerance tests (ETT) by investigating the relationship between dose-effect and clinical endpoint parameters (time of exercise duration –TED- and time to angina onset -TAO-).

Material and Method: In this present work, we used a PK/PD model previously developed from eight phase II/III studies to describe the impact of both ETT and plasma concentrations of ivabradine and S 18982 on heart rate.

In this model, the description of heart rate (HR) combines different parts: a linear effect of ETT on HR without treatment using two parameters: INIV (HR at rest) and SLOP (Impact of ETT on HR), and a treatment effect related to the plasma concentrations of ivabradine and S 18982 expressed through a parameter called INHIB (% of inhibition).

To determine impact of each patient’s covariates on the overall risk of experiencing a TED or TAO in pooled repeated oral phase II and phase III studies, a time-to-event analysis was proposed, taken into account the actual times at which event didn’t occur using a status “censored observation”.

In this analysis, the risk related to each clinical endpoint depends on a combination of a basic risk with a regression model selecting covariates among individual demographic parameters and HR description parameters (INIV, SLOP and INHIB).

This analysis was performed using NONMEM® Vers. V.1, the risk function being modelled using a Weibull regression.

Results: We have established the relationship between dose-effect and clinical endpoint parameters. This relationship shows that the effect of ivabradine on these parameters is driven by its impact on heart rate in a concentration-related manner.

Conclusion: Through this time-to–event analysis, we were able to take into account all available ETT information, including both observed clinical endpoints, and unobserved ones describing them as censored observations. Moreover, this analysis allows demonstration of a risk function difference between the different doses, even for ETT performed at the trough of concentrations.