Georg Hempel, Mats O. Karlsson*, Dinesh P. de Alwis John McNay, Nathalie Toublanc and Hans G. Schaefer
Eli Lilly and Company Limited, Clinical Pharmacology Windlesham, Surrey United Kingdom *Uppsala University, Sweden
Although approval of novel antihypertensive compounds is still based on manual blood pressure measurements, regulatory authorities ask for 24h-ambulatory blood pressure measurements (24h-ABPM) during the approval process. Analysis of 24h-ABPM tends to be difficult due to irregular time distribution of the data points and instability of the signal including artefacts or inherent spikes, i.e. noise associated with the data. Most often, the results of 24h-ABPM have been summarised as daily arithmetic mean values ignoring the important variability in blood pressure values.
Therefore, we have investigated the applicability of NONMEM to analyse 24h-ABPM. To describe the blood pressure profile without a drug effect, data from 142 patients derived from 2 studies were analysed. Different baseline models including cosine functions and polynomials have been tested. The baseline model was then implemented in a PK/PD model to describe the antihypertensive effect of moxonidine, a novel antihypertensive imidazoline I1, agonist.
Moxonidine plasma pharmacokinetics (PK) were described using a one-compartment model with first-order absorption and elimination. The final baseline model for blood pressure consists of two cosine functions allowing interoccasion variability, i.e. intraindividual deviations from day to day. For the PK/PD analysis of diastolic blood pressure measurements from 30 patients receiving either moxonidine or placebo an inhibitory Emax model with an effect link gave the best results. Emax was found to be 16.7% and EC50 was 0.95 pg/L.
We conclude that drug effects on 24h-ABPM data can be suitably described using NONMEM as the methodology can account for the different sources of variability.
Reference: PAGE 7 (1998) Abstr 688 [www.page-meeting.org/?abstract=688]
Poster: poster