S Chabaud, P Girard, P Nony , C. Laveille, JP Boissel.
Service de Pharmacologie Clinique Lyon, and Servier, Courbevoie FRANCE
Ivabradine (816257) is a new compound with specific bradycardic action and developed to prevent angina pectoris. In order to establish its PK-PD model in healthy volunteers, a double blind, placebo-controlled study was conducted in twelve healthy volunteers. Each volunteer received one simple blind placebo, then 3 single oral doses of 816257 (10, 20 and 30 mg) and a placebo (0) in a Latin square according to one of the 4 sequences: (10-20-0-30), (20-30-10- 0), (30-0-20-10), (0-10-30-20). Blood samples were taken before administration, and at time 0.5, 1, 1.5, 2, 3, 4, 6, 8, 10, 12 and 24 hours after administration. Parent drug S16257 and one of its metabolite S18982 were simultaneously assayed in plasma.
Bradycardic effects were measured using a solid memory ECG Holter recorder. Time intervals between two consecutive R waves (RR in ms) were recorded continuously from 5 minutes before administration to 24 hours after. This represented on average 100,000 measurements per volunteer per period, and effect, simply designed by RR, was expressed by the arithmetic mean of 10′ interval recording, which represented on average 600 measurements. Graphical plot of RR showed “chaotic” variation from 10′ to 10′ average, circadian ones reproducible between the placebos and active drugs and concentration effect related. The dose effect relationship (based on the 24h mean RR) was linear, which means that no plateau effect was reached. In order to eliminate “chaotic” and circadian variations, the RR measured under placebo was subtracted at each time and then filtered using a smoother (ΔRR).
Parent drug and metabolite concentrations expressed in nanomoles/liter were fitted simultaneously to a PK model using NONMEM and ADVAN 7. The best model included two compartments for the parent drug, a linear metabolism rate from 816257 to S18982 and a first-pass effect and two compartments for the metabolite. Plot of ARR versus fitted central compartments concentrations of both compounds showed a pronounced hysteresis plot. Use of peripheral compartments did not collapse the loops. Since there was no physiologic rationale for using an indirect action model, the effect compartment model was implemented with a sigmoid competitive PD model for the two molecules acting at the same receptor. Different models including one or two effect compartments were compared using Akaike informative criterion.
The best model was the one with 2 effect compartments, one for each compound. The equilibration was faster for metabolite than for parent compound (keo = 0.032 h-1 for S16257 versus 0.159 for S18982). However, the parent drug appeared more potent than the metabolite with an EC50 equals to 3.1 nmol/L versus 9.3 for the metabolite. Such a model that could include clinical trials simulations would help in designing phase III programmes.
Reference: PAGE 7 (1998) Abstr 683 [www.page-meeting.org/?abstract=683]
Poster: poster