J. DeJongh1/2, D. Eckland3, R. Heine4 and M. Danhof2.
1LAP&P Consultants, Leiden, The Netherlands; 2Leiden/Amsterdam Center for Drug Research, Leiden, The Netherlands; 3Takeda European Research and Development, London, U.K.; 4Institute for Endocrinology, Reproduction and Metabolism, Vrije Universiteit, Amsterdam, The Netherlands
Pioglitazone (Actos) is a novel drug for treatment of Non-Insulin Dependent (type 2) Diabetes Mellitus (NIDDM). Its efficacy has been evaluated by classical analysis of variance with a “last observation carried forward” approach. The present study was performed to identify a population model for glycaemic control by pioglitazone, including NIDDM disease progression. Data from a 6-month randomized, placebo-controlled monotherapy study (0, 7.5, 15, 30 or 45 mg/day) were evaluated. Therapy was preceded by a 5-7 week observation period. 3019 plasma samples from 382 subjects were considered. Glycaemic control was monitored through fasting blood glucose (FBG) and HbA1C levels. FBG levels were modelled by zero-order influx, and first-order decrease with a rate constant decreasing over time, leading to a pathological state. HbA1C levels were described as a delayed function of FBG, using an effect compartment. Pioglitazone treatment effects were expressed by counteracting the time-dependent decrease of glucose removal. An adequate description of the data was achieved by the model. The residual error for the model fit of Hb1Ac observations was smaller than for the FBG observations (3.2 vs. 7.7 %). Half-life for the Hb1Ac effect compartment was 2.5 weeks. In addition, simulations showed that a number of 50 patients per treatment group is sufficient for significance (P<0.05) testing between placebo and the lowest active drug treatment group. An accurate and precise description of glycaemic control before and during treatment of NIDDM was achieved with a model that brings observations for FBG and HbA1c under a single, common denominator.
Reference: PAGE 10 (2001) Abstr 189 [www.page-meeting.org/?abstract=189]
Poster: poster