Objectives: To estimate the turnover rate of cortical amyloid beta (Aβ) in a non-transgenic preclinical mouse model and to compare this rate with literature derived estimates for both other preclinical animal models and human subjects. Methods: Preclinical data from a non-transgenic mouse model of efficacy was used to develop a pharmacokinetic-pharmacodynamic model to describe the relationship between plasma and brain concentrations of a new chemical entity (NCE) and cortical Aβ levels. The linked PD model consisted of a semi-physiological Aβ turnover model which was used to provide an indirect estimate of Aβ turnover rates. The animal level data consisted of plasma and brain NCE levels and cortical Aβ levels. Each animal received a single oral NCE dose and administered doses spanned a 30-fold range of dose distributed across 5 distinct dosing groups. Animals were sacrificed at 0.25, 1, 3, 6, 10, 14 and 24 hours post dose with 5 animals per time point per dose group. A limited review of the scientific literature was completed and comparable Aβ turnover rates estimated for other preclinical animal species and human subjects. Data analysis and graphical displays were performed with S-PLUS, while model building was performed with NONMEM. Results: Robust NCE plasma and brain exposures were observed across all dose groups, leading to significant reduction of cortical Aß. Plasma and brain concentrations of the NCE were well described by a three-compartment PK model with plasma levels represented by the central compartment and the blood-brain barrier represented by the first peripheral compartment in series with the brain levels represented by a second peripheral compartment. The PD model consisted of a semi-physiological Aβ turnover model. Saturating NCE doses were used to estimate the Aβ turnover rate in the FB mouse at 1.27 hr-1 (t½ = 33 minutes). This estimated rate was consistent with the literature derived value of 38 minutes (Barten et al, 2005). When allometrically scaled to humans, the estimated Aß turnover rate (0.17 hr-1) was reasonably close to the recently estimated CSF Aβ turnover rate in healthy volunteers of 0.085 hr-1 (Bateman et al, 2006). Conclusions: Aβ turnover rates in the FB mouse model were estimated at 1.27 hr-1. This rate is consistent with that reported in another mouse model. When allometrically scaled to humans, this rate was approximately twice the rate recently estimated for cerebrospinal fluid Aβ turnover in healthy human subjects. References:
[1] Barten et al, The Journal of Pharmacology and Experimental Therapeutics, 312:635-643, 2005
[2] Bateman et al, Volume 12 , Number 7 , July 2006 Nature Medicine | 
