P1‐261: Modeling and simulation of plasma Aβ in humans after multiple oral doses of PF‐3084014, a potent gamma‐secretase inhibitor

PF-3084014, a selective and potent gamma secretase inhibitor, caused dose-dependent decreases in plasma Aβ after single and multiple dosing in humans. The initial reduction was followed by rebound elevation of 20-80%. Results from pharmacokinetic/pharmacodynamic (PK/PD) analyses of plasma drug and Aβ concentration data following oral administration of PF-3084014 in humans were used for the decision-making process in early development of PF-3084014. Population PK/PD models were fitted to plasma Aβ40 time-course data from a PF-3084014 multiple-dose study in healthy volunteers. The final model was selected based on the goodness of fit diagnostics and predictive check methods, which compared the observed mean response to data simulated from the parameter estimates and their uncertainty. Area-above-the-effect curve (AAEC) for plasma Aβ40 inhibition was selected as the primary PD parameter using a prespecified target for AAEC derived from preclinical PK/PD assessment of other gamma secretase inhibitors. Stochastic simulations using the final model were used to extrapolate to doses up to seven-fold higher. Percentiles of simulated data at the extrapolated doses provided an assessment of the probability that the drug can meet target effect levels. A composite indirect response model described the multiple-dose data well and provided AAEC aligned with observed values. Exposure levels at the predetermined exposure limits based on nonclinical toxicology findings were shown to have <1% probability of meeting the PD target on average. To attain an 80% probability of technical success, systemic exposures 2-3 fold higher than the imposed exposure limit would be needed. Doses yielding these exposures would be more than 3-fold higher than the highest dose previously studied, and therefore deemed unacceptable from a safety perspective. Based on PK/PD modeling and simulation, exposure levels needed to reach the PD target are 2-3 fold higher than exposure limits set on animal toxicology data. Therefore, doses substantially higher than those previously used would be needed to obtain the AAEC target with a high probability (>80%) of technical success. These findings influenced the early decision to halt further development of PF-3084014 for the Alzheimer's disease indication.