The accumulation of amyloid beta (Aβ) peptides is believed to be a central contributor to the neurodegeneration seen in the Alzheimer's disease (AD) brain. Given the central role of Aβ42 in AD pathogenesis, a therapeutic strategy to lower central Aβ42 (and Aβ40) levels via inhibition of BACE was adopted in a first in human trial in a 2-part adaptive design.
Part 1 evaluated PF-05297909 plasma PK and the PK/PD relationship for the reduction of plasma Aβ40, Aβ42 and AβX levels; Part 2 evaluated the exposure-response relationship between PF-05297909 and CSF levels of Aβ40, Aβ42 and AβX. Sufficient safety and tolerability, plasma exposure and reduction in plasma Aβ were necessary to initiate Part 2. Part 1 was a sequential parallel group dose escalation (25, 100, 250 and 325 mg) with n=8 (6:2, active:placebo) healthy volunteers (HV) in each cohort. Part 2 consisted of 3 cohorts of n=8 (6:2, active:placebo) HV. Doses selected for Part 2 started with the highest safe dose in Part 1 and then adapted for subsequent cohorts. The PK/PD relationship between PF-05297909 and Aβ42 was determined using a non-linear mixed effects (NLME) analysis. The doses for Part 2 - cohort 2 and 3 were to be chosen to improve the relative standard error in the estimate of the BACE IC50 as quantified by evaluating the determinant of the Fisher information matrix for the NLME model.
PF-05297909 was well-tolerated. Reduction in plasma Aβ (Aβ40 and Aβ42) was exposure related with an apparent maximum at the 250 mg dose with a greater duration of activity at the 325 mg dose of PF-05297909. A 325 mg dose was selected for Part 2 - cohorts 1 and 2 without further cohorts being run, as stopping criteria for futility were met following analysis of cohort 2. A PK/PD relationship in CSF was not observed.
The adaptive designed PF-05297909 FIH study allowed efficient testing of safety and of the PK/PD relationship between PF-05297909 exposure and Aβ (Aβ40 and Aβ42). PF-05297909 was safe and well tolerated in HV at exposures tested. A robust effect on plasma Aβ did not translate to CSF pharmacodynamic effects.
2013. Vol. 9, no 4, P287- p.