Chemistry Reference
In-Depth Information
prolongation was not predicted by in vitro assessments of hERG inhibition
since QT prolongation was observed with plasma drug concentrations of 2 mM
free drug while the whole-cell patch-clamp hERG IC
50
of BMS-641988 was
46 mM (free drug). Lack of activity in sodium or calcium channel whole-cell
patch assays, absence of any significant effects on action potential duration in
Purkinje fiber assays, or QT effects in suspended rabbit hearts were also in
contrast to the observed in vivo cardiovascular effects. The observed QT pro-
longation was unlikely to be due to the active metabolites BMS-949 and BMS-
511, since their combined concentrations did not exceed 2% of parent drug
C
max
in dogs. Furthermore, BMS-949 did not cause QT prolongation in dogs at
concentrations significantly higher than those achieved in the BMS-641988 dog
cardiovascular safety study.
In summary, BMS-641988 did not cause convulsions at any dose. It is an
antagonist of the GABA
A
receptor, but it did not induce seizures in dogs up to
the maximum dose tested (75mg kg
1
day
1
) with a plasma drug concentration
of 80 mM(13 mM free). It showed limited brain exposure across all species
(brain to plasma ratio of 0.02 to 0.15). Concentrations of the seizurogenic
active metabolite BMS-949 were higher in rodents than in dogs. A significant
margin (420-fold) was established, based on extensive preclinical seizure stu-
dies and the predicted human exposure of BMS-949 at the ecacious dose.
However, the therapeutic index for the observed QT prolongation was rela-
tively low (5-fold). In light of its impressive ecacy and tolerable safety profile,
BMS-641988 was selected for clinical evaluation of safety and ecacy in
humans. Simultaneously, a backup program was initiated to address the
potential liabilities of this clinical candidate.
6.4.3 Second-generation AR Antagonist: BMS-779333
The goal of the backup program was to identify an AR pan-antagonist with an
improved safety profile compared to BMS-641988 and with an equivalent or
superior in vivo ecacy in tumor models. The desired compound would exhibit
a significantly improved safety margin for cardiovascular effects, would be
devoid of chemical stability issues under physiologic conditions, and would
have no major active circulating metabolites in preclinical studies. The sulfo-
namide group in BMS-641988 was implicated in the formation of both the
major metabolites, whereas the imide functionality of BMS-641988 was
responsible for its chemical equilibrium with the ring-open forms. Based on
extensive SAR, neither of these two functionalities appeared to be essential for
potent AR antagonist activity in the lead series. Analysis of the available crystal
structures of the T877A AR LBD with imides in agonist mode further con-
firmed that the imide moiety generated no significant interactions within the
LBD. Therefore, initial medicinal chemistry efforts were focused on the tricyclic
lactams and sultams represented by the lead compounds 11 and 12, respectively
(Figure 6.7). However, both the lactam and sultam series were eventually
dropped, primarily due to suboptimal PK properties leading to a lack of robust
in vivo ecacy in tumor xenograft models.
Search WWH ::
Custom Search