Elevated
cytokine levels in inflammatory diseases are associated with downregulation of certain
cytochrome P450 (CYP)
enzymes. Upon treatment with some
cytokine-targeting therapeutic
proteins, the CYP
enzymes levels may be restored resulting in therapeutic
protein-mediated drug interactions (TP-DI). These analyses characterized the worst-case scenario for
CYP1A2, 2C9, and 3A-based TP-DI potential in patients with
psoriasis by comparing the pharmacokinetics of probe substrates between healthy volunteers and subjects with moderate to severe
psoriasis. Data for the CYP probe substrates
midazolam (
CYP3A),
caffeine (
CYP1A2), and S-
warfarin (
CYP2C9) from 7 drug interaction studies (1 in patients with
psoriasis and 6 in healthy subjects) were pooled to develop a population pharmacokinetics model for each substrate. A 2-compartment model with absorption lag time for
midazolam, a 1-compartment model with 5 transit absorption compartments for
caffeine, and a 3-compartment model with absorption lag time for S-
warfarin best described the observed data. Apparent oral clearance and relative bioavailability for
caffeine and S-
warfarin were not significantly different between the subject populations.
Psoriasis patients were estimated to have 17% lower
midazolam oral bioavailability than healthy volunteers. Compounded with other covariate effects, the ratio of median post hoc area under the plasma concentration-time estimates in subjects with
psoriasis relative to healthy subjects was 0.96, 1.13, and 0.65 for
midazolam,
caffeine, and S-
warfarin, respectively. Therefore,
inflammation in
psoriasis had no relevant effect on reducing
CYP1A2, 2C9, and 3A activities in vivo and no significant TP-DIs mediated through these
enzymes are expected in patients with
psoriasis. This approach can potentially be used in lieu of dedicated TP-DI studies to identify TP-DI risks within a disease area.