Emil Samara1, Gary Park2, Caroline Gundel2, Chrystelle Lamagna2, Vanessa Taylor2, and Andrew Chow2
1PharmaPolaris International, Inc., Maryland, USA, 2Rigel Pharmaceuticals Inc., California, USA
Objectives: Toll-Like Receptors (TLR) and Interleukin-1 Receptors (IL-1R) play a critical role in the innate immune response as microbial and tissue damage sensors, providing a bridge between the innate and adaptive immunity. Interleukin receptor associated kinases (IRAK) 1 and 4 are serine/threonine kinases that are essential for signaling downstream of most TLRs and IL-1Rs and the resulting production of pro-inflammatory cytokines. Suppression of TLR and IL-1R signaling through inhibition of IRAK1/4 kinases is a promising therapeutic approach for the treatment of inflammatory and autoimmune diseases. R835 is a potent and selective IRAK1/4 inhibitor currently under clinical investigation. A model-based approach is presented here to describe the concentration-response relationship of R835 in cytokine release inhibition induced by an intravenous (i.v.) administration of lipopolysaccharide (LPS, a TLR4 agonist) in healthy volunteers. The model developed was applied to describe the LPS effect in mice.
- To model the pharmacokinetics (PK) of R835 in healthy subjects and characterize the cytokine release profiles (PD) following a LPS challenge given with and without R835.
- To characterize the cytokine-release profile following a LPS challenge given with and without R835 in mice using the PK/PD model derived in humans
- To compare R835 potency between mice and humans
Methods: The human models were based on PK data of R835 after single and multiple dose oral administrations and the PD data in response to a 2 ng/kg i.v. LPS administered 2 hours after R835 (or placebo) administration. Nonlinear mixed effect modeling estimated with NONMEM 7.3.0 using the FOCE method was done in 3 steps. First, R835 plasma concentration-time data were fitted to a 1 or 2- compartment model with 1st order absorption and elimination. Secondly, the cytokine profiles in response to LPS challenge were characterized in the placebo subjects using a hypothetical one-compartment model with input and output rates. Finally, a combined model describing the placebo response and the effect of R835 on the inhibition of the cytokine release was developed. Due to model instability limited by the small N, the PD parameters in the combined model were fixed to the values derived from the placebo model. For both PK and PD parameters, exponential model was assumed for Interindividual variability, and a combined error model was used for the random effects. In mice, models were based on data from a single dose study, in which mice were administered R835 (15 mg/kg) or placebo, followed by a LPS challenge 30 minutes later. The same PK/PD modeling approach derived in humans was applied to describe the data in animals. Last, simulation analysis was conducted to compare the PD responses in mice and humans at similar systemic exposure for relative potency evaluation..
Results: In humans, a linear two-compartment model adequately fit R835 PK after both single and multiple dose. For the placebo model the longitudinal PK/PD model parameters included a zero-order release rate (kin0), a first-order release rate (kin1), and an output rate constant (kout). Once the placebo model was established, the drug effect was incorporated using an inhibitory model on kin1 that was parameterized using an intercept and a slope for R835 exposure effect. The same model adequately applied to the TNF-a, IL-8, and MIP-1b time profiles. The derived human model adequately fit the cytokine time profiles (IL6, IL8 and TNF-a) after LPS challenge administered with and without R835 in mice. The plasma concentrations of R835 after a 15 mg/kg oral dose in mice were similar to R835 concentrations after a 960 mg dose in humans. Simulation results show this R835 dose completely abolished the LPS effect on TNF and IL-8 release. This is consistent with the observations in mice after a single dose of 15 mg/kg.
Conclusions: A linear two-compartment model adequately described the PK profiles of R835 following single and repeated administration in humans. The PK of R835 appears dose and time independent across the doses evaluated. PK/PD modeling and simulation analysis reveals R835 can substantially attenuate the LPS pro-inflammatory cytokine response in a dose-dependent manner in both humans and mice and is related to circulating levels of R835 in plasma. The same mechanism-based PD model adequately described the cytokine release after LPS administration in both mice and humans. A similar pharmacodynamic response was attained at comparable systemic concentrations of R835 in humans and mice, showing R835 potency is similar in both species.
Reference: PAGE () Abstr 9308 [www.page-meeting.org/?abstract=9308]
Poster: Drug/Disease Modelling - Other Topics