P009. Prediction of AMG 181 pharmacokinetics and pharmacodynamics in humans
W.‑J. Pan1, D. Salinger1, R. Ponce2, P. Prince1, S. Patel1, D. Doherty1, H. Hsu3, W. Ress4, K. Misura4, B. Sullivan5, J. Wisler6, Z. Yu7, D. Borie8, M. Gibbs1, L. Wienkers1
1Amgen Inc., Pharmacokinetics and Drug Metabolism, Seattle, United States; 2Amgen Inc., Toxicology, Seattle, United States; 3Amgen Inc., Inflammation, Thousand Oaks, United States; 4Amgen Inc., Molecular Sciences and Computational Biology, Seattle, United States; 5Amgen Inc., Clinical Immunology, Thousand Oaks, United States; 6Amgen Inc., Toxicology, Thousand Oaks, United States; 7Amgen Inc., Early Development, Thousand Oaks, United States; 8Amgen Inc., Inflammation Clinical Research, Thousand Oaks, United States
Background: Use of AMG 181 PK/PD parameters obtained in cynomolgus monkeys to predict human PK/PD profiles and a safe starting clinical dose.
Methods: A 2‑compartment model with parallel linear/non-linear elimination components was developed using monkey AMG 181 IV and SC PK data. Allometric scaling was applied to predict human PK parameters which were used to simulate exposures across a range of proposed single and multiple IV and SC doses for the human study. An Emax model was used to quantify the relationship between AMG 181 concentrations and free α 4 β 7 receptor (on CD4+ T cells). AMG 181 FIH doses were selected based on: the observed exposures at the NOAEL dose of 80 mg/kg IV in monkeys and the predicted human exposures by allometric scaling of monkey data; and the anticipated serum AMG 181 concentration that would produce >50% α 4 β 7 receptor occupancy (RO) in humans.
Results: The predicted AMG 181 clearance (CL) and volume of distribution were 118 mL/day and 4100 mL, respectively. The Emax model estimated E0, Emax, and EC50 values were 531, 0.922, and 14 ng/mL, respectively. At the FIH starting single SC dose of 0.7 mg, the predicted exposure margins were 52600 for Cmax and 72700 for AUC. At the highest single 0.5 hour IV infusion of 420 mg, the predicted exposure margins were 30 for Cmax and 6 for AUC. The proposed starting dose of 0.7 mg SC could achieve concentrations that were above EC50 from Day 2 to 5, and subsequently return to below the assay lower limit of quantification (LLOQ in human serum: 10 ng/mL) by Day 8. Since receptor antagonism generally requires high occupancy to be clinically effective, it was anticipated that such a short duration of partial engagement with α 4 β 7 would have no meaningful downstream pharmacological effect. The highest dose of 420 mg IV would provide safety experience at exposure much higher than EC90, which has been correlated to efficacy based on published data.
Conclusions: A minimally pharmacologically active dose and an appropriate dose escalation sequence were proposed for testing of AMG 181 in healthy volunteers and those with inflammatory bowel diseases based on a PK-PD model developed with data from cynomolgus monkeys.