Mark Penney (1), Eleni Pefani (1), René van der Merwe (1) and Paul Newbold (2)
(1) MedImmune plc, Cambridge, UK, (2) MedImmune LLC, Gaithersburg, MD, USA.
Objectives: To determine the drop-out rate from neutropenia in a Phase 2a study of the monoclonal antibody MEDI8968 (aka AMG 108). To develop PKPD models for the exposure of MEDI8968 in COPD patients and its effect on neutrophils and C-Reactive Protein (CRP) to demonstrate pharmacology and analyse the PKPD of MEDI8968 in reducing the rate of Acute Exacerbations of COPD (AECOPD).
Methods: A PKPD model based on data from previous studies in healthy volunteers, osteoarthritis and rheumatoid arthritis patients was adapted for COPD patients by simulating from baseline neutrophil counts typical of a COPD population; MEDI8968 PK and its suppression of neutrophil count were assumed to be conserved. This model was combined with the study process following a low count and the percent of subjects predicted to have the outcomes of skipping a dose or dropping out calculated.
Model-predicted PK exposure and neutrophil counts were compared to the results of a Phase 2a study in COPD patients. Both PK and PKPD models were updated and tested for study-dependent covariates. The pharmacology of MEDI8968 was investigated further by constructing a PKPD model for MEDI8968-induced suppression of CRP. AECOPD rate was investigated using a time to event analysis, using a constant hazard model and introducing average PK exposure as a covariate.
Results: The model correctly predicted negligible drop out from neutropenia with a per-protocol subject recruitment cut-off of 2.5×10^9 neutrophils/L. PK and neutrophil count were well predicted. The updated model had minimal changes and study was not a covariate, confirming that COPD subjects had similar PKPD to those previously tested. Exposure in COPD subjects was sufficient to cause maximum suppression of the neutrophil count. Suppression of CRP was maximal as judged by the similar suppression obtained in [2] and [3]; however, no relationship to exposure could be derived due to the single dose level of MEDI8968 used in the study. AECOPD rate was described well with a constant hazard model. However, no effect of MEDI8968 on the time-to-exacerbation was found.
Conclusion: The model could account for the differences in blood neutrophil count between COPD and previously tested patients with a simple adjustment of the baseline counts and so confirm appropriate recruitment criteria to prevent subject drop-out. The updated model could confirm full pharmacology was achieved with the MEDI8968 dose, although no improvement in AECOPD rate was observed.
References:
[1] Beal SL, Sheiner LB, Boeckmann AJ & Bauer RJ (Eds.) NONMEM Users Guides. 1989-2011. Icon Development Solutions, Ellicott City, Maryland, USA.
[2] Cohen SB, Proudman S, Kivitz AJ, Burch FX, Donohue JP, Burstein D, Sun Y-N, Banfield C, Vincent MS, Ni L, Zack DJ. A randomized, double-blind study of AMG 108 (a fully human monoclonal antibody to IL-1R1) in patients with osteoarthritis of the knee. Arthritis Research & Therapy 2011, 13:R125
[3] Cardiel MH, Tak PP, Bensen W, Burch FX, Forejtova S, Badurski JE, Kakkar T, Bevirt T, Ni L, McCroskery E, Jahreis A, Zack DJ. A phase 2 randomized, double-blind study of AMG 108, a fully human monoclonal antibody to IL-1R, in patients with rheumatoid arthritis. Arthritis Research & Therapy 2010, 12:R192
Reference: PAGE 24 (2015) Abstr 3580 [www.page-meeting.org/?abstract=3580]
Poster: Drug/Disease modeling - Absorption & PBPK