Pinky Dua (1), Neil Benson (2, 4) and Piet H. van der Graaf (3)
Pfizer/Neusentis, (1) Clinical Pharmacology, (2) PDM & (3) Pharmacometrics, Sandwich/Cambridge & (4) Xenologiq, Canterbury, UK.
Objectives: The aim of this work was to integrate the knowledge of the biology of NGF response in a Quantitative Systems Pharmacology (QSP) framework to guide drug discovery and development programs, for novel drug target identification, target validation, biomarker and biomeasure [1] selection, dose prediction and ultimately patient selection (“precision medicine”).
Methods: In sensory neurons, NGF engages with its receptor known as tropomyosin receptor kinase A (TrkA) causing an accumulation of diphosphorylated extracellular signal regulated kinase (dppERK) in the nucleus and subsequently the expression of numerous genes related to neuronal survival and pain sensation; dppERK concentration could therefore be regarded as an upstream biomarker linked to pain response. Two Systems Biology models for describing the events leading to dppERK migration [2] and nuclear-cytoplasmic shuttling of ERK [3] were coupled and the resulting ordinary differential equation model (59 molecular species and 233 parameters) was converted into a QSP model by expressing it in three inter-connected compartments and integrating a pharmacokinetic component, allowing for investigations of small molecules as well as monoclonal antibodies. The model was calibrated using in-house and external preclinical and clinical data.
Results: Using sensitivity analysis it was concluded that after NGF itself, TrkA was one of the more sensitive druggable targets (target selection). With the focus on TrkA, the initial model was used to develop a reduced model to explore the characteristics required for a successful hypothetical TrkA inhibitor (target validation). The model could then be used to provide dose predictions for new entities without the need for any data from animal models of ‘disease’.
Conclusions: Using the model, a hypothetical drug with typical characteristics could inhibit the NGF pathway at conceivable plasma concentrations. Given the importance of the TrkA concentration to the observed response, TrkA concentrations were experimentally determined in a range of cell types including a mammalian neuronal cell line. Levels varied markedly between cell types, highlighting the need for a reliable estimate in a relevant human cell type. QSP models incorporating such biomeasure data could be of great utility at all stages in drug discovery and development.
References:
[1] Van der Graaf P & Benson N (2011) Pharm. Res. 28, 1460-1464
[2] Sasagawa S et al. (2005) Nature Cell Biol. 7, 365-372
[3] Fujioka A et al. (2006) J. Biol. Chem. 281, 8917-8926
Reference: PAGE 21 () Abstr 2481 [www.page-meeting.org/?abstract=2481]
Poster: New Modelling Approaches