GP3-6: Mixed effects PK/PD modelling of NNC0195-0092 - a noncovalent albumin binding growth hormone derivative in the hypophysectomised rat

Abstract

Introduction: The hypophysectomised rat is used during preclinical development to evaluate the pharmacokinetic and pharmacodynamic (PKPD) properties of new GH analogues . The animal model resembles the human GH deficiency with lack of GH secretion, low levels of serum IGF-I and growth retardation . Non-linear mixed effects modelling of the PKPD relationship may improve the mechanistic understanding of action, and form the basis for translation of pre-clinical parameters into meaningful and accurate human predictions for early clinical development . A mechanistic mixed-effects PKPD model for NNC0195-0092 a non-covalent albumin binding growth hormone derivative has been developed in hypophysectomised Sprague Dawley rats . Methods: Animal studies with single and multiple dosing of NNC0195-0092 was used to establish the PKPD relationship between NNC0195-0092 and insulin-like growth factor I (IGF-I) Non-linear mixed effects modelling software (NONMEM) was used to model the simultaneous relationship between NNC01950092, IGF-I and bodyweight gain . Derived parameters was scaled directly from model values using the allometric approach with exponents fixed to conventional values . Results: A final PKPD model was established, that described the PK of NNC0195-0092 as a two compartmental model with both linear and non-linear elimination terms, transit compartment absorption . The induction of IGF-I was modelled using an indirect response model with stimulation of kin with an Emax relationship to rhGH dose . Direct scaling form the PKPD model provided good predictions of the human PKPD except for subcutaneous bioavailability were the approaches failed . Conclusion: A PKPD model of pre-clinical animal data was developed . The model provides a quantitative way of comparing results from testing of new GH analogues in the hypophysectomised rat model to current treatment . Furthermore, direct scaling of parameters matched or improved results obtained when using multispecies scaling to predict human parameters .