M&S in Preclinical Development Predicting thyroid hormones side - PowerPoint PPT Presentation

M&S in Preclinical Development Predicting thyroid hormones side effects in human from preclinical toxicity studies EMA/EFPIA M&S Workshop BOS1, 1 December 2011 Sandra Visser Global Network Leader Non-Clinical Modeling & Simulation Implementation Leader Model Based Drug Discovery AstraZeneca, Innovative Medicines, Global DMPK CoE

Modelling and Simulation Continuum at AZ Target Discovery Preclinical Early Clinical Late Clinical LCM Selection Development Development Development Biology Pharmacology Disease Network Target Engagement Efficacy Target Safety Safety E max PD EC 50 C max Right Target Tissue Exposure Dose Schedule Trial Design Learning Learning Learning Learning Learning Learning Confirming Confirming Confirming Confirming Confirming Confirming Predictive Sciences Modeling Network Sandra Visser | 1 December 2011 Innovative Medicines | Non-Clinical Modeling and Simulation | MBDDx

Model Based Drug Discovery and Development Pharmacokinetic-Pharmacodynamic Principles PHARMACODYNAMICS PHARMACOKINETICS Dose k eo k on Outcome C p C e Target Target Disease Patho- Occupancy Mechanism Process physiology Plasma Target site k off Target Engagement Target Exposure Transduction to Efficacy/Safety Compound-specific properties System-specific properties Using a quantitative pharmacology approach to support decision making, by establishing a translational exposure – target engagement – efficacy/safety model in animals and humans and predicting the dose to man, optimal dosing schedule and clinical study design. 3 Sandra Visser | 1 December 2011 Innovative Medicines | Non-Clinical Modeling and Simulation | MBDDx

Biomarker Classification Map Target Engagement and Clinical Outcome represents quantitative Type 4B relationship between biomarkers Physiological Response Type 5 Type 0 Type 1 Type 2 Type 3 Type 4A Pathophysiology Type 6 Genotype/ Drug Target Target Physiological or Disease Outcome phenotype concentration Occupancy Mechanism Response Process Biomarkers for quantitative pharmacological support of the biological hypothesis PHC PoM PoP PoC PoM: degree, duration of target engagement sufficient for viable hypothesis test PoP: beneficial effect on targeted disease process or pathophysiology PoC: beneficial effect on clinical outcome Adapted from Danhof et al Pharmaceutical Research 22(9)1432. 2005 4 Sandra Visser | 1 December 2011 Innovative Medicines | Non-Clinical Modeling and Simulation | MBDDx

Model Based Drug Dx and Dv Strategy Aspiration and Benefits of applying Quantitative Pharmacology Strategy along value chain MBDDx MBDD Target Lead Lead Early Late Validation Generation Optimization Development Development Experimental Design, Modeling and Simulation of in vivo efficacy and safety Translational Science: back and forward translation at all stages Strategic translational PKPD and Rational optimization and selection Utilizing quantitative modeling for biomarker investment plan decision making. Predict human PKPD and dose PKPD-based Target Validation regimen and assess safety margins Optimal clinical design (cost savings and cost avoidance) for showing target engagement Confidence in vitro and in vivo models. Improved confidence in human target (PoM) and positive effect on disease process and Preliminary PKPD model for target engagement (PoM) and affecting disease engagement, efficacy and safety. process/pathophysiology (PoP/C) and support pathophysiology in the right patient population (PoP/C). Rational crititeria for candidate drug decision making 5 Sandra Visser | 1 December 2011 Innovative Medicines | Non-Clinical Modeling and Simulation | MBDDx

Case Study: Mechanism-based Pharmacokinetic- Pharmacodynamic Feedback Model of Thyroid Hormones after Inhibition of Thyroperoxidase in the Dog: Cross- species Prediction of Thyroid Hormone Profiles in Rats and Humans. Petra Ekerot, Douglas Ferguson, Sandra Visser PAGE 20 (2011) Abstr 2150 [www.page-meeting.org/?abstract=2150] Acknowledgements: Phil Mallinder, Steve Jordan, Elaine Cadogan, Matt Soars, Håkan Eriksson, Eva-Lena Glämsta, Lars B Nilsson, Anders Viberg, Olof Breuer, Susanne Rosqvist, Bert Peletier

Impact of TPO inhibition on Thyroid Hormones Thyroperoxidase (TPO) is a key enzyme involved in the synthesis of thyroxine (T4) and triiodothyronine (T3) thyroid hormones. The thyroid hormones T 4 and T 3 play important roles in metabolism, growth and development. T4 (& T3) inhibit the synthesis of TSH. TRH stimulates the pituitary to produce TSH which stimulates synthesis and secretion of T 4 and T 3 . – creating a negative regulatory feedback loop. Aim: To create a model to describe how TPO inhibition would impact on the position of homeostasis using data from toxicity studies in dogs. Sandra Visser | 1 December 2011 Innovative Medicines | Non-Clinical Modeling and Simulation | MBDDx

Model development AZD-1 1-month and 6-month dog safety study Study Dose groups (M/F) (µmol/kg) PK and hormone sampling (day) 1 Month (28d) 0, 15, 90, 700 Day: -5, 1, 4, 8 ,14, 28 0, 700 Day: 31 35, 42, 57 6 Month (27w) 0, 15, 60, 250 Week: -2, -1, 7, 13, 27 0, 250 Week: 28, 31, 40 Simultaneous fitting of plasma levels of T 4 , T 3 & TSH to characterize onset, intensity and return to baseline (including rebound) T4 T3 TSH 8 Sandra Visser | 1 December 2011 Innovative Medicines | Non-Clinical Modeling and Simulation | MBDDx

PKPD model of thyroid hormone regulation drug and system parameters Kin TSH K TSH preTSH preTSH preTSH preTSH preTSH TSH -/+ -/+ nn NF1 NF2 TSH T4 BL T4 T4 TSH BL T4 BL -/+ I max *C Inhib - 1 - K in T4 K T4 *(1-fr) IC 50 +C Inhib T 4 K T4 *fr thyroglobulin T 3 K in T3 K T3 ’Drug specific’ In vitro potency ’System-specific’ Inter-species differences Sandra Visser | 1 December 2011 Innovative Medicines | Non-Clinical Modeling and Simulation | MBDDx 9

Validation & Interspecies Translation Validation: - Successful prediction of T4, T3 and TSH profiles in dogs after 1-month dosing of AZD-2 based on in-vitro IC 50 for TPO inhibition and PK profile and model system parameter estimates from AZD-1 analysis. Cross-species translation: - Adjust rate-constants for known species differences in hormone half-lives & adjust in vivo IC50 for measured species differences in in vitro IC50 assays Species T 4 , half-life T 3 , half- % T 3 derived life from peripheral conversion of T 4 [1] Eisenberg et al., 2010: Thyroid, 20: 1215-1228 7 days [1] 1 day [1] 72 [2] [2] Nicoloff et al., 1972: J. Clin. Investigations, 51: 473-483 Man [3] Bianchi et al., 1983: J. Clin. Endocrinology, 56: 1152-1163 [4] Taroura et al., 1991: Fd Chem. Tox., 29: 595-599 21 hrs [3] 6 hrs [3] 65 [4] Rat [5] Kinlaw et al., 1985: J. Clin. Investigations, 75: 1238-1241 [6] Maddison, J.E. & Page S.W., ‘Small Animal Clinical Pharmacology; p499 14-16 hrs [5] 5-6 hrs [5] 37 [6] Dog 10 Sandra Visser | 1 December 2011 Innovative Medicines | Non-Clinical Modeling and Simulation | MBDDx

Prediction of rat hormone levels Revising safety screening cascade RAT Interspecies extrapolation – Accurate prediction of thyroid hormone levels in 1-month rat safety study - both in terms of absolute levels and time to steady state in-vitro/in-vivo correlation – Cross-compound correlation established relating in-vivo IC 50 to in-vitro IC 50 based on series of 3-day rat safety studies – Prediction of thyroid hormone levels in rat based on in-vitro TPO inhibition data -> reduced the need for in vivo safety screening of new drug candidates 11 Sandra Visser | 1 December 2011 Innovative Medicines | Non-Clinical Modeling and Simulation | MBDDx

Translation to man Predicting T4 and TSH after 21 days MAD with AZD-1 TSH T4 Translation to man – Model predicts minor effects in human 21 day safety – consistent with small (non-significant) effects observed – Builds confidence in ability to extrapolate pre-clinical safety results 12 Sandra Visser | 1 December 2011 Innovative Medicines | Non-Clinical Modeling and Simulation | MBDDx

Concluding remarks Key learnings The proposed mechanism-based PKPD feedback model provides a scientific basis for the prediction of TPO inhibition mediated effects on plasma thyroid hormones levels in humans based on results obtained in vitro and animals studies .. Benefits to Discovery Pre-clinical safety studies can predict effects in man, which improves screening and selection of drug candidates. In vitro-vivo correlations reduce the in vivo safety screening needs (saving animals and $) Benefits to Development Predicted timescale of anticipated T 4 effects allowed MAD for AZD-2 to proceed as planned without costly time delays. The model is currently used guiding dose selection and study design for Phase 2 studies by prediction the safety profile of AZD-2. 13 Sandra Visser | 1 December 2011 Innovative Medicines | Non-Clinical Modeling and Simulation | MBDDx

Confidentiality Notice This file is private and may contain confidential and proprietary information. If you have received this file in error, please notify us and remove it from your system and note that you must not copy, distribute or take any action in reliance on it. Any unauthorized use or disclosure of the contents of this file is not permitted and may be unlawful. AstraZeneca PLC, 2 Kingdom Street, London, W2 6BD, UK, T: +44(0)20 7604 8000, F: +44 (0)20 7604 8151, www.astrazeneca.com 14 Sandra Visser | 1 December 2011 Innovative Medicines | Non-Clinical Modeling and Simulation | MBDDx