Integrating Molecular Toxicology Earlier in the Drug Development Process Jonathan Hitchcock Molecular Toxicology, Drug Safety R&D Pfizer Global Research & Development Sandwich, Kent UK
4.5 4.0 3.5 3.0 2.5 2.0 PPM Shift in paradigm. Structure ~omics Pathways Toxicity O N O N O Cl O Integrated approach to Toxicology 2
Outline of Presentation Introduction Safety Attrition Value of earlier toxicology integration Molecular Toxicology: Past Perspective Examples Future challenges and opportunities Concluding remarks/summary 3
Compound Attrition on the R&D Process 3,000,000 Compounds Screened Preclinical Pharmacology Toxicity accounts for approximately 60% of attrition Late stage failures are more costly Preclinical Safety Clinical Pharmacology & Safety Average cost per drug: $800M - $1.4 billion 1-2 Products Discovery Exploratory Development Full Development Phase I Phase II Phase III 0 5 10 15 Idea 11-15 Years Drug 4
Late-Stage Attrition is Costly Early stage attrition Safety, Gentox, PK, Tox, QT Screens to manage it Late stage attrition Idiosyncratic tox. Very expensive. Difficult to screen for Before Candidate Selection : focus on attrition After CS : focus on preservation 800 File Launch Cumulative costs $M 600 400 200 Phase I Phase II II Phase III 0 0 1 2 3 4 5 6 7 8 9 10 Years 5
Value of Earlier Integration Improved Predictive Value balanced against Timeliness Bulk consumption Cost Predictive Toxicology (Source: Abbott Laboratories) From Drug Discovery & Development July 2003 6
Outline of Presentation Introduction Safety Attrition Value of earlier toxicology integration Molecular Toxicology integration: Past Perspective Examples Future challenges and opportunities Concluding remarks/summary 7
Why Molecular Toxicology? 2000 13 years, $250 million >1100 scientists 8
Molecular Technology 1985 Discovery of PCR (Kary B Mullis) 1995 TaqMan TM (Livak KJ) SAGE (Velculescu VE) DNA microarray (Schena M) 1996 Molecular Beacons (Tyagi S) SSH (Diatchenko L) 1998 RNA interference (Fire A & Mello C) 1999 Affymetrix Genechip (Lipshutz RJ) Luminex Multi-Analyte System 2000 Protein microarray (Snyder M) 2004 Inducible RNAi 9
PCR amplification 1985 Discovery of PCR (Kary B Mullis) 1995 TaqMan TM (Livak KJ) 1996 Molecular Beacons (Tyagi S) Ability to amplify very small amounts of genetic material. Quantitative gene expression analysis of specific genes. Measure multiple genes in the same sample. 10
Gene Expression Microarrays 1995 SAGE (Velculescu VE) DNA microarray (Schena M) 1996 SSH (Diatchenko L) 1999 Affymetrix Genechip (Lipshutz RJ) Ability to measure thousands of genes in one go! Improved analysis and pathway software to put results into context. RNA amplification techniques ensure even smallest of samples can be analysed. 11
RNA Interference (RNAi) 1998 RNA interference (Fire A & Mello C) 2004 Inducible RNAi Ability to knockdown expression of a particular gene of interest. Increased application in Drug Discovery for the evaluation of Target safety. Inducible RNAi can be used when evaluating key cell survival pathways. Gene X / Beta Actin Ratio (normalised to control) 1.2 0.8 0.6 0.4 0.2 Gene X Knockdown in Cell Line Y: Comparison of Lipofectamine & sirna concentrations 1 0 Media Neg 25nM 50nM 75nM 100nM Concentration sirna 5ul LF 7.5ul LF 10ul LF 12
Deeper Knowledge of Targets and Pathways Molecular understanding of biological processes is continually developing. Potential target related effects can be evaluated before compound has been synthesised. Endpoints from multiple assays can be analysed in an integrated manner. 13
Current Status Where Molecular Toxicology fits Clinical Safety Safety Management Clinical Safety Safety BioMarkers HT HT Safety Safety Screens Targeted Toxicity Screening Toxicity Mechanism Assessment Issue Resolution Interspecies Comparisons and Decision-making Target Target Organ Organ Studies Interpretation & Relevance Drug Discovery PreClinical Testing Clinical Development FDA Phase IV 14
Outline of Presentation Introduction Safety Attrition Value of earlier toxicology integration Molecular Toxicology integration: Past Perspective Examples Future challenges and opportunities Concluding remarks/summary 15
4.5 4.0 3.5 3.0 2.5 2.0 PPM Shift in paradigm. Structure ~omics Pathways Toxicity O N O N O Cl O Integrated approach to Toxicology 16
Predictive Screens Biolum Ames Re-engineered Salmonella reverse mutation assay Genetically modified AMES Salmonella strains Bioluminescent detection of bacterial reversion Simple, Robust & amenable to automation Proprietary technology, Patent Application No. 60/258,073 filed on December 22, 2000. Assay principle Bioluminescent sensor for revertant cells Measurement of metabolic activities Eliminates problems associated with high concentrations of cells, or contamination. 17
Bioluminescent detection of revertants 1 - Image of bioluminescent colonies of revertants TA100lux treated with MMS revertant colonies/well 80 70 60 50 40 30 20 10 0 0.1 1 10 100 Concentration [ug/plate] 2- Number of revertant colonies detected (n=4wells, ±SD) 18
Biolum Ames: Key attributes Sensitive & Predictive High concordance with standard assay Sensitive, high concordance with NTP data (94-95%) low false positive rate Quick & Economical Low bulk requirement (5mg instead of up to 1g) Increased throughput currently 24 well plate (potential for < 48 well) 2 day assay instead of 3, 20 compounds/wk Automation capability Fully automated scoring Software produces report 19
Use of In Silico Tools Predict before the Experiment! DEREK stands for Deductive Estimation of Risk based on Existing Knowledge DEREK (Lhasa Ltd) is an expert knowledge base system Several rule bases, consisting of descriptions of molecular substructures (structural alerts) have been associated with toxic end points. 20
Use of In Silico Tools What can DEREK tell me? DEREK has only been validated for genotoxicity Rules do exist for other toxicology endpoints, currently under assessment: Skin & respiratory sensitisation Carcinogenicity Reproductive & Developmental Irritancy Thyroid toxicity Alerts should always be put in context and DEREK should not be used as a sole source for go/no go decisions for compounds 21
Genomic Biomarker Development Vasculitis Vasculitis is a major safety issue associated with a variety of drugs; basic mechanism of toxicity is unknown Vasculitis can only be confirmed histopathologically Mechanistic insights are of great value towards understanding relevance and clinical significance Specific biomarkers needed for clinical risk management DermAtlas; http://www.dermatlas.org 22
Genomic Biomarker Development Example: Vasculitis In vivo Study Design PDE 4 inhibitor 3 days, 2 doses Male Sprague-Dawley rats: 10 per treatment, 6 control animals Standard Endpoints: Histopathology, Clinical Pathology Additional Endpoints: Genomics: Expression profiling (Affymetrix) & Taqman Proteomics Metabonomics 23
Histopathology: Degree of Vasculitis Observed Doses Rat number Vasculitis * Control 40 mg/kg 80 mg/kg M2 - M5 - M6 - M102 - M104 - M107 - M109 1 M110 1 M204 3+ M206 3+ M207 3+ M209 3+ Incidence of vasculitis: 40mg/kg/day:6/10 80mg/kg/day:10/10 + indicates fibrinoid necrosis 24
Hierarchical Clustering (Affymetrix Data) Control Low dose High dose M5 M107 M104 M6 M102 M2 M110 M206 M207 M204 M209 M109 No Vasculitis Induced vasculitis 25
Fold induction of mrna expression in mesenteric tissue (Taqman) Fold induction 2,5 2 1,5 1 0,5 0 TNF-α Sprague D. rats Fold induction 220 200 180 160 140 120 100 80 60 40 20 0 40 mg/kg/day 80 mg/kg/day IL-6 Sprague D. rats 26
Correlation between IL-6 mrna Expression and Vasculitis for Male Sprague-Dawley Rats 11 IL-6 mrna expression in individual rat (arbitrary unit) 10 9 8 7 6 5 4 V+ V V V 40 μγ/κγ/δαψ 80 μγ/κγ/δαψ V No vasculitis detected Control rats Low dose Detected vasculitis V V+ V V+ V+ V+ V+ V+ High dose 27
What makes a Good Biomarker? Predictive and specific. How? Well validated using a large data set and traditional endpoints Accessible and Translatable Non-invasive (blood, urine sample), measurable in animal models and in the clinic. Highly sensitive. can we pick up changes before traditional endpoints? Profile of gene/protein changes often used as opposed to individual biomarkers. 28
Future Challenges and Opportunities Freedom to operate (Regulatory Acceptance) In November 2003 the FDA released voluntary submission guidelines to cover regulatory use of genomic-based data generated and submitted under an IND Genuinely encourages the application of molecular data. Presents illustrative examples that cover many real life questions. Biomarker definitions need expansion. Limited understanding within the agency of dealing with data sets of this kind. Opportunity to educate and influence the regulators FDA now actually performs own analysis of molecular (specifically genomic) data. 29
Illustrative Example Relating to Safety Vasculitis is a major drug-related nonclinical safety signal and the basic mechanism of toxicity is unknown normally confirmed by histopathology. A sponsor can use new rat gene chip microarray technologies for expression profiling of 8000 known sequence genes to investigate the mechanism of toxicity and possibly see a pattern of genetic biomarkers in treated rats that is different from controls. THESE ARE RESEARCH DATA: VOLUNTARY SUBMISSION ENCOURAGED 30
Future Challenges and Opportunities Challenges: Linking to other omics i.e. metabonomics, proteomics. Deeper knowledge of Targets and Pathways improve understanding of large data sets Demonstration of added value of Molecular Toxicology Integration within development process Opportunities: Pharmacogenomics Drugs and dosage chosen based on genetic composition personalised medicine Potential Benefits: minimal / no side effects Discovery of novel targets Improved understanding will lead to new targets Improved predictivity of screens should lead to safer medicines. 31
4.5 4.0 3.5 3.0 2.5 2.0 PPM In Summary Shift in paradigm. Structure ~omics Pathways Toxicity O N O N O Cl O Integrated approach to Toxicology 32
Acknowledgements Fiona Spence Lab Lead, Molecular Toxicology DSRD - Sandwich Michael Lawton Jiri Aubrecht 33