Lessons from the Development and Regulatory Evaluation of a Clinical

Lessons from the Development and Regulatory Evaluation of a Clinical Trial Simulation Tool for Alzheimer’s Disease Klaus Romero MD MS FCP Director of Clinical Pharmacology Critical Path Institute Why do drug development programs fail?

Reason by phase 100

80 Clinical safety Efficacy 70 Formulation 60 Market potential 50 PK/Bioavailability Strategic % Failures % 40 Resources 30 Toxicology 20 COGS 10 Unknown 0 Other Preclinical Phase I Phase II Phase III Registration Development stage

2 Clinical Trial Simulations

Evaluate different design options

Drug-Disease-Trial Simulated results Models 퐴퐷퐴푆 Trial optimization 푖푝푘 휃푖푝푘 = E 70 patient푝 through simulations 푇푝푘~Weibull 훼, ℎ푝푘

Statistical Analysis Trial Execution

•X dose •N •Frequency of observations Design selection •Inclusion/exclusion criteria

3 C-Path: A Public Private Partnership

• Act as a trusted, neutral third party • Convene scientific consortia of industry, academia, and government for pre-competitive sharing of data/expertise The best science The broadest experience Active consensus building Shared risk and costs Industry

• Enable iterative EMA/FDA/PMDA participation in developing new methods to assess the safety and efficacy of medical products

• Official regulatory endorsement of novel methodologies and drug development tools

4 C-Path Consortia

Eight global consortia collaborating with 1,300+ scientists and 61 companies Coalition Against Major Diseases Focusing on diseases of the brain Critical Path to TB Drug Regimens Testing tuberculosis drug combinations Multiple Sclerosis Outcome Assessments Consortium  Biomarkers Measuring drug effectiveness in MS  Clinical Outcome Polycystic Kidney Disease Consortium Assessment New imaging biomarkers Instruments Patient-Reported Outcome Consortium  Clinical Trial Measuring drug effectiveness Simulation Electronic Patient-Reported Outcome Consortium Tools Electronic capture of drug effectiveness  Data Standards  In Vitro Tools Predictive Safety Testing Consortium Drug safety Coalition For Accelerating Standards and Therapies Data standard

5 C-Path Collaborators

Industry • Abbvie • ERT • Mitsubishi Tanabe Pharmaceutical • Acorda Therapeutics • Exco InTouch Commercialization, Inc. • Actelion Pharmaceuticals • Forest Laboratories, Inc. • Pharsight/Certara • Allergan • Fujirebio Diagnostics • Tanabe Pharma • Almac • GE Healthcare • Novo Nordisk • • Amgen • Orion Corporation Genentech • Oracle • AstraZeneca • Genzyme • • Otsuka Pharmaceutical Biogen Idec • • Boehringer Ingelheim GlaxoSmithKline • Pfizer • Bracket • Hoffmann-La Roche, Inc. • PMDA Pharmaceuticals • Bristol-Myers Squibb • Horizon Pharma • PHT • Celgene • ICON • Sanofi • Cepheid • Ironwood Pharmaceuticals • STC • CRF Health • Johnson & Johnson Pharmaceutical • Shire • CROnos Research & Development, LLC • Sunovion Pharmaceuticals • Daiichi Sanyko • Medical Care Corporation • TAG • • Edetek • Takeda Medidata Solutions • Teva Pharmaceuticals • Eisai • Meso Scale Discovery • • UCB Eli Lilly and Company • • EMD Serono Merck and Co., Inc. • Vertex • Ephibian • Millennium: The Takeda Oncology Company

Nonprofit Research Organization Government and Regulatory Agencies Government and Regulatory Agencies • Alzheimerís Association • Center for Disease Control • University of Arizona • Alzheimerís Foundation of America • European Medicines Agency • Arizona State University • Bill & Melinda Gates Foundation • Innovative Medicines Initiative • Baylor University • CDISC • International Genomics Consortium • University of California San Francisco • Engelberg Center for Health Care Reform • National Institute of Allergy and • University of Colorado-Denver • EDCTP Infectious Diseases • Emory University • Flinn Foundation • National Institute of Diabetes and • University of Florida • Foundation for National Institutes of Health Digestive and Kidney Diseases • Johns Hopkins • National MS Society • National Institutes of Health • Mayo Clinic • Parkinsonís UK • National Institute of Neurological • University of Texas Southwestern • PKD Foundation Disorders and Stroke • Tufts University • Reagan-Udall Foundation • U.S. Food and Drug Administration • Science Foundation Arizona • World Health Organization • SRI International • Stop TB Partnership • TB Alliance

6 Context of Use Summary

•What the tool is: • A clinical trial simulation tool to help optimize clinical trial design for mild and moderate AD, using ADAS-cog as the primary cognitive endpoint •What it is based on: • A drug-disease-trial model that describes disease progression, drug effects, dropout rates, placebo effect, and relevant sources of variability •What it is NOT intended for: • Approve medical products without the actual execution of well conducted trials in real patients

7 Maximizing data to develop quantitative tools

Romero K, et al. Striving for an integrated drug development process for neurodegeneration: The Coalition Against Major Diseases. Neurodegen Dis Manage 2011;1(5): 379-85. 8 CAMD database

Five companies: - 9 studies - 3179 Patients Additional Data: - ADNI - Literature

https://codr.c-path.org/main/login.html

9 3179 patients with mild and moderate AD dementia

Ito K, et al. Understanding Placebo Responses in Alzheimer's Disease Clinical Trials from the Literature Meta-Data and CAMD Database. J Alzheimers Dis. 2013 Jun 26. [Epub ahead of print]

10 Relevant variables

• Longitudinal cognitive instrument: - ADAS-Cog: 11items, 0-70points

• Basal cognitive instrument: - MMSE: 8items, 30-0points

• Demographics: - Baseline age and gender

• Genetics: - APOE4 allele

11 Data integration for model development

12 Rogers JA, et al. Alzheimer's disease modeling and simulation. J Pharmacokinet Pharmacodyn. 2012 Oct;39(5):479-98. 13 Disease progression: 75year-old men, by APOE4 and baseline severity

14 Some designs that can be simulated

ParallelCrossover

15 Balancing power, sample size and duration, given varying effect magnitudes

Crossover Parallel 91Week Crossover

Versus

78 Week Parallel

By effect magnitude

16 Evolving dropout likelihood by baseline age and severity

Rogers JA, et al. Alzheimer's disease modeling and simulation. J Pharmacokinet Pharmacodyn. 2012 Oct;39(5):479-98. 17 External validation: The model adequately describes independent data

Study 1014:

n 639 Age (years) 50-97 Females 359 (56%) Males 280 (44%) Follow-up range 479-700 (days)

Rogers JA, et al. Alzheimer's disease modeling and simulation. J Pharmacokinet Pharmacodyn. 2012 Oct;39(5):479-98. 18 Understanding completed trials

literature placebo Real Results model prediction-placebo model prediction-donepezil

10 Drug Y-placebo Drug Y-treatment

placebo worse

Drug Y

ADAS-cog (change from baseline) ADAS-cogfrom (change

0 12 26 39 52

WEEK Regulatory conclusions

This model adequately captures relevant information regarding disease progression, drug effects and clinical trial aspects (placebo effect and dropouts)

Clinical Trial Simulations based on this tool allows the objective, prospective and realistic evaluation of the operating characteristics of different trial designs.

FDA fit-for-purpose decision on CAMD CTS tool. 2013 EMA qualification opinion on CAMD CTS tool. 2013

20 AD Drug Disease Trial Model: The regulatory path

The total journey took 1317 days (3 years, 7 months and 9 days).

• On June 12, 2013 the FDA • On September 19, 2013 the EMA determined the CTS tool was determined the CTS tool was “Fit for Purpose.” “Qualified for Use.”

21 Main strategic lessons learned

• Use a consortium approach • Provide clear context of use • Establish partner relationship with regulators early in process: • Do not rush to submit a letter of intent, wait until there is clarity in position especially around the “context of use” • Think about model support, enhancements, support infrastructure, etc. • Role for organizations such as ISCTM/academia • User communities

22 - Ten companies - 24 studies - 9500 Patients - Accessible to >200 researchers

https://codr.c-path.org/main/login.html

23 Thank You www.c-path.org