An integrated natural disease progression model of nine cognitive and biomarker endpoints in patients with Alzheimer s Disease Angelica Quartino 1* Dan Polhamus 2*, James Rogers 2, Jin Jin 1 212, Genentech *authors contributed equally to the work presented 1. Genentech Inc. South San Francisco, CA 2. Metrum Research Group, Tariffville, CT
Background Alzheimer s Disease a progressive neuro-degenerative disease 2 Mild to moderate AD Progression of biomarker abnormality Max Min Time Jack et al, Lancet Neurology 12:27 (213) CSF biomarkers of disease (amyloid): Aβ 42, Brain amyloid load: amyloid PET imaging CSF biomarkers of disease (neuro-degeneration): t-tau and p-tau Brain atrophy: volumetric MRI ( whole brain volume, hippocampal volume, ventricle size) Cortical activity: FDG-PET Cognitive and functional impairment scales: ADAS-Cog 12-item, CDR-SOB, MMSE Trouble remembering recent events to inability to preform basic tasks and full time care Death within ~ 9 years from diagnosis
Objectives 3 To establish a natural disease progression model integrating multiple biomarkers and endpoints in patients with mild Alzheimer s Disease 1. è An enhanced ability to identify and understand disease progression and impact of covariates and drug treatment effect, rather than within endpoints è Ability to simulate realistic multivariate longitudinal data, to allow assessment of studies with co-primary endpoints 1) Polhamus D et al. AAIC 213 CDR-SOB, vmri in MCI
Data and Methods Alzheimer s Disease Neuroimaging Initiative Database 4 Natural history non-treatment study in USA/Canada 298 mild Alzheimer s Disease subjects* Baseline MMSE 2-26 Baseline covariates e.g. age, gender Up to 3 years longitudinal changes in: ADAS-Cog 12-item score CDR each of 6 items score volumetric MRI (hippocampal, ventricles) Software: OpenBUGS v. 3.2.2 MMSE: Mini-Mental State Examination * Data in the analysis was from ADNI database extracted on 22 January 214 : www.loni.ucla.edu/adni
Results Integrated Longitudinal Alzheimer s Disease Model Prognostic Factors Endpoints Age and gender ApoE4 genotype Baseline MMSE CSF Aβ 42, ptau, ttau FDG-PET Amyloid PET Treatment Effect Underlying AD disease progression Cognition ADAScog12 Cognition/Function CDR 6-items scores Biomarkers Hippocampal volume Ventricular volume Underlying Disease Status 6 Underlying disease status 4 2 6 4 2 1 2 3 4 Time 1 2 3 4 Time Endpoint Score 8 6 Endpoint raw score 4 2 8 6 4 2 2 4 6 Underlying disease status 2 4 6 Underlying Disease Status Endpoint Score 8 6 Endpoint raw score 4 2 8 6 4 2 1 2 3 4 Time 1 2 3 4 Time
Results Correlation between predicted disease status and observed endpoints 6 8 8 ADAS-Cog12 score 1 1 CDR SOB score Observed endpoint score 6 ADAS12 absolute score 4 2 1 Hippocampus absolute score 6 4 2 1 8 8 6 6 4 4 2.. 2.. Hippocampus Underlying disease volume status 1 1 16 16 12 Ventricles absolute score 12 8 8 4 4 Ventricular volume 2.. 2.. Underlying disease status -2.. 2.. 2.. 2.. Underlying disease status -2.. 2.. Predicted underlying disease status Worse
Results The model accurately captures the central trend of observed data 7 Change score from baseline Change from baseline 2 1 1 2 7 1 ADAS-Cog12 ADAS12score Hippocampus volume 6 4 2 3 2 1 CDR CDRSUM SOB score Ventricular Ventricles volume 1 2 3 1 2 3 1 2 3 Year 1 2 3 Time (Years) observed data (red dots) and simulated data (black lines) with 9% credible interval (shaded area)
Results The model maintains observed correlation between endpoints 8 Hippocampus volume Ventricular volume ADAS-Cog12 score 1 1 7. 7... 2. 2.. 1.. 1. 7. 7... 2. 2............ Spearman's rho -... -... -... Spearman s rho CDR SOB score CDRSum ADAS12 Ventricles ADAS-Cog12 score Ventricular volume Observed correlation (red line) Simulated correlation (black distribution)
Results Impact of prognostic factors on underlying disease progression rate 9 Slower disease progression Faster disease progression Reference subject ApoE4 negative Baseline MMSE 23 Age 7 Interaction high baseline MMSE- ApoE4 positive High age 11 89 1 High baseline MMSE 1 ApoE4 positive 18 82.1... Posterior estimated effect of normalized prognostic factors relative to reference subject (with uncertainty) MMSE: Mini-Mental State Examination
Results Change in disease progression and endpoints for subpopulations 1 Score Value 12 8 4 12 8 4 Underlying disease progression 8 6 Raw score 4 8 6 4 ADAS-Cog12 score ADAS cog progression of the average individual 2 2 6. 2.. 7. 1. Year Hippocampus progression of the average individu 6 Hippocampus volume 1 1 16 1 1 16 CDR SOB score CDR sum progression of the average individual Ventricular volume. 2.. 7. 1. Year Ventricular progression of the average individu. 2.. 7. 1. Year. 2.. 7. 1. Time (Years) Raw score 4 4 12 12 8 8 worse Baseline MMSE 2 26 ApoE4 genotype Negative Positive 3 3. 2.. 7. 1. Year. 2.. 7. 1. 4 4 Time (Years). 2.. 7. 1. Year. 2.. 7. 1.
Application Projected impact of a hypothetical treatment effect across endpoints 11 DM factor Change score Change from baseline - -1 1 ADAS-Cog12 ADAS12 score -1 Change from baseline -2-3 -4 1 2 3 4 CDR CDRsum SOB score % reduction ADNI AD progression % Change score -1 Percent change from baseline 1-2. 2. -.. -7.. 7... 1. 1. 2. Years Hippocampus volume Hippocampus - Percent change from baseline -1-2 1 2.. 1. 1. 2. Ventricles Years Ventricular volume -1. 1... 1. 1. 2. Years.. 1. 1. 2. -3 3 Time (Years).. 1. 1. 2. Years.. 1. 1. 2. Observed unadjusted mean of ADNI data (dots) Model predictions of a hypothetical treatment effect using baseline covariates (lines and shaded area)
Summary 12 Longitudinal PKPD model for Alzheimer s Disease Alzheimer s Disease Model Placebo Model Drug Effect Model Model properties Application Greater insights of disease progression, impact of patient covariates and drug treatment effect Comparison of novel-treatment outcomes and placebo response to historical data across endpoints Allows translation of information across endpoints and biomarkers Joint analysis of multiple endpoints (e.g. co-primary endpoints) Ability to assess the sensitivity of the different endpoints in subpopulations Trial design optimization for multiple endpoints Ability to simulate realistic multivariate longitudinal data
Acknowledgements Co-authors: Dan, Jim and Jin Colleagues and team members at Genentech Alzheimer s Disease Neuroimaging Initiative