The Impact Of Biomarkers On The Diagnosis Of Alzheimer’s

Disease

John Q. Trojanowski, M.D., Ph.D.

NIA Alzheimer’s Disease Core Center, NINDS Udall Center of Excellence For Parkinson’s Disease Research, Center for

Neurodegenerative Disease Research, Marin S. Ware Alzheimer Program, Institute on Aging, Department of Pathology and Laboratory

Medicine, University of Pennsylvania, Philadelphia, PA

Global Population Forecast: The Silver Tsunami & The Rising Tide (Epidemic) Of AD

AD is the most common dementia with ~5 M patients now & ~13 M by 2025, but a treatment now that will slow disease by ~5 years will decrease AD prevalence and cost ~50% by 2050.

Aging Related Neurodegenerative Diseases Characterized by Filamentous

Aggregates of Misfolded Proteins Disease Lesions Components

Alzheimer’s Disease (A multi-proteinopathy)

SPs (100%) NFTs (100%) LBs (50%)

TDP-43 (50%)

Aβ Tau

α-Synuclein TDP-43

Frontotemporal Lobar Degeneration Inclusions Tau, TDP-43, FUS

Amyotrophic Lateral Sclerosis Inclusions TDP-43, FUS, Tau

Dementia with Lewy Bodies LBs a-Synuclein

Multiple System Atrophy GCIs a-Synuclein

Prion diseases SPs Prions

Trinucleotide repeat diseases Inclusions Expanded polyglutamine repeats

Mechanisms and Consequences of Protein Misfolding

and Brain Amyloidosis in Neurodegenerative Diseases

(Forman M, Trojanowski, JQ, Lee VM-Y. Nat Med, 2004)

Hypothetical Timeline for the Onset and Progression of AD Neurodegeneration and Dementia: The Need For AD

Biomarkers Is Urgent

Shaw LM, Korecka M, Clark CM, Lee VM.-Y, Trojanowski JQ. Nat Rev Drug Discovery, 6:295-303, 2007.

Targets for Aβ and Tau Focused AD Drug Discovery

Shaw LM, Korecka M, Clark CM, Lee VM.-Y, Trojanowski JQ. Nat Rev Drug Discovery, 6:295-303, 2007.

M. Weiner, P. Aisen, R Peterson, C. Jack, W. Jagust, J Trojanowski, L. Shaw, A. Toga, L. Beckett, D. Harvey, C Mathis,

A. Gamst. R. Green. A. Saykin, S. Potkin, J Morris, L Thal (D) Neil Buckholz, David Lee, Holly Soares

Industry Scientific Advisory Board (ISAB) and Site PIs, Study Coordinators, and

821 subjects enrolled in 58 sites in US and Canada

FUNDED BY THE NATIONAL INSTITUTE ON AGING

ADNI-1 Governance Advisory NIA Staff Steering

Committee

PI: Weiner, UCSF Administrative Core

Executive Committee

Biomarkers Core

Trojanowski and Shaw,

Penn

Biostatistics Core

Beckett, UCD

Coordinating Center (CC) Aisen, UCSD

Peterson, Mayo

Clinical Sites: Clinical, MRI, PET,

Biomarkers

Informatics Core

Toga, UCLA

Neuroimaging Center (NC)

MRI Jack, Mayo

PET Jagust, Berkeley

ADNI Industry Scientific Advisory Board

PIB/PET Supplement : Alzheimer’s Association and GE Healthcare Cerebrospinal Fluid Extension: Alzheimer’s Association, AstraZeneca, Cure

Alzheimer’s Fund, Merck, Pfizer and an anonymous foundation Genome-Wide Genotyping :Gene Network Sciences, Merck, Pfizer and an

anonymous foundation Genome-Wide Genotyping Genetic Analysis: NIBIB, Merck, Pfizer and an

anonymous foundation

New members Abbott, Genentech, Roche, Bayer

GOALS OF ADNI-1 ($40 M From NIH, $25 M From ISAB, Foundations & FNIH; Funded

From 10/1/2004 To 9/30/2009 With 1 Year No Cost Extension To 9/30/2010)

•  Optimize and standardize biomarkers for clinical trials

•  Validate biomarkers as measures of change •  Validate biomarkers as diagnostics or

predictors •  Establish world-wide network for clinical AD

studies and treatment trials

ADNI-1: Naturalistic study of AD progression •  200 NORMAL 3 yrs •  400 MCI 3 yrs •  200 AD 2 yrs •  Visits every 6 mo

•  57 sites •  Clinical, blood, LP •  Cognitive Tests •  1.5T MRI

Some also have •  3.0T MRI (25%) •  FDG-PET (50%) •  PiB-PET (approx 100)

All data in public database: UCLA/LONI/ADNI No embargo of data

NIA GRAND OPPORTUNITES (GO) ARRA GRANT

($24 M From NIH; Funded From 9/30/2009 to 9/30/2011) •  Adds cohort of 200 very mild “early”

MCI, (EMCI) •  LPs on 100% of new subjects •  Follow ADNI-1 controls/MCI additional yr •  F18 amyloid imaging on ALL existing

and new ADNI/GO subjects (AV-45) •  Complete analysis of all ADNI data

SCOPE OF ADNI-2 ($40 M From NIH & $29 M From ISAB, Foundations & FNIH; Funded

From 10/1/2010 To 9/30/2015) •  Goal to continue to follow >400 controls and

MCI from ADNI-1 for 5 more years and enroll: –  100 additional EMCI (supplements 200 from GO) –  150 new controls, LMCI, and AD

•  MRI at 3,6, months and annually •  F18 amyloid (AV-45)/FDG baseline and

Yr 2 •  LP on 100% of subjects at enrollment •  Genetics

NA-ADNI

J-ADNI E-ADNI

WW-ADNI

A-ADNI

C-ADNI K-ADNI

T-ADNI

Arg-ADNI

Future ADNI sites

Criteria for Ideal Biochemical Biomarker for Early Detection of AD

•  Detect fundamental feature of neuropathology & validation in neuropathology-confirmed cases

•  Diagnostic sensitivity for detection of AD >80% & specificity >80% for distinguishing from other dementias

•  Reliable, reproducible test •  Non-invasive/minimally invasive •  Simple to perform •  Reasonable cost

Frank, RA et al, 2003; NIA Reagan Institute, et al, 1998

Diagnostic Performance of Laboratory Tests for AD

Sensitivity Proportion of patients with AD who are correctly identified by the laboratory test

Specificity Proportion of patients without AD who are correctly identified by the laboratory test

Prior probability Frequency of AD in a particular group of patients

Positive predictive value

Proportion of patients with a positive test who are correctly diagnosed as AD

Negative predictive value

Percentage of people with a negative test who are negative for AD at autopsy

Since there is significant variation in CSF biomarker levels between studies, there is an urgent need to standardize and validate AD biomarkers

Study design: All studies on CSF T-tau with >25 AD cases

Innogenetics T-tau ELISA

34 studies, 2600 AD cases

Comparison of: mean level of CSF T-tau

→ Need for standardization: CSF sampling / handling procedures

Laboratory procedures

External control program

CSF

T-ta

u p

g/m

L

0

100

200

300

400

500

600

700

800

900

1000

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34

336 pg/mL 919 pg/mL

Study no.

Blennow K, NYAS 2006

CSF through year 4 Plasma through year 4

N 948 2621

Average time (mins) 35.11 67.94

95% CI (mins) 32.46-37.78 66.32-69.57

•  After overnight fast •  Collect into polypropylene

tube •  Transfer to polypropylene

transfer tube •  No centrifugation •  Freeze at site, thaw & aliquot

at UPenn, storage at -80 0C

Number of biofluids collected as of 6/30/2010: 13,122 Number of aliquots in biofluid bank: 126,681

Pre-Analytical Issues are Critical: CSF & Plasma Collections For ADNI

Microspheres coupled with antibody

Each type of microsphere coded with fluorochromes

Up to 100 proteins can be analyzed at once Sample volume = 75 µL.

Microspheres pass 2 lasers:

a)  Identification

b)   Quantification

Olsson A, et al, Clin Chem 2005

Luminex xMAP Technology for Multiparameter Immunoassays

Analytical Issues Are Critical: ADNI CSF Biochemical Biomarker Inter-laboratory Study

•  University of Pennsylvania: Leslie M Shaw, John Trojanowski, Virginia M-Y Lee, Margaret Knapik-Czajka

•  Innogenetics: Hugo Vanderstichele •  Sahlgrenska University Hospital: Kaj Blennow •  Friedrich-Alexander-Universitat Erlangen-Nurnberg: Jens

Wiltfang, Piotr Lewczuk •  Pfizer Global Research & Development: Holly Soares,

Nancy Raha •  Eli Lilly & Company: Robert A Dean, Eric Siemers, Richard

Lachno, Brent Salfen, (Linco) •  Merck Research Laboratories: Adam Simon

Participating Centers & Investigators

CSF Biomarker Validation (All Data Are on ADNI Website) •  Calibration curve stability •  Aliquot reproducibility •  Short- & long-term within- and between-

day reproducibility •  Stability of biomarkers in CSF

– Freeze-thaw – Room temp –  +40C

• 

Within-Center Reproducibility and Between Center

Repeatability for CSF Ab1-42

CSF pool 00012 CSF pool 00040

Day to Day Variability of CSF Pools

Test Retest Sample Performance

Tau Aβ1-42 p-Tau181p

Tau/Aβ1-42

p-tau181p/Aβ1-42

LR TAA

ROC AUC 0.831 0.913 0.753 0.917 0.856 0.938

Threshold values

93 ng/mL

192 ng/mL

23 ng/mL 0.39 0.10 0.22

Sensitivity (%) 69.6 96.4 67.9 85.7 91.1 100

Specificity (%) 92.3 76.9 73.1 84.6 71.2 76.9

Test accuracy (%)

80.6 87.0 73.1 85.2 81.5 88.9

Positive predictive value (%)

90.7 81.8 67.9 85.7 77.3 82.4

Negative predictive value (%)

73.8 95.2 70.4 84.6 88.1 100

CSF Biomarker Cutpoints Established Using CSFs Collected from ADNI-Independent Autopsy-Based AD

and Age-Matched Cognitively Normal Subjects

CSF Aβ1-42 Is Most Informative Single AD Biomarker & LRTAA Best Delineates Mild AD from MCI and NC in ADNI Cohort

Non-ADNI AD cases vs Non-ADNI NC subjects

ADNI Probable AD vs NC subjects

Survival Analyses for ADNI MCI Subjects: Progression to AD for BASELINE CSF Biomarkers > or < Cut Points

Aβ42<192 pg/mL

t-tau/Aβ4 >0.39

riskTAA2 > 0.34

As of June 28, 2010

28

Analysis of the ADNI CSF Data Set By Unsupervised Mixture Modeling

De Meyer G, Shapiro F, Vanderstichele H, Vanmechelen E, Engleborghs B, De Deyn P-P, Coart E, Hanson O, Minthon L, Zetterberg H, Blennow K, Shaw LM, Trojanowski Q.,

The ADNI. Diagnosis-independent Alzheimer’s disease biomarker signature in cognitively normal elderly people. Arch. Neurol., 67:949-956, 2010.

•  Unsupervised mixture modeling method: – Assumes that the biomarker data are

obtained from 2 populations – US-ADNI dataset was modeled, without

using the group labels – Start with single biomarker models and

proceed to multiple biomarker models

29

Analysis of the ADNI CSF Data Set By Unsupervised Mixture Modeling De Meyer G, Shapiro F, Vanderstichele H, Vanmechelen E, Engleborghs B, De Deyn P-P, Coart E, Hanson O, Minthon L,

Zetterberg H, Blennow K, Shaw LM, Trojanowski Q., The ADNI. Diagnosis-independent Alzheimer’s disease biomarker signature in cognitively normal elderly people. Arch. Neurol., 67:949-956, 2010.

•  Mixture model classification based on Aβ1-42 and P-Tau181P gave best fit

•  Break down of the signatures obtained by mixture modelling over the clinical diagnosis groups: –  AD signature (in red):

•  36% of the normal subjects, 72% of the MCIs and 90% of the AD patients

Normal MCI AD

Aβ1-42 level (pg/ml)

P-T

au18

1 lev

el (p

g/m

l)

Results: Mixture Model for Multiple Biomarkers (2)

•  Validation of the multi-biomarker mixture model on 2 datasets:

P-T

au18

1 le

vel (

pg/m

l)

Aβ1-42 level (pg/ml)

P-T

au18

1 le

vel (

pg/m

l)

1.  Data Hansson et al. Lancet Neurol, 2006: 57 of 57 subjects with MCI progressing to AD had an AD-signature (sensitivity=100%)

2.  Data Engelborghs et al. Neurobiol Aging, 2008: Sensitivity on autopsy confirmed cases: 63 of 68 (94%)

ADNI Data Integration Is Off And Running

OBJECTIVES: Investigate effect of CSF abnormalities on rate of functional decline in NC, MCI, and mild AD.

DESIGN: T-tau, p-tau181, and Aβ42 assayed in CSF from ADNI participants. Random effects regressions to examine the relationship between CSF abnormalities, cognitive impairment (ADAS-Cog), and functional decline (Pfeffer’s FAQ);.

SETTING: ADNI. PARTICIPANTS: 114 NC, 195 MCI, 100 mild AD. OUTCOME MEASURE: Decline in Pfeffer’s FAQ.

RESULTS: Across all groups, persons with a combination of tau and Aβ42 abnormalities exhibited the steepest rate of functional decline.

CONCLUSIONS: CSF abnormalities are associated with functional decline, and the development of AD in NC and MCI subjects, and those persons with tau and Aβ42 abnormalities are at greatest risk of functional impairment.

Arch Neurol, 67:688-696, 2010

Temporal Ordering of AD Biomarkers Reflects Disease Progression

Shaw et al., 2007; Jack et al., 2010; Trojanowski JQ, et al, 2010

OTHER INITITATIVES – Michael J. Fox Foundation Parkinson’s Progression Marker Initiative (PPMI)

PPMI: Identify tools to inform PD clinical trial design and decisions

PPMI

Dataset/ sample

collection

Standardized protocols

Biomarker verification

studies

Identify progression

markers

PPMI comprises 4 core objectives

1.  Develop/collect comprehensive clinical/imaging dataset and biological samples, which is made available

2.  Establish standardized protocols for acquisition, transfer and analysis of clinical, imaging and biologic data

3.  Conduct preliminary verification and validation studies on imaging and biologic markers

4.  Identify and correlate clinical, imaging and biologic markers for use in future trials.

PPMI SC and Study Cores Steering

Committee

PI-K Marek, A Siderowf, C Scherzer, D Jennings, K Kieburtz, W Poewe, B Mollenhauer, C Tanner, B Ravina (core leaders, MJFF, ISAB)

Clinical Coordination Core

University of Rochester’s Clinical Trials Coordination Center PI: Bernard Ravina

Imaging Core Institute for Neurodegenerative Disorders PI: John Seibyl

Statistics Core University of Iowa PI: Chris Coffey

Bioinformatics Core

Laboratory of Neuroimaging (LONI) at UCLA PI: Arthur Toga

BioRepository Coriell/BioRep PI: Alison Ansbach, Pasquale De Blasio, Michele Piovella

Bioanalytics Core University of Pennsylvania PI: John Trojanowski, Les Shaw

Genetics Core National Institute on Aging/NIH PI: Andy Singleton

PPMI CLINICAL SITES

THE SEARCH FOR NEW AD BIOMARKERS: THE PFIZER-PENN ALLIANCE RULES BASED MEDICINE INITIATIVE

WT Hu, A Chen-Plotkin, SE Arnold , M Grossman, CM Clark, LM Shaw, E Pickering, M Kuhn, Y Chen, L McCluskey, L Elman, J Karlawish, HI Hurtig , A Siderowf, VM-Y Lee, H Soares, JQ Trojanowski.

Novel CSF Biomarkers for Alzheimer’s Disease and Mild Cognitive Impairment Acta Neuropath, 119:669-678, 1010.

WT Hu, A Chen-Plotkin, SE Arnold , M Grossman, CM Clark, LM Shaw, E Pickering, M Kuhn, Y Chen, L McCluskey, L Elman, J Karlawish, HI Hurtig , A Siderowf, VM-Y Lee, H Soares, JQ Trojanowski. Novel CSF Biomarkers for Alzheimer’s Disease and Mild Cognitive Impairment Acta Neuropath, 119:669-678, 1010.

Random Forests PAM

Traditional AD biomarkers

Traditional + MAP

biomarkers Traditional AD

biomarkers

Traditional + MAP

biomarkers

Sensitivity (%) 88.6 92.4 97.0 97.0

Specificity (%) 86.2 97.0 66.7 87.9

Accuracy (%) 87.9 93.9 86.9 93.9

Improved AD Diagnostic Accuracy

IMPACT OF ADNI (>150 Published And In Press Papers)

•  Provide new information on AD pathophysiology

•  Develop early detection methods –  Identify those at risk for future AD

•  Develop improved treatment trials – ADNI data indicate that biomarkers

increases statistical power over traditional cognitive measures, aid in subject selection, and could reduce clinical trial sample size

•  Lead to AD treatment/prevention

“A Scary Idea: Pre-emptive Brain Scans For Alzheimer's” FORBES, 7/15/2010 - Robert Langreth is a senior editor at Forbes, in

charge of health care coverage

“You feel fine and have no symptoms, but your brain is slowly rotting away. And there is nothing we can do about it. Have a nice day.”

Integrating AD Diagnosis and Therapy/Prevention Societal Cost: Prevalence: 5 M and rising rapidly; $150B- one of the largest single health care burdens

Genomic •  ApoE allele assessment: homozygous 4/4 increase risk

•  APP/PS1: familial AD risk

Symptomology •  Patient/family complaints about “forgetfulness”

Biomarkers •  Aβ42/40 ratio in CSF

•  CSF phosphotau

•  Deviation from normal suggest pathology of an amyloid type or neurodegeneration

Pet/MR Imaging •  Aggregated Aβ42 specific PET measure on PET-CT

•  Hippocampal localization; extent and severity of disease confirmed

•  Confirm therapy effectiveness

Therapy •  Beta secretase inhibitor (APP processing to Aβ)

Therapy Evaluation

Risk, Biomarker, Disease and Therapeutic Evaluation

•  Passive/active

immunization

Biomarkers Will Accelerate Efforts To Treat And Delay The Onset of AD

Delay (years)

WE AS A NATION HAVE THE PEOPLE, IDEAS AND CAPACITY TO ACCLERATE THE PACE OF FINDING CURES FOR AD, PD, ALS, FTLD AND OTHER AGING RELATED

NEURODEGENERATIVE DISORDERS NOW BUT IT IS A MATTER OF RESOURCES. THUS, WE NEED A PUBLIC-PRIVATE COMPACT TO CURE THESE DISEASES ALONG THE LINES OF

SIMILAR PAST INITIATIVES LIKE BUILDING THE PANAMA CANAL, LAUNCHING THE SPACE PROGRAM AND MANY OTHERS

So where do we go from here?

•  The US spends $53 Billion/year on anti-aging balms, salves, lotions, etc. with no proven efficacy.

•  The US spends $2.6 Billion/year on Viagra and Cialis.

•  The US spends $2 Billion/year on popcorn.

•  French President Sarkozy recently unveiled a plan to spend $480 Million per year – or $558 per person – for 5 years to fight AD which afflicts 860,000 people in France.

•  And in the US, the NIH spends only $600 Million/year – or $120 per person - for research on AD which afflicts 5,000,000 US citizens. Thus, France spends 4 times more than the US for its citizens with AD while the amount spent on healthy aging is far less and difficult to estimate.

ADNI Biomarker Team:

Les Shaw, Virginia M-Y Lee, Chris Clark, Steve Arnold, Hugo Vanderstichele, Magdalena Korecka, Margaret Knapik-Czajka, Uwe Christians, Kaj Blennow Holly Soares, Eric Siemers, Piotr Lewczuk, William Potter & many more collaborators •  NIA: Drs. R Hodes & N. Buckholz •  Industry Scientific Advisory Board •  Foundation for NIH •  ADNI Core leaders and Core

members •  ADNI Site PIs •  Our volunteer subjects

Supported by the NIH/NIA/NINDS, Michael J. Fox Foundation, The Marian S. Ware Alzheimer Program, The William Maul Measey-Truman G. Schnabel, Jr. Chair of Geriatric Medicine & Gerontology and the Families of our Patients

Penn Biomarker Team: William T Hu , Alice Chen-Plotkin , Murray Grossman, Steve Arnold, Chris Clark , Les Shaw, Leo McCluskey, Laura Elman , Virginia M-Y, Lee,

It Takes a Great Team!