Serial Combined PIB PET and Serial Combined PIB PET and MRI in Alzheimer's DiseaseMRI in Alzheimer's Disease

ororBiomarkers of the Alzheimer's Biomarkers of the Alzheimer's

pathological cascade and clinical pathological cascade and clinical expressionexpressionClifford R. Jack, Jr., Clifford R. Jack, Jr.,

Dept Radiology Dept Radiology Mayo ClinicMayo Clinic

Rochester MNRochester MN

Diagnosis of ADDiagnosis of AD Current criteria - McKhann 1984Current criteria - McKhann 1984

Dementia Dementia Progressive worsening of memory and other cognitive Progressive worsening of memory and other cognitive

functions - abstract thinking, judgment, language, functions - abstract thinking, judgment, language, etc.etc.

No disturbance of consciousnessNo disturbance of consciousness Absence of other brain disease that could account for Absence of other brain disease that could account for

decline -- ie diagnosis of exclusiondecline -- ie diagnosis of exclusion Loss of functional independenceLoss of functional independence based solely on clinical groundsbased solely on clinical grounds

Two-stateTwo-state model: AD pathology and clinical model: AD pathology and clinical symptoms are synonymoussymptoms are synonymous

normal

Dementia

Relationship between AD pathology and dementia syndromeThe two state view of early 1980’s

No pat

hology

patholo

gyD Knopman

AD

Control

Imaging of “Two State” View

4

3

2

1

0

New knowledge since 1984New knowledge since 1984 Demographics – 1% 60s; 30% 80s; 50% 90sDemographics – 1% 60s; 30% 80s; 50% 90s Overlap with other pathologies – CVD, Lewy body DzOverlap with other pathologies – CVD, Lewy body Dz Onset of dementia is gradual Onset of dementia is gradual intermediate cognitive intermediate cognitive

state (MCI), not normal but not dementedstate (MCI), not normal but not demented AD syndrome vs AD pathologyAD syndrome vs AD pathology 30% of cognitively normal elderly have AD pathology at 30% of cognitively normal elderly have AD pathology at

autopsy – esp. amyloid plaquesautopsy – esp. amyloid plaques Genetics & molecular pathways – amyloid cascade Genetics & molecular pathways – amyloid cascade

hypothesishypothesis Different types of AD path have different relationships to Different types of AD path have different relationships to

symptoms – amyloid plaques vs NFT vs symptoms – amyloid plaques vs NFT vs neurodegenerationneurodegeneration

Biomarkers of AD path developed and validatedBiomarkers of AD path developed and validated Temporal order of AD pathology and biomarkersTemporal order of AD pathology and biomarkers

New knowledge since 1984New knowledge since 1984 Demographics – 1% 60s; 30% 80s; 50% 90sDemographics – 1% 60s; 30% 80s; 50% 90s Overlap with other pathologies – CVD, Lewy body DzOverlap with other pathologies – CVD, Lewy body Dz Onset of dementia is gradual Onset of dementia is gradual intermediate cognitive intermediate cognitive

state (MCI), not normal but not dementedstate (MCI), not normal but not demented AD syndrome vs AD pathologyAD syndrome vs AD pathology 30% of cognitively normal elderly have AD pathology at 30% of cognitively normal elderly have AD pathology at

autopsy – esp. amyloid plaquesautopsy – esp. amyloid plaques Genetics & molecular pathways – amyloid cascade Genetics & molecular pathways – amyloid cascade

hypothesishypothesis Different types of AD path have different relationships to Different types of AD path have different relationships to

symptoms – amyloid plaques vs NFT vs symptoms – amyloid plaques vs NFT vs neurodegenerationneurodegeneration

Biomarkers of AD path developed and validatedBiomarkers of AD path developed and validated Temporal order of AD pathology and biomarkersTemporal order of AD pathology and biomarkers

normal

Dementia

Relationship between AD pathology and dementia syndromeThe two state view of early 1980’s

No pat

hology

patholo

gyD Knopman

normal

Dementia

The concept of Alzheimer’s disease across the cognitive continuum 2010

Biomarker Biomarker & Clinical

Biomarker & Clinical

Mild Cognitive Impairment

D Knopman

ObjectivesObjectives

to describe, and provide evidence in support, of a dynamic biomarker based model of AD progression

To place the role of biomarkers within this context Clinical Therapeutic trials

AD Pathology: 4 categoriesAD Pathology: 4 categories Amyloid plaquesAmyloid plaques Neurofibrillary Neurofibrillary

tanglestangles InflammationInflammation NeurodegenerationNeurodegeneration

Loss, shrinkage of Loss, shrinkage of dendritic tree, dendritic tree, synapses, neuronssynapses, neurons

Clinical symptomsClinical symptoms* NFT and neurodegeneration are both neuronal processes and occur in same topographic distribution

Imaging & CSF Biomarkers; Imaging & CSF Biomarkers; 4 classes4 classes

Brain Amyloidosis PET - amyloid plaque imaging CSF AB 1-42

Neuronal dysfunction and tau mediated injury CSF t-tau and p-tau FDG PET Functional MRI (activation and resting

state) Neurodegeneration

Structural MRI MR Spectroscopy Diffusion MRI Perfusion MRI

Inflammation – PET & MRS

Biomarker Reviews

Hampel, Alzheimer’s Dement 2008

Shaw, Nat Rev Drug Discov 2007

11C PIB and Structural MRI Provide Complementary 11C PIB and Structural MRI Provide Complementary Information in Imaging of AD and Amnestic MCI.Information in Imaging of AD and Amnestic MCI. Brain Brain 2008;131(Pt 3):665-6802008;131(Pt 3):665-680

Serial PIB and MRI in normal, MCI, and AD: implications Serial PIB and MRI in normal, MCI, and AD: implications for sequence of pathological events in AD.for sequence of pathological events in AD. Brain 2009 Brain 2009 132(Pt 5):1355-65132(Pt 5):1355-65

Objective:Objective: understand temporal relationships amyloid, understand temporal relationships amyloid, neurodegeneration, cognitionneurodegeneration, cognition

11C PIB11C PIB biomarker of amyloid load biomarker of amyloid load structural MRIstructural MRI biomarker of stage of neurodegeneration biomarker of stage of neurodegeneration Mormino et. al. Mormino et. al. Brain 2009; 132(Pt 5):1310-23Brain 2009; 132(Pt 5):1310-23

Model of disease staging based on PIB & Model of disease staging based on PIB & MRI MRI Publications in 2008 and early 2009Publications in 2008 and early 2009

Cross sectional group-wise comparison global cortical PiB and hippocampal volume

Jack et al, Jack et al, BrainBrain 2008;131:665-680 2008;131:665-680

Cross sectional group-wise comparison global cortical PiB and hippocampal volume

Jack et alJack et al, BrainBrain 2008;131:665-680 2008;131:665-680

Annual change in global PIB ratio and ventricular volume by clinical diagnosis

Mayo plus ADNI data

Jack et alJack et al , Brain 2009 132 (Pt 5):1355-65

Summary: Data derived from imaging Summary: Data derived from imaging consistent with model of typical late onset consistent with model of typical late onset AD with 3 main featuresAD with 3 main features

significant plaque deposition occurs prior to neuro degeneration and clinical decline

Dissociation: Change in cognition is closely coupled to rate of neurodegenerative progression, not to rate of amyloid deposition

Bi-phasic disease process: amyloid dynamic early vs. neurodegeneration dynamic mid to late stage

Brain 2008;131(Pt 3):665-680, Brain 2008;131(Pt 3):665-680, and Brain 2009 132(Pt Brain 2009 132(Pt 5):1355-655):1355-65

Time

Pre-symptomic Prodomal(MCI)

Dementia

Amyloid (PiB)

Cognition

Neuron / synapse numbers(MRI)

Proposed model relating imaging (pathology) and clinical presentation over an individual’s adult lifetime.

Brain 2009 132(Pt 5):1355-65

Graphical model of the dynamic biomarkers of AD Graphical model of the dynamic biomarkers of AD pathological progressionpathological progression

Jack et al, Jack et al, Brain 2009 132 (Pt 5):1355-65Brain 2009 132 (Pt 5):1355-65

Time

Pre-symptomic Prodomal(MCI)

Dementia

Amyloid (PiB)

Cognition

Neuron / synapse numbers(MRI)

Proposed model relating imaging (pathology) and clinical presentation over an individual’s adult lifetime.

Brain 2009 132(Pt 5):1355-65

Graphical model of the dynamic biomarkers of AD Graphical model of the dynamic biomarkers of AD pathological progressionpathological progression

Jack et alJack et al , Brain 2009 132 (Pt 5):1355-65Brain 2009 132 (Pt 5):1355-65

Expand Model: Imaging & CSF Expand Model: Imaging & CSF Biomarkers; “big 5”Biomarkers; “big 5”

Brain Amyloidosis PET - amyloid plaque imaging CSF AB 1-42

Neuronal dysfunction and tau mediated injury CSF t-tau and p-tau FDG PET Functional MRI (activation and resting

state) Neurodegeneration

Structural MRI MR Spectroscopy Diffusion MRI Perfusion MRI

Inflammation – PET & MRS

Biomarker Reviews

Hampel, Alzheimer’s Dement 2008

Shaw, Nat Rev Drug Discov 2007

PIB and CSF PIB and CSF AB42AB42

roughly equivalent measures roughly equivalent measures of brain amyloidof brain amyloid

Jagust 2009

Fagan 2006

Forsberg

2008

Evidence tau changes in normal Evidence tau changes in normal subjects “destined to decline”: PIB and subjects “destined to decline”: PIB and

CSF Tau, normal elderlyCSF Tau, normal elderlyFagan et al EMBO Molecular Medicine 2009Fagan et al EMBO Molecular Medicine 2009

Evidence brain volume changes in Evidence brain volume changes in normal subjects “destined to decline” - normal subjects “destined to decline” -

CSF AB and brain volume in cognitively CSF AB and brain volume in cognitively normal elderly (CDR 0):normal elderly (CDR 0): Fagan et al Annals Fagan et al Annals

20092009

Cortical Thickness in PIB + vs – elderlyCortical Thickness in PIB + vs – elderly controls: controls: Dickerson et al Cereb Cortex Dickerson et al Cereb Cortex

20092009

Spearman partial rank-order correlations adjusted for age (P-value) between CDR-SB or MMSE and each MRI or CSF biomarker

Pair of measurementsAll

(n=399)CN

(n=109)aMCI (n=192)

AD (n=98)

CDR-SB vs.

STAND score 0.59 (<0.001) 0.01 (0.88) 0.26 (<0.001) 0.34 (<0.001)

Aβ1-42 -0.37 (<0.001) 0.10 (0.28) -0.10 (0.18) -0.002 (0.99)

t-tau 0.39 (<0.001) -0.14 (0.15) 0.12 (0.11) 0.02 (0.81)

p-tau181P 0.36 (<0.001) -0.23 (0.02) 0.09 (0.24) -0.10 (0.33)

t-tau/Aβ1-42 0.45 (<0.001) -0.16 (0.09) 0.13 (0.08) 0.007 (0.95)

MMSE vs.

STAND score -0.50 (<0.001) 0.02 (0.80) -0.19 (0.01) -0.29 (0.004)

Aβ1-42 0.31 (<0.001) -0.11 (0.24) 0.11 (0.13) 0.03 (0.77)

t-tau -0.32 (<0.001) 0.11 (0.24) -0.05 (0.46) -0.13 (0.22)

p-tau181P -0.30 (<0.001) 0.03 (0.78) -0.05 (0.47) -0.09 (0.39)

t-tau/Aβ1-42 -0.37 (<0.001) 0.11 (0.28) -0.09 (0.22) -0.04 (0.71)

Note: Bold font indicates P-values < 0.01

MRI correlates with cognitive impairment MRI correlates with cognitive impairment better than CSF better than CSF Vemuri et al Vemuri et al Neurology 2009Neurology 2009

Serial Biomarker Measures: MRI Serial Biomarker Measures: MRI dynamic late, CSF tau intermediate:dynamic late, CSF tau intermediate:

Vemuri Neurology 2010Vemuri Neurology 2010

Non-linearityNon-linearity: Rate of atrophy : Rate of atrophy accelerates as approach dementia:accelerates as approach dementia:

Chan et al, Lancet 2003Chan et al, Lancet 2003

temporal ordering of biomarkerstemporal ordering of biomarkers Amyloid imagingAmyloid imaging [Mintun, 2006; Aizenstein, 2008; Klunk 2004; Rowe 2007; [Mintun, 2006; Aizenstein, 2008; Klunk 2004; Rowe 2007;

Mormino 2009] ] CSF ACSF A4242 [Peskind, 2006; Shaw, 2009; Fagan, 2007; Li, 2007; Fagan 2009; [Peskind, 2006; Shaw, 2009; Fagan, 2007; Li, 2007; Fagan 2009; Vemuri Vemuri

20092009] ] CSF tauCSF tau [Bouwman 2007; de Leon [Bouwman 2007; de Leon 2006; 2006; Wahlund 2003; Stefani 2006; Sluimer Wahlund 2003; Stefani 2006; Sluimer

2008; Hansson 2006; 2008; Hansson 2006; Sunderland Sunderland 1999; Blennow 1999; Blennow 2003; Vemuri 2009]2003; Vemuri 2009] FDG PETFDG PET [Minoshima, 1997; Chetelat, 2002; de Leon, 2001; Reiman, 1996; Small [Minoshima, 1997; Chetelat, 2002; de Leon, 2001; Reiman, 1996; Small

1995] 1995] MRIMRI [Fox 1997; Fox 1999; Kaye, 1997; Killiany 2000; Dickerson 2009] [Fox 1997; Fox 1999; Kaye, 1997; Killiany 2000; Dickerson 2009]

ConclusionsConclusions Biomarker abnormalities precede clinical symptoms Amyloid biomarkers become abnormal first Little evidence for ordering of amyloid imaging vs CSF AB42 FGD PET changes before MRI [Reiman 1998] Little evidence for ordering of FDG PET vs CSF tau MRI last onset but correlates with clinical Sx longest [Vemuri, 2009] Non-linear functions (over long period) [Chan 2003; Carlson 2008]

Dynamic Biomarkers of the Alzheimer’s Pathological Cascade

Ab Amyloid = CSF Ab42 or amyloid PET imaging; Tau Mediated Neuron Injury and Dysfunction = CSF tau or FDG PET; Brain Structure = structural MRI

Jack et al, Lancet Neurol 2010; 9: 119-28

Biomarker-based disease modeling Provides a framework that relates temporal Provides a framework that relates temporal

changes in AD biomarkers with clinical disease changes in AD biomarkers with clinical disease stage and with each other stage and with each other

Some details of model will undoubtedly changeSome details of model will undoubtedly change However, certain principles will stand up However, certain principles will stand up

Biomarkers measure specific aspects of AD pathBiomarkers measure specific aspects of AD path Temporally ordered: amyloid => neuronal => Temporally ordered: amyloid => neuronal =>

cognitioncognition Temporal ordering: both onset and ceiling; not start-Temporal ordering: both onset and ceiling; not start-

stop but displaced in timestop but displaced in time Combination of biomarkers needed for Combination of biomarkers needed for

comprehensive stagingcomprehensive staging Non linear function of timeNon linear function of time

ObjectivesObjectives

to describe, and provide evidence in support, of a dynamic biomarker based model of AD progression

To place the role of biomarkers within this context Clinical diagnosis Therapeutic trials

Diagnostic role of biomarkers in Diagnostic role of biomarkers in typical late onset ADtypical late onset AD

Pre clinical – pathology present, but no (minimal) Pre clinical – pathology present, but no (minimal) symptoms - biomarkers are only indicatorsymptoms - biomarkers are only indicator Amyloid biomarker required, others supportiveAmyloid biomarker required, others supportive

Mild Cognitive ImpairmentMild Cognitive Impairment identify etiology of impairment identify etiology of impairment likelihood that patient will progress to AD in short likelihood that patient will progress to AD in short

interval (i.e. time to event)interval (i.e. time to event) Dementia (meets clinical criteria for AD Dementia (meets clinical criteria for AD

dementia)dementia) increase (or decrease) confidence that clinically increase (or decrease) confidence that clinically

determined dementia is due to AD pathologydetermined dementia is due to AD pathology

Implications of the pathological Implications of the pathological specificity of biomarkers for anti amyloid specificity of biomarkers for anti amyloid

trials in MCI/ADtrials in MCI/AD initiating molecular pathway amyloid dysmetabolism initiating molecular pathway amyloid dysmetabolism inclusioninclusion should be based on evidence of the presence of should be based on evidence of the presence of amyloid in the brain amyloid in the brain amyloid PET imaging or CSF A amyloid PET imaging or CSF A42 42

Target efficacyTarget efficacy, amyloid reduction , amyloid reduction amyloid biomarker amyloid biomarker (PET amyloid imaging) (PET amyloid imaging)

However, objective is to treat brain amyloidosis with However, objective is to treat brain amyloidosis with intent to affect cognition. Change in Aβ amyloid load over intent to affect cognition. Change in Aβ amyloid load over time has little relationship to change in cognition in time has little relationship to change in cognition in demented (Elan study) demented (Elan study)

Cognitive outcome is expensive target – more so in mildly Cognitive outcome is expensive target – more so in mildly affected affected

Therapeutic outcome Therapeutic outcome neurodegenerative measures, neurodegenerative measures, MRI, CSF tau, FDG PETMRI, CSF tau, FDG PET

MRI,FDG PET, cognitive tests, in AD, n=30

LabLab ModalityModality VariableVariable SS/armSS/arm

Cog.Cog. MMSEMMSE 703703

Cog.Cog. ADAS-CogADAS-Cog 514514

FosterFoster PETPET Hypometab 2Hypometab 2 508508

Cog.Cog. CDR SBCDR SB 495495

JagustJagust PETPET ROI-avgROI-avg 396396

Schuff- FSSchuff- FS MRIMRI VentriclesVentricles 9595

ReimanReiman PETPET CV - fROICV - fROI 9191

ThompsonThompson MRIMRI CV % changeCV % change 5353

FoxFox MRIMRI BSI% changeBSI% change 5050

ADNI: sample size per arm to detect a 25% reduction in rate (0 -12 months) of

decline in AD

Implications of the pathological Implications of the pathological specificity of biomarkers for anti amyloid specificity of biomarkers for anti amyloid

trials in MCI/ADtrials in MCI/AD initiating molecular pathway amyloid dysmetabolism initiating molecular pathway amyloid dysmetabolism inclusion should be based on evidence of the presence of inclusion should be based on evidence of the presence of amyloid in the brain amyloid in the brain amyloid PET imaging or CSF A amyloid PET imaging or CSF A42 42

Target efficacy, amyloid reduction Target efficacy, amyloid reduction amyloid biomarker amyloid biomarker (PET amyloid imaging) (PET amyloid imaging)

However, objective is to treat brain amyloidosis with However, objective is to treat brain amyloidosis with intent to affect cognition. Change in Aβ amyloid load over intent to affect cognition. Change in Aβ amyloid load over time has little relationship to change in cognition in time has little relationship to change in cognition in demented (Elan study) demented (Elan study)

Cognitive outcome is expensive target – more so in mildly Cognitive outcome is expensive target – more so in mildly affectedaffected

Therapeutic outcomeTherapeutic outcome neurodegenerative measures, neurodegenerative measures, MRI, CSF tau, FDG PETMRI, CSF tau, FDG PET

Implications of pathological cascade for therapy: treatment vs prevention

Ab Amyloid = CSF Ab42 or amyloid PET imaging; Tau Mediated Neuron Injury and Dysfunction = CSF tau or FDG PET; Brain Structure = structural MRI

Jack et al, Lancet Neurol 2010; 9: 119-28

Implications of pathological cascade for therapy: treatment vs prevention

Ab Amyloid = CSF Ab42 or amyloid PET imaging; Tau Mediated Neuron Injury and Dysfunction = CSF tau or FDG PET; Brain Structure = structural MRI

Jack et al, Lancet Neurol 2010; 9: 119-28

Anti amyloid clinical trials in Anti amyloid clinical trials in cognitively normal cognitively normal P P AisenAisen

Are the right subjects enrolled? Are the right subjects enrolled? Amyloid biomarker for inclusionAmyloid biomarker for inclusion

Is intervention effective?Is intervention effective? Neurodegenerative biomarker for outcomeNeurodegenerative biomarker for outcome can be adequately powered with much can be adequately powered with much

smaller sample sizes (less expensive) than smaller sample sizes (less expensive) than trials in MCI and AD using conventional trials in MCI and AD using conventional clinical endpointsclinical endpoints

Dynamic Biomarkers of the Alzheimer’s Pathological Cascade – model for chronic

degenerative conditions

Ab Amyloid = CSF Ab42 or amyloid PET imaging; Tau Mediated Neuron Injury and Dysfunction = CSF tau or FDG PET; Brain Structure = structural MRI

Jack et al, Lancet Neurol 2010; 9: 119-28

Acknowledgments Acknowledgments

AG11378 AG19142 AG16574 AG06786 ADNI Robert H. and Clarice Smith and Abigail

Van Buren Alzheimer's Disease Research Program

Alexander Family Professorship in Alzheimer's disease research

Mayo Rochester ADRC Mayo Rochester ADRC and Study of Agingand Study of Aging

Ronald C. Petersen*Ronald C. Petersen* David KnopmanDavid Knopman Brad BoeveBrad Boeve Joe ParisiJoe Parisi Walter RoccaWalter Rocca Rosebud RobertsRosebud Roberts Bob IvnikBob Ivnik Glenn SmithGlenn Smith Shane PankratzShane Pankratz Yonas GedaYonas Geda Selam NegashSelam Negash

Mayo Jacksonville

Dennis Dickson

Neil Graff-Radford

Tannis Ferman

Mayo Aging and Dementia Mayo Aging and Dementia Imaging Research (ADIR) Lab Imaging Research (ADIR) Lab

20102010Kejal KantarciJeff GunterMatthew SenjemPrashanthi VemuriJennifer WhitwellMary MachuldaMatt Bernstein Heidi EdmonsonStephen WeigandHeather WisteScott PrzybelskiGuang ZengAnkit Master

Denise Reyes Bret BorowskiGreg PreboskeChad WardBrian GreggPaul LewisKaely SteinertRamesh AvulaDon GerhartDan HeardScott SquiresSamantha WilleAJ Spychalla

Hypothetical model of dynamic Hypothetical model of dynamic biomarkers of the Alzheimer's biomarkers of the Alzheimer's

pathological cascadepathological cascade

Clifford R Jack JrClifford R Jack JrDavid S. KnopmanDavid S. KnopmanWilliam J. Jagust William J. Jagust Leslie M. Shaw Leslie M. Shaw

Paul S. AisenPaul S. AisenMichael W. Weiner Michael W. Weiner Ronald C. Petersen Ronald C. Petersen John Q. TrojanowskiJohn Q. Trojanowski

Lancet Neurology 2010; 9: 119-28