The Relationship between Amyloid Deposition, Neurodegeneration, and Cognitive Decline in Dementia

NEUROIMAGING (DJ BROOKS, SECTION EDITOR)

The Relationship between Amyloid Deposition,Neurodegeneration, and Cognitive Decline in Dementia

Rik Vandenberghe

Published online: 16 September 2014# Springer Science+Business Media New York 2014

Abstract Amyloid imaging has been clinically approved formeasuring β amyloid plaque load in patients being evaluatedfor Alzheimer's disease or other causes of cognitive decline.Here we explore a multidimensional approach to cognitivedecline, where we situate amyloid plaque burden among anumber of other relevant dimensions, such as aging, volumeloss, other proteinopathies such as TDP43 and Lewy bodies,and functional reorganisation of cognitive brain systems. Themultidimensional model incorporates a 'pure AD' trajectory,corresponding to e.g. monogenic Alzheimer's disease, butleaves room for other combinations of biomarker abnormali-ties (e.g. volume loss without amyloid positivity) and othertrajectories. More tools will become available in the future thatallow one to carve out a causal-mechanistic space for explaingcognitive decline in a personalized manner, enhancingprogress towards more efficacious interventions.

Keywords Florbetapir . Flutemetamol . Florbetaben .

Alzheimer . Preclinical . MCI

Introduction

Thanks to amyloid positron emission tomography (PET) it hasbecome possible to measure in vivo the fibrillar plaque burdenin an individual's brain [1, 2••, 3, 4]. Very recently, large-scaleprospective longitudinal multimodal imaging datasets thatincorporate amyloid imaging in hundreds of subjects havestarted to illuminate fundamental questions about the patho-physiology of cognitive decline in older adults, in a way thatcould not be imagined even 15 years ago. Longitudinal mea-surements of amyloid load have also been applied to assessamyloid-lowering study drug interventions, which unfortu-nately did not yield clinical benefit [5–7]. Paradoxically, oneof the most remarkable discoveries based on amyloid imagingis the substantial portion of patients who have a clinicalphenotype of amnestic mild cognitive impairment (MCI) oreven clinically probable Alzheimer's disease (AD), but actu-ally do not have an increased amyloid load [7, 8, 9, 10].Until a few years ago clinicians may have been relativelyconfident about the positive predictive value of theirclinical diagnoses of amnestic MCI or clinically probableAD for the presence of neuritic amyloid plaque, never-theless 10-40 % of such cases actually do not haveincreased amyloid burden [8, 9, 10, 11••, 12].

The clinical use of amyloid imaging has been approved byregulatory authorities for the purpose of measuring β amyloidplaque load in patients under evaluation for AD. The distinc-tion between a disease-oriented (distinction AD from non-AD)and a process-oriented diagnostic tool (assessment of amyloidload) is important: It reflects a shift towards measurements ofpathophysiological processes in an individual, away fromassigning cases to conventional disease categories. Withoutdoubt a categorical diagnostic approach is appropriate for anumber of diseases encountered in a memory clinic, such asautosomal dominant AD, Huntington's disease, or syphiliticdementia, to name just a few. In the current review however we

This article is part of the Topical Collection on Neuroimaging

R. VandenbergheLaboratory for Cognitive Neurology, Department of Neurosciences,University of Leuven, Leuven, Belgium

R. Vandenberghe (*)Neurology Department, University Hospitals Leuven, Herestraat 49,3000 Leuven, Belgiume-mail: [email protected]

R. VandenbergheAlzheimer Research Centre KU Leuven, Leuven Institute ofNeuroscience and Disease, University of Leuven, Leuven, Belgium

Curr Neurol Neurosci Rep (2014) 14:498DOI 10.1007/s11910-014-0498-9

will explore themerits of a multidimensional approach [4]. Theterm 'multidimensional' refers to the traditional mathematicalapproach of multidimensional scaling, where a datapoint(a case) is plotted against a number of axes (dimensions).Such a data-driven approach respects the heterogeneity be-tween patients as well as the overlap between seeminglydistinct conventional clinical categories. According to thismultidimensional view, amyloid imaging is heralding a newera where we will be able to measure multiple relevant path-ophysiological dimensions in an individual on a personalizedbasis and tailor interventions accordingly.

A telling example for illustratory purposes is Lewy bodydementia. In Lewy body dementia, amyloid PET is oftenpositive [13, 14].Within amultidimensional mindset, a logicalapproach would be to apply multimodal imaging rather than toinsist on assigning the case to a single diagnostic category,either Lewy body dementia or AD [15]. Multimodal imagingof volume loss, 18F-fluorodeoxyglucose (FDG) PET metabo-lism, dopaminergic integrity and amyloid PET in a givenindividual enables a personalized approach to diagnosis and,hopefully, future interventions [15]. Every day clinicians en-counter many similar situations, e.g. instances where cerebro-vascular disease, cerebral amyloid angiopathy and AD over-lap, as in the vast majority of elderly patients with cognitivedecline above the age of 80.

One of the principal arguments in favor of the multidimen-sional model comes from genetic association studies in AD(for review see [16]) which implicate more than 30 polymor-phisms pointing to a dysregulation of a multiple molecularnetworks beside the APP processing pathway. These genes arerelated to inflammatory, lipid processing, and cell cycle path-ways, among others [16]. The most vulnerable nodes withinthese molecular networks are the hubs [16], one of which isthe Amyloid Precursor Protein (APP). The molecular hub-and-network approach may create favorable conditions fordeveloping more efficacious therapy, besides the amyloid-lowering strategy [7].

Population-based epidemiological studies point into thesame direction. Although a strong relationship exists betweendementia and the presence of neurofibrillary tangles and neu-ritic plaques up to the age of 75 years [17], this relationshipbecomes weaker above the age of 80, especially for neuriticplaques. In contrast, neuropathological measures of volumeloss retain their strong relationship with the presence or ab-sence of dementia over the entire age range [17]. The causes ofvolume loss are much broader than a raised amyloid burdenalone as will be discussed later in this review.

The purpose of the current review is to critically analyzethe tripartite amyloid load, neurodegeneration and cognitivedecline within a multidimensional framework. This multidi-mensional space encompasses the trajectory of what has beencalled 'pure' AD, as specified by the sequential model trig-gered by amyloid proposed by Jack et al. (2013) [18]. It

however also allows for combinations of biomarker abnor-malities that fall outside the boundaries of the sequentialmodel (e.g. hippocampal volume loss or episodic memorydecline without amyloid positivity). As we will review, suchcombinations are frequent and we expect their frequency toincrease as more tools become available to measure causaldimensions beyond amyloid alone.

Sequential Model with Amyloid as a Trigger

The tripartite relationship between amyloid aggregation, neu-rodegeneration and cognitive decline is the centerpiece of oneof the most influential heuristic models of AD pathogenesis todate, the Jack et al. (2013) [18] model. This model is atranslation of the amyloid cascade hypothesis [19, 20] into asequence of changes that can be measured by contemporarytechniques in vivo in humans. Its basic tenets are:

1. A number of measures are grouped under a commondenominator called 'neurodegeneration'. These measuresinclude volume loss as measured with structural magneticresonance imaging (MRI), FDG PET hypometabolism,and cerebrospinal fluid (CSF) total tau increase.

2. The panoply of in vivo measures evolve over time ac-cording to an orderly chronological sequence. The initialevent is β amyloid related (as measured by means ofamyloid PET and CSF Aβ42). This serves as a triggerfor 'neurodegeneration' (e.g. hippocampal volume loss).Positive markers of neurodegeneration are followed byand go along with clinical symptoms, as has been dem-onstrated e.g. for volume loss by numerous studies[17, 21–23, 24, 25••].

3. Changes over time of each of these measures are bestdescribed by sigmoidal curves. The steepness of the slopefor each measure can be considered as an indication of thephase of the disease [26]. This time dimension can beexpressed as years or on an 'Alzheimer's Disease Progres-sion Scale' [27•]. For clinical drug trials, ideally, onewould like to include subjects when they are at a stagewhere the primary outcome measure is near the point ofmaximal rate of change.

This sequential model has been successfully applied topopulation-based longitudinal multimodal imagingdatasets, such as the Mayo Clinic Study of Aging(MCSA) (n > 400, including cognitively intact subjectsas well as more than 100 MCI cases) and the BaltimoreLongitudinal Study of Aging (BLSA, n > 800), and alsoto the Alzheimer Disease Neuroimaging Initiative(ADNI) and the Australian Imaging, Biomarker and Life-style Flagship Study of Ageing (AIBL). The latter twostudies have recruited cognitively intact participants fromthe community combined with memory clinic recruited

498, of 9 Curr Neurol Neurosci Rep (2014) 14:498

patients with early or late MCI and early-to-moderatestage AD.

Modelling the sigmoidal changes for each of the measureson the basis of these large datasets has yielded exciting novelinsights. For instance, in BLSA the episodic memory measurewhich changed first in healthy older adults was the totallearning score of the 15-word list recall test [27•]. After aperiod of time the delayed recall also started to decline. Assubjects advanced to the clinical stage, the rate of change indelayed recall became more prominent than the rate of changeof the total learning score [27•]. Sigmoidal curves have alsobeen applied to quantitative measures of neocortical thinning,an established method for analyzing structural changes inneocortex. Cortical thinning of AD vulnerable neocorticalregions follows a sigmoidal curve, with the maximum speedat the inflection point around a MiniMental State Examination(MMSE) score of 21 out of 30 [24]. This is consistent acrossall AD vulnerable regions [28]. In contrast, hippocampalvolume continues to decline (positive acceleration) until anMMSE of at least 15 out of 30 (lower limit in study) [24].

Among healthy controls and MCI patients, a higherbaseline amyloid burden is associated with a higher rateof amyloid accumulation over the following years[29–32], regardless of the baseline hippocampal volume[32]. Over time, amyloid positivity is associated with ahigher rate of subsequent volume loss, mainly in tempo-ral neocortex and posterior cingulate [33], in particular incases who already have hippocampal volume loss to startwith [32]. When patients have evolved into the probableAD stage, amyloid retention values are at a plateau withalmost no subsequent change [31]. In the AIBL study,the timespan from a clearly negative amyloid PET to amanifestly positive, AD-like PET was estimated to be noless than 30 years [31].

The most powerful tests of the sequential model have comefrom longitudinal studies in subjects with autosomal dominantAD [34••, 35, 36•]. According to a large-scale internationalstudy of carriers of AD genetic mutations [34••], amyloiddeposition in precuneus occurs 28 years before symptomonset, hippocampal volume loss at about −15 years and glu-cose hypometabolism in precuneus at around −18 yrs. Inanother monogenic AD cohort, the Presenilin 1 (PS1)E280A mutation kindred (also called the Antioquia(Columbia) kindred), MCI starts on average at the age of 44and AD at the age of 49 years [35, 36•]. Fibrillar Aβ (asmeasured with 18F-florbetapir PET) begins to accumulate inmutation carriers at a mean age of 28 years, rises steeply overthe next 9 years and plateaus at a mean age of 38 years, about6 and 11 years before the expected MCI and dementia onset.Surprisingly, in presymptomatic monogenic AD spinal fluidAβ42 initially increases before it drops to a level below thatseen in normal controls at around −20 yrs [34••, 36•, 37]. Theinitial Aβ42 rise is surprising as prior to these studies the only

pathological values described for Aβ42 were decreases com-pared to controls.

The sequential model captures the sequence and nature ofchanges that are strictly Alzheimer related, and optimally fitsthe disease course in monogenetic cases. However, a numberof common pathophysiological processes are left out andunderspecified in the sequential model of AD-related changes.These are the 'confounding problem of non-AD pathophysi-ologies in elderly cohorts' [18]. Likewise, the neuropatholog-ical basis of the measures subsumed under the term 'neurode-generation' is complex. Several pathological processes con-verge on the same brain structures that mediate cognitivedecline [22, 25••]. While cross-sectional [38] and longitudinal[23] measures of MRI volume loss correlate with neurofibril-lary tangles and neuritic plaque load, MRI volume reflectsmore than neuronal loss caused by AD pathology alone. It isalso affected by aging, vascular changes or hippocampalsclerosis and many other factors [39, 40••, 22, 25••]. This isrelevant for trial design: while selective neuronal loss wouldbe an exquisite drug target, multicomposite measures such asMRI, FDG PET or CSF tau may have limited utility as thesemeasures are the integral outcome of a wide variety ofprocesses.

A Cellular Definition of Neurodegeneration

Neurodegeneration refers to neuronal loss which is precededby early neuronal dysfunction and synaptic loss. Its key fea-ture is the selectivity for specific types of neurons in specificregions of the brain [41–44]. Neuronal loss (neurodegenera-tion in its strict definition) is directly related to cognitivedecline [41, 45, 46, 47, 48]. For that reason, neuronal loss isa process against which one would definitely want to inter-vene. If we could preserve neurons, clinical outcomemeasureswould most likely be favorably affected.

Neurodegeneration is associated with abnormal aggrega-tion of proteins in extra- or intracellular space [49]. InAD, this not only consists of neuritic amyloid plaquesand neurofibrillary tangles. Tar DNA Binding Protein (TDP)-43 proteinopathy, for instance, occurs in 50 % of AD causesand is more extensive in AD than previously thought [40••]. Itis associated with cognitive impairment in multiple domainsand can result in a phenotype that closely resembles that seenin AD [40••]. Prospective community-based neuropathologi-cal studies have also highlighted the contribution of cerebro-vascular lesions [50] or Lewy bodies [51] to cognitive declinein AD. According to the Vienna Transdube Aging study,the different proteinopathies make partly independentcontributions to cognitive decline [49, 52]. Differentcombinations of proteinopathies may have a cumulativeeffect until a threshold for cognitive dysfunction isexceeded and mechanisms of compensation fail [52].

Curr Neurol Neurosci Rep (2014) 14:498 of 9, 498

One of the fundamental questions in the pathogenesis ofneurodegenerative diseases is the relationship between selec-tive neuronal loss and the intra- and extracellularproteinopathy. In AD, how does amyloid aggregation relateto selective neuronal loss? Does it cause neuronal loss or is it,like neuronal loss, a consequence of a third-party cause?Proteinopathies may also be markers of diseases that involvecomplex cascades of events that over time result in cognitivedecline and/or dementia [16, 53].

Recent longitudinal neuropathological studies in largepopulation-based cohorts have yielded critical insights intothe relationship between brain volume loss, proteinopathiesand cognitive decline that often go against commonly heldviews [17, 40••, 53]. In a large-scale prospective population-based clinicopathological study, the Rush Memory and Agingstudy combined with the Religious Order studies [40••, 53],AD pathologic indices explained 30 to 34 % of the between-subjects variance in cognitive decline, infarcts 1 to 3 %,neocortical Lewy bodies 4 to 8 %, and locus coeruleus nor-adrenergic neuronal density 3-6 % of variance. In other words,approximately 60 % of the variance in late-life cognitivedecline remained unexplained by classical neuropathologicalmarkers. The authors refer to this unexplained variance as the'pathology-cognition gap' [53]. There have been endeavoursby neuropathologists to try to measure compensatory mecha-nisms postmortem [53, 54]. In vivo measures however, suchas functional MRI, may be a more appropriate method tobridge the pathology-cognition gap.

In Vivo Measures Beyond What can be Measured PostMortem: Brain Circuits at Work (and 'at rest')

Structural or biochemical measures post mortem fall inherent-ly short of capturing the rich dimensions of the functioningbrain during life at a systems level. Functional MRI hasrevealed a number of important and novel insights in howcognitive brain systems at work react to amyloid-related inju-ry and how functional connectivity drives regional vulnera-bility for amyloid aggregation.

Evidence of Functional Compensation from Task-RelatedfMRI

Among the most popular episodic memory tasks for fMRI isface-name associative learning. Subjects have to encode andsubsequently retrieve a list of paired faces and names. Thistask activates the hippocampal formation in normal controls.A series of fMRI studies have evaluated how hippocampalactivity levels change over the course of AD: During the initialAD phase fMRI responses are higher than in normal controls,suggestive of compensatory mechanisms [55]. As the diseaseprogresses, this pattern is reversed so that at the late MCI andearly clinically probable AD stage fMRI activity levels in the

same region are lower than in normal controls [55]. Thissequence was also confirmed by a PS1 E280Amutation study[36•, 56]: Hippocampal and parahippocampal responses arehigher in presymptomatic mutation carriers aged between 18and 30 years than in controls and precuneus is less deactivated[36•, 56]. A similar sequence may occur in APOE ε4 carriersin the pre- or post-amyloid phase [57].

Adaptive mechanisms in AD have also been demonstratedusing fMRI in the language domain [58, 59]. In MCI andearly-stage AD, activity during associative-semantic process-ing in the left posterior superior temporal sulcus, a key regionfor lexical-semantic retrieval, is lower than in healthy controls[58, 59]. Response amplitude correlates with scores on con-frontation naming and with speed of identification of writtenwords. In the homotopical right-sided region, activity is higherin those AD subjects in whom confrontation naming is pre-served compared to controls, suggestive of compensatorymechanisms [59].

In a third cognitive domain, executive functions, prefrontaladaptive increases in activation have been described [60, 61].

Intrinsic Connectivity as a Determinant of Regional ADVulnerability

In vivo amyloid imaging has revealed the areas of predilectionfor amyloid aggregation in the earliest disease phase. Theseareas are the precuneus, orbitofrontal cortex and anteriorcingulate [62–65, 66•]. Intrinsic connectivity network analysis(ICN) [67–70] has shed light on one of the most fundamentalquestions in neurodegenerative disease, the determinants ofthis regional vulnerability. Originally, the emphasis was on theconcordance between one of the ICNs, the default modenetwork, and amyloid deposition [68]. However, amyloiddeposition extends beyond DMN, even in the earliest stagesof AD [57, 71]. More recently, the focus has shifted to deter-mining which nodes within networks are most affected andwhy. A node's vulnerability is best predicted by a greater totalconnectional flow through that node and by a shorter func-tional path to the disease-related epicenters [70, 71]. Regionalspecificity of amyloid deposition therefore does not onlydepend on which network but also on the hub status of nodeswithin those networks [57]. Apart from its hub status, a highaerobic glycolysis index may also predispose to selectivevulnerability [72].

Temporal Ordering of in Vivo Imaging Measures Priorto Clinical Manifestations

A positive amyloid PET scan in a subject who does not showovert cognitive deficits is a frequent occurrence, in particularwith increasing age, rising from 5 % between 50 and 60 yearsof age up to 50 % above the age of 82 [66•, 73, 74]. Thisasymptomatic stage [54] is referred to as the 'at risk of AD'


stage by the International Working Group-2 [75], which re-serve the term 'presymptomatic' for asymptomatic carriers offully penetrant mutations. Theoretically, a distinction shouldbe drawn between risk factors for cognitive decline, on the onehand, and biomarkers that reveal pathological changes beforethe disease becomes clinically manifest, on the other hand. Forinstance, APOE ε4 carrier status is a risk factor and increasesthe risk of AD, in particular in women [76]. APOE ε4 is alsorobustly associated with a higher risk of a positive amyloidscan in healthy controls [31, 66•, 73, 77]. Other geneticpolymorphisms, e.g. related to CR1 or Brain Derived Neuro-trophic Factor (BDNF) codon 66 may also affect the amyloidburden according to preliminary evidence from case–controlstudies in smaller groups of subjects [65, 78].

The National Institute for Aging-Alzheimer Association(NIA-AA) nomenclature has introduced the term of preclini-cal AD [79], which implies that if a subject with preclinicalAD lives long enough, he will develop clinical AD. Thesequential order of biomarker positivity as specified by Jacket al. [18] has laid the foundation for the NIA-AA criteria forpreclinical AD [79]: Stage 0 corresponds to cases with normalcognition, normal amyloid load and normal hippocampalvolume, stage 1 to cases with amyloid positivity, stage 2 tocases with amyloid positivity and hippocampal volume loss,and stage 3 to cases with subtle cognitive changes, amyloidpositivity and hippocampal volume loss.

Soon after this NIA-AA preclinical AD research stagingwas proposed, it was clear that the combinations of amyloidand structural MRI findings seen in reality on a relativelyfrequent basis did not conform to the predictions based onthe sequential model [18]. According to a support vectormachine analysis of the phase 2 18F-flutemetamol study,structural MRI scans more often show an AD-like pattern incognitively intact controls than amyloid scans [80]. In anMCSA series of 286 cognitively intact healthy controls, morethan half showed imaging abnormalities, with the largestgroup (n=69) showing hippocampal volume loss while beingamyloid negative [81], a phenomenon dubbed 'suspected non-Alzheimer Pathophysiology' (sNAP). The sNAP group doesnot differ in any other respect from the preclinical AD cases instage 2 or 3, e.g. in terms of regional brain volume loss orFDG PET regional hypometabolism [82]. In other words, theyare classified as non-AD because of the absence of amyloid-positivity but are otherwise not distinguishable from thepreclinical AD cases.

In amyloid-positive cognitively intact subjects, studies ofcortical thinning yields a rather different picture from hippo-campal volumetry. At a time when hippocampal volume isstill within the normal range, amyloid positivity in healthycontrols is associated with cortical thinning in posterior cin-gulate/precuneus, inferior parietal, rostral middle frontal, andsuperior temporal cortex [24, 83]. Early structural changes inneocortex which precede changes of the hippocampal

formation were also reported in the PS1 E280A kindred:Parietal volume was lower in the carriers even before amyloidload increased, similarly to what had been previously de-scribed in APOE ε4 carriers [36•]. Amyloid-positivity incognitively intact controls is also associated with a disruptionof intrinsic connectivity between the precuneus and a largenumber of other regions [84].

Amyloid Load and the Path Towards Cognitive Decline

sNAP subjects are as likely to progress within one year assubjects in NIA-AA preclinical AD stage 1 (10 %) and morelikely than subjects in phase 0 (5 %). If subjects are bothamyloid-positive and have hippocampal atrophy (stage 2),they are more likely to progress than subjects who have eitherof the two abnormalities (21 %) [85]. Genetic risk factorsbeyond APOE may also affect the risk of cognitive declinein amyloid-positive cognitively intact subjects. Amongamyloid-positive subjects, BDNF codon 66 met carriers ex-perience faster cognitive decline (episodic memory, executivefunction, language) over 36 months period [86].

The prevalence of a positive amyloid scan in MCI is 50-70 % [8, 10, 87]. Amyloid-positive subjects with MCI aremuch more likely to progress to dementia within 20–24months than amyloid-negative subjects (50-67 % versus 5-19 %) [9, 87, 88, 89, 90]. Among amyloid-positive subjectswith MCI, hippocampal atrophy predicts shorter time-to-progression while Aβ load does not [88]. In contrast, whenamyloid-positive and amyloid-negative subjects with MCI arecombined, hippocampal atrophy and Aβ load predict shortertime-to-progression with comparable power (hazard ratio foran inter-quartile difference of 2.6 for both) [88].

Among patients with MCI, a significant portion (15-30 %)have pathological hippocampal volumes with normal amyloidlevels [10, 91], calledMCI-sNAP [10]. These patients are alsoat risk for progression, and, according to at least one studyinvolving MCSA and ADNI cases, more so than amyloid-positive MCI cases who have normal hippocampal volumes[10]. Similar to the progression rates seen in sNAP, the rate ofcognitive decline in MCI-sNAP suggest that there are otherpathways to cognitive decline than the amyloid-triggeredpathway and that these pathways occur with a relatively highprevalence.

Conclusion

Amyloid positivity is by definition and convention considereda necessary condition for classifying a case as AD. Cases whodo not follow the course specified by the sequential model areassigned to a non-AD category [92]. The time course of theMRI volume changes are represented in the Jack sequential


model on the assumption that the MRI volume changes aredue to amyloid aggregation [92]. Other changes affect thebrain e.g. due to aging, vascular factors , otherproteinopathies… Hence the occurrence of MRI changes inthe absence of amyloid-positivity which, on occasion, resem-ble the MRI pattern seen in AD (sNAP and MCI-sNAP). Theproportion of patients with an amnestic MCI or a clinicallyprobable AD phenotype who are amyloid-negative was under-estimated prior to the advent of amyloid imaging. This high-lights the importance of non-amyloid triggered pathways tocognitive decline in the general population as well as inclinical practice.

The multidimensional model is complementary to thesequential model and attempts to accomodate the diver-sity of processes and trajectories that can be seen inolder adults, without the a priori restriction to a con-ventional category of 'pure AD'. Importantly, the multi-dimensional model is not meant as an alternative or acriticism of the sequential model but incorporates thetrajectories as specified by Jack and colleagues [18].Evidently, despite the huge progress enabled by amyloidimaging, our ability to measure relevant pathophysiolog-ical processes in vivo in humans is still in its infancy.As more tools become available to carve out the causal-mechanistic space of cognitive decline in vivo in indi-vidual cases, we expect a still wider variety of combi-nations that cross the boundaries between conventionaldisease categories.

Regardless of which model one prefers, the primary goal isto identify targets for therapies in individuals. The multidi-mensional model builds on the molecular hub-and-networkapproach to AD [16]. Starting from a descriptive level, it willhopefully pave the way for intervention in a personalizedmanner which is broader than using amyloid-loweringtreatment on its own.

Acknowledgements Supported by FWO grants G.0076.02 (R.V.), KULeuven Research grants OT/08/056, OT/12/097 (R.V.), FederaalWetenschapsbeleid belspo Inter-University Attraction Pole P6/29 andP7/11, Stichting Alzheimer Onderzoek grant 11020, Vlaams Initiatiefvoor Netwerken voor Dementie Onderzoek (IWT) and TransformationeelGeneeskundig Onderzoek BioAdaptAD (IWT). R.V. is a senior clinicalinvestigator of the Fund for Scientific Research, Flanders, Belgium(FWO). I thank Katarzyna Adamczuk for helpful comments.

Compliance with Ethics Guidelines

Conflict of Interest Rik Vandenberghe has received consultancy feesfrom GE Healthcare (Consultancy agreement with GEHC from 24 May2013 to 24 May 2014). Clinical trial agreement between GEHC and myinstitution (UZ Leuven) for the phase 1 and 2 18 F-flutemetamoltrial (PI Rik Vandenberghe).

Human and Animal Rights and Informed Consent This article doesnot contain any studies with human or animal subjects performed by anyof the authors.

References

Papers of particular interest, published recently, have beenhighlighted as:• Of importance•• Of major importance

1. Klunk WE, Engler H, Nordberg A, Wang Y, Blomqvist G, Holt D,et al. Imaging brain amyloid in Alzheimer's disease with Pittsburghcompound-B. Ann Neurol. 2004;55:306–19.

2.•• Clark C, Pontecorvo M, Beach T, Bedell B, Coleman R,Doraiswamy P, et al. Cerebral PETwith florbetapir compared withneuropathology at autopsy for detection of neuritic amyloid-βplaques: a prospective cohort study. Lancet Neurol.2012;11:669–78. This pivotal paper establishes the validity of18F-florbetapir as a marker of neuritic amyloid plaque density.

3. Rowe CC, Villemagne VL. Brain amyloid imaging. J Nucl MedTechnol. 2013;41:11–8.

4. Vandenberghe R, Adamczuk K, Dupont P, Van Laere K, ChételatG. Amyloid PET in clinical practice: Its place in the multidimen-sional space of Alzheimer's disease. Neuroimage Clin.2013;2:497–511.

5. Rinne J, Brooks DJ, Rossor MN, Fox NC, Bullock R, Klunk WE,et al. (11)C-PIB PETassessment of change in fibrillar amyloid-betaload in patients with Alzheimer's disease treated withbapineuzumab: a phase 2, double-blind, placebo-controlled,ascending-dose study. Lancet Neurol. 2010;9:363–72.

6. Ostrowitzki S, Deptula D, Thurfjell L, Barkhof F, Bohrmann B,Brooks DJ, et al. Mechanism of amyloid removal in patients withAlzheimer disease treated with gantenerumab. Arch Neurol.2012;69:198–207.

7. Salloway S, Sperling R, Fox NC, Blennow K, Klunk WE,Raskind M, et al. Two phase 3 trials of bapineuzumab inmild-to-moderate Alzheimer's disease. N Engl J Med.2014;370:322–33.

8. Vandenberghe R, Van Laere K, Ivanoiu A, Salmon E, Bastin C,Triau E, et al. 18F-flutemetamol amyloid imaging in Alzheimerdisease and mild cognitive impairment: a phase 2 trial. AnnNeurol. 2010;68:319–29.

9. Doraiswamy PM, SperlingRA, Coleman RE, JohnsonKA, ReimanEM, Davis MD, et al. Amyloid-β assessed by florbetapir F18 PETand 18-month cognitive decline: a multicenter study. Neurology.2012;79:1636–44.

10. Petersen RC, Aisen P, Boeve BF, Geda YE, Ivnik RJ,Knopman DS, et al. Mild cognitive impairment due toAlzheimer disease in the community. Ann Neurol. 2013;74:199–208.

11.•• Beach TG, Monsell SE, Phillips LE, Kukull W. Accuracy of theclinical diagnosis of Alzheimer disease at National Institute onAging Alzheimer Disease centers, 2005–2010. J Neuropathol ExpNeurol. 2012;71:266–73. This clinicopathological paper providescritical information about the accuracy of a clincal diagnosis ofclinically probable or possible AD in a prospective multicentreacademic memory clinic based series collected between 2005 and2010 (n=919). Compared to the standard-of-truth (neuritic plaquedensity and neurofibrillary tangle stage), the positive predictivevalue of a diagnosis of clinically probable AD ranged between 62and 84 %; specificity between 60 % to 71 %, with a sensitivityaround 73 %.

12. Vandenberghe R, Adamczuk K, Van Laere K. The interest ofamyloid PET imaging in the diagnosis of Alzheimer's disease.Curr Opin Neurol. 2013;26:646–55.

13. Edison P, Rowe CC, Rinne JO, Nq S, Ahmed I, Kemmpainen N,et al. Amyloid load in Parkinson's disease dementia and Lewy body


dementia measured with [11C]PIB positron emission tomography. JNeurol Neurosurg Psychiatry. 2008;79:1331–8.

14. Gomperts SN. Imaging the role of amyloid in PD dementia anddementia with Lewy bodies. Curr Neurol Neurosci Rep.2014;14:472.

15. Kantarci K, Lowe VJ, Boeve BF, Weigand SD, Senjem ML,Przybelski SA, et al. Multimodality imaging characteristics ofdementia with Lewy bodies. Neurobiol Aging. 2012;33:2091–105.

16. Bettens K, Sleegers K, Van Broeckhoven C. Current status onAlzheimer disease molecular genetics: from past, to present, tofuture. Hum Mol Genet. 2010;19:R4–R11.

17. Savva GM,Wharton SB, Ince PG, Forster G,Matthews FE, BrayneC, et al. Age, neuropathology, and dementia. N Engl J Med.2009;360:2302–9.

18. Jack CR, Knopman DS, Jagust WJ, Petersen RC, Weiner MW,Aisen PS, et al. Tracking pathophysiological processes inAlzheimer's disease: an updated hypothetical model of dynamicbiomarkers. Lancet Neurol. 2013;12:207–16.

19. Hardy J, Selkoe DJ. The amyloid hypothesis of Alzheimer's dis-ease: progress and problems on the road to therapeutics. Science.2002;297:353–6.

20. Oddo S, Caccamo A, Shepherd JD, Murphy MP, Golde TE, KayedR, et al. Triple-transgenic model of Alzheimer's disease withplaques and tangles: intracellular Aβ and synaptic dysfunction.Neuron. 2003;39:409–21.

21. Silbert LC, Quinn JF, Moore MM, Corbridge E, Ball MJ, MurdochG, et al. Changes in premorbid brain volume predict Alzheimer'sdisease pathology. Neurology. 2003;61:487–92.

22. Jagust WJ, Zheng L, Harvey DJ, Mack WJ, Vinters HV, WeinerMW, et al. Neuropathological basis of magnetic resonance imagesin aging and dementia. Ann Neurol. 2008;63:72–80.

23. Josephs KA, Whitwell JL, Ahmed Z, Shiung MM, Weigand SD,Knopman DS, et al. Beta-amyloid burden is not associated withrates of brain atrophy. Ann Neurol. 2008;63:204–12.

24. Sabuncu MR, Desikan RS, Sepulcre J, Yeo BTT, Liu H,Schmansky NJ, et al. The dynamics of cortical and hippocampalatrophy in Alzheimer disease. Arch Neurol. 2011;68:1040–8.

25.•• Erten-Lyons D, Dodge HH, Woltjer R, Silbert LC, Howieson DB,Kramer P, et al. Neuropathologic basis of age-associated brainatrophy. JAMA Neurol. 2013;70:616–22. From the Oregon Brainand Aging study, different measures (ventricular, total brain andhippocampal) were derived from longitudinal structural MRIsobtained in 70 cognitively intact older adults. These were corre-lated with clinical status, APOE status and neuropathologicalmeasures (NFT, neuritic plaques, different types of vascular le-sions). Strongest correlations were obtained for ventricular andtotal brain measures. The correlations between these structuralmeasures and cognition remained even after controlling for thedegree of neuropathology. Hippocampal volume correlated onlywith the degree of amyloid angiopathy.

26. MC Donohue, H Jacqmin-Gadda, M Le Goff, RG Thomas, RRaman, AC Gamst, et al. Estimating long-term multivariate pro-gression from short-term data. Alzheimers Dement. in press, 2014.

27.• M Bilgel, YAn, A Lang, J Prince, L Ferrucci, B Jedynak, and SMResnick. Trajectories of Alzheimer disease-related cognitive mea-sures in a longitudinal sample. Alzheimers Dement. in press, 2014.Ultimately, efficacy of any intervention must be proven in terms ofclinical parameters. From almost 900 BLSA participants, some ofthe most commonly used neuropsychological episodic memorymeasures were analyzed for their sensitivity to detect decline inthe preclinical AD stage. The importance of this work lies in thehigh familiarity worldwide of the test parameters examined and thenovel insight into the relative sensitivity of encoding versus retrievalparameters

28. Dickerson BC, Bakkour A, Salat DH, Feczko E, Pacheco J, GreveDN, et al. The cortical signature of Alzheimer's disease: regionally

specific cortical thinning relates to symptom severity in very mild tomild AD dementia and is detectable in asymptomatic amyloid-positive individuals. Cereb Cortex. 2009;19:497–510.

29. Vlassenko AG, Mintun MA, Xiong C, Sheline YI, Goate AM,Benzinger TLS, et al. Amyloid-beta plaque growth in cognitivelynormal adults: longitudinal [11C]Pittsburgh compound B data. AnnNeurol. 2011;70:857–61.

30. Villain N, Chételat G, Grassiot B, Bourgeat P, Jones G, Ellis KA,et al. Regional dynamics of amyloid-β deposition in healthy elder-ly, mild cognitive impairment and Alzheimer's disease: a voxelwisePIB-PET longitudinal study. Brain. 2012;135:2126–39.

31. Villemagne VL, Burnham S, Bourgeat P, Brown B, Ellis KA,Salvado O, et al. Amyloid β deposition, neurodegeneration, andcognitive decline in sporadic Alzheimer's disease: a prospectivecohort study. Lancet Neurol. 2013;12:357–67.

32. Jack CR,Wiste HJ, KnopmanDS, Vemuri P, MielkeMM,WeigandSD, et al. Rates of β-amyloid accumulation are independent ofhippocampal neurodegeneration. Neurology. 2014;82:1605–12.

33. Chételat G, Villemagne VL, Villain N, Jones G, Ellis KA, Ames D,et al. Accelerated cortical atrophy in cognitively normal elderlywith high β-amyloid deposition. Neurology. 2012;78:477–84.

34.•• Bateman RJ, Xiong C, Benzinger TLS, Fagan AM, Goate A, FoxNC, et al. Clinical and biomarker changes in dominantly inheritedAlzheimer's disease. N Engl J Med. 2012;367:795–804. From theDominantly Inherited Alzheimer Network, multimodal cross-sectional data were analyzed from 50 symptomatic and 50 asymp-tomatic mutation carriers and 100 noncarrier siblings. This studyprovides critical empirical evidence for the orderly sequence ofchanges in in vivo biomarkers in monogenic AD up to 30 yearsprior to expected disease onset.

35. Fleisher AS, Chen K, Quiroz YT, Jakimovich LJ, Gomez MG,Langois CM, et al. Florbetapir PET analysis of amyloid-β deposi-tion in the Presenilin 1 E280A autosomal dominant Alzheimer'sdisease kindred: a cross-sectional study. Lancet Neurol. 2012;11:1057–65.

36.• Reiman EM, Quiroz YT, Fleisher AS, Chen K, Velez-Pardo C,Jimenez-Del-Rio M, et al. Brain imaging and fluid biomarkeranalysis in young adults at genetic risk for autosomal dominantAlzheimer's disease in the Presenilin 1 E280A kindred: a case–control study. Lancet Neurol. 2012;11:1048–56. From theColumbian Alzheimer Prevention Initiative registry, 20 PS1 muta-tion carriers and 20 noncarrier siblings between 18 and 26 years ofage underwent task-related fMRI, structural MRI and CSF. Thisstudy provides essential insight in the functional organization ofcognitive brain systems prior to clinical disease expression.

37. Fagan AM, Xiong C, Jasielec MS, Bateman RJ, Goate AM,Benzinger TLS, et al. Longitudinal change in CSF biomarkers inautosomal-dominant Alzheimer's disease. Sci Transl Med. 2014;6:226–30.

38. B Kaur, JJ Himali, S Seshadri, AS Beiser, R Au, AC McKee, et al.Association between neuropathology and brain volume in theFramingham Heart Study. Alzheimer Dis Assoc Disord. in press,2014.

39. Fotuhi M, Do D, Jack C. Modifiable factors that alter the size of thehippocampus with ageing. Nat Rev Neurol. 2012;8:189–202.

40.•• Wilson RS, Yu L, Trojanowski JQ, Chen EY, Boyle PA, BennettDA, et al. TDP-43 pathology, cognitive decline, and dementia in oldage. JAMA Neurol. 2013;70:1418–24. In a consecutive series of130 cases with annual cognitive assessments for more than10 years, postmortem measures of TDP43 explained as muchvariability in the rate of the cognitive decline as neurofibrillarytangles. This is of high importance as it points to high clinicalrelevance of other proteinopathies apart from the classical ADhallmark lesions.

41. Gomez-Isla T, Price JL, McKeel DW, Morris JC, Growdon JH,Hyman BT. Profound loss of layer II entorhinal cortex neurons


occurs in very mild Alzheimer's disease. J Neurosci.1996;16:4491–500.

42. Braak H, Braak E. Neuropathological stageing of Alzheimer-related changes. Acta Neuropathol. 1991;82:239–59.

43. Hyman BT, Gomez-Isla T. Alzheimer's disease is a laminar, region-al, and neural system specific disease, not a global brain disease.Neurobiol Aging. 1994;15:353–4.

44. Braak H, Thal DR, Ghebremedhin E, Del Tredici K. Stages of thepathologic process in Alzheimer disease: age categories from 1 to100 years. J Neuropathol Exp Neurol. 2011;70:960–9.

45. DeKosky ST, Scheff SW. Synapse loss in frontal cortex biopsies inAlzheimer's disease: correlation with cognitive severity. AnnNeurol. 1990;27:457–64.

46. Terry RD, Masliah E, Salmon DP, Butters N, DeTeresa R, Hill R,et al. Physical basis of cognitive alterations in Alzheimer's disease:Synapse loss is the major correlate of cognitive impairment. AnnNeurol. 1991;30:572–80.

47. Davis DG, Schmitt FA,Wekstein DR,MarkesberryWR.Alzheimerneuropathologic alterations in aged cognitively intact subjects. JNeuropathol Exp Neurol. 1999;58:376–88.

48. Scheff SW, Price DA, Schmitt FA, Mufson EJ. Hippocampal syn-aptic loss in early Alzheimer's disease and mild cognitive impair-ment. Neurobiol Aging. 2006;27:1372–84.

49. Kovacs GG, Milenkovic I, Wöhrer A, Höftberger R, Gelpi E,Haberler C, et al. Non-Alzheimer neurodegenerative pathologiesand their combinations are more frequent than commonly believedin the elderly brain: a community-based autopsy series. ActaNeuropathol. 2013;126:365–84.

50. Snowdon DA, Greiner LH, Mortimer JA, Riley KP, Greiner PA,Markesbery WR. Brain infarction and the clinical expression ofAlzheimer's disease: The Nun study. JAMA. 1997;277:813–7.

51. Schneider JA, Arvanitakis Z, Yu L, Boyle PA, Leurgans SE,Bennett DA. Cognitive impairment, decline and fluctuations inolder community-dwelling subjects with Lewy bodies. Brain.2012;135:3005–14.

52. Kovacs G. Current concepts of neurodegenerative diseases. EurMed J Neurol. 2014;1:78–86.

53. Boyle PA, Wilson RS, Yu L, Barr AM, Honer WG, Schneider JA,et al. Much of late life cognitive decline is not due to commonneurodegenerative pathologies. Ann Neurol. 2013;74:478–89.

54. Iacono D, Resnick SM, O'Brien R, Zonderman AB, An Y,Pletnikova O, et al. Mild cognitive impairment and asymptomaticAlzheimer disease subjects: equivalent β-amyloid and tau loadswith divergent cognitive outcomes. J Neuropathol Exp Neurol.2014;73:295–304.

55. Dickerson BC, Sperling RA. Large-scale functional brain networkabnormalities in Alzheimer's disease: insights from functional neu-roimaging. Behav Neurol. 2009;21:63–75.

56. Quiroz YT, Budson AE, Celone K, Ruiz A, Newmark R, CastrillónG, et al. Hippocampal hyperactivation in presymptomatic familialAlzheimer's disease. Ann Neurol. 2010;68:865–75.

57. Jagust WJ, Mormino EC. Lifespan brain activity, β-amyloid, andAlzheimer's disease. Trends Cogn Sci. 2011;15:520–6.

58. Vandenbulcke M, Peeters R, Dupont P, Van Hecke P,Vandenberghe R. Word reading and posterior temporal dys-function in amnestic mild cognitive impairment. Cereb Cortex.2007;17:542–51.

59. Nelissen N, Vandenbulcke M, Fannes K, Verbruggen A, Peeters R,Dupont P, et al. Aβ amyloid deposition in the language system andhow the brain responds. Brain. 2007;130:2055–69.

60. Becker JT, Mintun MA, Aleva K, Wiseman MB, Nichols T,DeKosky ST. Compensatory reallocation of brain resourcessupporting verbal episodic memory in Alzheimer's disease.Neurology. 1996;46:692–700.

61. Grady CL, McIntosh AR, Beig S, Keightley ML, Burian H, BlackSE. Evidence from functional neuroimaging of a compensatory

prefrontal network in Alzheimer's disease. J Neurosci. 2003;23:986–93.

62. Mintun MA, Larossa GN, Sheline YI, Dence CS, Lee SY, MachRH, et al. [11C]PIB in a nondemented population: potential ante-cedent marker of Alzheimer disease. Neurology. 2006;67:446–52.

63. Pike KE, Savage G, Villemagne VL, Ng S, Moss SA, Maruff P,et al. Beta-amyloid imaging and memory in non-demented individ-uals: evidence for preclinical Alzheimer's disease. Brain. 2007;130:2837–44.

64. Aizenstein HJ, Nebes RD, Saxton JA, Price JC, Mathis CA,Tsopelas ND, et al. Frequent amyloid deposition without significantcognitive impairment among the elderly. Arch Neurol. 2008;65:1509–17.

65. Adamczuk K, De Weer AS, Nelissen N, Chen K, Sleegers K,Bettens K, et al. Polymorphism of brain derived neurotrophic factorinfluences β amyloid load in cognitively intact apolipoprotein e ε4carriers. Neuroimage Clin. 2013;2:512–20.

66.• Fleisher AS, Chen K, Liu X, Ayutyanont N, Roontiva A,Thiyyagura P, et al. Apolipoprotein E ε4 and age effects onflorbetapir positron emission tomography in healthy aging andAlzheimer disease. Neurobiol Aging. 2013;34:1–12. Based on apooled analysis of 245 subjects who had undergone 18 F-florbetapir and APOE genotyping, reliable age-dependent esti-mates are obtained for linear amyloid increase and for amyloid-positivity as a function of age and APOE.

67. Yeo BTT, Krienen FM, Sepulcre J, Sabuncu MR, Lashkari D,HollinsheadM, et al. The organization of the human cerebral cortexestimated by intrinsic functional connectivity. J Neurophysiol.2011;106:1125–65.

68. Buckner RL, Snyder AZ, Shannon BJ, LaRossa G, Sachs R, FotenosAF, et al. Molecular, structural, and functional characterization ofAlzheimer's disease: evidence for a relationship between defaultactivity, amyloid, and memory. J Neurosci. 2005;25:7709–17.

69. Seeley WW, Crawford RK, Zhou J, Miller BL, Greicius MD.Neurodegenerative diseases target large-scale human brain net-works. Neuron. 2009;62:42–52.

70. Zhou J, Gennatas ED, Kramer JH, Miller BL, Seeley WW.Predicting regional neurodegeneration from the healthy brain func-tional connectome. Neuron. 2012;73:1216–27.

71. Myers N, Pasquini L, Göttler J, Grimmer T, Koch K, Ortner M,et al. Within-patient correspondence of amyloid-β and intrinsicnetwork connectivity in Alzheimer's disease. Brain. 2014;137:2052–64.

72. Vlassenko AG, Vaishnavi SN, Couture L, Sacco D, Shannon BJ,MachRH, et al. Spatial correlation between brain aerobic glycolysisand amyloid-β deposition. Proc Natl Acad Sci U S A. 2010;107:17763–7.

73. Mathis CA, Kuller LH, Klunk WE, Snitz BE, Price JC, WeissfeldLA, et al. In vivo assessment of amyloid-β deposition innondemented very elderly subjects. Ann Neurol. 2013;73:751–61.

74. G Chételat, R La Joie, N Villain, A Perrotin, V de La Sayette, FEustache, and R Vandenberghe. Amyloid imaging in cognitivelynormal individuals, at-risk populations and preclinical Alzheimer'sdisease. NeuroImage: Clinical. 2013

75. Dubois B, Feldman HH, Jacova C, Hampel H, Molinuevo JL,Blennow K, et al. Advancing research diagnostic criteria forAlzheimer's disease: the IWG-2 criteria. Lancet Neurol. 2014;13:614–29.

76. Altmann A, Tian L, Henderson VW, Greicius MD, Alzheimer'sDisease Neuroimaging Initiative Investigators. Sex modifies theAPOE-related risk of developing Alzheimer disease. Ann Neurol.2014;75:563–73.

77. Morris JC, Roe CM, Xiong C, Fagan AM, Goate AM, HoltzmanDM, et al. APOE predicts amyloid-beta but not tau Alzheimerpathology in cognitively normal aging. Ann Neurol. 2010;67:122–31.


78. Thambisetty M, An Y, Nalls M, Sojkova J, Swaminathan S, ZhouY, et al. Effect of complement CR1 on brain amyloid burden duringaging and its modification by APOE genotype. Biol Psychiatry.2013;73:422–8.

79. Sperling RA, Aisen P, Beckett LA, Bennett DA, Craft S,Fagan AM, et al. Toward defining the preclinical stages ofAlzheimer's disease: Recommendations from the NationalInstitute on Aging-Alzheimer's Association workgroups ondiagnostic guidelines for Alzheimer's disease. AlzheimersDement. 2011;7:280–92.

80. Vandenberghe R, Nelissen N, Salmon E, Ivanoiu A, Hasselbalch S,Andersen A, et al. Binary classification of 18F-flutemetamol PETusing machine learning: Comparison with visual reads and struc-tural MRI. Neuroimage. 2012;64C:517–25.

81. Jack CR, Knopman DS,Weigand SD,Wiste HJ, Vemuri P, Lowe V,et al. An operational approach to National Institute on Aging-Alzheimer's Association criteria for preclinical Alzheimer disease.Ann Neurol. 2012;71:765–75.

82. Knopman DS, Jack Jr CR, Wiste HJ, Weigand SD, VemuriP, Lowe VJ, et al. Brain injury biomarkers are not depen-dent on β-amyloid in normal elderly. Ann Neurol. 2013;73:472–80.

83. Becker JA, Hedden T, Carmasin J, Maye J, Rentz DM, Putcha D,et al. Amyloid-β associated cortical thinning in clinically normalelderly. Ann Neurol. 2011;69:1032–42.

84. Sheline YI, Raichle ME, Snyder AZ, Morris JC, Head D, Wang S,et al. Amyloid plaques disrupt resting state default mode networkconnectivity in cognitively normal elderly. Biol Psychiatry.2010;67:584–7.

85. Knopman DS. CR Jack, Jr, H. J. Wiste, S. D. Weigand, P. Vemuri,V. Lowe, et al. Short-term clinical outcomes for stages of NIA-AApreclinical Alzheimer disease. Neurology. 2012;78:1576–82.

86. LimYY,VillemagneVL, Laws SM,AmesD, Pietrzak RH, Ellis KA,et al. BDNF val66met, Aβ amyloid, and cognitive decline in pre-clinical Alzheimer's disease. Neurobiol Aging. 2013;34:2457–64.

87. Nordberg A, Carter SF, Rinne J, Drzezga A, Brooks DJ,Vandenberghe R, et al. A European multicentre PET study offibrillar amyloid in Alzheimer's disease. Eur J Nucl Med MolImaging. 2013;40:104–14.

88. Jack CR, Wiste HJ, Vemuri P, Weigand SD, Senjem ML, Zeng G,et al. Brain beta-amyloid measures and magnetic resonance imag-ing atrophy both predict time-to-progression from mild cognitiveimpairment to Alzheimer's disease. Brain. 2010;133:3336–48.

89. Villemagne VL, Pike KE, Chételat G, Ellis KA, Mulligan RS,Bourgeat P, et al. Longitudinal assessment of Aβ and cognition inaging and Alzheimer disease. Ann Neurol. 2011;69:181–92.

90. Rowe CC, Bourgeat P, Ellis KA, Brown B, Lim YY, Mulligan R,et al. Predicting Alzheimer disease with β-amyloid imaging: resultsfrom the Australian Imaging, Biomarkers, and Lifestyle study ofageing. Ann Neurol. 2013;74:905–13.

91. Duara R, Loewenstein DA, Shen Q, Barker W, Potter E, Varon D,et al. Amyloid positron emission tomography with (18)F-flutemetamol and structural magnetic resonance imaging in theclassification of mild cognitive impairment and Alzheimer's dis-ease. Alzheimers Dement. 2013;9:295–301.

92. Jack Jr CR, Wiste HJ, Lesnick TG, Weigand SD, Knopman DS,Vemuri P, et al. Brain β-amyloid load approaches a plateau.Neurology. 2013;80:890–6.


The Relationship between Amyloid Deposition, Neurodegeneration, and Cognitive Decline in Dementia

Documents

Transcript of The Relationship between Amyloid Deposition, Neurodegeneration, and Cognitive Decline in Dementia