To investigate MRI and proton magnetic resonance spectroscopy (MRS) predictors of mild cognitive impairment (MCI) in cognitively normal older adults.
Subjects were cognitively normal older adults (n = 1,156) who participated in the population-based Mayo Clinic Study of Aging MRI/MRS study from August 2005 to December 2010 and had at least one annual clinical follow-up. Single-voxel MRS was performed from the posterior cingulate gyri, and hippocampal volumes and white matter hyperintensity volumes were quantified using automated methods. Brain infarcts were assessed on MRI. Cox proportional hazards regression, with age as the time scale, was used to assess the effect of MRI and MRS markers on the risk of progression from cognitively normal to MCI. Linear mixed-effects models were used to assess the effect of MRI and MRS markers on cognitive decline.
After a median follow-up of 2.8 years, 214 participants had progressed to MCI or dementia (estimated incidence rate = 6.1% per year; 95% confidence interval = 5.3%–7.0%). In univariable modeling, hippocampal volume, white matter hyperintensity volume, and N-acetylaspartate/myo-inositol were significant predictors of MCI in cognitively normal older adults. In multivariable modeling, only decreased hippocampal volume and N-acetylaspartate/myo-inositol were independent predictors of MCI. These MRI/MRS predictors of MCI as well as infarcts were associated with cognitive decline (p < 0.05).
Quantitative MRI and MRS markers predict progression to MCI and cognitive decline in cognitively normal older adults. MRS may contribute to the assessment of preclinical dementia pathologies by capturing neurodegenerative changes that are not detected by hippocampal volumetry.
We describe the operationalization of the National Institute on Aging–Alzheimer’s Association (NIA-AA) workgroup diagnostic guidelines pertaining to Alzheimer disease (AD) dementia in a large multicenter group of subjects with AD dementia.
Subjects with AD dementia from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) with at least 1 amyloid biomarker (n = 211) were included in this report. Biomarker data from CSF Aβ42, amyloid PET, fluorodeoxyglucose-PET, and MRI were examined. The biomarker results were assessed on a per-patient basis and the subject categorization as defined in the NIA-AA workgroup guidelines was determined.
When using a requirement that subjects have a positive amyloid biomarker and single neuronal injury marker having an AD pattern, 87% (48% for both neuronal injury biomarkers) of the subjects could be categorized as “high probability” for AD. Amyloid status of the combined Pittsburgh compound B–PET and CSF results showed an amyloid-negative rate of 10% in the AD group. In the ADNI AD group, 5 of 92 subjects fit the category “dementia unlikely due to AD” when at least one neuronal injury marker was negative.
A large proportion of subjects with AD dementia in ADNI may be categorized more definitively as high-probability AD using the proposed biomarker scheme in the NIA-AA criteria. A minority of subjects may be excluded from the diagnosis of AD by using biomarkers in clinically categorized AD subjects. In a well-defined AD dementia population, significant biomarker inconsistency can be seen on a per-patient basis.
To test the hypotheses predicted in a hypothetical model of Alzheimer disease (AD) biomarkers that rates of β-amyloid (Aβ) accumulation on PET imaging are not related to hippocampal neurodegeneration whereas rates of neurodegenerative brain atrophy depend on the presence of both amyloid and neurodegeneration in a population-based sample.
A total of 252 cognitively normal (CN) participants from the Mayo Clinic Study of Aging had 2 or more serial visits with both amyloid PET and MRI. Subjects were classified into 4 groups based on baseline positive/negative amyloid PET (A+ or A−) and baseline hippocampal volume (N+ or N−). We compared rates of amyloid accumulation and rates of brain atrophy among the 4 groups.
At baseline, 148 (59%) were amyloid negative and neurodegeneration negative (A−N−), 29 (12%) amyloid negative and neurodegeneration positive (A−N+), 56 (22%) amyloid positive and neurodegeneration negative (A+N−), and 19 (8%) amyloid positive and neurodegeneration positive (A+N+). High rates of Aβ accumulation were found in those with abnormal amyloid at baseline and were not influenced by hippocampal neurodegeneration at baseline. In contrast, rates of brain atrophy were greatest in A+N+.
We describe a 2-feature biomarker approach to classifying elderly CN subjects that is complementary to the National Institute on Aging–Alzheimer's Association preclinical staging criteria. Our results support 2 key concepts in a model of the temporal evolution of AD biomarkers. First, the rate of Aβ accumulation is not influenced by neurodegeneration and thus may be a biologically independent process. Second, Aβ pathophysiology increases or catalyzes neurodegeneration.
The new criteria for preclinical Alzheimer’s Disease (AD) proposed 3 stages: abnormal levels of β-amyloid (stage 1); stage 1 plus evidence of brain injury (stage 2); and stage 2 plus subtle cognitive changes (stage 3). However, a large group of subjects with normal β-amyloid biomarkers have evidence of brain injury; we labeled them as “suspected non-Alzheimer pathway” (sNAP) group. The characteristics of the sNAP group are poorly understood.
Using the preclinical AD classification, 430 cognitively normal subjects from the Mayo Clinic Study of Aging who underwent brain MR, 18fluorodeoxyglucose (FDG) and Pittsburgh compound B (PiB) positron emission tomography (PET) were evaluated with FDG PET regional volumetrics, MR regional brain volumetrics, white matter hyperintensity (WMH) volume and number of infarcts. We examined cross-sectional associations across AD preclinical stages, those with all biomarkers normal, and the sNAP group.
The sNAP group had a lower proportion (14%) with APOE ε4 genotype than the preclinical AD stages 2 + 3. The sNAP group did not show any group differences compared to stages 2 + 3 of the preclinical AD group on measures of FDG PET regional hypometabolism, MR regional brain volume loss, cerebrovascular imaging lesions, vascular risk factors, imaging changes associated with α-synucleinopathy or physical findings of parkinsonism.
Cognitively normal persons with brain injury biomarker abnormalities, with or without abnormal levels of β-amyloid, were indistinguishable on a variety of imaging markers, clinical features and risk factors. The initial appearance of brain injury biomarkers that occurs in cognitively normal persons with preclinical AD may not depend on β-amyloidosis.
Alzheimer’s disease; PET imaging; MR imaging; Epidemiology
To determine the association of conventional cardiovascular risk factors, markers of platelet activation, and thrombogenic blood-borne microvesicles with white matter hyperintensity (WMH) load and progression in recently menopausal women.
Women (n = 95) enrolled in the Mayo Clinic Kronos Early Estrogen Prevention Study underwent MRI at baseline and at 18, 36, and 48 months after randomization to hormone treatments. Conventional cardiovascular risk factors, carotid intima-medial thickness, coronary arterial calcification, plasma lipids, markers of platelet activation, and thrombogenic microvesicles were measured at baseline. WMH volumes were calculated using a semiautomated segmentation algorithm based on fluid-attenuated inversion recovery MRI. Correlations of those parameters with baseline WMH and longitudinal change in WMH were adjusted for age, months past menopause, and APOE ε4 status in linear regression analysis.
At baseline, WMH were present in all women. The WMH to white matter volume fraction at baseline was 0.88% (0.69%, 1.16%). WMH volume increased by 122.1 mm3 (95% confidence interval: −164.3, 539.5) at 36 months (p = 0.003) and 155.4 mm3 (95% confidence interval: −92.13, 599.4) at 48 months (p < 0.001). These increases correlated with numbers of platelet-derived and total thrombogenic microvesicles at baseline (p = 0.03).
Associations of platelet-derived, thrombogenic microvesicles at baseline and increases in WMH suggest that in vivo platelet activation may contribute to a cascade of events leading to development of WMH in recently menopausal women.
To model the temporal trajectory of β-amyloid accumulation using serial amyloid PET imaging.
Participants, aged 70–92 years, were enrolled in either the Mayo Clinic Study of Aging (n = 246) or the Mayo Alzheimer's Disease Research Center (n = 14). All underwent 2 or more serial amyloid PET examinations. There were 205 participants classified as cognitively normal and 55 as cognitively impaired (47 mild cognitive impairment and 8 Alzheimer dementia). We measured baseline amyloid PET-relative standardized uptake values (SUVR) and, for each participant, estimated a slope representing their annual amyloid accumulation rate. We then fit regression models to predict the rate of amyloid accumulation given baseline amyloid SUVR, and evaluated age, sex, clinical group, and APOE as covariates. Finally, we integrated the amyloid accumulation rate vs baseline amyloid PET SUVR association to an amyloid PET SUVR vs time association.
Rates of amyloid accumulation were low at low baseline SUVR. Rates increased to a maximum at baseline SUVR around 2.0, above which rates declined—reaching zero at baseline SUVR above 2.7. The rate of amyloid accumulation as a function of baseline SUVR had an inverted U shape. Integration produced a sigmoid curve relating amyloid PET SUVR to time. The average estimated time required to travel from an SUVR of 1.5–2.5 is approximately 15 years.
This roughly 15-year interval where the slope of the amyloid SUVR vs time curve is greatest and roughly linear represents a large therapeutic window for secondary preventive interventions.
In 2010, the authors published a hypothetical model of the major biomarkers of Alzheimer’s disease (AD). The model was received with interest because we described the temporal evolution of AD biomarkers in relation to each other and to the onset and progression of clinical symptoms. In the interim, evidence has accumulated that supports the major assumptions of this model. Evidence has also appeared that challenges some of the assumptions underlying our original model. Recent evidence has allowed us to modify our original model. Refinements include indexing subjects by time rather than clinical symptom severity; incorporating inter-subject variability in cognitive response to the progression of AD pathophysiology; modifications of the specific temporal ordering of some biomarkers; and, recognition that the two major proteinopathies underlying AD biomarker changes, Aβ and tau, may be initiated independently in late onset AD where we hypothesize that an incident Aβopathy can accelerate an antecedent tauopathy.
The appearance of β-amyloidosis and brain injury biomarkers in cognitively normal (CN) persons is thought to define risk for the future development of cognitive impairment due to Alzheimer’s disease (AD), but their interaction is poorly understood.
To test the hypothesis that the joint presence of β-amyloidosis and brain injury biomarkers would lead to more rapid neurodegeneration.
Longitudinal Cohort Study
Population-based Mayo Clinic Study of Aging.
191 CN persons (median age 77, range 71–93) in the Mayo Clinic Study of Aging who underwent MR, FDG PET and PiB PET imaging at least twice 15 months apart. Subjects were grouped according to the recommendations of the NIA-AA Preclinical AD criteria, based on the presence of β-amyloidosis, defined as a PiB PET SUVr >1.5, alone (Stage 1) or with brain injury (stage 2+3), defined as hippocampal atrophy or FDG hypometabolism. We also studied a group of MCI (n=17) and dementia (n=9) patients from the Mayo Clinic Study of Aging or the Mayo Alzheimer Center with similar follow-up times who had had comparable imaging and who all had PiB PET SUVr >1.5.
Main Outcome Measures
Rate of change of cortical volume on volumetric MR scans and rate of change of glucose metabolism on FDG PET scans.
There were 25 CN subjects with both high PiB retention and low hippocampal volume or FDG hypometabolism at baseline (Preclinical AD stages 2+3). On follow-up scans, the Preclinical AD stages 2+3 subjects had greater loss of medial temporal lobe volume and greater glucose hypometabolism in the medial temporal lobe compared to other CN groups. The changes were similar to the cognitively impaired participants. Extra-temporal regions did not show similar changes.
Higher rates of medial temporal neurodegeneration occurred in CN individuals who, on their initial scans, had abnormal levels of both β-amyloid and brain injury biomarkers.
Alzheimer’s disease; PET imaging; MR imaging; Epidemiology
With the aging of the population, the burden of Alzheimer’s disease (AD) is rapidly expanding. More than 5 million people in the US alone are affected with AD and this number is expected to triple by 2050. While men may have a higher risk of mild cognitive impairment (MCI), an intermediate stage between normal aging and dementia, women are disproportionally affected with AD. One explanation is that men may die of competing causes of death earlier in life, so that only the most resilient men may survive to older ages. However, many other factors should also be considered to explain the sex differences. In this review, we discuss the differences observed in men versus women in the incidence and prevalence of MCI and AD, in the structure and function of the brain, and in the sex-specific and gender-specific risk and protective factors for AD. In medical research, sex refers to biological differences such as chromosomal differences (eg, XX versus XY chromosomes), gonadal differences, or hormonal differences. In contrast, gender refers to psychosocial and cultural differences between men and women (eg, access to education and occupation). Both factors play an important role in the development and progression of diseases, including AD. Understanding both sex- and gender-specific risk and protective factors for AD is critical for developing individualized interventions for the prevention and treatment of AD.
Alzheimer’s disease; dementia; sex; gender; risk factors; dimorphic medicine
White matter hyperintensities (WMHs) associate with both cognitive slowing and motor dysfunction in the neurologically normal elderly. A full understanding of the pathology underlying this clinicoradiologic finding is currently lacking in autopsy-confirmed normal brains. To determine the histopathologic basis of WMH seen on MRI, we studied the relationship between postmortem fluid-attenuated inversion recovery (FLAIR) intensity and neuropathologic markers of white matter lesions (WMLs) that correspond to WMH in cognitively normal aging brains. Samples of periventricular (n = 24), subcortical (n = 26), and normal-appearing white matter (NAWM, n = 31) from 4 clinically and pathologically-confirmed normal cases were examined. FLAIR intensity, vacuolation, and myelin basic protein (MBP) immunoreactivity loss were significantly higher in periventricular WML vs. subcortical WML; both were higher than in NAWM. The subcortical WML and NAWM had significantly less axonal loss, astrocytic burden, microglial density, and oligodendrocyte loss than the periventricular WML. Thus, vacuolation, myelin density and small vessel density contribute to the rarefaction of white matter whereas axonal density, oligodendrocyte density, astroglial burden and microglial density did not. These data suggest that the age-related loss of MBP and a decrease in small vessel density, may contribute to vacuolation of white matter. The vacuolation enables interstitial fluid to accumulate, which contributes to the prolonged T2 relaxation and elevated FLAIR intensity in the white matter.
Digital microscopy; Fluid attenuated inversion recovery; Normal aging; Oligodendrocytes; Postmortem magnetic resonance imaging; White matter
To estimate the incidence of and to characterize cognitive and imaging findings associated with incident amyloid PET positivity.
Cognitively normal (CN) participants in the Mayo Clinic Study of Aging who had 2 or more serial imaging assessments, which included amyloid PET, FDG-PET, and MRI at each time point, were eligible for analysis (n = 207). Twelve subjects with Alzheimer disease dementia were included for comparison.
Of the 123 CN participants who were amyloid-negative at baseline, 26 met criteria for incident amyloid PET positivity. Compared to the 69 subjects who remained stable amyloid-negative, on average these 26 did not differ on any imaging, demographic, or cognitive variables except amyloid PET (by definition) and task-free functional connectivity, which at baseline was greater in the incident amyloid-positive group. Eleven of the 26 incident amyloid-positive subjects had abnormal hippocampal volume, FDG-PET, or both at baseline.
The incidence of amyloid PET positivity is approximately 13% per year among CN participants over age 70 sampled from a population-based cohort. In 15/26 (58%), incident amyloid positivity occurred prior to abnormalities in FDG-PET and hippocampal volume. However, 11/26 (42%) incident amyloid-positive subjects had evidence of neurodegeneration prior to incident amyloid positivity. These 11 could be subjects with combinations of preexisting non-Alzheimer pathophysiologies and tau-mediated neurodegeneration who newly entered the amyloid pathway. Our findings suggest that both “amyloid-first” and “neurodegeneration-first” biomarker profile pathways to preclinical AD exist.
To investigate the effect of intellectual and physical activity on biomarkers of Alzheimer’s disease (AD) pathophysiology and cognition in a non-demented elderly population. The biomarkers evaluated were brain Aβ-amyloid load via PIB-PET, neuronal dysfunction via FDG-PET and neurodegeneration via Structural-MRI.
We studied 515 non-demented (428 cognitively normal and 87 MCI) participants in the population based Mayo Clinic Study of Aging who completed a 3T MRI, PET scans, APOE genotype, had lifestyle activity measures and cognition data available. The imaging measures computed were global PiB-PET uptake; global FDG-PET and MRI based hippocampal volume. We consolidated activity variables into lifetime intellectual, current intellectual and current physical activities. We used a global cognitive Z-score as a measure of cognition. We applied two independent methods – partial correlation analysis adjusted for age and gender and path analysis using structural equations to evaluate the associations between lifestyle activities, imaging biomarkers and global cognition.
None of the lifestyle variables correlated with the biomarkers and the path associations between lifestyle variables and biomarkers were not significant (p>0.05). On the other hand, all the biomarkers were correlated with global cognitive Z-score (p<0.05) and the path associations between (lifetime and current) intellectual activities and global Z-score were significant (p<0.01).
Intellectual and physical activity lifestyle factors were not associated with AD biomarkers but intellectual lifestyle factors explained variability in the cognitive performance in this non-demented population. This study provides evidence that lifestyle activities may delay the onset of dementia but do not significantly influence the expression of AD pathophysiology.
Alzheimer’s disease; Imaging biomarkers; Lifestyle Activities
Secondary prevention trials in subjects with preclinical Alzheimer disease may require documentation of brain amyloidosis. The identification of inexpensive and noninvasive screening variables that can identify individuals who have significant amyloid accumulation would reduce screening costs.
A total of 483 cognitively normal (CN) individuals, aged 70–92 years, from the population-based Mayo Clinic Study of Aging, underwent Pittsburgh compound B (PiB)–PET imaging. Logistic regression determined whether age, sex, APOE genotype, family history, or cognitive performance was associated with odds of a PiB retention ratio >1.4 and >1.5. Area under the receiver operating characteristic curve (AUROC) evaluated the discrimination between PiB-positive and -negative subjects. For each characteristic, we determined the number needed to screen in each age group (70–79 and 80–89) to identify 100 participants with PiB >1.4 or >1.5.
A total of 211 (44%) individuals had PiB >1.4 and 151 (31%) >1.5. In univariate and multivariate models, discrimination was modest (AUROC ∼0.6–0.7). Multivariately, age and APOE best predicted odds of PiB >1.4 and >1.5. Subjective memory complaints were similar to cognitive test performance in predicting PiB >1.5. Indicators of PiB positivity varied with age. Screening APOE ε4 carriers alone reduced the number needed to screen to enroll 100 subjects with PIB >1.5 by 48% in persons aged 70–79 and 33% in those aged 80–89.
Age and APOE genotype are useful predictors of the likelihood of significant amyloid accumulation, but discrimination is modest. Nonetheless, these results suggest that inexpensive and noninvasive measures could significantly reduce the number of CN individuals needed to screen to enroll a given number of amyloid-positive subjects.
To determine the association between the focal atrophy measures on antemortem MRI and postmortem neuropathologic classification of dementia with Lewy bodies (DLB) using the Third Report of the DLB Consortium criteria.
We retrospectively identified 56 subjects who underwent antemortem MRI and had Lewy body (LB) pathology at autopsy. Subjects were pathologically classified as high (n = 25), intermediate (n = 22), and low likelihood DLB (n = 9) according to the Third Report of the DLB Consortium criteria. We included 2 additional pathologic comparison groups without LBs: one with low likelihood Alzheimer disease (AD) (control; n = 27) and one with high likelihood AD (n = 33). The associations between MRI-based volumetric measurements and the pathologic classification of DLB were tested with analysis of covariance by adjusting for age, sex, and MRI-to-death interval.
Antemortem hippocampal and amygdalar volumes increased from low to intermediate to high likelihood DLB (p < 0.001, trend test). Smaller hippocampal and amygdalar volumes were associated with higher Braak neurofibrillary tangle stage (p < 0.001). Antemortem dorsal mesopontine gray matter (GM) atrophy was found in those with high likelihood DLB compared with normal control subjects (p = 0.004) and those with AD (p = 0.01). Dorsal mesopontine GM volume decreased from low to intermediate to high likelihood DLB (p = 0.01, trend test).
Antemortem hippocampal and amygdalar volumes increase and dorsal mesopontine GM volumes decrease in patients with low to high likelihood DLB according to the Third Report of the DLB Consortium criteria. Patients with high likelihood DLB typically have normal hippocampal volumes but have atrophy in the dorsal mesopontine GM nuclei.
Acetylcholinesterase inhibitors are commonly used to treat patients with dementia with Lewy bodies. Hippocampal atrophy on magnetic resonance imaging and amyloid-β load on positron emission tomography are associated with the Alzheimer’s disease-related pathology in patients with dementia with Lewy bodies. To date, few studies have investigated imaging markers that predict treatment response in patients with dementia with Lewy bodies. Our objective was to determine whether imaging markers of Alzheimer’s disease-related pathology such as hippocampal volume, brain amyloid-β load on 11C Pittsburgh compound B positron emission tomography predict treatment response to acetylcholinesterase inhibitors in patients with dementia with Lewy bodies. We performed a retrospective analysis on consecutive treatment-naive patients with dementia with Lewy bodies (n = 54) from the Mayo Clinic Alzheimer’s Disease Research Centre who subsequently received acetylcholinesterase inhibitors and underwent magnetic resonance imaging with hippocampal volumetry. Baseline and follow-up assessments were obtained with the Mattis Dementia Rating Scale. Subjects were divided into three groups (reliable improvement, stable or reliable decline) using Dementia Rating Scale reliable change indices determined previously. Associations between hippocampal volumes and treatment response were tested with analysis of covariance adjusting for baseline Dementia Rating Scale, age, gender, magnetic resonance field strength and Dementia Rating Scale interval. Seven subjects underwent 11C Pittsburgh compound B imaging within 12 weeks of magnetic resonance imaging. Global cortical 11C Pittsburgh compound B retention (scaled to cerebellar retention) was calculated in these patients. Using a conservative psychometric method of assessing treatment response, there were 12 patients with reliable decline, 29 stable cases and 13 patients with reliable improvement. The improvers had significantly larger hippocampi than those that declined (P = 0.02) and the stable (P = 0.04) group. An exploratory analysis demonstrated larger grey matter volumes in the temporal and parietal lobes in improvers compared with those who declined (P < 0.05). The two patients who had a positive 11C Pittsburgh compound B positron emission tomography scan declined and those who had a negative 11C Pittsburgh compound B positron emission tomography scan improved or were stable after treatment. Patients with dementia with Lewy bodies who do not have the imaging features of coexistent Alzheimer’s disease-related pathology are more likely to cognitively improve with acetylcholinesterase inhibitor treatment.
dementia with Lewy bodies; acetylcholinesterase inhibitors; MRI; PiB; PET; amyloid
The hexanucleotide repeat in the chromosome 9 open reading frame 72 (C9ORF72) gene was recently discovered as the pathogenic mechanism underlying many families with frontotemporal dementia (FTD) and/or amyotrophic lateral sclerosis (ALS) linked to chromosome 9 (c9FTD/ALS). We report the clinical, neuropsychological, and neuroimaging findings of a family with the C9ORF72 mutation and clinical diagnoses bridging the FTD, parkinsonism and ALS spectrum.
To characterize the antemortem characteristics of a family with c9FTD/ALS associated with the GGGGCC repeat expansion in C9ORF72
Tertiary care academic medical center.
The members of the family affected by the mutation with features of FTD and/or ALS.
Main Outcome Measures
Clinical, neuropsychological, and neuroimaging assessments.
All three examined subjects had the hexanucleotide expansion detected in C9ORF72. All had personality/behavioral changes early in the course of the disease. One case had levodopa-unresponsive parkinsonism, and one had ALS. MRI showed symmetric bilateral frontal, temporal, insular and cingulate atrophy.
This report highlights the clinical and neuroimaging characteristics of a family with c9FTD/ALS. Further studies are needed to better understand the phenotypical variability and the clinico-neuroimaging-neuropathologic correlations.
To characterize the shape of the trajectories of Alzheimer’s Disease (AD) biomarkers as a function of MMSE.
Longitudinal registries from the Mayo Clinic and the Alzheimer’s Disease Neuroimaging Initiative (ADNI).
Two different samples (n=343 and n=598) were created that spanned the cognitive spectrum from normal to AD dementia. Subgroup analyses were performed in members of both cohorts (n=243 and n=328) who were amyloid positive at baseline.
Main Outcome Measures
The shape of biomarker trajectories as a function of MMSE, adjusted for age, was modeled and described as baseline (cross-sectional) and within-subject longitudinal effects. Biomarkers evaluated were cerebro spinal fluid (CSF) Aβ42 and tau; amyloid and fluoro deoxyglucose position emission tomography (PET) imaging, and structural magnetic resonance imaging (MRI).
Baseline biomarker values generally worsened (i.e., non-zero slope) with lower baseline MMSE. Baseline hippocampal volume, amyloid PET and FDG PET values plateaued (i.e., non-linear slope) with lower MMSE in one or more analyses. Longitudinally, within-subject rates of biomarker change were associated with worsening MMSE. Non-constant within-subject rates (deceleration) of biomarker change were found in only one model.
Biomarker trajectory shapes by MMSE were complex and were affected by interactions with age and APOE status. Non-linearity was found in several baseline effects models. Non-constant within-subject rates of biomarker change were found in only one model, likely due to limited within-subject longitudinal follow up. Creating reliable models that describe the full trajectories of AD biomarkers will require significant additional longitudinal data in individual participants.
Alzheimer’s disease biomarkers; Magnetic Resonance Imaging; cerebro spinal fluid; amyloid PET imaging; FDG PET imaging
A workgroup commissioned by the Alzheimer’s Association (AA) and the National Institute on Aging (NIA) recently published research criteria for preclinical Alzheimer’s disease (AD). We performed a preliminary assessment of these guidelines.
We employed Pittsburgh compound B positron emission tomography (PET) imaging as our biomarker of cerebral amyloidosis and 18fluorodeoxyglucose PET imaging and hippocampal volume as biomarkers of neurodegeneration. A group of 42 clinically diagnosed AD subjects was used to create imaging biomarker cut-points. A group of 450 cognitively normal (CN) subjects from a population based sample was used to develop cognitive cut-points and to assess population frequencies of the different preclinical AD stages using different cut-point criteria.
The new criteria subdivide the preclinical phase of AD into stages 1–3. To classify our CN subjects, two additional categories were needed. Stage 0 denotes subjects with normal AD biomarkers and no evidence of subtle cognitive impairment. Suspected Non-AD Pathophysiology (SNAP) denotes subjects with normal amyloid PET imaging, but abnormal neurodegeneration biomarker studies. At fixed cut-points corresponding to 90% sensitivity for diagnosing AD and the 10th percentile of CN cognitive scores, 43% of our sample was classified as stage 0; 16% stage 1; 12 % stage 2; 3% stage 3; and 23% SNAP.
This cross-sectional evaluation of the NIA-AA criteria for preclinical AD indicates that the 1–3 staging criteria coupled with stage 0 and SNAP categories classify 97% of CN subjects from a population-based sample, leaving just 3% unclassified. Future longitudinal validation of the criteria will be important.
Fully automated classification algorithms have been successfully applied to diagnose a wide range of neurological and psychiatric diseases. They are sufficiently robust to handle data from different scanners for many applications and in specific cases outperform radiologists. This article provides an overview of current applications taking structural imaging in Alzheimer's Disease and schizophrenia as well as functional imaging to diagnose depression as examples. In this context, we also report studies aiming to predict the future course of the disease and the response to treatment for the individual. This has obvious clinical relevance but is also important for the design of treatment studies that may aim to include a cohort with a predicted fast disease progression to be more sensitive to detect treatment effects.
In the second part, we present our own opinions on i) the role these classification methods can play in the clinical setting; ii) where their limitations are at the moment and iii) how those can be overcome. Specifically, we discuss strategies to deal with disease heterogeneity, diagnostic uncertainties, a probabilistic framework for classification and multi-class classification approaches.
Automated diagnosing; MRI; SVM; Dementia; Depression; Schizophrenia
Fully automated machine learning methods based on structural magnetic resonance imaging (MRI) data can assist radiologists in the diagnosis of Alzheimer’s disease (AD). These algorithms require large data sets to learn the separation of subjects with and without AD. Training and test data may come from heterogeneous hardware settings, which can potentially affect the performance of disease classification.
A total of 518 MRI sessions from 226 healthy controls and 191 individuals with probable AD from the multicenter Alzheimer’s Disease Neuroimaging Initiative (ADNI) were used to investigate whether grouping data by acquisition hardware (i.e. vendor, field strength, coil system) is beneficial for the performance of a support vector machine (SVM) classifier, compared to the case where data from different hardware is mixed. We compared the change of the SVM decision value resulting from (a) changes in hardware against the effect of disease and (b) changes resulting simply from rescanning the same subject on the same machine.
Maximum accuracy of 87% was obtained with a training set of all 417 subjects. Classifiers trained with 95 subjects in each diagnostic group and acquired with heterogeneous scanner settings had an empirical detection accuracy of 84.2±2.4% when tested on an independent set of the same size. These results mirror the accuracy reported in recent studies. Encouragingly, classifiers trained on images acquired with homogenous and heterogeneous hardware settings had equivalent cross-validation performances. Two scans of the same subject acquired on the same machine had very similar decision values and were generally classified into the same group. Higher variation was introduced when two acquisitions of the same subject were performed on two scanners with different field strengths. The variation was unbiased and similar for both diagnostic groups. The findings of the study encourage the pooling of data from different sites to increase the number of training samples and thereby improving performance of disease classifiers. Although small, a change in hardware could lead to a change of the decision value and thus diagnostic grouping. The findings of this study provide estimators for diagnostic accuracy of an automated disease diagnosis method involving scans acquired with different sets of hardware. Furthermore, we show that the level of confidence in the performance estimation significantly depends on the size of the training sample, and hence should be taken into account in a clinical setting.
Magnetic resonance imaging; MRI; Support vector machines (SVM); Alzheimer’s disease; Multi-site study
Progressive supranuclear palsy (PSP) is associated with pathological changes along the dentatorubrothalamic tract and in premotor cortex. We aimed to assess whether functional neural connectivity is disrupted along this pathway in PSP, and to determine how functional changes relate to changes in structure and diffusion. Eighteen probable PSP subjects and 18 controls had resting-state (task-free) fMRI, diffusion tensor imaging and structural MRI. Functional connectivity was assessed between thalamus and the rest of the brain, and within the basal ganglia, salience and default mode networks (DMN). Patterns of atrophy were assessed using voxel-based morphometry, and patterns of white matter tract degeneration were assessed using tract-based spatial statistics. Reduced in-phase functional connectivity was observed between the thalamus and premotor cortex including supplemental motor area (SMA), striatum, thalamus and cerebellum in PSP. Reduced connectivity in premotor cortex, striatum and thalamus were observed in the basal ganglia network and DMN, with subcortical salience network reductions. Tract degeneration was observed between cerebellum and thalamus and in superior longitudinal fasciculus, with grey matter loss in frontal lobe, premotor cortex, SMA and caudate. SMA functional connectivity correlated with SMA volume and measures of cognitive and motor dysfunction, while thalamic connectivity correlated with degeneration of superior cerebellar peduncles. PSP is therefore associated with disrupted thalamocortical connectivity that is associated with degeneration of the dentatorubrothalamic tract and the presence of cortical atrophy.
Resting state fMRI; functional connectivity; white matter tracts; atrophy; dentatorubrothalamic tract
To examine default mode and salience network functional connectivity as a function of APOE ε4 status in a group of cognitively normal age, gender and education-matched older adults.
Fifty-six cognitively normal APOE ε4 carriers and 56 age, gender and education-matched cognitively normal APOE ε4 non-carriers.
Main Outcome Measure
Alterations in in-phase default mode and salience network connectivity in APOE ε4 carriers compared to APOE ε4 non-carriers ranging from 63 to 91 years of age.
A posterior cingulate seed revealed decreased in-phase connectivity in regions of the posterior default mode network that included the left inferior parietal lobe, left middle temporal gyrus, and bilateral anterior temporal lobes in the ε4 carriers relative to APOE ε4 non-carriers. An anterior cingulate seed showed greater in-phase connectivity in the salience network, including the cingulate gyrus, medial prefrontal cortex, bilateral insular cortex, striatum, and thalamus in APOE ε4 carriers vs. non-carriers. There were no group-wise differences in brain anatomy.
We found reductions in posterior default mode network connectivity but increased salience network connectivity in elderly cognitively normal APOE ε4 carriers relative to APOE ε4 non-carriers at rest. The observation of functional alterations in connectivity in the absence of structural changes between APOE e4 carriers and non-carriers suggests that alterations in connectivity may have the potential to serve as an early biomarker.
Resting-state functional magnetic resonance imaging (fMRI) is emerging as an interesting biomarker for measuring connectivity of the brain in patients with Alzheimer's disease (AD). In this review, we discuss the origins of resting-state fMRI, common methodologies used to extract information from these four-dimensional fMRI scans, and important considerations for the analysis of these scans. Then we present the current state of knowledge in this area by summarizing various AD resting-state fMRI studies presented in the first section and end with a discussion of future developments and open questions in the field.
To empirically assess the concept that Alzheimer’s disease (AD) biomarkers significantly depart from normality in a temporally ordered manner.
Multi-site, referral centers
We studied 401 elderly cognitively normal (CN), Mild Cognitive Impairment (MCI) and AD dementia subjects from the Alzheimer’s Disease Neuroimaging Initiative. We compared the proportions of three AD biomarkers – CSF Aβ42, CSF total tau (t-tau), and hippocampal volume adjusted by intra-cranial volume (HVa) - that were abnormal as cognitive impairment worsened. Cut-points demarcating normal vs. abnormal for each biomarker were established by maximizing diagnostic accuracy in independent autopsy samples.
Main Outcome measures
Within each clinical group in the entire sample (n=401) CSF Aβ42 was abnormal more often than t-tau or HVa. Among the 298 subjects with both baseline and 12 month data, the proportion of subjects with abnormal Aβ42 did not change from baseline to 12 months in any group. The proportion of subjects with abnormal t-tau increased from baseline to 12 months in CN (p=0.05) but not in MCI or dementia. In 209 subjects with abnormal CSF AB42 at baseline, the percent abnormal HVa, but not t-tau, increased from baseline to 12 months in MCI.
Reduction in CSF Aβ42 denotes a pathophysiological process that significantly departs from normality (i.e., becomes dynamic) early, while t-tau and HVa are biomarkers of downstream pathophysiological processes. T-tau becomes dynamic before HVa, but HVa is more dynamic in the clinically symptomatic MCI and dementia phases of the disease than t-tau.
Alzheimer’s disease biomarkers; Magnetic Resonance Imaging; CSF tau; CSF Abeta; Alzheimer’s disease staging
Task-free functional magnetic resonance imaging (TF-fMRI) has great potential for advancing the understanding and treatment of neurologic illness. However, as with all measures of neural activity, variability is a hallmark of intrinsic connectivity networks (ICNs) identified by TF-fMRI. This variability has hampered efforts to define a robust metric of connectivity suitable as a biomarker for neurologic illness. We hypothesized that some of this variability rather than representing noise in the measurement process, is related to a fundamental feature of connectivity within ICNs, which is their non-stationary nature. To test this hypothesis, we used a large (n = 892) population-based sample of older subjects to construct a well characterized atlas of 68 functional regions, which were categorized based on independent component analysis network of origin, anatomical locations, and a functional meta-analysis. These regions were then used to construct dynamic graphical representations of brain connectivity within a sliding time window for each subject. This allowed us to demonstrate the non-stationary nature of the brain’s modular organization and assign each region to a “meta-modular” group. Using this grouping, we then compared dwell time in strong sub-network configurations of the default mode network (DMN) between 28 subjects with Alzheimer’s dementia and 56 cognitively normal elderly subjects matched 1∶2 on age, gender, and education. We found that differences in connectivity we and others have previously observed in Alzheimer’s disease can be explained by differences in dwell time in DMN sub-network configurations, rather than steady state connectivity magnitude. DMN dwell time in specific modular configurations may also underlie the TF-fMRI findings that have been described in mild cognitive impairment and cognitively normal subjects who are at risk for Alzheimer’s dementia.