Helen Hochstetler

Comparison of neuroimaging modalities for the prediction of conversion from mild cognitive impairment to Alzheimer's dementia.

In this study we compared Pittsburgh compound-B (PIB) positron emission tomography (PET) amyloid imaging, fluorodeoxyglucose PET for metabolism, and magnetic resonance imaging (MRI) for structure to predict conversion from amnestic mild cognitive impairment (MCI) to Alzheimers dementia using data from the Alzheimers Disease Neuroimaging Initiative cohort. Numeric neuroimaging variables generated by the Alzheimers Disease Neuroimaging Initiative-funded laboratories for each neuroimaging modality along with apolipoprotein-E genotype (n = 29) were analyzed. Performance of these biomarkers for predicting conversion from MCI to Alzheimers dementia at 2 years was evaluated in 50 late amnestic MCI subjects, 20 of whom converted. Multivariate modeling found that among individual modalities, MRI had the highest predictive accuracy (67%) which increased by 9% to 76% when combined with PIB-PET, producing the highest accuracy among any biomarker combination. Individually, PIB-PET generated the best sensitivity, and fluorodeoxyglucose PET had the lowest. Among individual brain regions, the temporal cortex was found to be most predictive for MRI and PIB-PET.

Clinical use of amyloid-positron emission tomography neuroimaging: Practical and bioethical considerations

Until recently, estimation of β‐amyloid plaque density as a key element for identifying Alzheimers disease (AD) pathology as the cause of cognitive impairment was only possible at autopsy. Now with amyloid‐positron emission tomography (amyloid‐PET) neuroimaging, this AD hallmark can be detected antemortem. Practitioners and patients need to better understand potential diagnostic benefits and limitations of amyloid‐PET and the complex practical, ethical, and social implications surrounding this new technology. To complement the practical considerations, Eli Lilly and Company sponsored a Bioethics Advisory Board to discuss ethical issues that might arise from clinical use of amyloid‐PET neuroimaging with patients being evaluated for causes of cognitive decline. To best address the multifaceted issues associated with amyloid‐PET neuroimaging, we recommend this technology be used only by experienced imaging and treating physicians in appropriately selected patients and only in the context of a comprehensive clinical evaluation with adequate explanations before and after the scan.

Association Between Clinical Measures and Florbetapir F18 PET Neuroimaging in Mild or Moderate Alzheimer’s Disease Dementia

Clinical diagnosis of Alzheimers disease (AD) is challenging, with 20% or more of patients misdiagnosed, even by expert clinicians. The authors conducted a retrospective, cross-sectional analysis comparing baseline neuropsychiatric and other clinical characteristics in 199 expert-diagnosed mild and moderate AD dementia patients participating in industry-sponsored clinical trials of an investigational therapy, where 18% lacked florbetapir positron emission tomography (PET) evidence of AD neuropathology. Significant differences were found only for cognition and ApoE ε4 status, but the large degree of score overlap would preclude using these measures to predict AD misdiagnosis. This study highlights the value of amyloid PET when evaluating patients with seemingly typical AD.

Relationship of Hippocampal Volume to Amyloid Burden across Diagnostic Stages of Alzheimer's Disease.

Aims: To assess how hippocampal volume (HV) from volumetric magnetic resonance imaging (vMRI) is related to the amyloid status at different stages of Alzheimers disease (AD) and its relevance to patient care. Methods: We evaluated the ability of HV to predict the florbetapir positron emission tomography (PET) amyloid positive/negative status by group in healthy controls (HC, n = 170) and early/late mild cognitive impairment (EMCI, n = 252; LMCI, n = 136), and AD dementia (n = 75) subjects from the Alzheimers Disease Neuroimaging Initiative Grand Opportunity (ADNI-GO) and ADNI2. Logistic regression analyses, including elastic net classification modeling with 10-fold cross-validation, were used with age and education as covariates. Results: HV predicted amyloid status only in LMCI using either logistic regression [area under the curve (AUC) = 0.71, p < 0.001] or elastic net classification modeling [positive predictive value (PPV) = 72.7%]. In EMCI, age (AUC = 0.70, p < 0.0001) and age and/or education (PPV = 63.1%), but not HV, predicted amyloid status. Conclusion: Using clinical neuroimaging, HV predicted amyloid status only in LMCI, suggesting that HV is not a biomarker surrogate for amyloid PET in clinical applications across the full diagnostic spectrum.

Empirically Defining Trajectories of Late-Life Cognitive and Functional Decline.

Background: Alzheimer’s disease (AD) is associated with variable cognitive and functional decline, and it is difficult to predict who will develop the disease and how they will progress. Objective: This exploratory study aimed to define latent classes from participants in the Alzheimer’s Disease Neuroimaging Initiative (ADNI) database who had similar growth patterns of both cognitive and functional change using Growth Mixture Modeling (GMM), identify characteristics associated with those trajectories, and develop a decision tree using clinical predictors to determine which trajectory, as determined by GMM, individuals will most likely follow. Methods: We used ADNI early mild cognitive impairment (EMCI), late MCI (LMCI), AD dementia, and healthy control (HC) participants with known amyloid-β status and follow-up assessments on the Alzheimer’s Disease Assessment Scale - Cognitive Subscale or the Functional Activities Questionnaire (FAQ) up to 24 months postbaseline. GMM defined trajectories. Classification and Regression Tree (CART) used certain baseline variables to predict likely trajectory path. Results: GMM identified three trajectory classes (C): C1 (n = 162, 13.6%) highest baseline impairment and steepest pattern of cognitive/functional decline; C3 (n = 819, 68.7%) lowest baseline impairment and minimal change on both; C2 (n = 211, 17.7%) intermediate pattern, worsening on both, but less steep than C1. C3 had fewer amyloid- or apolipoprotein-E ɛ4 (APOE4) positive and more healthy controls (HC) or EMCI cases. CART analysis identified two decision nodes using the FAQ to predict likely class with 82.3% estimated accuracy. Conclusions: Cognitive/functional change followed three trajectories with greater baseline impairment and amyloid and APOE4 positivity associated with greater progression. FAQ may predict trajectory class.

EMPIRICALLY DEFINING TRAJECTORIES OF LATE-LIFE COGNITIVE AND FUNCTIONAL DECLINE

OBJECTIVE: Define cognitive and functional decline trajectories using Growth Mixture Modeling (GMM) and determine predictors of potential class membership using Classification and Regression Trees (CART). BACKGROUND: Considerable variability exists in rate of cognitive and functional decline in Alzheimer’s Disease (AD), impeding forecasting an individual’s clinical course. Early prediction of clinical trajectory will permit maximizing care planning and potentially delay decline. DESIGN/METHODS: Subjects were elderly healthy controls (HC, n=325), persons with early mild cognitive impairment (EMCI, n=279), late MCI (LMCI, n=372), or AD dementia (n=216) from the Alzheimer’s Disease Neuroimaging Initiative (ADNI-1/GO/2), using first-available β-amyloid measurement ([ 11 C]PiB or [ 18 F]florbetapir PET SUVR, or cerebrospinal fluid Aβ 1-42 ) as baseline plus up to 24 months assessments using Alzheimer’s Disease Assessment Scale-Cognitive (ADAS-Cog 13 ) and Functional Activities Questionnaire (FAQ). GMM was conducted jointly on ADAS-Cog 13 and FAQ. Using baseline variables (e.g.; diagnosis, demographics, co-morbidities, FAQ, amyloid and APOE4 status) CART predicted likely class membership. RESULTS: GMM identified 3 trajectory classes: C1 (n=162, 13.6[percnt]) with highest baseline FAQ and ADAS-Cog 13 , and similar rates of cognitive and functional worsening; C2 (n=211, 17.7[percnt]) with intermediate baseline FAQ and ADAS-Cog 13 , but steeper functional than cognitive decline; C3 (n=819, 68.7[percnt]) with lowest baseline FAQ and ADAS-Cog 13 and minimal changes over time. C3 was least likely to be amyloid or APOE4 positive and most likely to involve HCs or EMCI cases. CART yielded 2 decision nodes for likely class membership: FAQ蠅13.5 for C1 (n=115, 9.6[percnt]), then either 4.5蠅FAQ<13.5 for C2 (n=241, 11.8[percnt]) or FAQ<4.5 for C3 (n=836; 70.1[percnt]). CONCLUSIONS: The higher the baseline FAQ and ADAS-Cog 13 scores, the more rapidly the class progressed, regardless of baseline diagnosis. Amyloid and APOE4 positive status were associated with faster decline. FAQ, easily performed in the clinic, was identified by CART as the only predictor for trajectory classes with 82.3[percnt] accuracy. Study Supported by: Lilly Disclosure: Dr. Hochstetler has received personal compensation for activities with Eli Lilly & Co. as an employee. Dr. Trzepacz has received personal compensation for activities with Eli Lilly & Co. as an employee. Dr. Wang has received personal compensation for activities with Eli Lilly & Company as an employee. Dr. Yu has received personal compensation for activities with Eli Lilly & Company as an employee. Dr. Case has received personal compensation for activities with Eli Lilly & Company as an employee. Dr. Leoutsakos has nothing to disclose. Dr. Henley has received personal compensation for activities with Eli Lilly & Co. as an employee. Dr. Henley holds stock and/or stock options in Eli Lilly & Co. Dr. Degenhardt has received personal compensation for activities with Eli Lilly & Company as an employee. Dr. Degenhardt holds stock and/or stock options in Eli Lilly & Company. Dr. Lyketsos has received personal compensation for activities with Pfizer Inc., Forest Laboratories, Eli Lilly & Company, Takeda Pharmaceutical Company, and Elan as a consultant.

IS FLORBETAPIR-PET OCCIPITAL SUVR A LATE BIOMARKER IN MILD OR MODERATE AD DEMENTIA AS COMPARED TO HIPPOCAMPAL VOLUME?

P4-311 IS FLORBETAPIR-PET OCCIPITAL SUVR A LATE BIOMARKER IN MILD OR MODERATE AD DEMENTIA AS COMPARED TO HIPPOCAMPAL VOLUME? Michael Witte, Peng Yu, Shufang Wang, Peter Castelluccio, Helen Hochstetler, Abhinay Joshi, Grazia Dell’Agnello, David Henley, Elisabeth Degenhardt, Sara Kollack Walker, Michael D. Devous, Sr,, Adam Devous, Sr., Paula Trzepacz, Eli Lilly and Company, Indianapolis, Indiana, United States; Eli Lilly and Company, Indianapolis, Indiana, United States; Bucher & Christian Consulting, Inc., Philadelphia, Pennsylvania, United States; Avid Radiopharmaceuticals, Philadelphia, Pennsylvania, United States; Eli Lilly Italia, Sesto Fiorentino, Italy; Eli Lilly and Company, Indianapolis, Indiana, United States; Avid Radiopharmaceuticals, Inc., Philadelphia, Pennsylvania, United States. Contact e-mail: witte_michael_m@lilly.com

Comparison of 11C-PiB-PET, FDG-PET and MRI modalities for the prediction of conversion from mild cognitive impairment to Alzheimer's dementia

will provide additional information about level of tracer uptake and may also help enable consistency in image assessment over time and across different centers. In this study, we investigate the performance of a fully automated quantification method by calculating sensitivity and specificity against histopathology and by comparing scan categorization using quantification and visual evaluation, respectively. Methods: [18 F]Flutemetamol images from 10 prior studies were used. A total of 276 scans were used from 172 AD, MCI and healthy volunteer (HV) subjects, 36 normal pressure hydrocephalus (NPH) subjects, and 68 end-of-life subjects from an autopsy trial, where a histopathology standard of truth (SOT) was available. Five visual readers, trained with an electronic training program, categorized all scans. A fully automated PET-only, MNI-space, quantification method was used to compute SUVR values for a composite neocortical region using pons as reference. An SUVR threshold was derived by ranking the PET scans from the autopsy cohort based on the composite SUVR and comparing data with SOT, a dichotomous Bielschowsky silver stain assessment of neuritic plaques based on CERAD criteria. The derived threshold was used to categorize all 276 scans into normal and abnormal. For the 68 scans from end-of-life subjects, sensitivity and specificity against SOT were computed. For all 276 scans quantification results were compared to categorization using visual assessment. Results: In the autopsy cohort, classification by quantification gave three false-positive and four false-negative scans, yielding 91% sensitivity and 88% specificity. All three false-positive cases were either borderline normal by SOT or had moderate to heavy cortical diffuse plaque burden. There was concordance between quantitative and visual read categorization for 268 of the 276 scans (97.1%). After excluding the autopsy scans, many of which were atypical with gross atrophy making both visual assessment and quantification difficult, quantitative and visual read categorization agreed in 206 of 208 remaining scans (99.0%). Conclusions: Categorization of [18 F]flutemetamol amyloid imaging data using an automated PET-only quantification method showed good agreement with histopathological classification of neuritic plaque density, and there was a strong concordance with visual read results.

Comparison of florbetapir positron emission tomography scans with cerebrospinal fluid biomarkers in healthy individuals and people with mild cognitive impairment or Alzheimer's disease dementia

differences between CNAband Ab+ subjects in themetabolic ROIs, but not in four volumetric and ten cortical thickness ROIs. Moreover, receiver operating characteristics analyses revealed significant differences between the areas under the curve for metabolic vs. structural ROIs, indicating larger metabolic than structural changes in CN Ab+ subjects. Conclusions: Our findings largely agreewith the proposed temporo-anatomical model of hypometabolic and atrophic changes in AD, supporting the notion that amyloid load may affect synaptic activity leading to synaptic loss.

Modeling of clinical measures to predict florbetapir F18 PET results in mild and moderate Alzheimer's dementia

P1-301 MODELING OF CLINICAL MEASURES TO PREDICT FLORBETAPIR F18 PET RESULTS IN MILD AND MODERATE ALZHEIMER’S DEMENTIA Michael Witte, Paula Trzepacz, Michael Case, Peng Yu, Helen Hochstetler, Sara Kollack Walker, Karen Sundell, David Henley, Eli Lilly and Company, Indianapolis, Indiana, United States; Lilly Research Laboratories, Indianapolis, Indiana, United States. Contact e-mail: witte_michael_m@lilly.com