Michael D. Greicius, M.D., M.P.H. - US grants

DESCRIPTION (provided by applicant): The long-term objective of this Mentored Patient-Oriented Research Career Development (K23) Award is to develop the candidate's skills in applying functional imaging to patient-oriented clinical research so that he may become an independent investigator in the fields of dementia and neuroimaging. Dr. Greicius' specific research goal is to develop proficiency in the application of neuroimaging techniques to the early and accurate diagnosis of Alzheimer's disease (AD). To accomplish this goal, the candidate will be mentored by experts in the fields of functional imaging, aging and cognitive disorders, neuropsychology and statistics. In addition, the candidate will pursue an educational program to include advanced seminars in behavioral neurology and functional imaging as well as formal coursework in the areas of computer programming, neuropsychology and statistics. Finally, the candidate will undertake a research project closely aligned with his research training plan. The goal of the proposed research project is to develop a novel fMRI technique that will predict, with a high degree of certainty, which subjects at risk for AD will subsequently develop the disease. The study will use a novel fMRI approach, functional connectivity MRI (fcMRI), to examine three subject groups: A group with mild cognitive impairment (MCI), a group with mild-moderate AD, and a group of healthy older adults age- and gender-matched to subjects in the AD and MCI groups. The MCI group will be imaged at baseline and followed longitudinally with clinical and neuropsychological evaluations. The functional connectivity analysis will focus on a specific resting-state network (RSN) that incorporates several brain regions affected early in the course of AD including the posterior cingulate cortex (PCC), inferior parietal lobes, and hippocampus. The main hypotheses are that 1) functional connectivity analyses will demonstrate abnormal RSN activity in the AD group compared to age-matched controls and 2) abnormal activity in the RSN will reliably distinguish MCI patients who convert to AD from those MCI patients who do not. By conducting this research project and implementing the research training plan, Dr. Greicius plans to become an independent investigator in the fields of dementia and neuroimaging and contribute to the early detection and treatment of AD and other dementing diseases of the elderly.

DESCRIPTION (provided by applicant): More than 100 years after it was first described, Alzheimer's disease (AD) is still diagnosed strictly on clinical criteria that are not sensitive to the early stages of disease and do not adequately distinguish AD from non-AD dementias. When comparing a clinical diagnosis to an autopsy-confirmed diagnosis, even the most seasoned clinicians are wrong 15-20% of the time. This diagnostic inaccuracy is assumed to be considerably worse in non-specialty settings where the initial diagnosis of AD is most often made. With advances in experimental therapeutics and the concomitant reminder that potent treatments may have serious side effects, the need for an accurate, non-invasive AD biomarker is more pressing than ever. Such a biomarker would be useful on several fronts. In the clinical setting it would allow for more diagnostic certainty in trying to distinguish AD from other dementias. An accurate biomarker with sufficient sensitivity should also help predict which patients with mild cognitive impairment (MCI) will go on to develop AD and, just as importantly, which will not. Lastly, an AD biomarker that detects disease in the earliest stages and tracks with clinical status would accelerate drug development by facilitating dose-response studies and enabling more rapid and objective assessment of efficacy. Despite considerable efforts, the field has yet to develop a biomarker that can meet these pressing needs. The current application will examine a relatively novel form of functional MRI (fMRI) as a candidate imaging biomarker in AD. Resting-state fMRI provides a measure of functional connectivity within specific brain networks and has shown promise in preliminary studies as an AD biomarker. The limitations of this approach currently are that it has not yet proven to be reliably interpretable at the single-subject level, its predictive value in MCI remains uncertain, and it has not been examined longitudinally. The current application, drawing on the strengths of a multi-site, longitudinal study, will attempt to address these limitations. The study will involve the acquisition of resting-state fMRI data from Stanford University and the University of California, San Francisco in four large cohorts of subjects: healthy aging, MCI, AD, and non-AD dementia. Subjects will be scanned at baseline and followed longitudinally. A subset of subjects will be scanned again at a 1-year interval. The aims of the study will be a) to enhance the sensitivity and specificity of resting-state functional connectivity measures in distinguishing AD from both healthy aging and non-AD dementia, b) to assess the utility of resting-state fMRI in predicting which patients with MCI subsequently convert to AD over the five-year course of the study and c) to assess the utility of resting-state fMRI in tracking disease progression over time.

Project Summary: Core F The overall goals of the Imaging Core are to obtain the highest caliber structural and functional imaging data on ADRC participants; to make these data available to a wide spectrum of investigators at Stanford; and to make the Alzheimer's Disease Neuroimaging Initiative (ADNI) data more accessible to investigators from non- imaging backgrounds. The Imaging Core, capitalizing on Stanford's strength in imaging brain networks, will focus on measures of functional and structural connectivity. Dr. Glover and Dr. Bammer, Core Co- Investigators, have made critical contributions to the acquisition and analysis of resting-state fMRI and diffusion-weighted imaging, respectively. Dr. Greicius, the Core Director, has made important advances in the application of functional connectivity measures to the study of Alzheimer's disease. A second major focus of the Imaging Core will be to make the acquired data available to and readily used by researchers across disciplines and schools at Stanford. To this end, the Imaging Core will take advantage of the neurobiological imaging management system (NIMS) built by Dr. Dougherty, the third Core Co-Investigator. NIMS has been in place for more than two years and provides a user-friendly interface for investigators to download brain imaging, and related, data. The Imaging Core will use the NIMS infrastructure to provide researchers with subject imaging data in three stages of processing: raw, processed, and abstracted summary measures. The imaging data will be linked to subjects' ancillary data (including neuropsychological measures, spinal fluid proteins, plasma proteomics, etc.) This will allow the widest possible set of Stanford investigators, with or without imaging expertise, to test hypotheses using the ADRC data. In parallel with the imaging data acquired through the ADRC, the Imaging Core will provide investigators across the Stanford campus with a curated, turnkey version of the ADNI dataset. Dr. Greicius' imaging group has extensive experience with the ADNI dataset and has developed a pipeline for quality control and data processing. The ADNI imaging data and associated ancillary data such as spinal fluid proteins, genotypes, etc. will also be available on NIMS. As with the ADRC data, this should greatly amplify the number of researchers who can benefit from the incredibly rich, multimodal ADNI data.

The Organization for Human Brain Mapping (OHBM) is the primary international organization dedicated to non-invasive neuroimaging research and the functional organization of the human brain. The Annual Meeting of OHBM is regarded as the premier venue for the integration of innovative brain imaging methods and cognitive neuroscience. It was started more than 20 years ago to provide a forum for the brain-mapping scientific community to disseminate findings and enable interactions among scientists investigating the functional organization of the brain with emerging imaging methods. Since 1994, the OHBM has sponsored twenty-two highly successful meetings, where attendees are exposed to cutting-edge neuroimaging data acquisition methods, emerging approaches to large-scale neuroimaging data analysis, the visualization of results, and their applications in health and disease. This award will fund travel awards for deserving students and trainees to attend the 2017 Annual Meeting to be held in Vancouver, Canada. Travel awards will be given to 30 students with high-ranking abstracts. Of these at least 15 travel awards will be awarded to the students with the top-ranked abstracts among 3 targeted groups (women, minorities, disabled.

The Annual Meeting provides a means for students of neuroimaging to attend educational courses, hear lectures from leading brain researchers, form new collaborations, and to present their original research. Historically, female representation in the mathematical and computational sciences has been low. In contrast, in both neuroscience and psychology it has been relatively very high. Hence, the ability to work in the cross-section of statistics, computer-science, psychology, and neuroscience promises to provide computational training to a diverse audience and help bridge the gender gaps in the STEM fields. In addition, improved understanding of the organization of the human brain is directly relevant to treating neurological and psychiatric disease, and the use of non-invasive imaging methods is increasingly important to translational investigation and training in clinical neuroscience. As neuroimaging is used in a wide-range of scientific disciplines, we would expect benefits for society in areas ranging from economic and social policies to medical, educational and psychological interventions.

Project Summary/Abstract Alzheimer's disease (AD) is a common, progressive, and ultimately fatal brain disease. Currently approved treatments provide only minimal symptomatic benefits and do not stop the disease from progressing. The field is in dire need of novel drug targets which could lead to disease-modifying therapies. The most common genetic risk factor for AD is the ?4 variant of the apolipoprotein E gene (APOE4). The current study will take advantage of the strong effect of APOE4 to discover new genetic variants that either increase risk for AD or reduce risk for AD. The research team?based at Stanford University but including collaborators at 13 other research centers?will recruit and study participants that fall into one two rare categories: cognitively normal people carrying one or two copies of the high risk APOE4 gene (protected APOE4 carriers) and young patients who early-onset AD (EOAD) before age 65 despite not carrying APOE4 gene. These subjects, the Stanford Extreme Phenotypes in AD (StEP AD) cohort, will undergo whole-exome sequencing and their exomes will be combined with large, publicly available exomes from ~4500 healthy older controls and ~5000 AD patients. In Aim 1 the research team will look for rare genetic variants seen more often in protected APOE4 carriers than in AD patients. In Aim 2 the team will look for rare genetic variants seen in APOE4-negative EOAD patients but not in healthy older controls. Most of the StEP AD cohort will undergo ?deep phenotyping? to include structural and molecular brain imaging, spinal fluid analysis, immunophenotyping, and culturing of participant-specific neurons. In Aim 3, the deep phenotyping data will be used to begin to understand the molecular effects of the rare protective or causal genetic variants identified in Aims 1 and 2. Rare but powerful genetic variants identified and characterized in this study will provide novel drug targets for the design of potentially disease- modifying treatments.

1. SUMMARY (Imaging Core) The overall goals of the Imaging Core are to obtain amyloid PET and high caliber MRI data to characterize Stanford ADRC participants along the AD and PD spectrums; to make these data available to a wide range of investigators at Stanford; and to provide training on the acquisition, interpretation, and analysis of imaging data relevant to neurodegenerative disease. The Imaging Core, capitalizing on Stanford's strength in brain imaging and radiochemistry, will focus on incorporation of amyloid PET imaging in conjunction with structural and functional MRI, and enable innovation for the exploration of novel PET targets. Dr. Greicius, the Core Director, has helped advance the application of functional connectivity measures to the study of AD. Dr. Mormino, the co-Director of the Imaging Core, has made contributions to characterizing the preclinical stage of AD using amyloid and tau PET. A second major focus of the Imaging Core will be to make the acquired data available to and readily used by researchers across disciplines and schools at Stanford. The Imaging Core will use the RedTree infrastructure to provide researchers with summary imaging data for PET (global and regional amyloid and tau PET values), structural measures (regional gray matter volume and thickness values), and resting state functional measures (network connectivity). The imaging data will be linked to subjects' ancillary data (including neuropsychological measures, spinal fluid proteins, plasma proteomics, etc.). This will allow the widest possible set of Stanford investigators, with or without imaging expertise, to test hypotheses using the ADRC data. The availability of amyloid PET and MRI data on the majority of clinical core participants will greatly enhance diagnostic accuracy which will fundamentally strengthen any clinical research undertaken at the Stanford ADRC. These imaging data, in conjunction with biofluid data from the Biomarker Core, will also enable classification according to the recent research framework of AD so that Stanford ADRC data can be more readily harmonized with data from other centers.

PROJECT SUMMARY/ABSTRACT Alzheimer?s disease (AD) is a common, progressive, and ultimately fatal brain disease. Currently approved treatments provide only minimal symptomatic benefits and do not stop the disease from progressing. The field is in dire need of novel drug targets which could lead to disease-modifying therapies. The most common genetic risk factor for AD is the ?4 variant of the apolipoprotein E gene (APOE4). The effect of APOE4 varies greatly between people of African ancestry and people of European ancestry. The current study?Illuminating the APOE Locus with Long-Read Sequencing and Targeted Genomics?will apply a new genome sequencing technology (long-read sequencing) to the study of APOE and several other AD-relevant genes including ABCA7. Long-read sequencing will be performed on DNA from roughly 2000 African-Americans with AD and 2000 healthy older African-American control subjects as well as DNA from roughly 5000 European-American AD patients and 5000 European-American controls. A subset of these patients will also have long-read sequencing of these genes? RNA derived from white blood cells, fibroblasts, or brain tissue. These analyses will help us understand how local genetic variants near the APOE4 variant can alter the type or amount of the APOE4 protein and how this affects risk of AD. Similar analyses will be done on ABCA7 and another 15-20 targeted genes that will be selected just before sequencing begins and following an up-to-date review of the AD genetics literature. In addition to understanding the local variants regulating a gene and the protein it produces, long-read sequencing will be useful in detecting large, damaging genetic mutations that are easily missed with standard whole-genome sequencing. The results will allow for more specific estimates of AD risk in individuals of diverse ancestral backgrounds and will provide novel targets for drug development.