MRI Imaging may identify a biomarker for early detection of Alzheimer’s disease

Anthony Wagner, Ph.D.

Stanford University School of Medicine

Funded in September, 2014: $200000 for 3 years


back to top

MRI Imaging may identify a biomarker for early detection of Alzheimer’s disease

Investigators will use high-powered MRI imaging in older adults to see whether thinning of a tiny ribbon of tissue in the brain’s hippocampus is a biomarker for pre-symptomatic Alzheimer’s disease (AD) and can be used for screening and early detection of AD.  

The two hallmarks of AD are plaques (build-up of the protein “amyloid” between brain cells) and tangles (build-up of the protein “tau” within cells).  Memory symptoms start being recognized when the accumulation of these plaques and tangles are sufficient to interrupt communication between cells in the brain’s hippocampus, where new memories are first formed. Yet that is relatively late in the disease process.

Microscopic signs of plaques and tangles are already evident in autopsies of most people who died in their early 60’s, so scientists know that the physical brain changes that lead to mild cognitive impairment (MCI) and eventually full-blown AD begin long before memory problems are noticed. Scientists are racing, therefore, to find ways to identify AD pre-symptomatically, when experimental treatments to arrest or slow accumulation of the plaques and tangles are most likely to be effective.

A possible biomarker for identifying AD early has been identified in adults with MCI by Stanford investigators using powerful 7T MRI imaging that has exceptionally strong resolution and can visualize microscopic changes. Previously in autopsy tissue studies of adults with AD who died, scientists had found a narrow band of tissue—less than a millimeter wide—within the hippocampus that early on is shriveled and has widespread loss of synapses, which are used by brain cells to communicate with one another.

Stanford investigators used 7T MRI to image this tiny area in people with MCI. They found that the degree of shriveling of this narrow tissue band strongly correlates with a patient’s memory performance—the narrower the band, the worse the memory. Additionally, patients with the narrowest tissue bands had more tau protein in their cerebral spinal fluid (CSF), suggesting that the bands contain high amounts of tau, which is implicated in disease progression. In aggregate, these factors suggest that the spread of tau tangles into this narrow band of hippocampal tissue and their destruction of brain cells leading to tissue thinning is a threshold event.  

The investigators hypothesize, therefore, that atrophy of this ribbon of tissue will be a reliable pre-clinical biomarker of early AD and that 7T-MRI can detect tissue narrowing in adults who are in the pre-symptomatic stage of AD.  To test their hypothesis, they will undertake 7T MRI in 100 adults who are cognitively healthy according to neuropsychological testing. They also will measure participants’ CSF tau levels and cognitive performance. They anticipate that 7T MRI will detect differences in tissue band thinning among the 100 participants, and that those with greater tissue atrophy will also have increased CSF tau and subtle cognitive deficiencies evidenced in challenging neuropsychological tests.  If so, the findings from this “cross-sectional” study (studying a group of people at one point in time) will provide initial evidence of this potential biomarker.

Additionally, since 7T MRI is available only in specialized academic health settings, investigators also will test participants using widely available 3T MRI to see if it too is sensitive enough to replicate findings as seen on powerful 7T MRI.  Following this cross-sectional study, researchers then would apply to NIH for funding for a “longitudinal” study to follow participants with MRI imaging over time to validate that this tissue thinning is a biomarker for early AD.

Significance:      If research ultimately validates that MRI imaging can reveal an AD biomarker, the research would lead to screening strategies to predict and detect AD at the earliest stage, when therapy is most likely to have an effect.     


back to top

Detection of Pre-Symptomatic Alzheimer's Disease Using 7-Tesla MRI

In Alzheimer’s disease (AD), pathological changes in the brain precede symptoms by many years. Newly emerging therapies have failed to help patients who already exhibit signs of dementia, fueling an urgent push for technologies that can detect AD in an early, preclinical stage. Aggregation of the microtubule-associated protein tau into neurofibrillary tangles is one of the hallmark neuropathological findings in AD. At autopsy, a large proportion of 60-year-olds exhibit the first signs of tau pathology; however, very few have symptoms at that age, implying a large reservoir of presymptomatic disease. Tau pathology propagates in an ordered fashion, beginning in the medial temporal lobe. After invading the entorhinal cortex, tangles appear in the hippocampal CA1 subfield, specifically in the CA1 stratum radiatum / stratum lacunosum-moleculare (CA1-SRLM), even before symptoms start. Spread of tangles to CA1-SRLM is a threshold event; postmortem tissue analysis reveals that synapse loss and atrophy eventually overtake this area and correlate with premortem memory function. CA1-SRLM is a narrow band of tissue measuring about 700 µm wide in a healthy young adult. Despite its small size, I have used 7.0-Tesla (7T) MRI to obtain high-resolution (220 µm in-plane) structural images of the hippocampus to observe CA1-SRLM atrophy in vivo among patients with mild AD dementia relative to age-matched controls. The width of CA1-SRLM correlates robustly with episodic memory performance among patients with mild AD dementia or amnestic Mild Cognitive Impairment, a prodromal form of AD. It is thinner, on average, in cognitively-healthy older adults than in young adults, consistent with a degree of preclinical pathological burden. Finally, preliminary data point to a specific correlation between CA1-SRLM width and cerebrospinal fluid (CSF) tau levels, consistent with the hypothesis that CA1-SRLM atrophy reflects tau-based pathology. What is missing is a large, definitive study of cognitively-healthy older adults, to determine whether CA1-SRLM atrophy is indeed a reliable preclinical marker of AD. Here, I propose to study a cohort of 100 adults aged 60 years and older who all score normally on a neuropsychological testing battery. Each will undergo 7T MRI and 3T MRI, as well as a lumbar puncture for AD protein biomarkers, including tau and phospho-tau levels. In Aim 1, I will test the hypothesis that CA1-SRLM width correlates with CSF tau, in order to validate the notion that this neuroimaging metric meaningfully reflects underlying AD-related pathology even in the absence of clinical symptoms. Behavioral outcomes hold the most significance in a clinical study, and in Aim 2, I will take advantage of the fact that despite normal performance on episodic memory tasks, these unimpaired older adults will nevertheless exhibit a range of below average to above average scores, allowing me to test the hypothesis that CA1-SRLM width (and likely CSF tau levels) will account for some of this variance. In other words, I predict that this neuroimaging metric will sensitively correlate to subtle evidence of memory dysfunction. Finally, in Aim 3, I will develop methods to measure CA1-SRLM width using 3T MRI data, to test whether 7T MRI findings may be replicated at a lower field strength, allowing translation of these discoveries to a widely-available platform. This work aims to contribute to a future in which presymptomatic detection of AD is routine, facilitating delivery of novel treatments to patients who stand the most to benefit.