Columbia University Medical Center

New MRI Approach Can Identify Sources Of Memory Loss In Humans

Regions of altered brain metabolism imaged with MRI can reveal causes of early memory impairment

New York, New York, Dec. 22, 2000—Researchers have found a way to pinpoint changes in brain activity that may underlie memory impairment, even before structural damage occurs. Dr. Scott Small and colleagues report in the December issue of Neuron that with a new high-resolution MRI technique, alterations in resting activity in subregions of the hippocampus, a brain circuit important for learning and memory. By enabling researchers to detect activity changes in mice genetically altered to model age-related memory loss, the approach may further understanding of the mechanisms of the disease in humans.

As a potentially valuable tool for diagnosing memory disorders from specific causes, the use of MRI to precisely map blood oxygenation in the brain at rest could also lead to more effective treatment for memory loss in its early stages. “This is encouraging data. We may be able to use this technique as an early diagnostic for Alzheimer’s disease,” says Dr. Small, Irving assistant professor of Neurology at Columbia University College of Physicians & Surgeons and first author of the paper. Dr. Small collaborated with Dr. Ed X. Wu, Assistant Professor of Radiology; Dr. Dusan Bartsch; Gerard Perera; Clay Lacefield; Dr. Robert DeLaPaz, Professor of Radiology; Dr. Richard Mayeux, Sergievsky Professor of Psychiatry, Neurology and Public Health; Dr. Yaakov Stern, professor of clinical neurology and psychiatry; and Dr. Eric Kandel, University Professor of Physiology & Cell Biophysics, Psychiatry, and Biochemistry & Molecular Biophysics.

The new technique is a variant of functional MRI, an approach frequently used by cognitive scientists to show rapid changes in brain activity during certain tasks in perception or thinking. Usually, researchers investigating how the brain works look at rapid, dynamic changes in blood oxygenation that reveal sudden increases in activity during a particular mental task. Functional MRI gives good information about what general areas are involved during a precise stage of the task but has difficulty pinpointing precise activity to small regions of the brain—such as the hippocampal subregions. “To get fast temporal resolution, you sacrifice spatial resolution,” says Dr. Small.

To precisely locate changes of resting activity over very small regions of the brain, however, Dr. Small and colleagues took a different approach to this trade-off. Stable, disease-related abnormalities in resting oxygenation, or ROXY, can be detected within tiny regions of the brain by analyzing similar MRI signals over longer periods. “In disease, you have static changes, so you can get better spatial resolution; you don’t need temporal resolution,” says Dr. Small. “We felt the technique had the resolution to narrow down damage to specific subregions.”

ROXY also offers advantages to researchers studying memory disorders: using functional MRI to evaluate a patient or subject requires them to understand and follow directions, but resting activity can be measured even with patients in the advanced stages of memory loss or in genetic mouse models of disease. Says Dr. Small, “In memory deficient mice that show similar physiological changes in memory formation, R(AB) mice, Eric Kandel and his colleagues Ted Abel and Rusiko Bourtchouladze found changes in brains, and used histology to look for lesions—there were no structural defects,” demonstrating that the high-resolution ROXY measurements with MRI reveal genetic alterations in activity even without visible tissue damage. Because the molecular workings of memory are well studied in mice, comparing physiological changes in human patients and mouse models might be able to suggest possible molecular mechanisms involved in human memory disorders. Likewise, ROXY imaging might be useful in evaluating other genetic mouse models for human brain disorders. Although human behavioral criteria may be hard to evaluate in mice, changes in physiology in particular brain regions might be easier to compare between patients and mouse models of a disease.

Although specific anatomical changes in the hippocampus are typically seen in the late stages of several forms of memory disorders, detectable structural changes are not prominent in early stages, when the impairment first presents itself. “When an elderly patient presents with mild forgetfulness, right now, we can’t tell if it’s normal or abnormal,” says Dr. Small. “Is it early Alzheimer’s disease, or is it something else? With this technique [using MRI to map resting activity], we may be able to detect Alzheimer’s disease more accurately.” Finding a way to determine exactly what part of the hippocampus isn’t working normally might give early cues about the origin of the memory deficit and, therefore, how best to treat it.

Memory disorders may differ in their underlying causes, as well as appropriate treatments. “Some mechanisms can cause structural changes, but others can cause changes in function without noticeable structural damage,” says Dr. Small. “The hippocampus is not a simple structure—it’s actually a circuit. This means the best way to evaluate hippocampal function is to evaluate each node in the circuit,” in other words, each subregion of the hippocampus. According to Dr. Small, different processes that accompany aging impair memory by targeting distinct regions of the hippocampal area, so the precise location of the problem spot may give clues about what caused the memory loss. “Vascular changes primarily target the CA1 region, and testosterone changes affect the dentate gyrus. If we can validate this technique, we’ll be able to tell that, if the physiological defect is in entorhinal cortex, we can diagnose the patient with Alzheimer’s disease. If, on the other hand, we find it in the dentate gyrus, we may suspect that hormonal changes are behind the deterioration. If the damage is principally in CA1, we might be inclined to do a cardiovascular workup.”

“The technique we used to gather human data doesn’t require new equipment. It has the potential for widespread use,” says Dr. Small of their ROXY measurements with MRI. Although measuring rapid changes in blood oxygenation requires a set-up generally only available in an academic research facility, commonly available clinical MRI facilities could be used to measure ROXY. “This would require collection of data normally, it would just be processed in a new way.” Potentially, data collected at conventional MRI centers without any retrofitting could then be processed at a central location to provide ROXY profiles for each patient, making it easier for patients virtually anywhere in the U.S. to gain access to the technique.

The work was supported by federal grants from the National Institutes of Aging and Mental Health, the Beeson Faculty Scholar Award from the American Federation of Aging, the Banbury Fund, the Taub Foundation, the G. Harold and Leila Y. Mathers Charitable Foundation, and the Howard Hughes Medical Institute.

# # #