Non-Invasive MRI Technique Distinguishes between Alzheimer's and Frontotemporal Dementia

A non-invasive magnetic resonance imaging (MRI) technique called arterial spin labeling is just as accurate as invasive scanning techniques in distinguishing Alzheimer’s disease from frontotemporal dementia (FTD) in the brains of elderly people, according to a new study at the San Francisco VA Medical Center (SFVAMC).

The study, led by Norbert Schuff, PhD, a Principal Investigator at SFVAMC, used arterial spin labeling to measure perfusion, or blood flow, in the areas of the brain affected by the two diseases. “Blood flow indicates brain activation,”
said Dr. Schuff. “So the area with less blood flow is the area affected by disease.” In the study, arterial spin labeling successfully distinguished between Alzheimer’s patients, FTD patients, and people without dementia.

Antao Du, PhD, SFVAMC Research Scientist and study co-author, is presenting the results at the first International Conference on Prevention of Dementia, which is being held June 18-21 in Washington, D.C. The conference is sponsored by the Alzheimer’s Association.

Frontotemporal dementia is a degenerative condition involving the front part of the brain. It is the second-most common dementia after Alzheimer’s disease, which mainly affects other brain areas such as the hippocampus and the temporal lobe. In their early stages, the two diseases present similar symptoms, making accurate diagnosis difficult, said Schuff, who is also an associate professor of radiology at the University of California, San Francisco (UCSF). In its later stages, FTD affects social conduct, social inhibitions, and personality, while Alzheimer’s is a progressive impairment of multiple cognitive functions, often involving memory decline. “Progression of frontotemporal dementia is usually faster than Alzheimer’s, and the underlying pathology is different, so it is important to know the difference,” Schuff observed.

Currently, brain blood flow can be measured using positron emission tomography
(PET) and single proton emission computerized tomography (SPECT). However, these techniques involve injecting patients with radioactive tracers. In addition, Schuff noted, they can be expensive—about $2,000 for a PET scan—can take up to half a day to perform, and are not widely available. “So if you can acquire blood flow information with MRI, that would be very beneficial. MRI is totally non-invasive, making it much safer for patients. It’s more widely available, it’s cheaper, and arterial spin labeling can be done in ten minutes together with a conventional MRI scan.”

In simple terms, MRI is a non-radioactive imaging technique that measures the magnetic alignment of protons in the body. In arterial spin labeling, a technique invented by researchers at the University of Pennsylvania, protons in arterial blood are magnetically aligned in the opposite direction from the rest of the protons in blood and brain tissue. By measuring the intensity of the magnetic signal from these so-called inversely polarized protons when they reach the brain, researchers can calculate the amount of blood flow, and thus neuron activity, in a particular section of the brain.

In the study, Schuff and his fellow researchers measured brain perfusion in 24 Alzheimer’s patients, 21 FTD patients, and 25 control subjects without dementia. The subjects were 62 to 90 years old, with an average age of just under 63. They were studied using an MRI system with a magnetic field strength of 1.5 Tesla, a common system in clinics and hospitals in the United States.
The researchers successfully used arterial spin labeling to replicate PET and SPECT data on brain perfusion in all subjects. They also found that the perfusion data, added to structural information about the brain obtained with conventional MRI, significantly improved the classification of FTD from normal aging. Thus, “we gained specificity and sensitivity,” said Schuff.

Schuff emphasized that because this was a research study, the aim of which was to accurately replicate PET and SPECT perfusion data, it did not prove that arterial spin labeling can be used to diagnose an individual patient. The next step for future research, he said, is to demonstrate that the perfusion abnormalities correlate with specific clinical symptoms. “At the moment, we have just dichotomized [patients] into Alzheimer’s and FTD,” he said. “But of course cognitive impairment is usually more complex—you have a range of impairments.”

The technique has the potential to distinguish other types of dementia as well.
Schuff plans to study a larger sample in a clinical setting, with the goal of determining whether this is possible. “Once we have a large database of images,” he predicted, “we can better determine what is normal, and then compare an individual subject with this normal range.”

Schuff intends to continue his studies using a more powerful MRI system that operates at a magnetic field strength of 4 Tesla. This state-of-the art system was recently installed at SFVAMC with support funds from the National Institutes of Health (NIH) and the Department of Defense. “At 4 Tesla, measurements of brain blood flow will be more accurate than at 1.5 Tesla due to a higher signal intensity and prolonged lifetime of the polarized protons,”
Schuff said.

Additional authors of the study are G.-H. Jahng, PhD, and Bruce Miller, MD, of UCSF; Sataru Hayasaka, PhD, of SFVAMC; Joel Kramer, MD, of UCSF; and Michael Weiner, MD, Director of the Magnetic Resonance Spectroscopy Unit at SFVAMC and professor of radiology, medicine, psychiatry, and neurology at UCSF.

The research was funded by grants from the NIH that were administered by the Northern California Institute for Research and Education (NCIRE), and by a grant from the U.S. Department of Veterans Affairs.