The most widely accepted method for detecting Alzheimer’s disease before symptoms arise is to take a PET scan of the brain, to search for the amyloid protein deposits that are a hallmark of the disease. Scientists from Washington University School of Medicine in St. Louis set out to develop a more efficient test—a blood test that could be administered to thousands of people per month at a much lower cost than the typical $4,000 price of a PET scan.
The first test they developed, which was designed to detect tiny amounts of amyloid in the blood, was accurate only 88% of the time when compared to PET images. So they decided to try combining the blood test with two other major risk factors for Alzheimer’s: age and the presence of the APOE4 gene variant.
It worked. The accuracy of the test rose to 94%, they reported in the Journal Neurology. Although gender is also a risk factor—two out of three Alzheimer’s patients are women—including it in the analysis did not improve the accuracy of the test, the researchers said.
The study involved 158 people over the age of 50, all but 10 of whom did not show any signs of cognitive decline when they enrolled. The researchers used mass spectrometry to measure the ratio of amyloid beta 42 and amyloid beta 40 in the blood, because that ratio is known to drop when deposits of the protein are increasing in the brain.
As part of their analysis, the researchers went back to study the false positives—the blood tests that initially showed the presence of amyloid even when brain scans showed no signs of Alzheimer’s. In some of those patients, when PET scans were repeated an average of four years later, early signs of amyloid deposits were detected. This could be early proof that the blood test may be useful for detecting Alzheimer’s long before symptoms appear, the researchers suggested in a statement.
RELATED: Eisai's prototype blood test for Alzheimer's shows links to CSF amyloid in early disease
The Washington University team is not the first to suggest that measuring the ratio of amyloid beta 42 and 40 in blood could be useful for detecting Alzheimer’s—in fact, Eisai is working on a prototype blood test that does just that. In July, the company reported that in a trial they found a strong correlation between the ratios of the two types of protein in both blood and cerebral spinal fluid taken from normal people and those with mild cognitive impairment. Eisai and its partner, Sysmex, now plan to compare the results with PET images.
The search for effective Alzheimer’s treatments continues, despite some notable failures in the field. Most recently, Amgen and Novartis halted trials of a BACE inhibitor after some patients showed a worsening of cognitive impairment.
Still, there are plenty of clinical trials underway for other treatments aimed at treating or preventing the disease, which is where an effective blood test for early detection could come in handy, the University of Washington researchers said.
"If you want to screen an asymptomatic population for a prevention trial, you would have to screen, say, 10,000 people just to get 1,500 or 2,000 that would qualify," said senior author and Washington University neurology professor Randall Bateman, M.D., in the statement. "Reducing the number of PET scans could enable us to conduct twice as many clinical trials for the same amount of time and money.”
In fact, Bateman’s team included a financial analysis in the study that was based on records from an Alzheimer’s prevention trial. That trial uses PET scans to confirm changes in the brain that would indicate the onset of Alzheimer’s. If those patients had been prescreened with a blood test, it would have reduced the number of PET scans needed by about 60%, the Washington University researchers concluded.