Toward A Blood Test For Alzheimer's

Doctors might one day be able to predict whether a person will develop Alzheimer’s disease with a few milliliters of extracted blood, according to a small study (Nat. Med. 2014, DOI: 10.1038/nm.3466). A team of U.S. clinicians and researchers, led by neurologist Howard J. Federoff of Georgetown University Medical Center, discovered a group of 10 lipids in blood that dip below normal levels years before a person exhibits symptoms such as memory loss.

Alzheimer’s, which now afflicts about 5 million people in the U.S., has so far been impervious to drug discovery efforts. Many scientists believe that late-stage clinical trials of experimental compounds have failed because study participants had too much brain cell damage for effective treatment.

As a result, “we want to identify individuals at the earliest stages of disease and intervene then,” when therapeutics might protect brain cells, says Mark E. Mapstone, a neuropsychologist at the University of Rochester School of Medicine who helped discover the new set of lipids.

To do so, scientists need techniques for diagnosing people with Alzheimer’s before symptoms appear. Some methods are already under development, such as positron emission tomography and cerebrospinal fluid analysis. But a blood test, Mapstone points out, would be a lot simpler and less expensive.

To find biomarkers for a possible blood test, Federoff, Mapstone, and coworkers began their study with about 500 human subjects. Over a five-year period, they periodically drew blood from these people. During the study, 28 subjects went from having healthy brains to exhibiting impaired memory. When the researchers screened blood samples of those patients with mass spectrometry and compared the results to individuals from the study group who were cognitively healthy, they uncovered the 10 predictive lipids. After establishing these biomarkers, the team’s mass spec experts analyzed fresh, unmarked blood samples from both subgroups. The researchers identified the patients who would eventually get Alzheimer’s with 90% accuracy.

“These findings are really excellent in their predictive power,” says Simon Lovestone, a psychiatrist at the University of Oxford. “But it’s important to note that the study is relatively preliminary.” The results will need to be replicated by other groups with much larger patient populations, he adds. “It’s the start, not the end.”