Until very recently, the only way to provide a firm diagnosis of Alzheimer's was through a brain autopsy.

Things are starting to change. Brain imaging and spinal taps have now started to look for the plaques and tangles that are the hallmarks of the neurodegenerative disease in living patients. These techniques are now being tested in clinical trials to detect the presence of Alzheimer's pathology years before symptoms appear, a prelude to administering drugs under development to stop progression of the disease. Many researchers have conjectured that by the time cognitive functioning starts to waver, damage to nerve cells may have proceeded too far for drugs to be of much benefit.

Ultimately, the vision of Alzheimer's researchers is to create a predictive test that could be administered, perhaps at the age of 50, about the time when someone is first counseled to start colonoscopies, bone scans and other tests to ward off diseases of aging. If a test is positive, a still-to-be-developed drug regimen could then be prescribed.

Any predictive test would need to be simple and relatively inexpensive—MRI, PET or spinal taps don't really fit that description. (MRI and PET have received approval in the U.S. for limited clinical use and spinal taps to sample cerebral spinal fluid are used in several European countries.)

A small step toward the kind of diagnostic that could be given in a doctor's office can be found in the March 9 online publication in Nature Medicine. The study highlights a test that looks at concentrations of certain fats circulating in blood to predict whether someone will become cognitively impaired or get Alzheimer's within three years. The test was devised by examining blood from 525 people over the age of 70 during a period of five years.

After the first three years had passed, the researchers examined the blood of 53 members of the group who had developed memory impairment or early-stage Alzheimer's and compared it with that of a healthy control group. They found levels of 10 fats whose presence seemed to indicate the breakdown of brain cell walls, possibly an indicator of neurodegeneration. They then tested these lipid concentrations in another 41 individuals from among the hundreds in the larger group, finding that 10 of them went on to develop dementia. “What we discovered is that these 10 circulating lipids in normal individuals could predict which of them would go on to have Alzheimer’s,” says Howard J. Federoff, a professor of neurology at Georgetown University Medical Center, who led the study. The changes in lipid levels determined with 90 percent accuracy (nine of 10 individuals) who progressed to dementia.

There have been previous attempts to develop Alzheimer’s blood tests to diagnose the disease. The main challenge has always been to replicate the assays in subsequent studies—and that is also a concern with the findings about this latest test. “The panel of markers identified are interesting and the greater than 90 percent accuracy is impressive,” says Ralph Martins, director of the WA Centre of Excellence for Alzheimer's Research and Care at Edith Cowan University, which has been involved in developing blood-based Alzheimer’s diagnostics. Martins adds in an e-mail: “Other groups such as mine studying the thoroughly characterized Australian Imaging Biomarkers and Lifestyle study of Ageing cohort will need to independently confirm these findings.”

William Hu of Emory University, also involved in developing Alzheimer’s blood tests, notes similar reservations: “Over the past six years (since the publication of a blood protein panel capable of diagnosing Alzheimer's disease also published in Nature Medicine), we have seen many initially exciting discoveries fail replication because of unsuspecting technical factors driving the main findings," Hu wrote. “These issues can include technical processing, over-the- counter supplements taken by those with MCI [mild cognitive impairment], or prescription drugs started on MCI/AD patients."

For his part, Federoff emphasizes that this work has a long way to go. Specifically, larger studies are needed among a more diverse study population and it must be determined as well whether this test can predict disease onset more than a decade before diagnosis. “This is really quite distal. I think it’s vastly premature to talk about commercializing this,” he says. Diagnostics of course, are not the only challenge. Also desperately needed are the still nascent pharmaceuticals that can be administered to patients who, once tested, know that they are likely to develop Alzheimer’s. So far, effective drugs have proven elusive as millions of Alzheimer’s patients and their caregivers watch intently as one drug after another has failed to stop the inexorable progression of the devastating disease.

Source: National Institutes of Health