Progress Cited in Alzheimer’s Diagnosis

http://www.nytimes.com/2007/10/15/science/15blood.html

By ANDREW POLLACK

Published: October 15, 2007

Scientists reported progress yesterday toward one of medicine’s long-sought goals: the development of a blood test that can accurately diagnose Alzheimer’s disease, and even do so years before truly debilitating memory loss.

A team of scientists, based mainly at Stanford University, developed a test that was about 90 percent accurate in distinguishing the blood of people with Alzheimer’s from the blood of those without the disease. The test was about 80 percent accurate in predicting which patients with mild memory loss would go on to develop Alzheimer’s disease two to six years later.

Outside experts called the results, published online yesterday by Nature Medicine, promising but preliminary. They cautioned that the work needed to be validated by others and in much larger studies, because there have been many disappointments in the past.

“Looking for biomarkers for Alzheimer’s disease is a very hot area,” said Dr. William Jagust, professor of public health and neuroscience at the University of California, Berkeley. “Things tend to get a lot of attention, and they are not always borne out.”

Right now, Alzheimer’s disease is diagnosed by a battery of mental and other tests, and even that diagnosis rests on the judgment of the physician. Doctors say it would be useful to have something like a pregnancy test for Alzheimer’s — one that is simple and definitive and can pick up the disease early, maybe even before symptoms appear.

At present, treatments for Alzheimer’s disease are not very effective, so some people might not want early notification that they have an incurable ailment. But others might want it.

“There are people who want to know what their future holds, so they can plan their estates and lives,” said Dr. Sam Gandy, a professor at Mount Sinai School of Medicine in New York who is chairman of the medical and scientific advisory council of the Alzheimer’s Association.

But the real usefulness of an early diagnostic test would come when drugs are developed that slow or halt the progression of Alzheimer’s disease. Several therapies that might be able to do that are now being tested. The drugs would be most valuable if they could be used before cognitive ability had declined too much.

Numerous efforts have been made to find an early marker in blood, urine, spinal fluid and eye movements, as well as through brain imaging using PET scans and M.R.I. A Norwegian company, DiaGenic, has presented some early results of a blood test that analyzes gene activity. Researchers at Cornell published early results last December using a pattern of 23 proteins in the spinal fluid. But no test has gained universal acceptance. The closest is a test of two proteins in the spinal fluid that is used in some European countries, said Dr. Dennis Selkoe, a professor of neurology in Boston at Brigham and Women’s Hospital and Harvard.

A blood test would be easier for patients than a spinal tap, Dr. Selkoe said. But a blood test is possible only if Alzheimer’s disease leads to changes outside the brain that can be detected in the blood.

Dr. Tony Wyss-Coray, an associate professor of neurology at Stanford and the senior author of the new paper, said there was evidence from animal studies that brains affected by Alzheimer’s sent out signals to the body’s immune system. So his team decided that rather than looking at all proteins in the blood, it would focus on those involved in communication between cells, hoping to eavesdrop, as it were, on dialogue related to Alzheimer’s.

The researchers gathered more than 200 blood samples from people with Alzheimer’s and those without. Using 83 of the samples, they measured the abundance of 120 proteins involved in cell signaling and found they could distinguish the Alzheimer’s samples from the controls using 18 of the proteins.

They then tested their 18-protein signature on an additional 92 samples. The tests agreed with the clinical diagnosis about 90 percent of the time.

Perhaps most intriguing were the results of the test on 47 blood samples taken from people with mild cognitive impairment, a minor loss of memory that can be a precursor of Alzheimer’s. The test was able to predict with about 80 percent accuracy whether a patient went on to develop Alzheimer’s two to six years after the blood sample had been collected.

Dr. Wyss-Coray, who is also at the Veterans Affairs Palo Alto Health System, said that monitoring communications between cells might be a way to develop diagnostic tests for other diseases. And understanding why the levels of the 18 proteins are different in Alzheimer’s patients might provide a better understanding of the disease.

The study was paid for by the National Institute on Aging, the John Douglas French Alzheimer’s Foundation, the Alzheimer’s Association and Satoris, a company co-founded by Dr. Wyss-Coray to commercialize the test.

The company said in a news release that it hoped to have a test available for research purposes next year. But even if the preliminary results are validated, it is likely to be a few years before a test is approved and ready for use by doctors.