March 9, 2014 at 2:01 PM ET
A blood test that can predict who is about to develop Alzheimer’s or related early memory loss up to two years before the first symptoms may be in early stages of development, researchers believe.
Many scientists have tried, and failed, to come up with a test that can predict Alzheimer’s. But this one, developed by teams at Georgetown University and the University of Rochester, seems especially accurate.
“This is the first time a highly sensitive and specific test has been able to predict who will become demented,” says Dr. Howard Federoff of Georgetown, who helped lead the research published in the journal Nature Medicine.
"Sensitive and specific" means the test misses very few people who did go on to develop Alzheimer’s or the memory loss that precedes it. And there were no “false negatives.”
“We did not find any individual that didn’t test positive that became demented,” Federoff said.
The need is urgent. Alzheimer’s affects more than 5 million Americans and it’s projected to soar to 13 million over the next 35 years. A study published earlier this month suggested it’s a big killer, taking down more than 500,000 Americans every year.
“They tried the therapies too late. We need to find people earlier.”
There’s no cure and not even a good treatment. Many different approaches have been unsuccessful. “It’s frustrating,” Federoff says. But Federoff and other experts think if people could be treated before they develop symptoms, before the brain damage is so bad, it might be possible to slow the onset of Alzheimer’s.
“They tried the therapies too late. We need to find people earlier,” Federoff said.
The first step would be a test that predicts who needs early treatment.
Federoff and colleagues looked at the blood of 525 people aged 70 or over who had no symptoms of Alzheimer’s at the start. They were all living on their own and generally healthy. They gave blood at the beginning of the five-year study and once a year after that.
Over the five years, 74 of the volunteers developed either mild cognitive impairment or Alzheimer’s, which can be diagnosed with psychological tests, brain scans or a spinal fluid test.
Federoff’s team analyzed their blood samples to see if there was any pattern at all of compounds that might change predictably in those who developed memory problems.
There was. It was a pattern of 10 lipids — fat-like compounds — that changed in the people who went on to develop memory loss or Alzheimer’s.
“We’re finding these lipids, 10 of them, that are highly predictive of who in the next two years will become cognitively impaired,” Federoff said in an interview.
The team is not exactly sure what the lipids mean. They may come from the debris caused when brain cells die.
Brain cells don’t die quietly, says Heather Snyder of the Chicago-based Alzheimer’s Association. And there is a lot going on in Alzheimer’s disease: buildups of protein called amyloid beta, tangles of nerves and also inflamed cells releasing signaling chemicals called cytokines, chemokines and destructive chemicals called reactive oxygen species.
It’s also known that Alzheimer’s is somehow associated with blood fats. Studies link Alzheimer’s risk with unhealthy cholesterol levels and a gene mutation that affects cholesterol, one called APOE4, also affects Alzheimer’s risk.
However, this particular blood test seemed to have nothing to do with APOE4, Federoff says.
As with any test, it will have to be tried out in a much larger group of people before anyone could think about developing it commercially.
"There is so much that we don’t understand about what is happening."
“This study addresses a critical need in the field, that of finding a relatively inexpensive and non-invasive biomarker to allow us to screen large groups of people,” said Dr. Ronald Petersen, director of the Mayo Clinic’s Alzheimer's Disease Research Center. “This study is an important step in that direction but needs to be validated on a larger group of subjects more representative of a more general aging population."
Federoff’s team is also looking to see if they can predict memory loss even sooner than the two to three years this test did it.
The test itself wouldn’t be hard to do, he says. “It’s an easy approach,” he said. “The technology exists in almost any clinical lab.”
Many people have tried to develop tests for Alzheimer’s, including blood tests.
“Many of those studies and those panels that are being identified are still very much in the early phases,” says the Alzheimer's Association Snyder. “They need to be validated and confirmed in much larger populations,” Snyder added in a telephone interview.
Part of the problem is that scientists don’t even agree on what causes Alzheimer’s. "There is so much that we don’t understand about what is happening, and one is what are those earliest changes and why,” she said.
It’s not even that easy to diagnose Alzheimer’s in people with symptoms. “Current biological markers for early disease, including levels of abnormal proteins in cerebrospinal fluid, structural and functional MRI of the brain, and brain PET amyloid imaging are limited by their inability to specifically diagnose (‘rule in’) the disease, or because they are invasive and can be expensive,” notes Maria Carrillo, Alzheimer’s Association vice president of medical and scientific relations.
Currently, people who worry they might be developing Alzheimer’s are often just left to fret, without knowing if they really ought to be getting their affairs in order.
Judy Silverman and Ami Schmitz of the NBC Medical Unit contributed to this report.