This past summer, the National Institute on Aging and the Alzheimer’s Association proposed new diagnostic criteria that would allow clinicians to use biomarkers—objective, measurable signs of the disease process—to aid in indentifying Alzheimer’s disease. Another workgroup of the same organizations explored the use of biomarkers for investigating “preclinical” Alzheimer’s—a stage before symptoms appear.
“We’re at an interesting transition point, from the basic level of biomarker research to a more meaningful applied level,” says Leslie Shaw, professor of pathology and laboratory medicine at the University of Pennsylvania Comprehensive Neuroscience Center.
While much of this research has looked into the brain itself, using structural and functional imaging, some of the most exciting work has focused on body fluids, with special attention to chemicals in cerebrospinal fluid (CSF): amyloid-beta and tau.
In the brain, these proteins make up the “plaques” and “tangles” that characterize Alzheimer’s. “There’s an emerging consensus that reduced amyloid-beta in the CSF correlates with increasing plaque burden in the brain,” says Shaw. “And that the degeneration of neurons is associated with elevated tau.”
Shaw was among the authors of a CSF study that garnered Page One headlines in national news media (including the New York Times) and in the scientific community. An August 28 Lancet editorial on the disappointing performance of experimental Alzheimer’s drugs talked of the need to identify at-risk individuals “before memory loss and issue destruction occurs” and singled out the CSF findings as a promising possibility.
A pattern emerges
The study, which was reported in Archives of Neurology in August, used CSF that had been collected from 102 people with Alzheimer’s, 200 with mild cognitive impairment (MCI), and 114 healthy controls, as part of the Alzheimer’s Disease Neuroimaging Initiative.
The data from the whole group was analyzed using a statistical approach in which the person running the data did not know which people had which diagnosis, says John Q. Trojanowski, co-director of the Center for Neurodegenerative Disease Research at University of Pennsylvania, member of the Dana Alliance for Brain Initiatives, and senior author of the paper.
“The distribution [of amyloid-beta and tau] was bimodal,” (two different results appeared with equal frequency) making it possible to derive thresholds for high tau and low amyloid “that crisply separated subjects into two meaningful groups”—one with a “healthy” and one with an “AD” CSF pattern.
When they analyzed a different set of CSF samples, taken from 68 people who later died of Alzheimer’s, the researchers found the predicted “AD” pattern in 64 (94 percent). In a third group, the pattern was a highly sensitive predictor of mild cognitive impairment that would worsen to become Alzheimer’s: it appeared in 100 percent of CSF samples from 57 people with mild cognitive impairment who within the next 5 years had been diagnosed with Alzheimer’s.
The study did not report the “false positive” rate— the number of people who had the AD pattern but failed to develop Alzheimer’s. “We didn’t pursue specificity,” says Shaw. “We’re not concluding that it’s a perfect test, but a meaningful signature, worth studying to the maximum degree.”
More information will emerge with time, says Trojanowski. “We’re following our original subjects, to see how they change. In 5 or 6 years, virtually all those with MCI who are going to convert, will convert. We’ll have a lot more statistics, with associated biomarker information, to sharpen our acumen.”
An even longer view may prove more enlightening. In the original control group, the researchers found the AD profile in the CSF of one-third of apparently normal, healthy people, which “suggests that AD pathology is active and detectable earlier than has heretofore been envisioned,” the authors wrote. To follow up on these findings would require “subsequent studies that include cognitively normal subjects who can be followed up for possibly 10 years or more.”
Reisa Sperling, director of the Center for Alzheimer's Research and Treatment of the Massachusetts Alzheimer's Disease Research Center and chair of the National Institute on Aging- Alzheimer’s Association preclinical Alzheimer’s workgroup, says that the study’s amyloid-beta findings are “very similar to what has been reported in PET studies” as an indication of amyloid accumulation in the brain. “Both are very close to being able to substitute for autopsy confirmation of amyloid,” she says.
The tau findings in the CSF study provide something more, however—“some evidence of neurodegeneration, suggesting that these downstream processes are beginning even before people have symptoms.”
As more CSF data become available, “it may be useful to break it down: What is the risk of AD if you have amyloid alone? How much do signs of neurodegeneration increase risk?” says Sperling, also a member of the Dana Alliance. “We may need to start talking about intervening at midlife, and not just in amyloid positive individuals,”
A blood test for amyloid
While its proximity to the brain makes CSF the most likely body fluid to yield Alzheimer’s biomarkers, the difficult and painful extraction procedure, a spinal tap, is a drawback. Other researchers are exploring whether a simple blood test might provide similar information.
At Columbia University, researchers measured plasma amyloid-beta in 880 people aged 65 and older, and repeated the test 4 ½ years later. By the second measurement, 481 remained cognitively healthy, 70 had developed Alzheimer’s, and 329 were cognitively impaired but not demented.
In a paper published online by Archives of Neurology in August, the researchers reported a significant association between high plasma amyloid-beta that remained stable or decreased over time, and cognitive deterioration. The findings “increase the potential utility of plasma amyloid-beta as a biomarker for AD progression,” says lead author Stephanie Cosentino, of Columbia’s Taub Institute for Research on Alzheimer’s Disease and the Aging Brain.
This association between plasma amyloid-beta and cognitive deterioration also applied to people who remained healthy as well as those who became impaired. “There are two possible interpretations: These individuals could be en route to dementia, and we didn’t follow them long enough, or amyloid-beta might be reflective of accelerated brain aging,” rather than Alzheimer’s, she said.
Cosentino noted that in this still-healthy group, cognitive decline was predominantly in the area of memory, which is where the symptoms of Alzheimer’s disease first appear.
“Plasma amyloid-beta is very exciting as a potential biomarker,” says Sperling. The problem is that wide variation between individuals makes it impossible to interpret single readings. “We need to understand factors that may be causing these differences.”
“This doesn’t mean that plasma amyloid-beta won’t ultimately prove useful,” she says. The Columbia study adds to “mounting data that intra-individual change is powerful.” Ultimately, she said, using biomarkers like CSF instead of relying on clinical diagnosis to define disease status and progression could make the results of more accessible tests like plasma amyloid clearer.
Pooling the power of diverse biomarkers is a growing trend in research. “We’re looking at data in a number of different ways, combining imaging data with CSF data,” says Neil Buckholtz, chief of the dementias of aging branch at the National Institute on Aging. “They’re complementary.”
At the same time, the focus is shifting to an earlier point in the disease process, he says. Both the CSF study and the Columbia plasma amyloid study found intriguing signs of change in people who were still in the cognitively normal range, and brain imaging studies also have found similar signs.
“Work is going on to figure out what’s happening in the pre-MCI group,” Buckholtz says. “The hope is that we will eventually be able to give drugs at the earliest stage of disease, to slow it or stop it.”