Advances in Neuroimaging Help Refine Search for Biomarkers for Alzheimer’s Disease


by Carl Sherman

April 9, 2010

Alzheimer’s disease inexorably devastates the brain. Tissues shrink as neurons die, while fluid-filled ventricles widen, in such critical regions as the hippocampus, vital for memory. But early on, the damage is subtle and slow to advance, which presents researchers with a formidable challenge—and opportunity.

If they could track the Alzheimer’s process as it unfolds, they might use this knowledge to create a tool to identify effective treatments. And by spotting the very earliest signs of disease—even before symptoms appear—they could maximize the benefit of such treatments.

 The quest for such biomarkers— patterns visible on brain scans and through cerebrospinal fluid analysis —has been a major focus of research over the past decade. As the search continues, refinements in magnetic resonance imaging (MRI), in particular, are revealing more and more of Alzheimer’s inside story.

Boosting the power of MRI

With high-resolution MRI, researchers can now view the brain in millimeter detail, making it possible to delineate subregions within cortical and subcortical structures.

“We have a fairly good idea of patterns of atrophy leading from mild cognitive impairment (MCI) to Alzheimer’s disease (AD), what regions are hit most,” says Dominic Holland of University of California, San Diego. “Synapse loss begins in the entorhinal cortex and spreads from there into the hippocampal region, then to the parietal and temporal lobes.”

Holland and Anders Dale, who led the study, analyzed MRI images in groups of healthy people and in those with Alzheimer’s or mild cognitive impairment who had been scanned every six months for two years. Using methods they had developed to rectify distortions that inevitably occur in MRI images, they tracked changes in key subregions of each individual brain, as well as in overall volume and ventricle size. When they looked for measures that had changed most, decline in brain volume and increased ventricle size stood out.

“The problem is, when you compare these changes in MCI or AD patients with changes in normal individuals, their significance drops way down,” Holland says. Much brain shrinkage and ventricle growth, in other words, could be attributed to the effects of aging. For identifying “disease-specific changes,” these global measurements were no better than neuropsychological testing.

But when the researchers reviewed patients’ serial brain scans after correcting for the normal effects of aging, changes in subregions of the cortex were much more revealing. Most sensitive of all to the disease process, it appeared, was the thin layer on the underside of the temporal lobe called the entorhinal cortex.

In the study, which was reported in December 2009 in the Proceedings of the National Academy of Science, the researchers interpreted their findings in terms of a most practical question: How large a clinical trial would be needed to show that a drug worked to slow brain deterioration in people with Alzheimer’s or mild cognitive impairment? 

These “power calculations” suggested that measuring the entorhinal cortex would be the most efficient way to identify a drug effective against disease-specific brain changes. And the researchers concluded that the number of subjects it would take to detect a 25 percent reduction among patients with mild cognitive impairment would be  241, less than half as many as would be needed using global brain volume or a clinical dementia rating scale; for Alzheimer’s patients, just 65.

 Although potential utility  in clinical trials is what drives much Alzheimer’s biomarker research,   “this is one of the first published studies to look directly at this kind of analysis, using power calculations to find the most sensitive ways of detecting change,” says Neil Buckholtz, chief of the dementias branch at the National Institute on Aging.    

 MRI findings may spotlight a valuable drug more quickly than would watching the progression of a person’s symptoms, but it would not be, in itself, enough to gain FDA approval, Buckholtz cautions. “The gold standard remains whether a drug has an effect on cognition, which is the hallmark of AD.”  

On the other hand, “a clinical endpoint can’t distinguish symptomatic improvement from true disease modifying effects,” as biomarkers that tap into pathophysiology aim to do. (For example, a drug like donepezil (Aricept), may temporarily improve cognition by increasing levels of a neurotransmitter, but doesn’t slow the loss of neurons.) The two kinds of data are complementary, he says.   

Very early warning    

   Another goal of Alzheimer’s research, Buckholtz says, is identifying at-risk individuals earlier and earlier in the disease progress. “There are a lot of interesting studies looking at what predicts which people still within the normal range will go on to progress to MCI. The hope is to develop drugs that will stop progression at the earliest stage of the disease.”

By using a relatively new variant of MRI, researchers at Tor Vergata University in Rome may have found such a means of prediction. They scanned the brains of healthy people aged 20 to 80 years, using standard MRI and also diffusion tensor imaging (DTI), which registers the movement of water in tissues. The participants also took tests of cognitive function.

As they reported in January in Neurology,  the researchers found that as total volume of the brain and of the hippocampus declined with age,  mean diffusivity—movement of water in all directions, as seen by DTI— increased in hippocampal tissue.

Like earlier investigators using structural MRI, the researchers didn’t see any link between loss of hippocampal or overall brain volume and cognitive decline. But DTI findings were more rewarding:  among subjects aged 50 or over, higher mean diffusity in the hippocampus was significantly associated with poorer performance on tests of verbal and visual memory.

“Conventional MRI only shows macroscopic alterations in brain anatomy,” says lead investigator Giovanni Carlesimo. “DTI is able to demonstrate microscopic changes in nerve fiber structure. Increased diffusivity is generally interpreted to mean an increase in extracellular space, as a consequence of neuron, axon, and dendrite loss.”

He noted that degenerative changes in the hippocampus occur early in Alzheimer’s, and memory deficits are among the first symptoms. “Our hypothesis is that in otherwise healthy elderly people, the association of memory deficit and [microscopic] hippocampal changes represents a very early indicator of the disease.“

The data, Carlesimo granted, are “preliminary… our hypothesis needs to be confirmed by the longitudinal evaluation of these individuals to confirm their higher susceptibility to developing AD.” A three-year follow-up study is underway.

Norbert Schuff, of the Center for Imaging of Neurodegenerative Diseases at the Veterans Affairs Medical Center and   University of California, San Francisco, agreed that “a future study will have to determine whether the changes the researchers saw are actually related to AD pathology; they may be a simple reflection of  aging.” That reservation notwithstanding, the study represents a good step forward, he says. “It has put DTI on the map as a potential imaging marker for early AD.”

“The paper shows that we now have another variant of MRI that seems more sensitive to capture the biological underpinning of memory deficits in a group without AD or MCI. … It could be [useful] to recruit for trials of a drug that might prevent the progression of these minor memory deficits,   and perhaps to stop progression to AD.”

Considering these two studies, Schuff says, “it’s clear that both techniques have validation; in future research, we will use high resolution structural MRI and DTI together. The combination of structural changes and DTI changes may be more powerful than either alone” for identifying biomarkers for early detection of Alzheimer’s.

slightly revised April 6, 2010