10 August 2012. Researchers think amyloid deposition, synaptic disruption, and neuronal glucose metabolism are all somehow connected as Alzheimer's disease (AD) unfolds, but they have not yet been able to explain just how, when, and where. Members of Alexander Drzezga’s lab from the Technical University in Munich, Germany, use longitudinal multimodal imaging to tackle aspects of that question. They presented their latest findings at the Alzheimer’s Association International Conference (AAIC), held 14-19 July 2012 in Vancouver, Canada.
Stefan Förster has set out to determine how plaque deposition and metabolism (a measure of synapse dysfunction) relate to each other over time. He published earlier this year that areas of plaque deposition at a baseline scan in people with mild AD became hypometabolic by about 27 months later (see Förster et al., 2012), suggesting that plaques cause a drop in metabolism. On an AAIC poster with his latest longitudinal data, he reported that metabolism predicts plaque deposition as well.
Förster followed 15 mild AD patients for about two years using [18F]-2-deoxy-2-fluoro-D-glucose (FDG) positron emission tomography (PET) scans to show metabolic function, and Pittsburgh compound B (PIB) to detect plaques. He found that areas of relatively preserved metabolism at baseline were overrun with amyloid two years later. In addition, as in his previous work, regions with baseline amyloid were less metabolically active at follow-up. The results point to a causal triangle, where initial neuron activity is followed by amyloid deposition, which in turn leads to functional neuronal decline. “This bidirectional finding suggests that there is a causal relationship of these two biomarkers in neurodegenerative AD,” Förster told Alzforum. “I think this is the first report of such a bidirectional relationship in early AD patients.”
But what could explain hypometabolism in regions that do not have much amyloid in the first place? Researchers have previously found that network activity in AD patients goes awry, especially in the default-mode network, which is active during sleep or when the mind wanders (see ARF related news story on Sperling et al., 2009; Hedden et al., 2009; Sheline et al., 2010). Perhaps a network breakdown is at play here, too, suggested Elisabeth Klupp, who spoke about her findings at an oral presentation. “We hypothesized that hypometabolism in non-amyloid affected brain regions could result from functional disconnection from remote brain areas affected by these pathologies,” she told her audience.
To find areas of hypometabolism with little plaque, she and her colleagues used multimodal imaging in 19 AD patients. An averaged FDG-PET scan from them compared with one from 15 age-matched healthy controls revealed areas that were abnormally hypometabolic in the AD patients. By superimposing that map on an averaged PIB-PET scan from the AD patients, the researchers saw areas with reduced metabolism and no significant plaque, most prominently the left superior frontal gyrus. Hypometabolism there exceeded what the team would expect, given the amount of amyloid. This region then functioned as a “seed” to figure out which areas were normally functionally connected to it in healthy controls.
In 17 elderly healthy controls at rest, functional magnetic resonance imaging (fMRI) scans revealed that parietal areas, especially on the left side of the brain, communicate with the left superior frontal gyrus. In the AD patients, according to PIB-PET, these same areas were overrun with amyloid. The overlap suggests that a breakdown in communication between areas with too much and very little amyloid may cause the unexplained hypometabolism, suggested Klupp.
“You might have amyloid in one place inducing neuronal damage that is mirrored at the other end,” Alexander Drzezga, principal investigator on the project, told Alzforum. Drzezga compared the situation to a faulty phone connection: If the talker suddenly goes mute, the listener hangs up. Likewise, if one brain region stops sending out signals, the recipient will not receive as many, and its metabolism will drop. The data appeared in a 2012 supplement of the Journal of Nuclear Medicine.
“This study for the first time explains distant effects on glucose metabolism through connectivity to areas of high amyloid deposition,” wrote Victor Villemagne, University of Melbourne, Australia, to Alzforum in an e-mail. Previous studies have started with areas of high amyloid deposition and examined how they were functionally connected to other areas, he said. “This study reverses that paradigm, placing the seed in a region of hypometabolism and tracking which regions are connected,” he told Alzforum. “This reversal is very original, adding to our knowledge of how different regions interact which other, and how amyloid deposition in one area might affect the functioning of a distant region.”
In the future, Drzezga’s research group plans to further explore a cross-sectional and longitudinal interrelation between amyloid burden, hypometabolism, and functional disconnection. In particular, they will check longitudinally for changes in AD patients’ connectivity and see how those changes relate to areas of amyloid deposition down the road. “The ultimate goal of this research is to understand the pathophysiology of neurodegeneration in Alzheimer’s disease,” Förster told Alzforum. “With these imaging modalities, we have the unique opportunity to follow the neuropathological changes in vivo and get a signature of this disease.”—Gwyneth Dickey Zakaib.