COPENHAGEN – Detection of tau neurofibrillary tangles in the brain via PET imaging highly correlates with memory decline and could serve as an accurate way to identify individuals at high risk of Alzheimer’s disease and track disease progression, according to the results of a prospective study.

In a group of cognitively normal elderly subjects, PET imaging with an investigational tau-specific radioligand showed a significant association between increased tau binding and progressive memory decline over 3 years. A second study found that increased binding effectively discriminated cognitively normal subjects from those with mild cognitive impairment (MCI) and Alzheimer’s dementia.

The studies hold great promise for the future of Alzheimer’s drug development, Dr. Keith Johnson said at the Alzheimer’s Association International Conference 2014.

"At this early stage of tau PET imaging, we do think we are able to visualize the second of the two cardinal pathologic processes," said Dr. Johnson, codirector of the Neuroimaging Program of the Massachusetts Alzheimer’s Disease Research Center at Harvard University, Cambridge, Mass. "Our initial findings indicate that the spread of tau to widespread cortical regions, in the presence of amyloid plaques, may be the smoking gun – the sign that indicates cognitive impairment is either imminent or underway."

The compound, [¹⁸F]T807, is being developed by Avid Radiopharmaceuticals, a subsidiary of Eli Lilly.

The first study by Dr. Johnson and his colleagues looked at [¹⁸F]T807 binding in a group of 56 cognitively normal subjects with a mean age of 72 years. All were members of the Harvard Aging Brain Study. The ongoing project tracks amyloid-beta deposition with PET imaging, as well as other Alzheimer’s biomarkers, in a group of cognitively normal, healthy subjects. Each participant also undergoes an annual test of cognition.

Dr. Johnson’s investigation examined the relationship between tau binding and memory performance as assessed retrospectively with the six-trial Selective Reminding Test over a median 3-year period. The linear regression model controlled for amyloid-beta plaque burden, age, and education.

Over the study period, the investigators found a highly significant relationship between memory performance and [¹⁸F]T807 binding in the entorhinal cortex (P less than .008) and the inferior temporal cortex (P less than .006). Amyloid-beta binding was significantly associated with tau binding in both regions.

When the model included tau and amyloid-beta plaque burden, the relationship between memory performance and inferior temporal tau burden remained significant in the inferior temporal region, whereas memory performance was no longer associated with tau binding in the entorhinal cortex. The amyloid burden was a significant, independent predictor of memory decline in both models.

The second study demonstrated the ligand’s ability to discriminate between normal and impaired memory. This study cohort comprised 51 individuals (mean age, 74 years); 40 were cognitively normal, 6 had MCI, and 5 had Alzheimer’s disease. They also underwent both tau and amyloid-beta PET imaging.

Among the cognitively normal subjects, [¹⁸F]T807 uptake was concentrated in the hippocampus. It varied considerably in the temporal allocortex, and occurred at low levels in the neocortex.

Compared with the cognitively normal subjects, the MCI and Alzheimer’s patients showed significantly greater binding in the neocortical regions, particularly in the temporal and occipital lobes. Binding extended into the parietal and frontal regions in patients with advanced symptoms.

Among all subjects, those with greater entorhinal binding performed significantly worse on a measure of delayed recall. Worse memory performance was also associated with higher amyloid-beta binding.

"We saw a progressive increase in cortical tau going from the normal subjects to those with Alzheimer’s dementia, and this pattern was different from amyloid," Dr. Johnson said. "The amyloid plaques were more diffuse and appeared at an earlier stage than the tau tangles. Our hypothesis is that the tau appears closer to the action, indicating incident cognitive impairment. It shows that these are the people who have or are about to have cognitive impairment, while the amyloid scan shows a much earlier stage of the pathology."

Although these studies are preliminary, they show that tau imaging may benefit research as much as amyloid imaging has, Dr. Johnson said.

"Ten years ago, amyloid plaque detection changed much of what we do in Alzheimer’s research, enabling us to identify the pathology years before any symptoms develop. It’s had a huge impact on drug development, allowing us to focus in on the right people for research cohorts. Unfortunately, amyloid has only been half of the story."

This new ability to see tau neurofibrillary tangles in vivo "should eventually enable us to more efficiently develop preventive therapies and even treatments for the actual symptoms," he said.

Dr. Johnson has been an adviser or consultant for Siemens Healthcare Diagnostics, Genzyme, and Piramal Healthcare. He has received research funding from Avid Radiopharmaceuticals, Eli Lilly, and other companies.

msullivan@frontlinemedcom.com

On Twitter @alz_gal

Investigational agent could accurately predict memory decline