This story has been modified from its original version.
Researchers have long known that tau pathology correlates more closely with the cognitive decline of Alzheimer’s disease than does Aβ pathology. They have predicted that being able to measure tau aggregates in the living brain would bring within reach a diagnostic and prognostic marker. At the Alzheimer's Association International Conference 2014, held July 12-17 in Copenhagen, Denmark, researchers reported that uptake of tau ligands into the brain as seen in positron emission tomography (PET) indeed correlates with memory decline. While this is good news, it is but a small step. "We are in the early days of tau PET and need to do much more work to be confident about what these ligands bind to in tissue. That said, tau imaging looks very promising," Brad Dickerson told Alzforum.
Dickerson, with Keith Johnson and colleagues at Massachusetts General Hospital in Boston and Charlestown, have performed some of the first clinical studies using the PET tau ligand T807, which is being developed by Eli Lilly. Johnson has examined binding of T807 in older adults who take part in the Harvard Aging Brain Study. This study tracks how approximately 250 cognitively normal people perform over time on challenging memory tests. It also peers into the brain using MRI and PET imaging of Aβ to see which pathological changes might precede or accompany changes in memory.
At AAIC, Johnson reported that among 56 people who underwent tau imaging, those who showed the greatest uptake of T807 in the inferior temporal lobe were also those who had deteriorated most in the previous three years on the selective reminding test. "Though preliminary, these findings suggest the spread of tau to the temporal neocortex signals memory decline in clinically normal older people," Johnson said. One implication is that tau imaging could be used for early diagnosis.
These emerging data fit with pathological staging of Alzheimer's disease described in the early 1990s (see Braak and Braak, 1991). Braak staging shows tau pathology spreading from the medial temporal lobe, mostly the hippocampus, into the cortex as the disease progresses. Interestingly, Johnson found that Aβ pathology also correlated with deterioration on the reminding test in this cohort of normal adults, but it did so independently of tau. The data suggest that tau spreads similarly in normal aging and in AD. "Basically, there is an age-related and an AD-related tau pathology, which share many common features," said Johnson. "The relationship between those two is of great interest."
Like Dickerson, Johnson emphasized that researchers still know too little about tau ligands and exactly what they bind to and where. Hints that certain ligands bind better to certain isoforms of tau, for example, raise the possibility of using specific tracers to diagnose specific diseases.
At AAIC, Makoto Higuchi from the National Institute of Radiological Sciences, Chiba, Japan, reported that the tau ligand PBB3 binds to tissue from Pick's disease patients, whose tau aggregates mostly comprise the protein’s three-repeat isoform. Other tauopathies, such as corticobasal degeneration and progressive supranuclear palsy, contain only four-repeat tau, while AD contains a mix.
Victor Villemagne, Austin Health, Melbourne, Australia, presented on THK-5117, a third tau ligand. While THK-5117 binds to PHF and the 4R lesions in PSP and CBD, it does not bind in Pick's disease. Then again, Higuchi reported that PBB3 does not bind to tau aggregates in hippocampal tissue from PS19 mice, which carry the human P301S tau mutation, yet it does bind to tau in the hippocampus of the rTG4510 model, which carries the P301L form. Interestingly, PBB3 bound to brainstem tau in PS19 animals. Higuchi suggested that there might be various toxic forms of tau that bind differently to the current crop of PET ligands.
"There is so much variability from case to case, and so many potentially confounding factors, that it will take a lot of effort to figure this out," Johnson agreed. He predicted that insight will come from postmortem correlations of tau pathology with prior PET scans taken when the study volunteer was alive. To date, researchers in the field have relied mostly on studying postmortem tissue using radiolabelled tau tracers, and in vivo studies are still gearing up.
As is the case with T807, researchers are getting a better handle on the properties and characteristics of PBB3 and THK-5117. In Copenhagen, Villemagne reported that THK-5117 has better binding kinetics than its forerunner, THK-5105. THK-5117 also seems to bind with greater specificity, since about 60 percent more gets taken up in AD brains than controls. That’s about twice as much as for THK-5105, which has but a 1.3-fold difference between cases and controls, said Villemagne. In 14 patients scanned with THK-5117 thus far, he observed higher binding in those whose MMSE scores were lower. Retention also correlated with loss of hippocampal volume. That correlation emerged in controls as well, though it was less robust.
As for PBB3, the number of patients scanned with this ligand has by now increased to 28—10 healthy controls, five people with MCI, and 13 with AD. As with THK-5117, binding of PBB3 correlated with MMSE scores, Higuchi said. Stay tuned on this fast-evolving topic.—Tom Fagan
Research Models Citations
- Braak H, Braak E. Neuropathological stageing of Alzheimer-related changes. Acta Neuropathol. 1991;82(4):239-59. PubMed.