Even though, overall, Alzheimer’s disease robs all people it afflicts of mind and memory, it progresses differently in different people. Its specific symptoms, their sequence, and how fast they worsen can vary quite a bit from one person to another. This used to be a tough nut to crack for researchers, but now, by combining data from multiple PET tracers in longitudinal study cohorts, they are beginning to decipher which regional pathologies provoke particular disease manifestations. The Human Amyloid Imaging conference, held January 15–17 in Miami, showcased the latest findings tying tangles, or plaques, to particular behavioral and cognitive impairments. Intriguingly, links between pathology and subtle symptoms emerged even in people who were still cognitively healthy. The data suggest that multitracer, and also multimodal, brain imaging may eventually allow researchers to predict how disease will progress in a given person.
- Tangles in specific brain regions cause depression; in others, psychiatric symptoms.
- Tangles in medial temporal regions underlie subtle cognitive deficits in unimpaired people.
- Plaques play a smaller role, but can correlate with very early deficits and apathy.
Scientists’ budding ability to parse which regional pathologies and cellular processes underlie particular symptoms will grow as new tracers come onto the scene.
At HAI alone, scientists introduced first-in-human or late preclinical data for tracers recognizing α-synuclein, the α7 nicotinic receptor, the muscarinic cholinergic receptor, microtubules, activated astrocytes, and cyclooxygenase enzymes that mark inflammation.
Regional Tangles and Behavior. Tau PET signal in highlighted brain regions correlates with mild behavioral impairment in 91 cognitively impaired older people. [Courtesy of Firoza Lussier, Serge Gauthier, and Pedro Rosa-Neto.]
Do Regional Tangles Cause Psychiatric Symptoms?
Several speakers tied behavioral symptoms to tangle accumulation in specific brain regions. Jennifer Gatchel and Bernard Hanseeuw of Massachusetts General Hospital, Boston, presented data from 252 participants in the Harvard Aging Brain Study, mean age 73. At the outset, all were cognitively healthy and not depressed. That said, those among them who at baseline had more tangles in their entorhinal or inferior temporal cortices on flortaucipir scans gradually worsened on the Geriatric Depression Scale over an average of six years of follow-up. Their baseline cortical amyloid had no effect on depression.
A separate, albeit cross-sectional, study confirms a link between tangles and depression. Beau Ances of Washington University in St. Louis examined the relationship between tau pathology and depression in 301 cognitively normal participants seen at the Knight Alzheimer’s Disease Research Center. Those whose flortaucipir signal was elevated across the brain were twice as likely to be depressed as were their peers with fewer tangles. In this group, too, amyloid plaques were not associated with depression.
Ances noted that 42 percent of depressed people in this study had tangles but no plaques, i.e., suspected non-AD pathology (SNAP). Perhaps mood disorders like depression characterize SNAP, he speculated. Oddly, antidepressant use correlated with an even higher risk of depression in this study. This could be because depressed people are more likely to be prescribed antidepressants. However, Ances noted that some participants were on antidepressants for other reasons. He believes the data hint at a deleterious role for antidepressants in people who are tau-positive, but noted that larger and longitudinal studies are needed to confirm this.
Other scientists linked tangles to psychiatric symptoms more generally. Cécile Tissot and Pedro Rosa-Neto of McGill University in Verdun, Canada, compared AZD4694 amyloid PET and MK-6240 tau PET to the global score on the neuropsychiatric inventory questionnaire (NPI-Q) in 35 people with mild cognitive impairment and 28 people with AD dementia. Tangle pathology, particularly in the precuneus, hippocampal formation, and frontal cortex, correlated with worse scores on the NPI-Q. Amyloid PET and volumetric MRI did not.
Besides depression, the NPI-Q measures delusions, hallucinations, agitation, disinhibition, and sleep problems. The researchers reported seeing links between those symptoms and regional tangles. Delusions correlated with tracer uptake in the ventromedial prefrontal cortex and cuneus, hallucinations with uptake in the cuneus. This makes sense, the authors noted, because the cuneus is involved in vision, and the ventromedial prefrontal cortex in inhibition. For its part, agitation correlated with tracer signal in the frontal cortex and precuneus, parts of the default mode network. Disinhibition correlated with uptake in the precuneus, dorsomedial prefrontal cortex, and temporal poles, areas involved in emotional processing. Unsurprisingly, motor disturbances were seen in people with tangles in the motor cortex. Getting up at night, or too much daytime napping, correlated with having tangles in the ventrolateral prefrontal cortex, another area involved in inhibition and, indeed, in sleep and memory consolidation (e.g., Cowan et al., 2020).
Firoza Lussier, from the same McGill team, linked mild behavioral impairment (MBI) in older adults to their tangle burden. The clinical term MBI recently has been proposed as denoting a cluster of symptoms, including mood changes, lack of motivation, weak impulse control, and socially inappropriate behavior, that precede or go along with cognitive decline. MBI is assessed via a 34-item checklist (Ismail et al., 2017).
Lussier analyzed 91 cognitively impaired participants in the McGill Translational Biomarkers of Aging and Dementia (TRIAD) cohort, who were on average 71 years old. A person’s total MBI score associated with his or her having tangles in the orbitofrontal cortex, posterior cingulate cortex, precuneus, cuneus, and lateral temporal lobes (see image above). In 94 cognitively healthy controls, by contrast, there was little MK-6240 uptake, or MBI.
While most talks linked tau pathology to behavior, a separate study from Tissot suggests plaques can play some role, too. She assessed 37 cognitively healthy, 18 MCI, and seven AD participants in the TRIAD study. All underwent AZD4694 amyloid PET, as well as scans with the PET tracer PBR28, which binds mitochondrial translocator protein and acts as a marker of inflammation (Wang et al., 2009; Donat et al., 2018). Uptake of both tracers in the mediofrontal cortex and the anterior nucleus of the thalamus picked out people who scored high on the Apathy Inventory. The mediofrontal cortex helps regulate behavior, while the thalamus relays communication between brain regions. The findings suggest that plaques and inflammation together may make a person apathetic, the researchers said. Apathy is the most common behavioral symptom of AD, and gets worse as the disease becomes more severe.
The Pattern of Plaques or Tangles Predicts Cognitive Impairment
Other research links regional pathology to particular cognitive deficits. Michelle Farrell and Reisa Sperling of Massachusetts General Hospital reported at HAI that low levels of plaque accumulation suffice to precipitate subtle cognitive deficits in people who are not measurably impaired. In 67 cognitively normal participants in the Harvard Aging Brain Study who initially had a low amyloid load by PiB PET, plaque accumulation over four years predicted subsequent declines on the Buschke Selective Reminding Test and the Digit Symbol Substitution Test over the next two years. These tests measure executive function and processing speed.
In this cohort, tangles alone, as seen by flortaucipir PET, did not affect decline. However, in participants whose baseline amyloid load was high—1.22 SUVR or more—a flortaucipir signal in the inferior temporal cortex did presage subsequent decline on the Selective Reminding Test. The findings strengthen previous data suggesting that amyloid by itself can trigger subtle cognitive problems early in disease, but that tangles are responsible for decline later on.
A sensitive and quick new cognitive test detects early damage wrought by both plaques and tangles. Danielle Mayblyum and Emma Thibault of Mass. General, working with Keith Johnson and Dorene Rentz, evaluated 52 cognitively healthy and nine cognitively impaired participants with the digital clock drawing test (Dec 2017 conference news). The more PiB a person retained in his or her frontal, lateral temporal, and retrosplenial cortices, the worse they performed on this test. MK-6240 signal in entorhinal and inferior temporal cortex, amygdala, and hippocampus also correlated with worse performance, suggesting that both pathologies blunt the skills required for this task.
What are the earliest cognitive consequences of having tau tangles? Working with Elizabeth Mormino of Stanford University School of Medicine, Tyler Toueg examined PI-2620 tau PET scans in 39 participants in the Stanford Memory and Aging Study. Of these, 30 were cognitively healthy. Nine had subtle deficits, which the scientists defined as scoring more than 1.5 standard deviations below the population mean on a neuropsych test, or having a family member who expressed concerns about his or her memory. These nine were roughly equivalent to stage 2 AD patients as defined by recent NIA-AA recommendations, i.e., they have some cognitive decline even though they still score in the normal range on tests (Apr 2018 news).
In Toueg’s study, the nine had more tau deposition in their medial temporal lobes, including the entorhinal cortex, hippocampus, and amygdala, than did the cognitively normal group. The data suggest that tangle accumulation in these regions can precipitate cognitive decline. As expected, seven patients from the Stanford Memory Disorders Clinic who had amyloid-positive amnestic MCI or AD dementia had even more tracer uptake in these regions.
There seem to be no hard, dichotomous switches in AD pathogenesis, however. Everything slowly creeps up on the unsuspecting brain. Case in point: Even among cognitively healthy people, a higher tangle burden in the hippocampus accounts for subtle memory deficits, reported Alexandra Trelle, who leads the Stanford Aging and Memory Study. In an overlapping group of 36 cognitively healthy study volunteers, Trelle found that more hippocampal tangles meant worse performance on tests of associative memory and mnemonic discrimination. Hippocampal tangles also came with low levels of CSF Aβ42, a presymptomatic marker of AD.
In this cohort, having tangles only in the entorhinal cortex did not correlate with test scores, nor did CSF Aβ42 and p-tau. Entorhinal tangles precede hippocampal tangles in Braak staging. The findings suggest that tangles start to dull memory once they reach the hippocampus, and also that tau tracer uptake in the hippocampus might be a more sensitive early marker of decline than CSF, the authors noted.
Tangles in a given region can also predict future cognitive decline. Adam Martersteck and Emily Rogalski of Northwestern University Feinberg School of Medicine in Chicago examined 17 people with primary progressive aphasia due to Alzheimer’s disease. Aphasia—problems understanding or expressing speech—results from damage to language centers, such as Broca’s and Wernicke’s areas, in the left hemisphere. At baseline, both tangle burden and cortical thinning in these regions correlated with difficulties on the Boston Naming Test, in which participants name objects from line drawings. In the eight participants who took the test again a year later, however, only the baseline tangle load correlated with decline. Tau PET better predicts decline than does atrophy, these authors concluded. In other words, tangles are an early marker of decline, and atrophy a later marker.
It’s not just objects. Apparently tangles also mess with one’s ability to remember names. Victoria Tennant and William Jagust of UC Berkeley tested 85 cognitively healthy people enrolled in the Berkeley Aging Cohort Study on the Northwestern Famous Faces task. It asks participants to name famous people when shown photos of their faces. The more entorhinal cortical flortaucipir signal a participant had, the worse he or she did. Thinning of the fusiform gyrus also correlated with naming difficulties.
Overall, research in multiple independent groups is converging on the point that tangles in the entorhinal cortex, one of the first regions affected in AD, can already harm cognition.
But wait. David Berron and Oskar Hansson at Lund University, Sweden, implicated an even earlier region. Before appearing in the entorhinal cortex, tangles pop up in the transentorhinal cortex, also known as area 35. From there, they spread to the entorhinal cortex and from there, to the hippocampus.
Berron examined RO-948 tau PET scans in 322 cognitively healthy participants in BioFINDER2. Of these, 241 were amyloid-negative, 81 amyloid-positive. Among the amyloid-positive group, tracer uptake in area 35 correlated with atrophy there, as well as in the anterior hippocampus. The findings suggest that tangles can have both local and remote effects on atrophy, the authors noted.
Area 35 tangles, but not cortical thickness, matched up with poor delayed word recall. Regression analysis showed that this effect was partially mediated by anterior hippocampal atrophy, which explained about 19 percent of the variance in memory. Intriguingly, in 107 amyloid-positive people with MCI, hippocampal atrophy explained 91 percent of the variance in delayed recall. The data imply that other disease mechanisms besides atrophy are at work in very early stages of AD.
On the face of it, this seems to contradict Trelle’s finding that tangles have to reach the hippocampus to harm memory. The studies used different tracers, PI-2620 versus RO-948, and focused on different regions of early tangle accumulation, entorhinal cortex versus area 35, so some of the difference could be methodological. But also, these are early days for the approach of tying symptoms to regional pathologies by PET. Further research will make clear which findings are robust and repeatable, and shed more light on how local pathology leads to functional loss.—Madolyn Bowman Rogers
- Cognitive Testing Is Getting Faster and Better
- New Definition of Alzheimer’s Hinges on Biology, Not Symptoms
- Cowan E, Liu A, Henin S, Kothare S, Devinsky O, Davachi L. Sleep spindles promote the restructuring of memory representations in ventromedial prefrontal cortex through enhanced hippocampal-cortical functional connectivity. J Neurosci. 2020 Jan 17; PubMed.
- Ismail Z, Agüera-Ortiz L, Brodaty H, Cieslak A, Cummings J, Fischer CE, Gauthier S, Geda YE, Herrmann N, Kanji J, Lanctôt KL, Miller DS, Mortby ME, Onyike CU, Rosenberg PB, Smith EE, Smith GS, Sultzer DL, Lyketsos C, NPS Professional Interest Area of the International Society of to Advance Alzheimer’s Research and Treatment (NPS-PIA of ISTAART). The Mild Behavioral Impairment Checklist (MBI-C): A Rating Scale for Neuropsychiatric Symptoms in Pre-Dementia Populations. J Alzheimers Dis. 2017;56(3):929-938. PubMed.
- Wang M, Yoder KK, Gao M, Mock BH, Xu XM, Saykin AJ, Hutchins GD, Zheng QH. Fully automated synthesis and initial PET evaluation of [11C]PBR28. Bioorg Med Chem Lett. 2009 Oct 1;19(19):5636-9. Epub 2009 Aug 15 PubMed.
- Donat CK, Mirzaei N, Tang SP, Edison P, Sastre M. Imaging of Microglial Activation in Alzheimer's Disease by [11 C]PBR28 PET. Methods Mol Biol. 2018;1750:323-339. PubMed.