Subscribe to eTOC

A Blood Test for Tau Is Consistent with PET and CSF Measures for Alzheimer's Dementia

Article In Brief

Researchers found in two different studies that blood tests, which indicated abnormally high levels of a tau protein closely linked to dementia due to Alzheimer's disease (AD), distinguished AD from frontotemporal lobar degeneration. In one study, it also predicted which patients who were cognitively normal or had mild cognitive impairment upon initial evaluation would later develop Alzheimer's dementia.

A blood test for Alzheimer's disease (AD) pathology appears to be as accurate as more invasive measures and equally well correlated with AD dementia, according to two studies published simultaneously in the March issue of Nature Medicine.

The studies measured plasma phosphorylated-tau-181 (p-tau181), a form of the tau protein, and found that abnormally high levels are as closely linked to dementia due to AD as are PET scans of amyloid protein and measures of p-tau181 in cerebrospinal fluid.

Not only did a high level of plasma p-tau181 distinguish AD from frontotemporal lobar degeneration, but in one of the studies it also predicted which patients who were cognitively normal or had mild cognitive impairment (MCI) upon initial evaluation would later develop Alzheimer's dementia.

While neurologists and officials at the National Institutes of Health (NIH) applauded the findings, they emphasized that the results need to be replicated and the p-tau181 test further investigated before it can be made available in the neurology clinic.

“I think we are still a few years away from having this sort of test available to the general neurologist in a clinic setting or in their office,” said Eliezer Masliah, MD, director of the neuroscience at NIH's National Institute on Aging. “But they and their patients should know that these less expensive, less invasive tests are coming. We are hopeful that they will be approved in the near future by the FDA.”

A longtime AD researcher called the findings a “triumph.”

“I've been in this field since 1980 and having a blood test for Alzheimer's has been one of the holy grails,” said Steven T. DeKosky, MD, FAAN, the Aerts-Cosper Professor of Alzheimer's Research at the University of Florida College of Medicine, and associate director of the Florida Alzheimer's Disease Research Center. “This is a remarkable accomplishment. Right now, I'm sure all of us who do PET scan studies would love to be able to use this plasma test as a screen.”

The principal investigator of one of the two studies told Neurology Today that his group is already working to validate plasma p-tau181 in a primary-care setting and to develop a clinical-grade test that could be used in any laboratory.

The test “could be used in patients with either dementia or MCI to improve the diagnostic work-up, especially at specialized clinics and in patients where lumbar puncture or amyloid PET imaging cannot be done,” said Oskar Hansson, MD, PhD, professor in the department of clinical memory research at Lund University in Sweden.


“We and others think that while plasma p-tau181 seems to be very useful, its possible there may be other epitopes that may be even slightly more accurate.”—DR. ADAM L. BOXER

Study Details

Dr. Hansson's group measured plasma p-tau181 levels in three cohorts comprising a total of 589 individuals. The first group included 64 participants who were cognitively unimpaired (with or without a positive amyloid-beta scan), 28 who had MCI, 38 who had AD dementia, and 52 with non-AD neurodegenerative disease. The second cohort, which was followed up for eight years to track conversion to AD dementia, included 219 cognitively unimpaired participants (42 percent of whom were positive for amyloid-beta), and 125 with MCI (65 percent of whom were amyloid-beta positive). A third cohort included pathology from post-mortem tests on 33 individuals:16 had confirmed AD dementia and 47 were non-AD.

The study found a clear association between plasma and CSF levels of p-tau181 in both cohorts one and two (p<0.001). In 174 participants from cohort one, plasma P-tau181 levels predicted positive tau PET scans (area under the curve (AUC) = 0.87-0.91 for different brain regions). Increased plasma p-tau181 was also associated with increased amyloid-beta PET (using a global cortical composite measure) in both cohort one and two (p< 0.001).

In the autopsy-confirmed cohort three, plasma p-tau181 distinguished AD dementia from non-AD neurodegenerative diseases as accurately as did tau PET and CSF p-tau181 (AUC = 0.94-0.98).

The results “show that plasma p-tau181 may be increased early in AD, potentially even in some Aβ+ cognitively unimpaired individuals (preclinical AD),” the paper reported. “Plasma p-tau181 then increased further during the symptomatic (prodromal and dementia) stages of AD. In contrast, plasma p-tau181 was not increased in non-AD. These characteristics mark out plasma p-tau181 as a promising biomarker to track disease progression in AD and to differentiate AD from non-AD conditions, with utility for patient management in clinical practice, research and trials.”

In cognitively unimpaired and MCI subjects, a level of p-tau181 above 1.81 pg mL–1 at baseline was associated with a dramatically increased risk of future AD dementia (p< 0.001). None of the other plasma biomarkers analyzed in the study, including total tau, amyloid-beta42/amyloid-beta 40 and neurofilament light protein, were independently related to risk of AD dementia.

The second study in Nature Medicine, part of the ongoing Advancing Research and Treatment for Frontotemporal Lobar Degeneration study, was led by Adam L. Boxer, MD, PhD, professor and director of the Alzheimer's Disease and Frontotemporal Degeneration Clinical Trials Program at the University of California, San Francisco, Memory and Aging Center. The cohort of 362 participants includes 69 healthy controls, 47 people with MCI, 56 with clinically diagnosed AD, and the remainder with various forms of frontotemporal lobar degeneration.

Dr. Boxer and colleagues found that plasma p-tau181 was increased by 3.5-fold in AD compared with controls and differentiated AD from both clinically diagnosed (AUC = 0.894) and autopsy-confirmed frontotemporal lobar degeneration (AUC = 0.878). The measure also identified individuals who were amyloid-beta positive on PET scans regardless of clinical diagnosis and correlated with cortical tau protein deposition measured by 18F-flortaucipir PET.

As promising as the results were, Dr. Boxer told Neurology Today that his group is looking at the predictive value of other variants of phosphorylated tau.

“We and others think that while plasma p-tau181 seems to be very useful, it's possible there may be other epitopes that may be even slightly more accurate,” he said.

Expert Commentary

Both papers credited a 2018 study published in Alzheimer's & Dementia as the first to find that plasma p-tau181 increases with AD clinical severity and is associated with tau- and amyloid-PET scans. The first author of that study said that she still wants to see the results replicated in more diverse cohorts before the test becomes available outside of research trials.


“The fact that these two studies both nicely complement each other and replicate our findings is really promising. But I would like to see more replication in community-based studies rather than just in studies from memory clinics.”—DR. MICHELLE M. MIELKE

“The fact that these two studies both nicely complement each other and replicate our findings is really promising,” said Michelle M. Mielke, PhD, professor of epidemiology and neurology at the Mayo Clinic. “But I would like to see more replication in community-based studies rather than just in studies from memory clinics.”

Her 2018 study, part of the ongoing Mayo Clinic Study of Aging, is now looking not only at plasma p-tau181 but also at plasma p-tau217.

“Previously there weren't the technologies available to look at [p-tau] 217,” she said. “Historically, p-tau181 was the first to come out, and so that tended to be the focus. Which one is more accurate is still to be determined. And other phosphorylated tau fragments are also being looked at as well.”

Two studies, which were published last year, analyzed data from the Framingham Heart Study and found that total tau in serum is a biomarker not only for dementia but also for stroke risk.

“We wanted to also look at phosphorylated tau in plasma but at the time the measures were not sensitive enough,” said the senior author of both studies, Sudha Seshadri, MD, FAAN, professor of neurology at Boston University School of Medicine and founding director of the Glenn Biggs institute for Alzheimer's & Neurodegenerative Diseases at UT Health San Antonio.

“Total tau was a very good marker in our hands for all types of dementia and stroke risk. But it does seem that total tau is sensitive but perhaps not as specific. It seems to go up with a variety of insults to the brain.”

Based on the two new papers, she said, plasma P-tau181 looks to be more specific for AD dementia.”

Dr. Seshadri will be measuring plasma p-tau181 in the Framingham Study to see how it compares in a community setting. “I would also like to see how it works in Hispanics in south Texas, in the African-American population, and in other groups,” she said.

While the new studies involved hundreds of patients, the results still need to be replicated in far larger groups, Dr. Masliah said. NIA, which supported both studies in part, will likely fund such follow-up studies in its consortium of 32 AD centers.

“What we're looking into is pulling samples from all those centers so they can be tested and verified by other technologies and then followed longitudinally,” Dr. Masliah said. “These are very, very exciting results, very important results, something we need to follow up on.” If replicated in larger, more diverse cohorts, the plasma ptau181 test “could be a game changer for clinical trials,” he said.


Dr. Hansson has received research support (for his institution) from Roche, GE Healthcare, Biogen, AVID Radiopharmaceuticals, and Euroimmun. He has received consultancy/speaker fees from Biogen and Roche. Dr. Boxer has served as a consultant for Aeton, Abbvie, Alector, AGTC, Amgen, Arkuda, Arvinas, Asceneuron, Esai, Ionis, Lundbeck, Novartis, Passage BIO, Sangamo, Samumed, Third Rock, Toyama and UCB. He has received research support from Avid, Biogen, BMS, C2N, Cortice, Eisai, Eli Lilly, Forum, Genentech, Janssen, Novartis, Pfizer, Roche, and TauRx.

Link Up for More Information

• Thijssen EH, La Joie R, Wolf A, et al; for the Advancing Research and Treatment for Frontotemporal Lobar Degeneration (ARTFL) investigators. Diagnostic value of phosphorylated tau181 in Alzheimer's disease and frontotemporal lobar degeneration Nat Med 2020;26(3):387–397.
    • Janelidze S, Mattsson N, Palmqvist S, et al. Plasma p-tau181 in Alzheimer's disease: Relationship to other biomarkers, differential diagnosis, neuropathology and longitudinal progression to Alzheimer's dementia Nat Med 2020;26(3):379–386.
      • Mielke MM, Hagen CE, Xu J, et al. Plasma phospho-tau181 increases with Alzheimer's disease clinical severity and is associated with tau- and amyloid-positron emission tomography Alzheimers Dement 2018;14(8):989–997.
        • Pase MP, Beiser AS, Jimali JJ, et al. Assessment of plasma total tau level as a predictive biomarker for dementia and related endophenotypes JAMA Neurol 2019;76(5):598–606.
          • Pase MP, Himali JJ, Aparicio HJ, et al. Plasma total-tau as a biomarker of stroke risk in the community Ann Neurol 2019;86(3):463–467.