05 Apr 2025

Positive signals in tau PET scans and p-tau217 assays both suggest cognitive decline. Which performs better? Neither, according to a study published March 28 in Nature Aging—at least in people who are cognitively normal. Scientists led by Rik Ossenkoppele, University of Amsterdam, and Oskar Hansson at Lund University, Sweden, directly compared the two markers. Both equally predicted change in a composite cognitive test. Combining the two was slightly better. Still, using the plasma test first, then PET, reduced the number of people needed for a clinical trial more than did using plasma p-tau217 alone. The scientists concluded that while the two markers perform similarly, they yield different and complementary information. Sequential testing, they think, would work best to screen candidates for clinical trials.

Previously, Hansson and others had reported that tau-PET beat out p-tau217 in predicting further decline among people who were already cognitively impaired (Smith et al., 2023; Mundada et al., 2023). But PET measures neurofibrillary tangles, which accumulate relatively late in the disease process, whereas p-tau217 reflects amyloid-driven phosphorylation of soluble tau, which occurs earlier. How would the two markers perform in people who are cognitively healthy?

First author Ossenkoppele tested this in 1,474 cognitively unimpaired people, 408 of whom had tested positive for amyloid in the brain. All had the plasma and PET measures taken at baseline. The volunteers came from nine cohorts: Amsterdam Dementia Cohort; BioFinder-1; BioFinder-2; Knight ADRC; Australian Imaging Biomarkers and Lifestyle Study of Ageing; Mayo Clinic Olmsted Study of Aging; Pre-symptomatic Evaluation of Experimental or Novel Treatments for Alzheimer’s Disease; Translational Biomarkers in Aging and Dementia; and the Wisconsin Registry for Alzheimer’s Prevention. Ossenkoppele used z scores to account for any heterogeneity in measurements across the different clinical centers and assays involved. Z scores measure difference from a mean, rather than absolute values.

Over eight years, people in the highest p-tau217 quartile were much likelier to score worse on the modified preclinical Alzheimer cognitive composite (mPACC5), a sensitive test of cognition often used in healthy cohorts. Falling into the highest quartile for tau-PET signals in the medial temporal lobe (MTL) or the neocortex similarly predicted a declining mPACC5 score. Linear regression coefficients for the plasma and two PET measures were 0.33, 0.34, and 0.33, respectively, showing virtually no difference. Likewise, all three markers similarly predicted a person’s chance of being diagnosed with mild cognitive impairment over the next six years (image below). Data from the 408 amyloid-positive people alone painted the same picture.

Chances of MCI. People who scored in the highest quartile for plasma p-tau217 (left), MTL tau-PET (center), or neocortical tau-PET (right), were more likely to develop mild cognitive impairment over six years. [Courtesy of Ossenkoppele et al., Nature Aging, 2025.]

What about combining the markers? Lumping MTL or neocortical tau-PET in with the plasma marker increased performance slightly, bumping up the regression coefficient by 0.02 to 0.35. Though small, this was statistically significant. Further, the plasma and PET measure independently accounted for 14 and 17 percent of the variance in the mPACC5, the remainder being due to both markers combined (22 percent) and other covariates, such as age, sex, years of education, and APOE genotype. The upshot was that plasma and MTL PET measures complemented each other to account for 53 percent of the change in cognition (image below). Combining neocortical PET with p-tau217 similarly increased the predictive value slightly. The authors concluded that while the plasma and PET measures similarly predict cognitive decline, they reflect different aspects of pathology and complement each other.

Given this finding, is there value in using both? The field has been considering a two-step approach for identifying cognitively impaired people who would benefit from treatment. In this scenario, someone whose plasma p-tau217 seems neither definitively negative nor positive would be sent for further screening using tau-PET (Nov 2023 news). Ossenkoppele and colleagues propose that a similar strategy could work to identify healthy people for inclusion in prevention trials. Someone who has high plasma p-tau217 would be sent for brain imaging. If their PET scan came back positive, this would suggest a high likelihood for dementia (image below).

The Two-Step. Sequential screening for plasma p-tau217 and tau-PET could dramatically reduce the number of people needed for a secondary prevention trial. [Courtesy of Ossenkoppele et al., Nature Aging, 2025.]

Ossenkoppele and colleagues calculated that this strategy could reduce the number of cognitively unimpaired (CU) people needed for a prevention trial by up to 94 percent if only those people in the top quartiles were selected. Using plasma p-tau217 alone would reduce the number by 76 percent. “This information would become even more crucial if treatments such as lecanemab and donanemab are found to be effective in preclinical AD, as this would require large-scale screening of CU populations for AD pathology,” they write.—Tom Fagan

Comments

No Available Comments

Make a Comment

To make a comment you must login or register.

References

News Citations

1. Plasma p-Tau-217 Assays Work Well, But No Home Run for Diagnosis 30 Nov 2023

Paper Citations

1. Smith R, Cullen NC, Pichet Binette A, Leuzy A, Blennow K, Zetterberg H, Klein G, Borroni E, Ossenkoppele R, Janelidze S, Palmqvist S, Mattsson-Carlgren N, Stomrud E, Hansson O, Alzheimer's Disease Neuroimaging Initiative. Tau-PET is superior to phospho-tau when predicting cognitive decline in symptomatic AD patients. Alzheimers Dement. 2023 Jun;19(6):2497-2507. Epub 2022 Dec 14 PubMed.
2. Mundada NS, Rojas JC, Vandevrede L, Thijssen EH, Iaccarino L, Okoye OC, Shankar R, Soleimani-Meigooni DN, Lago AL, Miller BL, Teunissen CE, Heuer H, Rosen HJ, Dage JL, Jagust WJ, Rabinovici GD, Boxer AL, La Joie R. Head-to-head comparison between plasma p-tau217 and flortaucipir-PET in amyloid-positive patients with cognitive impairment. Alzheimers Res Ther. 2023 Sep 22;15(1):157. PubMed.

Further Reading

No Available Further Reading