Barthélemy · Nature medicine 2024 · multi-cohort diagnostic accuracy study · n=1,759

Highly accurate blood test for Alzheimer's disease is similar or superior to clinical cerebrospinal fluid tests.

Cited 440 times in the scientific literature.

Level 2 - randomized trial

Multi-cohort diagnostic accuracy study validating a blood biomarker against established reference standards (PET imaging and CSF assays).

PubMed 38382645 · doi:10.1038/s41591-024-02869-z · record verified 2026-08-28

What was done

Researchers evaluated the diagnostic accuracy of a mass spectrometry-based blood test measuring plasma %p-tau217 (ratio of phosphorylated tau-217 to non-phosphorylated tau) against FDA-approved cerebrospinal fluid (CSF) assays (Aβ42/40 and p-tau181/Aβ42). The test was evaluated in the Swedish BioFINDER-2 cohort (n = 1,422) and the US Knight ADRC cohort (n = 337), using brain amyloid-β (Aβ) and tau positron emission tomography (PET) imaging as the reference standards. Primary analyses focused on cognitively impaired participants (BioFINDER-2: n = 720; Knight ADRC: n = 50).

What was found

Plasma %p-tau217 matched FDA-approved CSF tests in classifying Aβ-PET status, with both achieving areas under the curve (AUC) between 0.95 and 0.97. For tau-PET classification, plasma %p-tau217 was generally superior to CSF tests, yielding AUCs of 0.95 to 0.98. In cognitively impaired individuals, plasma %p-tau217 demonstrated 89% to 90% accuracy, positive predictive value, and negative predictive value for Aβ-PET, and 87% to 88% for tau-PET, which improved to 95% using a two-cutoffs approach.

Why it matters

High-performing plasma biomarkers offer a minimally invasive, scalable alternative to lumbar punctures and PET scans for identifying Alzheimer pathology, facilitating timely patient selection for disease-modifying therapies.

Limits

The Knight ADRC cognitively impaired subset was small (n = 50). Mass spectrometry platforms can be less accessible in routine clinical laboratories than standard immunoassays, and performance in broader community-based primary care populations was not evaluated.

Cited by