Skip to main content
Premium Trial:

Request an Annual Quote

Lund Team IDs Plasma Phosphorylated Tau as Protein Biomarker for Alzheimer's


NEW YORK – A team led by researchers at Lund University have found that plasma levels of the protein phosphorylated tau181 (p-tau181) are indicative of tau aggregates in the brain and linked to Alzheimer's disease.

The findings, detailed in a study published last week in Nature Medicine, suggest that blood-based testing for Alzheimer's-linked tau brain deposits is feasible and that such tests could be combined with blood-based assays for markers like amyloid-β 40 and amyloid-β 42 to improve Alzheimer's testing, said Oskar Hansson, professor of clinical memory research at Lund University and senior author on the study.

Hansson said that he and his colleagues are now working to validate the marker in a clinical study enrolling primary care patients throughout southern Sweden.

One characteristic of Alzheimer's disease are tau aggregates in the brain, and total tau and phosphorylated tau are commonly measured in the cerebrospinal fluid of patients suspected of having Alzheimer's. Tau aggregates can also be detected using positron emission tomography.

These markers have drawbacks, however. Sample CSF requires a lumbar puncture, an invasive procedure that many patients, especially asymptomatic ones, wish to avoid. PET imaging, meanwhile, is an expensive and resource-intensive procedure and therefore poorly suited to wide-scale use, for instance for population screening or selecting patients for drug trials.

This has driven interest in blood-based markers for Alzheimer's disease. Blood is a relatively cheap and easy sample to collect. However, the relevant markers are likely present in lower concentrations in blood than in CSF, meaning blood-based assays may need to be more sensitive. Additionally, there are questions about whether the levels of these markers in blood accurately reflect their presence in the brain.

In recent years, researchers have found that markers like amyloid-β can be accurately measured in blood and do correlate with CSF-based tests and PET imaging. For instance, diagnostics firm C2N Diagnostics has presented data on its blood-based amyloid-β test showing that in 415 blood samples collected at six sites across the US the test predicted the presence of amyloid in the brain with an area under the receiver operator curve of .86 and an AUC of .90 when the test results were combined with age and ApoE4 status.

For the Nature Medicine study looking at p-tau181, Hansson and his team set out to determine how well plasma p-tau181 levels correlated with the presence of tau aggregates in the brain and with a patient's Alzheimer's status.

They looked at three separate cohorts. The first included 182 subjects, 64 cognitively normal patients (60 percent of whom were positive for PET amyloid-β), 28 patients with mild cognitive impairment, 38 patients with Alzheimer's dementia, and 52 with non-Alzheimer's neurodegenerative conditions. The second included 219 cognitively normal patients (42 percent positive for PET amyloid-β), 125 patients with mild cognitive impairment (65 percent positive for PET amyloid-β). Both of these cohorts were followed longitudinally for up to eight years to track their progression.

The third cohort consisted of 16 subjects with Alzheimer's confirmed by autopsy and 47 subjects confirmed by autopsy not to have Alzheimer's.

The study found that blood-based p-tau181 measurements correlated both with CSF p-tau181 measurements and positive PET tau results. It also distinguished between Alzheimer's dementia and non-Alzheimer's dementia with accuracy similar to that provided by CSF and PET tau measurements. Additionally, elevated blood levels of p-tau181 were linked to development of Alzheimer's dementia in cognitively normal individuals and subjects with mild cognitive impairment.

Comparing p-tau181 to other markers, Hansson said he and his colleagues found that plasma p-tau181 was more accurate than plasma amyloid-β for distinguishing between Alzheimer's and other neurodegenerative conditions but that combination of plasma p-tau181 and plasma amyloid-β was better than either measure alone at identifying cognitively normal individuals with amyloid brain pathology.

A separate Nature Medicine study led by researchers at the University of California, San Francisco and also published last week similarly found that plasma p-tau181 levels correlated with PET tau measurements and were able to distinguish between Alzheimer's and other neurodegenerative conditions. The UCSF study also found that plasma p-tau181 levels were 3.5 times higher in patients with Alzheimer's compared to controls.

The Lund researchers are now working to further validate the marker and have launched a study called ADetect that has recruited 600 individuals with mild to moderate cognitive symptoms from 15 primary care centers in southern Sweden, Hansson said.

Evaluation of cognitive symptoms along with blood samples will be done at the primary care centers while all subjects will also undergo assessments at the Lund University memory clinic consisting of neuropsychological testing, brain MRI, and measurement of CSF biomarkers. Hansson said he hoped to finish the study in the next two years.

He added that in addition to clinical validation, researchers would need to establish assays for measuring the marker on a widely used clinical analyzer platform and noted that "such work is currently ongoing."

The researchers used the Meso Scale Discovery platform for the p-tau181 measurements. They also used Quanterix's Simoa technology in the study to measure the marker neurofilament light chain (Nfl), which has shown promise as a biomarker for Alzheimer's and other neurodegenerative diseases. While the Simoa technology could provide high sensitivity measurements of plasma p-tau181, Hansson said he and his colleagues had not seen any advantage of a high-sensitivity assay for the marker.

The researchers are also exploring whether other phosphorylated tau forms might be effective Alzheimer's markers. Hansson cited a paper the team has in press at Nature Communications and which they published as a MedRxix preprint in January that found that CSF p-tau217 was more effective than p-tau181 at predicting brain amyloid burden as measured by PET and at distinguishing between Alzheimer's dementia and other neurodegenerative disorders.

He said that they are now looking into whether p-tau217 can be reliably measured in plasma and if it offers the same advantages compared to plasma p-tau181.