Local research has shown measuring brain-derived p-tau217 in blood has superior accuracy than measuring other p-tau species for detecting Alzheimer’s disease (AD)-related amyloid pathology in older adults in Singapore.
Plasma brain-derived p-tau217 identified amyloid-beta positivity on PET with an area under the curve (AUC) of 0.965 (95 percent confidence interval [CI], 0.942–0.987), outperforming total-p-tau217 and total or brain-derived p-tau181 and p-tau231 (AUCs, 0.823–0.937; p≤0.008 across all comparisons). [J Prev Alzheimers Dis 2026;13:100615]
At a cutoff point of 11.691 NULISA Protein Quantification (NPQ) units, plasma brain-derived p-tau217 achieved a sensitivity and specificity of 92 percent, a negative predictive value (NPV) of 96 percent, and a positive predictive value (PPV) of 86 percent.
“No other plasma brain-derived and total p-tau species reached >90 percent in both sensitivity and specificity, although these p-tau species achieved high NPVs between 89 percent and 95 percent, as well as PPVs between 73 percent and 83 percent, except for total-p-tau181, which yielded a PPV of 57 percent,” the authors noted.
When applying a three-range approach using a threshold of 95-percent sensitivity and specificity, plasma brain-derived p-tau217 achieved a PPV of 90 percent and NPV of 97 percent. The proportion of individuals classified as having intermediate risk based on amyloid positivity was only 7 percent with brain-derived p-tau217 as opposed to between 20 percent and 57 percent with other plasma brain-derived p-tau and total p-tau species.
The three-range approach, according to the authors, was more stringent than the 90-percent sensitivity and specificity currently recommended by the Global CEO Initiative (GCI) on AD for blood-based biomarkers to be used as diagnostic tests for amyloid pathology.
“Since intermediate results are not informative of brain amyloid status, the GCI Workgroup recommended that a blood test should have intermediate values in no more than 15 percent of the cohort, a benchmark met only by brain-derived p-tau217,” the authors pointed out.
“Taken together, our data suggest that of the six biomarkers we evaluated in the current study, plasma brain-derived p-tau217 stood out as the only biomarker that achieved all the CGI Workgroup’s proposed benchmarks for both the three-range and single cut-off diagnostic performance for brain amyloid pathology,” they said.
A total of 213 older adults (mean age 73 years, 56.8 percent female), recruited from memory clinics and the community in Singapore, participated in the study. Of these participants, 44 were cognitively normal, 107 were cognitively impaired with no dementia, and 62 had dementia.
PET imaging data showed amyloid-beta positivity in 74 participants. The cohort had a high baseline burden of cerebrovascular disease, regardless of amyloid-beta status (67 percent in the amyloid-beta–negative group, 66 percent in the amyloid-beta–positive group).
When predicting the risk of progression to dementia, participants in the high-risk group—categorized according to amyloid-beta–positivity via plasma brain-derived p-tau217—exhibited a faster, longitudinally-assessed cognitive decline compared with those in the low-risk group (hazard ratio, 5.62, 95 percent CI, 2.39–13.2; p<0.001).
Plasma brain-derived p-tau217 has potential clinical utility for both the diagnosis of amyloid pathology and prognosis of AD-associated cognitive impairments, according to the authors.
“Our results support rigorous validation of this biomarker in larger and more ethnically diverse cohorts with varying comorbidities and cerebrovascular disease profiles, particularly in light of the imminent adoption of blood biomarkers as part of the routine workup for AD-associated cognitive decline and dementia,” they said.