Blood phosphorylated tau for the diagnosis of Alzheimer's disease: a systematic review and meta-analysis
Bibliographic record
Abstract
BACKGROUND: Plasma phosphorylated tau (p-tau) biomarkers show promise to transform the clinical management of Alzheimer's disease by providing more accessible and cost-effective diagnostic tools. p-tau biomarkers have emerged as leading contenders for clinical implementation; however, there have been no comprehensive meta-analyses of their diagnostic performance. We aimed to evaluate the diagnostic performance of plasma p-tau biomarkers and individual p-tau assays to identify biologically defined Alzheimer's disease. METHODS: For this systematic review and meta-analysis, we searched Embase, MEDLINE, PubMed, Scopus, and Web of Science for articles published from July 1, 1984 up to Dec 9, 2024, that reported on the discriminative accuracy of plasma p-tau biomarkers for amyloid-PET, tau-PET, CSF, and neuropathological reference standards. We included cohort, case-control, cross-sectional, and randomised controlled studies that recruited adults from any setting. Articles were excluded if they did not contain data on a p-tau blood biomarker, did not contain an appropriate biological reference standard, did not report diagnostic accuracy data, included participants younger than 18 years, or reported duplicate or overlapping data from another publication. Summary data were independently extracted by eight authors. Risk of bias was assessed using QUADAS-2. The primary outcome was the diagnostic performance of plasma p-tau biomarkers for Alzheimer's disease. We used a bivariate random-effects meta-analysis to estimate pooled sensitivity, specificity, diagnostics odds ratio and area under the receiver operating characteristic curve. We assessed the certainty of evidence using GRADE. This study was done following PRISMA-DTA guidelines and is registered with PROSPERO as CRD42023422143. FINDINGS: Of the 6429 studies identified by our search, 312 studies were assessed for eligibility, with 113 studies included in the final analysis, comprising 29 625 unique individuals. Plasma p-tau217 was the highest-performing biomarker for identifying biologically defined Alzheimer's disease, with pooled sensitivity of 88·1% (95% CI 86·7-89·5, moderate certainty of evidence), specificity of 88·7% (87·4-89·9, moderate certainty of evidence), area under the receiver operating characteristic curve (AUROC) of 91·1% (88·9-92·4, moderate certainty of evidence), and diagnostic odds ratio of 50·7 (40·6-63·4). p-tau181 pooled sensitivity was 80·5% (78·4-82·4, low certainty of evidence), specificity was 76·4% (74·1-78·6, low certainty of evidence), AUROC was 81·5% (80·2-82·9, low certainty of evidence), and diagnostic odds ratio was 13·4 (11·4-16·7). p-tau205 pooled sensitivity was 76·6% (70·7-81·6, moderate certainty of evidence), specificity was 86·0% (78·6-91·2, moderate certainty of evidence), AUROC was 85·1% (80·7-89·6, moderate certainty of evidence), and diagnostic odds ratio was 20·2 (10·5-38·7). p-tau212 pooled sensitivity was 84·5% (75·5-90·6, moderate certainty of evidence), specificity was 87·3% (79·5-92·5, moderate certainty of evidence), AUROC was 90·3% (86·6-94·1, moderate certainty of evidence), and diagnostic odds ratio was 41·2 (22·0-77·3). p-tau231 pooled sensitivity was 75·2% (71·3-78·8, moderate certainty of evidence), specificity was 75·3% (71·2-78·9, moderate certainty of evidence), AUROC was 80·2 (77·6-82·7, moderate certainty of evidence), and diagnostic odds ratio was 9·3 (7·0-12·2). Approximately 90% of studies were rated as high risk of bias for not having used predefined or externally derived thresholds. INTERPRETATION: Plasma p-tau217 is a highly sensitive and specific biomarker for Alzheimer's disease pathology, despite the high risk of bias of many studies. Prospective clinical implementation studies in real-world settings are needed to characterise the effect of plasma p-tau217 on Alzheimer's disease diagnosis and clinical management. FUNDING: McGill Faculty of Medicine Fellowship.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.008 | 0.002 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".