Robustness of Plasma P-Tau217 Diagnostic Thresholds for Alzheimer Disease Across Various Clinical Populations and Laboratory Environments
In brief
Plasma p-tau217 thresholds showed over 92% sensitivity and specificity for Alzheimer disease
In two external memory-clinic cohorts, plasma p-tau217 thresholds identified clinical-biological Alzheimer disease with 93% to 96% sensitivity and 92% to 93% specificity. Results held across centers and laboratory conditions, but specificity was lower in people with inherited frontotemporal dementia, especially with ALS; second-line spinal-fluid testing resolved most gray-zone results. Broader, more diverse validation is still needed.
- Journal
- Neurology (Q1)
- Published
- 16 November 2026
- Study design
- Cross-sectional study
- Evidence level
- Level 3, Low (CEBM 3b)
- Authors
- Jean-Louis Bayart, Nicolas Villain, Vincent Planche, Emilien Boyer, Lise Colmant, Isabelle Le Ber, et al.
- PMID
- 43038487
- DOI
- 10.1212/WNL.0000000000218784
Why clinicians should know about it
- Picked for Biochemistry (medical) (paper of the day, 18 November 2026): Multicenter study tests p-tau217 thresholds across laboratories
Abstract
BACKGROUND AND OBJECTIVES: Blood-based biomarkers (BBMs), especially plasma tau phosphorylated at threonine 217 (p-tau217), could scale Alzheimer disease (AD) clinical-biological diagnosis, but clinical use requires thresholds transferable across laboratories and populations. We compared Lumipulse p-tau217 with other BBMs; evaluated amyloid-anchored thresholds across centers, analyzers, reagent lots, and clinical populations; and assessed second-line CSF adjudication of gray zone results. METHODS: This multicenter, cross-sectional diagnostic-classification study included an exploratory Brussels cohort and 3 external validation cohorts. Brussels participants were cognitively unimpaired or impaired and had Aβ status determined by amyloid-PET or CSF Aβ42 and tau status by tau-PET or CSF p-tau181. Validation cohorts comprised consecutive Paris memory-clinic patients, patients with early-onset neurodegenerative dementia in Bordeaux, and individuals with monogenic frontotemporal lobar degeneration (FTLD). Plasma p-tau217 was measured locally on Lumipulse analyzers. BBMs and ratios were compared using receiver operating characteristic (ROC) analyses. In Brussels, a Youden cutoff and dual 95% or 97.5% sensitivity/specificity cutoffs were derived; intermediate values defined gray zones. RESULTS: Among 411 participants (mean age 70.5 years; 51.6% female), p-tau217 showed the best diagnostic performance; p-tau217/Aβ42 did not improve performance or reduce gray zones. In Brussels, p-tau217 discriminated Aβ+ from Aβ- status (AUC 0.96, 95% CI 0.93-0.98; p < 0.0001). The 0.193 pg/mL Youden cutoff (95% CI 0.178-0.213) yielded 91.5% sensitivity (95% CI 84.7-95.5) and 91.6% specificity (95% CI 84.8-95.5) and detected neocortical tau-PET positivity with 98.5% sensitivity and 81.8% specificity. In Paris, sensitivity/specificity for clinical-biological AD were 93.2% (95% CI 85.1-97.1)/93.0% (95% CI 81.4-97.6); in Bordeaux, sensitivity/specificity for clinical-biological AD were 95.7% (95% CI 79.0-99.2)/92.3% (95% CI 66.7-98.6). Specificity in monogenic FTLD was 88.4% (95% CI 75.5-94.9). False-positive classifications were enriched in FTLD with amyotrophic lateral sclerosis (ALS). Dual thresholds yielded gray zones in 15.4%-24.8% of Paris patients; second-line CSF testing resolved 86.2%-88.9% of gray zone cases. DISCUSSION: Lumipulse plasma p-tau217 thresholds were transferable across centers, analyzers, reagent lots, and heterogeneous clinical populations, supporting evaluation of plasma-first pathways with confirmatory testing for gray zone values. Values below 0.193 pg/mL made neocortical tauopathy unlikely. Limitations include cross-sectional design, specialized Western European recruitment, unavailable race/ethnicity data, heterogeneous reference standards, incomplete AD copathology adjudication in monogenic FTLD, and no cross-platform comparison.
Abstract as published, via PubMed.
For healthcare professionals. The summary is generated by AI from the published abstract, and the evidence level is assigned automatically from the study design on the Oxford CEBM hierarchy. Neither is medical advice. Read the full paper before changing practice.