Integration of plasma eMTBR-tau243 and p-tau217 in the diagnosis and stratification of Alzheimer's disease: a prospective cohort study.

backgroundAlzheimer's disease pathology can be accurately identified with blood biomarkers (core 1 biomarkers), including plasma p-tau217 tests. Biomarkers that change longitudinally after diagnosis (core 2 biomarkers) can increase confidence that Alzheimer's disease pathology is contributing to cognitive symptoms. Plasma eMTBR-tau243 reflects tau tangle pathology and is a promising core 2 biomarker. We aimed to examine whether, among patients positive for plasma p-tau217, plasma eMTBR-tau243 improves classification of established Alzheimer's disease and predicts cognitive decline and tau tangle accumulation.

methodsIn this prospective cohort study, we included adults aged 40 years or older with subjective cognitive decline, mild cognitive impairment, or dementia, based on cognitive test results and clinical assessments, consecutively recruited at two memory clinics in Sweden, as part of the Swedish BioFINDER-2 study. Data were collected prospectively specifically for the purposes of this study. We validated our findings in a cohort of patients from the Knight Alzheimer Disease Research Center (ARDC), which included individuals determined to be cognitively unimpaired or to have cognitive impairment. Using mass spectrometry techniques, plasma p-tau217 was measured as a ratio to a non-phosphorylated peptide (%p-tau217), and eMTBR-tau243 was measured as a concentration. A priori derived thresholds for positivity were used. The primary outcome was established Alzheimer's disease based on both Alzheimer's disease pathology-specific CSF or PET biomarkers and clinical assessment, in accordance with the International Working Group definition. This study was registered with ClinicalTrials.gov, NCT03174938.

findingsBetween May 18, 2017, and Sept 8, 2023, we enrolled 572 patients with cognitive symptoms (142 with subjective cognitive decline, 259 with mild cognitive impairment, and 171 with dementia). 291 (51%) of 572 patients were female and 281 (49%) were male. All potentially eligible participants were included. Follow-up data collection was done until Sept 9, 2024. 350 (61%) of 572 patients had positive plasma %p-tau217. Among these, 341 (97%) were amyloid β (Aβ)-positive by CSF biomarkers or PET and 199 (57%) had established Alzheimer's disease (positive predictive value [PPV] 57%, 95% CI 51-61, for %p-tau217 alone). 194 (55%) of 350 %p-tau217-positive patients were also positive for eMTBR-tau243, with an accuracy of 81% (95% CI 76-84), PPV 84% (95% CI 78-88), negative predictive value (NPV) 77% (95% CI 68-82), and sensitivity 82% (95% CI 76-87) for established Alzheimer's disease, in cross-sectional analyses. Findings were similar in the validation cohort. Positive eMTBR-tau243 was associated with worse longitudinal cognitive decline and longitudinal tau tangle accumulation in %p-tau217 positive patients. Among %p-tau217 positive individuals, eMTBR-tau243 had an accuracy of 87%, PPV 76%, and NPV 90% in identifying individuals with high tau-PET load.

interpretationPlasma %p-tau217 can confirm Aβ-positivity, while eMTBR-tau243 can confirm clinically symptomatic Alzheimer's disease. Initial testing with %p-tau217 followed by eMTBR-tau243 in %p-tau217-positive patients could help assess whether Alzheimer's disease pathology is causing symptoms. Plasma eMTBR-tau243 might also help in determining the severity of tau tangle burden, which could be useful in guiding therapeutic decisions.

fundingNational Institute of Aging, European Research Council, Alzheimer's Association, GHR Foundation, Swedish Research Council, ERA PerMed, Knut and Alice Wallenberg foundation, Strategic Research Area MultiPark at Lund University, Swedish Alzheimer Foundation, Swedish Brain Foundation, Parkinson foundation of Sweden, Cure Alzheimer's fund, Rönström Family Foundation, Berg Family Foundation, Konung Gustaf V:s och Drottning Victorias Frimurarestiftelse, Skåne University Hospital Foundation, Michael J Fox Foundation, Lilly Research Award Program, Regionalt Forskningsstöd (Södra sjukvårdsregionen), Wallenberg AI; Autonomous Systems and Software Program and Data-Driven Life Science joint call for research projects, Greta and Johan Kock Foundation, and Swedish federal Government under the ALF agreement.