Abstract
INTRODUCTION
Cross‐cohort validation studies for plasma phosphorylated tau (p‐tau) 217 are limited. We evaluated the Janssen plasma p‐tau217+ assay and proposed a cutpoint value in three independent community‐based cohorts.
METHODS
We included n = 441 participants (age = 70.3 ± 7.3) from three independent community‐based cohorts with amyloid‐beta–positron emission tomography (Aβ‐PET), tau‐PET, clinical, and cognitive information.
RESULTS
The cohorts had low pre‐test probability (%Aβ positivity = 14.9–24.7) and were predominantly cognitively normal (> 73%). Plasma p‐tau217+ had high accuracy for abnormal Aβ PET (areas under the curve [AUCs] = 81‐86%), good correlation with Aβ‐PET burden (0.336‐0.397) that was highest in the cohort with the most Aβ‐PET‐positive participants, and the biomarker concentrations were highest in the joint Aβ‐PET and tau‐PET positive group. Negative predictive value (NPV) was high across cohorts (≤93%) but positive predictive value (PPV) was consistently poor (< 57%). Sensitivity and specificity averaged 75% and 84%, respectively. A combined cohort cutpoint of 0.05pg/ml gave AUC = 84.5%, NPV = 94%, PPV = 50%, sensitivity = 75%, and specificity = 84%.
DISCUSSION
Plasma p‐tau217+ can rule out Aβ pathophysiology due to Alzheimer's disease at the population level. Cohort‐level %Aβ‐PET positivity influences accuracies.
Highlights
Plasma phosphorylated tau (p‐tau) 217 assays have demonstrated potential to identify older adults with brain amyloid pathology due to Alzheimer's disease.
Community‐based studies are crucial for accessing the clinical validity of fluid biomarkers and their real‐world applicability.
We evaluated the Janssen plasma p‐tau217+ assay in three community‐based cohorts.
Janssen plasma p‐tau217+ identified amyloid‐beta abnormalities across three diverse cohorts of community‐dwelling older adults.
High NPV indicates p‐tau217+ as a good tool for initial screening to enrich for high‐risk individuals particularly for cohort studies and clinical trial participation.
Keywords: amyloid pathology, blood‐based biomarker, community‐based cohort, preclinical Alzheimer's disease, p‐tau217, tau pathology
1. INTRODUCTION
The use of cerebrospinal fluid (CSF) and imaging techniques like positron emission tomography (PET) for the antemortem diagnosis of Alzheimer's disease (AD), has led to significant advancements in the field. 1 , 2 , 3 This has enabled the detection of AD in cognitively unimpaired individuals years before symptom onset, and has provided a foundation for interventions, particularly anti‐amyloid therapies, which are known to reduce amyloid‐beta (Aβ) plaques from the brain. 2 , 3 , 4 , 5 These techniques have brought hope in the fight against AD. However, there are limitations to the use of such approaches, including invasiveness, cost, and the need for specialized expertise. 2 , 3 , 6 As a result, researchers have developed blood‐based biomarkers as potential alternatives. These biomarkers show promise for diagnosing and predicting AD and are currently being considered for clinical use. For instance, in May 2025, the United States Food and Drug Administration (FDA) cleared the Lumipulse G pTau217/β‐Amyloid 1‐42 Plasma Ratio for clinical use, marking it the first blood‐based biomarker for the early detection of amyloid plaques.
Several blood‐based biomarkers are gaining popularity in AD research, such as glial fibrillary acidic protein (GFAP), ratio of Amyloid beta 42 to Amyloid beta 40 (Aβ42/Aβ40 ratio), neurofilament light chain (NfL), phosphorylated tau (p‐tau) 181, p‐tau217, and p‐tau231. 1 , 7 , 8 , 9 These biomarkers are being extensively studied for their potential applications in memory clinics, primary care settings, and in clinical trials. 7 , 9 Among the tau biomarkers, p‐tau217 has shown the most consistent and most promising results. The plasma p‐tau217 assay has been found to perform similarly to the CSF p‐tau217 assay in discriminating between AD and other neurodegenerative disorders, 3 , 10 , 11 making it a promising but less invasive alternative. This has led to its widespread acceptance, resulting in different diagnostic companies using various techniques, including mass spectrometry and immunoassay to develop assays for measuring p‐tau217 in blood. Most plasma p‐tau217 assays utilize immunoassays. 12 , 13 The Janssen p‐tau217+ assay, developed by Janssen Research and Development, uses the Single Molecule Array immunoassay (SIMOA) technology to quantify minute amounts of p‐tau217 in human blood samples. Unlike other assays, this method specifically uses the PT3 antibody which has a unique binding affinity for tau protein phosphorylated at Thr217. The binding is further enhanced when there is simultaneous phosphorylation at the neighboring sites Thr212 and Ser214, hence the name p‐tau217+. 14 , 15 The Janssen p‐tau217+ assay is currently being marketed as a Laboratory Developed Test by the Quanterix Corporation subsidiary diagnostic company Lucent Diagnostics. Since this assay is being used clinically, it is necessary to widely validate it in community‐dwelling older adults who are less likely to join classical research cohorts but may form part of patients to be evaluated in primary or secondary care with this assay.
While there have been studies evaluating the technical performance and utility of the Janssen p‐tau217+ assay to monitor and detect the pathology of AD, especially in Aβ positive participants, and comparing it to other commercially available plasma p‐tau217 assays, 7 , 15 , 16 there is a lack of studies that assess the p‐tau217+ assay in community‐based cohorts with particular interest in the preclinical AD phase. Therefore, this study aimed to investigate the association of the Janssen plasma p‐tau217+ assay with Aβ PET and tau PET and its classification accuracies across three largely cognitively normal community cohorts. Additionally, we took advantage of the multicohort approach to generate and cross‐validate preliminary cutpoints for plasma p‐tau217+.
2. METHODS
2.1. Cohort studies and blood sampling
In this study, three community‐based independent cohort studies representing diverse groups of participants were used. The methods have been described extensively previously. 17 , 18 , 19 , 20 , 21
RESEARCH IN CONTEXT
Systematic review: We performed a systematic literature review using PubMed and Google Scholar with a focus on articles that have used the Janssen plasma phosphorylated tau (p‐tau) 217+ assay. Previous studies have focused more on cohorts based on volunteers recruited from urban and academic medical centers in the study of p‐tau217 clinical utility. Only a few previous studies have investigated the diagnostic potential of the assay using community‐based cohort like we employed in this study.
Interpretation: Our study showed Janssen plasma p‐tau217+ assay as a reliable indicators of brain amyloid‐beta (Aβ) pathophysiology in community‐based cohorts including mostly cognitively unimpaired individuals. The consistently high negative predictive values (NPVs) highlight the assay utility for community screening to rule out Aβ pathophysiology due to Alzheimer's disease (AD).
Future directions: Longitudinal studies are needed to evaluate assay utility as a monitoring biomarker tool. Moreover, further studies are needed to explore the potential of using the assay to examine the effects of comorbidity on AD detection.
2.1.1. Cohort 1 (MYHAT‐NI)
The Monongahela Youghiogheny Healthy Aging Team‐Neuroimaging (MYHAT‐NI) is a sub‐cohort of the parent MYHAT study, which is a community‐based sample of older adults without dementia, designed to investigate mild cognitive impairment in older adults in an economically depressed Rust Belt region in southwestern Pennsylvania. 17 , 18 The parent MYHAT study recruited participants aged 65 and above, utilizing age‐stratified random sampling from voter registration lists publicly available. The MYHAT‐NI study recruited a subgroup of MYHAT participants who had a Clinical Dementia Rating (CDR) Sum of Boxes score of < 1.0 for neuroimaging assessments to study the distribution and functional correlations of AD pathologies. For this reason, all MYHAT‐NI participants had normal or only very mildly impaired cognition at the time of enrollment. The only exclusion criterion was a contraindication to neuroimaging. Sociodemographic information and cognitive test data were collected at the parent study visit. Blood samples were collected during the baseline neuroimaging visit, which included [11C] Pittsburgh Compound B (PiB) PET imaging of Aβ plaques and [18F]AV‐1451 PET imaging of tau neurofibrillary tangles. Detailed study designs for MYHAT‐NI, including subject recruitment strategies, multi‐domain cognitive assessments, neuroimaging, apolipoprotein E (APOE) ε4 genotyping and data processing have been published by our group previously. 17 , 18 , 19
2.1.2. Cohort 2 (CoBrA)
The Connectomics in Brain Aging and Dementia (CoBrA) study, under the auspices of the Human Connectome Project (HCP) is an ongoing community‐based longitudinal study that aimed to recruit an equal number of self‐identified Black/African American and non‐Hispanic White participants aged 50‐89 years old. 20 The study aims to accelerate and address questions related to the structure and function of AD and other neurovascular diseases leveraging the comparative advantage of different imaging techniques, including magnetic resonance imaging (MRI), magnetoencephalography, and PET imaging. The participants underwent neuroimaging assessment with [11C]PiB PET imaging of Aβ plaques. Demographic information was collected and neuropsychological assessment completed. Neuropsychological test battery assessments used include the Montreal Cognitive Assessment (MoCA), verbal fluency, a 30‐item visual naming test, Trail making verbal free recall and the Rey‐Osterreith Complex Figure to measure their motor functions, cognitive status, emotions and personality traits. Blood samples were collected and processed for biomarker analysis, DNA analysis and APOE ε4 genotyping. Further information about this cohort can be found in a previous publication. 20
2.1.3. Cohort 3 (HeartSCORE)
The Heart Strategies Concentrating on Risk Evaluation (HeartSCORE) study began in 2003 as a community‐based cohort study of 2000 participants recruited in Allegheny County, Pennsylvania. Participants were 45 to 75 years old at baseline. The exclusion criteria include comorbid conditions that can limit life expectancy to < 5 years and inability to undergo annual follow‐up visits. Demographic and medical histories were collected including information on self‐reported race, marital/co‐habiting status, education, income, smoking, physical activity, and dietary habits. A subset of participants from the HeartSCORE study were recruited for neuroimaging and neurocognitive testing. All participants underwent a Mini‐Mental State Examination. Detailed information about the study design for HeartSCORE, including data collection and processing, study recruitment, retention strategies, cognitive assessments, and neuroimaging can be found in previous publications. 21
2.2. Neuroimaging
[11C]PiB PET and [18F]AV‐1451 PET were used to assess brain Aβ and tau pathology respectively. The A status was based on a global [11C]PiB standardized uptake value ratio (SUVR) computed by volume‐weighted averaging of nine composite regional outcomes (anterior cingulate, posterior cingulate, insula, superior frontal, cortex, orbitofrontal cortex, lateral temporal cortex, parietal, precuneus, and ventral striatum). Participants were classified as A+ or A‐, with A+ reported as SUVR > 1.346 and A‐ reported as SUVR < 1.346. For T status, we used a MetaTemporal composite region 22 . Participants were classified as T+ and T‐, with SUVR > 1.18 considered as T+ while SUVR ≤1.18 considered as T‐. For the neuroimaging analysis, this study utilized the same tracers, scanners, and protocol throughout the study to ensure consistent results.
2.3. Plasma biomarker assay
The plasma p‐tau217+ assay measurements were conducted at Quanterix (Billerica, MA) using the Janssen assay (aka LucentAD p217). 7 The plasma p‐tau217+ measurements were performed as in‐kind research collaboration with the University of Pittsburgh (Principal Investigator, T.K. Karikari) under material transfer agreement. However, the scientists conducting the measurements were blinded to the sample information until the measurements were completed. Plasma p‐tau217+ measurements were performed on a Quanterix HD‐X instrument, utilizing the monoclonal anti‐human p‐tau217 antibody PT3, the binding affinity of which towards p‐tau217 is enhanced by additional phosphorylation at aa212, as the capture antibody. Anti‐human tau antibody HT43 was used as the detector. 7 Before starting the assay measurement, the plasma samples were thawed at room temperature and then centrifuged at 4000xg for 10 minutes to remove any particulate matter. The intra‐assay and inter‐assay %coefficients of variation (CVs) for the assay were 3.9% and 7.3% respectively.
2.4. Statistical Analysis
Plasma p‐tau217+ data were independently unblinded and matched with the clinical, neuropsychological and neuroimaging datasets and statistical analysis performed at the University of Pittsburgh without any involvement of Janssen or Quanterix staff. All analyses were conducted using MATLAB (version R2021b) or R statistical software version 4.2.1 (R Foundation for Statistical Computing, Vienna, Austria; http://www.r‐project.org/). For demographic characteristics, we presented the continuous variables as medians and interquartile ranges (IQRs) and categorical variables as counts and percentages. Wilcoxon rank‐sum tests were used to compare continuous variables, and Fisher's exact tests were used for categorical variables between the A+ and A‐ groups. Additionally, Wilcoxon rank‐sum tests were employed for comparing p‐tau217+ levels dichotomized by A status, T status, and cognitive impairment. The Kruskal–Wallis test was employed to compare plasma p‐tau217+ level among four groups stratified by A and T statuses. Spearman's rank‐based correlation measured the relationship between plasma p‐tau217+ levels, and neuroimaging‐assessed AD pathology burden. The area under the curve (AUC) of the receiver operating characteristic (ROC) curve, with and without covariates (age, sex, and APOE ε4 carrier status), was determined using linear regression, with log2‐transformed p‐tau217+ levels as the dependent variable. The predictive performance, including sensitivity, specificity, negative predictive value (NPV), and positive predictive value (PPV), was based on the Youden index. Linear regression models were also used to evaluate the significance of the association between p‐tau217+ levels and demographic and genotypic risk factors.
3. RESULTS
3.1. Cohort characteristics
We included a total of 441 participants from the MYHAT‐NI (n = 93), CoBrA (n = 201) and HeartSCORE (n = 147) cohorts.
The MYHAT‐NI cohort had a median (IQR) age of 76 (72‐81) years, 48 (51.6%) females, 89 (95.7%) non‐Hispanic Whites (NHW), and 13 (14%) APOE ε4 carriers. Among them, 23 (24.7%) participants were A+, 33 (35.5%) were T+, 29 (31.2%) were N+, and 85 (91.4%) had no dementia (CDR global score of 0) (Table 1a). The A+ participants were older (p = 0.025), had higher CDR global scores (i.e., more cognitive impairment; p = 0.021), and a higher prevalence of APOE ε4 carriers (p < 0.001) than the A‐ group. They also had a higher likelihood of being T+ (p < 0.001). There were no significant differences in sex, education, race, Mini‐Mental State Examination (MMSE) scores, or N status (whether analyzed as a dichotomous or continuous variable) between the A+ and A‐ groups.
TABLE 1.
Participant characteristics in the MYHAT‐NI, CoBrA, and HeartSCORE cohorts according to Aβ PET positivity.
| a. Cohort: MYHAT‐NI | ||||
|---|---|---|---|---|
| MYHAT‐NI cohort | ||||
| Total | Aβ PET‐ | Aβ PET+ | p‐value | |
| N (%) | 93 | 70 (75.3%) | 23 (24.7%) | |
| Age (years) | 76.0 (72.0‐81.0) | 76.0 (71.0‐80.0) | 80.0 (76.0‐83.8) | 0.025 |
| Sex | 0.344 | |||
| Female | 48 (51.6%) | 34 (48.6%) | 14 (60.9%) | |
| Male | 45 (48.4%) | 36 (51.4%) | 9 (39.1%) | |
| Education (years) | 12 (12‐16) | 12 (12‐16) | 13 (12‐16) | 0.970 |
| Race | 1.000 | |||
| NHW | 89 (95.7%) | 67 (95.7%) | 22 (95.7%) | |
| B/AA | 4 (4.3%) | 3 (4.3%) | 1 (4.3%) | |
| MMSE | 0.434 | |||
| > = 24 | 91 (97.85%) | 69 (98.6%) | 22 (95.7%) | |
| 19‐23 | 2 (2.15%) | 1 (1.4%) | 1 (4.3%) | |
| CDR global score | 0.021 | |||
| CDR = 0 | 85 (91.4%) | 67 (95.7%) | 18 (78.3%) | |
| CDR = 0.5 | 8 (8.6%) | 3 (4.3%) | 5 (21.7%) | |
| APOE ε4 carrier | <0.001 | |||
| Yes | 13 (14.0%) | 3 (4.3%) | 10 (43.5%) | |
| No | 80 (86.0%) | 67 (95.7%) | 13 (56.5%) | |
| PiB SUVR | 1.14 (1.01‐1.33) | 1.10 (1.07‐1.15) | 1.70 (1.48‐1.89) | <0.001 |
| Tau PET | <0.001 | |||
| Negative | 60 (64.5) | 52 (74.3%) | 8 (34.8%) | |
| Positive | 33 (35.5) | 18 (25.7%) | 15 (65.2%) | |
| Tau PET SUVR | 1.16 (1.11‐1.21) | 1.14 (1.09‐1.18) | 1.21 (1.17‐1.26) | <0.001 |
| b. Cohort: CoBrA | ||||
|---|---|---|---|---|
| CoBrA COHORT | ||||
| Total | Aβ PET‐ | Aβ PET+ | p‐value | |
| N (%) | 201 | 171 (85.1%) | 30 (14.9%) | |
| Age (years) | 62.0 (57.0‐71.0) | 60.0 (56.0‐68.0) | 71.5 (68.0‐78.0) | <0.001 |
| Sex | 0.004 | |||
| Female | 129 (64.2%) | 117 (68.4%) | 12 (40.0%) | |
| Male | 72 (35.8%) | 54 (31.6%) | 18 (60.0%) | |
| Education (years) | 14 (12‐17) | 14 (12‐16) | 16 (14‐18) | <0.001 |
| Race | <0.001 | |||
| Asian | 3 (1.5%) | 3 (1.8%) | 0 (0%) | |
| B/AA | 105 (52.2%) | 99 (57.9%) | 6 (20.0%) | |
| NHW | 93 (46.3%) | 69 (40.4%) | 24 (80.0%) | |
| MoCA | 25 (23‐27) | 25 (23‐27) | 26 (23‐27) | 0.427 |
| APOE ε4 carrier | 0.131 | |||
| Yes | 61 (30.4%) | 48 (28.1%) | 13 (43.3%) | |
| No | 139 (69.2%) | 122 (71.4%) | 17 (56.7%) | |
| Missing | 1 (0.5%) | 1 (0.6%) | 0 (0%) | |
| PiB SUVR | 1.10 (1.05‐1.17) | 1.09 (1.05‐1.14) | 1.66 (1.51‐1.91) | <0.001 |
| N Status | 0.216 | |||
| Negative | 129 (64.2%) | 113 (66.1%) | 16 (53.3%) | |
| Positive | 72 (35.8%) | 58 (33.9%) | 14 (46.7%) | |
| Cortical thickness | 2.74 (2.66‐2.79) | 2.74 (2.67‐2.80) | 2.71 (2.58‐2.76) | 0.018 |
| c. Cohort: HeartSCORE | ||||
|---|---|---|---|---|
| HeartSCORE COHORT | ||||
| Total | Aβ PET‐ | Aβ PET+ | p‐value | |
| N (%) | 147 | 121 (82.3%) | 26 (17.7%) | |
| Age (years) | 73.0 (70.0‐77.0) | 73.0 (70.0‐76.0) | 75.0 (70.3‐79.0) | 0.206 |
| Sex | 0.497 | |||
| Female | 99 (67.4%) | 83 (68.6%) | 16 (61.5%) | |
| Male | 48 (32.7%) | 38 (31.4%) | 10 (38.5%) | |
| Education (years) | 16.0 (13.0‐18.0) | 16.0 (14.0‐18.0) | 16 (12.0‐18.0) | 0.642 |
| Race | 0.340 | |||
| NHW | 105 (71.4%) | 84 (69.4%) | 21 (80.8%) | |
| B/AA | 42 (28.6%) | 37 (30.6%) | 5 (19.2%) | |
| MMSE | 0.585 | |||
| > = 24 | 134 (91.1%) | 109 (90.1%) | 25 (100.0%) | |
| 19‐23 | 5 (3.4%) | 5 (4.1%) | 0 (0%) | |
| Missing | 4 (2.7%) | 4 (3.3%) | 0 (0%) | |
| CDR global score | 0.071 | |||
| CDR = 0 | 108 (73.5%) | 93 (76.9%) | 15 (57.7%) | |
| CDR = 0.5 | 23 (15.7%) | 17 (14.1%) | 6 (23.1%) | |
| CDR = 1.0 | 1 (0.7%) | 1 (0.8%) | 0 (0%) | |
| Missing | 15 (10.2%) | 10 (8.3%) | 5 (19.2%) | |
| APOE ε4 carrier | 0.002 | |||
| Yes | 40 (27.2%) | 26 (21.5%) | 14 (53.8%) | |
| No | 107 (72.3%) | 95 (78.5%) | 12 (46.2%) | |
| PiB SUVR | 1.17 (1.13‐1.24) | 1.15 (1.12‐1.18) | 1.60 (1.51‐1.78) | <0.001 |
| Tau PET | <0.001 | |||
| Negative | 97 (66.0%) | 89 (73.6%) | 8 (30.8%) | |
| Positive | 49 (33.3%) | 32 (26.4%) | 17 (65.4%) | |
| Missing | 1 (0.7%) | 0 (0%) | 1 (3.9%) | |
| Tau PET SUVR | 1.15 (1.11‐1.21) | 1.4 (1.10‐1.19) | 1.23 (1.16‐1.27) | <0.001 |
Note: Numerical variables are reported as median (IQR), and categorical variables as count (percentage). All p‐values were calculated using the Wilcoxon rank sum test for numerical variables and Fisher's exact test for categorical variables, with missing values excluded. Tau pathology (T) was assessed using the PET tracer [18F]AV‐1451, with a SUVR cutoff of 1.18 to define tau positivity (≥1.18 as positive).
Abbreviations: Aβ, amyloid beta; APOE, apolipoprotein E; B/AA, Black/African American; CDR, Clinical Dementia Rating; CoBrA, Connectomics in Brain Aging and Dementia; IQR, interquartile range; MYHAT‐NI, Monongahela Youghiogheny Healthy Aging Team‐Neuroimaging; MMSE, Mini‐Mental State Examination; MoCA, Montreal Cognitive Assessment; NHW, non‐Hispanic White; PET, positron emission tomography; PiB, [11C] Pittsburgh Compound B; SUVR, standardized uptake value ratio;
The CoBrA cohort had a median (IQR) age of 62 (57‐71) years, comprising 129 (64.2%) female participants, 61 (30.4%) APOE ε4 allele carriers, 30 (14.9%) A+ participants, and 72 (35.8%) N+ participants (Table 1b). Among them, 105 (52.2%) identified themselves as Black/African American (B/AA), 3 (1.5%) as Asian, and the rest as non‐Hispanic White (NHW). The A+ group was older (p < 0.001), had fewer female participants (p = 0.004), and had a higher likelihood of being NHW (p < 0.001) than the A‐ group, but the two groups had no significant difference in terms of APOE ε4 carriership or both dichotomized and numerical N status (Table 1b).
The HeartSCORE cohort had a median (IQR) age of 73 (70‐77) years, 99 (67.4%) were females, 40 (27.2%) were APOE ε4 carriers. Additionally, 26 (17.7%) were A+, 49 (33.3%) were T+, 39 (26.5%) were N+, and 108 (73.5%) were dementia‐free, with a CDR global score of 0 (Table 1c). Among them, 105 (71.4%) were self‐identified as NHW and 42 (28.6%) were B/AA. There was no significant difference in terms of age, sex, years of education, or race between the A+ and A‐ groups, but there was a difference in APOE ε4 carrier status (p = 0.002), with a higher proportion in the A+ group. Similar to the MYHAT‐NI cohort, A+ participants had a higher likelihood of being T+ but not N+.
We also classified the participants based on their tau pathology, and the results are presented in Supplementary Table S1. Among all variables examined, the T+ and T‐ participants showed significant differences only in their Aβ PET status (both dichotomous and continuous) and Tau PET SUVR.
Significant differences in plasma p‐tau217+ levels were observed across the three cohorts, with a p value of < 0.001 in the Kruskal–Wallis test (Figure S1). Post hoc analysis revealed that the MYHAT‐NI cohort differed most notably from both the CoBrA and HeartSCORE cohorts. MYHAT‐NI exhibited higher overall p‐tau217+ levels, with a median (IQR) of 0.052 (0.037–0.074) pg/ml, compared with 0.035 (0.026–0.050) pg/ml in CoBrA and 0.038 (0.027–0.055) pg/ml in HeartSCORE. No significant difference was found between the CoBrA and HeartSCORE cohorts. The significant difference across cohorts remained after adjusting for common demographic/genotypic factors, including age, sex, race, education, and APOE ε4 carrier status, with an adjusted N‐way ANOVA p‐value of 0.012.
3.2. Association with amyloid pathology
3.2.1. Association with Aβ PET positivity: Individual cohorts
We first evaluated the association of plasma p‐tau217+ with dichotomized Aβ PET status, separately for each cohort (Figure 1). Significant differences in p‐tau217+ levels were observed when comparing A+ to A‐ groups across all three cohorts, with Wilcoxon rank‐sum p‐values < 0.001 in all three cohorts (Figure 1A). Compared to the A‐ participants, p‐tau217+ levels were 1.80‐fold, 2.11‐fold, and 1.81‐fold higher in the A+ participants for MYHAT‐NI, CoBrA, and HeartSCORE, respectively.
FIGURE 1.
Association of plasma p‐tau217+ with Aβ‐PET status. (A) Boxplot distribution of p‐tau217+ levels by Aβ‐PET status, with overlaid data points. The central box represents the IQR, the line inside the box indicates the median, and the whiskers extend to the most extreme data points within 1.5 times the IQR from the quartiles. All p‐values were determined using the Wilcoxon rank‐sum test, and fc represents the ratio of mean p‐tau217+ levels in the A+ group relative to the A− group. (B–D) Results from ROC analyses evaluating the ability of p‐tau217+ to discriminate Aβ‐PET positivity across the CoBrA, HeartSCORE, MYHAT‐NI, and combined cohorts. (B) Forest plot of AUC values for p‐tau217+ only models. (C) Forest plot of AUC values for models including p‐tau217+ along with covariates (age, sex, race, and APOE ε4 carrier status). (D) Table summarizing AUC values and their corresponding 95% CIs. Aβ, amyloid beta; APOE, apolipoprotein E; AUC, area under the curve; CI, confidence interval; CoBrA, Connectomics in Brain Aging and Dementia; fc, fold change; IQR, interquartile range; MYHAT‐NI, Monongahela Youghiogheny Healthy Aging Team‐Neuroimaging; PET, positron emission tomography; ROC, receiver operating characteristic curve.
We utilized ROC analysis to evaluate the predictive performance of p‐tau217 in distinguishing A+ participants from A‐ individuals. Strong predictive performance was observed in all three cohorts, with AUCs of 0.857 (95% confidence interval [CI]: 0.755 – 0.922), 0.850 (95% CI: 0.719 – 0.927), and 0.814 (95% CI: 0.688 – 0.892) for CoBrA, MYHAT‐NI, and HeartSCORE, respectively, based on the p‐tau217+ alone (Figure 1B). Inclusion of covariates, including sex, age, race, and APOE ε4 carrier status, improved the AUCs of CoBrA to 0.926 (95% CI: 0.879 – 0.961), MYHAT‐NI to 0.907 (95% CI: 0.804 – 0.961), and HeartSCORE to 0.865 (95% CI: 0.779 – 0.926) (Figure 1C).
3.2.2. Cutoff value estimation
Table 2 shows the classification performance based on cutoff values determined using the Youden index. When using p‐tau217+ alone, the optimal thresholds ranged from 0.049 to 0.064 pg/ml, yielding high negative predictive values (NPV: 0.929–0.966) but relatively modest positive predictive values (PPV: 0.463–0.559), with overall accuracy ranging from 0.796 to 0.837. Incorporating covariates generally improved classification performance, except in the HeartSCORE cohort, where accuracy declined from 0.837 to 0.653 after adjusting for covariates.
TABLE 2.
Classification performance of p‐tau217+ assay in predicting amyloid PET positivity.
| Cohort | Model | Sensitivity | Specificity | PPV | NPV | Accuracy | p‐tau217+ threshold | Youden Index |
|---|---|---|---|---|---|---|---|---|
| MYHAT‐NI | p‐tau217+ only | 0.826 | 0.786 | 0.559 | 0.932 | 0.796 | 0.064 | 0.612 |
| MYHAT‐NI | p‐tau217+ plus covariates | 0.783 | 0.943 | 0.818 | 0.930 | 0.903 | NA | 0.725 |
| HeartSCORE | p‐tau217+ only | 0.692 | 0.868 | 0.529 | 0.929 | 0.837 | 0.057 | 0.560 |
| HeartSCORE | p‐tau217+ plus covariates | 0.962 | 0.587 | 0.333 | 0.986 | 0.653 | NA | 0.548 |
| CoBrA | p‐tau217+ only | 0.833 | 0.830 | 0.463 | 0.966 | 0.831 | 0.049 | 0.664 |
| CoBrA | p‐tau217+ plus covariates | 0.900 | 0.830 | 0.482 | 0.979 | 0.841 | NA | 0.726 |
| Combined | p‐tau217+ only | 0.747 | 0.840 | 0.504 | 0.938 | 0.823 | 0.055 | 0.587 |
| Combined | p‐tau217+ plus covariates | 0.709 | 0.906 | 0.622 | 0.934 | 0.871 | NA | 0.614 |
Note: Generalized linear models with Aβ‐PET status as the outcome and log2‐transformed p‐tau217+ as the predictor, with or without the inclusion of covariates (age, sex, race, and APOE ε4 carrier status), were used for the classification analysis. The optimal cutoff for classification was determined using the Youden index, and corresponding NPV, PPV, sensitivity, and specificity were calculated at this threshold.
Abbreviations: Aβ, amyloid beta; APOE, apolipoprotein E; CoBrA, Connectomics in Brain Aging and Dementia;MYHAT‐NI, Monongahela Youghiogheny Healthy Aging Team‐Neuroimaging; NPV, negative predictive value; PET, positron emission tomography; PPV, positive predictive value.
3.2.3. Correlation with Aβ burden measured by PiB SUVR
As depicted in Figure 2, plasma p‐tau217+ levels showed a strong positive correlation with PiB SUVR across all three cohorts, with Spearman correlation coefficients (ρ) of 0.336 (p = 0.001), 0.384 (p < 0.001), and 0.397 (p < 0.001) for MYHAT‐NI, CoBrA, and HeartSCORE, respectively. Overall, a stronger correlation was observed among A+ participants, with ρ values of 0.426 (p = 0.044), 0.333 (p = 0.073), and 0.642 (p < 0.001) for MYHAT‐NI, CoBrA, and HeartSCORE, respectively. The corresponding ρ values among A‐ participants were ‐0.136 (p = 0.262), 0.162 (p = 0.035), and 0.164 (p = 0.072).
FIGURE 2.
Scatterplots depict the correlation between plasma p‐tau217+ levels and amyloid burden measured by PiB SUVR. Spearman correlation was used to assess associations. Blue dots represent A− individuals, and red dots represent A+ individuals. The red line indicates the linear regression across the entire cohort, with the gray shaded area representing the 95% confidence interval. Spearman's rho (ρ) values are shown for A‐, A+, and the entire cohort. Asterisks indicate significance levels: *, p < 0.05; **, p < 0.01; ***, p < 0.001. PiB, [11C] Pittsburgh Compound B; SUVR, standardized uptake value ratio.
3.2.4. Association in the combined cohort
When combining all three cohorts, the overall fold change in p‐tau217+ levels between A+ and A− groups was 1.97. The AUC for distinguishing A+ from A− individuals was 0.845 (95% CI: 0.791–0.890) using p‐tau217+ alone and improved to 0.887 (95% CI: 0.849–0.918) when covariates were included. Consistent with findings from individual cohorts, the optimal threshold determined by Youden index yielded a high NPV but less robust PPV. Specifically, the NPV and PPV were 0.938 and 0.504, respectively, using p‐tau217+ alone, and 0.934 and 0.622 when covariates were incorporated. The combined cohort gave a threshold of 0.05 pg/ml. Although p‐tau217+ levels were elevated in the MYHAT‐NI cohort compared to CoBrA and HeartSCORE, there was no significant cohort‐specific association with Aβ PET status, as indicated by the non‐significant interaction between cohort and A status on p‐tau217+ levels (data not shown).
3.3. Association with tau pathology
We further analyzed the correlation between plasma p‐tau217+ and tau PET status to determine its potential for detecting tau neurofibrillary tangle pathology (Figure 3). A similar fold change difference was observed when comparing T+ vs. T‐ individuals for MYHAT‐NI, HeartSCORE, and the combined cohorts, with values of 1.31 (p = 0.032), 1.34 (p = 0.007), and 1.33 (p < 0.001), respectively. ROC analysis revealed modest diagnostic performance, with AUC values of 0.635 (95% CI: 0.494‐0.746) for MYHAT‐NI and 0.638 (95% CI: 0.538‐0.722) for HeartSCORE, when using plasma p‐tau217+ only. Incorporating covariates led to slight improvements, with AUCs increasing to 0.666 (95% CI: 0.533‐0.767) and 0.654 (95% CI: 0.558‐0.749), respectively. When the two cohorts were combined, the AUC was 0.633 (95% CI: 0.571‐0.681) using plasma p‐tau217+ alone, and 0.646 (95% CI: 0.592‐0.697) after adjusting for covariates.
FIGURE 3.
Association of plasma p‐tau217+ with Tau PET status. (A) Boxplot distribution of p‐tau217+ levels by Tau PET status, with overlaid data points. All p‐values were determined using the Wilcoxon rank‐sum test, and fc represents the ratio of mean p‐tau217+ levels in the T+ group relative to the T− group. (B,C) present results from ROC analyses evaluating the ability of p‐tau217+ to discriminate Tau PET positivity across the HeartSCORE, MYHAT‐NI, and combined cohorts. (B) Forest plot of AUC values for p‐tau217+ only models. (C) Forest plot of AUC values for models including p‐tau217+ along with covariates (age, sex, race, and APOE ε4 carrier status). APOE, apolipoprotein E; AUC, area under the curve; fc, fold change; MYHAT‐NI, Monongahela Youghiogheny Healthy Aging Team‐Neuroimaging; PET, positron emission tomography; ROC, receiver operating characteristic.
We then evaluated the strength of the association between plasma p‐tau217+ levels and tau tangle load determined by tau PET SUVR (Figure 4). The p‐tau217+ levels showed a moderate correlation with Tau‐PET SUVR, with ρ values of 0.195 (p = 0.061), 0.336 (p < 0.001), and 0.265 (p < 0.001) for MYHAT‐NI, HeartSCORE, and combined cohort, respectively. Similar to the association with Aβ burden, a stronger correlation was observed among T+ participants, with ρ values of 0.354 (p = 0.044), 0.466 (p < 0.001), and 0.384 (p < 0.001) for MYHAT‐NI, HeartSCORE, and combined cohort, respectively. The corresponding ρ values among T‐ participants were ‐0.080 (p = 0.544), 0.227 (p = 0.025) and 0.095 (p = 0.239).
FIGURE 4.
Scatterplots depict the correlation between plasma p‐tau217+ levels and tau tangle load measured by tau PET SUVR. Spearman correlation was used to assess associations. Blue dots represent T‐ individuals, and red dots represent T+ individuals. The red line indicates the linear regression across the entire cohort, with the gray shaded area representing the 95% confidence interval. Spearman's rho (ρ) values are shown for T‐, T+, and the entire cohort. Asterisks indicate significance levels: *, p < 0.05; **, p < 0.01; ***, p < 0.001. PET, positron emission tomography; SUVR, standardized uptake value ratio.
3.4. Plasma p‐tau217+ levels by A/T classification
To gain a better understanding of how plasma p‐tau217+ varies across different stages of AD pathology, we compared its profiles across four groups stratified by A and T status. As shown in Figure 5, median p‐tau217+ levels increased stepwise across the A‐T‐, A‐T+, A+T‐, and A+T+ groups in both individual and combined cohorts. However, in the individual cohorts, significant differences were observed only between A+T+ and A‐T‐ or A‐T+ (all p < 0.001). In the combined cohort, the A+T‐ also differed significantly from the A‐T‐ (p = 0.005) and A‐T+ (p = 0.019) groups. No significant differences were found between T‐ and T+ groups when A status was held constant (i.e., A‐T‐ vs. A‐T+ or A+T‐ vs. A+T+), suggesting that A status primarily drives the observed changes in p‐tau217+ levels. Consistent with this, T status was not significantly associated with p‐tau217+ levels after adjusting for A status in either cohort (data not shown).
FIGURE 5.
Plasma p‐tau217+ profiles by A/T classification. All p‐values were calculated using post hoc analysis following the Kruskal–Wallis test, with Bonferroni correction for multiple comparisons. Asterisks indicate significance levels: *, p < 0.05; **, p < 0.01; ***, p < 0.001.
3.5. Association of plasma p‐tau217+ with cognitive impairment
We evaluated the association of p‐tau217+ with cognitive impairment using different assessment tools across three cohorts: the CDR global score in MYHAT‐NI (CDR 0 = normal; CDR 0.5 = impaired), the MoCA in CoBrA (≥26 = normal; < 26 = impaired), and the MMSE in HeartSCORE (≥24 = normal; < 24 = impaired). As shown in Figure S2, among the three cohorts, only the MYHAT‐NI cohort demonstrated a marginally significant elevation in p‐tau217+ levels in the impaired group, with a fold change of 1.38 (p = 0.086). In contrast, both CoBrA and HeartSCORE showed lower p‐tau217+ levels in the impaired groups, with fold changes of 0.581 (p = 0.996) and 0.838 (p = 0.527), respectively; however, these differences were not statistically significant.
3.6. Associations of plasma p‐tau217+ levels with demographic and genetic risk factors
Age, sex, race, education, and APOE ε4 genotype are commonly recognized as important confounding factors in AD research. Table S2 illustrates the relationship between p‐tau217+ levels and these factors in our cohorts. In univariate analyses, we observed a significant positive correlation between age and p‐tau217+ levels in both the MYHAT‐NI and CoBrA cohorts, with marginal significance in the HeartSCORE cohort. APOE ε4 carrier status was associated with higher p‐tau217+ levels in MYHAT‐NI and HeartSCORE (p = 0.028 and 0.014, respectively), but not in CoBrA (p = 0.516). In the CoBrA cohort, race, sex, and education also showed significant univariate associations with p‐tau217+ levels. However, in multivariable models that included all these factors, only age remained significantly associated with p‐tau217+ levels (p < 0.001), while sex showed marginal significance (p = 0.051).
4. DISCUSSION
The transformative utility of blood‐based biomarkers in the field of AD and related dementias continues to become increasingly realistic due to the introduction of various tests that have high accuracies, facilitating enhanced patient choices and diagnostic market competition for predicting ATN status. Among these, the p‐tau tests—specifically p‐tau217, p‐tau231, and p‐tau181—have garnered the most attention. Notably, p‐tau217 has demonstrated the highest reliability and potential for diagnosing AD. Companies such as Eli Lilly, SpearBio, C2N, Quanterix, Alamar, Roche, Janssen, Beckman Coulter, and Fujirebio have developed p‐tau217 assay platforms that show promising results for clinical applications. The Janssen plasma p‐tau217+ assay used in this study has yielded results comparable to the Quanterix ALZpath p‐tau217 in a translational study cohort, showing utility for monitoring and detecting AD pathology. 15 In another study comparing multiple plasma p‐tau217 assays, the Janssen p‐tau217+ assay demonstrated strong capability to identify abnormal Aβ and tau PET status. 13 Janssen has also transferred non‐exclusive commercialization rights for its p‐tau217+ assay to Quanterix, the same company which manages the ALZpath assay. This assay is now commercially available as a Lab Developed Test through Quanterix as the LucentAD p217 assay. 23
This study evaluated the Janssen plasma p‐tau217+ assay in mostly asymptomatic individuals using three independent community‐based cohorts. We examined the assay's diagnostic capability for detecting AD and its association with both Aβ and tau PET. Plasma p‐tau217+ differentiated Aβ PET‐negative from Aβ PET‐positive participants, as well as when classified according to combined Aβ and tau PET status. When the amyloid and tau statuses were stratified together, the results showed that the A status was the main driver of the changes seen in the p‐tau217+ assay as the levels increased from A‐T‐ to A‐T+ to A+T‐, and finally A+T+. This pattern was consistent across cohorts. The differences between A+ and A‐ participants were consistently observed across the cohorts and the AUCs for the Aβ PET were comparable, indicating reliability for classification purpose. This underscores a strong relationship between p‐tau217+ and Aβ burden examined on the Aβ PET scan.
In contrast, plasma p‐tau217+ performed moderately in correlating with tau PET results compared to Aβ PET results. While plasma p‐tau217+ showed significant difference between T+ and T‐ participants, it only achieved an average AUC of about 65%. In addition, T status no longer showed a significant association with p‐tau217+ levels after adjusting for A status. This finding aligns with previous reports indicating that further understanding of site‐specific patterns of tau phosphorylation is necessary before the p‐tau217+ assay can be considered as a potential surrogate for tau PET. 11 , 24 Ongoing efforts have led to the discovery of other assays such as p‐tau serine‐262 and serine‐356 which correlate more strongly with tau PET. 25 In contrast to our findings, Warmenhoven et al., 23 reported higher AUCs and accuracies for the Janssen plasma p‐tau217+ assay in identifying individuals with abnormal tau PET status. The discrepancies in our outcomes may arise from the fact that our study comprised community‐setting cohorts, while the other study focused on participants from a memory clinic with a higher prevalence of AD pathology. While the Warmenhoven study was composed of 48% cognitively impaired participants, our cohorts were composed of less than 10% combined. Additionally, there were only a few participants in the present study cohorts who were tau PET positive. Among those who were, many had SUVR values comparatively lower than those typically seen in symptomatic cohorts, which makes it challenging to establish a meaningful correlation between p‐tau217+ and tau PET.
A key takeaway from this study is the relatively high specificity and sensitivity of the Janssen plasma p‐tau217+ assay across all cohorts. This is particularly noteworthy given that the vast majority of the cohorts consisted of cognitive normal older adults. In terms of the performance values, there was high NPV which makes p‐tau217+ a good tool for the initial screening of high‐risk individuals or clinical trial participants. However, the low PPV suggests that an orthogonal test is needed to confirm the presence of AD pathology among those with high p‐tau217+. The low PPV could also be attributed to the fact that many participants in the cohorts were cognitively unimpaired. We adopted the optimal cutpoint strategy from the Alzheimer's Association workgroup revised criteria for diagnosing and staging of AD, 9 establishing an average threshold of 0.05 pg/ml across cohorts and when the cohorts were combined. To our knowledge, this is the first study to assess the cutoff in the Janssen plasma p‐tau217+ assay.
Regarding demographics, the results effectively highlighted the influence of demographics factors on biomarker testing. When adjusted for covariates such as APOE ε4 genotype, education, sex and age, the p‐tau217+ assay showed resilient associations within the cohort. This further emphasizes that these covariates have a minimal impact on p‐tau217 assay compared to other assays, like p‐tau181 which are more influenced by such factors. 12
Many studies on plasma p‐tau217+ have been limited by a lack of diversity, often overlooking its performance in real‐world settings. However, this study addresses this gap by examining various ethnicities, races, and socio‐economic backgrounds. For instance, all cohorts analyzed differences between NHW and B/AA, while the CoBrA cohort even included Asian participants despite being a very small proportion. In conclusion, this study provides strong evidence that the Janssen plasma p‐tau217+ assay is a reliable biomarker for prognostic purposes, even in community settings.
This study has a key limitation: the lack of long‐term follow‐up data to assess the relationship between the p‐tau217+ assay and the progression of AD. The current analysis is based solely on baseline assessment data, highlighting the critical need for follow‐up information. Without follow‐up data, it is challenging to evaluate the long‐term effectiveness and reliability of the p‐tau217+ assay as a biomarker for AD pathology. Future research should include follow‐up assessments to deepen our understanding and validate the role of the p‐tau217+ assay in monitoring AD progression.
While the Janssen plasma p‐tau217+ assay correlates with both Aβ and tau pathology, our results show that it is particularly effective in identifying individuals with abnormal Aβ plaques, as indicated by its strong correlation with Aβ PET imaging findings. This study suggests that p‐tau217+ could be a valuable biomarker for the early detection of AD pathophysiology with high sensitivity, specificity, and predictive values. Together, this may enable earlier detection, and access to interventions—both pharmacological and non‐pharmacological—potentially slowing disease progression.
CONFLICT OF INTEREST STATEMENT
X.Z. is an inventor on University of Pittsburgh provisional patents on anti‐tau antibodies and plasma amyloid‐beta peptide biomarker assays by immunoprecipitation‐mass spectrometry. G.T.B. and H.K. are employees of Janssen Research and Development. The plasma p‐tau217+ measurements were performed at Quanterix, and managed by Janssen Research and Development, but both parties were blinded to sample ID and were not involved in the data analysis. The co‐authors employed by Janssen provided comments on the manuscript and provided approval for submission of the manuscript. They are listed as inventors on the pending patent for the Janssen plasma p217+tau assay (JAB7064USPSP2). TKK has consulted for Quanterix Corporation, SpearBio Inc., Neurogen Biomarking LLC., and Alzheon, has served on advisory boards for Siemens Healthineers and Neurogen Biomarking LLC., outside the submitted work. Over the past 3 years, he has received in‐kind research support from Janssen Research Laboratories, SpearBio Inc., and Alamar Biosciences, as well as meeting travel support from the Alzheimer's Association and Neurogen Biomarking LLC., outside the submitted work. T.K.K. has received royalties from Bioventix for the transfer of specific antibodies and assays to third party organizations. T.K.K. is an inventor on patents and provisional patents regarding biofluid biomarker methods, targets and reagents/compositions, that may generate income for the institution and/or self should they be licensed and/or transferred to another organization. The other authors report no conflict of interest. Author disclosures are available in the Supporting Information.
CONSENT STATEMENT
All the studies were approved by the Institutional Review Board at the University of Pittsburgh. All subjects provided written informed consent.
Supporting information
Supporting Information
Supporting Information
ACKNOWLEGMENT
We thank all study participants, their families and caretakers in this cohort studies. The MYHAT, CoBrA, and HeartSCORE studies were funded by NIH grants R37AG023651, R01AG072641, R01AG052521, and R01HL089292. This study used biomarker testing infrastructure established with support from NIH/NIA R01AG083874. T.K.K. and the Karikari Laboratory were further supported by NIH/NIA (U24AG082930, P30AG066468, RF1AG077474, R01AG083156, R01AG025516, R01AG073267, R01AG075336, P01AG025204), NIH/NINDS (U01NS131740, U01NS141777), NIH/NIMH (R01MH108509), Aging Mind Foundation (DAF2255207), the Department of Defense (HT94252320064), the Anbridge Charitable Fund, and a professorial endowment from the Department of Psychiatry, University of Pittsburgh. The content of this article is solely the responsibility of the authors and does not necessarily represent the official views of the funders.
Balogun WG, Zeng X, Triana‐Baltzer G, et al. Cross‐cohort validation and cutpoint estimation of the Janssen plasma p‐tau217+ assay in predominantly cognitively normal community studies. Alzheimer's Dement. 2025;21:e70916. 10.1002/alz.70916
REFERENCES
- 1. Hansson O, Blennow K, Zetterberg H, et al. Blood biomarkers for Alzheimer's disease in clinical practice and trials. Nat Aging. 2023;3:506‐519. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Therriault J, Servaes S, Tissot C, et al. Equivalence of plasma p‐tau217 with cerebrospinal fluid in the diagnosis of Alzheimer's disease. Alzheimers Dement. 2023a;19(11):4967‐4977. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Barthélemy NR, Salvadó G, Schindler SE, et al. Highly accurate blood test for Alzheimer's disease is similar or superior to clinical cerebrospinal fluid tests. Nat Med. 2024;30:1085‐1095. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Fagan AM, Roe CM, Xiong C, et al. Cerebrospinal fluid tau/β‐amyloid42 ratio as a prediction of cognitive decline in nondemented older adults. Arch Neurol. 2007;64:343‐349. [DOI] [PubMed] [Google Scholar]
- 5. van Dyck CH, Swanson CJ, Aisen P, et al. Lecanemab in Early Alzheimer's Disease. N Engl J Med. 2023;388(1):9‐21. [DOI] [PubMed] [Google Scholar]
- 6. Schindler SE, Atri A. The role of cerebrospinal fluid and other biomarker modalities in the Alzheimer's disease diagnostic revolution. Nat Aging. 2023;3:460‐462. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Groot C, Cicognola C, Bali D, et al. Diagnostic and prognostic performance to detect Alzheimer's disease and clinical progression of a novel assay for plasma p‐tau217. Alz Res Therapy. 2022;14:67. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Balogun WG, Zetterberg H, Blennow K, et al. Plasma biomarkers for neurodegenerative disorders: ready for prime time? Curr Opin Psychiatry. 2023;36:112‐118. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Jack CR Jr, Andrews JS, Beach TG, et al. Revised criteria for diagnosis and staging of Alzheimer's disease: alzheimer's Association Workgroup. Alzheimers Dement. 2024;20(8):5143‐5169. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Palmqvist S, Janelidze S, Quiroz YT, et al. Discriminative Accuracy of Plasma Phospho‐tau217 for Alzheimer Disease vs Other Neurodegenerative Disorders. JAMA. 2020;324(8):772‐781. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Mielke MM, Dage JL, Frank RD, et al. Performance of plasma phosphorylated tau 181 and 217 in the community. Nat Med. 2022;28:1398‐1405. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Ashton NJ, Brum WS, Di Molfetta G, et al. Diagnostic Accuracy of a Plasma Phosphorylated Tau 217 Immunoassay for Alzheimer Disease Pathology. JAMA Neurol. 2024;81(3):255‐263. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Janelidze S, Bali D, Ashton NJ, et al. Head‐to‐head comparison of 10 plasma phospho‐tau assays in prodromal Alzheimer's disease. Brain. 2023;146(4):1592‐1601. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Doré V, Doecke JD, Saad ZS, et al. Plasma p217+tau versus NAV4694 amyloid and MK6240 tau PET across the Alzheimer's continuum. Alzheimers Dement Diagn Assess Dis Monit. 2022;14:1‐11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Therriault J, Ashton NJ, Pola I, et al. Comparison of two plasma p‐tau217 assays to detect and monitor Alzheimer's pathology. EBioMedicine. 2024;102:105046. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Feizpour A, Doré V, Krishnadas N, et al. Alzheimer's disease biological PET staging using plasma p217+tau. Commun Med. 2025;5:53. doi: 10.1038/s43856-025-00768-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Sullivan KJ, Liu A, Chang CH, et al. Alzheimer's disease pathology in a community‐based sample of older adults without dementia: the MYHAT neuroimaging study. Brain Imaging Behav. 2021;15:1355‐1363. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Zeng X, Lafferty TK, Sehrawat A, et al. multi‐analyte proteomic analysis identifies blood‐based neuroinflammation, cerebrovascular and synaptic biomarkers in preclinical Alzheimer's disease. Mol Neurodegener. 2024;19(1):68. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Ganguli M, Chang CC, Snitz BE, et al. Prevalence of mild cognitive impairment by multiple classifications: the Monongahela‐Youghiogheny Healthy Aging Team (MYHAT) project. Am J Geriatr Psychiatry. 2010;18:674‐683. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Cohen AD, Bruña R, Chang YF, et al. Connectomics in Brain Aging and Dementia—The Background and Design of a Study of a Connectome Related to Human Disease. Front Aging Neurosci. 2021;13:669490. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Bambs CE, Kip KE, Mulukutla SR, et al. Sociodemographic, clinical, and psychological factors associated with attrition in a prospective study of cardiovascular prevention: the Heart Strategies Concentrating on Risk Evaluation study. Ann Epidemiol. 2013;23(6):328‐333. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Gogola A, Lopresti BJ, Tudorascu D, et al. Biostatistical Estimation of Tau Threshold Hallmarks (BETTH) Algorithm for Human Tau PET Imaging Studies. J Nucl Med. 2023;64(11):1798‐1805. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Warmenhoven N, Salvadó G, Janelidze S, et al. A comprehensive head‐to‐head comparison of key plasma phosphorylated tau 217 biomarker tests. Brain. 2025;148(2):416‐431. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Therriault J, Vermeiren M, Servaes S, et al. Association of Phosphorylated Tau Biomarkers with Amyloid Positron Emission Tomography vs Tau Positron Emission Tomography. JAMA Neurol. 2023b;80(2):188‐199. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Islam T, Hill E, Abrahamson EE, et al. Phospho‐tau serine‐262 and serine‐356 as biomarkers of pre‐tangle soluble tau assemblies in Alzheimer's disease. Nat Med. 2025;31:574‐588. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supporting Information
Supporting Information