Plasma p-tau181 Level Predicts Neurodegeneration and Progression to Alzheimer's Dementia: A Longitudinal Study

Background: Plasma-based biomarkers would be potential biomarkers for early diagnosis of Alzheimer's disease (AD) because they are more available and cost-effective than cerebrospinal fluid (CSF) or neuroimaging. Therefore, we aimed to evaluate whether phosphorylated tau181 (p-tau181) in plasma could be an accurate AD predictor. Methods: Participants from the ADNI database included 185 cognitively unimpaired subjects with negative Aβ (CU–), 66 subjects with pre-clinical AD (CU with positive Aβ), 164 subjects with mild cognitive impairment with negative Aβ (MCI–), 254 subjects with prodromal AD (MCI with positive Aβ), and 98 subjects with dementia. Multiple linear regression models, linear mixed-effects models, and local regression were used to explore cross-sectional and longitudinal associations of plasma p-tau181 with cognition, neuroimaging, or CSF biomarkers adjusted for age, sex, education, and APOE genotype. Besides, Kaplan–Meier and adjusted Cox-regression model were performed to predict the risk of progression to dementia. Receiver operating characteristic analyses were performed to evaluate the predictive value of p-tau181. Results: Plasma p-tau181 level was highest in AD dementia, followed by prodromal AD and pre-clinical AD. In pre-clinical AD, plasma p-tau181 was negatively associated with hippocampal volume (β = −0.031, p-value = 0.017). In prodromal AD, plasma p-tau181 was associated with decreased global cognition, executive function, memory, language, and visuospatial functioning (β range −0.119 to −0.273, p-value < 0.05) and correlated with hippocampal volume (β = −0.028, p-value < 0.005) and white matter hyperintensity volume (WMH) volume (β = 0.02, p-value = 0.01). In AD dementia, increased plasma p-tau181 was associated with worse memory. In the whole group, baseline plasma p-tau181 was significantly associated with longitudinal increases in multiple neuropsychological test z-scores and correlated with AD-related CSF biomarkers and hippocampal volume (p-value < 0.05). Meanwhile, CU or MCI with high plasma p-tau181 carried a higher risk of progression to dementia. The area under the curve (AUC) of the adjusted model (age, sex, education, APOE genotype, and plasma p-tau181) was 0.78; that of additionally included CSF biomarkers was 0.84. Conclusions: Plasma p-tau181 level is related to multiple AD-associated cognitive domains and AD-related CSF biomarkers at the clinical stages of AD. Moreover, plasma p-tau181 level is related to the change rates of cognitive decline and hippocampal atrophy. Thus, this study confirms the utility of plasma p-tau181 as a non-invasive biomarker for early detection and prediction of AD.

The diagnostic criteria for AD was established based on amyloidosis, tau pathology, and neurodegeneration derived from cerebrospinal fluid (CSF), positron emission tomography (PET), and magnetic resonance imaging (MRI) proposed by the National Institute of Aging-Alzheimer Association (5,6). Though PET and CSF biomarkers are invaluable in AD-related brain pathology, the use of PET imaging and lumbar puncture is restricted to limited centers because of high prices, radiopharmaceuticals, and invasiveness. Conventional structural MRI neuroimaging is often used to evaluate AD progression. However, the advanced MR techniques are of limited value in AD diagnosis due to high heterogeneity, low sensitivity, and low signal-to-noise ratios (7).
Therefore, an affordable, non-invasive means is essential for large-scale screening programs and longitudinal studies. The previous study has developed a new ATN framework, focused on plasma biomarkers including Aβ 42 /Aβ 40 ratio (A), plasma p-tau181 (T), and neurofilament light (N) (8). Plasma Aβ 42 /Aβ 40 ratio was decreased in cognitively normal subjects with subjective cognitive decline, the earliest stage of AD, and plasma Aβ 42 /Aβ 40 level can predict amyloid-PET status (9,10). Nevertheless, peripheral production of Aβ peptides makes it difficult for clinical practice (11). Plasma neurofilament light level was increased in patients with AD (12,13), but not specific for AD given that many other neurodegenerative diseases such as amyotrophic lateral sclerosis, Creutzfeldt-Jakob disease, and frontotemporal dementia showed significantly high values (14)(15)(16). Recent prospective studies found that blood p-tau181 can be a potential diagnostic marker of AD (17)(18)(19). A study of familial AD indicated that plasma p-tau181 level was increased before symptom onset (20). However, longitudinal analysis is scarce on plasma p-tau181 prognostic value for progression to dementia and relationships between plasma p-tau181 and multiple ADassociated cognitive domains remain unclear.
In the present study, we investigated longitudinal associations of plasma p-tau181 with AD-related CSF biomarkers, multiple cognitive domains, and volumes of hippocampus and white matter hyperintensity (WMH). Moreover, we assessed the value of plasma p-tau181 in predicting progression to dementia and compared the diagnostic values of different models.

Study Population
Participants were selected from the Alzheimer's Disease Neuroimaging Initiative (ADNI) data (http://adni.loni.usc.edu) in October 2020. The ADNI database was initiated in 2003 to detect MCI and early AD dementia progression based on the combinations of CSF or plasma biomarkers, serial MRI, PET, and clinical and neuropsychological information. This project was approved by institutional review boards of all participating institutions and written informed consent was obtained from all participants before inclusion in the study.
The study included 767 ADNI participants with baseline plasma p-tau181. Definitions of the participant classifications have been described in the previous study, including cognitively unimpaired (CU), mild cognitive impairment (MCI), and dementia (21). Besides, in this study, the threshold values of Aβ were 880 pg/ml for CSF Aβ1-42 or 1.11 SUVR for AV45-PET scan (22). Then, CU and MCI subjects were split into four subgroups: subjects with negative Aβ-related pathological changes (CUand MCI-) and subjects with positive Aβ-related pathological changes (CU+ and MCI+). CU+ subjects were regarded as patients with pre-clinical AD and MCI+ subjects were regarded as patients with prodromal AD. The enrollment flow chart is given in Supplementary Figure 1.

Plasma p-tau181 Quantification
Level of plasma p-tau181 was quantified with the Single Molecule array (Simoa) technique combined with two monoclonal antibodies (Tau12 and AT270), as described previously (17). Moreover, the analysis of plasma p-tau181 was carried out at the Clinical Neurochemistry Laboratory, University of Gothenburg, Sweden.

Cognitive Evaluation
The general cognition was assessed by Mini-Mental State Examination (MMSE) and Montreal Cognitive Assessment (MoCA), and composite scores were used to reflect memory, executive function, and language-related and visuospatial domains. All these neuropsychological tests were evaluated at baseline and yearly thereafter.

Neuroimaging Analyses
The average of the mean florbetapir standard uptake value ratio (SUVR) was calculated using four cortical regions (frontal, anterior cingulate, precuneus, and parietal cortex) normalized to the whole cerebellum. WMH volume was computed with the segmentation of high-resolution 3D T1-weighted and FLAIR sequences based on a Bayesian approach. The average volume of the right and left hippocampus was obtained using FreeSurfer software (http://surfer.nmr.mgh.harvard.edu/fswiki).

CSF Measurements
Levels of Aβ, total tau (t-tau), and p-tau in CSF were measured using Elecsys immunoassays at the Clinical Neurochemistry Laboratory, University of Gothenburg, Sweden (BioFINDER) or at the Biomarker Research Laboratory, University of Pennsylvania, USA (23).

Statistical Analyses
The level of plasma p-tau181 was compared between five subgroups with the Mann-Whitney U test. The rate of change in plasma p-tau181 was calculated in linear mixedeffects models with age, sex, education, and APOE genotype as covariates. Baseline associations of plasma p-tau181 with cognition and other biomarkers were tested by linear regression models adjusting for age, sex, education, and APOE genotype.
Associations of baseline plasma p-tau181 with longitudinal level of other biomarkers were tested by linear mixed-effects models. The local regression was employed for associations between baseline plasma p-tau181 and the change rate of other biomarkers. Furthermore, clinical progression to dementia was estimated using Kaplan-Meier survival curves with log-rank test. The receiver operating characteristic (ROC) curves and the area under the curve (AUC) of the key biomarkers of plasma and CSF were calculated to predict progression to dementia. All tests used

Demographic Data
We included 767 subjects (mean ± SD age, 72.3 ± 7.1 years; education, 16.3 ± 2.6 years), of whom 366 were women (47.7%). Other descriptive statistics of the subjects are given in Table 1. The cohort were divided into five subgroups: 185 from the CUgroup (cognitively unimpaired subjects with negative Aβ), 164 from the MCI-group (MCI-subjects with negative Aβ), 66 from the CU+ group (pre-clinical AD, CU-subjects with positive Aβ), 254 from the MCI+ group (prodromal AD, MCI-subjects with positive Aβ), and 98 from the dementia group. Figure 1 shows that the level of plasma p-tau181 is highest in dementia, followed by that in prodromal AD and pre-clinical AD and the lowest level were found in MCI-(p < 0.05), and there is no significant difference between MCI-and CU-.
Cross-sectional Associations of Plasma p-tau181 With CSF Biomarkers, Cognition, and Imaging Markers  p-tau (β = 0.25, p = 0.016) were found in the dementia group. As for the cohort, plasma p-tau181 was significantly associated with CSF Aβ and above biomarkers (Supplementary Table 1). Figure 3 indicates that there may be non-linear relationship between plasma p-tau181 and cognition. In the plasma p-tau181 medium subgroup ( Table 2). There was no significant associations between plasma p-tau181 and white matter hyperintensity (WMH) volume at baseline.

Analyses of Longitudinal Conversion to Dementia
Kaplan-Meier survival analysis demonstrated that subjects with higher levels (>14.982 pg/ml) of plasma p-tau181 were associated with higher progression compared to those with lower levels (<14.981 pg/ml) of plasma p-tau181 (log-rank p < 0.0001; Figure 5). The basic model (model 1: age, sex, education, and APOE ε4 genotype) discriminated clinical progression with the FIGURE 3 | (A-H) Cross-sectional associations of plasma p-tau181 with cognition and neuroimaging biomarkers. Plasma p-tau181 is categorized into three 33% terciles: low <12.03 pg/ml, medium 12.04-19.63 pg/ml, and high >19.65 pg/ml. Significant correlations (p < 0.05) are indicated by "*" in different levels of plasma p-tau181.

DISCUSSION
This longitudinal study demonstrated that plasma p-tau181 level was elevated during the clinical progression of dementia, and associations with CSF biomarkers, different cognitive domains, and imaging markers were validated. Our findings highlighted that plasma p-tau181 was (1) positively associated with CSF ttau and p-tau in pre-clinical AD, prodromal AD, and dementia stage; (2) longitudinally correlated with change and change rate of global cognition, executive function, memory, languagerelated, and visuospatial domains and hippocampal volume; (3) significantly predictive of clinical conversion to dementia.
We showed that plasma p-tau181 of subjects with positive Aβ were higher than subjects with negative Aβ. Consistent with previous investigations, increased plasma p-tau181 level was associated with increased Aβ pathology (24). Moreover, the change of plasma p-tau181 occurred before the onset of abnormal  Aβ pathology, indicating that accumulation of Aβ could stimulate the early increase in plasma p-tau181 (25). As expected, strong associations were found between plasma and CSF levels of p-tau181, indicating that plasma p-tau181 might be used as a surrogate marker of tau pathology, as well as CSF p-tau (17,18). However, the DELCODE study had demonstrated that plasma tau was not related to core markers of CSF in the stage of subjective cognitive decline (26). In line with the previous study (27), there was no association between plasma p-tau181 and CSF p-tau in subjects with negative Aβ. Inconsistently, we also found that plasma p-tau181 was associated with CSF Aβ in CU with negative Aβ, caused by the more specific subgroups (27). Additionally, plasma p-tau181 was significantly related to CSF t-tau/Aβ and p-tau/Aβ ratios, but not CSF Aβ, in pre-clinical AD and prodromal AD. This was supported by previous studies that ratios were more accurate biomarkers than single ones in predicting the progression of AD (28,29). Previous studies had demonstrated that plasma p-tau181 was correlated with cognitive decline (18,30), and we found consistent associations between increased plasma p-tau181 and decreased cognitions in the whole cohort. Interestingly, the baseline plasma p-tau181 had an S-shaped relationship with global cognition and multiple cognitive domains in our study. Besides, in the stage of prodromal AD, the baseline plasma p-tau181 showed obvious correlations with changes of cognitions and hippocampal atrophy. The associations of plasma p-tau181 with hippocampal atrophy and gray matter loss indicated that plasma p-tau181 was related to Alzheimer's neuronal loss (30). Furthermore, the linear mixed-effects model analysis suggested that plasma p-tau181 related to WMH volume in prodromal AD (p = 0.01). From this, it could be speculated that plasma p-tau181 was relevant to cerebrovascular disease. What is more, associations are also presented as S-shaped patterns between plasma p-tau181 and change rates of global cognition, executive function, memory, language-related, and visuospatial abilities, as well as hippocampal atrophy rate. Therefore, it could be hypothesized that there is a threshold period for p-tau181 and only moderate levels of p-tau181 are associated with change rates of the above indicators. After a specific range of p-tau181, these effects are gradually leveling out.
The changes of phosphorylation sites are different at different stages of AD progression, besides plasma p-tau217 and p-tau181 are increased 20 years before the formation of tau aggregates (31)(32)(33). Consistent with prior research, plasma p-tau181 was a potential biomarker for diagnosing AD and predicting AD progression (19,34,35). Plasma p-tau181 was a specific marker in AD dementia, which can distinguish AD from frontotemporal dementia, vascular dementia, progressive supranuclear palsy, FIGURE 6 | Receiver operating characteristic analysis curves denoting CSF biomarkers and plasma p-tau 181 for progressing to dementia. Model 1 includes age, sex, years of education, and APOE ε4 genotype. Model 2 additionally includes plasma p-tau 181. Model 3 additionally accounting for CSF biomarkers including Aβ1-42, t-tau, and p-tau. All above factors are contained in model 4.
corticobasal basal syndrome, primary progressive aphasia, Parkinson's disease, and multiple system atrophy (17). In patients with familial AD, the level of plasma p-tau181 was comparable among APP and PSEN1 mutation carriers (20). Though plasma p-tau181 changed without a specific gene, it was associated with APOE ε4 (36). The association between tau and APOE ε4 was also independent of Aβ, primarily mediated by activated microglia (37).
Nevertheless, several limitations should be addressed in this study. Firstly, the ADNI study is mainly based on Caucasoid populations, and consequently, the results may not be directly applied to other racial/ethnic groups. Secondly, due to the strict inclusion criteria, individuals with severe vascular pathology were excluded, which probably attenuated relevance between plasma p-tau181 and cerebrovascular diseases. Also, follow-up data of biomarkers were not complete, and therefore, we cannot test whether plasma p-tau181 was congruent with CSF p-tau181 and tau PET.
In summary, plasma p-tau181 was increased from the preclinical stage of AD and longitudinally associated with multiple cognitive domains decline and hippocampal atrophy. Thus, plasma p-tau181, a readily accessible biomarker, could be of high predictive and diagnostic values for AD. However, international multicenter, large-sample, and longitudinal studies are needed to validate these hypotheses.

DATA AVAILABILITY STATEMENT
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/Supplementary Material.

ETHICS STATEMENT
The ADNI study was conducted according to the Good Clinical Practice guidelines, the Declaration of Helsinki, and US 21 CFR: Part 50 (Protection of Human Subjects) and Part 56 (Institutional Review Boards). The ADNI study was conducted in compliance with HIPAA regulations. The first author of this paper was granted administrative permissions to access the anonymized ADNI data in May, 2019. Ethics approval for data collection in ADNI was obtained by each ADNI participating institution's institutional review board (https:// adni.loni.usc.edu/wp14%20content/uploads/how_to_apply/ ADNI_Acknowledgement_List.pdf). All study participants or authorized representatives provided written informed consent. Ethics approval was obtained from the institutional review