Validation of the robustness of blood-based biomarkers for predicting amyloid-β positivity in Chinese populations

Background: Blood-based biomarkers (BBBs) of Alzheimer’s disease (AD) provide a promising, minimally invasive alternative for detecting cerebral amyloid-β (Aβ) pathology. However, a lack of robust validation across diverse platforms and populations has hindered their broader clinical adoption.

Objective: This study aimed to cross-platform validation of the robustness of BBBs for predicting Aβ positivity in a Chinese population.

Methods: The whole cohort (N = 1,254) of AD clinical spectrum underwent cognitive assessments, cranial MRI scans, and Aβ PET scans. Subcohort 1 (N = 504) underwent Simoa-based quantification of peripheral blood Aβ40, Aβ42, p-tau181, and NfL. Subcohort 2 (N = 262) underwent additional single molecule assays (Simoa) based quantification of p-tau217 and GFAP. The whole cohorts (Subcohort 1, Subcohort 2, and the remaining population) were measured for the aforementioned six biomarkers (Aβ40, Aβ42, p-tau181, p-tau217, NfL, and GFAP) using light-initiated chemiluminescent assays (LiCA). We validated the robustness of BBBs for predicting Aβ positivity in Chinese populations, with a focus on p-tau217.

Results: The BBBs of Aβ42/40, p-tau181, p-tau217, GFAP, and NfL have demonstrated remarkable robustness in identifying Aβ positivity within the Chinese population, as evidenced by both LiCA and Simoa assays. Among these markers, p-tau217 has emerged as the most accurate, performing robustness in both the whole cohort and cognitively normal individuals. Utilizing a dual-threshold approach for p-tau217, only 16% of samples fell into the intermediate range, thus requiring additional Aβ PET testing.

Conclusion: Blood-based biomarkers have demonstrated good robustness for predicting Aβ pathology in the Chinese population, with plasma p-tau217 standing out as the most promising marker for early detection of AD-related changes.

Introduction

Dementia is a rapidly growing and increasingly complex global public health challenge, with two-thirds of affected individuals residing in low- and middle-income countries (Kalaria et al., 2024). China, with the largest population of dementia patients in the world (Jia et al., 2020), faces a particularly acute burden, as Alzheimer’s disease (AD) and other dementias is the fifth leading cause of death in the country (Ren et al., 2022; Yang et al., 2024). This places a significant strain on both public resources and healthcare systems. Early detection of AD is essential, as it can delay disease progression, reduce healthcare costs, and improve quality of life for patients.

The current diagnostic criteria for AD rely on the detection of brain amyloid-β (Aβ) through cerebrospinal fluid (CSF) analysis and Aβ positron emission tomography (PET) imaging, which are used to enhance diagnostic certainty (Jack et al., 2018). However, the widespread application of these biomarkers in clinical settings is severely limited by their invasiveness, high costs, and restricted availability (Fargo et al., 2016). As a result, there is an urgent need to validate and implement accessible, scalable, and cost-effective biomarkers that can be readily applied in clinical practice.

Blood-based biomarkers (BBBs) represent a promising alternative to Aβ PET and AD CSF biomarkers, offering a less invasive and more cost-effective approach to detecting AD pathology (Leuzy et al., 2022). Proposed AD BBBs include the blood Aβ42/40, phosphorylated tau (p-tau) species—such as p-tau181, p-tau217, and p-tau231—as well as non-specific biomarkers like glial fibrillary acidic protein (GFAP) and neurofilament light chain (NfL) (Jack et al., 2024). Advances in highly sensitive detection techniques, such as the single molecule assay (Simoa) and the Fujirebio lumipulse assay, have enabled accurate measurement of distinct brain-derived proteins in blood (Figdore et al., 2024b; Rissin et al., 2010). However, reported diagnostic accuracy remains inconsistent, with significant variability in the performance of plasma Aβ42/40 and p-tau assays across different platforms and manufacturers (Janelidze et al., 2021, 2023; Warmenhoven et al., 2025). This inconsistency has hindered widespread replication of these findings.

Over the past 3 years, large-scale Chinese cohorts have successively validated plasma p-tau217 and p-tau181 as robust indicators of amyloid-PET positivity. In a 2024–2025 nationwide study covering Shenzhen (GHABS, N = 425) and Shanghai (Huashan, N = 297), Lan et al. (2025) showed that p-tau217 (Simoa-ALZpath) classified Aβ-PET positivity with AUC 0.90–0.91 and outperformed p-tau181 (AUC 0.80–0.84) across the entire AD continuum; a two-cutoff strategy further reduced the indeterminate zone to 23%–28%, suggesting that ≈75% of scans could be avoided. Lin et al. (2025) reported the first East-Asian cross-cultural validation (Korea-Taiwan, N = 270) using identical Simoa assays; p-tau217 achieved AUC 0.936 for Aβ PET, with NPV 97.5% in the low-risk stratum and PPV 86% in the high-risk stratum, performances that were reciprocally confirmed when Taiwanese-derived models were tested in Korean participants.

In this study, we aimed to validate the robustness of BBBs for predicting Aβ positivity in Chinese populations. Our primary objective was to assess the performance of BBBs, including Aβ42/40, p-tau181, p-tau217, GFAP, and NfL, as detected by different platforms, in predicting Aβ positivity in two subgroups. We further validated the robustness of these biomarkers in a larger, full cohort and conducted robustness analyses in cognitively normal individuals to evaluate the inter-assay reproducibility with a particular focus on the potential of p-tau217 in predicting Aβ positivity. Additionally, by implementing a two-cutoff LiCA p-tau217 algorithm to minimize confirmatory PET/CSF testing in a large Shanghai memory-clinic and community sample.

Materials and methods

Participants

Between March 2019 and October 2024, a total of 1,254 individuals were recruited from the Shanghai Sixth People’s Hospital Affiliated to Shanghai Jiao Tong University School of Medicine through the outpatient memory clinic and community outreach efforts, as part of the Chinese Preclinical Alzheimer’s Disease Study (CPAS; Cui et al., 2023). Clinical diagnoses were subsequently established according to established criteria for cognitively unimpaired (CU) individuals, mild cognitive impairment (MCI), and AD dementia. CU participants were defined as those without cognitive impairment. MCI participants were diagnosed in accordance with the Petersen criteria (Petersen et al., 1997; Petersen, 2004), while the diagnosis of AD was made by experienced neurologists, following the National Institute on Aging and Alzheimer’s Association criteria for probable AD dementia (McKhann et al., 2011). A detailed description of the inclusion and exclusion criteria for participant recruitment is provided in our previous studies (Cui et al., 2023; Pan et al., 2022).

Complete datasets were obtained for all participants, including cognitive assessments, apolipoprotein E (APOE) genotyping (with APOE ε4 carriage defined as the presence of at least one ε4 allele), cranial MRI scans, Aβ PET scans and BBBs analyses. All tests, blood collections, and imaging procedures were completed within a 1-month period. Blood samples were collected from all participants using anticoagulant-containing vacutainers, centrifuged, aliquoted, and stored for blood biomarker analysis. All participants provided written informed consent after being fully informed of the study procedures and objectives.

Clinical assessments

Each participant underwent a comprehensive cognitive assessment. The Chinese version of Addenbrooke’s Cognitive Examination III (ACE-III-CV) and the Chinese version of the Montreal Cognitive Assessment-Basic (MoCA-BC) were used to evaluate global cognitive function. An extensive neuropsychological battery was administered to assess three key cognitive domains: memory, language, and attention/executive function, as previously described (Cui et al., 2023). Specifically, memory function was assessed using the long-delayed recall and recognition components of the Auditory Verbal Learning Test (AVLT). Language ability was evaluated with the Animal Fluency Test (AFT, total score) and the 30-item Boston Naming Test (BNT, total score). Executive function was measured using parts A and B of the Shape Trail Test (STT, time to completion).

Blood biomarker assessment

Light-initiated chemiluminescent assays (LiCA)

A total of 1,254 participants underwent blood testing for six biomarkers (Aβ42, Aβ40, p-tau181, p-tau217, GFAP, and NfL) using LiCA assays conducted by Chemclin Diagnostics, Beijing, China, in December 2024. These participants were collectively designated as the whole cohort, which was further stratified into two subcohorts (Figure 1A).

FIGURE 1

Figure 1. (A) Flowchart of the study population. (B–E) ROC curves illustrating the discriminative performance of LiCA versus Simoa assays for identifying abnormal Aβ PET in subcohort 1, with corresponding AUC values. ROC, receiver operating characteristic; AUC, area under the curve; Aβ, amyloid-β; PET, positron emission tomography; p-tau, phosphorylated tau; NfL, neurofilament light; ns, not significant. LiCA represents the Chemclin LiCA^® kits. Subcohort 1 [N = 504; mean age 67.8; Female, N (%) = 309 (61%)].

Testing was performed on the Chemclin LiCA 800 automated immunoassay analyzer (Chemclin Diagnostics, Beijing, China) using the Chemclin LiCA^® plasma Aβ42, Aβ40, p-tau181, p-tau217, NfL, and GFAP kits according to the manufacturer’s instructions (Ullman et al., 1996; Wang et al., 2022). EDTA-plasma samples were analyzed directly from the aliquot tubes without the need for cleanup, centrifugation, or dilution prior to measurement. LiCA employs a dual-antibody sandwich method with two-step incubations for blood biomarker detection. One antibody was pre-conjugated to nano-scale Chemibeads, while the other antibody was biotinylated. During the first step of incubation, target proteins within the sample form a sandwich complex by binding to both the Chemibead-immobilized detection antibody and the biotinylated capture antibody. In the subsequent step, streptavidin-coated Sensibeads are introduced, leveraging the high-affinity biotin-streptavidin interaction to bind the sandwich complex. Upon laser excitation, the Sensibeads generated singlet oxygen molecules that diffused into the Chemibeads, triggering a chemiluminescent reaction. In the absence of the target analyte, no sandwich complex is formed, resulting in unbound beads separated by a distance exceeding 200 nm. This spatial separation prevents singlet oxygen-mediated energy transfer, thereby eliminating background signal generation. The LiCA assays operates entirely in a homogeneous phase, eliminating the need for wash steps typically required in conventional immunoassay protocols. The principle of LiCA was illustrated in Supplementary Figure 1.

Quanterix Simoa assays

Within the whole cohort, 504 participants underwent additional testing for four AD blood biomarkers (Aβ40, Aβ42, p-tau181, and NfL) using the Simoa assay at Company A, forming subcohort 1. Additionally, 262 participants were tested for six AD blood biomarkers (Aβ40, Aβ42, p-tau181, p-tau217, GFAP, and NfL) using the Simoa assay at Company B, constituting subcohort 2. The remaining 488 participants did not undergo Simoa-based biomarker detection. The laboratory technicians were blinded to both the clinical diagnosis and the Aβ PET results.

The plasma biomarker testing was performed in Shanghai, China, by Company A and Company B using the Simoa platform (Quanterix, Billerica, MA, USA) according to the manufacturer’s instructions. At Company A, the Neurology 3-Plex A Assay Kit was used to measure Aβ42 and Aβ40, the P-tau 181 V2 Assay Kit was used for p-tau181, and the NF-light Assay Kit was used for NfL, as previously reported (Pan et al., 2023). At Company B, the Neurology 4-Plex E Advantage kit was employed to measure Aβ42, Aβ40, GFAP, and NfL, while the P-tau 181 V2 Assay Kit and the ALZpath p-Tau 217 V2 Assay Kit were used to measure p-tau181 and p-tau217, respectively, as described previously (Wu et al., 2024). The Simoa measurements for subcohorts 1 and 2 were both conducted on the Quanterix platform. All Simoa-based biomarker testing was completed within 1 year of blood collection. Subcohort 1 (N = 504) and Subcohort 2 (N = 262) were assembled by consecutive enrollment during March 2019–December 2021 and January 2023–January 2024, respectively. The assignment of all subcohorts was random. The measurement of p-tau217 and GFAP was not conducted in Subcohort 1 because the corresponding commercial assay kits were not available at the time of analysis. Consequently, Subcohort 2 was tested for six AD-related blood biomarkers following the commercial release of p-tau217 and GFAP kits.

PET image acquisition and processing

Amyloid imaging was performed using [¹⁸F]florbetapir PET/CT scans on a Biography 64 PET/CT scanner (Siemens, Erlangen, Germany) at the PET Center of Huashan Hospital Affiliated to Fudan University. The imaging protocol followed well-established parameters as described in previous studies (He et al., 2024; Ren et al., 2023). Approximately 370 MBq ( ± 10%) of [¹⁸F]florbetapir was administered intravenously, followed by a 20-min PET scan acquisition after a 50-min uptake period. The acquired PET data were reconstructed using a filtered back-projection algorithm with comprehensive corrections for attenuation, normalization, dead time, photon attenuation, scatter, and random coincidences. The reconstructed PET images had a matrix size of 168 × 168 × 148 voxels, with a voxel dimension of 2.04 mm × 2.04 mm × 1.50 mm.

Using the Amyvid reading protocol, Aβ PET images were interpreted via visual inspection by three experienced raters who were blinded to the clinical diagnosis (He et al., 2024). Consensus among the raters determined the classification of participants as either Aβ positive (Aβ+) or Aβ negative (Aβ−) based on the PET images.

Statistical analysis

Statistical analyses were conducted using SPSS version 23 (IBM SPSS) and MedCalc statistical software. For comparisons of characteristics between PET Aβ− and Aβ+ groups, independent-sample t-tests were used for normally distributed continuous variables, while Mann–Whitney U tests were applied for skewed continuous variables. Chi-square tests were used for categorical variables. Receiver operating characteristic (ROC) analysis was performed to evaluate the sensitivity and specificity of each blood-based biomarker in identifying individuals with abnormal Aβ PET status, with the area under the curve (AUC) calculated as a measure of diagnostic performance. Differences in AUC values among individual biomarkers were tested using the DeLong method. Optimal ROC cut points were determined based on the sensitivity and specificity.

Calculation method of dual cut point analysis: Only samples with values below the lower threshold were classified as negative and those above the upper threshold as positive; subsequent comparison with the PET gold standard yielded the Sensitivity (Sens)/Specificity (Spec). Intermediate values were not included in the calculation of Sens/Spec, but were incorporated into the gray zone proportion and the rate of requiring further examination.

Results

Participant characteristics by clinical and Aβ status

Table 1 summarizes the clinical characteristics and blood biomarker distribution of the participants. The study included 1254 participants, consisting of 554 CU individuals, 407 with MCI, and 293 with AD dementia, as determined by consensus clinical diagnosis. Across all groups, the proportion of females was 63% in the CU group, 60% in the MCI group, and 59% in the AD dementia group. The mean age of participants ranged from 65.2 years in the CU group to 69.2 years in the AD dementia group. The prevalence of both APOE ε4 carriage and Aβ positivity increased progressively with clinical severity. In the CU group, 19% were APOE ε4 carriers, compared to 29% in the MCI group and 47% in the AD dementia group. Similarly, the proportion of participants with abnormal Aβ PET were observed in 10% of the CU group, 28% of the MCI group, and 75% of the AD dementia group.

TABLE 1

Table 1. The clinical characteristics, and blood biomarker distribution of participants.

The participants were divided into two groups based on their Aβ PET status: Aβ negative and Aβ positive. Table 2 shows the demographic and clinical characteristics, as well as neuroimaging and plasma biomarker data. Significant differences were observed between the Aβ− and Aβ+ groups. Specifically, Aβ+ participants were older (p < 0.001 in subcohort 1 and the whole cohort; p < 0.01 in subcohort 2), had a higher prevalence of APOE ε4 carriers (all p < 0.001), and exhibited lower cognitive scores. Additionally, Aβ+ participants had lower education levels compared to Aβ− participants in subcohort 1 (p < 0.01) and the whole cohort (p < 0.001). In terms of plasma biomarkers, the Aβ+ group exhibited elevated levels of p-tau181, p-tau217, GFAP, and NfL, as well as a reduced Aβ42/40 ratio (all p < 0.001). No statistically significant differences were observed between the Aβ− and Aβ+ groups in terms of sex. Cognitive scale results are presented in Table 1: the Aβ+ group scored an average of 2.1 points lower on the MMSE, 2.7 points lower on the MoCA-BC, and 1.8 points lower on the AVLT delayed-recall subscore, all p < 0.001; the STT-B completion time was prolonged by 14 s, p < 0.001.

TABLE 2

Table 2. The characteristics of the subcohort 1, subcohort 2, and the whole cohort.

The performance of BBBs detected by different platforms in predicting Aβ positivity

We performed a comprehensive comparison of three BBBs (Aβ42/40, p-tau181, and NfL) measured by LiCA and Simoa assays to assess their predictive performance for Aβ status in subcohort 1. As illustrated in Figures 1B–E, ROC curves were generated to evaluate the performance of these biomarkers in detecting abnormal Aβ PET. The LiCA Aβ42/40 assay demonstrated significantly superior discriminative accuracy in identifying Aβ PET-positive individuals, achieving an AUC of 0.81 (95% CI: 0.77–0.84), compared to the Simoa Aβ42/40 assay, which yielded an AUC of 0.72 (95% CI: 0.68–0.76; p < 0.01). For p-tau181, the discriminative performance of the LiCA and Simoa assays was comparable, with AUCs of 0.80 (95% CI: 0.76–0.84) and 0.77 (95% CI: 0.73–0.81), respectively, showing no significant difference. Similarly, the LiCA and Simoa NfL assays exhibited equivalent performance, both achieving AUCs of 0.73 (95% CI: 0.68–0.77 for LiCA and 0.69–0.77 for Simoa). Notably, a comparison of p-tau217 and GFAP levels between the LiCA and Simoa assays was not conducted, as the Simoa assays were not utilized for the detection of these biomarkers in subcohort 1.

Next, we performed a comparison of five BBBs—Aβ42/40, p-tau181, p-tau217, GFAP, and NfL—measured by LiCA and Simoa assays to evaluate their predictive performance for Aβ PET status in subcohort 2. Figure 2 displays the ROC curves for detecting abnormal Aβ PET in this subcohort using both assays. When identifying individuals with abnormal Aβ PET, the LiCA Aβ42/40 assay demonstrated significantly superior discriminative accuracy, achieving an AUC of 0.87 (95% CI: 0.82–0.91), compared to the Simoa Aβ42/40 assay, which yielded an AUC of 0.77 (95% CI: 0.71–0.82; p < 0.05). For p-tau181, the LiCA assay showed a higher AUC (0.86; 95% CI: 0.81–0.90) than the Simoa assay (0.81; 95% CI: 0.75–0.85), although this difference was not statistically significant. The discriminative performance of LiCA and Simoa assays for p-tau217 was comparable, with AUCs of 0.95 (95% CI: 0.92–0.98) and 0.94 (95% CI: 0.91–0.97), respectively. Similarly, the LiCA and Simoa GFAP assays exhibited equivalent performance, with AUCs of 0.90 (95% CI: 0.85–0.93) and 0.89 (95% CI: 0.85–0.93), respectively. Likewise, the LiCA and Simoa NfL assays demonstrated comparable performance, with AUCs of 0.68 (95% CI: 0.61–0.73) and 0.69 (95% CI: 0.63–0.75), respectively. The discriminative performance of LiCA and Simoa assays for p-tau217/Aβ42 was comparable, with AUCs of 0.95 (95% CI: 0.89–0.98) and 0.93 (95% CI: 0.87–0.97), respectively.

FIGURE 2

Figure 2. ROC curves illustrating the discriminative performance of LiCA versus Simoa assays for identifying abnormal Aβ PET in subcohort 2 (A–F), with corresponding AUC values (G). ROC, receiver operating characteristic; AUC, area under the curve; Aβ, amyloid-β; PET, positron emission tomography; p-tau, phosphorylated tau; GFAP, glial fibrillary acidic protein; NfL, neurofilament light; ns, not significant. LiCA represents the Chemclin LiCA^® kits. Subcohort 2 [N = 262; mean age 68.2; Female, N (%) = 161 (61.5%)].

Blood-based biomarkers detected by different platforms have shown robustness in predicting Aβ positivity.

Robustness analysis on the performance of BBBs in predicting Aβ positivity in the whole cohort

To confirm the robustness of findings observed in the subcohorts, we further evaluated the diagnostic performance of BBBs measured by LiCA assays in differentiating Aβ status in the larger whole cohort. ROC curve analysis revealed that p-tau217 demonstrated the highest diagnostic accuracy for classifying Aβ PET status (Aβ+ vs. Aβ−), achieving an AUC of 0.92 (95% CI: 0.90–0.94). Among the biomarkers evaluated, p-tau217 emerged as the most accurate predictor of Aβ PET status, followed by p-tau217/Aβ42 and GFAP, which achieved AUCs of 0.91 (95% CI: 0.90–0.93) and 0.87 (95% CI: 0.84–0.89), respectively. Notably, Plasma p-tau217 levels exhibit significant differences among CU, MCI, and AD cohorts (Supplementary Figure 2). In contrast, p-tau181 and Aβ42/40 showed moderate diagnostic performance, with identical AUC values of 0.80 (95% CI: 0.77–0.82 for p-tau181 and 95% CI: 0.78–0.83 for Aβ42/40). NfL exhibited significantly lower diagnostic accuracy compared to the other biomarkers, with an AUC of 0.74 (95% CI: 0.71–0.76). The ROC curves for all biomarkers are illustrated in Figure 3A.

FIGURE 3

Figure 3. (A,B) ROC curves with Youden index cutpoints of blood-based biomarkers by LiCA for prediction of Aβ PET positivity in the whole cohort. (C,D) ROC curves with Youden index cutpoint of p-tau217 by LiCA for prediction of Aβ PET positivity in the CU individuals of the whole cohort. ROC, receiver operating characteristic; AUC, area under the curve; Aβ, amyloid-β; PET, positron emission tomography; p-tau, phosphorylated tau; GFAP, glial fibrillary acidic protein; NfL, neurofilament light; CU, cognitively unimpaired. LiCA represents the Chemclin LiCA^® kits.

We have further established the cut-off values for these five biomarkers, among which p-tau217 has the highest sensitivity and specificity. An optimal cutoff value of 0.47 pg/mL for p-tau217 was established, yielding 85.8% sensitivity and 89.6% specificity in predicting Aβ status. The optimal cutoff values for all biomarkers are illustrated in Figure 3B. These findings highlight the superior performance of p-tau217 in predicting Aβ status.

Diagnostic performance of p-tau217 in predicting Aβ positivity in CU individuals

Amyloid-β deposition exists even before the onset of clinical symptoms, and early identification of Aβ positive individuals allows for timely intervention or participation in clinical trials. Therefore, we further analyzed the diagnostic performance of plasma p-tau217 in CU individuals to validate the robustness of blood-based p-tau217 in identifying those with abnormal Aβ PET. Supplementary Table 1 summarizes the clinical characteristics and blood biomarker distribution of the participants. Compared to Aβ− participants, Aβ+ participants were older (p < 0.01), had a higher prevalence of APOE ε4 carriers (p < 0.001), exhibited lower cognitive scores in MoCA-BC (p < 0.05), and had elevated levels of p-tau181, p-tau217, GFAP, and NfL, as well as a reduced Aβ42/40 (p < 0.001 for Aβ42/40, p-tau181, p-tau217, GFAP, and p < 0.01 for NfL). No statistically significant differences were observed between the Aβ− and Aβ+ groups in terms of sex and education levels. As illustrated in Figure 3C, ROC curve analysis demonstrated that p-tau217 achieved an AUC of 0.83 (95% CI: 0.80–0.86) for identifying abnormal Aβ PET status in CU individuals. An optimal cutoff value of 0.37 pg/mL for p-tau217 was established, yielding 81.1% sensitivity and 72.5% specificity in predicting Aβ status (Figure 3D). As shown in Supplementary Figure 3, among other blood biomarkers, only p-tau 217/Aβ42 achieves AUC = 0.85 (comparable to p-tau 217 alone) in the CU population, while all other indicators yield AUC < 0.80.

The results indicate that p-tau217 demonstrates robust performance in predicting Aβ positivity among CU individuals.

Diagnostic threshold determination (dual cutoffs) of LiCA p-tau217 for Aβ PET positivity

Although LiCA-measured p-tau217 demonstrated significantly superior diagnostic performance compared to other LiCA-based biomarkers, it failed to achieve both sensitivity and specificity of ≥90% simultaneously using the single cutpoint method (Schindler et al., 2024). To enhance diagnostic precision and reduce false-positive outcomes relative to the single cutpoint method, we adopted a dual-threshold approach for the LiCA p-tau217 assay (Brum et al., 2023; Figdore et al., 2024a). Specifically, to achieve 90% sensitivity and specificity—thereby creating a practical clinical tool that minimizes the number of subjects falling into the intermediate “gray zone” (Sarto et al., 2023)—we established a lower cutoff value of p-tau217 ≤ 0.37 pg/mL and an upper cutoff value of p-tau217 > 0.47 pg/mL. Values below 0.37 pg/mL were deemed to have a high probability of being Aβ−, while those above 0.47 pg/mL were highly likely to be Aβ+. Figure 4A illustrates the distribution of LiCA p-tau217 stratified by Aβ status, with positive, negative, and intermediate “gray zones” demarcated by assay cutpoints corresponding to the 90% sensitivity and 90% specificity thresholds. Using this dual-threshold approach, 51% of all individuals had p-tau217 levels below the 0.37 pg/mL threshold, 33% had levels above the 0.47 pg/mL cutoff, and 16% fell into the intermediate range, requiring additional Aβ testing (CSF/Aβ PET) to confirm the presence of Aβ pathology (Figure 4B). In summary, by applying lower and upper cutoffs for plasma p-tau217, the Aβ status in 84% of all subjects could be predicted with reasonably high accuracy, potentially avoiding the need for more advanced biomarker testing (CSF/Aβ PET) in these cases.

FIGURE 4

Figure 4. (A) LiCA p-tau217 assay performance for prediction of Aβ PET positivity at 90% sensitivity and 90% specificity cut-points in the whole cohort. (B) The clinical interpretation thresholds (dual cutoffs) for p-tau217 levels in the identification of abnormal Aβ PET status in the whole cohort. Aβ–, amyloid-β PET-negative; Aβ+, amyloid-β PET-positive; PET, positron emission tomography; p-tau, phosphorylated tau. LiCA represents the Chemclin LiCA^® kits.

Additionally, p-tau217 (dual cutoffs) measured by LiCA was found to perform well in predicting Aβ positivity across CU, MCI and AD participants, with gray-zone rates of 16.8%, 17.2% and 8.5% (Supplementary Table 2).

Discussion

Alzheimer’s disease is a progressive neurodegenerative disorder that not only profoundly impacts patients and their families, but also imposes a significant socioeconomic burden on healthcare systems worldwide. China is rapidly entering an aging society, with the prevalence of AD on the rise. Early diagnosis by BBBs biomarker have thus become crucial strategies to mitigate its socioeconomic impact. This study validates the robustness of BBBs for predicting Aβ positivity in Chinese populations and confirms that p-tau217 is a promising blood biomarker to reduce diagnostic costs and enhance screening efficiency of AD.

Previous studies utilizing Simoa platform have established the diagnostic accuracy of BBBs for detecting Aβ pathology, with reported AUC ranges as follows: Aβ42/40 (0.69–0.78) (Janelidze et al., 2021), p-tau181 (0.76–0.88) (Karikari et al., 2020), p-tau217 (0.92–0.93) (Ashton et al., 2024), GFAP (0.80–0.82) (Arslan et al., 2025; Yang et al., 2023), and NfL (0.65–0.66) (Arslan et al., 2025; Yang et al., 2023). We performed a robustness comparison of blood biomarkers evaluated by LiCA and Simoa assays for predicting Aβ status in the Chinese population. Our findings indicated that plasma p-tau217, GFAP, p-tau181, and NfL detected by both Simoa and LiCA assays exhibited comparable efficiency in the entire AD spectrum in identifying individuals with Aβ PET positivity. Notably, the AUC for LiCA Aβ42/40 was significantly higher than that obtained from two central laboratories using different Simoa assay kits. This improved performance may be attributed to the LiCA platform’s streamlined workflow, which eliminates the need for sample dilution, thereby reducing potential adsorption or degradation of Aβ42 and Aβ40 during processing. The results revealed that plasma p-tau217, GFAP, p-tau181, Aβ42/40, and NfL detected by different platforms achieved robustness in predicting Aβ positivity within the clinical AD continuum. Among all BBBs analyzed by LiCA, p-tau217 emerged as the most accurate, with an optimal single cutoff value of 0.47 pg/mL yielding 85.8% sensitivity and 89.6% specificity in predicting Aβ status. This was followed by GFAP, p-tau181, and Aβ42/40, while NfL showed slightly inferior performance. Remarkably, p-tau217 achieved an AUC of 0.83 for identifying abnormal Aβ PET status in CU individuals, underscoring its potential as a reliable biomarker for the early detection of Aβ pathology during preclinical stages, even among CU individuals. Furthermore, the application of a dual-threshold combination for p-tau217 using LiCA revealed that only 16% of samples fell into the intermediate range, necessitating further Aβ PET testing to confirm Aβ pathology. This finding positions p-tau217 as an excellent candidate for distinguishing Aβ+ from Aβ− individuals as a single biomarker.

In our study, BBBs measured using Simoa exhibited performance in predicting Aβ PET positivity that was largely consistent with findings reported in the existing literature. Additionally, BBBs detected by different platforms (Simoa and LiCA) exhibited robustness in predicting Aβ positivity within two subcohorts comprising a total of 766 attendees, particularly highlighting p-tau217 as a robust discriminator of Aβ pathology (AUC in subcohort 2: LiCA 0.95 vs. Simoa 0.94). Furthermore, in the whole cohort of 1254 attendees and CU individuals, the pivotal role and robustness of p-tau217 measured by LiCA in detecting Aβ deposition was further confirmed. In addition, we validated the robustness on the performance of p-tau217 in predicting Aβ positivity in CU individuals. This real-world clinical setting enhances the translational potential of our findings by reflecting the actual diagnostic challenges encountered in routine practice.

In the present study, the AUC of the standalone biomarker p-tau217 on the LiCA platform was 0.92, which is consistent with findings from a contemporary Chinese cohort (Simoa-ALZpath p-tau217 AUC = 0.91) (Lin et al., 2025). We subsequently computed the p-tau217/Aβ42 ratio (N = 1,254), and its AUC was 0.91—slightly inferior to that of the standalone p-tau217 assay. This aligns with the findings of a 2025 meta-analysis published in Lancet Neurology: among 11 head-to-head comparison studies, the pooled AUC difference between standalone p-tau217 and the p-tau217/Aβ42 ratio was <0.02 (I² = 0%). Thus, utilization of standalone p-tau217 in Chinese populations achieves discriminative efficacy comparable to that of the ratio-based model, while obviating the additional variability and costs inherent to Aβ42 quantification.

The dual cutoffs established for LiCA p-tau217 in this study (≤0.37 pg/mL for Aβ negativity, >0.47 pg/mL for Aβ positivity) categorized 16% of participants into the gray zone, reflecting an absolute 3% reduction relative to recent multicenter Simoa-ALZ path data (Lan et al., 2025), which reported a 19% gray zone. Currently, PET scans incur a relatively high cost. The application of the dual-threshold strategy enables 84% of subjects to forgo PET scanning, thereby achieving substantial cost savings. While Aβ PET remains the gold standard for detecting Aβ deposition, our preliminary research revealed that only 33.3% of Chinese cognitive clinic patients opt for this examination (Cui et al., 2023), with even greater implementation barriers in cognitively normal populations. P-tau217 holds substantial clinical promise by facilitating the early identification of at-risk subjects during the initial stages of pathological progression, thereby creating critical therapeutic windows for the timely implementation of disease-modifying interventions. Furthermore, plasma p-tau217 measured by LiCA could potentially serve as inclusion criteria (including enrichment and stratification) or be used to evaluate treatment efficacy in clinical trials for disease-modifying therapies (Teunissen et al., 2022). A compelling illustration of this biomarker-driven approach is demonstrated in the TRAILBLAZER-ALZ 2 trial (Donanemab, Eli Lilly), where p-tau217 was strategically integrated into the trial framework (Sims et al., 2023). This biomarker-stratified detection strategy not only offers significant economic advantages but also addresses the practical challenges currently impeding early screening practices for AD.

Our study also has several limitations. Firstly, although the participants represented a diverse patient population, the single-center design of this study, conducted in China, resulted in a racially homogeneous cohort. This limits the generalizability of our findings to other racial and ethnic groups. Secondly, the accuracy and true predictive power of the cutoffs derived from our findings need to be validated in diverse datasets with similar characteristics to ensure broader applicability and generalizability. Thirdly, the lack of longitudinal follow-up restricts the value of blood biomarkers in predicting disease progression. These limitations highlight the need for further multicenter studies involving diverse populations to strengthen the validity and generalizability of our conclusions.

Conclusion

Our study demonstrates that, among the Chinese population, BBBs exhibit strong robustness in predicting Aβ positivity. Notably, plasma p-tau217 stands out as the most promising and superior biomarker for detecting AD pathology. Utilizing BBBs to determine Aβ positivity can enable more precise identification of early-stage patients, which holds significant importance for effectively addressing the growing burden of AD.

Data availability statement

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Ethics statement

The studies involving humans were approved by the Ethical Committee for Medical Research at the Shanghai Sixth People’s Hospital Affiliated to Shanghai Jiao Tong University School of Medicine (approval number 2019-041). The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.

Author contributions

YW: Funding acquisition, Writing – original draft, Data curation, Formal analysis, Methodology. LC: Writing – review & editing, Methodology, Software, Funding acquisition, Conceptualization. F-FP: Data curation, Formal analysis, Writing – review & editing. ZZ: Writing – review & editing, Methodology, Data curation. LH: Data curation, Methodology, Writing – review & editing. YM: Formal analysis, Data curation, Funding acquisition, Writing – review & editing. Y-HG: Supervision, Writing – review & editing. Q-HG: Funding acquisition, Conceptualization, Supervision, Writing – review & editing.

Funding

The author(s) declare financial support was received for the research and/or publication of this article. This study was funded by the STI2030-Major Projects (2022ZD0213800), the National Natural Science Foundation of China (82171198), the Shanghai Municipal Science and Technology Major Project (2018SHZDZX01), the Shanghai Municipal Health Commission (202140042, 202340173), and Scientific Research Project of Shanghai Municipal Health Commission (202340080).

Acknowledgments

We would like to thank Jie-Hua Zhu, Xiang-Qing Xie, and Yun Yang for their help with neuropsychological tests.

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The authors declare that no Generative AI was used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fnagi.2025.1660755/full#supplementary-material

References

Arslan, B., Brum, W. S., Pola, I., Therriault, J., Rahmouni, N., Stevenson, J., et al. (2025). The impact of kidney function on Alzheimer’s disease blood biomarkers: Implications for predicting amyloid-β positivity. Alzheimers Res. Ther. 17:48. doi: 10.1186/s13195-025-01692-z

PubMed Abstract | Crossref Full Text | Google Scholar

Ashton, N. J., Brum, W. S., Di Molfetta, G., Benedet, A. L., Arslan, B., Jonaitis, E., et al. (2024). Diagnostic accuracy of a plasma phosphorylated tau 217 immunoassay for Alzheimer disease pathology. JAMA Neurol. 81, 255–263. doi: 10.1001/jamaneurol.2023.5319

PubMed Abstract | Crossref Full Text | Google Scholar

Brum, W. S., Cullen, N. C., Janelidze, S., Ashton, N. J., Zimmer, E. R., Therriault, J., et al. (2023). A two-step workflow based on plasma p-tau217 to screen for amyloid β positivity with further confirmatory testing only in uncertain cases. Nat. Aging 3, 1079–1090. doi: 10.1038/s43587-023-00471-5

PubMed Abstract | Crossref Full Text | Google Scholar

Cui, L., Huang, L., Pan, F. F., Wang, Y., Huang, Q., Guan, Y. H., et al. (2023). Chinese preclinical Alzheimer’s disease study (C-PAS): Design and challenge from PET acceptance. J. Prevent. Alzheimer’s Dis. 10, 571–580. doi: 10.14283/jpad.2023.49

PubMed Abstract | Crossref Full Text | Google Scholar

Fargo, K. N., Carrillo, M. C., Weiner, M. W., Potter, W. Z., and Khachaturian, Z. (2016). The crisis in recruitment for clinical trials in Alzheimer’s and dementia: An action plan for solutions. Alzheimers Dement. 12, 1113–1115. doi: 10.1016/j.jalz.2016.10.001

PubMed Abstract | Crossref Full Text | Google Scholar

Figdore, D. J., Griswold, M., Bornhorst, J. A., Graff-Radford, J., Ramanan, V. K., Vemuri, P., et al. (2024a). Optimizing cutpoints for clinical interpretation of brain amyloid status using plasma p-tau217 immunoassays. Alzheimer’s Dement. 20, 6506–6516. doi: 10.1002/alz.14140

PubMed Abstract | Crossref Full Text | Google Scholar

Figdore, D. J., Wiste, H. J., Bornhorst, J. A., Bateman, R. J., Li, Y., Graff-Radford, J., et al. (2024b). Performance of the Lumipulse plasma Aβ42/40 and pTau181 immunoassays in the detection of amyloid pathology. Alzheimer’s Dement. Diag. Assess. Dis. Monit. 16:e12545. doi: 10.1002/dad2.12545

PubMed Abstract | Crossref Full Text | Google Scholar

He, K., Li, B., Huang, L., Zhao, J., Hua, F., Wang, T., et al. (2024). Positive rate and quantification of amyloid pathology with [(18)F]florbetapir in the urban Chinese population. Eur. Radiol. 34, 3331–3341. doi: 10.1007/s00330-023-10366-z

PubMed Abstract | Crossref Full Text | Google Scholar

Jack, C. R. Jr., Andrews, J. S., Beach, T. G., Buracchio, T., Dunn, B., Graf, A., et al. (2024). Revised criteria for diagnosis and staging of Alzheimer’s disease: Alzheimer’s association workgroup. Alzheimers Dement. 20, 5143–5169. doi: 10.1002/alz.13859

PubMed Abstract | Crossref Full Text | Google Scholar

Jack, C. R. Jr., Bennett, D. A., Blennow, K., Carrillo, M. C., Dunn, B., Haeberlein, S. B., et al. (2018). NIA-AA research framework: Toward a biological definition of Alzheimer’s disease. Alzheimers Dement. 14, 535–562. doi: 10.1016/j.jalz.2018.02.018

PubMed Abstract | Crossref Full Text | Google Scholar

Janelidze, S., Bali, D., Ashton, N. J., Barthélemy, N. R., Vanbrabant, J., Stoops, E., et al. (2023). Head-to-head comparison of 10 plasma phospho-tau assays in prodromal Alzheimer’s disease. Brain: A J. Neurol. 146, 1592–1601. doi: 10.1093/brain/awac333

PubMed Abstract | Crossref Full Text | Google Scholar

Janelidze, S., Teunissen, C. E., Zetterberg, H., Allué, J. A., Sarasa, L., Eichenlaub, U., et al. (2021). Head-to-Head comparison of 8 plasma Amyloid-β 42/40 assays in Alzheimer disease. JAMA Neurol. 78, 1375–1382. doi: 10.1001/jamaneurol.2021.3180

PubMed Abstract | Crossref Full Text | Google Scholar

Jia, L., Quan, M., Fu, Y., Zhao, T., Li, Y., Wei, C., et al. (2020). Dementia in China: Epidemiology, clinical management, and research advances. Lancet Neurol. 19, 81–92. doi: 10.1016/s1474-4422(19)30290-x

PubMed Abstract | Crossref Full Text | Google Scholar

Kalaria, R., Maestre, G., Mahinrad, S., Acosta, D. M., Akinyemi, R. O., Alladi, S., et al. (2024). The 2022 symposium on dementia and brain aging in low- and middle-income countries: Highlights on research, diagnosis, care, and impact. Alzheimers Dement. 20, 4290–4314. doi: 10.1002/alz.13836

PubMed Abstract | Crossref Full Text | Google Scholar

Karikari, T. K., Pascoal, T. A., Ashton, N. J., Janelidze, S., Benedet, A. L., Rodriguez, J. L., et al. (2020). Blood phosphorylated tau 181 as a biomarker for Alzheimer’s disease: A diagnostic performance and prediction modelling study using data from four prospective cohorts. Lancet Neurol. 19, 422–433. doi: 10.1016/s1474-4422(20)30071-5

PubMed Abstract | Crossref Full Text | Google Scholar

Lan, G., Wang, M., Gonzalez-Ortiz, F., Zhang, L., Li, A., Li, B., et al. (2025). Comprehensive evaluation of plasma tau biomarkers for detecting and monitoring Alzheimer’s disease in a multicenter and multiethnic aging population. Nat. Aging 5, 1601–1618. doi: 10.1038/s43587-025-00904-3

PubMed Abstract | Crossref Full Text | Google Scholar

Leuzy, A., Mattsson-Carlgren, N., Palmqvist, S., Janelidze, S., Dage, J. L., and Hansson, O. (2022). Blood-based biomarkers for Alzheimer’s disease. EMBO Mol. Med. 14:e14408. doi: 10.15252/emmm.202114408

PubMed Abstract | Crossref Full Text | Google Scholar

Lin, Y. S., Kwon, H. S., Lee, W. J., Hwang, M., Jeong, J. H., Koh, S. H., et al. (2025). Cross-cultural validation of plasma p-tau217 and p-tau181 as precision biomarkers for amyloid PET positivity: An East Asian study in Taiwan and Korea. Alzheimers Dement. 21:e14565. doi: 10.1002/alz.14565

PubMed Abstract | Crossref Full Text | Google Scholar

McKhann, G. M., Knopman, D. S., Chertkow, H., Hyman, B. T., Jack, C. R. Jr., Kawas, C. H., et al. (2011). The diagnosis of dementia due to Alzheimer’s disease: Recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement. 7, 263–269. doi: 10.1016/j.jalz.2011.03.005

PubMed Abstract | Crossref Full Text | Google Scholar

Pan, F., Huang, Y., Cai, X., Wang, Y., Guan, Y., Deng, J., et al. (2023). Integrated algorithm combining plasma biomarkers and cognitive assessments accurately predicts brain beta-amyloid pathology. Commun. Med. 3:65. doi: 10.1038/s43856-023-00295-9

PubMed Abstract | Crossref Full Text | Google Scholar

Pan, F. F., Huang, Q., Wang, Y., Wang, Y. F., Guan, Y. H., Xie, F., et al. (2022). Non-linear character of plasma amyloid beta over the course of cognitive decline in Alzheimer’s continuum. Front. Aging Neurosci. 14:832700. doi: 10.3389/fnagi.2022.832700

PubMed Abstract | Crossref Full Text | Google Scholar

Petersen, R. C. (2004). Mild cognitive impairment as a diagnostic entity. J. Intern. Med. 256, 183–194. doi: 10.1111/j.1365-2796.2004.01388.x

PubMed Abstract | Crossref Full Text | Google Scholar

Petersen, R. C., Smith, G. E., Waring, S. C., Ivnik, R. J., Kokmen, E., and Tangelos, E. G. (1997). Aging, memory, and mild cognitive impairment. Int. Psychogeriatr. 9 (Suppl. 1), 65–69. doi: 10.1017/s1041610297004717

PubMed Abstract | Crossref Full Text | Google Scholar

Ren, R., Qi, J., Lin, S., Liu, X., Yin, P., Wang, Z., et al. (2022). The China Alzheimer report 2022. Gen Psychiatr. 35:e100751. doi: 10.1136/gpsych-2022-100751

PubMed Abstract | Crossref Full Text | Google Scholar

Ren, S., Li, J., Huang, L., Huang, Q., Chen, K., Hu, J., et al. (2023). Brain functional alterations and association with cognition in people with preclinical subjective cognitive decline and objective subtle cognitive difficulties. Neuroscience 513, 137–144. doi: 10.1016/j.neuroscience.2023.01.004

PubMed Abstract | Crossref Full Text | Google Scholar

Rissin, D. M., Kan, C. W., Campbell, T. G., Howes, S. C., Fournier, D. R., Song, L., et al. (2010). Single-molecule enzyme-linked immunosorbent assay detects serum proteins at subfemtomolar concentrations. Nat. Biotechnol. 28, 595–599. doi: 10.1038/nbt.1641

PubMed Abstract | Crossref Full Text | Google Scholar

Sarto, J., Ruiz-García, R., Guillén, N., Ramos-Campoy, Ó,Falgàs, N., Esteller, D., et al. (2023). Diagnostic performance and clinical applicability of blood-based biomarkers in a prospective memory clinic cohort. Neurology 100, e860–e873. doi: 10.1212/wnl.0000000000201597

PubMed Abstract | Crossref Full Text | Google Scholar

Schindler, S. E., Galasko, D., Pereira, A. C., Rabinovici, G. D., Salloway, S., Suárez-Calvet, M., et al. (2024). Acceptable performance of blood biomarker tests of amyloid pathology — recommendations from the Global CEO Initiative on Alzheimer’s disease. Nat. Rev. Neurol. 20, 426–439. doi: 10.1038/s41582-024-00977-5

PubMed Abstract | Crossref Full Text | Google Scholar

Sims, J. R., Zimmer, J. A., Evans, C. D., Lu, M., Ardayfio, P., Sparks, J., et al. (2023). Donanemab in early symptomatic Alzheimer disease: The TRAILBLAZER-ALZ 2 randomized clinical trial. JAMA 330, 512–527. doi: 10.1001/jama.2023.13239

PubMed Abstract | Crossref Full Text | Google Scholar

Teunissen, C. E., Verberk, I. M. W., Thijssen, E. H., Vermunt, L., Hansson, O., Zetterberg, H., et al. (2022). Blood-based biomarkers for Alzheimer’s disease: Towards clinical implementation. Lancet Neurol. 21, 66–77. doi: 10.1016/s1474-4422(21)00361-6

PubMed Abstract | Crossref Full Text | Google Scholar

Ullman, E. F., Kirakossian, H., Switchenko, A. C., Ishkanian, J., Ericson, M., Wartchow, C. A., et al. (1996). Luminescent oxygen channeling assay (LOCI): Sensitive, broadly applicable homogeneous immunoassay method. Clin. Chem. 42, 1518–1526. doi: 10.1093/clinchem/42.9.1518

Crossref Full Text | Google Scholar

Wang, M., Li, J., Huang, Y., Chen, T., Dong, S., Zhang, R., et al. (2022). Analytical validation of the LiCA^® high-sensitivity human thyroid stimulating hormone assay. Clin. Biochem. 101, 42–49. doi: 10.1016/j.clinbiochem.2021.11.018

PubMed Abstract | Crossref Full Text | Google Scholar

Warmenhoven, N., Salvadó, G., Janelidze, S., Mattsson-Carlgren, N., Bali, D., Orduña Dolado, A., et al. (2025). A comprehensive head-to-head comparison of key plasma phosphorylated tau 217 biomarker tests. Brain: A J. Neurol. 148, 416–431. doi: 10.1093/brain/awae346

PubMed Abstract | Crossref Full Text | Google Scholar

Wu, J., Li, B., Wang, J., Huang, Q., Chen, X., You, Z., et al. (2024). Plasma glial fibrillary acid protein and phosphorated tau 181 association with presynaptic density-dependent tau pathology at 18F-SynVesT-1 brain PET imaging. Radiology 313:e233019. doi: 10.1148/radiol.233019

PubMed Abstract | Crossref Full Text | Google Scholar

Yang, K., Yang, X., Yin, P., Zhou, M., and Tang, Y. (2024). Temporal trend and attributable risk factors of Alzheimer’s disease and other dementias burden in China: Findings from the Global Burden of Disease Study 2021. Alzheimers Dement. 20, 7871–7884. doi: 10.1002/alz.14254

PubMed Abstract | Crossref Full Text | Google Scholar

Yang, Z., Sreenivasan, K., Toledano Strom, E. N., Osse, A. M. L., Pasia, L. G., Cosme, C. G., et al. (2023). Clinical and biological relevance of glial fibrillary acidic protein in Alzheimer’s disease. Alzheimers Res. Ther. 15:190. doi: 10.1186/s13195-023-01340-4

PubMed Abstract | Crossref Full Text | Google Scholar