The PrecivityAD2 Blood Test: A Simple Blood Draw That May Help Detect Alzheimer's Disease Brain Changes

Table of Contents

• Key Points
• Background: Why This Research Matters
• Study Methods: How the Research Was Conducted
• Key Findings: Detailed Results With All Numbers
• Clinical Implications: What This Means for Patients
• Limitations: What This Study Couldn't Prove
• Recommendations: What Patients Should Know
• Frequently Asked Questions
• Source Information

Key Points

• PrecivityAD2 measures %p-tau217 and the Aβ42/40 ratio to produce an Amyloid Probability Score 2.
• In 583 older adults with mild cognitive impairment or dementia, PrecivityAD2 agreed with amyloid PET scans 88% of the time.
• Blood sampling is more convenient, less costly, and less invasive than PET scans or spinal taps.
• PrecivityAD2 detects amyloid plaques, but Alzheimer's involves multiple processes, so it is part of a larger diagnostic picture.

Background: Why This Research Matters

• Amyloid PET (positron emission tomography) imaging—a brain scan that uses a radioactive tracer to visualize amyloid plaques
• Cerebrospinal fluid (CSF) analysis—a spinal tap procedure that measures amyloid and tau proteins in the fluid surrounding the brain

Study Methods: How the Research Was Conducted

Study Participants

Researchers analyzed blood samples from a total of 583 individuals with suspected Alzheimer's disease. The study combined participants from two different research studies: the PARIS study (224 participants) and the MissionAD study (359 participants). All participants were symptomatic, meaning they had either mild cognitive impairment (MCI) or dementia at the time of enrollment.

The PARIS study was a substudy of the larger IDEAS study (Imaging Dementia—Evidence for Amyloid Scanning), which evaluated the clinical utility of amyloid PET scans in Medicare beneficiaries with MCI or dementia meeting Appropriate Use Criteria for amyloid PET. Participants enrolled in PARIS between October 1, 2018, and January 4, 2019, and they provided blood samples that were processed and frozen for later analysis.

The MissionAD program consisted of two global phase 3 clinical trials that tested the safety and efficacy of elenbecestat, a BACE inhibitor drug, in participants with MCI due to Alzheimer's or mild Alzheimer's dementia with amyloid positivity confirmed by PET visual read or CSF test. A total of 359 baseline screening plasma samples were selected from this study, resulting in approximately 50% positive prevalence of amyloid PET results.

All participants (or their legally authorized representatives) reviewed and signed an approved informed consent document. The study followed the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) reporting guidelines. The research received ethical approval from central or local ethics and safety review committees.

How Amyloid PET Images Were Analyzed

Three different amyloid PET tracers were used in the PARIS Discovery cohort: [18F]florbetapir (Amyvid), [18F]florbetaben (Neuraceq), and [18F]flutemetamol (Vizamyl). In the IDEAS study, PET scans were originally interpreted visually by local radiologists and nuclear medicine physicians. However, for the PARIS study specifically, all PET images were obtained and processed centrally by the American College of Radiology, with two board-certified radiologists specially trained for each amyloid PET tracer.

Imaging results were quantified using two standard measures: the standardized uptake value ratio (SUVR) and the Centiloid (CL) scale. For the MissionAD study, two tracers were used—florbetapir and florbetaben—and images were processed centrally by Bioclinica. The researchers defined amyloid positivity as a Centiloid value greater than 25, which is a more sensitive threshold for detecting brain amyloid plaques than visual read methods.

How Blood Samples Were Analyzed

The researchers used a highly sophisticated laboratory technique called liquid chromatography-mass spectrometry (LC-MS/MS) to measure the blood biomarkers. This technology is capable of detecting extremely small amounts of proteins in blood samples with high precision.

For the Aβ42/40 ratio determination: On the day of analysis, plasma samples were thawed along with quality control samples and frozen calibrators. Each 450 μL sample was combined with an immunoprecipitation buffer containing known amounts of isotopically labeled (15N) Aβ40 and Aβ42 internal standards. After immunocapture, the amyloid proteins bound to magnetic beads were digested using an enzyme called Lys-N metalloendoprotease. The digested peptides (Aβ28-40 and Aβ28-42) were purified and analyzed by LC-MS/MS. The concentrations were calculated by comparing the peak areas of the natural (14N) peptides to the added internal standards. Concentrations were expressed as the plasma Aβ42/40 ratio.

For the p-tau217 and np-tau217 determination: A similar immunocapture approach was used, this time using an anti-tau antibody attached to magnetic beads. After washing, tau proteins were digested with an enzyme called trypsin, which cut the tau protein into specific peptides. The digested samples were purified using solid phase extraction and analyzed by LC-MS/MS. The researchers measured both phosphorylated tau at threonine 217 (p-tau217) and non-phosphorylated tau at threonine 217 (np-tau217). The percentage of phosphorylated tau was calculated using the formula: %p-tau217 = p-tau217 / np-tau217 × 100%.

Quality control was rigorous throughout. Each 96-well plate included a six-point calibration curve, six quality control samples, and participant samples. Runs were rejected if quality control concentrations fell outside strict statistical limits. The total allowable error for np-tau217 was set at 30% or 15 pg/mL (whichever was greater), and for p-tau217 it was set at 30% or 0.3 pg/mL.

Statistical Analysis

All data analysis was performed using R version 4.3.1 statistical software. The researchers used receiver operating characteristic (ROC) analyses to evaluate how well the biomarkers distinguished between people with and without brain amyloid. The area under the ROC curve (AUC-ROC) is a measure of diagnostic accuracy, where 0.5 means the test is no better than chance and 1.0 means perfect accuracy. Optimal cutoff values were determined using the Youden index, which maximizes both sensitivity and specificity simultaneously. Confidence intervals (95% CI) were calculated using the DeLong method, and logistic regression models were used to predict amyloid positivity. Model robustness was compared using the Akaike information criterion (AIC), Bayesian information criterion (BIC), and likelihood ratio tests. Comparisons of accuracy and participant distributions were performed using Fisher's exact test, and 95% confidence intervals on accuracy measures were calculated using Wilson's method.

Key Findings: Detailed Results With All Numbers

Who Was in the Study?

The combined study group of 583 participants had a mean age of 72.6 years (standard deviation [SD] 6.9), with ages ranging from 55 to 91 years. Women made up 48.7% of the participants (284 of 583). The majority of participants were White (91.3%), and 30.4% were of Hispanic or Latino origin. Notably, almost 50% of participants from the MissionAD study were of Hispanic or Latino origin, contributing to this significant representation.

In terms of cognitive status, 81.6% of participants had a diagnosis of mild cognitive impairment (MCI), and 18.4% had dementia. The mean Mini-Mental State Examination (MMSE) score was 26.1 (SD 3.2), indicating mild cognitive difficulties overall. About 41.5% of participants carried one or more copies of the apolipoprotein E ε4 (apoE4) gene variant, which is a known genetic risk factor for Alzheimer's disease. Of the 583 participants, 52.7% (307 people) were positive for brain amyloid as defined by a Centiloid score greater than 25.

Participants were divided into two groups based on their amyloid status:

• Amyloid-negative group (CL less than 25): 276 participants
• Amyloid-positive group (CL greater than 25): 307 participants

Several notable differences emerged between these two groups. The amyloid-positive group was older on average (74.3 years vs. 70.7 years), had a higher proportion of participants with dementia (25.7% vs. 10.1%), and had a much higher rate of apoE4 carrier status (60.5% vs. 20.4%). The amyloid-positive group also had significantly different blood biomarker levels, which is exactly what the test is designed to detect.

Blood Biomarker Results

When the researchers measured amyloid proteins in the blood, they found important differences between the two groups:

• Aβ40 levels were not significantly different between groups (amyloid-negative: 483 pg/mL, SD 101; amyloid-positive: 487 pg/mL, SD 93.2; p = 0.59). This makes sense because Aβ40 is the more common form of amyloid and is not specifically related to Alzheimer's pathology.
• Aβ42 levels were significantly lower in the amyloid-positive group (48 pg/mL vs. 43 pg/mL, p < 0.0001). This reflects the fact that Aβ42 tends to be deposited in the brain, reducing its levels in the blood.
• The Aβ42/40 ratio was 0.12-fold lower in amyloid-positive participants compared to amyloid-negative participants. The mean Aβ42/40 ratio was 0.1001 (SD 0.0150) in the amyloid-negative group and 0.0878 (SD 0.0101) in the amyloid-positive group.
• The diagnostic accuracy of the Aβ42/40 ratio (AUC-ROC = 0.75, 95% CI: 0.71 to 0.79) was significantly better than that of Aβ42 alone (AUC-ROC = 0.64, 95% CI: 0.60 to 0.69), with p < 0.001 by DeLong comparison of ROC curves. The improved performance of the ratio likely comes from normalizing the Aβ42 concentration to the overall amount of amyloid in the sample.

For the tau protein measurements, the results were even more striking:

• The mean %p-tau217 was 3.02% (SD 3.10%) in amyloid-negative participants versus 11.14% (SD 6.42%) in amyloid-positive participants—a highly significant difference.
• The mean p-tau217 concentration was 1.57 pg/mL (SD 1.90) in amyloid-negative participants versus 6.24 pg/mL (SD 4.31) in amyloid-positive participants.
• The mean np-tau217 concentration was 50.63 pg/mL (SD 23.83) in amyloid-negative versus 55.39 pg/mL (SD 19.54) in amyloid-positive participants, a smaller difference.
• The optimal cutoff for %p-tau217 was determined to be 4.2% using the Youden index.

Handling Very Low Measurements: Value Imputation

Some participants had p-tau217 concentrations so low that they fell below the limit of detection (LOD) of the mass spectrometer, which was 1.3 pg/mL. In the amyloid-negative group, 149 of 276 participants (54%) had p-tau217 below this threshold, compared to only 11 of 307 amyloid-positive participants (3.6%). For these samples, the concentration was imputed as half the LOD (0.65 pg/mL), and the %p-tau217 was calculated based on this value. The researchers considered this approach important because providing an actionable %p-tau217 result is more useful to doctors than reporting "undetectable" for patients whose p-tau217 is very low but confirmed present.

Head-to-Head Comparison: %p-tau217 vs. p-tau217 Concentration

One of the key questions this study addressed was whether the percentage of phosphorylated tau is a better diagnostic marker than the absolute concentration of phosphorylated tau. The researchers found a statistically significant advantage for the percentage measure.

The AUC-ROC for %p-tau217 was 0.94, which was statistically significantly higher than the AUC-ROC for p-tau217 concentration at 0.91. This means that normalizing the amount of phosphorylated tau to the total amount of tau available to be phosphorylated provides better diagnostic information. The researchers noted that two people with the same absolute p-tau217 concentration could have very different percentages if they had different total tau levels—and the percentage appears to better reflect what is actually happening in the brain.

The PrecivityAD2 Test and the Amyloid Probability Score 2 (APS2)

The researchers then developed a predictive algorithm called the PrecivityAD2 test that combines two blood measurements: %p-tau217 and the Aβ42/40 ratio. This algorithm produces a score called the Amyloid Probability Score 2 (APS2), which ranges from 0 to 100 (the underlying model generates a likelihood between 0 and 1, which is then multiplied by 100 and rounded).

In the logistic regression model using %p-tau217 alone, this biomarker was highly significant (p < 0.0001). A 3 percentage point increase in %p-tau217—roughly one standard deviation in the amyloid-negative group—was associated with an odds ratio of 5.4 (95% CI: 4.1 to 7.2) for identifying brain amyloidosis. In plain terms, this means that people with a 3-point higher %p-tau217 were about 5.4 times more likely to have amyloid plaques in their brains.

When both %p-tau217 and the Aβ42/40 ratio were combined in the PrecivityAD2 model, both analytes were highly significant contributors (p < 0.0001 for each). The odds ratio for a 3 percentage point increase in %p-tau217 in this combined model was 4.6 (95% CI: 3.5 to 6.0), and a 0.015 decrease in the Aβ42/40 ratio (approximately one standard deviation) was associated with an odds ratio of 2.1 (95% CI: 1.5 to 2.9).

The overall diagnostic performance of the PrecivityAD2 test was excellent:

• AUC-ROC for APS2: 0.94
• Agreement with amyloid PET: 88%

Importantly, diagnostic performance was consistent across different demographic groups. The PrecivityAD2 test performed similarly regardless of ethnicity, sex, age, and apoE4 status, which suggests it can be used broadly across diverse patient populations.

Clinical Implications: What This Means for Patients

This study has significant implications for the diagnosis and management of Alzheimer's disease. For patients experiencing memory problems, the path to diagnosis currently often involves expensive PET scans or invasive spinal taps. A simple blood test that can accurately detect brain amyloid—with 88% agreement with PET scans—could transform the diagnostic process.

Here is what this finding means in practical terms:

• Earlier and easier diagnosis: A routine blood draw at a standard clinic could provide important information about whether Alzheimer's pathology is present, potentially earlier in the disease course.
• Treatment eligibility: With the approval of lecanemab and other amyloid-targeting therapies, confirming the presence of amyloid pathology is now a prerequisite for treatment. The PrecivityAD2 test could help identify which patients are appropriate candidates for these medications.
• Reduced cost: Blood tests are far less expensive than PET scans, which cost approximately $6,500 for research use. This could make amyloid testing more accessible to a broader population, including those in rural or underserved areas.
• Clinical trial enrollment: The test could also help expedite enrollment of participants with confirmed amyloid pathology in clinical trials of new Alzheimer's treatments.
• Patient stratification: The ability to distinguish between patients with and without Alzheimer's pathology is crucial because conditions with similar symptoms (such as vascular dementia or frontotemporal dementia) require different management approaches.

The study authors noted that this test represents a significant advance over measuring p-tau217 concentration alone. The finding that %p-tau217 outperforms the raw concentration suggests that individual differences in total tau levels matter—two people with the same concentration of phosphorylated tau may have very different disease states if their total tau differs. This normalization makes the test more robust across diverse patient populations.

Limitations: What This Study Couldn't Prove

While these results are encouraging, the researchers acknowledged several important limitations that patients should understand:

• Study population characteristics: The participants in this study were predominantly White (over 90%), although nearly one-third were of Hispanic or Latino origin. However, the relatively small numbers of Black or African American participants (5.5%) and Asian participants (0.5%) mean that additional validation in more diverse populations is needed. Notably, the researchers reported that diagnostic performance was similar across ethnicity, sex, age, and apoE4 status within this study.
• The gold standard itself is imperfect: The study used amyloid PET as the reference standard, but PET scans are not 100% accurate either. The Centiloid > 25 threshold was chosen based on published evidence, but different thresholds could yield slightly different agreement rates.
• Correlation with previous sample analysis: For the PARIS participants, the researchers had to re-analyze blood samples for Aβ concentrations because they wanted both Aβ and p-tau217 measured at the same time. While the correlation between the two analyses was strong (Pearson correlation coefficients of 0.87 for Aβ40, 0.86 for Aβ42, and 0.62 for the Aβ42/40 ratio), the need to use previously frozen samples could introduce some variability.
• Overlap with previous study: 90 of the 359 MissionAD participants (25%) had also been part of the previous analysis for the original PrecivityAD test validation, which means there was not complete independence from the earlier work.
• This is a diagnostic study, not a treatment study: The research validates that the blood test can identify brain amyloid, but it does not address whether using this test to guide treatment leads to better patient outcomes. That question requires separate long-term studies.

Recommendations: What Patients Should Know

Based on this research, here are practical takeaways for patients and their families:

1. Talk to your doctor about memory concerns early. The earlier Alzheimer's disease is identified, the more options may be available. If you or a loved one are experiencing memory problems, discuss them with a healthcare provider rather than waiting.
2. Understand that diagnosis is a multi-step process. A blood test like PrecivityAD2 provides important information, but it is not yet a substitute for a comprehensive evaluation. Doctors typically use a combination of cognitive testing, medical history, physical examination, and now potentially blood tests to make a diagnosis.
3. Ask about blood biomarker testing. If you are undergoing evaluation for memory concerns and are being considered for amyloid PET or CSF testing, ask your doctor whether blood-based biomarker testing might be appropriate as an initial screening step.
4. Know that treatment eligibility requires amyloid confirmation. If you are considering disease-modifying therapies like lecanemab, your doctor will need to confirm the presence of amyloid pathology before prescribing. The PrecivityAD2 test could potentially help with this confirmation, though the exact clinical workflow will depend on further regulatory decisions and practice guidelines.
5. Consider participating in research. Advancing our understanding of Alzheimer's disease requires ongoing clinical studies. The researchers specifically called for additional data from more diverse cohorts and from participants at different stages of the disease process.
6. Understand the gene connection. The study confirmed that apoE4 carriers are significantly more likely to have brain amyloid—41.5% of the combined cohort carried at least one copy of this gene variant, and 60.5% of amyloid-positive participants were carriers. Discuss your family history and potential genetic testing with your doctor to understand your personal risk.

It is also worth noting what this test does not do. The PrecivityAD2 test detects the presence of amyloid plaques, but Alzheimer's disease is complex and involves multiple pathological processes. A positive result on this test is an important piece of information, but it is part of a larger diagnostic picture—not a complete answer on its own.

Frequently Asked Questions

What is the PrecivityAD2 blood test?

PrecivityAD2 is a blood test that measures two Alzheimer's-related markers: %p-tau217 and the Aβ42/40 ratio. It combines these into a score called the Amyloid Probability Score 2 (APS2). In a study of 583 older adults with mild cognitive impairment or dementia, it agreed with PET brain scans 88% of the time for detecting amyloid plaques.

How accurate is this blood test compared to a PET scan?

In a study of 583 older adults with mild cognitive impairment or dementia, the PrecivityAD2 blood test agreed with amyloid PET scans in 88% of cases. Its diagnostic accuracy, measured by AUC-ROC, was 0.94. This suggests it may be a reliable option to detect Alzheimer's brain changes, though PET scans themselves are not perfect either.

Who might be eligible for this blood test?

The study included adults aged 55 to 91 who had mild cognitive impairment or dementia. The test may help doctors identify whether Alzheimer's amyloid plaques are present. However, diagnosis is a multi-step process, and this test is not yet a substitute for a full medical evaluation, which includes cognitive testing and history.

What are the limitations of the PrecivityAD2 test?

The study's participants were mostly White, so more diverse validation is needed. The test detects amyloid plaques, but Alzheimer's involves multiple processes. Also, the PET scan used as the reference standard is not 100% accurate. A positive result is one piece of information, not a complete diagnosis.

Should I ask my doctor about this blood test?

If you're being evaluated for memory concerns and considering amyloid PET or spinal fluid testing, ask if blood-based biomarker testing might be an initial screening step. The PrecivityAD2 test could help confirm amyloid presence, which is required for certain Alzheimer's treatments. Always discuss your situation with a healthcare provider.

Source Information

Original Article Title: Clinical validation of the PrecivityAD2 blood test

DOI: 10.1002/alz.13764

Authors: Matthew R. Meyer, Kristopher M. Kirmess, Stephanie Eastwood, Traci L. Wente-Roth, Faith Irvin, Mary S. Holubasch, Venky Venkatesh, Ilana Fogelman, Mark Monane, Lucy Hanna, Gil D. Rabinovici, Barry A. Siegel, Rachel A. Whitmer, Charles Apgar, Randall J. Bateman, David M. Holtzman, Michael Irizarry, David Verbel, Pallavi Sachdev, Satoshi Ito, John Contois, Kevin E. Yarasheski, Joel B. Braunstein, Philip B. Verghese, and Tim West. Matthew R. Meyer and Kristopher M. Kirmess contributed equally to this study.

Publication Details: This article was accepted on January 31, 2024, and published in Alzheimer's & Dementia (2024; Volume 20, pages 3179–3192). DOI: 10.1002/alz.13764. The research was supported by funding from the National Institutes of Health (Grant/Award Number: R44AG059489), BrightFocus, the Gerald and Henrietta Rauenhorst Foundation, and the Alzheimer's Drug Discovery Foundation.

Study Registration: The PARIS study was registered at ClinicalTrials.gov (Identifier: NCT02420756), and the MissionAD study was also registered at ClinicalTrials.gov (Identifier: NCT02956486).

This patient-friendly article is based on peer-reviewed research published in a scientific journal. It is intended for educational purposes and should not replace professional medical advice. Always consult your healthcare provider with questions about your personal health situation.