Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat

The DOI matches its Crossref record; 68 of the 68 references it cites were found in a registry or an index.

A two-step workflow based on plasma p-tau217 to screen for amyloid β positivity with further confirmatory testing only in uncertain cases

Open access

Abstract

Cost-effective strategies for identifying amyloid-β (Aβ) positivity in patients with cognitive impairment are urgently needed with recent approvals of anti-Aβ immunotherapies for Alzheimer’s disease (AD). Blood biomarkers can accurately detect AD pathology, but it is unclear whether their incorporation into a full diagnostic workflow can reduce the number of confirmatory cerebrospinal fluid (CSF) or positron emission tomography (PET) tests needed while accurately classifying patients. We evaluated a two-step workflow for determining Aβ-PET status in patients with mild cognitive impairment (MCI) from two independent memory clinic-based cohorts (n = 348). A blood-based model including plasma tau protein 217 (p-tau217), age and APOE ε4 status was developed in BioFINDER-1 (area under the curve (AUC) = 89.3%) and validated in BioFINDER-2 (AUC = 94.3%). In step 1, the blood-based model was used to stratify the patients into low, intermediate or high risk of Aβ-PET positivity. In step 2, we assumed referral only of intermediate-risk patients to CSF Aβ42/Aβ40 testing, whereas step 1 alone determined Aβ-status for low- and high-risk groups. Depending on whether lenient, moderate or stringent thresholds were used in step 1, the two-step workflow overall accuracy for detecting Aβ-PET status was 88.2%, 90.5% and 92.0%, respectively, while reducing the number of necessary CSF tests by 85.9%, 72.7% and 61.2%, respectively. In secondary analyses, an adapted version of the BioFINDER-1 model led to successful validation of the two-step workflow with a different plasma p-tau217 immunoassay in patients with cognitive impairment from the TRIAD cohort (n = 84). In conclusion, using a plasma p-tau217-based model for risk stratification of patients with MCI can substantially reduce the need for confirmatory testing while accurately classifying patients, offering a cost-effective strategy to detect AD in memory clinic settings.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1038/s43587-023-00471-5

Cites

68 of 68 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig. 1 Development and validation of a two-step workflow for Aβ-PET status capable of reducing further confirmatory tests while accurately classifying patients. a, Distribution of predicted probabilities of Aβ-PET positivity based on a logistic regression model including plasma p-tau217,age and APOE ε4 status as predictors. The predicted probabilities are displayed for the BioFINDER-1 (model training; left), BioFINDER-2 (model validation; middle) and both combined cohorts (right), with blue dots corresponding to individuals who are Aβ-PET negative and red dots to individuals who are Aβ-PET positive. On the right y axis, the probability values corresponding to the evaluated risk thresholds are demonstrated and accompanied by the metric used to define them (90%, 95%, 97.5% sensitivity or 90%, 95%, 97.5% specificity). The lower dashed line demonstrates where the 95% sensitivity low-risk threshold falls on the probability distribution, with the upper line corresponding to the 95% specificity high-risk threshold. b, Flowchart recapitulating results from the first step of the workflow (blood-biomarker-based risk stratification) and demonstrating the accuracy for the second step of the clinical workflow, when intermediate-risk individuals are referred to lumbar puncture (LP) to perform a CSF Aβ42/Aβ40 test for predicting Aβ-PET status based on the 95% Se/Sp strategy, with the flowchart for the two other strategies presented in Supplementary Information. c, The accuracy of both workflow steps combined, corresponding to the proportion of correct classifications for the low- and high-risk groups, along with the proportion of correct CSF Aβ42/Aβ40 classifications in the intermediate-risk group, according to each of the strategies, computed in the BioFINDER-1 and BioFINDER-2 MCI combined populations (n = 348). The error bars correspond to 95% CIs. d, Dots and lines indicating the observed percentage of reduction in further tests (here using CSF Aβ42/Aβ40) by applying the blood-based risk stratification strategy, based on each of the risk threshold strategies (90% Se/Sp, n = 301; 95% Se/Sp, n = 247; 97.5% Se/Sp, n = 205).
Fig. 1 Development and validation of a two-step workflow for Aβ-PET status capable of reducing further confirmatory tests while accurately classifying patients. a, Distribution of predicted probabilities of Aβ-PET positivity based on a logistic regression model including plasma p-tau217,age and APOE ε4 status as predictors. The predicted probabilities are displayed for the BioFINDER-1 (model training; left), BioFINDER-2 (model validation; middle) and both combined cohorts (right), with blue dots corresponding to individuals who are Aβ-PET negative and red dots to individuals who are Aβ-PET positive. On the right y axis, the probability values corresponding to the evaluated risk thresholds are demonstrated and accompanied by the metric used to define them (90%, 95%, 97.5% sensitivity or 90%, 95%, 97.5% specificity). The lower dashed line demonstrates where the 95% sensitivity low-risk threshold falls on the probability distribution, with the upper line corresponding to the 95% specificity high-risk threshold. b, Flowchart recapitulating results from the first step of the workflow (blood-biomarker-based risk stratification) and demonstrating the accuracy for the second step of the clinical workflow, when intermediate-risk individuals are referred to lumbar puncture (LP) to perform a CSF Aβ42/Aβ40 test for predicting Aβ-PET status based on the 95% Se/Sp strategy, with the flowchart for the two other strategies presented in Supplementary Information. c, The accuracy of both workflow steps combined, corresponding to the proportion of correct classifications for the low- and high-risk groups, along with the proportion of correct CSF Aβ42/Aβ40 classifications in the intermediate-risk group, according to each of the strategies, computed in the BioFINDER-1 and BioFINDER-2 MCI combined populations (n = 348). The error bars correspond to 95% CIs. d, Dots and lines indicating the observed percentage of reduction in further tests (here using CSF Aβ42/Aβ40) by applying the blood-based risk stratification strategy, based on each of the risk threshold strategies (90% Se/Sp, n = 301; 95% Se/Sp, n = 247; 97.5% Se/Sp, n = 205).CC BY · PubMed Central
Fig. 2 A model with z-transformed plasma p-tau217 levels enables interassay and geographical application of the two-step workflow. a, Distribution of predicted probabilities of Aβ-PET positivity based on a logistic regression model including z-transformed plasma p-tau217 levels, combined with age and APOE ε4. The z-transformation was done using a CU reference sample from each specific cohort, based on the mean and s.d. of each specific assay in its corresponding population. The predicted probabilities are displayed for the BioFINDER-1 (model training; left), BioFINDER-2 (model validation; middle) and TRIAD (geographical and interassay validation; right), with blue dots corresponding to individuals who are Aβ-PET negative and red dots to individuals who are Aβ-PET positive. On the right y axis, the probability values corresponding to the evaluated risk thresholds are demonstrated and accompanied by the metric used to define them (90%, 95%, 97.5% sensitivity or 90%, 95%, 97.5% specificity), and the original thresholds from main analyses were used to evaluate their robustness. The lower dashed line demonstrates where the 95% sensitivity low-risk threshold falls on the probability distribution, with the upper line corresponding to the 95% specificity high-risk threshold. b, The accuracy of both workflow steps combined, corresponding to the proportion of correct classifications for the low- and high-risk groups along with the proportion of correct CSF Aβ42/Aβ40 classifications in the intermediate-risk group. The dots correspond to the point estimates for observed accuracy and the lines to 95% CIs, computed based on each cohort’s full sample (BioFINDER-1, n = 136; BioFINDER-2, n = 212; TRIAD, n = 84). Each of the threshold strategies is represented by a color as indicated on the right. c, The percentage of reduction in further tests by applying the blood-based risk stratification strategy, based on each of the risk threshold strategies. The dots and lines correspond to the observed reduction in needed confirmatory tests (cohort (number of tests avoided according to 90%, 95% and 97.5% strategies, respectively): BioFINDER-1 (115, 92, 71); BioFINDER-2 (179, 151, 112); and TRIAD (72, 57, 46)).
Fig. 2 A model with z-transformed plasma p-tau217 levels enables interassay and geographical application of the two-step workflow. a, Distribution of predicted probabilities of Aβ-PET positivity based on a logistic regression model including z-transformed plasma p-tau217 levels, combined with age and APOE ε4. The z-transformation was done using a CU reference sample from each specific cohort, based on the mean and s.d. of each specific assay in its corresponding population. The predicted probabilities are displayed for the BioFINDER-1 (model training; left), BioFINDER-2 (model validation; middle) and TRIAD (geographical and interassay validation; right), with blue dots corresponding to individuals who are Aβ-PET negative and red dots to individuals who are Aβ-PET positive. On the right y axis, the probability values corresponding to the evaluated risk thresholds are demonstrated and accompanied by the metric used to define them (90%, 95%, 97.5% sensitivity or 90%, 95%, 97.5% specificity), and the original thresholds from main analyses were used to evaluate their robustness. The lower dashed line demonstrates where the 95% sensitivity low-risk threshold falls on the probability distribution, with the upper line corresponding to the 95% specificity high-risk threshold. b, The accuracy of both workflow steps combined, corresponding to the proportion of correct classifications for the low- and high-risk groups along with the proportion of correct CSF Aβ42/Aβ40 classifications in the intermediate-risk group. The dots correspond to the point estimates for observed accuracy and the lines to 95% CIs, computed based on each cohort’s full sample (BioFINDER-1, n = 136; BioFINDER-2, n = 212; TRIAD, n = 84). Each of the threshold strategies is represented by a color as indicated on the right. c, The percentage of reduction in further tests by applying the blood-based risk stratification strategy, based on each of the risk threshold strategies. The dots and lines correspond to the observed reduction in needed confirmatory tests (cohort (number of tests avoided according to 90%, 95% and 97.5% strategies, respectively): BioFINDER-1 (115, 92, 71); BioFINDER-2 (179, 151, 112); and TRIAD (72, 57, 46)).CC BY · PubMed Central
Fig. 3 A potential workflow for incorporating a plasma p-tau217 risk prediction model for predicting Aβ status in clinical practice. Conceptual flowchart for future implementation of the proposed two-step diagnostic workflow. Participants with cognitive impairment in specialized settings could be screened for the risk of underlying Aβ pathology based on a high-performance plasma p-tau biomarker model also incorporating clinically relevant variables, such as age and APOE ε4 status. Importantly, a clinical assessment would determine the need for an AD biomarker assessment. Comorbidities potentially affecting circulating biomarker levels should also be taken into consideration. Based on probability thresholds, chosen according to the decision to be made by the physician, patients could be stratified into low, intermediate and high risk of harboring underlying cerebral Aβ pathology. This biomarker-supported risk stratification could enable highly accurate decisions for individuals in the low- and high-risk groups. Individuals falling within the intermediate-risk group should be forwarded for further testing to determine their Aβ status with a confirmatory PET or CSF Aβ test, depending on center preference and availability. Such a strategy would largely reduce the number of further tests needed, while maintaining a high classification accuracy.
Fig. 3 A potential workflow for incorporating a plasma p-tau217 risk prediction model for predicting Aβ status in clinical practice. Conceptual flowchart for future implementation of the proposed two-step diagnostic workflow. Participants with cognitive impairment in specialized settings could be screened for the risk of underlying Aβ pathology based on a high-performance plasma p-tau biomarker model also incorporating clinically relevant variables, such as age and APOE ε4 status. Importantly, a clinical assessment would determine the need for an AD biomarker assessment. Comorbidities potentially affecting circulating biomarker levels should also be taken into consideration. Based on probability thresholds, chosen according to the decision to be made by the physician, patients could be stratified into low, intermediate and high risk of harboring underlying cerebral Aβ pathology. This biomarker-supported risk stratification could enable highly accurate decisions for individuals in the low- and high-risk groups. Individuals falling within the intermediate-risk group should be forwarded for further testing to determine their Aβ status with a confirmatory PET or CSF Aβ test, depending on center preference and availability. Such a strategy would largely reduce the number of further tests needed, while maintaining a high classification accuracy.CC BY · PubMed Central
Extended Data Fig. 1 Sensitivity and specificity across possible probability thresholds in both cohorts and derivation of probability thresholds. (a) Sensitivity and specificity across probability thresholds in BioFINDER-1 and BioFINDER-2, separately presented. The x-axis corresponds to the full range of possible thresholds for the probabilities of Aβ-PET positivity based on a plasma p-tau217-based model for Aβ-PET positivity. Solid lines correspond to the observed sensitivities and specificities point estimates for the range of possible probability thresholds, and ribbons to 95% confidence intervals, with BioFINDER-1 in light green and BioFINDER-2 in dark green. (b) Since sensitivity and specificity overlapped across the range of possible thresholds in both cohorts, we derived risk stratification thresholds based on predictions from both datasets combined. The lower-risk probability thresholds (left) evaluated were 42% (resulting in a sensitivity of 90%), 31% (resulting in a sensitivity of 95%) and 20% (resulting in a sensitivity of 97.5%), while the higher-risk probability thresholds (right) evaluated were 70% (resulting in a specificity of 90%), 80% (resulting in a specificity of 95%) and 85% (resulting in a specificity of 97.5%). (c) Calibration plot showing external validation in BioFINDER-2 of the model derived in BioFINDER-1. The solid black line shows smoothed associations between the predicted probabilities and observed frequencies of Aβ-PET-positivity. The closer this line is to the dotted grey identity line, the better performing and more generalizable a prediction model is. Aβ = Amyloid-β. PET = Positron emission tomography. P-tau217 = tau phosphorylated at threonine 217. Se = Sensitivity. Sp = Specificity.
Extended Data Fig. 1 Sensitivity and specificity across possible probability thresholds in both cohorts and derivation of probability thresholds. (a) Sensitivity and specificity across probability thresholds in BioFINDER-1 and BioFINDER-2, separately presented. The x-axis corresponds to the full range of possible thresholds for the probabilities of Aβ-PET positivity based on a plasma p-tau217-based model for Aβ-PET positivity. Solid lines correspond to the observed sensitivities and specificities point estimates for the range of possible probability thresholds, and ribbons to 95% confidence intervals, with BioFINDER-1 in light green and BioFINDER-2 in dark green. (b) Since sensitivity and specificity overlapped across the range of possible thresholds in both cohorts, we derived risk stratification thresholds based on predictions from both datasets combined. The lower-risk probability thresholds (left) evaluated were 42% (resulting in a sensitivity of 90%), 31% (resulting in a sensitivity of 95%) and 20% (resulting in a sensitivity of 97.5%), while the higher-risk probability thresholds (right) evaluated were 70% (resulting in a specificity of 90%), 80% (resulting in a specificity of 95%) and 85% (resulting in a specificity of 97.5%). (c) Calibration plot showing external validation in BioFINDER-2 of the model derived in BioFINDER-1. The solid black line shows smoothed associations between the predicted probabilities and observed frequencies of Aβ-PET-positivity. The closer this line is to the dotted grey identity line, the better performing and more generalizable a prediction model is. Aβ = Amyloid-β. PET = Positron emission tomography. P-tau217 = tau phosphorylated at threonine 217. Se = Sensitivity. Sp = Specificity.CC BY · PubMed Central
Extended Data Fig. 2 Flowchart illustrating plasma-based risk stratification and further testing of intermediate-risk individuals with CSF Aβ42/Aβ40 for the 90 and 97.5% strategies. Flowchart recapitulating results from the first step of the workflow (blood biomarker-based risk stratification) and demonstrating the accuracy for the second step of the clinical workflow, when intermediate-risk individuals are referred to CSF Aβ42/Aβ40 test for predicting Aβ-PET status. (a) Shows results for the 90% Se/Sp risk stratification strategy, and (b) for the 97.5% Se/Sp strategy. The 95% Se/Sp strategy is represented in Fig. 1b of the main text. Aβ = Amyloid-β. PET = Positron emission tomography. CSF = cerebrospinal fluid. P-tau217 = tau phosphorylated at threonine 217. Se = Sensitivity. Sp = Specificity. LP = lumbar puncture.
Extended Data Fig. 2 Flowchart illustrating plasma-based risk stratification and further testing of intermediate-risk individuals with CSF Aβ42/Aβ40 for the 90 and 97.5% strategies. Flowchart recapitulating results from the first step of the workflow (blood biomarker-based risk stratification) and demonstrating the accuracy for the second step of the clinical workflow, when intermediate-risk individuals are referred to CSF Aβ42/Aβ40 test for predicting Aβ-PET status. (a) Shows results for the 90% Se/Sp risk stratification strategy, and (b) for the 97.5% Se/Sp strategy. The 95% Se/Sp strategy is represented in Fig. 1b of the main text. Aβ = Amyloid-β. PET = Positron emission tomography. CSF = cerebrospinal fluid. P-tau217 = tau phosphorylated at threonine 217. Se = Sensitivity. Sp = Specificity. LP = lumbar puncture.CC BY · PubMed Central
Extended Data Fig. 3 Separate-step performance. For each of the graphics, the x-axis corresponds to the three evaluated strategies for blood-based biomarker risk stratification (Se/Sp 90%; Se/Sp 95%; Se/Sp 97.5%), with dots representing point estimates and bars corresponding to 95% confidence intervals, computed for the BioFINDER-1 and BioFINDER-2 combined population (n = 348). (a) Indicates the overall accuracy for the low- and high-risk groups for the workflow’s first step, that is blood-based biomarker risk stratification. This metric was calculated based on the number of Aβ-PET negative individuals classified to the low-risk group and of Aβ-PET positive individuals classified to the high-risk group (90% Se/Sp: n = 265; 95% Se/Sp: n = 229; 97.5% Se/Sp: n = 197), divided by the total individuals in the high and low-risk groups. (b) Displays the accuracy for the second step of the workflow. Individuals in the intermediate-risk group were assumed to be forwarded to a lumbar puncture to test CSF Aβ42/40 test, and the accuracy corresponding to the overall concordance of a CSF-negative result with a negative Aβ-PET scan and of a CSF-positive results with a positive Aβ-PET scan (90% Se/Sp: n = 42; 95% Se/Sp: n = 87; 97.5% Se/Sp: n = 143). Aβ = Amyloid-β. PET = Positron emission tomography. CSF = cerebrospinal fluid. P-tau217 = tau phosphorylated at threonine 217. Se = Sensitivity. Sp = Specificity.
Extended Data Fig. 3 Separate-step performance. For each of the graphics, the x-axis corresponds to the three evaluated strategies for blood-based biomarker risk stratification (Se/Sp 90%; Se/Sp 95%; Se/Sp 97.5%), with dots representing point estimates and bars corresponding to 95% confidence intervals, computed for the BioFINDER-1 and BioFINDER-2 combined population (n = 348). (a) Indicates the overall accuracy for the low- and high-risk groups for the workflow’s first step, that is blood-based biomarker risk stratification. This metric was calculated based on the number of Aβ-PET negative individuals classified to the low-risk group and of Aβ-PET positive individuals classified to the high-risk group (90% Se/Sp: n = 265; 95% Se/Sp: n = 229; 97.5% Se/Sp: n = 197), divided by the total individuals in the high and low-risk groups. (b) Displays the accuracy for the second step of the workflow. Individuals in the intermediate-risk group were assumed to be forwarded to a lumbar puncture to test CSF Aβ42/40 test, and the accuracy corresponding to the overall concordance of a CSF-negative result with a negative Aβ-PET scan and of a CSF-positive results with a positive Aβ-PET scan (90% Se/Sp: n = 42; 95% Se/Sp: n = 87; 97.5% Se/Sp: n = 143). Aβ = Amyloid-β. PET = Positron emission tomography. CSF = cerebrospinal fluid. P-tau217 = tau phosphorylated at threonine 217. Se = Sensitivity. Sp = Specificity.CC BY · PubMed Central
Extended Data Fig. 4 Reduced renal function does not seem to influence plasma p-tau217 levels between Aβ-PET negative and positive individuals classified as low- or high-risk. (a) The dots represent plasma p-tau217 concentrations (y-axis), with the x-axis representing the Aβ-PET status in combination with chronic kidney disease (CKD) as determined by an eGFR below 60 mL/min/1.73m2¬. Only participants classified as low- or high-risk at step-1 of the workflow based on the plasma p-tau217 model-derived probabilities and the 95% Se/Sp strategy are included. The colors indicate whether patients were correctly classified (blue) or misclassified (red). P-values come from t-tests (two-sided, alpha 0.05) were used to assess whether plasma p-tau217 levels were altered by the presence of CKD among Aβ-negatives and Aβ-positive participants. Plasma p-tau217 levels did not were not significantly altered by CKD among Aβ-negative participants or among Aβ-positive participants. (b) The y-axis represents plasma p-tau217 levels and the x-axis represents continuous values of eGFR, with the with the colors indicating whether patients were correctly classified (blue) or misclassified (red) at the step-1 of the workflow based on the plasma p-tau217 model-derived probabilities and the 95% Se/Sp strategy. The plot shows that misclassifications occur throughout all the span of renal function. Further, the dashed line indicates most of the misclassified individuals with CKD were, in fact, very close to the eGFR cutoff for CKD of 60 mL/min/1.73m2¬. Aβ = Amyloid-β. PET = Positron emission tomography. eGFR = estimated glomerular filtration rate. CKD = chronic kidney disease. P-tau217 = tau phosphorylated at threonine 217. Se = Sensitivity. Sp = Specificity.
Extended Data Fig. 4 Reduced renal function does not seem to influence plasma p-tau217 levels between Aβ-PET negative and positive individuals classified as low- or high-risk. (a) The dots represent plasma p-tau217 concentrations (y-axis), with the x-axis representing the Aβ-PET status in combination with chronic kidney disease (CKD) as determined by an eGFR below 60 mL/min/1.73m2¬. Only participants classified as low- or high-risk at step-1 of the workflow based on the plasma p-tau217 model-derived probabilities and the 95% Se/Sp strategy are included. The colors indicate whether patients were correctly classified (blue) or misclassified (red). P-values come from t-tests (two-sided, alpha 0.05) were used to assess whether plasma p-tau217 levels were altered by the presence of CKD among Aβ-negatives and Aβ-positive participants. Plasma p-tau217 levels did not were not significantly altered by CKD among Aβ-negative participants or among Aβ-positive participants. (b) The y-axis represents plasma p-tau217 levels and the x-axis represents continuous values of eGFR, with the with the colors indicating whether patients were correctly classified (blue) or misclassified (red) at the step-1 of the workflow based on the plasma p-tau217 model-derived probabilities and the 95% Se/Sp strategy. The plot shows that misclassifications occur throughout all the span of renal function. Further, the dashed line indicates most of the misclassified individuals with CKD were, in fact, very close to the eGFR cutoff for CKD of 60 mL/min/1.73m2¬. Aβ = Amyloid-β. PET = Positron emission tomography. eGFR = estimated glomerular filtration rate. CKD = chronic kidney disease. P-tau217 = tau phosphorylated at threonine 217. Se = Sensitivity. Sp = Specificity.CC BY · PubMed Central
Extended Data Fig. 5 Patients with CKD misclassified as false-positives often presented high CSF p-tau levels or CSF-positivity for Aβ42/Aβ40. This figure represents CSF p-tau181 levels and CSF Aβ42/Aβ40 status measured for BioFINDER-1 and BioFINDER-2 patients according to their classification status at the 95% Se/Sp risk stratification strategy with the main analysis plasma p-tau217-based model. The y-axis and dots displays CSF p-tau181, with colors representing CKD status (CKD-, blue; CKD + , red) and the shapes correspond to CSF Aβ42/Aβ40 status. In the x-axis, patients are stratified into true-negatives (low-risk label at step-1 who were also Aβ-PET-negative), false-negatives (low-risk label at step-1 who were Aβ-PET-positive), true-positives (high-risk label at step-1 who were also Aβ-PET-positive), false-positives (high-risk label at step-1 who were Aβ-PET-negative), with intermediate-risk individuals excluded from the plot (assumed to be referred for a CSF test with no applicable correct/incorrect classification label). (a) Displays CSF biomarker results measured with Elecsys for BioFINDER-1 and most of BioFINDER-2 patients, with the horizontal line corresponding to a previously validated cut-off for p-tau181 of 28 pg/mL (ref. 56). (b) Displays CSF biomarker results measured with Lumipulse a subset of BioFINDER-2 patients, with the horizontal line corresponding to a previously validated cut-off for p-tau181 of 50.2 pg/mL (ref. 59). For both assays, CSF Aβ42/Aβ40 was handled as described in the methods. Given the false-positive group (n = 7; x-axis, in bold) had demonstrated a higher rate of CKD after classification with a plasma p-tau217 risk stratification model, this figure indicates that n = 3 out of the n = 4 false-positives with CKD had elevated CSF p-tau181 levels (very close to indicated clinical cutoffs), with n = 2 of these patients also being positive for CSF Aβ42/Aβ40. This suggests a peripheral increase in plasma p-tau217 in the absence of Aβ-PET-positivity could be related to an underlying disease process (since CSF changes might occur earlier than PET) rather than peripheral impaired clearance. Aβ = Amyloid-β. CSF = cerebrospinal fluid. PET = Positron emission tomography. CKD = chronic kidney disease. P-tau181 = tau phosphorylated at threonine 181. P-tau217 = tau phosphorylated at threonine 217. Se = Sensitivity. Sp = Specificity.
Extended Data Fig. 5 Patients with CKD misclassified as false-positives often presented high CSF p-tau levels or CSF-positivity for Aβ42/Aβ40. This figure represents CSF p-tau181 levels and CSF Aβ42/Aβ40 status measured for BioFINDER-1 and BioFINDER-2 patients according to their classification status at the 95% Se/Sp risk stratification strategy with the main analysis plasma p-tau217-based model. The y-axis and dots displays CSF p-tau181, with colors representing CKD status (CKD-, blue; CKD + , red) and the shapes correspond to CSF Aβ42/Aβ40 status. In the x-axis, patients are stratified into true-negatives (low-risk label at step-1 who were also Aβ-PET-negative), false-negatives (low-risk label at step-1 who were Aβ-PET-positive), true-positives (high-risk label at step-1 who were also Aβ-PET-positive), false-positives (high-risk label at step-1 who were Aβ-PET-negative), with intermediate-risk individuals excluded from the plot (assumed to be referred for a CSF test with no applicable correct/incorrect classification label). (a) Displays CSF biomarker results measured with Elecsys for BioFINDER-1 and most of BioFINDER-2 patients, with the horizontal line corresponding to a previously validated cut-off for p-tau181 of 28 pg/mL (ref. 56). (b) Displays CSF biomarker results measured with Lumipulse a subset of BioFINDER-2 patients, with the horizontal line corresponding to a previously validated cut-off for p-tau181 of 50.2 pg/mL (ref. 59). For both assays, CSF Aβ42/Aβ40 was handled as described in the methods. Given the false-positive group (n = 7; x-axis, in bold) had demonstrated a higher rate of CKD after classification with a plasma p-tau217 risk stratification model, this figure indicates that n = 3 out of the n = 4 false-positives with CKD had elevated CSF p-tau181 levels (very close to indicated clinical cutoffs), with n = 2 of these patients also being positive for CSF Aβ42/Aβ40. This suggests a peripheral increase in plasma p-tau217 in the absence of Aβ-PET-positivity could be related to an underlying disease process (since CSF changes might occur earlier than PET) rather than peripheral impaired clearance. Aβ = Amyloid-β. CSF = cerebrospinal fluid. PET = Positron emission tomography. CKD = chronic kidney disease. P-tau181 = tau phosphorylated at threonine 181. P-tau217 = tau phosphorylated at threonine 217. Se = Sensitivity. Sp = Specificity.CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 65 references

Cited by

See all 267

Metadata

Authors
Brum, W. S., Cullen, N. C., Janelidze, S., Ashton, N. J., Zimmer, E. R., Therriault, J., Benedet, A. L., Rahmouni, N., Tissot, C., Stevenson, J., Servaes, S., Triana-Baltzer, G., Kolb, H. C., Palmqvist, S., Stomrud, E., Rosa-Neto, P., Blennow, K., & Hansson, O.
Affiliations
  • Universidade Federal do Rio Grande do Sul (BR)
  • Lund University (SE)
  • King's College London (GB)
  • McGill University Research Centre for Studies in Aging (CA)
  • Janssen Research & Development (United States) (US)
  • Sahlgrenska University Hospital (SE)
Corresponding author
  • Brum, Wagner S.
  • Hansson, Oskar
Venue
Nature Aging
Published
2023-08-31
Volume
3
Issue
9
Pages
1079-1090
ISSN
2662-8465 (ISSN-L)
OpenAlex
W4386324919
Cited by
248
References
65
Citations by year
2026: 68 · 2025: 98 · 2024: 74 · 2023: 8
FWCI
50.50
Type
journal article
Language
en
Open access
Licence
Locations
  • Nature Aging (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • LA Referencia (Red Federada de Repositorios Institucionales de Publicaciones Científicas) (submittedVersion, cc-by-nc-sa)
  • Lund University Publications (Lund University) (submittedVersion)
Funders
  • Alzheimer's Association (ZEN-21-848495, MOP-11-51-31, 159815, 2021 Zenith Award, NIRP-12-259245, RFN 152985, #NIRG‐12‐92090, 34874, 33397, 162303)
  • Cure Alzheimer's Fund
  • Fondation Brain Canada (ZEN-21-848495, 2020-VICO-279314, 34874, 33397)
  • Stiftelsen för Gamla Tjänarinnor
  • Weston Brain Institute (MOP-11-51-31, NIRP‐12‐259245)
  • Australian Government (2022-Projekt0080)
  • Consortium canadien en neurodégénérescence associée au vieillissement (MOP-11-51-31)
  • Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (MOP-11-51-31, 88887.596742/2020-00, 88887, NIRP-12-259245, RFN 152985, 159815, 2020-00, 162303, NIRG-12-92090)
  • Lunds Universitet (AF-980907)
  • Hjärnfonden (ALZ2022-0006), FO2021-0293, FO2017‑0243, FO2017, ZEN-21-848495, ALFGBG-715986, #ALFGBG-965240, 1412/22)
  • Knut och Alice Wallenbergs Stiftelse (2017‐0383, 2022-Projekt0080, 2022-00775, AF-980907, 2022-1259, FO2021-0293, 2020-O000028, 1412/22)
  • Vetenskapsrådet (ALFGBG-965240, ALZ2022-0006, ERAPERMED2021-184, 2017-00915, (#ALFGBG-715986, 2022-00775, ZEN-21-848495, 2022-Projekt0080, FO2021-0293, AF-980907, 1412/22, 2020-O000028, 2017-00383, FO2017-0243, 2022-1259)
  • Konung Gustaf V:s och Drottning Victorias Frimurarestiftelse (2020-O000028, FO2017-0243)
  • Skånes universitetssjukhus (2022-1259, 2020-O000028, 2022-Projekt0080)
  • Parkinsonfonden (1412/22)
  • Canadian Institutes of Health Research (#MOP-11-51-31, 33397, NIRP-12-259245, NIRG-12-92090, 152985, 159815, RFN 152,985, 162303, 92090)
Topics
  • Health Sciences › Medicine › Psychiatry and Mental health › Dementia and Cognitive Impairment Research
  • Life Sciences › Neuroscience › Cellular and Molecular Neuroscience › Alzheimer's disease research and treatments
  • Health Sciences › Medicine › Oncology › Cancer-related cognitive impairment studies
Keywords
plasma p-tau217, Alzheimer's disease, mild cognitive impairment, blood biomarkers, risk stratification, APOE ε4, cost-effective diagnosis, memory clinics
MeSH terms
Alzheimer Disease, Blood Proteins, Humans, Tomography, X-Ray Computed, Amyloid beta-Peptides, Workflow

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: A two-step workflow based on plasma p-tau217 to screen for amyloid β positivity with further confirmatory testing only in uncertain cases. https://app.cochat.ai/papers/10.1038/s43587-023-00471-5
The paper
Brum, W. S., Cullen, N. C., Janelidze, S., Ashton, N. J., Zimmer, E. R., Therriault, J., Benedet, A. L., Rahmouni, N., Tissot, C., Stevenson, J., Servaes, S., Triana-Baltzer, G., Kolb, H. C., Palmqvist, S., Stomrud, E., Rosa-Neto, P., Blennow, K., & Hansson, O. (2023). A two-step workflow based on plasma p-tau217 to screen for amyloid β positivity with further confirmatory testing only in uncertain cases. Nature Aging. https://doi.org/10.1038/s43587-023-00471-5

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.