Blood test for Alzheimer’s could replace costly and invasive spinal taps

By Dr. Priyom Bose, Ph.D.Reviewed by Benedette Cuffari, M.Sc.Feb 20 2025

A new blood test for Alzheimer’s shows high accuracy in detecting disease biomarkers—potentially reducing the need for invasive spinal taps by over 50%.

Study: Clinical performance of the fully automated Lumipulse plasma p-tau217 assay in mild cognitive impairment and mild dementia. Image Credit: Pickadook / Shutterstock.com

A fully automated Lumipulse plasma phosphorylated tau 217 (p-tau217) assay has the potential to effectively diagnose mild cognitive impairment (MCI) and mild dementia. In a recent Alzheimer’s and Dementia study, researchers indicate that this method can potentially reduce over 50% of lumbar punctures performed for diagnostic purposes.

How is Alzheimer's disease diagnosed?

The clinical diagnosis of Alzheimer's disease (AD) is based on the detection of amyloid beta (Aβ) and tau (T) proteins as biomarkers in cerebrospinal fluid (CSF) and positron emission tomography (PET). The rapid progression of neurodegenerative diseases like AD emphasizes the need to establish Aβ positivity before anti-amyloid immunotherapies are initiated.

Previous studies have highlighted greater benefits associated with blood-based biomarkers (BBMs) as compared to CSF and PET for AD diagnosis. For example, blood sampling improves the ease of trial recruitment, diagnoses, and monitoring of AD progression, whereas CSF sampling involves an invasive lumbar puncture (LP) that increases the risk of bleeding, infection, and nerve damage.

Of all BBMs, p-tau217 has the potential to differentiate AD from other neurodegenerative diseases through its significant change and tracks worsening cortical atrophy over time. Although mass spectrometry is used to quantify low-abundant plasma proteins, newer automated immunoassay techniques are needed to accurately determine plasma p-tau217 and Aβ pathology.

Previous studies have confirmed the utility of fully automated methods, such as Fujirebio Lumipulse, in diagnostic laboratories to reduce variability in assay performance. However, the effectiveness of these assays must be validated through real-world clinical data.

About the study

The current study examined the performance of the fully automated Fujirebio research-use-only (RUO) Lumipulse G plasma p-tau217 assay to determine CSF-defined Aβ pathology in individuals subjected to lumbar puncture (LP) for disease diagnosis in a regional specialist memory clinic (RSMC).

RSMC performs detailed cognitive/medical assessments, neuroimaging, and routine laboratory investigations for 400 to 500 individuals every year. If required, patients undergo CSF testing to detect AD pathology.

Standard curves and supplied calibrators were used for optimal assay performance on a Lumipulse G600 II analyzer. To estimate the inter-assay coefficient of variation (CV), six plasma samples were analyzed twice on two separate days.

Study findings

A total of 148 participants were recruited for the current study, 54.1% of whom were female with a mean age of 69.4 years. Approximately 68% of the participants were diagnosed with MCI, whereas 31% had mild dementia. The mean time required from initial diagnosis to confirmation was about 120 days.

CSF testing estimated 60.8% of the study cohort to be Aβ positive. The suspected etiology of Aβ negative individuals was Lewy body disease, frontotemporal dementia, and vascular cognitive impairment. In some individuals who were negative for Aβ, neurodegenerative etiology was not clear after diagnostic tests; however, each of these study participants were diagnosed with MCI.

Plasma sample analyses using the Lumipulse p-tau217 assay revealed that median plasma p-tau217 levels were four times higher in Aβ-positive study participants than Aβ-negative individuals. Plasma p-tau217 had an area under the curve (AUC) value of 0.92 in Aβ-positive patients.

A two-threshold approach was used to provide more stringent thresholds under which, in theory, only intermediate results would proceed to further testing. In the current analysis, 90%, 95% and 97.5% of sensitivity/specificity thresholds for plasma p-tau217 may have obviated the need for more than half of LPs.

Fewer false positive results were generated at a specificity of 90%. These individuals had MCI without specific neurodegenerative etiology.

Conclusions

The current study confirmed the excellent performance of the fully automated Lumipulse assay in detecting Aβ positivity in a real-world clinical cohort of patients diagnosed with MCI and mild dementia. Taken together, these findings demonstrate that this device can be used in laboratory settings for diagnostic purposes with higher specificity and sensitivity than current analytical approaches.

In the future, the potential impact of renal function and vascular risk factors on plasma p-tau217 performance should be assessed at a large-scale and by using a more diverse sample.