Skip to main content
. 2019 Dec 12;8(Suppl 2):129–145. doi: 10.1007/s40120-019-00166-3

Table 1.

Comparison of different aspects relevant for implementation of Aβ-PET, CSF, and blood-based analysis methods

Topic Aβ-PET imaging CSF analysis Blood analysis
Context of use (COU) Used in clinical trials as an entry point for patient stratification with the aim to identify subjects with Aβ load in the brain [73] More applications: early diagnosis, differential diagnosis, treatment follow-up

Multiple applications in the diagnostic or therapeutic area, including (pre-) clinical studies (PK/PD testing)

Biomarkers can be disease-related or treatment-related, classified by the BEST biomarker category (Web-1)

Clinical

Method comparison

Both methods have a high clinical accuracy to identify Aβ load in the brain. Concordance between results is high, although discrepant results have been published Correlation with CSF for some proteins (e.g., NF-L, Aβ ratio), not for other proteins
Comparison needs to be done with respect to the clinical utility or for a specific COU
Possibility for use of blood protein panels in combination with other diagnostic tools (e.g., CSF) [21]

Clinical

Change in function of disease process

Aβ-PET scan differences are visualized after changes in CSF biomarker profiles Changes in concentrations of CSF proteins are detectable more than 20 years before cognitive symptoms appear To be defined, as compared to Aβ-PET imaging or CSF analysis
Comparison is dependent on the COU
Clinical trials Used for patient stratification Exploratory biomarker Not integrated in trials yet
Analytes_number

One hallmark.

Different outcome in function of the selected tracer

Several relevant proteins can be included in the panel (e.g., tau, Aβ, NF-L, …) A huge number of proteins can be included in the panel
Most proteins are detectable with classical technology platforms Concentrations are lower than in CSF (for most of the analytes), but technologies allow low limit of detection
Cost High cost, limiting repeated measures Limited cost/sample Low cost/sample
Repeated sampling is possible, but more difficult Repeated sampling is possible
Safety Exposure to radiation. Problems with claustrophobia for demented subject Lumbar Puncture is perceived as invasive and with a risk of 1–5% for severe headache problems [59] Safe, routine procedure
Instrumentation

PET scanners: not available worldwide

Cyclotrons: tracers need to be on-site within 1 h of preparation

Random-access immunoassay analyzers have been qualified for accurate and reliable quantification of CSF proteins [9] Point-of-care test systems for multiple protein analysis will become available soon. Random-access analyzers are already available in many labs
Technology is only available in specialized centers

Standardization

Training

It needs to be performed independent of a patient’s clinical feature. Only by readers who successfully complete a special training program

CSF-Aβ1–42 data can be harmonized between technology platforms using CSF-based certified reference materials [32]

Efforts are ongoing for a worldwide implementation of a SOP for collection and storage of CSF. Run-validation criteria can be established easily. Each lab technician, skilled in the art, can perform these immunoassays

No special training required, except availability of technicians, skilled in the art to perform immunoassays
Data interpretation Visual interpretation or software. Flashy outcome Reader and curve fit algorithms. Objective outcome. Interpretation is more difficult. No universal cutoff value available
Regulatory The FDA approved Aβ-PET imaging by visual interpretation (‘not software’) to identify subjects with ongoing amyloidopathy (= exclusion criterium) [73]

EMA qualified the combined use of CSF Aβ and tau proteins as a tool for patient enrollment in clinical trials (Web-2)

A letter of support from the FDA encourages their use in clinical trials (Web-3). The method must be used in adjunct to other diagnostic evaluations. Algorithms based on CSF proteins, together with imaging methods, were integrated into revised criteria for AD diagnosis [18] and research proposals for the amyloid/tau/neurodegeneration (A/T/N) classification of AD pathological processes [27]

No qualification proposals submitted
No regulatory approved assay formats

Website references: Web-1 https://www.ncbi.nlm.nih.gov/books/NBK326791/; Web-2 https://www.ema.europa.eu/en/documents/regulatory-procedural-guideline/qualification-opinion-novel-methodologies-predementia-stage-alzheimers-disease-cerebrospinal-fluid_en.pdf; Web-3 https://www.fda.gov/drugs/cder-biomarker-qualification-program/letter-support-los-initiative

 β-amyloid, COU context of use, CSF cerebrospinal fluid, EMA the European Medicine Agency, FDA The Food and Drug Administration, NF-L neurofilament light, PET positron emission tomography, SOP standard operating procedure