vlog

[Skip to Navigation]
Sign In
Viewpoint
January 4, 2021

The Future Is P-Tau—Anticipating Direct-to-Consumer Alzheimer Disease Blood Tests

Author Affiliations
  • 1Department of Medical Ethics and Health Policy, University of Pennsylvania Perelman School of Medicine, Philadelphia
  • 2Department of Medicine, University of Pennsylvania Perelman School of Medicine, Philadelphia
JAMA Neurol. 2021;78(4):379-380. doi:10.1001/jamaneurol.2020.4835

Recent studies suggest blood levels of phosphorylated tau (p-tau) isoforms can detect both the tau and amyloid pathologies that define Alzheimer disease (AD).1 More research is needed to replicate these results in large, diverse cohorts and quantify a p-tau blood test’s ability to prognosticate the onset of dementia. Nevertheless, we anticipate that the ability to receive diagnostic and prognostic information with the ease of a blood test will revolutionize AD research and care. We also anticipate that it will be the basis of a profitable direct-to-consumer (DTC) test, one that highlights considerable ethical and social challenges of DTC testing and AD.

Add or change institution
1 Comment for this article
EXPAND ALL
P-Tau as a diagnostic test for Alzheimer Disease
Eleftherios Diamandis, MD,PhD | University of Toronto
Largent et al. advocate the use of direct-to-consumer (DTC) testing for p-tau as a predictive and early diagnostic test for Alzheimer Disease (AD). While in theory this could be possible, there are many issues that may stand in the way. For example, p-tau levels in serum range between 0.05-0.1 ng/mL. DTC assays are usually much less sensitive than standard laboratory tests, usually not being able to quantify less than 1 ng/mL. Consequently, it is unlikely that any current DTC test will be able to accurately quantify p-tau in serum. Second, any predictive test for AD must be quantitative and highly precise (e.g. target coefficient of variation <15%) since changes of p-tau in AD are relatively small (e.g. levels may double or triple, at best). The proposed application is much different from current applications of DTC, when the goal is to detect absence or presence of a biomarker (such as in pregnancy). Third, and probably most important, if a test is used to uncover a disease with relatively low prevalence, its sensitivity must be high (e.g. >90%) and its specificity even higher (e.g. >99%) in order for the positive predictive value (PPV) to be acceptable (e.g. >50%) (1). The PPV is defined as the chance of somebody having a disease (in this case AD) if the test is positive. We recently calculated the PPV and other parameters at various disease prevalences and various test sensitivities and specificities (2). To illustrate our point, we mention here that at a prevalence of 1%, a test with a sensitivity of 90% and a specificity of 95% will have a PPV of only 15%. At a prevalence of 5%, the PPV becomes 50% (considered as barely acceptable) only when the sensitivity of the test is 98% and the specificity is 95%.
I conclude that the use of p-tau as a DTC test for AD does not seem to be promising, given what we know about disease prevalence and test sensitivity, specificity and predictive value.

References

1. Fiala C, Diamandis EP. A multi-cancer detection test: focus on the positive predictive value. Ann Oncol. 2020 Sep;31(9):1267-1268. doi: 10.1016/j.annonc.2020.05.028. Epub 2020 Jul 30. PMID: 32741675.

2. Prassas I, Fiala C, Diamandis EP. Assay requirements for COVID-19 testing: serology vs. rapid antigen tests. Clin Chem Lab Med. 2021 Mar 18. doi: 10.1515/cclm-2021-0234. Epub ahead of print. PMID: 33730769.
CONFLICT OF INTEREST: None Reported
READ MORE
×