Measurement of plasma phosphorylated tau 217 offers transformed paradigms for clinical trials of treatments for Alzheimer’s disease

Introduction

The recent report characterising the utility of a commercially available immunoassay of phosphorylated tau 217 (p-tau217) in plasma to detect Alzheimer’s disease (AD) pathology was greeted with widespread enthusiasm (Ashton et al., 2024). The report itself, and much of the comment in response to it, focused on its potential use in clinical diagnosis: augmenting the ability to distinguish AD from other causes of dementia, to distinguish early AD from mild cognitive impairment (MCI) or to detect AD in its preclinical stages. However, the full benefit of these functions would only be realised once effective drugs to prevent AD progression became available. Here, I focus on a much more immediate application, the way that p-tau217 testing allows for the development of clinical trials which are dramatically cheaper, less demanding and shorter than those which have to date been used to assess the efficacy of proposed disease-modifying treatments for AD.

Key characteristics of the performance of the plasma immunoassay for p-tau217

In order to fully understand the impact which the plasma immunoassay for p-tau217 might have on clinical trial design, it is helpful to be aware of some key aspects of its performance in relation to other more invasive measures of AD progression and to clinical manifestations of disease.

Previously, the optimum biomarkers measures to identify Aβ and tau features of AD pathology consisted of invasive and/or demanding investigations requiring positron emission tomography (PET) scans or CSF sampling (Hansson et al., 2023). Levels of p-tau217 have been shown to identify individuals with MCI who have features of Aβ pathology and also to be good predictors for progression of MCI to AD (Gonzalez-Ortiz et al., 2024; Janelidze et al., 2023; Mendes et al., 2024). In the study of the commercially available assay, plasma p-tau217 had high accuracy for identifying patients with elevated CSF Aβ and tau pathology, with area under the curve (AUC) of 0.92 and 0.93 respectively (Ashton et al., 2024). Levels of plasma p-tau217 increased across tau-PET–defined Braak stages and increased over time in Aβ-positive individuals, with the highest increase in those with tau positivity.

Overall, it seems that plasma p-tau217 levels agree very well with other indicators of the pathological processes leading to clinical AD. The levels do also discriminate between AD cases and controls and do also predict clinical progress. Of course, even the best measures of pathology do not correlate completely with clinical status. As discussed below, in the context of early clinical trials for drugs targeting AD pathology, the ability to assess the underlying pathology may be of key importance, ahead of effects on clinical outcomes.

Possible advantages for use of plasma p-tau217 levels in clinical trials

A typical clinical trial for a possible treatment for Alzheimer’s disease would involve both a CSF and/or PET assessment of pathology and a clinical assessment of cognitive function. It would require a placebo control group and sample sizes would need to be large enough to plausibly provide adequate power, typically running into hundreds (van Dyck et al., 2023). And it would need to take place over a period of time long enough for there to be a reasonable expectation that a clinical effect might be observed, for example 18 months, in order to demonstrate that the rate of deterioration in cognitive function was lower in treated participants than in controls. All these features mean that there are requirements for specialist expertise, a research centre which can support the specialist investigations, substantial resources and of course financial support for all of these. Here it is argued that with some not unreasonable assumptions it might be possible to carry out very small scale phase 2 trials which could be done at any clinical research centre with no requirement for any specialist ability to carry out investigations or even to perform cognitive testing.

No need for PET scanning or CSF sampling

This benefit is self-evident. If we accept that plasma p-tau217 provides a reasonable indicator of underlying pathology with performance on a par with PET or CSF based assessments, as seems reasonable, then we can use it as our sole biomarker.

No need for clinical assessment

Again, if we accept that plasma p-tau217 is a marker of disease activity then we can use it as the sole outcome measure and we do not need to carry out detailed neuropsychological assessments for this purpose. Furthermore, we do not even necessarily need to carry out any clinical assessment in the process of recruitment. One might recruit volunteers from a memory clinic who were already diagnosed with MCI or early AD. Alternatively, one might simply advertise for volunteers aged 55-65, measure plasma p-tau217 in all of them and recruit the 10% or so who were found to have moderately raised levels without carrying out any assessment of their cognitive function. Thus in theory it might be feasible to carry out clinical trials in research centres with no access to neuropsychiatric services. The research could be delivered by anybody authorised to carry out drug trials.

No need for large cohorts or a protracted assessment period

A desirable property of plasma p-tau217, which does not apply to clinical indicators of AD progress, is that the levels are dynamic and might be expected to go down if a treatment effectively impacts the disease process. By contrast, the expectation is that an effective treatment will not restore cognitive function but will at best slow its decline. This implies that one would need a large sample assessed over a prolonged period in order to expect to see any difference in cognitive function between treatment and placebo groups. By contrast, since measured plasma p-tau217 levels will reflect a balance between production and elimination from the peripheral circulation, it is not unreasonable to think that an effective treatment might produce a rapid, perhaps immediate, decline in these levels. If one were in a position to assume that levels do not usually decline spontaneously then a fall might be taken as a promising indication of a treatment effect which could lead to more intensive investigation. Again, in this situation it might be reasonable use only a relatively small number of participants.

No need for a placebo group

If one is ready to make the assumption that p-tau217 levels do not decline spontaneously then arguably there is no need for a placebo group. One could simply administer a treatment to all participants to see if a decrease occurred.

Suggested protocol

To summarise, a minimalist protocol to screen a number of promising treatments which could be carried out at any clinical trials centre might be as follows:

Advertise for volunteers aged between 55 and 65. Measure plasma p-tau217 in 500. Select 50 subjects with the highest levels. Administer the trial treatment for 3 months. Measure the levels again. If levels have fallen then carry out further investigations of this treatment. Otherwise, move on to trial the next treatment using the same cohort.

Such an approach would allow a quick screen for an effect on AD progression of 8 different treatments in just 2 years.

Practical implications for trialling new treatments

Following the procedures outlined above, to a greater or lesser extent, could dramatically reduce the requirements for a trial to discover whether a proposed treatment seems to impact the AD process and hence reduces the necessity to have substantial pre-existing evidence before embarking on such a trial. To take a concrete example, we and others have suggested that vanadium might impact AD progression (Curtis et al., 2019; Curtis and Bandyopadhyay, 2021; Gonzalez-Cano et al., 2024; He et al., 2022; Tavares et al., 2023; Yao et al., 2023). The rationale for this is based on animal studies, genetic studies and an understanding of the effects of vanadate on biological processes, in particular the insulin receptor activated pathway which seems to inhibit tau phosphorylation. Admittedly, the evidence is not overwhelming but it certainly seems sufficient to justify at least one trial to see whether it does in fact have any impact on AD progression. However, vanadate, at least in some forms, is readily available and sold as a dietary supplement and could not be patented. Thus, a pharmaceutical company would have no interest in sponsoring a trial of vanadate no matter how strong was the preclinical evidence that it might be effective. And the huge sums of money which to date have been required for AD clinical trials have meant that no independent researcher has successfully applied to grant-giving bodies to carry out such a trial.

The ability to use a marker in peripheral blood which can cheaply and effectively provide an indicator of underlying AD process now enables researchers to carry out initial phase 2 screening trials of vanadate and perhaps dozens of other substances in a similar situation, for which there is only modest preclinical evidence for an effect and/or which are not attractive prospects for pharmaceutical companies. Hopefully, as the potential benefits of this approach are recognised such trials will indeed start to take place.

References

Ashton, N.J., Brum, W.S., Molfetta, G. Di, Benedet, A.L., Arslan, B., Jonaitis, E., Langhough, R.E., Cody, K., Wilson, R., Carlsson, C.M., Vanmechelen, E., Montoliu-Gaya, L., Lantero-Rodriguez, J., Rahmouni, N., Tissot, C., Stevenson, J., Servaes, S., Therriault, J., Pascoal, T., Lleó, A., Alcolea, D., Fortea, J., Rosa-Neto, P., Johnson, S., Jeromin, A., Blennow, K., Zetterberg, H. (2024) Diagnostic Accuracy of a Plasma Phosphorylated Tau 217 Immunoassay for Alzheimer Disease Pathology. JAMA Neurol 81, 255–263.

Curtis, D., Bakaya, K., Sharma, L., Bandyopadhay, S. (2019) Weighted burden analysis of exome-sequenced late onset Alzheimer’s cases and controls provides further evidence for involvement of PSEN1 and demonstrates protective role for variants in tyrosine phosphatase genes. Ann Hum Genet 84, 291–302.

Curtis, D., Bandyopadhyay, S. (2021) Mini‐review: Role of the PI3K/Akt pathway and tyrosine phosphatases in Alzheimer’s disease susceptibility. Ann Hum Genet 85, 1–6.

Gonzalez-Cano, S.I., Flores, G., Guevara, J., Morales-Medina, J.C., Treviño, S., Diaz, A. (2024) Polyoxidovanadates a new therapeutic alternative for neurodegenerative and aging diseases. Neural Regen Res 19, 571–577.

Gonzalez-Ortiz, F., Ferreira, P.C.L., González-Escalante, A., Montoliu-Gaya, L., Ortiz-Romero, P., Kac, P.R., Turton, M., Kvartsberg, H., Ashton, N.J., Zetterberg, H., Harrison, P., Bellaver, B., Povala, G., Villemagne, V.L., Pascoal, T.A., Ganguli, M., Cohen, A.D., Minguillon, C., Contador, J., Suárez-Calvet, M., Karikari, T.K., Blennow, K. (2024) A novel ultrasensitive assay for plasma p-tau217: Performance in individuals with subjective cognitive decline and early Alzheimer’s disease. Alzheimers Dement 20, 1239–1249.

Hansson, O., Blennow, K., Zetterberg, H., Dage, J. (2023) Blood biomarkers for Alzheimer’s disease in clinical practice and trials. Nat Aging 3, 506–519.

He, Z., Zheng, L., Zhao, X., Li, X., Xue, H., Zhao, Q., Ren, B., Li, N., Ni, J., Zhang, Y., Liu, Q. (2022) An Adequate Supply of Bis(ethylmaltolato)oxidovanadium(IV) Remarkably Reversed the Pathological Hallmarks of Alzheimer’s Disease in Triple-Transgenic Middle-Aged Mice. Biol Trace Elem Res 200, 3248–3264.

Janelidze, S., Bali, D., Ashton, N.J., Barthelemy, N.R., Vanbrabant, J., Stoops, E., Vanmechelen, E., He, Y., Dolado, A.O., Triana-Baltzer, G., Pontecorvo, M.J., Zetterberg, H., Kolb, H., Vandijck, M., Blennow, K., Bateman, R.J., Hansson, O. (2023) Head-to-head comparison of 10 plasma phospho-tau assays in prodromal Alzheimer’s disease. Brain 146, 1592–1601.

Mendes, A.J., Ribaldi, F., Lathuiliere, A., Ashton, N.J., Janelidze, S., Zetterberg, H., Scheffler, M., Assal, F., Garibotto, V., Blennow, K., Hansson, O., Frisoni, G.B. (2024) Head-to-head study of diagnostic accuracy of plasma and cerebrospinal fluid p-tau217 versus p-tau181 and p-tau231 in a memory clinic cohort. J Neurol 271.

Tavares, C.A., Santos, T.M.R., da Cunha, E.F.F., Ramalho, T.C. (2023) Parameterization and validation of a new AMBER force field for an oxovanadium (IV) complex with therapeutic potential implications in Alzheimer’s disease. J Mol Graph Model 122.

van Dyck, C.H., Swanson, C.J., Aisen, P., Bateman, R.J., Chen, C., Gee, M., Kanekiyo, M., Li, D., Reyderman, L., Cohen, S., Froelich, L., Katayama, S., Sabbagh, M., Vellas, B., Watson, D., Dhadda, S., Irizarry, M., Kramer, L.D., Iwatsubo, T. (2023) Lecanemab in Early Alzheimer’s Disease. New England Journal of Medicine 388, 9–21.

Yao, J., He, Z., You, G., Liu, Q., Li, N. (2023) The Deficits of Insulin Signal in Alzheimer’s Disease and the Mechanisms of Vanadium Compounds in Curing AD. Curr Issues Mol Biol 45, 6365–6382.