In February, the Foundation for the National Institutes of Health’s (FNIH) Biomarkers Consortium published a study in Nature Medicine describing a predictive “clock” model for Alzheimer’s disease (AD). By measuring plasma phosphorylated tau-217 (p-tau217) levels, researchers could estimate the years until the onset of AD cognitive symptoms with remarkable accuracy, decades before they appear.
Today's Clinical Lab spoke with Alessio Travaglia, PhD, director, Translational Science, Neuroscience and Rare Diseases, FNIH, about the new model and the future of Alzheimer’s research.
What motivated the Consortium to develop a plasma p-tau217 clock model as a biomarker for Alzheimer’s disease?
Alessio Travaglia: The FNIH Biomarkers Consortium has been asking "what’s next" in the Alzheimer’s space for over 20 years. Beyond considering what to measure as a reliable biomarker, like p-tau217, amyloid-beta, neurofilament light chain (NfL), or other candidate biomarkers, the real clinical and research hurdle was designing a clinical trial knowing not only if someone will develop Alzheimer’s but when?
We wanted to make clinical trials more effective, and to do that, we needed to ensure that the right people are enrolled in the right trials at the right time. By predicting AD onset, we ensure that participants are at a stage where they can benefit most from a drug, giving the trial the highest chance of success.
How much could this "clock" model of AD reduce the cost and duration of Phase 3 prevention trials, which currently take years and involve thousands of participants?
AT: A helpful analogy here is that of developing a drug to reduce fever. You don’t want to give this drug to someone with no fever, nor to someone whose fever has already peaked. You want to give it to the person who is about to “spike” a fever in a few hours.
Similarly, for AD, the p-tau217 clock model has an accuracy margin of about three years. Knowing that margin allows for much more accurate enrollment, and therefore, improves the likelihood of the drug showing efficacy.
The time from %p-tau217 positivity to onset of Alzheimer’s symptoms is markedly shorter in older individuals. Do older individuals require different thresholds for diagnosis?
AT: That was a surprising result, but the data is clear. If I were designing a clinical trial today, I could no longer use a wide age range for inclusion. Moving forward, we have to be very specific with age of enrollment because this study shows that people progress to symptom onset at different speeds depending on their age.
For clinical labs, the takeaway is that the same biomarker level means something different at age 60 than it does at age 80. The inclusion criteria for trials and eventually diagnostic guidelines will need to be refined based on these age-related trajectories.
Many labs currently rely on positron emission tomography scans or spinal taps. Will blood biomarkers replace these more invasive and expensive technologies?
AT: For decades, diagnosis was based on physicians’ observations, which had a 30–50 percent misdiagnosis rate. Positron emission tomography (PET) tracers changed the game, but they are expensive and not globally accessible.
In contrast to PET and spinal taps, blood biomarkers are scalable, cost-effective, and can be done at a primary care level.
In many ways, p-tau217 correlates so well with PET scans that you have to ask—is the scan even needed if the blood work tells you the same thing? We may see a future in neurology similar to oncology, with bloodwork as the first screen for diseases followed by imaging only if there is ambiguity.
Since this test can predict symptoms decades in advance, what are the ethical implications of giving a patient a "countdown" to AD, especially when we don't have a cure yet?
AT: It’s a tough question, and the answer will require input from clinicians and others in the medical community and beyond. From my own perspective, I would like to know an AD diagnosis in advance, and there are two reasons to that:
First, we have the first generation of drugs now, which are not cures but they are a start.
Second, research shows that lifestyle interventions, including sleep, diet, and exercise, can impact disease trajectory for various neurological diseases.
For example, the Finnish Geriatric Intervention Study to Prevent Cognitive Impairment and Disability, or the FINGER trial, showed that we can slow cognitive decline in at-risk older adults through these interventions. In my opinion, the advanced notice provided by the p-tau clock gives patients a window to act.
What further studies is the Consortium conducting to evaluate the efficacy of the p-tau217 clock model?
AT: To understand if these results are generalizable, we are now performing follow-up studies with more cohorts beyond the two cohorts in the published study to see if the p-tau217 clock model works the same across many populations.
We’re also looking for “unusual suspects,” which are additional biomarkers that can further narrow that three- to four-year margin of error down to six months or less. One candidate we are working on is eMTBR-tau243, and we have additional promising candidates.
Now that we can predict the timing of onset, where should the field focus its energy next?
AT: The future of AD research will continue to embrace a multi-pronged approach: drug development and clinical trials can focus on using the p-tau clock to streamline trials; non-profits like FNIH will work on guidance documents and ethical frameworks; meanwhile, basic scientists can continue to focus on the underlying biological mechanisms, e.g., they can now ask what starts the clock by triggering p-tau217 elevation in the first place, and how do we halt or reset it?
This study is published as the FNIH Biomarkers Consortium celebrates 20 years. It’s a testament to our vision of team science—working together across industry, academia, and government—as the best way to tackle a problem as complex as Alzheimer’s disease.
