Spotlight on Alzheimer’s: The biomarker race
Volume 11, Issue 18 | May 1, 2025
ALSO IN THIS ISSUE
Bird Flu Update: Global Virus Network scientists urge surveillance
Measles Outbreaks: KS cases spike, NM cases slow
This week, the Alzheimer’s Association published their annual Alzheimer’s Disease Facts and Figures report. Along with the usual daunting numbers (for the first time, over 7 million folks in the US have the disease) is a special report with the results of a nationwide survey of older Americans’ “awareness and attitudes” about Alzheimer’s testing, early detection, and treatment. (Read the Facts and Figures executive summary here or download the whole magilla.)
Some highlights, as reported by ABC News:
79% would want to know they had the disease before symptoms began to interfere with their daily activities
80% said they would ask a doctor to be tested, rather than waiting for a test to be offered
91% said they'd want a blood biomarker test, or some other simple test
The Association plans to release guidelines for the use of blood-based biomarker tests this year and continues to advocate for state-based legislation requiring insurance coverage for “comprehensive biomarker testing.” It’s also working with public-health agencies to help educate both clinicians and the public about the latest diagnostic research and best practices.
In light of these reports, this week’s issue will tell you what the latest science has to say about early Alzheimer’s diagnosis. Which biomarkers are the best? (Spoiler alert: pTau217 tops the list so far.) And do these biomarkers tell us what we need to know about how best to treat this frustrating disease? (Maybe, maybe not.)
While research remains slow and expensive, its pace has been accelerating. But we still don’t know what causes this thing, and we have some suspicions as to why that is. Read on to find out.
pTau leads the liquid biopsy race so far
In the quest for a reliable blood-based biomarker for early Alzheimer’s disease diagnosis, pTau has emerged as a frontrunner, especially pTau217. Recent papers and reviews have added to the evidence backing it up.
Plasma pTau217 making its way to clinical acceptance
The biomarker that currently gets the most love is pTau 217. At this year’s AD/PD 2025 Vienna Conference, the pTau217 plasma test was the star of the show. Its accuracy is at least as high as any other more expensive and/or more invasive tests for both initial diagnosis and later staging. Some highlights from the conference:
pTau217's PPV was high in both referral (89 - 95%) and primary-care settings (82%). (That’s because all patients were symptomatic to some degree. The lower percentage in primary care is likely due to milder/earlier symptomatic cases. In a non-symptomatic over-60 population the PPV would be much lower, because disease prevalence is below 20%.)
Plasma pTau217 can diagnose Alzheimer’s, but cognition does not always decline at the same rate in everyone. A few quite “advanced” cases have only minor cognitive symptoms, and 25% of folks without symptoms have Aβ plaques.
Lower-cost immunoassay tests are as good as traditional mass spectrometry at detecting clinical levels of pTau217.
Adding certain other markers to pTau217 (e.g., digital cognitive scoring at earlier stages, and/or eMTBR243-tau or Aβ42 for later stages) can further improve its clinical accuracy.
pTau217 and pTau181 lead the pack, but might best be used as a rule-out diagnostic
Sweden’s high level of data collection has made it a research leader in Alzheimer’s disease diagnostics research. A Nature Medicine study of 2,148 Swedish adults evaluated six different blood-based biomarkers, including neurofilament light chain (NfL, goes up when neurons die), and glial fibrillary acidic protein (GFAP, increased by brain inflammation). Of those, pTau217 and pTau181 were the best, and the two were essentially equivalent.
The problem is that in this study their positive predictive values (PPVs) were low: 18% for pTau217, 16% for pTau181. Negative predictive value (NPV) was signifcantly higher, at about 96%. (Remember, PPV = the likelihood that someone who tests positive has the disease, and NPV = the likelihood that someone who tests negative doesn’t have the disease.) This study was done in a community setting, where prevalence isn’t that high - typically below 20% - which would lower the PPV.
COMMENTARY: A PPV of 18% is not even close to good enough - it is very close to the average prevalence in this population. Of course, the high NPV may help rule out Alzheimer’s, enabling a more effective diagnosis elsewhere.
A biomarker to predict future cognitive impairment?
No matter how good pTau217 or pTau181 is or isn’t at diagnosing Alzheimer’s, the fact remains - knowing you have this disease doesn’t tell you how fast your symptoms will progress. A March 2025 Nature Medicine paper aimed to solve this issue by finding a better biomarker to predict cognitive impairment.
The best candidate they found was a ratio of two proteins in cerebrospinal fluid (YWHAG1 to NPTX2, involved in the activity of the synapses between neurons). An increase of one standard deviation in the ratio made progression from cognitively normal to mildly impaired 3x more likely, and from mild cognitive impairment to dementia 2.2x more likely.
Add immune-system problems to the list of things that could cause Alzheimer’s
A Cell Reports Medicine paper added autoimmunity and general immune dysregulation to the list of Alzheimer’s root causes. A study of over 300,000 health records revealed that patients with a history of autoimmune disorders had a 40 - 70% higher risk of developing the disease (1.4% in healthy folks vs. about 2.7% in those with an autoimmunity history). The engagement of the immune system is supported by a recent epidemiological report that shingles vaccination reduces dementia incidence by roughly 20%.
COMMENTARY: A strength of this work on the immune system is that it came at the association from both directions – how many autoimmune patients develop Alzheimer’s, and how many Alzheimer’s patients had a history of autoimmunity – so the data are compelling.
Work like this questions whether the root cause of Alzheimer’s disease really is the result of a 20-plus-year development of plaques and tangles of amyloid beta and tau proteins. If it isn’t, then all this research into the proteins that plaques and tangles are made of is just a case of “looking for our car keys under the lamppost.”
We hypothesized previously that plaques and tangles are created from excess proteins that neurons and glia (cells that support neurons) cannot cope with in any normal way. Once protein aggregation is underway, it can become a problem in its own right (but not for everyone - plaques and tangles are found in many non-impaired individuals). (For an interesting perspective on what brought us to the plaque/tangle hypothesis, and some concerns with it, see this thoughtful 2022 editorial.)
The underlying problem is that there is no consensus on either the underlying driving causes of this disease, nor of how many different disease subtypes may be lumped under this one label. We believe that the data will show that there is more than just one version of Alzheimer’s disease, so personalized medicine will become important.
Bird Flu Update: GVN scientists urge robust surveillance
CDC funds dairy worker prevalence research, cancels PPE workshop
Idaho surge continues
Dairy strain spread undetected for months
Scientists from the Global Virus Network emphasized “the importance of robust surveillance systems in a Lancet Regional Health - Americas commentary this week. In addition to recommendations regarding vaccines, protections for farm workers, genetic surveillance and prediction, and response preparation, the group advocated for “testing milk, wastewater, and people working with infected animals” and “self-administered diagnostic tests for farm workers,” CIDRAP reported.
The CDC has awarded more than $3 million to a research team led by the Texas A&M University School of Public Health to assess the presence of H5N1 among Texas dairy workers. However, the agency has canceled a National Academy of Sciences workshop on best practices for the use of PPE by workers at high risk of exposure to the disease.
H5N1 outbreaks in Idaho dairy cattle continue to rise. The 15 herds reported this week bring the state total to 86, second in the nation after California.
According to new research published in Science, the dairy cattle zoodemic originated from a single spillover from birds, probably in Texas in 2023. It then spread undetected among cows for several months, moving across the country as infected cattle were transported from state to state. Since then, it has also spilled over to cats and several other mammal species, and has also spread back into poultry and wild birds. The strain now present in cattle has mutations that make the virus more adapted to mammals, but they show up only at a low frequency so far.
Measles Outbreaks: KS cases spike, NM cases slow
According to Caitlin Rivers’ Force of Infection (paywall), the outbreak in New Mexico “appears to be slowing.” However, Kansas cases spiked this week, with nine confirmed new cases.
The current larger outbreaks are shown below. The cases in Texas, New Mexico, Kansas, and Oklahoma all came from the same source; the outbreaks in Indiana, Michigan, Montana, and Ohio are unrelated.
