Blood-protein study identifies candidates for tracking Alzheimer’s treatment

Washington University School of Medicine in St. Louis announced September 24 that researchers identified blood-protein changes in patients receiving lecanemab for early Alzheimer’s disease. The findings point toward possible tools for monitoring treatment, an important problem in care aimed at preserving cognition and independence, but remain preliminary.

The university reports that the study appeared that day in The Lancet Neurology. Its central question concerns what blood measurements can reveal during treatment: whether they track changes in brain amyloid, cognitive performance or different biological processes.

### Following patients through treatment

According to WashU Medicine, the longitudinal study followed 197 patients with early Alzheimer’s symptoms and evidence of amyloid buildup at its Memory Diagnostic Center for up to two years. Patients provided blood samples approximately every six months and underwent cognitive testing; a subset also had brain imaging.

The researchers compared the treated patients with 458 cognitively unimpaired, amyloid-negative people and 1,312 cognitively impaired people who were not receiving anti-amyloid treatment. These were observational comparisons, rather than randomized treatment assignments.

Alamar Biosciences, whose measurement technology was used, reports that investigators analyzed 2,385 samples from 1,967 participants. Its NULISAseq panel measured 130 plasma proteins covering amyloid and tau pathology, inflammation, neuronal injury and synaptic function.

The company’s account identifies 34 proteins whose levels changed significantly with the number of lecanemab infusions. The changes ran in different directions. It also reports that proteins associated with amyloid clearance largely differed from those associated with subsequent cognitive decline.

That distinction matters when interpreting a potential monitoring test. A measurement associated with removal of a drug’s biological target answers a different question from a measurement associated with future memory or thinking. Neither association, by itself, establishes how much an individual patient benefits from treatment.

### Earlier research shows why interpretation is difficult

A separate study published January 10 in Alzheimer’s Research & Therapy illustrates the challenge. Researchers at Peking Union Medical College Hospital studied 42 lecanemab-treated patients, with 29 completing the six-month treatment evaluation. Assessments included blood biomarkers, cognitive testing and brain imaging, alongside untreated comparison groups.

That study reported reduced amyloid burden and changes in some plasma markers. Nevertheless, cognitive and daily-living measures worsened over follow-up, although progression was slower than in the comparison group. Its blood analyses did not find statistically significant changes in p-tau217 or p-tau181 over six months.

These findings do not directly confirm or contradict the new WashU results: the studies differ in size, follow-up, measurement methods and patient selection. They show why a biomarker’s behavior needs to be tested in its intended clinical setting before it can support decisions.

### From association to a useful test

The practical next step is validation: determining whether a specified measurement, taken at a specified point, reliably answers a clinically useful question in new patients. A promising association discovered across many proteins is the beginning of that process.

For example, a proposed test of amyloid clearance would need comparison with an appropriate imaging reference. A proposed predictor of cognitive decline would need testing against subsequent clinical outcomes. Researchers would also need to establish whether using either test improves decisions beyond information already available.

The present evidence supports continued investigation of blood-based monitoring. It does not establish a replacement for imaging, a rule for stopping treatment or a dependable forecast for an individual patient. For healthspan, the consequential outcome remains preservation of function; changes in laboratory measurements require a demonstrated connection to that goal.