Game-Changer for Asymptomatic Alzheimer’s Trials?

Blood tests to screen for amyloid-beta could cut the need for PET scans in half in asymptomatic Alzheimer’s research trials, a British cohort study suggested.

To find 100 amyloid PET-positive, dementia-free older adults from a population with similar amyloid PET positivity prevalence to the study cohort without plasma screening, 543 PET scans would need to be performed, said Jonathan Schott, MD, of University College London in England, and co-authors.

Using mass spectrometry measures of plasma amyloid-beta would require 623 people to have a blood test, and of those, 243 would proceed to a scan, they wrote in Brain.

“To identify Alzheimer’s before symptoms, the gold standard tests available right now are amyloid PET scans and lumbar punctures, which are either expensive or relatively invasive,” said co-author Ashvini Keshavan, PhD, also of University College London.

“Our paper shows that blood tests could be used in screening to recruit to future trials in asymptomatic individuals and reduce the required number of PET scans by roughly half,” Keshavan told MedPage Today. “This could translate to cheaper and faster trial recruitment.”

The study compared three types of measurements — liquid chromatography-mass spectrometry measures of plasma amyloid-beta, and single molecule array (Simoa) measures of plasma amyloid-beta and of phosphorylated tau181 (phospho-tau181) — in dementia-free members of Insight 46, a substudy of a population-based British 1946 birth cohort, to see how well they predicted a positive cortical ¹⁸F-florbetapir amyloid PET scan.

“We found that amyloid-beta measured by mass spectrometry performs better than amyloid-beta or phospho-tau181 measured by these particular Simoa immunoassays in plasma, in the setting of identifying a positive amyloid PET scan in cognitively normal individuals,” Keshavan said.

“Efficacy is one important aspect, but there are other considerations when looking at using a test for screening, like how well it scales up to large numbers,” she added. “This might still mean there is a role for these Simoa immunoassays in this setting.”

The researchers studied 441 dementia-free older adults with an average age of 71. About half (51%) were women and 18.6% were amyloid PET-positive. APOE4 carriers made up 22% of the amyloid PET-negative group and 57% of the amyloid PET-positive group.

A base model that included age, sex, and APOE4 carrier status had an area under the receiver operating characteristics curve (AUC) of 0.695 (95% CI 0.628-0.762). The two best-performing Simoa plasma biomarkers were amyloid-beta_42/40 (AUC 0.620, 95% CI 0.548-0.691) and phospho-tau181 (AUC 0.707, 95% CI 0.646-0.768), but neither outperformed the base model.

Mass spectrometry plasma tests performed significantly better than any other measure. For amyloid-beta_1-42/_1-40, the AUC was 0.817 (95% CI 0.770-0.864); for an amyloid-beta composite, it was 0.820 (95% CI 0.775-0.866). Mass spectrometry measures of amyloid-beta_1-42/_1-40 detected amyloid PET positivity with 86.6% sensitivity and 71.9% specificity.

Across theoretical ranges of amyloid PET positivity prevalence of 10% to 50%, mass spectrometry measures of amyloid-beta_1-42/_1-40 consistently reduced the number of people who would proceed to scans, with greater cost savings occurring at lower prevalence, Schott and co-authors said.

The study had several limitations, they pointed out. Participants were exclusively white and British, and findings may not apply to ethnically diverse populations. All people in the study were born in the same year, which may have influenced the predictive capacity of models that incorporated age, sex, and APOE4 carrier status. Cerebrospinal fluid (CSF) testing was not available; using CSF as the comparator instead of PET might reduce the cost savings of blood tests, the researchers noted.