Blood Test Predicts Which People With Amyloid Are Likely to Decline Cognitively

A blood test identified which cognitively intact people with brain amyloid pathology were most likely to deteriorate in the next 6 years, longitudinal data showed.

Compared with other measures, plasma phosphorylated tau 217 (p-tau217) best predicted decline on the modified Preclinical Alzheimer Cognitive Composite (mPACC) and the Mini-Mental State Examination (MMSE) with correlation coefficients of 0.41 and 0.34, respectively, reported Niklas Mattsson-Carlgren, MD, PhD, and Oskar Hansson, MD, PhD, both of Lund University in Sweden, and colleagues.

Baseline plasma p-tau217 also was associated with progression to Alzheimer's dementia (HR 2.03, 95% CI 1.57-2.63, P<0.001), the researchers wrote in JAMA Neurology.

Alzheimer's disease starts with a long period of amyloid-beta accumulation without symptoms. P-tau217 may help identify which cognitively unimpaired people with amyloid pathology (preclinical Alzheimer's disease) might benefit most in clinical trials, the researchers suggested.

"It is this subgroup of individuals with preclinical Alzheimer's disease that really needs effective disease-modifying therapies in the future," Hansson told MedPage Today.

"We are convinced that our results have immediate implications for clinical trials evaluating novel therapies in preclinical Alzheimer's disease populations, because the number of included participants can be substantially reduced if only including those with elevated plasma p-tau217 levels," Hansson said.

"However, we also envision using plasma p-tau217 in preclinical Alzheimer's disease in clinical practice in the future when a disease-modifying therapy is approved for clinical use at this early disease stage," he added.

The researchers studied 171 people with preclinical Alzheimer's from the Swedish BioFINDER-1 cohort and validated their findings in 52 people from the Wisconsin Registry for Alzheimer Prevention (WRAP). Mean ages were about 73 and 64, respectively. Some people in BioFINDER-1 had subjective cognitive decline, but all participants in both cohorts were objectively cognitively unimpaired.

Participants had brain amyloid pathology defined by cerebrospinal fluid (CSF) in BioFINDER-1 and by PET scans in WRAP. Besides p-tau217, other plasma measures included p-tau181, p-tau231, glial fibrillary filament protein (GFAP), and neurofilament light (NfL). Data were collected from 2010 to 2020 in BioFINDER-1 and 2011 to 2021 in WRAP.

Primary outcomes were MMSE and mPACC scores over a median of 6 years (range 2-10 years). Both tests measure global cognition; the mPACC also assesses episodic memory and timed executive function.

The researchers adjusted models for age, sex, years of education, apolipoprotein E ε4 (APOE4) allele status, and baseline cognition. They derived cognition slopes using linear regression models with cognitive score as the outcome and time as the predictor, and tested combinations of covariates and biomarkers.

Adjusting for covariates, most biomarkers were associated with mPACC slopes, and all biomarkers except plasma p-tau231 were associated with MMSE slopes. Plasma p-tau217 was the strongest biomarker to predict cognitive decline on both the mPACC (R² 0.41 vs 0.23 for a covariates-only model, P<0.001) and the MMSE (R² 0.34 vs 0.04 for the covariates-only model, P<0.001) in BioFINDER-1. Similar patterns emerged in WRAP.

Sample sizes were reduced in hypothetical clinical trials enriched for individuals with elevated plasma p-tau 217. In clinical trial simulations using mPACC slopes as the outcome in BioFINDER-1, relative sample sizes -- compared with including all eligible participants -- were 79% when including the three highest quartiles of baseline plasma p-tau217, 55% when including the two highest quartiles, and 42% when including the highest quartile. Similar results were seen using MMSE slopes and were validated in WRAP.

The findings complement other data about p-tau217, which has shown strong diagnostic performances for Alzheimer's disease and has discriminated Alzheimer's from other neurodegenerative diseases and normal cognition. Recent trials, including the TRAILBLAZER-ALZ study of donanemab for early Alzheimer's, have included plasma p-tau217 as a metric.

The study has several limitations, Hansson and colleagues acknowledged. Plasma p-tau217 has been associated with both amyloid and tau accumulation. In this study, tau PET was not assessed and the extent to which it was linked with high plasma p-tau217 was unknown.

"However, in cognitively unimpaired individuals, tau PET uptake is usually mild and not readily detectable at the individual level, although there have been group-level increases and associations with future cognitive decline," the researchers noted.

Comment