Independent study demonstrates amyloid probability score accurately indicates amyloid pathology

Abstract

Background

The amyloid probability score (APS) is the model read‐out of the analytically validated mass spectrometry‐based PrecivityAD ^® blood test that incorporates the plasma Aβ42/40 ratio, ApoE proteotype, and age to identify the likelihood of brain amyloid plaques among cognitively impaired individuals being evaluated for Alzheimer's disease.

Purpose

This study aimed to provide additional independent evidence that the pre‐established APS algorithm, along with its cutoff values, discriminates between amyloid positive and negative individuals.

Methods

The diagnostic performance of the PrecivityAD test was analyzed in a cohort of 200 nonrandomly selected Australian Imaging, Biomarker & Lifestyle Flagship Study of Aging (AIBL) study participants, who were either cognitively impaired or healthy controls, and for whom a blood sample and amyloid PET imaging were available.

Results

In a subset of the dataset aligned with the Intended Use population (patients aged 60 and older with CDR ≥0.5), the pre‐established APS algorithm predicted amyloid PET with a sensitivity of 84.9% (CI: 72.9–92.1%) and specificity of 96% (CI: 80.5–99.3%), exclusive of 13 individuals for whom the test was inconclusive.

Interpretation

The study shows individuals with a high APS are more likely than those with a low APS to have abnormal amounts of amyloid plaques and be on an amyloid accumulation trajectory, a dynamic and evolving process characteristic of progressive AD pathology. Exploratory data suggest APS retains its diagnostic performance in healthy individuals, supporting further screening studies in the cognitively unimpaired.