Although amyloid PET imaging is the gold standard for identifying amyloid positivity to diagnose Alzheimer’s disease (AD), testing for a panel of plasma biomarkers is useful for screening. A study published in Alzheimer’s Research & Therapy used machine learning to find that a single biomarker, phosphorylated tau 217 (pTau217), offers an alternative means of screening that is more cost-effective than a panel of biomarkers.
Last year, plasma biomarkers were added to updated criteria for diagnosing Alzheimer’s disease. These included pTau217, pTau181, Aβ42/40, glial fibrillary acidic protein, and neurofilament light chain, with pTau217 showing the strongest link to amyloid PET positivity. The authors of the current study speculated that testing for a single biomarker could reduce the cost and simplify interpretation if research shows that it is accurate.
To investigate, the authors used machine learning models to measure the performance of a full panel of plasma biomarkers and rank the importance of each individual biomarker for: