Reducing unnecessary tau-PET scans in a memory clinic setting via blood-based biomarker screening

Abstract

Tau-PET provides valuable prognostic information in Alzheimer’s disease but using blood-based biomarkers as a first-line screening tool could reduce costs. In this study, plasma Aβ42/Aβ40, pTau181, pTau217, pTau231, NfL, and GFAP were measured along with tau-PET in 550 memory clinic patients who had subjective cognitive decline, mild cognitive impairment, or any form of dementia from the Swedish BioFINDER-2 study. For each biomarker, cutoffs were established to achieve 90% sensitivity for detecting tau-PET positivity. The percentage of participants who fell below the cutoff was calculated along with the tau-PET positive rate among participants with abnormal biomarker values. Results showed that between 18–51% of tau-PET scans could be saved with plasma screening, with P-tau217 performing best. Performance of GFAP, P-tau181, and P-tau231 was in a second tier. The PET positive rate among those who would receive a tau-PET scan was ~ 75% with plasma screening compared to 37% in the entire study population.