Test-retest variability of plasma biomarkers in Alzheimer's disease and its effects on clinical prediction models

INTRODUCTION The effect of random error on the performance of blood-based biomarkers for Alzheimer's disease (AD) must be determined before clinical implementation. METHODS We measured test-retest variability of plasma amyloid beta (A beta)42/A beta 40, neurofilament light (NfL), glial fibrillary acidic protein (GFAP), and phosphorylated tau (p-tau)217 and simulated effects of this variability on biomarker performance when predicting either cerebrospinal fluid (CSF) A beta status or conversion to AD dementia in 399 non-demented participants with cognitive symptoms. RESULTS Clinical performance was highest when combining all biomarkers. Among single-biomarkers, p-tau217 performed best. Test-retest variability ranged from 4.1% (A beta 42/A beta 40) to 25% (GFAP). This variability reduced the performance of the biomarkers (approximate to Delta AUC [area under the curve] -1% to -4%) with the least effects on models with p-tau217. The percent of individuals with unstable predicted outcomes was lowest for the multi-biomarker combination (14%). DISCUSSION Clinical prediction models combining plasma biomarkers-particularly p-tau217-exhibit high performance and are less effected by random error. Individuals with unstable predicted outcomes (gray zone) should be recommended for further tests.

Automated summary

The study found that blood-based biomarkers have high performance in clinical prediction of Alzheimer's disease, with p-tau217 performing the best. Test-retest variability has some impact on the performance of biomarkers, but it has less effect on models with p-tau217. Further testing is recommended for individuals with unstable predicted outcomes.