Alzheimer's disease is the leading cause of human dementia. The lack of diagnostic tests and limited therapeutic options has driven the search for endogenous biomarkers. The INNO-BIA AlzBio3 assay is a multiplex flow-based immunoassay measuring Aβ42, tau, and p-tau in cerebrospinal fluid (CSF). This study assesses assays performance under varying bead count (BC) parameters. Original method validation parameters at 100 BC were acceptable. Reanalyses performed at 3, 10, 25, and 50 BCs were compared to 100 BC data by ANOVA, Bland-Altman analysis, evaluation of concordance correlation coefficients, and frequency distribution of coefficient of variation (CV) ranges. Method validation characteristics were acceptable with 100 BCs. Equivalency for 25 and 50 versus 100 BCs was demonstrated, but not for 3 and 10 BCs. A general trend of decreasing agreement between decreasing BCs and the 100 BC reference resulted in decreases in concordance coefficients ρ(c) . The frequency of CV values greater than 20% increased with decreasing BCs, and CV values of 5% or less decreased with decreased BCs. Statistical analyses demonstrate that BCs of 3 and 10 are not equivalent with the reference and should not be used as a basis for determination of Aβ42, tau, and p-tau concentration in human CSF.
© 2011 Wiley-Liss, Inc.