Recent advances in neuroimaging and cerebrospinal fluid (CSF) biomarker assays have provided evidence of a long preclinical stage of Alzheimer's disease (AD). This period is being increasingly targeted for secondary prevention trials of new therapies. In this context, the interest of a noninvasive, cost-effective amyloid-β (Aβ) blood-based test does not need to be overstated. Nevertheless, a thorough validation of these bioanalytical methods should be performed as a prerequisite for confident interpretation of clinical results. The aim of this study was to validate ELISA sandwich colorimetric ABtest40 and ABtest42 for the quantification of Aβ40 and Aβ42 in human plasma. The validation parameters assessed included precision, accuracy, sensitivity, specificity, recovery, and dilution linearity. ABtest40 and ABtest42 proved to be specific for their target peptide using Aβ peptides with sequence similar to the target. Mean relative error in the quantification was found to be below 7.5% for both assays, with high intra-assay, inter-assay, and inter-batch precision (CV <9.0% on average). Sensitivity was assessed by determination of the limit of quantification fulfilling precision and accuracy criteria; it was established at 7.60 pg/ml and 3.60 pg/ml for ABtest40 and ABtest42, respectively. Plasma dilution linearity was demonstrated in PBS; however, dilution in a proprietary formulated buffer significantly increased the recovery of both Aβ40 and Aβ42 masked by matrix interactions, allowing a more comprehensive assessment of the free and total peptide levels in the plasma. In conclusion, both assays were successfully validated as tools for the quantification Aβ40 and Aβ42 in plasma.
Keywords: Alzheimer’s disease; amyloid-β peptide; biomarker; immunoassay validation; plasma.