A new LC‐MS/MS assay for the quantification of Aβ40 and Aβ42 in plasma: validation and clinical performance

Abstract Background Early detection of Alzheimer’s disease (AD) represents an unmet clinical need. Cerebrospinal fluid (CSF) biomarkers can potentially fulfil this need, but specimen collection is invasive. Blood‐derived biomarkers are preferable, and several studies have shown an association between the Aβ42/40 ratio and amyloid PET status. However, blood diagnostic tests for AD are challenging due to interlaboratory differences in analytical techniques and preanalytical handling. Here we describe the development and validation of a LC‐MS/MS assay to quantify Aβ40 and Aβ42 from plasma and assess clinical performance. Method Aβ40 and 42 peptides were simultaneously immunoprecipitated prior to protein digestion and LC‐MS/MS analysis. Blood collection studies were conducted to determine the optimal specimen type, tube type, and storage conditions. Clinical validation studies were performed in a cohort of 209 clinically assessed individuals with PET data (67 cognitively normal, 66 with early MCI, 37 with late MCI, and 39 with AD). Result Both Aβ40 and 42 showed a linear measurement range of 10‐2500 pg/mL, with R 2 values ≥0.99. Assay precision ranged from 2.6‐9.0% for Aβ40, Aβ42, and Aβ42/40. Plasma specimens were stable for up to 5 freeze‐thaw cycles, 6 hours at room temperature, 5 days refrigerated, at least 32 days frozen and 9 months at ultra‐low temperature. Compared to other tube types, blood collected in K 2 EDTA tubes had an average analyte response that was 1.6‐fold higher for both Aβ40 and 42 (range 1.4‐4.0 fold). Peptide recovery was higher in plasma than in serum (1.2‐1.4 fold higher for Aβ40 and 42). Variation in the internal standard signal, a measure of overall assay variability, was <10% for K 2 EDTA compared to >20% for other tube types (range 24.6‐33.3%). The test accurately differentiated PET‐positive from PET‐negative individuals (AUROC=0.862 95% CI 0.814‐0.910). At the Aβ42/40 cut‐point of 0.160, the assay has a sensitivity of 71% and specificity of 89% for detecting PET‐positivity, which is suggestive of AD. Conclusion We optimized the analytical performance of an LC‐MS/MS assay for plasma Aβ40 and Aβ42. Preanalytical factors, including specimen collection tube types, were key for reproducible and accurate quantification. Performance characteristics were similar to those reported in the literature using CSF specimens.