Background: Cerebrospinal fluid (CSF) amyloid-β (Aβ42) is a well-established biomarker for Alzheimer disease. Several immunoassays for Aβ42 exist but differ in absolute concentrations and may suffer from matrix interference, thereby hampering interlaboratory comparisons and the use of general cutoff levels. Together with the IFCC Working Group on CSF Proteins, we developed a candidate reference measurement procedure (RMP) for Aβ42.
Methods: The antibody-independent candidate RMP was based on solid-phase extraction and isotope-dilution LC-MS/MS. The candidate RMP used 2 differently stable isotope-labeled Aβ42 peptides for calibration in human CSF, an important aspect since there was no analyte-free matrix available. Because no CSF certified reference material (CRM) exists, we used a nonlabeled Aβ42 standard, the concentration of which was determined by amino acid analysis. We performed measurements on a high-resolution quadrupole-Orbitrap hybrid instrument. The results were compared with a method run in a second laboratory with triple quadrupole instrumentation.
Results: The candidate RMP allowed quantification of CSF Aβ42 from 150 to 4000 pg/mL. Validation of the method showed a recovery of 100% (15%), intraassay and interassay imprecision of 5.0% and 6.4%, respectively, and an expanded uncertainty of 15.7%. No analytical interferences or carryover were detected.
Conclusions: This method will help set the value of CSF Aβ42 in a CRM, which could be used to harmonize Aβ42 assays and facilitate the introduction of general cutoff concentrations for CSF Aβ42 in clinical trials and practice.
© 2014 The American Association for Clinical Chemistry.