MS Assay Precision and Its Implication for Longitudinal Biomarkers
Mon 3:00 PM - Track 2: Computational, Statistical, and Epidemiological Considerations for Mass Spectrometry in the Clinical Laboratory
Steve Skates
Harvard Medical School
Steve Skates
Recently, MS assays for proteins in plasma have been developed using techniques such as selective reaction monitoring (SRM). SRM is an appealing approach to assaying new targets identified through discovery proteomics due to its relatively rapid development time compared to the long expensive process of producing monoclonal antibody pairs that recognize distinct epitopes of the protein required for development of immunoassays. Serial dilutions of a target protein spiked-in at a known concentration and measured with replication provide data from which to fit SRM calibration curves on the log-log scale. Estimated concentrations of the target protein in new biospecimens and their standard errors are derived from the calibration curves.

One application of MS assays is to evaluating biomarker candidates of early cancer detection. Measurement in longitudinal specimens will identify candidates that rise earliest above background variation. However, in proteins measured longitudinally, background variation spans a wide range, in one biomarker from 7% to 40% CV. When SRM assay CVs exceed 15%, a large fraction of cancer cases that produce a biomarker will have substantially delayed identification, leading possibly to not identifying the candidate as an early detection biomarker. This application emphasizes the importance of achieving single digit CVs for SRM and related MS assay techniques.