Comprehensive Biomarker Discovery Platform Reveals Qualified MRM Assay for Prediction of INF/Ribavirin Treatment Response in Hepatitis C Patients
Year: 2011
Type: article
Abstract: The current standard of care for chronic Hepatitis C (CHC) results in sustained viral response in only half of patients and also has significant side effects. Improved pre-treatment clinical predictors of response are needed to individualize treatment and select individuals most appropriate for new therapy. We have identified 10 peptides to 6 proteins which can differentiate non-responders from responders with high accuracy based on pre-treatment samples. Serum samples were randomized, immunodepleted (MARS 14), and digested with trypsin prior to LC-MSE or LC-MRM analysis. LC-MSE analyses were performed on a Waters nanoAcquity LC and QToF Premier, while MRM analyses were performed on nanoAcquity and Xevo TQ. Rosetta Elucidator® was used to quantify all LC-MSE data. PLGS v2.4 was used to make peptide identifications. MRM method generation and sample quantitation was performed with Skyline v0.6. The response signature was initially identified from the analysis of 96 samples (Duke Hepatology Clinical Research database) by LC-MSE on a QToF mass spectrometer, using sparse latent factor regression analysis. This yielded a group of almost 400 candidate peptides which were part of 3 metaprotein predictors; these peptides were then curated based on set selection criteria to give 86 target peptides for MRM analysis. MRM analysis of the original 96 samples yielded a final group of 10 peptides which maintained Bonferroni-corrected statistical significance for predicting treatment response. The assay has been qualified by a blinded analysis of an all comers clinical trial sample set (PEDS-C, NIH, n=51) using the MRM method, which yielded an AUROC of 0.91 with a sensitivity of 0.828 and specificity of 0.786. The analysis of an additional CHC clinical trial cohort (Chariot, n = 243) is ongoing. This presentation will utilize CHC treatment response as a case study on the critical analytical, statistical, and clinical cohort requirements to successfully perform biomarker discovery and verification.
Source: Europe PMC (PubMed Central)
Authors: Dubois, L.G., Patel, K., Lucas, J.E., McCarthy, J., McHutchison, J.G., Moseley, M.A., Thompson, J.W.
Institutions: Duke University, Duke Clinical Research Institute
Language: English
Subfield: Spectroscopy
Field: Chemistry
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W286271040