Using Stepwise Pharmacogenomics and Proteomics to Predict Hepatitis C Treatment Response in Difficult to Treat Patient Populations. Issue 3 (22nd August 2018)

• Record Type: Journal Article
• Authors: Naggie, Susanna; Clement, Meredith; Lusk, Sam; Osinusi, Anu; Himmel, Tiffany; Lucas, Joseph E.; Thompson, Will J.; Dubois, Laura; Moseley, M. Arthur; Clark, Paul J.; Kottilil, Shyam; Patel, Keyur
• Is Part Of: Proteomics. Volume 13:Issue 3(2019)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2018-08-22
• Subjects: hepatitis C virus -- human Immunodeficiency virus -- IL28B -- pharmacogenomics -- proteomics; Proteomics -- Periodicals; 572.605
• DOI: 10.1002/prca.201800006
• Languages: English
• ISSNs: 1862-8346

Abstract

Abstract : Purpose: In the interferon era of hepatitis C virus (HCV) therapies, genotype/subtype, cirrhosis, prior treatment failure, sex, and race predicted relapse. Our objective is to validate a targeted proteomics platform of 17 peptides to predict sustained virologic response (SVR). Experimental design: Stored plasma from three, open‐label, trials of HIV/HCV‐coinfected subjects receiving interferon‐containing regimens is identified. LC‐MS/MS is used to quantitate the peptides directly from plasma, and IL28B genotyping is completed using stored peripheral blood mononuclear cells (PBMC). A logistic regression model is built to analyze the probability of SVR using responders and nonresponders to interferon‐based regimens. Results: The cohort ( N = 35) is predominantly black (51.4%), male (86%), and with median age 48 years. Most patients achieve SVR (54%). Using multivariable models, it is verified that three human corticosteroid binding globulin (CBG) peptides are predictive of SVR in patients with the unfavorable IL28B genotypes (CT/TT). The model performs better than IL28B alone, with an area under the curve of 0.870. Conclusions and clinical relevance: In HIV/HCV‐coinfected patients, three human CBG peptides that accurately predict treatment response with interferon‐based therapy are identified. This study suggests that a stepwise approach combining a genetic predictor followed by targeted proteomics can improve the accuracy of clinical decision‐making.