Background:  Pegylated interferon (pegIFN) -α + ribavirin (RBV) is the standard of care treatment for chronic hepatitis C (HCV). IFN-induced psychiatric toxicity can impair quality of life, reduce treatment adherence, and increase drop-out rates. We investigated the utility of gene-expression profiles to predict development of psychiatric toxicity in HIV/HCV-co-infected patients undergoing treatment for HCV.

Methods:  A total of 33 HCV/HIV-co-infected patients were treated with pegIFN-α2b (1.5 µg/kg/week) and RBV (1 to 1.2 g/day for 48 weeks). Peripheral blood mononuclear cells (PBMC) transcriptional profiles were analyzed using Affymetrix HG U133A arrays and class prediction performed using K-nearest neighbor prediction analysis with 10x cross-validation. The patients had regular interviews, and depression scores were evaluated with Beck`s Depression Inventory (BDI) regularly. Patients were classified into 2 groups, those with (n = 21) and without (n = 12) any psychiatric toxicity after initiation of therapy. Psychiatric toxicity was defined as onset of symptoms that met Diagnostic and Statistical Manual for Mental Disorders (DSM-IV) criteria for a new diagnosis during therapy, a BDI score >15 or an increase in BDI score of ≥50% at any time during treatment, symptoms requiring dose adjustment of existing psychotropic medications or addition of new psychotropic medications during therapy, or worsening symptoms that threatened IFN dose reduction or discontinuation.

Results:  Between the 2 groups of patients, 138 genes were differentially regulated with a p-value of <0.05 using one-way ANOVA. Using class prediction analysis, the development of an IFN-induced psychiatric symptom could be predicted prior to initiation of therapy by 55 genes with 90% accuracy. Of the 138 genes that were differentially expressed, several were identified as typical IFN-response genes.   

Conclusions: These data suggest that PBMC gene expression levels can be used to predict development of IFN-induced psychiatric symptoms in HCV/HIV-co-infected patients undergoing combination therapy. Validation of these findings in larger clinical trials will be invaluable to develop selective preventive therapeutic strategies and improve anti-HCV responses. Finally, the functional analysis of the differentially regulated genes that were identified in this study could lead to the discovery of novel drug targets to improve the efficacy of and adherence to HCV therapy.