Homeostasis model assessment of insulin resistance does not seem to predict response to telaprevir in chronic hepatitis C in the REALIZE trial

Abstract

Baseline homeostasis model assessment-estimated insulin resistance (HOMA-IR), a marker for insulin resistance, has been associated with poor virologic response to peginterferon alpha/ribavirin (PR) in chronic hepatitis C. We evaluated the association between baseline HOMA-IR and pretreatment factors on sustained virologic response (SVR) to telaprevir (TVR) in genotype 1 patients with hepatitis C and prior peginterferon/ribavirin (PR) treatment failure. Patients were randomized to 12 weeks of TVR (750 mg q8h) plus peginterferon (180 μg/week) and ribavirin (1,000-1,200 mg/day) (with or without a 4-week lead-in) followed by PR, or PR alone (PR48), for 48 weeks. Univariate and multiple logistic regression analyses explored the prognostic significance of baseline HOMA-IR alone and adjusted for other pretreatment factors and SVR. The TVR arms were pooled for the purposes of this analysis. In all, 662 patients were randomized; 578 had baseline HOMA-IR and other prognostic data and were included in this analysis. Median baseline HOMA-IR was 2.6 (interquartile range [IQR] 1.7-4.3); 207 (36%), 206 (36%), and 165 (29%) patients had baseline HOMA-IR <2, 2 to <4, and ≥ 4, respectively. Male gender, higher body mass index, triglycerides, gamma-glutamyl transpeptidase, maximum alanine aminotransferase/aspartate aminotransferase, and fibrosis stage were associated with higher baseline HOMA-IR. Baseline HOMA-IR was associated with SVR in univariate analysis, but not after adjustment for other baseline prognostic factors (TVR: OR = 0.95, 95% confidence interval [CI]: 0.71,1.29; PR48: 0.60; 95% CI: 0.25,1.43). In patients with prior PR treatment failure, baseline HOMA-IR correlated with SVR in univariate but not multivariate analyses, suggesting other factors have a more direct causal relationship with virologic response to TVR-based therapy than HOMA-IR.