Identification of Predictors for Treatment Failure in Hepatitis C Virus Patients Treated With Ledipasvir and Sofosbuvir

Abstract

Background: New hepatitis C virus (HCV) therapies report cure rates of ~90% but are expensive. Identification of predictors for treatment failure could help decrease health care costs, limit unnecessary drug exposure and adverse events, and prevent drug-drug interactions. Failure to achieve rapid viral response (RVR), defined as detectable viral load at 4 weeks, has previously been identified as a predictor of treatment failure with some previous HCV therapies. Objective: To evaluate RVR, and other potential variables, as predictors of treatment failure in patients treated with ledipasvir and sofosbuvir (LDV/SOF). Methods: A retrospective, case-control analysis of adult veterans treated with LDV/SOF was conducted. Included patients had a viral load obtained between weeks 3 and 6 of therapy (RVR) and between weeks 12 and 16 after the end of therapy for sustained viral response (SVR12) evaluation. Identified SVR12 failures (viral load detectable) constituted the case population. The control population was randomly selected in a 1:4 case: control ratio from all identified SVR12 successes. Results: In all, 12 SVR12 failures were identified; 48 of 144 SVR12 successes were randomly selected for inclusion. Overall failure rate was 7.7% (12/156). Univariate analysis identified histamine-2 receptor antagonist, previous treatment failure, and pretreatment creatinine clearance (CrCl; Cockcroft-Gault) >90 mL/min as potential predictors of SVR12 failure; these variables were included in the regression model with RVR per protocol. In multivariate analysis, pretreatment CrCl >90 mL/min was independently associated with SVR12 failure (odds ratio = 7.27; 95% CI = 1.33 to 39.72; P = 0.022). Conclusions: Patients with pretreatment CrCl >90 mL/min were more likely to fail SVR12 than patients with CrCl <90 mL/min.