Background:  Ribavirin (RBV) plasma concentration was shown to correlate with hemoglobin decrease, and with early virological response and sustained virological response in HIV⁺ patients infected by HCV genotype 1 or 4. Cut-offs of RBV concentration for efficacy and toxicity has been investigated only in the early phase of treatment. Therefore, our aim was to study RBV concentration cut-offs of sustained virological response and of hemoglobin decrease throughout the whole course of therapy.

Methods:  HIV⁺/HCV⁺ patients treated with RBV and pegylated interferon alpha (pegIFNα) -2a or -2b were prospectively evaluated. Qualitative and quantitative HCV RNA, hemoglobin levels, and RBV C_trough were measured at baseline and week 2, 4, 12, 24, 36, and 48. HCV RNA was also measured at 4, 12, and 24 weeks after the end of therapy. Efficacy analysis was performed on patients with a definitive virological outcome (sustained virological response, relapse, and non-response), whereas for toxicity analysis drop-outs were considered until the last available observation. Logistic and time-weighed linear regression were used. ROC was used to calculate cut-off values.

Results:  We included 42 patients, of which 17 (40.5%) with genotype 1 or 4. RBV weight-adjusted dose was 12.7 mg/kg (11.8 to 14.3). HCV RNA and hemoglobin at baseline were 6.32 log (5.94 to 6.53) and 14.9 g/dL (14.1 to 16), respectively. Overall, 20 patients (47.6%) achieved early virological response and 18 (42.8%) sustained virological response:  12 (28.6%) were non-responders and 12 (28.6%) dropped out. HCV genotype resulted to be an independent predictor of early virological response and sustained virological response (p = 0.001 and p = 0.002, respectively). Overall, no correlation between RBV C_trough and early virological response nor sustained virological response was found. However, in patients with genotypes 1 or 4, RBV C_trough independently predicted early virological response (p = 0.049), whereas a trend toward significance was found for sustained virological response (p = 0.07). A RBV C_trough cut-off of 1600 ng/mL was found to be associated to both early virological response (χ² = 7.1, p = 0.015) and sustained virological response (χ² = 4.89, p = 0.027). Overall hemoglobin variation was –1.8 g/dL (–3.2 to –1.0), and the lowest hemoglobin value reached was 12.1 g/dL (11.3 to 12.9). Higher RBV C_trough correlated with hemoglobin decrease (R = –0.468, p = 0.007), and predicted hemoglobin decrease >3 g/dL at logistic regression analysis (p = 0.038). ROC analysis provided a RBV C_trough cut-off of 2300 ng/mL for  prediction of toxicity.

Conclusions:  Our study confirmed a relationship between RBV exposure and both efficacy and toxicity. Moreover, we found RBV C_trough cut-offs for both sustained virological response in genotypes 1 and 4 and long-term overall hematological toxicity.  Further studies are warranted in order to evaluate the possible use of such values as a therapeutic drug-monitoring tool in the clinical setting.