We thank Wang et al. for their interest in our manuscript and their helpful comments. The question of the low sustained virological response (SVR) rate in females belonging to the placebo group in our study is worth further exploration. Patients infected by hepatitis C virus (HCV) genotype 1 and showing insulin resistance belong to a very difficult-to-treat group. In a recent study,1 only 30% of patients with HCV genotype 1 and fasting glucose levels >5.6 mmol/L (which could be a threshold for insulin resistance) achieved sustained response. Moreover, in our initial cohort, only a quarter of females infected by genotype 1 and showing homeostasis model assessment (HOMA) score >2 reached SVR.2 Thus, the SVR rate in females with hepatitis C genotype 1 and HOMA >2 treated with standard of care was as expected.
Taking their second and third points together, Wang et al. highlighted several confounding factors involving the beneficial effect of metformin in females: metformin promotes weight loss and improves insulin sensitivity; insulin resistance depends on HCV replication and metabolic syndrome; and patients with weight-loss could be receiving higher doses of ribavirin and, thus, reach higher SVR rates. Definitive responses to these questions are beyond the aims and the statistical power of our study. Average daily ribavirin dose was similar in females taking metformin or placebo: 15.9 (standard deviation [SD] = 1.69) mg/kg/day versus 15.4 (SD = 1.26) mg/kg/day (P = not significant). Ribavirin dose had been modified only to correct for anemia. As stated in the original manuscript, females receiving metformin had a mean body weight decrease of 1.3 kg (SD = 5.4) versus 1.1 kg (SD = 5.6) in women receiving standard care (P = not significant). Although changes in body weight or ribavirin dose did not influence SVR, improving insulin sensitivity was related to SVR. Lastly, >80% of females receiving metformin reached virus clearance at week 24, and this did not correlate with weight loss or with insulin sensitivity at this time point. A relationship between metformin use and viral decline was seen in females. Metformin does not have an antiviral effect and, thus, it seems to be due to an improvement in the antiviral activity of the combined peginterferon and ribavirin therapy. Females that reached HOMA <2 but had no weight loss showed significantly higher viral decline (−5.3 [SD = 0.79] versus −4.2 [SD = 1.8] log10 HCV RNA; P = 0.004).
In hepatitis C, insulin resistance depends on viral factors such as genotype and HCV replication as well as host factors such as body weight and/or steatosis. However, the weighting of these factors differs between patients.3 Improving viral and/or host factors can help improve insulin sensitivity. The final mechanism by which metformin improved SVR in females remains unclear, and further studies are warranted to elucidate how insulin-sensitizer drugs could improve response in patients with chronic hepatitis C treated with pegylated interferon plus ribavirin.