Pharmacokinetic/pharmacodynamic and time-to-event models of ribavirin-induced anaemia in chronic hepatitis C

Clin Pharmacokinet. 2005;44(4):417-28. doi: 10.2165/00003088-200544040-00006.


Background: Interferon (IFN)-alpha and ribavirin combination therapy is the standard treatment for patients with chronic hepatitis C. However, ribavirin induces anaemia, especially by haemolysis, an adverse effect that is dose-limiting.

Objectives: The aim of this study was to determine the relationships between ribavirin exposure and haemoglobin time-course, the time-to-anaemia and the covariates influencing these relationships in a population of patients treated for chronic hepatitis C. In addition, we also intended to establish a simple rule defining the need for dosage adjustment, using data obtained during the first month of treatment.

Methods: A retrospective analysis of 99 patients treated with IFNalpha plus ribavirin, with known dosage administration history, liver biopsy, demographic data, red blood cell counts, haemoglobin level (1037 measurements, median 10 per patient, range 2-31) and serum creatinine during the entire treatment period (178 days, range 53-382 days) was conducted. The data were analysed by a pharmacokinetic/pharmacodynamic population model and Weibull time-to-anaemia model. The rule defining the need for dosage adjustment was as follows: adjustment was needed if haemoglobin at steady state (H(ss)), estimated by the Bayesian method based on data obtained during the first month of treatment, was <12 g/dL for men or <11 g/dL for women.

Results: In both models, anaemia was related to the exposure of erythrocytes to ribavirin at time t (RT in mg/kg/day) by a maximum effect model, with RT(50) (dosage administration rate at which half the maximal effect is reached) approximately 12 mg/kg/day, and the significant covariates were initial haemoglobin level and bodyweight. Performances of a Bayesian prediction of H(ss) based on two early haemoglobin level measurements were encouraging (mean prediction error 0.12 g/dL, precision 0.85 g/dL). The proposed rule for the need of dosage adjustment was able to predict the actual evolution of the dosage regimen in 76% of non-adapted patients and 69% of adapted patients.

Conclusion: The current guidelines for ribavirin dosage administration, based on bodyweight, are adequate, at least in the 45-105 kg range. Results indicate that Bayesian therapeutic monitoring could be helpful in controlling ribavirin-induced anaemia.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adult
  • Aged
  • Anemia / blood
  • Anemia / chemically induced*
  • Antiviral Agents / adverse effects
  • Antiviral Agents / pharmacokinetics
  • Antiviral Agents / pharmacology*
  • Bayes Theorem
  • Creatinine / blood
  • Dose-Response Relationship, Drug
  • Drug Therapy, Combination
  • Erythrocyte Count
  • Female
  • Hemoglobins
  • Hepatitis C, Chronic / drug therapy*
  • Humans
  • Interferon-alpha / therapeutic use*
  • Male
  • Middle Aged
  • Models, Biological
  • Retrospective Studies
  • Ribavirin / adverse effects
  • Ribavirin / pharmacokinetics
  • Ribavirin / pharmacology*


  • Antiviral Agents
  • Hemoglobins
  • Interferon-alpha
  • Ribavirin
  • Creatinine