A model to predict long-term sustained response to interferon therapy in chronic hepatitis C

Area of Science:

• Hepatology
• Medical Statistics
• Virology

Background:

• Interferon therapy is a common treatment for chronic hepatitis C.
• However, only a small percentage of patients achieve a sustained virologic response.
• Predicting treatment outcomes is crucial for optimizing patient management.

Purpose of the Study:

• To develop and validate a mathematical model for predicting sustained response to interferon-alpha (IFN-alpha) therapy in chronic hepatitis C patients.
• To provide a tool for estimating individual patient probability of treatment success.

Main Methods:

• Logistic regression analysis was employed to develop the predictive model.
• The model incorporated patient demographics (age, sex), disease characteristics (duration), and pretreatment laboratory values (gamma-glutamyl-transpeptidase, alanine aminotransferase), along with hepatitis C virus genotype.
• Model performance was assessed using goodness-of-fit and discrimination metrics (area under the receiver operating curve) on independent development and validation cohorts.

Main Results:

• The developed model demonstrated good performance in both the development (goodness-of-fit P = 0.71, AUC = 0.79) and validation (goodness-of-fit P = 0.15, AUC = 0.78) datasets.
• Key predictors included age, sex, disease duration, pretreatment liver enzyme levels, and HCV genotype.

Conclusions:

• The validated mathematical model effectively estimates the probability of sustained response to interferon-alpha therapy in chronic hepatitis C.
• This predictive model can serve as a valuable decision support tool for clinicians managing chronic hepatitis C patients treated with interferon-alpha.