Statistical models for predicting response to interferon-alpha and spontaneous seroconversion in children with...

Area of Science:

• Hepatology
• Virology
• Pediatric Medicine

Background:

• Chronic hepatitis B (CHB) affects children worldwide.
• Predicting treatment response and spontaneous seroconversion in pediatric CHB is challenging.
• Accurate prognostic models are needed to guide therapy decisions.

Purpose of the Study:

• To develop and validate prognostic models for predicting outcomes in children with CHB.
• Identify children likely to respond to interferon-alpha (IFN-alpha) therapy.
• Identify children likely to achieve spontaneous seroconversion.

Main Methods:

• Utilized data from a multinational controlled trial of 70 children treated with IFN-alpha and 74 untreated children.
• Employed SMILES (similarity of least squares) network analysis for multidimensional data integration.
• Incorporated clinical data including age, gender, liver enzymes (AST, ALT), HBV DNA levels, IFN-alpha dose, pretreatment trends, liver biopsy, race, and transmission mode.

Main Results:

• SMILES models achieved high prospective prediction accuracy: 89% for IFN-alpha responders, 96% for non-responders, 100% for seroconverters, and 93% for non-seroconverters.
• Key predictors for IFN-alpha response included HBV DNA pretreatment direction, baseline HBV DNA, IFN-alpha dose, and gender.
• Key predictors for spontaneous seroconversion included ALT pretreatment direction, baseline HBV DNA, age, and AST pretreatment direction.

Conclusions:

• Prognostic models using SMILES demonstrate significant potential for predicting IFN-alpha response and spontaneous seroconversion in children with CHB.
• These models can assist clinicians in determining the likelihood of treatment success or natural viral clearance.
• Personalized treatment strategies for pediatric hepatitis B can be enhanced through these predictive tools.