Machine learning-based 2-year risk prediction tool in immunoglobulin A nephropathy

Area of Science:

• Nephrology
• Machine Learning in Medicine
• Predictive Analytics

Background:

• Immunoglobulin A nephropathy (IgAN) is a leading cause of chronic kidney disease.
• Early identification of rapid disease progression is crucial for timely intervention.
• Existing models may not fully capture the complexity of IgAN progression.

Purpose of the Study:

• To develop and validate a machine learning (ML) model for predicting 2-year risk of rapid IgAN progression.
• To assess the model's ability to predict long-term kidney-related outcomes in IgAN patients.
• To identify key predictors of IgAN progression using ML techniques.

Main Methods:

• Retrospective cohort study of 1,301 patients with biopsy-proven IgAN.
• Development and external validation of a random forest-based ML model.
• Prediction of primary (30% eGFR decline or ESRD) and secondary (proteinuria improvement) outcomes within 2 years.

Main Results:

• The ML model demonstrated reliable performance in predicting 2-year primary and secondary outcomes.
• Baseline proteinuria was the most significant predictor for both outcomes.
• Predicted risk categories (low, moderate, high) based on the 2-year model correlated significantly with 10-year kidney outcome risks.

Conclusions:

• Machine learning-based 2-year risk prediction models for IgAN progression show reliable performance.
• These models effectively predict long-term kidney outcomes, enabling risk stratification.
• The findings support the use of ML for early identification and management of high-risk IgAN patients.