Predicting the Progression from Asymptomatic to Symptomatic Multiple Myeloma and Stage Classification Using Gene...

Area of Science:

• Computational biology
• Genomics
• Machine learning in medicine

Background:

• Accurate staging of multiple myeloma (MM) is crucial for treatment optimization.
• Predicting progression from monoclonal gammopathy of undetermined significance (MGUS) to symptomatic MM is challenging due to limited data.
• This study addresses the need for improved MM staging and risk stratification of asymptomatic patients.

Purpose of the Study:

• To develop machine learning models for enhanced multiple myeloma staging.
• To stratify asymptomatic patients based on their risk of progression to symptomatic MM.
• To identify key molecular pathways involved in MM pathogenesis and progression.

Main Methods:

• Utilized gene expression microarray datasets for model development.
• Trained and validated multiple myeloma staging models across five independent datasets using AUC metrics.
• Employed two distinct modeling approaches to predict progression in asymptomatic patients, incorporating feature selection and pathway enrichment analyses.

Main Results:

• Multiple myeloma staging models achieved high efficacy with consistent AUC values of 0.9.
• Models for predicting asymptomatic progression demonstrated AUC values exceeding 0.8 across various algorithms.
• Enrichment analyses identified PI3K-Akt, MAPK, Wnt, and mTOR pathways as central to MM pathogenesis.

Conclusions:

• This study presents a novel methodology using gene expression data for MM staging and progression prediction.
• The developed machine learning models show significant potential for clinical application in patient monitoring.
• Early intervention strategies can be enhanced through improved patient stratification and risk assessment.