Multi-omics profiling and AI-driven clinically deployable risk models in MGUS and smoldering myeloma

Area of Science:

• Hematology
• Oncology
• Bioinformatics

Background:

• Monoclonal gammopathy of undetermined significance (MGUS), smoldering multiple myeloma (SMM), and multiple myeloma (MM) represent a spectrum of plasma cell disorders.
• Current risk stratification relies on clinical, laboratory, and cytogenetic markers, which inadequately capture disease complexity and limit predictive accuracy.

Purpose of the Study:

• To explore the limitations of existing risk stratification models for MGUS and SMM.
• To examine the potential of multi-omics data and artificial intelligence (AI) in enhancing predictive accuracy for these plasma cell disorders.

Main Methods:

• Review of current literature on multi-omics technologies (genomics, transcriptomics, proteomics, metabolomics).
• Analysis of artificial intelligence (AI) and machine learning (ML) applications in integrating multi-omics data for disease progression prediction.
• Discussion of challenges in clinical adoption, including data integration, standardization, privacy, bias, and regulatory aspects.

Main Results:

• Multi-omics technologies provide deeper insights into the molecular drivers of plasma cell disorders.
• AI and ML, when combined with multi-omics data, show potential for identifying novel biomarkers and improving the precision of disease outcome prediction.
• Advancements offer a path toward dynamic, personalized risk stratification frameworks.

Conclusions:

• AI and multi-omics hold significant promise for revolutionizing risk stratification in MGUS, SMM, and MM.
• Overcoming technical, ethical, and regulatory challenges is crucial for the clinical implementation of these advanced predictive tools.
• Future integration into patient care requires a roadmap addressing data complexity, standardization, and evolving AI capabilities.