Medical News Observer

Key Points:

• Researchers built IRMMa using genetic, clinical, and treatment data from 1933 patients, incorporating deep learning to adapt and improve with new data.
• The model outperformed existing prognosis tools with a c-index of 0.726.
• IRMMa’s unique approach of combining various data types aims to revolutionize clinical trials and personalized treatment options for multiple myeloma patients.

Overview

A multicenter collaboration spearheaded by Sylvester Comprehensive Cancer Center at the University of Miami Miller School of Medicine developed an Individualized Risk Model for Myeloma, IRMMa, for predicting individualized prognoses in newly diagnosed multiple myeloma patients. The Journal of Clinical Oncology published this research. 

IRMMa Model

Researchers built IRMMa using genetic, treatment, and clinical data from 1933 newly diagnosed patients, identifying key driver genes influencing tumor growth. They identified 12 distinct groups by analyzing genomic drivers, expanding on previous molecular classifications based on gene expression. IRMMa model focuses on individual patient risk and specific treatments. Thanks to its deep learning capabilities, the model can improve by adding new data, including recent antibody-based therapies. 

The IRMMa model demonstrated superior accuracy in predicting overall survival compared to existing models, with a c-index of 0.726. A clinical trial further validated the model’s effectiveness, revealing significant treatment variance across the 12 genomic groups, which could influence the selection of therapies.

This IRMMa model is available online, mainly for researchers. In the future, as more patient data and emerging treatments are integrated, It can improve its accuracy and usefulness. IRMMa is a pioneering approach that enables personalized therapeutic decisions for newly diagnosed multiple myeloma patients. It incorporates various data types such as clinical, demographic, genomic, and therapeutic data.

Reference

Maura, Francesco, Arjun Raj Rajanna, Bachisio Ziccheddu, Alexandra M. Poos, Andriy Derkach, Kylee Maclachlan, Michael Durante, et al. 2024. “Genomic Classification and Individualized Prognosis in Multiple Myeloma.” Journal of Clinical Oncology, January, JCO.23.01277. https://doi.org/10.1200/JCO.23.01277.