Site Editors

Shaji K. Kumar, MD

Prashant Kapoor, MD, FACP


Individualized, Risk-Prediction Model for Patients With Newly Diagnosed Multiple Myeloma

By: Amy MacDonald, MS
Posted: Wednesday, May 8, 2024

Overall survival outcomes for newly diagnosed patients with multiple myeloma vary widely. Francesco Maura MD, of the Sylvester Comprehensive Cancer Center, University of Miami, and colleagues sought to create a comprehensive risk assessment model to help predict patient outcomes. Their data, published in the Journal of Clinical Oncology (JCO), used clinical and genomic data to create a first-of-its-kind computerized individualized risk-prediction model—called IRMMa—to inform treatment decisions for newly diagnosed patients.

“By integrating 20 highly relevant genomic features, IRMMa allows better identification of primary refractory and early progressive myeloma patients compared to current staging systems such as R-ISS and R2-ISS. IRMMa subsequently boosts overall survival prediction accuracy and could guide clinicians in adjusting for treatment and consolidation strategies,” stated JCO Associate Editor, Suzanne Lentzsch, MD, PhD, of Herbert Irving Comprehensive Cancer Center, New York.

For their study, the researchers assembled a large training set (n = 1,933) and a validation set (n = 256) of patients with newly diagnosed multiple myeloma. By entering clinical, demographic, genomic, and therapeutic data from these patients into an artificial intelligence–assisted computer model, the researchers created their patient-specific prediction tool.

While building this model, the researchers screened 132 myeloma-related genomic features and identified 20 whose inclusion improved the tool’s accuracy. (These features included chr.1q21 gain or amplifications, deletion of chr.1p32, TP53 loss, NSD2 translocation, APOBEC mutational signatures, and several key copy number signatures.) Of note, however, the prediction tool was found to have the capacity to generate survival estimates even in the absence of genomic data, which may ultimately surpass the predictive accuracy of standard assessment tools (eg, variations of the International Staging System [ISS]). Finally, among the numerous clinical features tested, age and ISS were the most important for the model’s accuracy, whereas the impact of sex, Eastern Cooperative Oncology Group (ECOG) score, and race seemed to be limited.

Disclosure: For full disclosures of the study authors, visit

By continuing to browse this site you permit us and our partners to place identification cookies on your browser and agree to our use of cookies to identify you for marketing. Read our Privacy Policy to learn more.