Dr Landgren on Evolving AI Risk-Stratification Methods in Multiple Myeloma

C. Ola Landgren, MD, PhD, a professor, chief of the Division of Myeloma in the Department of Medicine, director of the Sylvester Myeloma Institute, co-leader of the Translational and Clinical Oncology Program, and the Paul J. DiMare Endowed Chair in Immunotherapy at the University of Miami Miller School of Medicine, discussed evolving risk-stratification approaches in the management of multiple myeloma in honor of Multiple Myeloma Awareness Month, which is observed annually in March. He also highlighted the development of the CORAL artificial intelligence (AI) tool and its potential role in guiding personalized treatment decisions for this patient population.

Landgren explained that researchers at the University of Miami have long focused on developing individualized prediction models that integrate disease biology, clinical characteristics, and treatment data to better estimate patient outcomes. He noted that these models were trained on large datasets of prior cases, allowing clinicians to input a patient’s specific features and receive predictions about treatment responses or alternative therapeutic strategies. According to Landgren, this approach supported more informed and dynamic clinical decision-making in multiple myeloma care.

He further described the evolution of the CORAL platform, which is a histomorphologic model that incorporated multimodal data. Although clinical and treatment data are often readily available, Landgren explained that comprehensive genomic profiling is not always accessible due to logistical and resource limitations, along with the time needed to conduct these tests. To address this gap, CORAL was designed to extract meaningful biological insights directly from routine bone marrow biopsy images. By analyzing these images, the model could infer underlying molecular characteristics and translate them into information typically obtained through advanced sequencing technologies.

By leveraging deep learning to interpret standard histopathology slides and predict genomic abnormalities while simultaneously integrating clinical and treatment variables, this version of CORAL could enable real-time, individualized risk assessment and treatment guidance, even in settings where detailed molecular data are unavailable, Landgren continued. The scalable architecture of CORAL allows for continuous refinement and expansion, positioning it as a potentially transformative tool in the management of multiple myeloma and possibly other cancers, he concluded.