Researchers at Johns Hopkins University have developed a computational platform that can accurately simulate how tumors respond to combination therapy, allowing oncologists to select optimal treatment regimens without lengthy trial and error. This is particularly important for rapidly progressing hepatocellular carcinoma, the most common form of liver cancer, reports infohub.kz.
The platform integrates mathematical modeling of pharmacokinetics with an agent-based model that tracks individual cell behavior. It accounts for the spatial arrangement of elements within the tumor. When testing the combination of the targeted drug cabozantinib and the immunotherapeutic agent nivolumab, the AI's predictions closely matched actual clinical trial outcomes.
Detailed simulations revealed why treatment was ineffective in some patients: fibroblasts can form a dense physical barrier around the tumor, preventing protective T-lymphocytes from reaching the disease site. Assessing the density and structure of this barrier before therapy could help oncologists identify patients who are unlikely to respond to immunotherapy.