Researchers at The University of Texas MD Anderson Cancer Center have developed a spatial atlas of tertiary lymphoid structures (TLSs) across multiple cancer types, using artificial intelligence. The study was published in the journal Science on May 29, 2026.
TLSs are specialized immune structures that form in tissues, including tumors, and are associated with improved responses to immunotherapy. The atlas, created with a deep learning model, analyzed tissue samples from over 1,300 patients across 13 cancer types, including lung, colorectal, and melanoma.
The AI model, named TLS-Profiler, identified distinct TLS maturation states and their spatial organization within tumors. The researchers found that the presence and maturity of TLSs correlated with patient survival and response to immune checkpoint inhibitors.
This resource is publicly available to help researchers understand how TLSs influence cancer immunity and to guide the development of new immunotherapies. The study was led by Dr. Kunal Rai and Dr. Andrew Futreal.
