In a new study published today, scienceresearchers at the University of Texas MD Anderson Cancer Center have developed a spatial atlas of specialized immune structures called tertiary lymphoid structures (TLS) across multiple cancer types. This first-ever atlas reveals that TLS maturation status, spatial location within tumors, and composition can provide clinically meaningful information about cancer prognosis and treatment response.
This research was led by Linghua Wang, MD, Professor of Genomic Medicine, Executive Director and Director of the Center for Cellular and Linguistic Intelligence, Associate Member of the James P. Allison Institute™, and Co-Director of the Area of Focus with the Oncology Data Science Institute at UT MD Anderson.
In this study, the research team developed a scalable artificial intelligence (AI) framework to detect, profile, and classify TLS from spatial-omics data and routine pathology slides. They also created a composite scoring system to more effectively stratify patients by prognosis and treatment response across different cancer types and treatment settings.
Prior to this study, most of the focus on TLS as a biomarker centered solely on whether TLS was present and, in some cases, mature. Here we show you that you can dig even deeper. TLS within tumor tissue is much more complex than that. Their maturational state, spatial location, and composition within the tumor can tell us important information about the tumor immune microenvironment, therapeutic response, and clinical outcome. ”
Linghua Wang, MD, Professor of Genomic Medicine, UT MD Anderson
What is TLS and why is it important in cancer?
The immune system response to tumors is a highly coordinated effort within the tumor microenvironment. In some tumors, immune cells come together to form organized structures called tertiary lymphoid structures (TLS). These structures act as local immune “hubs”, recruiting B cells, T cells, antigen-presenting cells, and other supporting cells that help coordinate anti-tumor immune responses.
Previous studies have shown that TLS, especially the more mature ones, are often associated with improved patient outcomes and improved responses to immunotherapy in various cancer types. However, the presence of TLS alone doesn’t tell the whole story.
This study takes some of that understanding further. Tumors can contain TLS with vastly different levels of organization, cellular composition, and spatial relationships within tumor cells, and researchers have shown that these differences convey important biological and clinical information.
What does this research bring to our understanding of TLS?
Although it is well known that TLS is important in cancer, our understanding of the cellular and molecular heterogeneity of TLS remains limited, especially in its natural spatial context across large cohorts of human tumor samples.
This study addresses that gap by developing a scalable computational framework to accurately detect, comprehensively profile, and classify TLS from spatial-omics data. Leveraging this framework, the team built a pan-cancer spatial atlas of TLS across 340 samples from 12 cancer types. This atlas allowed us to explore the landscape of TLS in tumor tissues, define how key features of TLS change, and identify transcriptional programs associated with TLS maturation.
This study found that TLS varies widely across organizations. As the TLS matures, it becomes more organized and its immune, stromal, and vascular components undergo coordinated changes. Furthermore, their proximity to tumor cells is associated with spatial gradients of tumor signaling.
These findings suggest that TLS maturation and spatial context are associated with distinct tumor signaling environments and may reflect important features of the tumor immune microenvironment.
To make these insights more scalable, the team also developed an AI framework to rapidly identify and classify TLS from pathology images routinely used in daily clinical care. Training this AI model will allow you to easily translate your TLS analytics process into the clinic, making the process significantly faster and more scalable. The AI framework allowed the researchers to go a step further and assess 25,088 TLSs from over 3,000 full-slide images across 10 independent cohorts to develop a TLS “composition score” for a given patient’s tumor.
This composition score captures not only the number of TLS but also the maturation status of TLS within the tumor. This method significantly outperformed traditional TLS measurements in stratifying patients by prognosis and treatment response, suggesting that a more detailed view of TLS biology that takes maturation status into account may provide more clinically meaningful information than the presence of TLS alone.
What are the next steps for this research?
The TLS composite scoring approach needs to be validated in prospective clinical trials. If successful, this framework could support broader integration of TLS profiling into routine pathology workflows as it uses routine pathology images.
This finding also raises important biological and therapeutic questions. One important observation from this study is that many TLS within tumor tissue remain immature, with some located away from the tumor region rather than within or adjacent to tumor cells. This suggests that future studies should investigate how to promote TLS towards a more mature functional state and how to enhance its spatial interactions with tumor cells and the broader tumor microenvironment. These efforts may help identify therapeutic strategies that promote effective TLS formation and maturation and enhance TLS-associated antitumor immune responses.
sauce:
University of Texas MD Anderson Cancer Center
Reference magazines:
Cho K.S. Others. (2026). Cancer-wide spatial atlas of tertiary lymphoid structures. science. DOI: 10.1126/science.adz2742. https://www.science.org/doi/10.1126/science.adz2742
