Artificial Intelligence and Multimodality Data Integration Decipher Tertiary Lymphoid Structure Maturity in Gastric...

Area of Science:

• Oncology
• Immunology
• Computational Biology

Background:

• Tertiary lymphoid structures (TLS) are crucial in the gastric cancer tumor microenvironment.
• Clinical assessment of TLS is difficult, hindering research into their maturation.
• Automated tools are needed for TLS identification and understanding their role.

Purpose of the Study:

• To develop an AI tool for quantitative characterization of TLS maturity in gastric cancer.
• To investigate the correlation between TLS maturity and patient survival.
• To elucidate the immune circuits involved in mature TLS formation and function.

Main Methods:

• A transformer-based deep learning model was created for whole-slide image analysis.
• The model was applied to a cohort of 253 gastric cancer patients.
• Integrated analysis included single-cell RNA sequencing, multiplex IHC, flow cytometry, and coculture assays.

Main Results:

• Higher TLS maturity was significantly correlated with improved patient survival.
• A key immune circuit involving CD8+ tissue-resident memory T cells and B cells (CXCL13-CXCR5 axis) was identified in mature TLS.
• This circuit enhances T cell-mediated cytotoxicity via granzyme B upregulation.

Conclusions:

• A clinically applicable AI tool for TLS maturity assessment was established.
• Key immune interactions regulating TLS maturation and antitumor immunity in gastric cancer were uncovered.
• Findings offer potential for guiding treatment strategies in gastric cancer.