Abstract 2456: Artificial intelligence-powered spatial analysis of tumor-infiltrating lymphocytes predicts prognosis in gallbladder cancer after resection

Abstract Background: Tumor-infiltrating lymphocytes (TIL) play an essential role in various cancer types. However, their prognostic significance in resected gallbladder cancer remains largely unknown. Methods: This study included 225 patients who underwent curative resection for gallbladder cancer. H&E-stained whole slide images were analyzed using the Lunit SCOPE IO, an artificial intelligence (AI)-powered spatial TIL analyzer. Based on the density of intratumoral TILs in the cancer epithelium and stromal TILs in the cancer stroma, three immune phenotypes (IP) were defined: immune-inflamed phenotype (IIP), immune-excluded phenotype (IEP), and immune-desert phenotype (IDP). Survival outcomes were analyzed, with subgroup analysis conducted for early-stage (stage 0-II) and advanced-stage (stage III-IV) disease. Results: Of the 225 patients, 58 (25.8%) had IIP, 151 (67.1%) had IEP, and 16 (7.1%) had IDP. During a median follow-up of 71 months, 82 deaths (36.4%) were recorded. IIP and IEP were significantly associated with more prolonged disease-free survival (DFS) and overall survival (OS) compared to IDP. Conclusions: Immune phenotypes identified through AI-powered spatial TIL analysis were associated with DFS and OS in resected gallbladder cancer, particularly in late-stage disease. These findings highlight the prognostic potential of immune phenotyping, especially in advanced disease. Univariate and Multivariate Cox Regression Analyses for DFS and OS Citation Format: Hyemin Kim, Young Hoon Choi, So Jeong Yoon Yoon, Cheolyong Joe, Se-Hoon Lee, Sang Hyun Shin, In Woong Han, Jin Seok Heo, Yo Han Jeon, Kee-Taek Jang, Kwang Hyuck Lee, Jong Kyun Lee, Kyu Taek Lee, Hongbeom Kim, Jookyung Sophie Park. Artificial intelligence-powered spatial analysis of tumor-infiltrating lymphocytes predicts prognosis in gallbladder cancer after resection [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 2456.