Non-invasive tumor microenvironment evaluation and treatment response prediction in gastric cancer using deep learning radiomics

• DOI: 10.1016/j.xcrm.2023.101146
• Publication Date: 2023-08-08T15:10:56Z
• Authors: Yuming Jiang; Kangneng Zhou; Zepang Sun; Hongyu Wang; Jingjing Xie; Taojun Zhang; Shengtian Sang; Md Tauhidul Islam; Jen-Yeu Wang; Chuanli Chen; Qingyu Yuan; Sujuan Xi; Tuanjie Li; Yikai Xu; Wenjun Xiong; Wei Wang; Guoxin Li; Ruijiang Li
• Keywords: Deep Learning; Stomach Neoplasms; Chemotherapy, Adjuvant; Tumor Microenvironment; Humans; Immunotherapy; Article; 3. Good health

The tumor microenvironment (TME) plays a critical role in disease progression and is key determinant of therapeutic response cancer patients. Here, we propose noninvasive approach to predict the TME status from radiological images by combining radiomics deep learning analyses. Using multi-institution cohorts 2,686 patients with gastric cancer, show that model accurately predicted an independent prognostic factor beyond clinicopathologic variables. further predicts benefit adjuvant chemotherapy for localized disease. In treated checkpoint blockade immunotherapy, clinical improves predictive accuracy when combined existing biomarkers. Our enables assessment TME, which opens door longitudinal monitoring tracking therapy. Given routine use radiologic imaging oncology, our can be extended many other solid types.