PO.MCB08.04 · 分子与细胞生物学

Integrative transcriptomic profiling reveals stromal activation patterns in gastric cancer beyond HER2 status

海报缩略图:Integrative transcriptomic profiling reveals stromal activation patterns in gastric cancer beyond HER2 status
编号 5913 展板 1 时间 4/21 02:00–05:00 区域 Section 21 主讲 Woo Sun Kwon, PhD
分会场 Genetic and Transcriptomic Dissection of Cancer Evolution
查看完整资料 下载 PDF 登录后可访问当前开放资料 AACR 官方页面 ↗

作者与单位

Woo Sun Kwon1, Chan Hee Park1, Sejung Park1, Choong-Kun Lee2, Poh Zhong Wee3, Jia Hui Liew3, Jingming Chew3, Sun Young Rha1

1Yonsei University College of Medicine, Seoul, Korea, Republic of,2Yonsei University Health System, Seoul, Korea, Republic of,3Auristone Pte. Ltd, Singapore, Singapore

摘要 Abstract

Advanced gastric cancer (AGC) continues to show suboptimal outcomes despite therapeutic progress. Among TCGA-defined subtypes, the genomically stable (GS) subtype remains particularly challenging due to the absence of actionable drivers and its limited responsiveness to immune checkpoint inhibitors. GS tumors frequently exhibit stromal activation-characterized by cancer-associated fibroblasts, desmoplastic fibrosis, and angiogenic remodeling-which contributes to immune exclusion and therapeutic resistance. Although recent clinical activity of trastuzumab deruxtecan has refined HER2 classification, biological determinants of treatment response in non-HER2-driven subtypes, especially GS-like disease, remain poorly understood. Fibrosis- and angiogenesis-related transcriptional scores may capture key tumor-microenvironment programs relevant to AGC biology. We analyzed bulk RNA sequencing data generated from pre-treatment FFPE tumor samples of 120 patients diagnosed with stage IV AGC at Yonsei Cancer Center between 2015 and 2024 (IRB no. 4-2014-0349), categorized into HER2 positive (n=62) and negative (n=58). Differential expression analysis (DEG) and GSVA-based pathway enrichment were used to characterize transcriptomic differences across HER2 groups. Fibrosis and angiogenesis scores were computed and compared among the subgroups, and key findings were validated in the TCGA GC cohort. Across the entire AGC cohort, fibrosis scores (median 29.72, range 0-100) and angiogenesis scores (median 17.92, range 0-98.06) exhibited broad interpatient variability and showed a strong positive correlation (R² = 0.501, p < 0.0001). When samples were mapped to TCGA molecular subtypes, fibrosis scores were highest in GS-likely and were significantly elevated compared with CIN-likely (p = 0.04), a trend reproduced in the TCGA cohort. Low PD-L1 expression was associated with lower angiogenesis scores (p = 0.0059), indicating a relationship between angiogenic activation and immune phenotype. No significant differences in stromal scores were observed across HER2 subgroups, suggesting that stromal activation is largely independent of HER2 status. Within HER2-defined subgroups, HER2-low showed distinct transcriptomic programs characterized by KRAS pathway activation, loss of SNF-related tumor-suppressor activity, enrichment of proliferation-related signaling, and a marked reduction in inflammatory and interferon-related pathways compared with the other two groups. Stromal remodeling, represented by fibrosis and angiogenesis scores, is a major biological feature of AGC and is particularly enriched in GS-like tumors, independent of HER2 status. These findings suggest that fibrosis- and angiogenesis-based stromal programs may improve biological stratification and identify tumor-microenvironment-driven vulnerabilities in advanced gastric cancer.
利益披露 Disclosure
W. Kwon, None.. P. Wee, None.. J. Liew, None.. J. Chew, None.

在会议检索中打开