Related Experiment Video
Updated: Oct 29, 2026

Detection of a CDH1 Rare Transcript Variant in Fresh-frozen Gastric Cancer Tissues by Chip-based Digital PCR
Published on: February 5, 2018
High-resolution Spatial Transcriptomics Reveals Contrasting Tumor States in Two Early Gastric Cancers
Akira Ishikawa1, Hidehiko Takigawa2, Shiro Oka2
1Department of Molecular Pathology, Graduate School of Biomedical and Health Sciences, Hiroshima University, Hiroshima, Japan; a-ishikawa@hiroshima-u.ac.jp.
Background/Aim:
Intratumoral heterogeneity is a major feature of gastric cancer (GC), but its spatial organization remains incompletely understood. We applied high-resolution spatial transcriptomics to two histologically contrasting early GCs.
Materials And Methods:
One formalin-fixed, paraffin-embedded specimen from a gastric adenocarcinoma of fundic-gland type and one from a signet-ring cell carcinoma with signet-ring and non-signet poorly cohesive components were analyzed using Visium HD. All comparisons were made between clusters within each section.
Results:
Spatial clustering recapitulated histological architecture in both tumors. The fundic-gland type adenocarcinoma formed a predominantly single epithelial cluster with lower predicted developmental potential than the normal isthmus, and limited inferred signaling with adjacent tissues. Within the signet-ring cell carcinoma, the poorly cohesive component showed higher predicted developmental potential, increased stem cell-associated markers, and oxidative phosphorylation enrichment, whereas the signet-ring component showed epithelial-mesenchymal transition enrichment (likely reflecting stromal admixture) and higher ANXA10 expression. Distinct APP- and CEACAM-related communication patterns were predicted.
Conclusion:
In these two tumors, high-resolution spatial transcriptomics resolved histologically distinct tumor states and microenvironmental compartments within individual lesions. The findings are descriptive and hypothesis-generating rather than subtype-defining and require validation in larger cohorts.

