Re-Clustering and Profiling of Digestive System Tumors According to Microenvironment Components

Yongwei Wang, Sen Guo, Zhihong Chen, Bing Bai, Shuo Wang, Yaxian GaoFrontiers in Oncology 2021 · DOI

Abstract

Our research provides a new strategy based on the microenvironment components for the reclassification of digestive system tumors, which could provide guidance for prognosis judgment and treatment response prediction like Immune Checkpoint Blockade (ICB). By identifying six distinct tumor microenvironment (TME) clusters, we demonstrate that stromal components are critical determinants of clinical outcomes.

Introduction: Beyond Organ-Centric Classification

Digestive system tumors, including those of the esophagus, stomach, liver, pancreas, and colon, are aggressive and often diagnosed at advanced stages. While Immune Checkpoint Blockade (ICB) therapy has revolutionized oncology, it remains effective for only a small fraction of patients with these cancers. Traditional classification by organ of origin fails to explain the vast differences in treatment response between patients with the same cancer type. This study addresses this gap by focusing on the tumor microenvironment (TME)—the complex cellular neighborhood that shapes tumor behavior.

Methods and TME Profiling

The researchers analyzed bulk RNA-sequencing data from 1,526 patients across eight digestive tumor types from The Cancer Genome Atlas (TCGA). Using the ESTIMATE package to calculate immune and stromal scores, and algorithms like CIBERSORT and MCP-counter to quantify specific cell populations, they moved toward a holistic TME-centric view.

Distribution of TME cell type scores in the READ (rectum adenocarcinoma) cohort, illustrating the diversity of the microenvironment components quantified in the study.

Unsupervised K-means clustering was then used to group patients into six distinct clusters based on their TME composition, regardless of their primary tumor site.

Key Insights: The Stromal Paradox

The study revealed that the "neighborhood" often trumps the "address." Tumors from different organs frequently shared identical TME profiles. A critical finding was the identification of Cluster 3, which exhibited high immune infiltration but also high stromal scores.

Typically, high immune infiltration is associated with better prognosis. However, Cluster 3 showed significantly poorer survival. This "stromal paradox" suggests that dense stromal components, such as fibroblasts and endothelial cells, create a physical or chemical barrier that induces an "inhibitory immune status," effectively exhausting immune cells before they can attack the tumor.

Survival analysis in the STAD (stomach adenocarcinoma) cohort demonstrates that high fibroblast infiltration—a major stromal component—is associated with significantly poorer survival (p = 0.036).

Results and Clinical Utility

  • Stromal Score as a Biomarker: In validation cohorts (IMvigor210 and GSE78220), patients with high stromal scores had significantly lower response rates to anti-PD1/PDL1 treatments. The stromal score outperformed traditional immune scores in predicting immunotherapy response in these contexts.
  • Pathway Enrichment: Cluster 3 was heavily enriched in pathways related to TGF-beta, VEGF, and focal adhesion—mechanisms known to promote tumor invasion and immune evasion.

Conclusion

This TME-based reclassification offers a potential new biomarker for patient stratification. While these bioinformatic findings require prospective clinical validation, they suggest that targeting the stroma (e.g., with TGF-beta inhibitors) could be a key strategy to "unlock" the microenvironment and improve the efficacy of immunotherapy in resistant digestive tumors.

Summary updated

Abstract

Our research provides a new strategy based on the microenvironment components for the reclassification of digestive system tumors, which could provide guidance for prognosis judgment and treatment response prediction like ICB.
BibTeX
@inproceedings{wang2021re,
  author = {Yongwei Wang and Sen Guo and Zhihong Chen and Bing Bai and Shuo Wang and Yaxian Gao},
  title = {{Re-Clustering and Profiling of Digestive System Tumors According to Microenvironment Components}},
  booktitle = {Frontiers in Oncology},
  year = {2021},
  doi = {10.3389/fonc.2020.607742},
}
Figure