The Transcriptomic Landscapes of Appendiceal Primary and Metastatic Tumors are Distinct.

BACKGROUND

Improved understanding about the pathobiology of appendiceal cancers (AC) and resulting metastasis is required for the development of novel treatments. The tumor microenvironment in AC is heterogeneous and incompletely characterized. The objective of this study was to leverage spatial high-plex technology to evaluate the transcriptomic landscape of epithelial and stromal cells in primary AC tumors, adjacent normal appendix, and corresponding peritoneal metastasis.

METHODS

A tissue microarray (TMA) containing cores from 14 unique patients having matched primary tumor, adjacent normal appendix, and peritoneal metastases was analyzed with digital spatial profiling (NanoString, GeoMx) using pancytokeratin (PCK) to delineate stroma (PCK-) from epithelium (PCK+). Then RNA sequencing was performed to measure transcript abundance separately within the stromal and epithelial compartments.

RESULTS

Transcriptomic analysis demonstrated differences between tumor and stromal compartments in both primary tumor and metastatic sites. Primary and metastatic tumor stroma (PCK-) demonstrated greater expression of ribosomal biogenesis pathways than normal appendiceal tissue. Primary and metastatic tumors were generally similar with respect to transcription. However, within the epithelial compartment (PCK+), peritoneal metastases exhibited upregulated cytoskeletal and collagen metabolism pathways/genes compared with primary tumor.

CONCLUSIONS

The study data indicated that although appendiceal peritoneal disease is transcriptionally similar to the primary tumor, potentially important distinctions exist between metastatic and primary disease. Differences appear to be driven predominantly by changes in collagen metabolism at the peritoneal site. A better understanding of both tumor and stromal compartments of metastatic disease will be essential to improving therapeutic options, specifically systemic treatment, which is characteristically ineffective.