OA08.4. Spatial Transcriptomic Analysis of Non-Dysplastic Barrett's Esophagus From Patients Who Did and Did Not Progress to Advanced Disease
TL;DR
This study suggests early remodeling of the stromal microenvironment and possible differences in the immune makeup with increased myeloid numbers and activation in progressors in patients who progress.
Abstract
Esophageal Cancer: Molecular Biology/Pathology Barrett's esophagus forms and progresses in the setting of chronic inflammation and repeated bouts of tissue damage and regeneration. It has long been speculated the stromal and inflammatory background in Barrett's esophagus contribute to progression, but the cells and pathways exerting this influence are poorly understood. Using a subset of a well clinically annotated cohort of patients with a baseline diagnosis of non-dysplastic Barrett's esophagus and known progression history we performed spatially resolved whole transcriptome analysis using the Nanostring (now Bruker) GeoMx digital spatial profiler. In total 25 patients who later progressed to more advanced disease and 20 patients who did not were analyzed. The selected regions of interest were segmented into epithelial and stromal compartments based on cytokeratin immunostaining and were analyzed separately. Standard quality control and processing of the data was performed using the GeoMx analysis software. For differential gene expression analysis, a linear mixed model (adjusted pvalues< 0.05) and Benjamin Hackenberg multiple testing correction were used. The differential gene expression data was then used to perform Reactome pathway analysis. To estimate cell type composition the SpatialDecon (v1.3) package on the GeoMx analysis suite using the safeTME for tumor reference dataset was used. After all quality control and filtering, 168 epithelial areas of interest (84 from patients who progressed and 75 from patients who did not) were profiled, and 158 stromal areas of interest (86 from progressors and 72 from non-progressors) were profiled. For the epithelium, genes enriched in progressors or non-progressors were mostly involved in cell states and differentiation, except for potential immune-associated genes REG4 and ANPEP. In progressors, pathways related to EMT, MYC signaling, inflammatory and interferon response, and reactive oxygen species were upregulated. In the stroma, ~200 genes were differentially expressed, although the overall effect size was small for most. Several stromal/fibroblast related genes (FBLN1, COL1A1/2, COL6A2, LUM, and IGFBP7) and immune related genes (B2M, HLA-DRA, and C1GC) were upregulated in progressors. Neutrophil degranulation, interleukin and interferon signaling, and pathways associated with stromal remodeling were upregulated in progressors. Macrophage, NK cell, and fibroblast numbers were increased in progressors. Even in histologically non-dysplastic Barrett's esophagus years before progression, some differences in immune and stromal organization pathways can be seen in patients who will later progress compared to those who did not. This study suggests early remodeling of the stromal microenvironment and possible differences in the immune makeup with increased myeloid numbers and activation in progressors. Our (not yet proven) working hypothesis is that in patients who progress, increased neutrophil activation leads to the release of DNA-damaging reactive oxygen species, driving genomic evolution, and increased macrophages promote stromal remodeling to a microenvironment supportive of progression.
