72. Utilizing Spatial Genomics to Explore Early Molecular Pathways in Radiation-Induced Fibrosis

INTRODUCTION: Breast cancer is the most commonly diagnosed cancer, affecting one in eight women. Radiation therapy after surgical removal of the tumor provides survival benefits. However, a potential side effect of radiation is radiation-induced fibrosis (RIF), which can cause pain and lead to poor aesthetic outcomes. The biological mechanisms underlying RIF are not fully understood, warranting further investigation to identify potential intervention pathways for preventing the symptom severities. In this preliminary study, we performed spatial transcriptomics analysis of skin and capsule tissues collected post-radiation from breast reconstruction patients who were later found to have RIF in a 7-year chart review. The study aimed to identify the spatial location of differentially expressed genes previously identified using bulk RNA sequencing. MATERIALS AND METHODS: We collected tissue samples from two breast cancer patients who underwent bilateral mastectomy and reconstruction and unilateral radiation. At the time of final reconstruction (post-radiation), samples were taken from the non-radiated and radiated reconstructed breast tissues. These patients were later identified as having RIF. Initially, these samples underwent bulk RNA sequencing and histological analysis, followed by spatial transcriptomics using the Visium CytAssist platform. RESULTS: Bulk RNA sequencing revealed that CLCA2, COL4A5, and COL6A6 were differentially expressed within the capsules, potentially indicating reduced micro vascularization—a known outcome of radiation therapy. Genes CXCL9 and PTCHD4 were upregulated in the skin, suggesting a possible enhancement of the immune response to radiation. Conversely, multiple keratin-related genes (KRT6B, KRT17, KRT25, KRT28, and KRT75) were downregulated. While bulk RNA sequencing gives an overall gene expression change in the studied tissue, spatial transcriptomics allows one to identify local expression levels. DISCUSSION: The spatial transcriptomic data revealed that 26 genes were overexpressed, and 10 were underexpressed by 1 log-fold or more within the radiated skin tissue. Among them, several keratin genes, including KLF, KRT1, KRT10, KRT2, KRT14, KRT15, FLG, FLG2, and LOR, were overexpressed and spatially segregated within the epidermis of the radiated skin tissue. Interestingly, cell-cell desmosome junctional genes—DSC1, DSC3, DSP, PKP1, PERP, and COL17A1—were also overexpressed within the same region. These results indicate that radiation affects epidermal gene expression, potentially contributing to the mechanism for the clinically observed epithelial thickening. In the radiated capsule, a set of genes related to contractile fibers, including TTN, TNNT1, TNNT3, TNNC1, TNNC2, TNNI1, TNNI2, ACTA1, ACTN2, MYH2, MYLPF, NEB, and DES, were differentially expressed by 5-fold or more, indicating possible mechanistic pathways to reduce the effects of RIF. Many troponin and sarcomere organization genes were also differentially expressed within this tissue. Interestingly, both techniques identified that KRT17—a type I keratin involved in epidermal development and repair after tissue injury—was downregulated, suggesting it could be a potential target protein for combating radiation-related effects of epithelial thickening.