Regulation of tumor-associated fibroblasts by TACSTD2 in esophageal squamous cell carcinoma and its potential role in immunotherapy resistance.

e16045 Background: Esophageal squamous cell carcinoma (ESCC) has high global incidence and mortality, with poor prognosis due to late diagnosis and limited chemotherapy efficacy. Immune checkpoint inhibitors targeting PD-1/PD-L1 improve survival in some ESCC patients, but benefits are limited to a subset, lacking biomarkers to identify responsive populations. Identifying biomarkers to predict immunotherapy efficacy is crucial for advancing treatment. Methods: Biopsy tissue samples were collected from four ESCC patients before treatment. All patients received three cycles of neoadjuvant therapy with TC regimen plus sintilimab. Post-treatment pathological responses were classified as pCR (pathological complete response), or tumor regression of grades 1 to 3 (P0, P1, P2, and P3). Surgical resection tissue samples were also collected from patients with pCR (P0-after) and tumor regression grade 3 (P3-after). Spatial transcriptomics was performed on these six tissue samples. Additionally, single-cell RNA sequencing data from two untreated patients and one immune therapy-treated ESCC patient from the GSE221561 dataset were included, and cell types were annotated. Results from spatial transcriptomics and single-cell sequencing were analyzed to identify key genes and pathways influencing immunotherapy efficacy, as well as examine cellular distribution, communication, and type-specific effects. The distribution and characteristics of immune cells before and after immunotherapy were explored. Results: (1) Differential genes between non-pCR (P1-P3) and pCR (P0) groups, as well as between Response-good (P0-P1) and Response-poor (P2-P3) groups, included TACSTD2, KRT family, S100A family, membrane proteins, and DSCs. TACSTD2 expression increased across 4 groups, correlating with immunotherapy resistance. Pathway analysis highlighted extracellular matrix organization. (2) Fibroblasts were the second most abundant cell type, with significant differences across 4 groups. These differences matched extracellular matrix-related gene expression, suggesting that tumor-associated fibroblasts (CAFs) influence immunotherapy efficacy by secreting extracellular matrix. (3) TACSTD2 spatial expression significantly overlapped with α-SMA, a CAF marker, suggesting TACSTD2 may induce CAF differentiation. Cell communication analysis showed interactions between ESCC cells and CAFs via TGF-β receptors and ligands. (4) Comparison of non-responding (P3-after) and complete response (P0-after) patients showed differential gene enrichment in extracellular matrix composition and T-cell infiltration, induced by different CAF types. Conclusions: TACSTD2 may regulate CAF differentiation via the TGF-β pathway, promoting immunotherapy resistance through specific extracellular matrix production and T-cell imbalance.