Abstract Disclosure: M. Ilie: None. M. Lepetit: None. M. Chanal: None. A. Vasiljevic: None. E. Jouanneau: None. O. Gandrillon: None. F. Picard: None. G. Raverot: None. P. Bertolino: None. Context: While gonadotroph tumors lack both prognostic markers and medical treatment options, the tumor microenvironment (TME) stands as a promising therapeutic and prognostic tool. Aim: Identify new TME players in gonadotroph tumors, gain insight into the mechanisms of tumorigenesis driven by these TME cells, and identify new TME-related prognostic markers and therapeutic targets. Methods: Data analysis of single-cell transcriptomics of 6 gonadotroph tumors and spatial transcriptomics of 3 gonadotroph tumors. Immunohistological validation on a cohort of 50 gonadotroph tumors, including 10 paired recurrent tumors. Results: 24,471 sequenced cells from the 6 tumors were analyzed. As expected, we found all previously TME-reported cell types (including stromal, endothelial, monocytes/macrophages, stem, and T cells) and identified mast cells as a new population in gonadotroph tumors. Tryptase immunostaining and whole slides analysis confirmed the presence of mast cells in all 50 gonadotroph tumors (range 0.01-4.33%). Ligand-receptor networks analysis, done with CellPhoneDB, identified bilateral crosstalk between mast cells and endothelial cells. VEGFA was the top ranked secreted ligand signaling from mast cells towards endothelial cells, suggesting a proangiogenic role for mast cells. Consistent with this hypothesis, the percentage of mast cells was very strongly correlated with the percentage of endothelial cells in the 6 tumors (Spearman’s rho=0.94, p=0.01). Spatial transcriptomics analysis showed that mast cells and endothelial cells were frequently colocalized, which was confirmed by double tryptase/CD34 immunofluorescence. We further found mast cells to express more VEGFA than TNF in the 6 sequenced tumors, which is suggestive of a pro-tumorigenic role. Using a clinical cohort, we confirmed the association of mast cells with a bad prognosis, as tumors progressing/relapsing more rapidly after surgery had more mast cells (p=0.002). In addition, more mast cells were present in the recurrent compared to the initial tumor for a same patient (p=0.007). Finally, bioinformatic analysis using Nichenet pinpointed KITLG (acting via KIT) to be one of the probable drivers of VEGFA expression in mast cells. Interestingly, KITLG-KIT was also the top ranked interaction between endothelial and mast cells. Conclusion: We identified and validated the presence of mast cells in gonadotroph tumors. The phenotype of mast cells was indicative of a bad prognosis, which was confirmed in a clinical cohort of 40 patients. Moreover, bilateral crosstalk between mast cells and endothelial cells was indicative of a mutual potentiation, suggesting that KIT and VEGFA targeting might be beneficial for the treatment of gonadotroph tumors. Presentation: 6/1/2024