Genotype-driven tumor ecosystems drive immune evasion and immunotherapy resistance in melanoma

Genetic and transcriptional alterations in cancer cells shape their interactions with immune and stromal compartments, thereby influencing tumor progression, immune evasion, and response to immune checkpoint inhibitors (ICIs). Yet, how these interactions are spatially organized within melanoma and how they relate to clinical outcome remains incompletely understood. We used spatial multi-omic profiling on melanoma tissues from patients, compared those with loss-of-function mutations in the Neurofibromin 1 ( NF1 ) tumor suppressor gene to those with intact NF1, and integrated these results with functional studies and subsequent gene expression changes observed in patient-derived melanoma models and an immunotherapy-resistant syngeneic mouse model. Spatial multi-omic analysis of melanoma tissues from patients identified 12 meta-niches composed of distinct cell populations with unique molecular features. Although all tumors contained these meta-niches, those enriched in immunosuppressive cancer-associated fibroblasts (CAFs) and macrophages were significantly more common in tumors with NF1 loss. Conversely, niches enriched in cytotoxic CD8 + T cells were significantly reduced. In both human and mouse melanoma, NF1 loss was associated with impaired antigen-presentation programs and reduced CD8 + T cell infiltration. Mechanistically, EGFR signaling emerged as a prominent feature of NF1 -deficient melanoma ecosystems and correlated with reduced expression of antigen-presentation genes and T cell exclusion. EGFR inhibition restored the expression of antigen-presentation programs and enhanced antitumor immune responses in a syngeneic Nf1 knockdown model resistant to ICIs. These findings define spatially organized, functionally distinct meta-niches enriched in NF1 -mutant melanoma that may contribute to their aggressive disease features and poor ICI response. Our study links an understudied melanoma driver to specific immune-evasion mechanisms and identifies EGFR inhibition as a candidate strategy to improve outcomes in patients with NF1 -mutant melanoma resistant to immunotherapy.