Prognostic Risk Model for Melanoma Based on Dysplastic Nevi Features: Insights into Immune Microenvironment and Treatment

Introduction: Malignant melanoma (MM), a highly aggressive form of skin cancer, is characterized by poor survival outcomes and dysplastic nevi (DN) are welldefined precursors for melanoma. This study developed a prognostic model using molecular features common to DN and malignant melanoma (MM) to evaluate melanoma progression. Methods: RNA-sequencing and clinical data were retrieved from the TCGA and GEO databases. Differentially expressed genes (DEGs) among common melanocytic nevi (CN), DN, and MM were identified using the "limma" R package (|log2 fold change| >1, P < 0.05). The LASSO regression method was used to build the risk model. Immune infiltration, genomic mutations, and drug sensitivity were assessed with CIBERSORT and "maftools" R packages. Results and Discussion: Sixty common DEGs were identified in both MM and DN, clearly distinct from CN. Fourteen prognostic genes were used to construct a risk score (RS) model that independently predicted melanoma outcomes (P < 0.05). High-risk patients showed a profoundly altered tumor immune microenvironment, with reduced CD8⁷ T cells and M1 macrophages, but increased M0 and M2 macrophage infiltration. The immune checkpoint gene NRP1 was strongly correlated with most model genes, while C3, CP, CSF1R, ADA2, and EPSTI1 were broadly associated with multiple immune checkpoints. Together, these immune microenvironment differences likely underlie the divergent clinical outcomes between risk groups. Conclusion: The novel RS model, derived from shared molecular features of DN and MM, effectively predicts prognosis and immune responses in melanoma patients. By linking DN biology to melanoma progression, this model provides actionable insights for risk stratification and therapeutic strategies.