Abstract
BACKGROUND: Despite therapeutic advances, metastatic melanoma, and particularly brain metastasis (MBM), remains a lethal burden for patients. Existing single-cell studies offer a more detailed view of melanoma and its microenvironment, which is crucial to improve diagnosis and treatment. RESULTS: We here present a computational reanalysis of single-nucleus data comparing 15 MBM and 10 extracranial melanoma metastases (ECM), considering recent best practice recommendations. We used cell type-specific pseudobulking and omit imputation during patient integration to gain complementary insights. Interestingly, our analysis revealed high homogeneity in tumor cell expression profiles within and between MBM and ECM. However, MBM displayed even higher homogeneity but a more flexible energy metabolism, suggesting a specific metastatic adaptation to the putatively more restricted brain microenvironment. While tumor cells were homogeneous, the metastasis microenvironment, especially lymphocytes and related immune-tumor interaction pathways, exhibited greater divergence between MBM and ECM. Overall, this suggests that major differences between MBM and ECM are potentially driven by variations in their microenvironment. Finally, a comparison of single-cell data to previous bulk studies, including their deconvoluted putative cell types, showed significant differences, potentially causing divergent conclusions. CONCLUSION: Our study contributed to refine the understanding of differences between MBM and ECM, suggesting these are potentially more influenced by their local microenvironments. Future research and therapies could possibly focus on the metabolic flexibility of melanoma brain metastases and patient-specific immune pathway alterations.