According to this study, melanoma metastases show distinct metabolic and hypoxic profiles across organs, which may contribute to differences in response to immunotherapy. The findings suggest that targeting metabolic pathways could help overcome resistance to immune checkpoint inhibitors in patients with liver metastases.
Abstract
Background: Different metastatic sites in advanced melanoma show distinct responses to immune checkpoint inhibitors (ICIs). In patients treated with ICI, liver metastases are associated with reduced responses, while lung metastases are associated with improved responses. Metabolic reprogramming, including enhanced glycolysis and hypoxia-associated signaling, is a hallmark of tumor progression. Increased hypoxia-inducible factor 1 (HIF1α) expression has been associated with poorer prognosis and reduced response to ICI across several solid malignancies. We characterized and compared metabolic and hypoxic profiles across melanoma metastatic sites to provide insight into organ-specific responses to ICI.
Methods: We described metabolic and hypoxic profiles in tumor samples from 182 melanoma patients across metastatic sites, including liver (n=24), lung (n=35), brain (n=41), lymph node (n=25), and subcutaneous tissue (n=57). Multiplex immunohistochemistry staining (n=92; CD3, SOX10, CAIX, HIF1α, GLUT1) and transcriptomic analysis of bulk RNA sequencing (n=158), were performed to compare cell densities and gene expression across these sites.
Results: Liver metastases showed upregulation of glycolytic and metabolic pathways compared with other sites, though 97% of pathways were also characteristic of normal liver tissue; SLC2A1 (GLUT1), however, was selectively elevated in liver metastases relative to normal liver, consistent with tumor-specific metabolic reprogramming overlayed on the hepatic niche. Liver, lung, and brain metastases exhibited higher proportions of GLUT1+CD3+ T cells compared with lymph node and subcutaneous metastases. In both liver and lung, GLUT1+ tumor regions had lower T cell infiltration than GLUT1- regions. Lung metastases had the highest HIF1α expression, while liver showed reduced expression of CAIX compared with other sites.
Conclusions: These results demonstrate that the metabolic features of melanoma metastases and their accompanying immune cells differ across anatomical sites, possibly reflecting both the intrinsic biology of the host organ as well as tumor-specific metabolic reprogramming. This variability may contribute to organ-specific differences in response to ICI. They highlight the potential for therapies targeting metabolic pathways to overcome liver-mediated ICI resistance.
Reference:
Conway JW, Paver EC, Beckwith J, Yaseen Z, Gide TN, Quek C, et al. Site-specific metabolic and hypoxia-associated profiles in metastatic melanoma. Journal for ImmunoTherapy of Cancer. 2026;14:e015623. https://doi.org/10.1136/jitc-2026-015623