Dissecting the tumour immune microenvironment in merkel cell carcinoma based on a machine learning framework

Area of Science:

• Oncology
• Immunology
• Computational Biology

Background:

• Merkel cell carcinoma (MCC) is a rare neuroendocrine skin cancer with known molecular heterogeneity.
• The tumor microenvironment (TME) of MCC is complex, but its cellular states and ecosystems remain poorly understood.

Purpose of the Study:

• To systematically identify and characterize the cellular states and ecotypes within MCC.
• To elucidate the functional roles and regulatory mechanisms of these cellular components in MCC pathogenesis.

Main Methods:

• A machine learning framework was employed to analyze MCC data.
• Identification and characterization of distinct cellular states across immune cell types, including B cells, CD8 T cells, fibroblasts, and monocytes/macrophages.
• Functional profiling and transcriptional regulation analysis were performed.

Main Results:

• 30 distinct cellular states from 9 immune cell types were identified.
• Functional analysis revealed enrichment of immune and cancer hallmark pathways.
• Four distinct MCC ecotypes with varying patient compositions were discovered.
• Key transcription factors (E2F1, E2F3, E2F7) regulating the TME were identified.

Conclusions:

• This research provides a comprehensive landscape of cellular states and ecotypes in MCC.
• Findings offer insights into intrinsic MCC subtypes and oncogenic pathways.
• The study advances understanding for developing targeted therapeutic strategies for MCC subtypes.