Pan-cancer neural regulation pattern with implications for patient stratification and immunotherapy.

The interplay between the nervous system and cancer plays a central role in regulating oncogenesis. We depict the pan-cancer landscape of neuro-tumor interactions by integration of >10,000 The Cancer Genome Atlas (TCGA) bulk and 239 pan-cancer spatial transcriptomics, in combination with 115 bulk and >120,000 single-cell checkpoint-inhibitor-treated transcriptomes. Neural signals are significantly correlated with various clinical features, and neural cancer subtypes have been characterized. High neural signals are associated with cancer-associated fibroblast infiltration and immune exclusion. Spatiotemporal transcriptome analysis reveals that tumor core regions present significantly higher neural signals and further unveils the pivotal modulators essential for tumor growth via ligand-receptor interaction networks. Neural signals are predictive biomarkers for drug treatment responses, particularly cancer immunotherapy. We constructed a comprehensive data resource Neuro_Cancer, to investigate neuro-tumor interactions in diverse cancer types. Our systematic dissection of the landscape of the neuronal regulation pattern provides a deeper understanding of tumorigenesis and potential therapeutic opportunities.