Druggable tumor-nerve-immune circuits in cancer pain: Model systems, human translatability and therapeutic paradoxes

Cancer pain is increasingly recognized as an active output of tumor-nerve-immune biology rather than a passive consequence of tissue damage. Tumor-derived algogens, neurotrophic factors, stress signals and immune mediators can sensitize nociceptors, facilitate perineural invasion, remodel myeloid and lymphoid compartments and recruit central glial mechanisms that may maintain persistent pain. However, the evidence base is asymmetric: many causal circuits were defined in selected bone-metastasis, perineural-invasion, oral-cancer, sarcoma or chemotherapy-induced neuropathy models, whereas applicability to human cancer pain varies by tumor type, pain phenotype, immune contexture and treatment exposure. This critical narrative Review evaluates peripheral, central, immune and pharmacological evidence using explicit model-to-human boundaries. Recent patient studies in oral squamous cell carcinoma link pain with CGRP-positive nerve density, circulating or tissue CGRP, perineural invasion and reduced CD8-positive T-cell density, but these associations do not establish human causality; validated biomarkers of spinal glial mechanisms are still lacking. We examine druggable nodes including NGF-TrkA, SLIT2-ROBO1, CGRP-CALCRL/RAMP1, CSF1R-dependent myeloid or glial signaling, voltage-gated sodium channels and opioid-neuroimmune interactions. Clinical trial and cohort examples illustrate therapeutic paradoxes in which analgesia must be balanced against joint safety, tissue repair, host defense and potentially confounded associations with immunotherapy outcomes. We finally propose a phenotype-matched model-selection framework and a druggable circuit-state strategy for patient-stratified pharmacological development.