Eosinophils in cancer immunity: context-dependent plasticity, biomarker potential, and therapeutic opportunities

Eosinophils, traditionally associated with type 2 immunity, are increasingly recognized as context-dependent regulators of cancer immunity. Their effects are not intrinsically anti-tumorigenic or pro-tumorigenic but are shaped by five partially distinct dimensions of context: tumor type and stage, anatomical compartment and microlocalization, cytokine milieu and activation state, the circulating-versus-tissue compartment, and treatment exposure. This Review critically synthesizes mechanistic and clinical evidence defining this functional plasticity, with explicit separation of murine and human evidence throughout. Eosinophils can directly kill tumor cells through granule proteins and reactive oxygen species, recruit and activate cytotoxic lymphocytes, promote vascular normalization, and participate in interleukin-33-dependent anti-tumor circuits. In other settings, they can facilitate metastatic dissemination, angiogenesis, tissue remodeling, and immune suppression. These opposing functions help explain the heterogeneous clinical associations reported across malignancies. Tissue eosinophils at the invasive tumor front are associated with longer time to recurrence in colorectal cancer, whereas peripheral eosinophil dynamics have emerged as candidate biomarkers during immune checkpoint inhibitor therapy in non-small cell lung cancer; eosinophilic malignant pleural effusion may also define a distinct prognostic subgroup. We critically distinguish prognostic, predictive, and pharmacodynamic biomarker claims, and emphasize that most reported associations remain prognostic rather than demonstrated predictors of treatment benefit. We further discuss therapeutic opportunities involving the interleukin-5/interleukin-33/eotaxin axis, induced eosinophils, and eosinophil-assisted cellular therapy, graded by their level of clinical maturity and weighed against the risks of systemic eosinophil activation. We argue that clinically useful eosinophil biomarkers will require integration of abundance with phenotype, spatial localization, tissue compartment, and treatment context. This framework clarifies when eosinophils are protective, harmful, or therapeutically exploitable and highlights priorities for prospective human studies in cancer immunology and precision immunotherapy.