Eosinophils in non-small cell lung cancer immunotherapy: from dual clinical associations and antitumor mechanisms to a prospective validation framework

Immune checkpoint inhibitors (ICIs) are an important treatment for advanced non-small cell lung cancer (NSCLC), but the lack of reliable biomarkers for outcome stratification limits their precise clinical application. Peripheral blood eosinophil-related parameters, including absolute eosinophil count (AEC), relative eosinophil count (REC) and dynamic changes during treatment, have been studied as candidate biomarkers in recent years because of their convenient detection and dynamic monitoring. At the same time, tumor tissue eosinophilia (TATE), as a direct manifestation of a local immune response, has also attracted attention. This review focuses on the clinical associations of the above indicators in the immunotherapy of NSCLC: the association between baseline AEC and REC levels and patient survival, the association of early on-treatment eosinophil elevation with clinical benefit, the association between baseline AEC and immune-related adverse events, particularly checkpoint inhibitor-related pneumonitis, and whether TATE may provide supplementary information to peripheral blood indicators. In addition, the potential role of eosinophils in the association of immune-related adverse events (irAEs) with efficacy is explored, raising the hypothesis that eosinophils may lie at the intersection of antitumor immunity and immune-mediated toxicity. Based on the available evidence, we further propose a framework for prospective validation that incorporates prespecified baseline stratification, early on-treatment dynamic monitoring, endpoint assessment, and multivariable modeling, with the aim of facilitating the transition of eosinophil-related biomarkers from retrospective associations to prospective clinical validation. In addition to clinical evidence, this paper also summarizes the potential mechanisms through which eosinophils are recruited, activated, and participate in antitumor immunity within the tumor microenvironment. Existing evidence suggests that the above indicators have potential clinical utility, but problems such as inconsistent cut-off values, lack of standardization of dynamic monitoring, and insufficient prospective verification have not yet been resolved. The NSCLC-specific mechanisms underlying eosinophil-associated effects require further investigation.