Prognostic role of pre-treatment eosinophilia in NSCLC patients undergoing immune checkpoint inhibitor therapy: A meta-analysis.

Abstract

e20620 Background: Precision management of non-small cell lung cancer (NSCLC) requires biomarkers to predict immunotherapy response. Though eosinophilia can occur as a part of immune-related adverse events, it has also been suggested as a useful simple biomarker for favorable response to immune checkpoint inhibitors (ICI). Given the sparse knowledge and data for the same, we conducted meta-analysis to examine the prognostic significance of eosinophilia in ICI-treated NSCLC patients for progression-free survival (PFS) and overall survival (OS). Methods: A systematic search of PubMed, Google Scholar, and Scopus was done to identify studies published up to 2025 to examine the prognostic value of pre-treatment eosinophilia in NSCLC patients undergoing ICIs. The random effects model was for pooled analyses. Hazard ratios and 95% confidence intervals assessed OS and PFS. Subgroup analyses were based on initial eosinophil levels, location, and treatment. Results: A total of 8 studies with 13,600 patients with a mean/median age of 65 years were included. Pre-treatment eosinophil count showed mixed associations with survival outcomes. For overall survival (OS), unadjusted hazard ratio was 0.79 (95% CI: 0.42–1.51, p=0.48, I²=82.79%), while adjusted hazard ratio was 0.74 (95% CI: 0.53–1.03, p=0.07, I²=73.31%). For progression-free survival (PFS), unadjusted hazard’s ratio was 0.78 (95% CI: 0.54–1.13, p=0.19, I²=70.21%), and adjusted hazard ratio was 0.68 (95% CI: 0.58–0.80, p<0.01, I²=0%), suggesting a significant association between higher eosinophil counts and improved PFS in adjusted analyses. Excluding Takeuchi's study in leave one out sensitivity analysis, the overall odds ratio shifts slightly upward, and the confidence interval narrows, becoming statistically significant. This suggests that Takeuchi's study may have contributed to the broader confidence interval and lack of significance in the overall analysis for OS. Conclusions: Higher eosinophil count is associated with 32% reduced risk of progression. Pre-treatment eosinophilia in NSCLC patients receiving ICIs is significantly associated with improved PFS but shows a borderline association with OS. These results support eosinophilia as a potential prognostic biomarker of PFS for immunotherapy response in NSCLC.