Eosinophil counts as a prognostic marker in glioblastoma: a retrospective cohort study

BACKGROUND: Glioblastoma is a highly aggressive primary brain malignancy characterized by poor median survival despite trimodal therapy. While established prognostic factors exist, significant heterogeneity in patient outcomes persist. This study aims to evaluate whether white-cell differentials predict oncologic outcomes in glioblastoma (GBM). METHODS: We retrospectively analyzed 85 GBM patients treated with standard surgery, radiotherapy (60 Gy in 30 fractions), and temozolomide. Clinical data (age, Karnofsky performance status, extent of resection, MGMT methylation, corticosteroid use, dosimetric data) and hematologic indices (absolute lymphocyte count (ALC), neutrophil-to-lymphocyte ratio (NLR), absolute eosinophil (AEC) and basophil counts, etc.) were collected at baseline and up to 3 months post-chemoradiation. RESULTS: Median overall survival (OS) in our cohort was 22.5 months (95% CI, 20.4-32.2). On univariate analysis, higher AEC correlated with improved OS at baseline (HR 0.68; p = 0.013) and at 2 months (HR 0.58; p = 0.019). In the combined multivariable model-controlling for age, extent of resection, MGMT status, KPS, steroid use, and % of brain volume receiving 60 Gy) -baseline AEC remained independently prognostic (aHR 0.64; p = 0.021). CONCLUSIONS: Baseline eosinophils were independently associated with improved survival. Such findings highlight the importance of host immunity and baseline eosinophils as a potential prognostic marker of OS in GBM and warrant larger, prospective studies on prognostic marker validation.