ABSTRACT: Background: Glioma, the most common brain tumor, is a heterogeneous group of glia-derived tumors, the majority of which have characteristics of diffuse infiltration and immunosuppression. The LGALS protein family is a large class of sugar-binding proteins. Among them, LGALS3 has been reported to promote tumor development and progression in some cancers. However, the clinical significance and biological functions of LGALS3 in glioma remain virtually unknown. The purpose of our research is to detect LGALS3 expression and its prognostic value in glioma and reveal the relationship between its expression and the clinico/molecular-pathological features of patients and immune cell infiltration. Methods: LGALS3 protein expression was examined by immunohistochemistry. The mRNA expression data of LGALS3 was downloaded and analyzed from TCGA and Rembrandt datasets. The association between LGALS3 and glioma clinically relevant diagnostic/molecular markers (IDH, 1p19q, ATRX, MGMT, and TERT) was examined using the Chi-Squared (?2) test. The correlation between LGALS3 expression and the infiltration of multiple intra-tumoral immune cell types, including B cells (CD20), T cells (CD4 and CD8), macrophages (CD68), and M2 tumor-associated macrophages (CD163), was evaluated by Spearman correlation analysis. Kaplan-Meier analysis and the Cox regression analysis were applied to evaluate the prognostic value of LGALS3 in glioma. The log-rank test was used to evaluate Kaplan-Meier results for significance. Results: Out of all 304 glioma cases, LGALS3 protein was expressed in 125 glioma cases (41.1%, 125/304), with 69.2% (9/13) in WHO I, 9.8% (8/82) in WHO II, 34.2% (26/76) in WHO III, and 61.7% (82/133) in WHO IV. The expression of LGALS3 was correlated with patient age, WHO grade, PHH3 (mitosis), Ki67 index, IDH, 1p/19q codeletion, and TERT promoter status. LGALS3 was an independent poor prognostic marker in diffusely infiltrating gliomas and was positively correlated with immune cell infiltration, particularly CD163+ tumor-associated macrophages in the TCGA dataset, Rembrandt dataset, and our SYSUCC cohort (R = 0.419, 0.627, and 0.724). Conclusion: LGALS3 was highly expressed in pilocytic astrocytoma, GBM, and IDH wild-type LGG. It served as a poor prognostic marker in diffusely infiltrating gliomas. Based on its prognostic significance and strong correlation with CD163+ TAMs, it may act as an important therapeutic target for human glioma.
