A Nomogram Model Incorporating PVALB Expression for Individualized Prognostic Prediction in Glioma

Background Glioma is a malignant brain tumor with poor prognosis. The tumor microenvironment (TME), particularly immunosuppressive macrophages, drives progression. Parvalbumin (PVALB), a calcium-binding protein, shows antitumor and immunomodulatory potential in other cancers, but its role in glioma is unclear. Objectives To investigate the relationship between PVALB expression, immune infiltration, and clinical outcomes in glioma, and to develop a prognostic nomogram. A retrospective multi-center study integrating bioinformatics analysis of public databases and validation using an independent retrospective patient cohort. Methods Exercise and PVALB-treatment were performed in tumor-bearing mice to examine their anti-tumor effcts. RNA-seq data from 1,987 glioma samples (CGGA, TCGA, Rembrandt) were analyzed. PVALB expression was validated by immunohistochemistry in 210 patient samples from two hospitals. Immune cell infiltration was quantified using ssGSEA. A nomogram incorporating PVALB, IDH1 status, and WHO grade was developed and validated for predicting 1-, 2-, and 3-year overall survival (OS). Results PVALB expression decreased with increasing glioma grade. High PVALB expression correlated with favorable OS across all datasets and was an independent protective factor (HR = 0.31). It was associated with reduced infiltration of immunosuppressive cells (MDSCs, Tregs) and lower expression of immune checkpoints (PD-1, CTLA-4). The nomogram demonstrated good calibration and discrimination (3-year AUC: 0.89–0.90). Conclusion Exercise or PVALB-treatment presented remarkedly anti-tumor effects. PVALB is a novel, independent favorable prognostic biomarker in glioma, associated with a less immunosuppressive TME. The validated nomogram integrating PVALB, IDH1 status, and WHO grade provides a practical tool for risk stratification.