Selenoprotein GPX3 is a novel prognostic indicator for stomach adenocarcinoma and brain low-grade gliomas: Evidence from an integrative pan-cancer analysis.

ABSTRACT

Background: The antioxidant enzyme GPX3 is a selenoprotein that transports selenium in blood and maintains its levels in peripheral tissues. Aberrant GPX3 expression is strongly linked to the development of some tumors. However, there is a scarcity of studies examining the pan-cancer expression patterns and prognostic relevance of GPX3. Methods: GPX3 expression levels in normal tissues and multiple tumors were analyzed using TCGA, CCLE, GTEx, UALCAN and HPA databases. Forest plots and KM survival curves were utilized to evaluate the correlation between GPX3 expression and the outcome of tumor patients. The prognostic value of GPX3 in LGG was assessed utilizing the CGGA datasets, and that in STAD was tested by TCGA and GEO databases. A nomogram was then constructed to predict OS in STAD using R software. Additionally, the impact of GPX3 on post-chemoradiotherapy OS in patients with LGG and STAD was evaluated using the KM method. The multiplicative interaction of GPX3 expression, chemotherapy and radiotherapy on STAD and LGG was analyzed using logistic regression models. The correlation of GPX3 with the immune infiltration, immune neoantigens and MMR genes were investigated in TCGA cohort. Results: GPX3 exhibited downregulation across 21 tumor types, including STAD, with its decreased expression significantly associated with improved OS, DFS, PFS and DSS. Conversely, in LGG, low levels of GPX3 expression were indicative of a poorer prognosis. Univariate and multivariate Cox models further identified GPX3 as an independent predictor of STAD, and a nomogram based on GPX3 expression and other independent factors showed high level of predictive accuracy. Moreover, low GPX3 expression and chemotherapy prolonged the survival of STAD. In LGG patients, chemoradiotherapy, GPX3 and chemotherapy, and GPX3 and chemoradiotherapy may improve prognosis. Our observations reveal a notable connection between GPX3 and immune infiltration, immune neoantigens, and MMR genes. Conclusions: The variations in GPX3 expression are linked to the controlling tumor development and could act as a promising biomarker that impacts the prognosis of specific cancers like STAD and LGG.