Anoikis-related long non-coding RNA signatures to predict prognosis and small molecular drug response in glioma: based on multi-database analysis

Abstract

Abstract Background Gliomas are intrinsic brain tumors that originate from neuroglial progenitor cells. Traditional treatments, including surgery, chemotherapy, and radiotherapy, have limited improvement in prognosis for patients with gliomas, and recurrence rates remain high.The construction of prognostic model can predict the development and treatment effect of glioma, which is of great clinical significance. Methods Anoikis play a key role in the critical stages of tumor development, metastasis, and cancer cell dissemination. Based on TCGA database and CGGA database, the LASSO model is constructed with Anoikis-related lncRNAs as the core. Combined with clinicopathological features, univariate- and multivariate COX analysis were used to verify the effectiveness of the model. Despite the existence of various prognostic models, none of them are truly suitable for clinical application. The model we have constructed provides an option for clinical application. Results We constructed a risk model with 8 ARlncRNAs(LINC00519, AC140481.1, LINC00928, HOXA-AS2, CRNDE, ACAP2-IT1, USP30-AS1, TMPO-AS1) at its core and validated their high accuracy in predicting overall survival. We also confirmed their association with clinicopathological features. Studies of drug sensitivity and immunological associations suggest that it could provide more precise guidance to clinicians. Conclusion Our study elucidated a prognostic prediction model of glioma by 8 Anoikis-related long non-coding RNAs.High-risk patients have a short survival time and a pro-tumor immune microenvironment.