A Innovative Prognostic Symbol Based on Autophagy-Related Long Noncoding RNAs Signature Predicts in Low-Grade Glioma

Abstract Autophagy is a highly conserved lysosomal degradation process essential in tumorigenesis. However, the involvement of autophagy-related lncRNA in low-grade gliomas (LGG) remains a pending question. Efforts were made to establish an autophagy-related lncRNA signature prognostic in LGG patients, and to explore the behind potential function. We used Univariate Cox, Least Absolute Shrinkage and Selection Operator (Lasso), and Multivariate Cox regression models were designed to establish an autophagy-related lncRNA prognostic signature. Kaplan-Meier survival analysis, receiver operating characteristic (ROC) curve, nomogram, C-index, calibration curve and clinical decision-making curve were adopted to assess the predictive capability of the identified signature.A signature comprising 9 autophagy-related lncRNAs (AL136964.1, ARHGEF26-AS1, PCED1B-AS1, AS104072.1, PRKCQ-AS1, LINC00957, AS125616.1, PSMB8-AS1, and AC087741.1) was identified as a prognostic model. LGG patients were allocated into high- and low-risk cohorts depending on the median model-based Riskscore. The survival analysis showed a 10-year survival rate of 9.3% (95% CI: 1.91-45.3%) in high-risk LGG patients and 48.4% (95% CI: 24.7-95.0%) in low-risk individuals in the training set. While those of patients in the validation set were 13.48% (95% CI: 4.52-40.2%) for high-risk LGG patients and 48.4% (95% CI: 28.04-83.4%) for low-risk LGG patients, respectively. This suggested a relatively low survival in high-risk cases compared to low-risk individuals. In addition, LncRNA signature was independently prognostic and potentially associated with the progression of LGG. Taken together, our constructed 9 autophagy-related lncRNA signature may play a crucial part in the diagnosis and treatment for LGG, which may guide to open up a new avenues for tumor targeted therapy.