Development of a Novel Prognostic Model of Glioblastoma Based on m6A-Associated Immune Genes and Identification of a New Biomarker

Abstract

BackgroundAccumulating evidence shows that m6A regulates oncogene and tumor suppressor gene expression, thus playing a dual role in cancer. Likewise, there is a close relationship between the immune system and tumor development and progression. However, for glioblastoma, m6A-associated immunological markers remain to be identified.MethodsWe obtained gene expression, mutation, and clinical data on glioblastoma from The Cancer Genome Atlas and Chinese Glioma Genome Atlas databases. Next, we performed univariate COX–least absolute shrinkage and selection operator (LASSO)–multivariate COX regression analyses to establish a prognostic gene signature and develop a corresponding dynamic nomogram application. We then carried out a clustering analysis twice to categorize all samples according to their m6A-regulating and m6A-associated immune gene expression levels (high, medium, and low) and calculated their m6A score. Finally, we performed quantitative reverse transcription-polymerase chain reaction, cell counting kit-8, cell stemness detection, cell migration, and apoptosis detection in vitro assays to determine the biological role of CD81 in glioblastoma cells.ResultsOur glioblastoma risk score model had extremely high prediction efficacy, with the area under the receiver operating characteristic curve reaching 0.9. The web version of the dynamic nomogram application allows rapid and accurate calculation of patients’ survival odds. Survival curves and Sankey diagrams indicated that the high-m6A score group corresponded to the groups expressing medium and low m6A-regulating gene levels and high m6A-associated prognostic immune gene levels. Moreover, these groups displayed lower survival rates and higher immune infiltration. Based on the gene set enrichment analysis, the pathophysiological mechanism may be related to the activation of the immunosuppressive function and related signaling pathways. Moreover, the risk score model allowed us to perform immunotherapy benefit assessment. Finally, silencing CD81 in vitro significantly suppressed proliferation, stemness, and migration and facilitated apoptosis in glioblastoma cells.ConclusionWe developed an accurate and efficient prognostic model. Furthermore, the correlation analysis of different stratification methods with tumor microenvironment provided a basis for further pathophysiological mechanism exploration. Finally, CD81 may serve as a diagnostic and prognostic biomarker in glioblastoma.

Systematic integration of m6A regulators and autophagy-related genes in combination with long non-coding RNAs predicts survival in glioblastoma multiforme

“…Some evidence suggests that m6A alteration may play a key role in gliomas through a variety of mechanisms, providing more opportunities for early diagnosis and targeted therapy of gliomas 2 . Moreover, some m6A based risk score models have also been proposed for predicting the prognosis of glioma patients 32 , 33 . Recently, m6A methylation regulatory genes have been used to classify patients with low-grade gliomas into high- or low-risk subgroups 34 .…”

Thrombosis risk prediction in lymphoma patients: A multi‐institutional, retrospective model development and validation study

“…We used the Least Absolute Shrinkage and Selection Operator (LASSO) regression for parsimonious feature selection beyond lymphoma subtypes. [19][20][21] We then constructed a Fine and Gray subdistribution hazard model using the final selected variables to predict the cumulative incidence with death as competing events. The native sub-distribution hazard models and beta coefficients were saved and shared on GitHub for subsequent external validation.…”

“…For the RAM derivation, we mandated the classification of lymphoma subtypes based on the logistic regression odds ratios (ORs) of VTE at 6 months. We used the Least Absolute Shrinkage and Selection Operator (LASSO) regression for parsimonious feature selection beyond lymphoma subtypes 19–21 . We then constructed a Fine and Gray sub‐distribution hazard model using the final selected variables to predict the cumulative incidence with death as competing events.…”

Impact of cytotoxic chemotherapy on aldo-keto reductase family 1 member B10 expression

“…Figure 1 illustrates the AKR1B10 levels in our patients before, during, and after treatment in comparison to those in 49 patients reported in our previous study. 18 Based on our prior study, the optimal diagnostic cutoff of serum AKR1B10 was 3.456 ng/ mL, which yielded an area of the curve of 0.9426 AE 0.0324 (95% confidence interval [CI] ¼ 0.8790-1.006; P < 0.0001), sensitivity of 90.63% (95% CI ¼ 74.98-98.02), and specificity of 93.88% (95% CI ¼ 83.13-98.72). To increase the sensitivity, we considered AKR1B10 levels exceeding 5 ng/mL as positive.…”