Stiffiness Of Adult Glioblastoma:Biological Processes Characteristics And Predicting Model Of Prognosis

Background:Glioblastoma(GBM)is the most common malignant tumor in central nervous system with an extremely poor prognosis,and researches on the mechanisms of GBM progression and screening therapy targets and biomarkers are of vital importance.Tissue stiffiness is common among solid tumors.For GBM,an increasing stiffiness of extracellular matrix can promote tumor cells proliferation,invasion,epithelial-mesenchymal transition and chemoresistance,which indicate the tumor stiffiness as a potential therapy targets and biomarkers.This study utilized dataset from database,bioinformatics tools and machine learning algorithm to clarify the biological process of tumor stiffiness of GBM,identify the signature gene set and evaluate its efficacy as a prediction of prognosisMethods:The RNA-sequencing data and clinical data of newly diagnosed IDH-wildtype GBM patients were obtained from The Cancer Genome Atlas database.The classification of GBM samples was done by consensus clustering algorithm and GBM samples were divided into different subtypes.Differentially expressed genes were identified between all subtypes.Pathway enrichment analysis,gene set enrichment analysis and protein-protein interaction analysis were preformed to find biological processes that were associated with stiffiness of tumor.Random forest model was used to screen the signature gene set that was related with stiffiness of GBM and set up predicting model of stiffiness of GBM and prognosis.CIBERSORT algorithm and ESTIMATE algorithm were used to analysis the proportion of immune cells that infiltrated in tumor microenvironment and the enrichment of stromal component.Results:One hundred and forty-one patients(n=141)with newly diagnosed IDH-wildtype GBM were involved,of which median age was 62 years old and ratio of different sex was 1.86(male:female).Based on the key molecules,signal pathways and biological processes that published in previous researches,I set up a tumor stiffiness associated gene set and divided GBM samples into 2 subtypes:soft tumor that evaluated by biological process(referred as "BP?soft")and stiff tumor that evaluated by biological process(referred as"BP?stiff").According to survival analysis,the BP?soft subtype had a significantly prolonged median overall survival(10.6 m vs 14.7 m,p=0.044)and median progression free survival(4.4 m vs 9.8 m,p=0.001).Based on bioinformatics analysis,I found regulation of tissue stiffiness of GBM was associated with biological process in bellowing:extracellular matrix component and structure,epithelial proliferation,collagen metabolism,glycosaminoglycan binding,integrin binding,growth factor and matrix metalloproteinase.Also,down-regulation of angiogenesis was found in BP?soft subtype.Immune cells infiltrating analysis shown that GBM tissue of BP?soft subtype had higher proportions of activate CD4+T memory cell,follicular helper T cells,resting CD4+T memory cell and MO macrophage,and BP soft subtype had a lower proportion of M2 macrophage,when compared with BP stiff subtype.13 signature genes were characterized as ANKRD1?CLU?FGF9?KRT19?L1CAM?LOX?LOXL1?MLPH?MMP19?NCKAP1L?RAB27A?SDC4?WNT2,which were associated with stiffiness of GBM.Based on the signature gene set of stiffiness of GBM,predicting model of tumor stiffiness and prognosis were built up.Conclusions:This study classified GBM into BP soft subtype and BP stiff subtype.A signature gene set of stiffiness of GBM was screened,and predicting model of tumor stiffiness and prognosis were built up based on random forest model.According to the gene set enrichment analysis and immune cells infiltrating analysis,I hypothesized that BP?soft subtype may benefit from immunotherapy and BP?stiff subtype may benefit from antiangiogenic therapy.