Tumor Purity as an Underlying Key Factor in Glioma

Abstract

Glioma tissues consist of not only glioma cells but also glioma-associated nontumor cells, such as stromal cells and immune cells. These nontumor cells dilute the purity of glioma cells and play important roles in glioma biology. Currently, the implications of variation in glioma purity are not sufficiently clarified. Here, tumor purity was inferred for 2,249 gliomas and 29 normal brain tissues from 5 cohorts. Based on the transcriptomic profiling method, we classified CGGA and TCGA-RNAseq cohorts as the RNAseq set for discovery. Cases from TCGA-microarray, REMBRANDT, and GSE16011 cohorts were grouped as a microarray set for validation. Tissues from the CGGA cohort were reviewed for histopathologic validation. We found that glioma purity was highly associated with major clinical and molecular features. Low purity cases were more likely to be diagnosed as malignant entities and independently correlated with reduced survival time. Integrating glioma purity into prognostic nomogram significantly improved the predictive validity. Moreover, most recognized prognostic indicators were no longer significantly effective under different purity conditions. These results highlighted the clinical importance of glioma purity. Further analyses found distinct genomic patterns associated with glioma purity. Low purity cases were distinguished by enhanced immune phenotypes. Macrophages, microglia, and neutrophils were mutually associated and enriched in low purity gliomas, whereas only macrophages and neutrophils served as robust indicators for poor prognosis. Glioma purity and relevant nontumor cells within microenvironment confer important clinical, genomic, and biological implications, which should be fully valued for precise classification and clinical prediction. .

Costimulatory checkpoint SLAMF8 is an independent prognosis factor in glioma

“…According to histopathologic classifications, GBM had the highest SLAMF8. In five molecular entities, as previously reported, SLAMF8 was enriched in tumors of patients with wild‐type IDH lower grade glioma (LGG) and IDH wild‐type GBM (Figure A). These expression trends of SLAMF8 were validated in the TCGA cohort (Figure S1A).…”

“…Glioma purity and infiltrated immune cell components are reported to have important clinical and biological implications. Lower purity means malignancy progression, poor prognosis, and enhanced immune phenotype . SLAMF8 was closely related to immune score, stromal score, and glioma purity.…”

Amino acid metabolism‐related gene expression‐based risk signature can better predict overall survival for glioma

“…Lowpurity gliomas were characterized by intensive local immune phenotypes and correlated with a poor prognosis. 23 Therefore, we applied the ESTIMATE algorithm to predict tumor purity using gene expression profiles 26 and found a significant increase in ESTIMATE scores in the high-risk group (Figure S7), indicating that a greater presence of inflammatory microenvironment components is associated with progressive tumorigenesis. 27 In addition, we analyzed the 30-amino acid metabolism-related genes and proteins in detail.…”

“…Functional annotation of the 30‐gene signature showed that biological functions of angiogenesis, cell adhesion and immune response may contribute to patients’ high risk and poor clinical outcome. Low‐purity gliomas were characterized by intensive local immune phenotypes and correlated with a poor prognosis . Therefore, we applied the ESTIMATE algorithm to predict tumor purity using gene expression profiles and found a significant increase in ESTIMATE scores in the high‐risk group (Figure ), indicating that a greater presence of inflammatory microenvironment components is associated with progressive tumorigenesis …”

Combinations of four or more CpGs methylation present equivalent predictive value for MGMT expression and temozolomide therapeutic prognosis in gliomas

“…RNA sequencing library was constructed as previously published study . Briefly, RNA was extracted from the frozen tissue sample, and the RNA‐seq library was constructed and subsequently sequenced on the Illumina HiSeq 2000 platform (Illumina, San Diego, CA, USA) using 101‐bp pair‐end sequencing strategy.…”