Based on immunology analysis of tumor distype, this idea is worth learning from!

Today, we share an article published in molecular oncology in May 2020, with the Title Classification of diffusion lower ‐ grade glioma based on empirical profiling, with an impact factor of 5.962.Background diffuse low-grade gliomas (LGG) are invasive brain tumors caused by glial cells or precursor cells. Compared with glioblastoma (GBM, grade IV), the course of disease is softer, and the survival range of patients is wide (from 1 to 15 years), and there is a great difference between the two groups according to tumor types.studies have shown that different immunotherapies can successfully treat many malignant tumors, but not all patients can benefit from immunotherapy.for successful immunotherapy, a better understanding of tumor specific immune microenvironment is needed.more and more studies have shown that TME is still highly different in different subtypes of glioma.at present, the immune microenvironment of diffuse LGG has not yet been fully characterized. Deepening the understanding of LGG immune microenvironment will provide valuable information for personalized immunotherapy.data download 1) RNA SEQ data, somatic mutation and copy number data, clinical information data of diffuse low-grade glioma samples (n = 402) were obtained from TCGA database - training set 2) RNA SEQ data, clinical information and survival data of diffuse low-grade glioma samples (n = 171) were obtained from CGGA database - validation set (cohort 1) 3) RNA SEQ data, clinical information and survival data of diffuse low grade gliomas (n = 274) were obtained from CGGA database The microarray data, clinical information and survival data of diffuse low-grade gliomas (n = 171) were obtained from CGGA database. The results showed that: (1) identification and validation of immune subtypes of diffuse low-grade gliomas, based on 782 previously reported immune related genes, the authors used consensus clustering method to identify the immune subtypes of diffuse low-grade gliomas The data sets of 402 diffuse LGG samples were clustered (50 iterations, 80% resampling rate).cumulative distribution function and consensus thermography were used to evaluate the best k = 3 (three subclasses were defined as IM1, im2 and IM3), i.e. diffuse low-grade glioma was divided into three immune subtypes: IM1, im2 and IM3.in order to verify the immune subtypes in the CGGA cohort, the authors first identified 751 immune related genes shared by the training and validation sets.in the training set, microarray prediction analysis (PAM) was used to train genes to construct gene based classifiers.in the validation set, the PAM classifier constructed based on the validation set was used to classify the belonging immune subtypes of each validation cohort sample, and the molecular and clinical information were annotated (Fig. 1b-d); principal component analysis (PCA) showed that there were significant differences in gene expression patterns among different immune subtypes (Fig. 2a). The three immune subtypes showed similar prognosis and pathological characteristics in the training set and validation set (Fig. 2B, c).the authors evaluated the similarity and repeatability of acquired immune subtypes between training set and validation set by IGP method. Cellular and molecular characteristics of immune subtypes. In order to reveal the immune heterogeneity among the three subtypes, several immune related tools were used for analysis.firstly, matrix score, immune score and tumor purity were calculated according to estimate method (Fig. 3a). It was found that im2 subtype had higher immune score, matrix score and lower tumor purity; then, cibersort analysis was performed to calculate the proportion of lymphocytes and M2 macrophages, and IM1 was found However, im2 and IM3 showed higher levels of M2 macrophages (Fig. 3b). Finally, the enrichment levels of immune cell types and functions in the three immune subtypes were analyzed by ssgsea (Fig. 3C).in addition, the authors also analyzed the expression levels of HLA and checkpoint genes among the three subtypes (the results are included in the supplementary materials). The correlation between invasive immune cells and prognosis was analyzed by Cox regression analysis. It was found that when stratified by immune subtype, the survival rate was correlated with cytolytic activity in im2, and patients with high neutrophil or Treg scores tended to have better prognosis (FIG. 4A, b).in addition, the authors evaluated the prognostic value of the main immune examination points, and found that the high expression of PDCD1, havcr2 and ido1 means that the prognosis is worse (the results are included in the supplementary materials).these results suggest that infiltrating immune cells may affect the prognosis of patients and provide valuable targets for immunotherapy of diffuse LGG. Genomic variation of immune subtypes, the differences of genomic changes among the three immune subtypes were discussed.the new im2 subtype has more obvious aneuploid variation, homologous recombination defects, and higher copy number load and tumor mutation load (Fig. 5a). correlation analysis of genomic mutations and immune subtypes showed that IM1 was rich in IDH1, 1p / 19q co deletion, CIC, fubp1 and Notch1 mutations; im2 was rich in driving gene mutations, such as PTEN, EGFR and NF1; IM3 was rich in IDH1, ATRX and TP53 mutations (Fig. 5b). gistic2.0 analysis showed that im2 showed more frequent deletion or amplification regions, such as CDKN2A / cdkn2b, egfr1, CDK4, kit and PDGFRA. these findings suggest that tumors with high immune infiltration may have higher levels of genomic changes. In order to define the immune status of diffuse LGG, the authors found that im2 could be further divided into two subgroups: IM2A and im2b (Fig. 6a, b), in which IM2A had better outcome (Fig. 6c). ssgsva analysis showed that IM2A was associated with higher levels of immune cells and function (Fig. 6D). cibersort analysis showed that the percentage of M2 macrophages in IM2A was increased (Fig. 6e). these data indicate that there is obvious intra cluster heterogeneity in immune subtypes. 3.6 using Cox proportional hazard model to develop and validate immune related features, considering the close relationship between prognosis and immune infiltration, the authors used Cox proportional risk model to develop immune related features to predict prognosis. Sam analysis showed that there were 421 differentially expressed immune genes between im2 and IM1 / IM3 subtypes. Univariate Cox regression analysis showed that 359 differentially expressed genes were significantly associated with OS of patients (Fig. 7a). then, the Cox proportional hazard model was used to select the genes with the best prognostic value (Fig. 7b). eight gene markers were identified (Figure 7C), and the risk score was calculated based on their expression levels and regression coefficients. high risk scores were found to be concentrated in im2, classic and mesenchymal gliomas, grade III or IDH wild type tumors (Figure 7D). Kaplan – Meier analysis showed that high risk score meant poor prognosis (Fig. 7e, f). the authors further evaluated the accuracy of the prediction by calculating the risk score, AUC of age and grade (area under the curve), and found that the AUC of risk score (85.5%) was much higher than that of age (77.8%) and grade (68.5%) (Fig. 7G). these data demonstrate the powerful function of immune markers in predicting prognosis and highlight the importance of immune TME in predicting patient survival. the analysis process has been shared here. If you still have not finished, you may want to review the analysis process of the full text: it can be seen that the analysis process of this paper is not complicated. The author has studied the tumor microenvironment differences of diffuse low-grade gliomas by using a variety of analysis methods from multiple perspectives, identified and verified the tumor classification, and developed It has important guiding significance for immunotherapy of LGG patients. the author's research ideas and analysis methods are worth learning, and interested partners can read them carefully! The latest thinking 618 activity last day increases the quantity also reduces the price, sweeps the code to lock immediately!