Identification Of Prognostic Signature And Hierarchical Cluster Analysis Of Immunogenomic Profiling For Diffuse Large B-cell Lymphoma Based On Bioinformatics

BackgroundDiffuse large B-cell lymphoma(DLBCL)is the most common type of non-Hodgkin lymphoma(NHL)worldwide,accounting for about 30-40%of all lymphomas.DLBCL has heterogeneity of pathophysiological,genetic and clinical characteristics.Patients are often characterized by rapidly growing masses in single or multiple lymph nodes or extranodal sites.In the current World Health Organization(WHO)lymphoma classification,approximately 80%of DLBCL cases are designated as not otherwise specified(NOS).With the significant development of gene sequencing technology,two molecularly different DLBCL types have been identified through gene expression patterns,including activated B-cell like(ABC)and germinal centre B-cell like(GCB)type.Rituximab-based immunochemotherapy combined with chemotherapy R-CHOP has become the standard first-line treatment for patients with diffuse large B-cell lymphoma.The International Prognostic Index(IPI)risk scoring system that includes clinical parameters(age,lactate dehydrogenase(LDH),sites of involvement,Ann Arbor stage,ECOG performance status)has been widely used.Although the clinical,morphological and molecular parameters used to classify diffuse large B-cell lymphoma are diverse today,However,30%-40%of patients will have relapse or refractory disease after initial treatment,which will greatly reduce the survival time of patients.Therefore,new prognostic biomarkers should be developed for risk stratification and treatment optimization for early patients.Bioinformatics is an interdisciplinary discipline that applies mathematics and computer science to the indexing,classification and analysis of biomolecular information.High-throughput sequencing,as a new experimental technology,is applied to Bioinformatics analysis.Tumor whole genome sequencing can not only analyze tumor transcriptome and epigenetics,but also further analyze the immune microenvironment of tumor tissue and the state of tumor immune infiltration.At present,bioinformatics analysis has become an important technical support for diffuse large B-cell lymphoma,providing the basis for molecular targeted therapy of diffuse large B-cell lymphoma.In addition to cancer cells,the tumor microenvironment also contains immune cells,fibroblasts,capillaries,extracellular matrix,and biomolecules infiltrating them.With the development of tumor immunotherapy and the in-depth research on immune checkpoint inhibitor therapy,the complex microenvironment within the tumor may be one of the important reasons for the low response rate of immunotherapy.Therefore,the research on the tumor microenvironment needs to be improved urgently.Combining deep learning and data analysis of the DLBCL microenvironment,it provides a novel calculation method for quantitatively characterizing the tumor microenvironment.Based on bioinformatics methods,this topic has identified a new set of predictable prognostic markers for diffuse large B-cell lymphoma and discussed its immune microenvironment.The content of this subject is mainly divided into two parts:The first part is to determine a set of 35 gene prediction models that can effectively predict the survival outcome of DLBCL patients.The combination of 35 gene prognostic models and other prognostic indicators IPI and cell of origin(COO)and other existing risk models may help to better assess the prognosis of patients.In the second part,an immunophenotype of DLBCL was constructed based on the enrichment level of immune-related gene sets.This immunophenotyping is of great significance for predicting the prognosis of patients with diffuse large B-cell lymphoma and the efficacy of new immune checkpoint inhibitors.Part One:Identification of a Prognostic Relevant Gene Set and Related Tumor Microenvironment Analysis in Diffuse Large B-Cell LymphomaMethod1.Univariate Cox regression was used to screen the genes related to the overall survival of DLBCL patients in the GSE117556,GSE10846 and GSE31312 datasets.Hazard ratio(HR)>1 is defined as risky genes.HR<1 is defined as protective genes,and P<0.01 as the cut-off point.2.The risky genes and protective genes shared by the three datasets were intersected and combined with Lasso regression and Cox regression to build the final gene risk prediction model.3.The NCICCR datasets were utilized to verify the robustness and transferability of the risk score model.Results1.we succeeded in constructing a 35-gene prognostic signature which robustly and reliably predicted the clinical outcome in our training and validation groups.2.We found that the immune and stromal scores of DLBCL were extremely associated with overall survival and risk scores.3.There were significant differences in the level of immune cell infiltration in the high and low risk groups of DLBCL samples.4.A stratified analysis of clinical characteristics such as cell of origin(COO),Age,ECOG,IPI,Stage and,Double expression showed that OS were obvious statistical significance between the high-risk group and low-risk group.5.The 35-gene prognostic signature achieved AUC values of 0.788 and 0.609 in GSE117556 and NCICCR,respectively,which means a high OS prediction performance.6.On the basis of multivariate Cox regression,the risk score and clinical factors are combined to construct a nomogram model.Conclusion1.This study identified a 35-gene prognostic signature that can effectively predict DLBCL patient survival outcomes.2.The 35 genes prognostic model related to tumor microenvironment(TME)in combination with other prognostic indicators IPI and COO might be useful in better assessing the prognosis of patients and selecting treatments that could be active only in specific subtypes of DLBCL.Part Two:Hierarchical Clustering of Diffuse Large B-Cell Lymphoma Based on Immunogenomic ProfilingMethod1.This study included 928 patients with diffuse large cell lymphoma in GSE117556 in the Gene Expression Omnibus(GEO)database.For each sample lymphoma diffuse large B cell,the level of enrichment according to 29 kinds of immune-related genes will be set diffuse large B-cell lymphoma hierarchical clustering performed.2.The biological characteristics of gene expression are used to infer the immune score and matrix score.3.The CIBERSORT package was used to calculate the distribution of immune cell types in each subset,and the proportion of immune cell types in the subtypes of diffuse large B-cell lymphoma was compared by Kruskal-Wallis test.4.The Kaplan-Meier survival curve was drawn based on the patient's survival information to visualize the survival difference between immune subtypes.5.Finally,use the single sample gene set enrichment analysis(ssGSEA)method for GO(Gene Ontology)and KEGG(Kyoto Encyclopedia of Genes and Genomes)enrichment analysis.Results1.An unsupervised cluster analysis was performed on 29 immune-related gene sets.According to the ssGSEA score of the genome,928 patients were divided into three subgroups:Immunity-H,Immunity-M and Immunity-L.The expression level of immune-related genes in Immunity-H group was significantly higher than that in Immunity-L group.2.We evaluated the prognostic value of each immune subtype for patient survival.The results found that the survival curves of the three subgroups had significant statistical differences.It also proved that immunophenotyping has good predictability for the survival of diffuse large B-cell lymphoma.The Immunity-H group had the best prognosis,the Immunity-L group had the worst prognosis,and the Immunity-M group was somewhere in between.3.We also explored the relationship between the expression of genes such as PD-1,PD-L1,CD3D,HIF1A,MUM1 and immune subgroups,and found the expression of genes such as PD-1,PD-L1,CD3D,HIF1A,MUM1,etc.There are significant differences in the Immunity-H and Immunity-L groups.4.Finally,based on the enrichment scores in each sample,the differential genes in the Immunity-H and Immunity-L groups were screened.KEGG analysis showed that the differential genes in the Immunity-H and Immunity-L groups were mainly enriched in allograft rejection and iron.PD-L1 expression and PD-1 checkpoint pathway,protein export,T cell receptor signaling pathway in disease,cancer.GO analysis showed that Immunity-H and Immunity-L group differential genes were significantly enriched and immune synapse formation,positive regulation of interleukin-2 biosynthesis process,positive regulation of nitric oxide synthase biosynthesis process,tolerance induction The regulation,T cell receptor complex.Conclusion1.We constructed an immunophenotyping of diffuse large B-cell lymphoma based on the enrichment level of 29 immune-related gene sets(immune cell types,functions and pathways).2.This immunophenotyping is of great significance for predicting the prognosis of patients with diffuse large B-cell lymphoma and the efficacy of new immune checkpoint inhibitors.