Clinical Prognosis Of Breast Cancer Prediction Based On The Pyroptosis-related LncRNA Model

Objectives:Pyroptosis is a programmed cell death characterized by cell membrane rupture and the release of inflammatory substances,which plays an important role in the initiation and development of cancer.Long non-coding ribonucleic acids(Long non-coding RNA,lncRNA)are involved in the pathological process of various diseases by either directly or indirectly acting on proteins related to the pyroptosis signaling pathway.However,there are few studies on lncRNA related to pyroptosis in breast cancer(Breast Cancer,BC),and the prognostic value of pyroptosis-related lncRNA in BC is unclear.In this paper,a prognostic evaluation model was constructed using lncRNA associated with pyroptosis of breast cancer cells,and then the immunotherapy and sensitivity of drug therapy in patients in the prognostic model was analyzed.Methods:1.In this paper,1226 BC transcriptomic data and 1097 BC clinical data were downloaded from cancer Genome Atlas(The Cancer Genome Atlas,TCGA),and 436pyroptosis genes were obtained in Msig DB database,Gene Cards database and literature.By co-expression analysis of differential pyroptosis genes with BC lncRNA(|cor|>0.6,P<0.001),the resulting pyroptosis-related lncRNA was combined with BC clinical data.2.We performed univariate COX analysis on lncRNAs associated with pyroptosis in breast cancer,preliminarily screened for survival significant genes(P<0.05),and then performed lasso regression on these genes to obtain lasso significant genes.Finally,we performed multivariate COX regression analysis to construct a risk prognostic model(P<0.05).3.We used independent prognostic analysis,C-index curve,ROC(the Receiver Operating Characteristic)curve,survival analysis,and nomogram to predict the ability of the model to distinguish high-risk groups between low-risk groups and the effectiveness of the prediction model(P<0.001).In addition,we performed co-expression analysis on model lncRNAs to construct a lncRNA-m RNA co-expression network(|cor|>0.6,P<0.001),and explored the functions and enrichment pathways of model lncRNAs.4.We divide BC patients into high and low risk groups by risk model score,|log2FC|>1 and FDR<0.05 as screening conditions,analyze the difference between the high risk group and the low risk group,and then performed functional enrichment analysis and pathway enrichment analysis with the differential genes.We next analyzed immune cell-related functions and immune checkpoints between high and low risk groups.5.Finally,we evaluated the drug sensitivity in the high and low risk groups(P<0.05)according to the semi-inhibitory concentration(Half Maximal Inhibitory Concentration,IC50)available in the cancer drug sensitivity genomics(Genomics of Drug Sensitibity in Cancer,GDSC)database,so as to obtain potential therapeutic drugs with differences in high and low risk groups.Results:1.In this study,304 lncRNAs were obtained based on the transcriptome data of BC in the TCGA database and the coexpression of different pyroptosis genes in breast cancer.We divided BC into a training set(n=722)and a validation set(n=311)according to its clinical characteristics.Then,304 lncRNAs were analyzed by univariate COX analysis through training set,and 24 lncRNAs related to survival were obtained.Then,the optimization model was constructed by Lasso-COX analysis on these 24 lncRNAs,multivariate COX analysis yielded risk prognostic models based on five pyroptosis-associated lncRNAs(AC073316.1,MSC-AS1,AC137932.2,AL021578.1,LINC01871)(P<0.05).The high and low risk groups were distinguished by the median risk score of the training set.2.Survival analysis of the training and validation sets showed that the model could well distinguish between the high and low risk groups,and that the high-risk group had a worse prognosis(P<0.001).Subsequently,the training and validation sets were analyzed by ROC curves to evaluate the model efficacy(AUC=0.774 for training set,AUC=0.694 for validation set),demonstrating that the model had good predictive power.The results of the C-index curve analysis of the model and other clinical traits showed that the risk model is more accurate than other clinical traits to predict the survival prognosis of BC patients(Risk Score0.7<C-index<0.8),and the model has moderate accuracy.The results of independent prognostic analysis showed that the model was independent of other clinical traits(univariate COX:risk Score(P<0.001),Age(P<0.001),Gender(P<0.884),Stage(P<0.001),T(P<0.001),multivariate COX:risk Score(P<0.001),Age(P<0.001),Gender(P<0.584),Stage(P<0.001),T(P=0.311)).Finally,we combined the risk scores and other clinical traits to construct nomograms predicting 1-year,3-year,and 5-year survival.3.Model lncRNA co-expression analysis constructs 77 lncRNA-m RNA target co-expression networks for functional and pathway enrichment of these m RNA.The results showed that these m RNA were involved in biological processes such as signal transduction,cytolysis,immune response,and G protein receptor coupling,and constructed the biological components of extracellular region and cell membrane,which were enriched in pathways such as programmed cell death.Meanwhile,the differential genes of the high and low risk groups were enriched in immune-related functions.4.The results of immune-related function analysis indicated that T cell co-inhibition,killing activity,T cell co-stimulation,and proinflammatory response were lower in the low-risk group,and the expression of immune cells CD8~+T-Cells,Tfh,Th1-Cells was also lower in the high-risk group than in the low-risk group.Furthermore,the expression of the immune checkpoint genes CD276,TNFSF4,and NRP 1 was significantly higher in the high-risk group than in the low-risk group.5.The results of the drug sensitivity analysis indicated that BC patients in the high-risk group were more sensitive to RO-3306 and BI 2536,while patients in the low-risk group were more sensitive to Dihydrorotenone and Navitoclax.Conclusion:In this study,a risk prognostic model composed of AC073316.1,MSC-AS1,AC137932.2,AL021578.1,and LINC01871 was selected based on the BC TCGA database and pyroptosis gene,which has good predictive power and can provide some references for patient immunotherapy and drug sensitivity prediction.