Immune landscape of a prognostic immunogenic cell death (ICD)-related long noncoding RNAs (lncRNAs) signature in colon adenocarcinoma (COAD)

Abstract Background Immunogenic cell death (ICD), as a form of regulated cell death, demonstrates promising roles in immunogenic tumor microenvironment. However, the expression and roles of ICD-related long noncoding RNAs (lncRNAs) in COAD were largely unknown. Methods RNA-sequencing data and clinical data were downloaded from The Cancer Genome Atlas (TCGA) database. A list of 34 immunogenic cell death-related genes was retrieved from published literatures. ICD-related lncRNAs were retrieved using R language and Perl software. Univariate and multivariate cox analyses were used to construct an ICD-related lncRNA risk model. The performance of thissignature was validated using Kaplan-Meier curve and receiver operating curve (ROC). The immune landscape of the risk signature was revealed by analyzing the correlations with immune cells infiltration, immune function, immune checkpoint inhibitors. Correlations with RNA stemness score (RNAss) were analyzed using Spearman correlation coefficient. Correlation with m6A regulators was investigated using Wilcoxon test. Moreover, expression and survival analysis of the risk lncRNAs were revealed based on TCGA database. Results We identified 23 ICD-related lncRNAs with prognostic values to construct a risk model. The risk model demonstrated good performance in predicting survival. The area under curve (AUC) of the ROC curve at 1-, 2- and 3-year were 0.684, 0.715, and 0.747, respectively. Further univariate and multivariate analysis showed that the risk score was an independent prognosis factor. Further validation showed that there were significant differences in immune cells infiltration, immune functions, and immune checkpoint inhibitor expression between high- and low-risk groups. In addition, there was negative correlation between the risk signature and RNAss, and there was difference in the expression of m6A regulators in high- and low-risk groups. A nomogram including the risk score also demonstrating significant prognostic value. Moreover, upregulated expression of MIR4435-2HG, LINC02257, AL161729.4, AC007128.1 and AP001619.1 predicted poor overall survival in COAD. Conclusions We established an ICD-related lncRNA signature for the overall survival of COAD, which demonstrating high sensitivity and specificity. The correlations between ICD-related lncRNAs and tumor immune microenvironment still need to be further investigated.