Identification of an Individualized Immune-Related Prognostic Risk Score in Lung Squamous Cell Cancer

Abstract

Background: Lung squamous cell carcinoma (LUSC) is one of the most common histological subtypes of non-small cell lung cancer (NSCLC), and its morbidity and mortality are steadily increasing. The purpose of this study was to study the relationship between the immune-related gene (IRGs) profile and the outcome of LUSC in patients by analyzing datasets from The Cancer Genome Atlas (TCGA). Methods: We obtained publicly available LUSC RNA expression data and clinical survival data from The Cancer Genome Atlas (TCGA), and filtered IRGs based on The ImmPort database. Then, we identified risk immune-related genes (r-IRGs) for model construction using Cox regression analysis and defined the risk score in this model as the immune gene risk index (IRI). Multivariate analysis was used to verify the independent prognostic value of IRI and its association with other clinicopathological features. Pearson correlation analysis was used to explore the molecular mechanism affecting the expression of IRGs and the correlation between IRI and immune cell infiltration. Results: We screened 15 r-IRGs for constructing the risk model. The median value of IRI stratified the patients and there were significant survival differences between the two groups (p = 4.271E-06). IRI was confirmed to be an independent prognostic factor (p < 0.001) and had a close correlation with the patients' age (p < 0.05). Interestingly, the infiltration of neutrophils or dendritic cells was strongly upregulated in the high-IRI groups (p < 0.05). Furthermore, by investigating differential transcription factors (TFs) and functional enrichment analysis, we explored potential mechanisms that may affect IRGs expression in tumor cells. Conclusion: In short, this study used 15 IRGs to build an effective risk prediction model, and demonstrated the significance of IRGs-based personalized immune scores in LUSC prognosis.

Keywords: TCGA; immune-related genes; lung squamous cell carcinoma (LUSC); prognosis; transcriptome.

Figure 1

Differentially expressed immune-related genes. Heatmap (A) and volcano plot (B) of differentially expressed genes between lung squamous cell carcinoma (LUSC) and adjacent non-tumor tissues. Heatmap (C) and volcano plot (D) demonstrating differentially expressed immune-related genes (IRGs). The black dots represent the undifferentiated genes, while the red and green dots represent the differentiated genes. (E) Forest plot of hazard ratios showing the prognostic values of immune-related genes.

Figure 2

Gene functional enrichment of IRGs. Gene ontology analysis (GO) of differentially expressed immune-related genes (A) and the significant Kyoto Encyclopedia of Genes and Genomes pathways (KEGG) (B). Gene ontology analysis (GO) of prognostic-associated immune-related genes (C) and the most significant Kyoto Encyclopedia of Genes and Genomes pathways (KEGG) (D).

Figure 3

Transcription factor-mediated regulatory network. Heatmap (A) and volcano plot (B) showing differentially expressed transcription factors (TFs). (C) Regulatory network based on clinically relevant TFs and IRGs.

Figure 4

Construction of the prognostic risk model based on immune-related genes. (A) Rank of prognostic risk score and distribution of groups. (B) Survival status of patients in different groups. (C) Heatmap of expression profiles of included genes.

Figure 5

The prognostic value of the risk score. (A) The overall survival (OS) time of patients in the high-risk group and low-risk group. (B) The ROC curves of OS for the 15-gene immune-related risk score. Forest plot of hazard ratios showing the prognostic values of immune-related genes involved in the risk model based on Univariate analysis (C) and multivariate analysis (D).

Figure 6

The relationships between the risk immune gene and clinicopathologic characters. (A) The correlation between the Immune gene Risk Index (IRI) and patients' age. The relationships between AGTR2 and (B) tumor stage; (C) distant metastasis; (D) lymph node metastasis. The relationships between distant metastasis and (E) AMH; (F) FGFR4; (G) GCCR. The relationships between patients' age and APLN (H) and MMP12 (I). The relationships between patients' gender and ENG (J) and FGFR4 (K). The relationships between PLAU and (L) age; (M) T stage; (N) lymph node metastasis. The relationships between RNASE7 and (O) tumor stage and (P) distant metastasis.

Figure 7

Relationships between the Immune gene Risk Index (IRI) and infiltration abundances of immune cells. (A) neutrophils; (B) dendritic cells; (C) B cells; (D) CD4 T cells; (E) CD8 T cells; and (F) macrophages.