A 20-Gene Signature Predicting Survival in Patients with Clear Cell Renal Cell Carcinoma Based on Basement Membrane

Abstract

Objectives: The most common subtype of renal cell carcinoma, clear cell renal cell carcinoma (ccRCC), has a high heterogeneity and aggressive nature. The basement membrane (BM) is known to play a vital role in tumor metastasis. BM-related genes remain untested in ccRCC, however, in terms of their prognostic significance.

Methods: BM-related genes were gleaned from the most recent cutting-edge research. The RNA-seq and clinical data of the ccRCC were obtained from TCGA and GEO databases, respectively. The multigene signature was constructed using the univariate Cox regression and the LASSO regression algorithm. Then, clinical features and prognostic signatures were combined to form a nomogram to predict individual survival probabilities. Using functional enrichment analysis and immune-correlation analysis, we investigated potential enrichment pathways and immunological characteristics associated with BM-related-gene signature.

Results: In this study, we built a model of 20 BM-related genes and classified them as high-risk or low-risk, with each having its anticipated risk profile. Patients in the high-risk group showed significantly reduced OS compared with patients in the low-risk group in the TCGA cohort, as was confirmed by the testing dataset. Functional analysis showed that the BM-based model was linked to cell-substrate adhesion and tumor-related signaling pathways. Comparative analysis of immune cell infiltration degrees and immune checkpoints reveals a central role for BM-related genes in controlling the interplay between the immune interaction and the tumor microenvironment of ccRCC.

Conclusions: We combined clinical characteristics known to predict the prognosis of ccRCC patients to create a gene signature associated with BM. Our findings may also be useful for forecasting how well immunotherapies would work against ccRCC. Targeting BM may be a therapeutic alternative for ccRCC, but the underlying mechanism still needs further exploration.

Figure 1

Identification of the candidate BM-related genes in the TCGA cohort. (a) Differentially expressed genes associated with BM are shown using a heatmap. (b) BM-related genes having significant predictive value based on OS are visualized in a forest plot. (c) Candidate gene interactions are mapped out by the PPI network retrieved from the STRING database.

Figure 2

Distribution and prognostic analyses of the 20-gene signature in the TCGA cohort and GSE29609 cohort. (a, b) The distributions of the risk scores and corresponding survival status of KIRC patients in the TCGA cohort. (c) KM curves for the OS of ccRCC patients in the high- and low-risk group in the TCGA cohort. (d) The AUC of time-dependent ROC curves confirmed the risk score's prognostic efficacy in the TCGA cohort. (e, f) The distributions of the risk scores and corresponding survival status of the GSE29609 dataset. (g) KM curves for the OS of patients in the high- and low-risk groups in the GSE29609 dataset. (h) The AUC of time-dependent ROC curves confirmed the risk score's prognostic efficacy in the GSE29609 dataset.

Figure 3

Results of Cox regression for risk factors for ccRCC. (a) Outcomes from a univariate Cox regression study of OS in a cohort of patients with ccRCC based on risk signature score and clinical factors. (b) Results of stepwise multivariate cox regression analysis. (c, d) Correlation of risk group and clinical traits.

Figure 4

Stratified by age, gender, race, grade, T stage, N stage, or M stage, KM curves demonstrate OS disparities between high- and low-risk groups.

Figure 5

Building a nomogram of 20 BM-related genes. (a) A predictive nomogram for predicting 1, 3, and 5 years OS in ccRCC patients. (b) The calibration plots for predicting 1, 3, 5 years OS.

Figure 6

Analyses of GO and KEGG with typical findings. (a) Top 5 significant BP, MF, and CC terms in GO analyses. (b) Top 10 significant KEGG signaling pathways.

Figure 7

The correlation between BM-related DEGs and immune. (a) Immune cell infiltration between high- and low-risk groups. (b) The connection between prognostic signature and immune checkpoints.