An anoikis-based gene signature for predicting prognosis in malignant pleural mesothelioma and revealing immune infiltration

Abstract

Introduction: Malignant pleural mesothelioma (MPM) is an aggressive, treatment-resistant tumor. Anoikis is a particular type of programmed apoptosis brought on by the separation of cell-cell or extracellular matrix (ECM). Anoikis has been recognized as a crucial element in the development of tumors. However, few studies have comprehensively examined the role of anoikis-related genes (ARGs) in malignant mesothelioma.

Methods: ARGs were gathered from the GeneCard database and the Harmonizome portals. We obtained differentially expressed genes (DEGs) using the GEO database. Univariate Cox regression analysis, and the least absolute shrinkage and selection operator (LASSO) algorithm were utilized to select ARGs associated with the prognosis of MPM. We then developed a risk model, and time-dependent receiver operating characteristic (ROC) analysis and calibration curves were employed to confirm the ability of the model. The patients were divided into various subgroups using consensus clustering analysis. Based on the median risk score, patients were divided into low- and high-risk groups. Functional analysis and immune cell infiltration analysis were conducted to estimate molecular mechanisms and the immune infiltration landscape of patients. Finally, drug sensitivity analysis and tumor microenvironment landscape were further explored.

Results: A novel risk model was constructed based on the six ARGs. The patients were successfully divided into two subgroups by consensus clustering analysis, with a striking difference in the prognosis and landscape of immune infiltration. The Kaplan-Meier survival analysis indicated that the OS rate of the low-risk group was significantly higher than the high-risk group. Functional analysis, immune cell infiltration analysis, and drug sensitivity analysis showed that high- and low-risk groups had different immune statuses and drug sensitivity.

Conclusions: In summary, we developed a novel risk model to predict MPM prognosis based on six selected ARGs, which could broaden comprehension of personalized and precise therapy approaches for MPM.

Keywords: Anoikis; Immune infiltration landscape; Malignant pleural mesothelioma; Prognosis; Risk model.

Fig. 1

Flowchart of this study

Fig. 2

Characteristics and differences of anoikis-related genes of MPM. A Volcano diagram shows the DEGs. B Univariate Cox regression analysis of the ARGs. C Network diagram showed the correlations between the most relative ARGs–OS. D CNV variation frequency of the ARGs–OS in TCGA–MESO. E Chromosome localization and alteration of ARGs–OS

Fig. 3

Result of consensus clustering and immune infiltration landscape analysis. A Consensus clustering heatmap shows the optimal classification of MPM samples with K = 2. B–D PCA, tSNE, UAMP analysis shows a significant distribution pattern of patients in Cluster A and Cluster B. E ARGs expression in two clusters. F Overall survival of two clusters (p < 0.001)

Fig. 4

Construction and validation of the risk model. A Screening of prognostic ARGs via cross validation. B Curve of error rate tenfold cross-validation. C Heatmap diagram displays the expression of the prognostic ARGs in the low- and high-risk groups. D, E Kaplan–Meier survival curve analysis shows the OS rate of patients in the low- and high-risk group in the training set and validation set. F, G Time-dependent ROC curve shows the AUC at 1, 2, and 3 years in the training set and validation set

Fig. 5

Construction and validation of the risk model. A Univariate and multi-variate COX analyses to assess risk scores and clinical features B Nomogram of risk score and clinical characteristics predicting 1-, 2-, and 3-year survival. C Calibration curves of the nomogram. D Risk score in two clusters established before. E Alluvial diagram is about the relationship between A and B clusters, high- and low-risk groups, and living status

Fig. 6

Functional analysis of cluster A and cluster B. A, B GSVA analysis between cluster A and cluster B. C–F GSEA between cluster A and cluster B

Fig. 7

Immune microenvironment landscape analysis. A Immune infiltration patterns in cluster A and cluster B. B Differences in immune cell composition between the high-risk and low-risk groups. C Proportional percentage of immune cells in each sample. D Correlation between immune cells. E Association between the six ARGs together with risk score and immune cells

Fig. 8

Correlation between risk score and immune cells. A–E Correlation analysis between risk score and the proportion of Macrophages M0, Macrophages M2, cor.T cells gamma delta, cor.NK cells activated, cor.Plasma cells

Fig. 9

Tumor environment and drug sensitivity analysis. A StromalScore, ImmuneScore, and ESTIMATEScore between high- and low-risk groups. B–I IC50 values of Afatinib, Afuresertib, Bortezomib, Cisplatin, Crizotinib, Erlotinib, Osimertinib and Dasatinib between high- and low-risk groups