Identification of biomarkers associated with the invasion of nonfunctional pituitary neuroendocrine tumors based on the immune microenvironment

Abstract

Introduction: The invasive behavior of nonfunctioning pituitary neuroendocrine tumors (NF-PitNEts) affects complete resection and indicates a poor prognosis. Cancer immunotherapy has been experimentally used for the treatment of many tumors, including pituitary tumors. The current study aimed to screen the key immune-related genes in NF-PitNEts with invasion.

Methods: We used two cohorts to explore novel biomarkers in NF-PitNEts. The immune infiltration-associated differentially expressed genes (DEGs) were obtained based on high/low immune scores, which were calculated through the ESTIMATE algorithm. The abundance of immune cells was predicted using the ImmuCellAI database. WGCNA was used to construct a coexpression network of immune cell-related genes. Random forest analysis was used to select the candidate genes associated with invasion. The expression of key genes was verified in external validation set using quantitative real-time polymerase chain reaction (qRT‒PCR).

Results: The immune and invasion related DEGs was obtained based on the first dataset of NF-PitNEts (n=112). The immune cell-associated modules in NF-PitNEts were calculate by WGCNA. Random forest analysis was performed on 81 common genes intersected by immune-related genes, invasion-related genes, and module genes. Then, 20 of these genes with the highest RF score were selected to construct the invasion and immune-associated classification model. We found that this model had high prediction accuracy for tumor invasion, which had the largest area under the receiver operating characteristic curve (AUC) value in the training dataset from the first dataset (n=78), the self-test dataset from the first dataset (n=34), and the independent test dataset (n=73) (AUC=0.732/0.653/0.619). Functional enrichment analysis revealed that 8 out of the 20 genes were enriched in multiple signaling pathways. Subsequently, the 8-gene (BMP6, CIB2, FABP5, HOMER2, MAML3, NIN, PRKG2 and SIDT2) classification model was constructed and showed good efficiency in the first dataset (AUC=0.671). In addition, the expression levels of these 8 genes were verified by qRT‒PCR.

Conclusion: We identified eight key genes associated with invasion and immunity in NF-PitNEts that may play a fundamental role in invasive progression and may provide novel potential immunotherapy targets for NF-PitNEts.

Keywords: WGCNA; biomarkers; immune microenvironment; invasive; nonfunctioning pituitary neuroendocrine tumors (NF-PitNEts).

Figure 1

Screening for immune and invasion-related genes in pituitary tumors. (A) Volcano plot showing the differentially expressed genes (DEGs) between high Immune-score samples and low Immune-score samples. (B) The functional enrichment analysis of the Immune-related DEGs. (C) Volcano plot showing the differentially expressed genes (DEGs) between patients with and without invasion. (D) The functional enrichment analysis of the Invasion-related DEGs. (E) Multiple immune cell differences between patients with and without invasion.

Figure 2

WGCNA of crucial immune cells in pituitary tumors. (A) Soft power selection of the WGCNA network. Here, we selected 20 as the power. (B) Clustering dendrogram of genes with dissimilarity based on topological overlap and assigned module colors. (C) The relationships between gene modules and immune cells. The P value is shown in parentheses.

Figure 3

Random Forest analysis. (A) Venn diagram showing the candidate genes between immune-related DEGs, invasion-related DEGs and WGCNA modules. We selected 81 common genes for random forest analysis. (B) Random Forest analysis of the 81 genes. (C) Top 50 genes with the lowest MeanDecreaseAccuracy in random forest analysis. (D) The lowest error rate model contains 20 candidate genes based on random forest analysis.

Figure 4

Validation of model classification performance. (A) The ROC curve of the RF model with 20 genes in the training dataset; the AUC reached 0.732. (B) The RF model’s classification performance in the self-test dataset; the AUC reached 0.653. (C) The RF model’s classification performance in the independent test dataset; the AUC reached 0.619.

Figure 5

Twenty crucial genes were enriched in multiple pathways. (A) Functional enrichment analysis of the 20 genes. The results showed that 8 genes played an important role in these pathways. (B) The expression levels of the 8 crucial genes between patients with and without invasion. *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 6

The expression levels of the 8 crucial genes in high Immune-score samples and low Immune-score samples. The results are expressed as the means ± SD (Student’s t test. **p < 0.01, ***p < 0.001, ****p < 0.0001). ns, no significance.

Figure 7

Consensus cluster of the 8 crucial genes. (A) Consensus index of the consensus cluster analysis. The 8 crucial genes could divide the patients into two groups. (B) Heatmap of the two groups divided by the 8 genes. (C) The new model constructed by the 8 crucial genes performed with good classification effectiveness in pituitary tumors. (D) Validation of the 8 crucial genes between 8 CS invasive NF-PitNEts and 8 noninvasive NF-PitNEts. The results are expressed as the means ± SD (Student’s t test. *p < 0.05; **p < 0.01).