Distinct immunological features of oropharyngeal cancer peritumoral tonsillar tissues from inflammatory tonsils and regional lymph nodes: A pilot study

Abstract

Background Cancer immune responses are generated in secondary lymphoid organs, such as the lymph nodes and tonsils. In the current study, transcriptional profiles of peritumoral tonsillar tissues (PTTs) from oropharyngeal cancers (OPCs) were assessed and compared with those of inflammatory tonsils and regional lymph nodes (rLNs). Methods RNA samples of PTTs and rLNs from 13 OPCs, and 4 inflammatory tonsils were subjected to microarray analysis, and differentially expressed genes (DEGs) identified from 730 nCounter Panel immune-related genes. Gene Set enrichment Analysis (GSEA) was used for DEG profiling of PTTs and rLNs between lymph node metastasis-negative and metastasis-positive cases. The top 20 genes, as ranked by GSEA metric scores, were extracted and subjected to principal component analysis (PCA). The correlation of each patient’s PCA score with lymph node status was assessed by Receiver Operating Characteristics (ROC) analysis. Results Comparing DEG analyses of PTTs with those of inflammatory tonsils and rLNs revealed 144 and 45 upregulated genes, respectively. ClueGO, a widely used Cytoscape plug-in, revealed activated pathways in PTTs, including lymphocyte proliferation (followed by T cell activation involved in the immune response) and positive regulation of leukocyte migration (followed by antimicrobial humoral immune response mediated by antimicrobial peptides) as the most significantly enriched immune system process functions in the gene ontology when comparing inflammatory tonsils and rLNs. The area under the ROC curves of PTTs and rLNs were 0.806 and 0.389, and were significant by DeLong’s test (p = 0.025). Conclusion PTTs exhibit unique immunological features distinguishing them from inflammatory tonsils and rLNs. Gene expression analysis of PTTs is useful for investigating the mechanism of OPC lymphatic spread, even compared with analysis of rLNs.