CXCL12 and CXCR4, but not CXCR7, are primarily expressed by the stroma in head and neck squamous cell carcinoma.

The CXCL12/CXCR4 axis is involved in numerous models of metastatic dissemination, including head and neck squamous cell carcinoma (HNSCC). We assessed the relative expressions of CXCL12, CXCR4 and CXCR7 in the stroma and the tumour of HNSCC, and evaluated the methylation status of the CXCL12 promoter.Snap-frozen, HPV negative HNSCC samples were micro-dissected to isolate the tumoural and stromal compartments. The expression levels of CXCL12, CXCR4 and CXCR7 were assessed by qRT-PCR, and the methylation level of the CXCL12 promoter was evaluated by pyrosequencing.In total, 23 matched tumour/stroma samples were analysed. Higher expressions of CXCR4 and CXCL12 were observed in the stroma (p = 0.012 and p < 0.0001, respectively). No significant difference in expression was observed for CXCR7. A high methylation level (>40%) of the CXCL12 promoter was observed in only a few tumoural samples (5/23) and was associated with a lower expression of the gene (p = 0.03).Stromal cells, rather than the tumour itself, are mainly responsible for the expression of both CXCL12 and CXCR4 expression in HNSCC. CXCR7 expression did not differ between the two compartments and was not related to CXCL12 or CXCR4 expression. Finally, the methylation of the CXCL12 promoter could only explain the low intra-tumoural expression of this gene in 20% of cases.

The gene expression profile of inflammatory, hypoxic and metabolic genes predicts the metastatic spread of human head and neck squamous cell carcinoma.

OBJECTIVES: To assess the prognostic value of the expression profile of the main genes implicated in hypoxia, glucose and lactate metabolism, inflammation, angiogenesis and extracellular matrix interactions for the metastatic spread of head and neck squamous cell carcinoma. PATIENTS AND METHODS: Using a high-throughput qRT-PCR, we performed an unsupervised clustering analysis based on the expression of 42 genes for 61 patients. Usual prognostic factors and clustering analysis results were related to metastasis free survival. RESULTS: With a median follow-up of 48months, 19 patients died from a metastatic evolution of their head and neck squamous cell carcinoma and one from a local recurrence. The unsupervised clustering analysis distinguished two groups of genes that were related to metastatic evolution. A capsular rupture (p=0.005) and the "cluster CXCL12 low" (p=0.002) were found to be independent prognostic factors for metastasis free survival. Using a Linear Predictive Score methodology, we established a 9-gene model (VHL, PTGER4, HK1, SLC16A4, DLL4, CXCL12, CXCR4, PTGER3 and CA9) that was capable of classifying the samples into the 2 clusters with 90% accuracy. CONCLUSION: In this cohort, our clustering analysis underlined the independent prognostic value of the expression of a panel of genes involved in hypoxia and tumor environment. It allowed us to define a 9-gene model which can be applied routinely to classify newly diagnosed head and neck squamous cell carcinoma. If confirmed by an independent prospective study, this approach may help future clinical management of these aggressive tumors.