Gene-expression of metastasized versus non-metastasized primary head and neck squamous cell carcinomas: A pathway-based analysis

Abstract

Abstract Background Regional lymph node metastasis is an important prognostic factor in head and neck squamous cell carcinoma (HNSCC) and plays a decisive role in the choice of treatment. Here, we present an independent gene expression validation study of metastasized versus non-metastasized HNSCC. Methods We used a dataset recently published by Roepman et al. as reference dataset and an independent gene expression dataset of 11 metastasized and 11 non-metastasized HNSCC tumors as validation dataset. Reference and validation studies were performed on different microarray platforms with different probe sets and probe content. In addition to a supervised gene-based analysis, a supervised pathway-based analysis was performed, evaluating differences in gene expression for predefined tumorigenesis- and metastasis related gene sets. Results The gene-based analysis showed 26 significant differentially expressed genes in the reference dataset, 21 of which were present on the microarray platform used in the validation study. 7 of these genes appeared to be significantly expressed in the validation dataset, but failed to pass the correction for multiple testing. The pathway-based analysis revealed 23 significant differentially expressed gene sets, 7 of which were statistically validated. These gene sets are involved in extracellular matrix remodeling (MMPs, MMP regulating pathways and the uPA system), hypoxia and angiogenesis (HIF1α regulated angiogenic factors and HIF1α regulated invasion). Conclusion Pathways that are differentially expressed between metastasized and non-metastasized HNSCC are involved in the processes of extracellular matrix remodeling, hypoxia and angiogenesis. A supervised pathway-based analysis enhances the understanding of the biological context of the results, the comparability of results across different microarray studies, and reduces multiple testing problems by focusing on a limited number of pathways of interest instead of analyzing the large number of probes available on the microarray.