Abstract
PURPOSE: Microarray analysis was used to bring a comprehensive insight into underlying molecular mechanisms and obtain a whole assessment of aberrant gene expression in nasopharyngeal carcinoma (NPC). METHODS: Combined with microdissection, gene expression profiles in 23 NPCs and 10 nontumor nasopharyngeal epithelial tissue samples were analyzed. RESULTS: Gene expression patterns suggested the dysregulation of the GTP/GDP-bound Ras cycle and an abnormal hyperactivity of cell cycle in NPC. Alterations in the WNT pathway suggest that this pathway may be activated in NPC. A 6-feature weighted-voting model was chosen because it represented the main characteristics of NPCs and predicted NPCs most accurately from the nontumor tissues (33 of 34 correct calls; 97.1% accuracy, Fisher's exact test, P value = 8.389 x 10(-8)). CONCLUSIONS: The data generated in this study represent a comprehensive list of genes aberrantly regulated in NPC. The 6-feature weighted-voting model may provide an extensive list of potential molecular markers for early diagnosis.