Oral and oropharyngeal squamous cell carcinomas (OSCC) are among the most common cancers worldwide. The overall 5-yr survival rate of OSCC patients is approximately 50% but is higher than 80% in early-stage patients. The high mortality of OSCC is due in part to late diagnosis and high recurrence rates. Thus, early detection of OSCC and identification of a good predictor of disease progression are important in improving OSCC outcomes. To identify potential markers for early diagnosis and disease progression, we used Affymetrix U133 plus 2.0 arrays to examine the gene expression profiles of 167 primary OSCC from newly diagnosed patients, 58 “normal” oral mucosa from oral cancer patients (NCA) and 45 normal oral mucosa from patients without cancer (controls), all of whom were enrolled at three University of Washington-affiliated medical centers from 2003 to 2008. We first compared gene expression in the 45 controls to the 167 tumors using ANOVA implemented in Partek® Genomics Suite™ software, adjusting for age, sex, cigarette-smoking, alcohol use and HPV status. We found 2,596 probe sets differentially expressed between tumors and controls, using the criteria of a FDR of 0.05 and at least a two-fold difference in the expression level. We then used linear regression to compare the gene expression level of these 2,596 probe sets between controls and NCA, and between NCA and tumors. Among 2,596 probe sets, 60 were significantly up-regulated and 11 were significantly down-regulated in both comparisons using the Bonferroni correction with a p-value of 1.93x10-5 and a difference in gene expression greater than 1 standard deviation of the expression in NCA. A cluster analysis on gene expression of the 71 probe sets identified a subgroup of 45 OSCC patients with highest/lowest expression of these probe sets. These 45 OSCC patients had a higher risk of disease progression after treatment (defined by persistence or recurrence of OSCC) and higher OSCC-specific mortality when compared with the remaining 122 OSCC patients; the respective Cox regression hazard ratio (HR) and 95% CI, adjusted for age, sex, and stage for disease progression and OSCC-specific mortality were 2.7 (1.5-4.6) and 3.8 (2.1-7.0). The top gene associated with both outcomes was dermatan sulfate epimerase (DSE). A high level of DSE expression in tumor (higher than mean expression among tumors) was associated with a higher risk of disease progression (HR 2.8, 95% CI: 1.5-5.1) and of OSCC-specific mortality (HR 6.0, 95% CI: 2.9-12.2) after adjusting for age, sex and stage. DSE encodes a tumor-rejection antigen that could potentially serve as a target for cancer immunotherapy. In conclusion, we identified 71 probe sets in which dysregulation occurred in both NCA and cancer samples. Some of these genes could potentially serve as early diagnostic markers, markers to predict disease progression, or as potential targets for new therapy.

Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 724. doi:1538-7445.AM2012-724