Squamous cell carcinoma of the head and neck (SCCHN) is the most common form of head and neck cancer, and accounts for approximately 3% of malignancies in the USA. Despite significant advances in early detection and treatment options, limited improvement in patient survival has been achieved. The 5-year survival rate for all stages of SCCHN is approximately 35% - 50%. One reason for this poor prognosis is that SCCHN is frequently diagnosed at an advanced stage. DNA methylation has recently been recognized as a useful marker for diagnosing and monitoring the progression of many types of cancer. CpG island sites frequently become hypermethylated during tumorigenesis, and often exhibit distinctive tumor type and stage related patterns. A multi-endpoint ongoing molecular epidemiologic project was designed to evaluate whether there is a panel of methylation loci significantly correlated to SCCHN patient survival. We have conducted a preliminary DNA methylation screening analysis using the Illumina GoldenGate Methylation Cancer Panel I with DNA isolated from primary SCCHN tumor tissues (n=40) from patients who were followed for a minimum of two years from the time of diagnosis. Due to the sensitivity limits of the assay, CpG loci with \#946; scores < 0.25 or > 0.75 in at least 75% of the samples were excluded from the analysis. Similarly, in order to screen for differential methylation patterns, we excluded the loci with small \#946; score ranges which were defined as the mean score ± 12.5% for at least 75% of the tumors. Following this screening process, 458 loci were included in the association analysis to examine the relationship of the degree of methylation with survival among SCCHN patients. Using these loci in an unsupervised correlation-based machine learning feature selection algorithm, we further reduced the relevant data pool to 10 significant loci; 5 of which were hyper-methylated and 5 were hypo-methylated. \#946; scores of these 10 loci were then used in the multiple linear regression optimization that predicted days of survival. Predictions significantly correlated with observed survival (r=0.73, p<0.0001). Using this model in a 10-fold cross validation scheme, we were able to achieve 87.5% predictability of separating the cases into two classes: short survival (< 850 days) and long survival (> 850 days) with 89% sensitivity and 83% specificity. This preliminary finding suggests that methylation status may not only represent a useful biomarker for monitoring and early detection of SCCHN, but may also aid in the identification of genes differentially regulated that closely correlate with patient survival, which in turn could lead to novel strategies for molecular targeted therapies and patient quality of life. A larger study population with extended follow-up data is required to further validate the observed correlation between DNA methylation and survival. [NIH 1P50 CA097190].

Citation Information: In: Proc Am Assoc Cancer Res; 2009 Apr 18-22; Denver, CO. Philadelphia (PA): AACR; 2009. Abstract nr 4842.

100th AACR Annual Meeting-- Apr 18-22, 2009; Denver, CO