Abstract A8: Identification of serum biomarkers for head and neck squamous cell carcinoma

Head and neck cancer remains one of the most devastating cancers in the United States. African American males develop head and neck squamous cell carcinoma (HNSCC) at significantly higher rates with stage at diagnosis for all individuals disproportionately skewed towards stages III and IV. Delayed detection results in decreased survival and increased morbidity while delayed detection of recurrence often results in a situation where there are no curative therapeutic options. This project seeks to identify serum biomarkers for early detection of HNSCC and targets a population that is a racial and socioeconomic cross section from urban Nashville, TN and adjacent suburban and rural areas. Serum proteome analysis using LC-MS/MS identified ten candidate proteins present at higher frequencies in the serum from late stage HNSCC patients compared to controls with a probability of p<0.05. Currently, antibody-based techniques are being used to confirm and validate these potential serum biomarkers in the original study subjects (from the Vanderbilt Ingram Cancer Center) as well as for patients from Nashville General Hospital at Meharry. Preliminary analyses, including antibody selection, are being performed on HNSCC cell lines, tonsil (normal oropharynx) and HNSCC tumor extracts. Three of the proteins identified, S100A8, S100A9 and Fetuin-A (AHSG), are produced by the liver during the systemic response to disease so it is important to delineate between their presence in serum due to that phenomenon and protein being produced by the tumor itself. Western blot (WB) analysis of S100A8 and S100A9 show that detectable expression is limited to 1/6 HNSCC cell lines tested for S100A8 and only normal oropharynx for S100A9. WB analysis for Fetuin-A showed 2/6 HNSCC cell lines positive while tonsil expressed large amounts of the protein and tumor tissue even larger amounts. Interestingly, Fetuin-A detected in cell lines appears larger in size than that in tonsil or tumor tissue. The coxsackie and adenovirus receptor (CAR) has been identified as a potential biomarker for a number of cancers but interestingly, it is in the context of down-regulation as tumor cells become more de-differentiated. As a result it is somewhat unexpected that high serum levels would be a cancer biomarker, however this is the indication by MS analysis. A specific isoform is prognostic for poor outcome in ovarian cancer. WB analysis does identify two isoforms of CAR in HNSCC cell lines that express it (3/6) with high levels in normal tonsil and low to none in tumor tissue. These data tend to correlate with that in the literature in regards to tissue expression. In conclusion, preliminary identification of antibody reagents appropriate for quantitative ELISA of serum has focused, initially, on tissues due to anticipated higher levels of protein expression there. These studies have, in themselves, shown some interesting trends in quantity, expression pattern and apparent “isoforms” that remain to be interpreted within the context of the serum proteome analysis and the disease.