Proof of Principle Gene Signature Reported For Head and Neck Cancer

December 11, 2006

Science is replete with metaphors, and one of the popular ones these days is “gene expression signatures.” The term is rooted in the much-debated question:  Is it possible to select a dozen or more highly informative genes whose expression patterns, when characteristically altered, provide a distinctive pattern, or genetic signature, of a developing tumor?  If so, gene signatures would be a boon to diagnose cancer earlier, when it likely is more amenable to treatment.  Among some head and neck cancers, the answer seems to be yes, although work is ongoing to refine the sensitivity and specificity of the first generation signatures.  But as this work unfolds, some scientists have wondered whether they might also take the next logical step and assemble gene signatures that are prognostic of aggressive head and neck tumors that will spread to nearby lymph nodes.

In the November issue of the journal Neoplasia, NIDCR grantees and colleagues report proof of principle success with two signatures comprised of genes that are broadly known to be involved in tumor metastasis. In the study, the scientists obtained 25 primary oral tongue squamous cell carcinoma samples -of which 11 had spread to the lymph node and 14 were known to be non-metastatic.  Using signatures of 33 and later 22 genes, the scientists found their gene sets provided an overall accuracy rate of 85 percent in correctly classifying metastatic and non-metastatic tumors.  “Our results demonstrated that lymph node metastasis and extracapsular spread can be predicted by gene expression analyses,” the authors concluded.  “This provides a foundation for further validation of the prediction/classification power of identified molecular markers in a large independent set.”