Thyroid carcinoma is collectively the most common of the endocrine cancers and overall has a good prognosis if diagnosed early. However, the molecular events underlying development of the two most common types, follicular (FTC) and papillary thyroid carcinoma (PTC), is not well understood. Furthermore, pre-operative distinction between FTC and benign follicular adenomas (FA) by fine-needle aspiration remains a challenge. Microarray expression analysis is a powerful tool that measures the levels of expression of thousands of different genes simultaneously. Statistical algorithms are then used to cluster samples according to their expression profiles. We have conducted a series of microarray expression analyses on PTC, FTC, FA and normal thyroid tissue. Comparison of PTC and FTC showed that while there are many changes in common between the two tumour types, there are also clear differences in their expression profiles. Most intriguingly, a small group of genes showed over-expression in PTCs but were under-expressed in FTCs, relative to normal thyroid tissue. Several genes have been newly-implicated in the aetiology of FTC by these studies, including caveolin-1 and the maternally-imprinted tumour suppressor gene aplysia ras homolog I (ARHI). ARHI is of particular interest, since normal expression was retained in most FAs examined. This gene may therefore act as a marker that distinguishes between benign thyroid nodules and those prone to malignant transformation. Comparison of the expression profiles of FTC and FA has identified three additional genes that can predict the diagnosis with high sensitivity and specificity. Thus, microarray expression analysis has already provided new insight into the genetic basis of different types of thyroid carcinoma and identified promising targets for further investigation that might ultimately improve pre-operative diagnosis.
07 - 09 Nov 2005
Society for Endocrinology