Towards improving the utility of fine-needle aspiration biopsy for the diagnosis of thyroid tumours

Standard cytopathology is an excellent tool for distinguishing benign from malignant thyroid nodules. When a thyroid FNAB is benign or malignant, the sensitivity (97%) and specificity (99%) are nearly perfect (Belfiore et al., 1992). An indeterminate or suspicious FNAB, however, is much less helpful in distinguishing benign from malignant thyroid nodules. Many potential molecular markers have been studied to improve the accuracy of FNAB, particularly those indeterminate or suspicious biopsies (Table 1). Two markers, thyroid peroxidase and galectin-3, seem to be particularly promising as practical tools to improve the accuracy of FNAB. These markers accurately distinguish benign from malignant thyroid nodules, even in the subset of cytologically indeterminate FNABs. These results have been confirmed by several investigators. The antibodies are commercially available, and the assays can be applied to products of FNABs in a standard clinical pathology laboratory. Other potential markers are promising, but do not yet fulfil the above criteria required for use in clinical practice. One promising marker, Pax8-PPAR, is a genetic rearrangement between a thyroid-selective gene (Pax8) and a nuclear transcription factor (PPAR) that may be a useful marker for follicular thyroid carcinoma (Kroll et al., 2000). This genetic rearrangement may also provide insight into mechanisms of progression to follicular carcinoma and behaviour of this thyroid cancer subtype. More studies are needed to define the role of this fusion protein and its use as a molecular marker for thyroid neoplasms. We predict that molecular markers will be used in the near future to improve diagnostic accuracy of FNAB for thyroid nodules, particularly those which are cytologically indeterminate.