Proffered papers
14.00–15.00 Tuesday 16 September 2003 11 Galectin-3 in thyroid nodules: a cytohistological correlation

Aim To assess sensitivity and specificity of galectin-3 to identify malignant lesions of thyroid in cytological specimens.

Methods Fifteen FNA biopsies with a histological correlation out of 392 have been considered suitable for immunostaining with galectin-3. Galectin-3 (mouse monoclonal antibody supplied by Novocastra Laboratories Ltd, Newcastle-upon-Tyne) primary antibody solution was used at a concentration of 1 : 400. Immunostaining was assessed using a score from 0 to 3 considering 0–1, negative and 2–3, positive.

Results Four cases were cytologically reported as malignant (27%), five as suspicious (33%), six as benign (40%). Histologically, six were follicular adenomas, two hyperplastic nodules, three papillary carcinomas, two follicular carcinomas and one was a medullary carcinoma.

Only in one case galectin-3 immunostaining was positive in cytology, which resulted in histology to be a benign hyperplastic nodule. In histology, the galectin-3 immunostaining was positive both in benign (three of nine) and malignant (three of six) nodules and negative in half of the carcinomas (three of six). Only in one case, galectin-3 immunostaining was consistently positive both in cytology and histology (hyperplastic colloid nodule).

Conclusion In our experience, galectin-3 staining is poorly reproducible in cytological and histological samples and is of little help in the differential between malignant and benign lesions of the thyroid.