[Low galectin-3 capacity to discriminate thyroid lesions]

Galectin-3 is considered useful in identifying lesions included in the term follicular tumour.

To evaluate galectin-3's power to discriminate among diverse thyroid lesions and to identify a set point of expression at which there is the least possible false positive results.

26 follicular carcinomas and 104 assorted thyroid lesions were analysed (8 Hashimoto's thyroiditis, 18 goiter, 44 adenoma, 27 papillary carcinoma and 7 normal glands). Immunohistochemistry for galectin-3 was performed way (NCL-GAL-3, 1:100, Novocastra, UK). In a blinded manner intensity and percentage of expression were evaluated, as was its capacity to identify the previously mentioned thyroid entities with 2 x 2 tables. Minimum reactivity set point values were identified and the one which presented the least false positive cases was selected.

A reaction was considered positive when 25% of the cells were marked, in which case neither the Hashimoto's thyroiditis, goiter, nor the residual tissue were positive for galectin-3. A positive galectin-3 reaction was identified in 4/44 adenomas, 5/26 follicular carcinomas and 18/27 papillary carcinomas. From a total of 53 carcinomas, 23 showed positive and 30 a negative reaction to galectin-3. Thus, galectin-3's sensitivity to discriminate between benign and malignant lesions was 94% while its specificity was 43%. When comparing follicular carcinoma with adenoma, sensibility was 19% while specificity was 91%. The positive predictive value was 56% and the negative predictive value 66%. Galectin-3 was most useful for identifying papillary thyroid carcinoma.

Galectin-3 has limited value to distinguish benign from malignant thyroid lesions. The reaction must be considered positive when at least 25% of cells are marked.