Expression of CK-19, galectin-3 and HBME-1 in the differentiation of thyroid lesions: systematic review and diagnostic meta-analysis.

To distinguish between malignant and benign lesions of the thyroid gland histological demonstration is often required since the fine-needle aspiration biopsy method applied pre-operatively has some limitations. In an attempt to improve diagnostic accuracy, markers using immunocytochemistry and immunohistochemistry techniques have been studied, mainly cytokeratin-19 (CK-19), galectin-3 (Gal-3) and Hector Battifora mesothelial-1 (HBME-1). However, current results remain controversial. The aim of the present article was to establish the diagnostic accuracy of CK-19, Gal-3 and HBME-1 markers, as well as their associations, in the differentiation of malignant and benign thyroid lesions. A systematic review of published articles on MEDLINE and The Cochrane Library was performed. After establishing inclusion and exclusion criteria, 66 articles were selected. The technique of meta-analysis of diagnostic accuracy was employed and global values of sensitivity, specificity, area under the summary ROC curve, and diagnostic odds ratio (dOR) were calculated. For the immunohistochemistry technique, the positivity of CK-19 for the diagnosis of malignant thyroid lesions demonstrated global sensitivity of 81% and specificity of 73%; for Gal-3, sensitivity of 82% and specificity of 81%; and for HBME-1, sensitivity of 77% and specificity of 83%. The association of the three markers determined sensitivity of 85%, specificity of 97%, and diagnostic odds ratio of 95.1. Similar results were also found for the immunocytochemistry assay. This meta-analysis demonstrated that the three immunomarkers studied are accurate in pre- and postoperative diagnosis of benign and malignant thyroid lesions. Nevertheless, the search for other molecular markers must continue in order to enhance this diagnostic accuracy since the results found still show a persistency of false-negative and false-positive tests. Http://www.diagnosticpathology.diagnomx.eu/vs/3436263067345159.