Combined immunohistochemistry for thyroid peroxidase, galectin-3, CK19 and HBME-1 in differential diagnosis of thyroid tumors.

We evaluated some proposed molecular thyroid tumor markers: thyroid peroxidase (TPO), galectin-3, cytokeratin-19, and HBME-1, individually and in combination, by immunohistochemistry in a total of 242 archival thyroid tissue sections. The expression of each individual marker was most helpful for the diagnosis of papillary carcinoma and its follicular variant. However, none of them was sensitive and specific enough to discriminate between Hürthle adenoma and carcinoma. Galectin-3 and HBME-1 could be used as single discriminators between follicular thyroid adenoma and carcinoma, but HBME-1 is the better choice. As a single test, all analyzed tumor markers had sufficient power to predict differentiated thyroid cancer, with sensitivities ranging from 66.5% to 82.2%. The sensitivity was improved by using combinations of some proposed markers. Only two antigens, HBME-1 and TPO, had distinct predictive values for different diagnostic alternatives i.e. a sequential combination improved diagnostic accuracy between follicular thyroid adenoma and the follicular variant of papillary thyroid carcinoma to 92.6% and consequently, between overall benign and malignant thyroid tumors to 89.1%. HBME-1 is the most accurate ancillary stain in discriminating well-differentiated thyroid carcinomas from benign tumors, although the addition of TPO did improve accuracy and served as a useful confirmatory marker.

Serum Cyfra 21.1 and galectin-3 protein levels in relation to immunohistochemical cytokeratin 19 and galectin-3 expression in patients with thyroid tumors.

PURPOSE: The aim of this study was to assess the clinical utility of circulating preoperative Cyfra 21.1 [soluble fragment of cytokeratin (CK) 19] and galectin-3 (gal-3) in patients with thyroid tumors, to compare their serum values with tissue expression and to analyze the prognostic significance of these markers in relation to the clinical status of postsurgical differentiated thyroid carcinoma (DTC) patients. PATIENTS: Concentrations of Cyfra 21.1 and gal-3 were evaluated by immunoassays in sera of 9 healthy subjects, 97 preoperative patients with diverse thyroid tumors (10 FTA, 63 PTC, 11 FTC, 5 PDTC, 4 ATC, 4 LNM) and 25 postoperative DTC patients (14 remissions and 11 metastases). RESULTS: Low Cyfra 21.1 values were found in all subgroups, but with a tendency toward higher values in poorly differentiated DTC patients. Compared to the control (0.23 ng/mL), serum levels of gal-3 were significantly elevated in patients with thyroid tumors but with overlapping between adenoma (4.16 ng/mL) and carcinoma (3.85, 4.37, 4.64, 6.07 ng/mL for PTC, PDTC, ATC and LNM, respectively). The tissue expression of CK19 and gal-3 was immunohistochemically determined on 45 matched paraffin-embedded sections. Most thyroid carcinomas showed positive CK19 (27/35) and gal-3 immunostaining (31/35), while adenomas were mostly immunonegative (8/10 and 7/10, respectively). However, there was no significant correlation between their serum and tissue levels. Clinical status of postoperative DTC patients had no influence on serum concentrations of the tested markers. CONCLUSIONS: While CK19 and gal-3 are accurate as tissue markers, their serum levels could not be used as reliable markers for identification of thyroid malignancy or in thyroid cancer follow-up. On the other hand, a tendency toward higher serum levels of Cyfra 21.1 in the small number of PDTC patients examined adds weight to previous reports postulating a role for cytokeratins in predicting a high degree of malignancy.

Thyroid peroxidase and galectin-3 immunostaining in differentiated thyroid carcinoma with clinicopathologic correlation.

Thyroperoxidase and galectin-3 have been reported as useful immunohistochemical markers of thyroid malignancy. In this study, we evaluated the relationship between immunohistochemical staining results for these markers and clinicopathologic features of patients with differentiated thyroid cancer. A total of 193 archival thyroid samples including 28 follicular adenomas, 18 follicular carcinomas, and 147 papillary carcinomas with 114 adjacent thyroid tissues were analyzed by immunohistochemistry. Thyroperoxidase was underexpressed (<50% stained thyrocytes), and galectin-3 was expressed (>5% stained thyrocytes) in most carcinomas. The sensitivity for diagnosis of differentiated thyroid carcinoma was 86.1% for thyroperoxidase and 82.4% for galectin-3, whereas the combination of both markers increased the sensitivity up to 94.5%. Thus, the combination of thyroperoxidase and galectin-3 immunohistochemistry may help to ascertain the malignant nature of the lesion. Furthermore, tumor size, nodal involvement, extrathyroidal invasion, and high tumor-node-metastasis stage in patients with papillary carcinoma were related to thyroperoxidase absence and high galectin-3 expression in most cases (P < .05). In patients with follicular carcinoma, the extent of invasiveness was associated with galectin-3 positivity. Thus, expression of these markers is related to more or less aggressive biological behavior of differentiated thyroid carcinomas. Although thyroperoxidase presence may indicate favorable prognosis of papillary cancer, expression of galectin-3 illustrates the potential importance of this protein in the pathogenesis and/or progression of differentiated thyroid carcinomas.

Thyroid peroxidase immunohistochemistry in differential diagnosis of thyroid tumors.

Thyroperoxidase (TPO) is a thyroid-specific enzyme expressed by differentiated thyroid cells. Initial immunohistochemical studies claimed that TPO expression, detected by the monoclonal antibody mAb 47, may be a potentially important diagnostic tool in differentiating malignant from benign lesions. However, some recent studies have failed to reproduce the earlier results, suggesting the limitations for TPO immunohistochemistry. To assess these observations we have evaluated the immunohistochemical expression of TPO in thyroid tissue from 215 patients. The studied material included 87 nonmalignant thyroid lesions and 128 thyroid carcinomas. TPO expression was investigated using newly available mAb 47 and staining of less than 80% of the follicular cells/specimen as the threshold indicating a malignant lesion. We found that TPO had a sensitivity of 89.9% for cancer and a specificity of 64.4% for nonmalignant lesions, showing that it does not give a sufficient degree of diagnostic certainty that the lesion is benign. In addition, the variability in the degree of TPO expression found within and between follicular carcinomas, and the significant number of benign adenomas having similar immunostaining patterns, assured us that TPO immunostaining is not sufficiently discriminatory in the differential diagnosis of thyroid cancer versus benign lesions.