Résumé
BACKGROUND: Currently, in follicular lesion of aspirates of thyroid, pathologic evaluation of surgical specimen is the only diagnostic method whether the patient had follicular thyroid malignancy or not. The aim of this study is the evaluation of the clinical utility of HBME-1 immunostaining in the diagnosis of follicular thyroid malignancy in surgical specimen, and to establish the diagnostic guideline of HBME-1 immunostaining. METHODS: From 1994 to Sep. 1999, the 72 paraffin embedded tissue, which was already diagnosed as thyroid follicular carcinoma or adenoma through the pathologic evaluation of surgical specimen, were studied. Among 72 specimens, the 29 follicular carcinoma were included, and the others were follicular adenoma. The specimens were stained with HBME-1 monoclonal antibody by standard avidin-biotin peroxidase complex methods. One limited pathologist had read the findings of the immunostaining with a basis such as percent of tumor area. These percentage were divided to 4 grade as follows: 1) Grade 0: negative stained, 2) Grade 1: stained area or = 60%. After we had set a basis of follicular carcinoma as more than Grade 2, defined the clinical utility of HBME-1 immunostaining. The clinical utility was based that the concordance rate between pathologic diagnosis and the findings of immunostaining was more than 80% in both groups. RESULTS: 1) There was significant difference between two groups in intensity of cellular staining (p=0.04, x2). But, there might not be helpful to rule out follicular carcinoma of thyroid from adenoma in fine-needle aspirates. 2) In both groups, the percent of stained area of tumor was very diverse from 0% to 100%, and was statistically significant different (p=0.007). 3) Because the only 5 cases of normal tissue in both groups were stained weakly, the HBME-1 immunostaining was like to specific reaction with tumor tissue in both groups. 4) When we had set a basis of follicular thyroid carcinoma as more than Grade 2 (> or = 30%), the concordance rate between pathologic diagnosis and the findings of immuno- staining was 69.7% in follicular adenoma, 65.5% in follicular carcinoma, respectively. CONCLUSION: The HBME-1 immunostaining may not be help to differentiate follicular carcinoma from adenoma.