Abstract

Objective The purpose of this study is to investigate CD56 staining in ovarian granulosa cell tumor and its morphological mimics in order to determine the value of CD56 staining in a diagnostic setting. Materials and methods Tissue samples taken from 82 ovarian tumors, 26 extra-ovarian tumors and 11 normal ovaries were immunohistochemically stained using monoclonal anti-CD56 antibody. Ovarian tumors comprised 32 granulosa cell tumors, 3 Sertoli-stromal cell tumors, 14 fibrothecomas, 6 carcinoid tumors, 1 large cell neuroendocrine carcinoma, 17 endometrioid adenocarcinomas and 9 poorly differentiated serous adenocarcinomas. Extra-ovarian tumors comprised 22 uterine endometrial stromal sarcomas and 4 pulmonary small cell carcinomas. Normal ovaries contained 47 ovarian follicles. Results All of the 32 granulosa cell tumors, all of the 3 Sertoli-stromal cell tumors, all of the 4 small cell carcinomas, 1 of 1 large cell neuroendocrine carcinoma, 11 of 14 fibrothecomas, 5 of 6 carcinoid tumors, 17 of 22 endometrial stromal sarcomas and 7 of 9 poorly differentiated serous adenocarcinomas were positive for CD56. No immunoreactive cells were observed in 17 endometrioid adenocarcinomas or 47 ovarian follicles. All the immunoreactive cells showed membranous staining except for fibrothecomas where vague cytoplasmic staining was seen. Conclusion CD56, known as a neuroendocrine marker, is a sensitive marker of granulosa cell tumors, but since granulosa cell tumors and neuroendocrine tumors may be morphologically similar, CD56 positivity represents a significant diagnostic pitfall. CD56 is useful in distinguishing between granulosa cell tumor and normal ovarian follicles or endometrioid adenocarcinoma. Lack of membranous CD56 expression in fibrothecoma may help differentiate it from granulosa cell tumor. However, CD56 is of limited use for distinguishing between granulosa cell tumor and poorly differentiated carcinoma or endometrial stromal sarcoma. Appropriate and cautious interpretation of CD56 expression should lead to a more accurate diagnosis of granulosa cell tumor.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call