273 Metastatic Breast Carcinoma With CK7, GATA3, and ER Expression? A Diagnostic Pitfall

Abstract

Immunohistochemical stains (IHC) are often used for work-up of carcinoma of unknown primary, and in this context, coexpression of CK7, GATA3, and ER is highly suggestive of metastatic breast cancer. GATA3 is expressed in >90% breast carcinomas, and is more sensitive than mammaglobin and GCDFP. However, specificity is approximately 70%, with expression reported in the majority of urothelial and choriocarcinomas. Estrogen receptor (ER) is often expressed in neoplasms of mammary, uterine, and ovarian origins. This case highlights a major diagnostic pitfall in the interpretation of these stains. A woman in her fifth decade presented with a possible thoracic vertebral lesion following a compression fracture. The biopsy showed dense bone and a fragment of tissue with predominantly plump polygonal cells, with eosinophilic cytoplasm, and hyperchromatic atypical nuclei with mild pleomorphism. IHC showed that the cells were strongly positive for CK7, GATA3, patchy positive for ER, and negative for CDX2, CK20, TTF1, thyroglobulin, mammaglobin, Melan A, and PAX8. The case was signed out as metastatic breast adenocarcinoma. Whole body imaging and PET scans failed to identify a primary tumor. The patient received hormonal therapy for stage 4 ER-positive breast cancer. On review, we noted prominent large blood vessels in the tissue and a few compressed glands on the periphery positive for PAX8 resembling endometrial glands. This tissue was separate from the bone. The morphology, along with the knowledge that intermediate and cytotrophoblasts express GATA3, cytokeratins, and ER led us to conclude that the separate tissue fragment likely represents a tissue contaminant (“floater”) from an endometrial implantation site. This case reminds us that trophoblasts are also positive for GATA3, cytokeratins, and ER, and products of conception can be a dangerous contaminant leading to erroneous metastatic breast cancer diagnosis with severe consequences. IHC should always be interpreted with caution in context of morphology.