Morphological and phenotypical features of ovarian metastases in breast cancer patients

Abstract

Background: Autotransplantation of frozen-thawed ovarian tissue is a method to preserve ovarian function and fertility in patients undergoing gonadotoxic therapy. In oncology patients, the safety cannot yet be guaranteed, since current tumor detection methods can only exclude the presence of malignant cells in ovarian fragments that are not transplanted. We determined the need for a novel detection method by studying the distribution of tumor cells in ovaries from patients with breast cancer. Furthermore, we examined which cell-surface proteins are suitable as a target for non-invasive tumor-specific imaging of ovarian metastases from invasive breast cancer.

Methods: Using the nationwide database of the Dutch Pathology Registry (PALGA), we identified a cohort of 46 women with primary invasive breast cancer and ovarian metastases. The localization and morphology of ovarian metastases were determined on hematoxylin-and-eosin-stained sections. The following cell-surface markers were immunohistochemically analyzed: E-cadherin, epithelial membrane antigen (EMA), human epidermal growth receptor type 2 (Her2/neu), carcinoembryonic antigen (CEA), αvβ6 integrin and epithelial cell adhesion molecule (EpCAM).

Results: The majority of ovarian metastases (71%) consisted of a solitary metastasis or multiple distinct nodules separated by uninvolved ovarian tissue, suggesting that ovarian metastases might be overlooked by the current detection approach. Combining the targets E-cadherin, EMA and Her2/neu resulted in nearly 100% detection of ductal ovarian metastases, whereas the combination of EMA, Her2/neu and EpCAM was most suitable to detect lobular ovarian metastases.

Conclusions: Examination of the actual ovarian transplants is recommended. A combination of targets is most appropriate to detect ovarian metastases by tumor-specific imaging.

Keywords: breast cancer; fertility preservation; ovarian metastases; ovarian tissue autotransplantation; tumor markers; tumor-specific imaging.

Fig. 1

Patient selection and composition of the study population. The source population was compiled by the Dutch histopathology and cytopathology network. The exclusion criteria are indicated in the dotted boxes

Fig. 2

Localization of ovarian metastases derived from patients diagnosed with invasive breast cancer. Three examples are shown: (a) a solitary metastasis, (b) multiple distinct nodules separated by uninvolved ovarian tissue and (c) diffuse seeding without any discernable pattern. In order to clearly display the solitary metastasis in (a) and the multiple distinct nodules in (b), a green line is drawn that delineates the metastases in the ovary. Scale bars represent 5 mm

Fig. 3

The correlation between tumor marker expression in breast tumors and ovarian metastases for individual patients. Upper panel (a) shows invasive ductal breast cancer and lower panel (b) represents invasive lobular breast cancer. For each patient, the percentage of positive tumor cells in primary and locally recurrent breast tumors (if applicable) was set against the percentage of positive tumor cells in their corresponding ovarian metastases. EMA, epithelial membrane antigen; Her2/neu, human epidermal growth receptor type 2; EpCAM, epithelial cell adhesion molecule

Fig. 4

Detection of ovarian metastases by a combination of markers. Representative image of a lobular ovarian metastasis stained with DAPI counterstain and triple immunofluorescence for EpCAM (a), EMA (b), Her2/neu (c), and the three stainings combined (d). The solid arrow indicates tumor cells that are positive for EpCAM, but negative for EMA and Her2/neu. The dashed arrow indicates tumor cells that are positive for Her2/neu, but negative for EpCAM and EMA. Scale bars represent 100 μm. EpCAM, epithelial cell adhesion molecule; EMA, epithelial membrane antigen; Her2/neu, human epidermal growth receptor type 2