Clear cell carcinoma (CCC) of the ovary is the surface epithelial neoplasm most often confused with primitive germ cell tumors, particularly yolk sac tumor (YST) and dysgerminoma. OCT3/4 has proven to be a sensitive and relatively specific marker for the latter entity, but existing markers for YST are limited. Recent studies suggest that glypican-3 (GPC3), an oncofetal protein expressed in fetal liver and malignant tumors of hepatocytic lineage, is also expressed in germ cell tumors, particularly YST. To investigate whether GPC3 is useful in distinguishing YST from ovarian CCC, we studied the expression of GPC3 in a large series of ovarian neoplasms and compared it to the expression profiles of CK7 and alpha-fetoprotein. Tissue microarrays containing over 400 benign and malignant ovarian neoplasms, including 34 CCCs were stained with monoclonal GPC3 (clone 1G12, Biomosaics, Burlington, VT). These arrays contained a wide assortment of ovarian surface epithelial neoplasms and sex cord stromal neoplasms, as well as germ cell tumors. Full paraffin tissue sections from 32 YSTs and 10 CCCs were also assessed. All but one YST (97%), including those associated with mixed germ cell tumor were positive for GPC3, whereas all teratomas and embryonal carcinomas were negative. Both cytoplasmic and membrane staining were present in the positive cases, with no background staining. The syncytiotrophoblastic cells in the germ cell tumors and placental villi included in the arrays were also positive for GPC3. Most CCCs (83%) were completely negative for GPC3, as were 99% serous, 94% endometrioid, and 100% mucinous tumors. Five CCCs exhibited focal, moderate to strong GPC3 expression and in 2 the expression was focal and weak. All other tissues, including normal ovary were negative for GPC3. GPC3 seems to be a promising diagnostic marker for differentiating YST from ovarian CCC (P < 0.0001). Because GPC3 may be associated with alpha-fetoprotein expression, further studies are required to determine the utility of GPC3 in differentiating YST from CCC with hepatoid differentiation.

Purpose: We recently described two types of stromal response in breast cancer derived from gene expression studies of tenosynovial giant cell tumors and fibromatosis. The purpose of this study is to elucidate the basis of this stromal response-whether they are elicited by individual tumors or whether they represent an endogenous host reaction produced by the patient.Experimental Design: Stromal signatures from patients with synchronous dual primaries were analyzed by immunohistochemistry on a tissue microarray (n ¼ 26 pairs) to evaluate the similarity of stromal responses in different tumors within the same patient. We also characterized the extent to which the stromal signatures were conserved between stromal response to injury compared to the stromal response to carcinoma using gene expression profiling and tissue microarray immunohistochemistry.Results: The two stromal response signatures showed divergent associations in synchronous primaries: the DTF fibroblast response is more likely to be similar in a patient with multiple breast primaries (permutation analysis P ¼ 0.0027), whereas CSF1 macrophage response shows no significant concordance in separate tumors within a given patient. The DTF fibroblast signature showed more concordance across normal, cancer, and biopsy site samples from within a patient, than across normal, cancer, and biopsy site samples from a random group of patients, whereas the CSF1 macrophage response did not.Conclusions: The results suggest that the DTF fibroblast response is host-specific, whereas the CSF1 response may be tumor-elicited. Our findings provide further insight into stromal response and may facilitate the development of therapeutic strategies to target particular stromal subtypes.