Classification and terminology of non-low-grade endometrial sarcomas, which show significant nuclear atypia, have been controversial. Currently, these tumors seem to be classified all together into "undifferentiated endometrial sarcoma (UES)." However, it remains unclear whether these non-low-grade sarcomas are universally "undifferentiated." We divided these sarcomas morphologically into undifferentiated endometrial sarcoma with nuclear uniformity (UES-U) and undifferentiated endometrial sarcoma with nuclear pleomorphism (UES-P), and compared their molecular genetic and immunohistochemical profiles. Eighteen low-grade endometrial stromal sarcomas (ESS-LG), 7 UES-U, and 6 UES-P were examined. All the patients with ESS-LG were still alive, either with or without disease, whereas 4 of the 5 patients with advanced stage UES-U and all 3 of the patients with advanced stage UES-P had died of the disease. JAZF1-JJAZ1 fusion transcript was detected in 6 (50%) out of 12 ESS-LG and in 1 (33%) of 3 UES-U, whereas it was not detected in any of the cases of UES-P. ESS-LG and UES-U frequently showed positive immunoreaction for estrogen receptor (ESS-LG: 94%, UES-U: 57%) and progesterone receptor (ESS-LG: 94%, UES-U: 57%), whereas all the UES-P were negative for these receptors. Nuclear beta-catenin expression was more frequently recognized in ESS-LG (47%) and UES-U (85%), compared with UES-P (33%). Moreover, nuclear accumulation of p53 and TP53 gene missense mutations were limited to 3 UES-P cases. Our data suggest that UES-U shares some molecular genetic and immunohistochemical characteristics with ESS-LG, but UES-P considerably differs from ESS-LG.

Aberrant activation of the Wnt signaling pathway has been implicated in tumorigenesis of a wide range of tumors, including colorectal cancer. Regarding endometrial stromal tumors and related high-grade sarcomas, there have been some reports regarding nuclear accumulation of b-catenin. To clarify the function of the aberrant Wnt signaling pathway in these tumors, we searched for mutations of the CTNNB1 (b-catenin) gene and APC gene by PCR direct sequencing and analyzed the methylation status of SFRP genes. We also examined overexpression of cyclin D1 and MMP-7, which are direct target genes of b-catenin. Eight endometrial stromal nodules, 16 low-grade endometrial stromal sarcomas, and 13 undifferentiated endometrial sarcomas were examined. PCR and direct sequencing revealed no mutation of the b-catenin gene or the APC gene. Concerning the promoter methylation status of SFRP genes, methylation-specific PCR revealed no significant difference between the group with nuclear b-catenin expression and that without nuclear b-catenin expression. Immunohistochemistry revealed overexpression of cyclin D1 in 2 out of 8 endometrial stromal nodules, 1 out of 17 low-grade endometrial stromal sarcomas, and 6 out of 13 undifferentiated endometrial sarcomas, and these 6 undifferentiated endometrial sarcomas simultaneously expressed nuclear b-catenin. Interestingly, all six undifferentiated endometrial sarcoma cases with cyclin D1 overexpression histologically featured rather uniform nuclei. In contrast, the six cases of undifferentiated endometrial sarcoma with highly pleomorphic nuclei were all negative for cyclin D1. In conclusion, among endometrial stromal tumors and related sarcomas, undifferentiated endometrial sarcomas featuring uniform nuclei were characterized by frequent coincident expression of b-catenin and cyclin D1. This finding raises the possibility that cyclin D1 is upregulated by b-catenin in these high-grade sarcomas previously called high-grade endometrial stromal sarcoma.

No abstract

We have confirmed frequent nuclear localization of E-cadherin but not beta-catenin in human ovarian GCT, ovarian follicles and stroma. There is no evidence of misregulated Wnt/beta-catenin signalling (represented by nuclear expression of beta-catenin) in human ovarian GCT. Nuclear translocation of E-cadherin might contribute to ovarian folliculogenesis or granulosa/stromal cell differentiation.

SCC in MCT may be derived from metaplastic squamous epithelium.