Cytogenetic findings in malignant mixed mesodermal tumors of the uterus

Genomics of gynaecological carcinosarcomas and future treatment options

Gynaecological carcinosarcomas are the most lethal gynaecological malignancies that are often highly resistant to standard chemotherapy. They are composed of both carcinomatous and sarcomatous components and are associated with high rates of metastatic disease. Due to their rarity, molecular studies have been carried out on relatively few tumours, revealing a broad spectrum of heterogeneity. In this review, we have collated the gene mutations, gene expression, epigenetic regulation and protein expression reported by a number of studies on gynaecological carcinosarcomas. Based on these results, we describe potential therapeutics that may demonstrate efficacy and present any pre-clinical studies that have been carried out. We also describe the pre-clinical models currently available for future research to assess the potential of molecularly matched therapies. Interestingly, over-expression of many biomarkers in carcinosarcoma tumours often doesn't correlate with a worse prognosis. Therefore, we propose that profiling the mutational landscape, gene expression, and gene amplification/deletion may better indicate potential treatment strategies and predict response, thus improving outcomes for women with this rare, aggressive disease.

Primary Peritoneal Malignant Mixed Mesodermal (Müllerian) Tumor

Aims and background

Malignant mixed mesodermal tumor (MMMT) is a biphasic neoplasm (carcinosarcoma) composed of both epithelial and mesenchymal elements. Extragenital MMMT, including primary peritoneal MMMT, is an extremely rare tumor with features consistent with its origin from the secondary Müllerian system. The neoplastic elements of extragenital MMMT presumably arise directly from the mesothelium or submesothelial stroma and hence parallel the biphasic pattern of the genital (uterine or ovarian) counterpart.

Methods and study design

Here we report on the clinical, pathological, and immunohistochemical features of a case of peritoneal MMMT in a 65-year-old woman. The patient presented with abdominal fullness and pain. Gynecological examination revealed a huge pelvic abdominal mass. On histology, the tumor consisted of poorly differentiated carcinomatous and sarcomatous (rhabdomyosarcoma) components. Further immunohistochemical analysis revealed positive reactivity for both epithelial (cytokeratin and epithelial membrane antigen) and mesenchymal (vimentin, S-100, and desmin) markers. The patient refused treatment and died of the disease three months later.

Results and conclusions

Based on the present case and on previous studies, primary peritoneal MMMT seems to be a rare but highly malignant neoplasm with an aggressive behavior and poor prognosis. Its exact origin, histogenesis and molecular alterations are poorly understood.

Genomic aberrations in carcinomas of the uterine corpus

Endometrial carcinoma, the most common invasive neoplasm of the female genital tract, occurs either in a hormone-related, less virulent form (type I) or in a hormone-independent, more aggressive form (type II). Another cancer of the uterine corpus is carcinosarcoma, a biphasic or mixed epithelial-mesenchymal tumor, now classified as metaplastic carcinoma. We examined by karyotyping and comparative genomic hybridization a consecutive series of 67 endometrial carcinomas and 15 carcinosarcomas and compared the cytogenetic features of the different carcinoma subtypes. All three subtypes of uterine carcinoma had in common a nonrandom gain of material from 1q and 8q but differed from one another in other respects. Endometrial carcinomas of type I mostly presented gains from chromosome arms 1q and 8q and losses from Xp, 9p, 9q, 17p, 19p, and 19q, whereas endometrial carcinomas of type II showed a more complex imbalance picture, with gains from chromosome arms 1q, 2p, 3q, 5p, 6p, 7p, 8q, 10q, and 20q and losses from Xq, 5q, and 17p. The carcinosarcomas mostly showed gains of or from 1q, 5p, 8q, and 12q but losses from 9q, that is, they were much more similar to endometrial carcinomas in their pattern of acquired genomic changes than to sarcomas of the uterine corpus. It was also possible to identify different copy number changes among the different grades of type I carcinomas, between serous papillary and clear-cell carcinomas of type II, as well as between homologous and heterologous carcinosarcomas. Specifically, type I adenocarcinomas that were highly differentiated mostly showed gains from 1q and 10p; those that were moderately differentiated showed gains from 1q, 7p, 7q, and 10q as well as losses from Xp, 9p, 9q, 17p, 19p, and 19q; whereas those poorly differentiated showed gains from 1q, 2p, 2q, 3q, 6p, 8q, and 20q but losses from Xp, Xq, 5q, 9p, 9q, 17p, and 17q. The serous papillary carcinomas showed gains from 1q, 2p, 2q, 3q, 5p, 6p, 6q, 7p, 8q, 18q, 20p, and 20q but losses from 17p, whereas the clear-cell carcinomas showed gains from 3q, 7p, 8q, 10q, 16p, and 20q but losses from 6q. Finally, the homologous carcinosarcomas presented gains from 1p, 1q, 8q, 12q, and 17q as well as losses from 9q and 13q, whereas the heterologous tumors showed gains from 1q, 8p, and 8q. The reproducibility of the observed correlations between karyotypic aberration patterns and histological differentiation was underscored by the fact that those carcinosarcomas whose epithelial component resembled type I endometrial carcinomas also exhibiting a type I aberration profile, whereas carcinosarcomas with a type II carcinoma differentiation had karyotypic abnormalities similar to those of type II endometrial carcinomas.