Transition of epithelial toward mesenchymal differentiation during ovarian carcinosarcoma tumorigenesis

Background. It was the purpose of this study to test the monoclonal theory in ovarian carcinosarcoma. Methods. Twenty-six women with a diagnosis of ovarian carcinosarcoma were subjected to a clinicopathologic analysis. Biopsies from metastatic lesions obtained at primary surgery and surgery for recurrent disease were reviewed. Special attention was paid to the composition of metastatic lesions and to florid desmoplastic reaction as a potential pitfall for the detection of sarcomatous areas. Results. Biopsies derived from metastatic disease at primary surgery (n = 107) consisted of carcinoma cells only (n = 71, 66%), >50% carcinoma cells (n = 21, 20%), >50% sarcoma cells (n = 13, 12%), or sarcoma cells only (n = 2, 2%). The microscopic analysis demonstrated a preponderance of epithelial cells in the primary setting and suggested the epithelial component to drive the tumor, a finding consistent with the monoclonal theory. Biopsies derived from surgery for recurrent disease (n = 8) consisted of carcinoma cells only (0%), >50% carcinoma cells (n = 1, 13%), >50% sarcoma cells (n = 4, 50%), or sarcoma cells only (37%). Since sarcomatous cells dominated the tumorigenic cell population in the recurrent setting, this analysis revealed a change of the composition of metastatic lesions in time when compared to the data in the primary setting. This change was supported by the observation of a threefold higher incidence of sarcoma-dominated metastatic lesions at interval debulking when compared to primary debulking (24 vs 8%, respectively). The potential of a phenotypic change during ovarian cancer progression was further highlighted by the detection of two cases of carcinosarcoma that presented as a recurrence of epithelial ovarian carcinoma. Conclusion. Our results are consistent with the monoclonal theory of ovarian carcinosarcoma histogenesis, but suggest that there is a tendency toward a sarcomatous differentiation during disease progression. These data are important to understand the tumor biology and might have implications for a tailored treatment of ovarian carcinosarcoma. (C) 2003 Elsevier Science (USA). All rights reserved.