[Molecular abnormalities in epithelial ovarian tumors: present and future]

Ovarian cancer is the fourth most common cancer in women. Its pronostic is dreadful and, in spite of numerous studies, the steps of ovarian carcinogenesis are unclear. Histologically, three sub-types of ovarian tumors (benign, borderline and invasive) are distinguished, suggesting the existence of a continuum. However, as each sub-type presents its own biologic characteristics, the hypothesis of the progression of a pre-neoplastic precursor (benign or borderline tumor) into an invasive tumor is still open to discussion. Numerous molecular biological studies have been conducted on ovarian tumors, with the aims of identifying their molecular abnormalities and better understanding the process of ovarian carcinogenesis. Synthesis of the published data (concerning oncogene amplification and/or surexpression, loss of heterozygosity, tumor suppressor gene inactivation, microsatellite instability) shows that there are numerous abnormalities, confirming the heterogeneity and the complexity of these tumors. Hence, it remains very difficult to draw a scheme of ovarian carcinogenesis. Nevertheless, in a near future the new technology of laser microdissection may improve the quality of the results and the study of early ovarian lesions. Indeed, with this technique, it becomes possible to isolate well-defined and homogeneous cell populations and to study small or architecturally complex (surface lesions) tumors. In the next years, the results obtained may allow the identification of early events of the ovarian carcinogenesis and the development of diagnostic and therapeutic tools.