Pathology of ovarian cancer precursors

Ninety percent of ovarian cancers in the Western world are epithelial cancers derived from the surface epithelium of the ovary and its inclusion cysts. The so-called surface epithelium is mesothelium that comes to resemble epithelium as it is reflected over the surfaces of the ovaries. At various ages, but particularly in women in the reproductive, menopausal, and postmenopausal age groups, this epithelium migrates into the ovarian stroma to form inclusion cysts. These cysts probably results from a dynamic interplay of surface epithelium and underlying ovarian stroma, but can also develop as a result of periovarian adhesions. There is abundant evidence that their formation is not related to repair of ovulation. It is generally accepted that benign and malignant ovarian epithelial tumors arise from surface epithelium and its cystic derivatives because they both, but particularly the latter, have a potential to differentiate into epithelia similar to those of normal müllerian derivation (tubal, endometrial, and endocervical epithelia) and their tumors resemble those of the fallopian tube, endometrium, and endocervix. Also, both intraepithelial carcinomas and precarcinomatous lesions can be observed in the surface epithelium and its cystic derivatives. These carcinomas may arise de novo or as a transformation of pre-existing benign tumors and non-neoplastic lesions of similar derivation. Surface epithelial inclusion cysts have a greater propensity to undergo neoplasia than does the surface epithelium itself. This difference has been recognized for many years most epithelial ovarian tumors are intraparenchymal, rather than being located on the ovarian surface. More recent evidence includes the immunohistochemical demonstration of various ovarian carcinoma antigens far more frequently in inclusion cyst epithelium than in surface epithelium; and the much more frequent presence of tubal metaplasia in the cyst epithelium than in the surface epithelium. Tubal metaplasia is encountered in non-neoplastic ovaries contralateral to ovarian carcinomas two to three times as frequently as in control ovaries, suggesting that the metaplastic epithelium is more prone to the development of carcinoma that non-metaplastic epithelium. Carcinoma precursors occur in the ovary, as in the cervix and endometrium, but have been reported only rarely because they are easily overlooked and have not been searched for by pathologists.