ABSTRACT
In the past two decades, clinicopathological, immunohistochemical, and molecular genetic studies have provided data to allow for the development of a dualistic model of endometrial
carcinogenesis. In this model, there are two types of endometrial carcinoma that have been designated Type I and Type II (Table 1). Factors associated with unopposed estrogenic stimulation, such as obesity and exogenous hormone use, as well as the presence of endometrial hyperplasia, are related to the development of the most common form of endometrial carcinoma; that is, the endometrioid subtype that represents the Type I carcinomas (1). More recent studies have confirmed this association by demonstrating elevated serum estrogen levels in patients with endometrioid carcinoma. It also has been recognized that some forms of endometrial carcinoma appear to be unrelated to hormonal factors and hyperplasia (2). Serous carcinoma is the prototypic endometrial carcinoma that is not related to estrogenic stimulation and represents the Type II carcinoma. Most of the other subtypes of endometrial carcinoma can be classified as variants of either Type I or II on the basis of clinicopathologic and immunohistochemical features. Thus, other low-grade carcinomas, which are associated with endometrial hyperplasia and estrogenic stimulation, such as mucinous or low-grade endometrioid with squamous differentiation, are Type I carcinomas. In contrast,
most clear cell carcinomas share features with serous carcinoma and are thus considered Type II carcinoma. Some carcinomas display mixed patterns; for example, endometrioid and serous carcinoma. It is likely that, in these neoplasms, endometrioid carcinomas develop a serous component as a progressive step in tumorigenesis. In a similar fashion, it is conceivable that carcinosarcomas (malignant mesodermal mixed tumors, MMMTs) evolve from endometrioid or serous carcinomas. Thus, the proposed dualistic origin of endometrial carcinoma is a model that best explains the development of these tumors with diverse morphological appearances. As in all biological systems, there may be rare or unusual examples that cannot be accounted for using this model.
