ESTROGENS, ENDOMETRIAL HYPERPLASIA, AND ENDOMETRIAL CARCINOMA

‘The subject of this paper can be stated very simply: are estrogens an iinportant and common etiologica! factor in the development. of endometrial carcinoma in the human? Such a relationship is certainly inferred or suggested in numerous publications. The following is a quotation from Willis’ book The P a t l d o g y OJ Tzurzor.7 (1948) : “The obvious relationship of hyperplasia and neoplasia i n certain cases, however, must not be allowed to mislead us. There are legion (sic) hyperplasias which never end in tumor formation and there itre many tumors which develop without any recognizable previous hyperplasia. (‘learly, then, t he causation of a tumor involves more than the mere persistence of hyperplasia evoking stimuli. The specific nature of the stimulus is all-important ; for example, on the experimental side both c x t m n tetrachloride ;tnd xminoazotoluene cause liver damage followed by regenerative hyperplasia but tumor formation is rarely seen following the former and very frequently following the latter.” Prior to pursuing this subject further i t seems prudent to detine our own present position on this subject. We believe t h a t some (and probably the minority) of endometrial carcinomas are preceded by and possibly induced i t i or developed from areas of endometrial hyperplasia. The exact initiating factor in such instances is unknown t u us, but certainly therc is evidence that endometrial carcinoma has developed i n patients with past evidences of hyperplasia (Hertig arid Sommers, 1949; Specrt, 1952). ntial to define what is meant by liyperplasia. I’athologically, self-explanatory: an increase i n size of an organ clue to reduplication of cellular elements. To our way of thinking this presupposes or necessitates the presence of at least ;t few mitotic figures. I n addition, i n the endometrium there is usually variation in size and shape of the glands and pseudostratification of the nuclei. This latter phenomenon should not be confused with early atrophic changes in which (as in the inyometrium) there is a loss of cytoplasm and apparent crowding together of nuclei which, in endometrial glands, can produce an appearance of pseudostratification. Artually, this marked diminution in cytoplasm and lack of mitotic figurcs clc:Lrlv distinguish this atrophic condition from hyperplasia. It is unfortunate that. the terms “Swiss cheese” and *‘cystic’’ hyperplasia of the endometrium were ever suggested. Hyperplasia may or may not include an appreciable distention of the glands, but cystic distention of glands is not of itself evidence that hyperplasia is present. ‘The increasing frequency of cystically distended glands with advancing postmenopausal years will be discussed later. We are more than a little unclear as to the exact status of the stroma will1 hyperplasia. Certainly with active hyperplasia (estrogen induced) stromal