Distinction between endometrial and endocervical adenocarcinoma.

Abstract

To the Editor: The results of the studies published in this journal by Castrillon et al. (1) and by McCluggage et al. (2) provoke us to make the following comments. We had almost identical results and came to very similar conclusions after examining several hundred uterine adenocarcinomas immunohistochemically (3,4). We also agree with Castrillon et al. that most, if not all, endocervical adenocarcinomas of endometrioid type are CEA-positive and vimentin-negative, like those of mucinous (endocervical) type. We assume that endometrioid-type carcinomas of the endocervix are most likely of endocervical histogenetic origin with defective cytoplasmic differentiation (no mucin production), similar to the intestinal type of mucinous adenocarcinoma of the ovary. In contrast to McCluggage et al. we would not consider an endometrioid carcinoma of the endometrium CEA-positive if “positivity was confined to benign squamous elements” as illustrated in their Figure 3. On the other hand, the glands of mucinous adenocarcinomas of the endometrium that we studied were positive for CEA and negative for vimentin, thereby reacting as endocervical carcinomas do. Most had arisen in an atrophic endometrium without preceding hyperplasia but with regions of endocervical metaplasia, suggesting they had originated from these metaplastic endocervical cells. Twenty-two of our patients with these endocervical type endometrial carcinomas had received either tamoxifen or synthetic gestagens (3). In conclusion, we agree with Castrillon et al. and McCluggage et al. that vimentin positivity and CEA negativity of uterine adenocarcinomas points to an endometrial origin. On the other hand, vimentin-negative and CEA-positive mucinous adenocarcinomas may not only originate from the endocervix, but in rare cases also from the endometrium, most likely from foci of endocervical metaplasia. Gisela Dallenbach-Hellweg Dietmar Schmidt