To the Editor:
Zaino et al. 2 report their findings in a clinicopathologic review of 61 cases of villoglandular carcinomas of the uterus. The authors found the behavior of villoglandular carcinoma to be similar to that of nonvilloglandular endometrioid carcinomas and concluded that villoglandular carcinomas should be considered to carry a prognosis similar to that of endometrioid carcinomas without villoglandular differentiation. The authors state that their findings do not support conclusions drawn by our study, in which we found villoglandular carcinomas to be a more virulent form of endometrioid carcinoma. 1 We would like to caution these investigators against attempting to compare their results with ours without taking case definition into consideration. Similar to the findings of Zaino et al., 2 we did not find any differences in the behavior of villoglandular carcinomas compared with nonvilloglandular endometrioid carcinomas when the topography of papillary differentiation was not considered. However, when the definition of villoglandular carcinoma was limited to tumors showing papillae in the myometrial component of the tumor and excluded tumors that showed papillae in the endometrial component of the tumor but a nonpapillary myometrial component, villoglandular carcinoma became a meaningful entity. Once papillae were found in the myometrial component of endometrioid carcinomas, tumors were frequently aggressive, regardless of what percentage of the tumor occurred in papillary form. Unlike our study, Zaino et al. did not consider the topography of papillary formation when defining a case as a mixed villoglandular carcinoma (personal communication, Robert J. Kurman, September 19, 1996). Because no assumptions can be made on the topography of villoglandular differentiation in the mixed villoglandular carcinomas, a meaningful comparison between the two studies would be limited to their cases that were pure villoglandular carcinomas with myometrial invasion and our cases of carcinoma showing villoglandular differentiation in the myometrium. Thus, only 19 of their 61 cases can be compared with our study. Even this comparison is skewed because it can be argued that “pure” villoglandular carcinomas would be more frequently of lower grade than carcinomas showing mixed patterns. At any rate, whereas 5 (20.8%) of the 24 patients in our study died of disease at a mean follow-up time of 54 months, only 2 (10.5%) of 19 patients with carcinoma which clearly had a myometrial villoglandular component in the study by Zaino et al. died of disease (mean follow-up time is unknown to us). Although the results of Zaino et al. do not appear to confirm our findings, their patient population appropriate for comparison is relatively small (n = 19) and because a comparison between “pure” and predominantly mixed villoglandular carcinomas is somewhat skewed, we do not believe this study provides significant evidence that villoglandular differentiation in the myometrium of an endometrioid carcinoma is not of prognostic or biologic significance. We encourage other investigators to examine this issue.
Robert A. Ambros M.D.
John H. Malfetano M.D.
1. Ambros R, Ballouk F, Malfetano J, Ross J. Significance of papillary (villoglandular) differentiation in endometrioid carcinoma of the uterus. Am J Surg Pathol 1994; 18:569–75.
2. Zaino RJ, Kurman RJ, Brunetto VL, Morrow CP, Bentley RC, Cappellari JO, Bitterman P. Villoglandular adenocarcinoma of the uterus: a clinicopathologic study of 61 cases. A Gynecologic Oncology Group Study. Am J Surg Pathol 1998; 22:1379–85.