Endometrial cytology in early diagnosis of adenocarcinoma arising from adenomyosis uteri

Adenocarcinoma arising in adenomyosis uteri is a rare disease and diagnosed by endometrial curettage usually at an advanced stage, when the cancer has invaded to the surface endometrium as well as to the serosa of the uterus. Previous reports pointed out that most cases have a poor prognosis because preoperative diagnosis at an early stage is difficult.1-5 A 70-year-old woman, gravida 2, was not on hormone replacement therapy. She had atypical vaginal bleeding and underwent cytologic examination of the endometrium thrice and endometrial curettage twice. The level of tumor marker was within normal limits. Endometrial cytologic examinations revealed that there were many small, three-dimensional clusters, consisting of malignant cells (Figure 1). These cells were irregularly arranged, their polarities were altered, and cell boundaries were indistinct. Their nuclei were enlarged and nucleoli prominent. The background was dirty, with tumor diathesis. However, inflammatory cells were not conspicuous. These cytologic findings were characteristics of adenocarcinoma. In contrast, repeat endometrial curettage did not reveal malignant changes in the endometrium. The patient received a total abdominal hysterectomy, bilateral salpingo-oophorectomy and pelvic lymphadenectomy, and the peritoneal cytology was normal. The cancer, about 1 cm in diameter, was located close to the adenomyotic glands. The endometrioid adenocarcinoma lesions were infolded in some lumina of the atrophic glands (Figure 2). The normal adenomyotic endometrial cell layer was adjacent to three or more cell layers of malignant cells with enlarged nuclei and prominent nucleoli (Figure 3). No endometrial alteration was recognized in the covering surface endometrium on any serial sections of the hysterectomy specimen. The FIGO stage of endometrial cancer was IC (pT1CN0M0). As it is known that adenomyotic glands in the myometrium are in contact with surface endometrial glands,6 the malignant cells collected in the endometrial cavity might be derived from the surface holes, connecting carcinoma foci of adenomyotic glands in the myometrium. The cytologic features of our case were very small clusters, implying that these cancer cell clusters could slip through small holes from carcinoma foci in the adenomyotic glands. Repeat endometrial cytology may be useful for the detection of early cancer lesions arising from adenomyosis.