Histologic study of patterns of cervical involvement in FIGO stage II endometrial carcinoma

The pathology of cervical involvement in endometrial carcinoma has not been fully defined previously. We reviewed the histopathology of 66 hysterectom ies of women with stage II endometrial carcinoma. Cervical spread was categ orized as macroscopic or microscopic; stage IIA or IIB; direct spread, surf ace or lymphvascular metastasis; and size, number, and location. The cervic al tumor was macroscopically identified in 15 (23%) women and microscopical ly identified in 51 (77%). Twenty-one patients (32%) were stage IIA acid 45 (68%) stage TIE. The method of spread was direct spread in 28 patients, su rface metastases in 27, lymphovascular in 3, both direct spread and surface metastases in 7 and both direct spread and lymphovascular in 1. The cervic al tumor had a mean horizontal dimension of 3 mm and a median of 2 mm. Ther e were multiple sites of cervical tumor in 31 (47%) patients and single in 35 (53%). The sites of spread, including cases with multiple sites, were th e endocervix in 60 women (90%), transformation zone in 24 (37%), and ectoce rvix in 3 (5%). Most patients had minimal microscopic cervical tumor. Small examples of direct spread may be an artifact of definition depending on th e histology of the isthmian-endocervical junction and many surface metastas es appear to follow dilatation and curettage, In 7 of 66, 11%, however, the cervical tumor was greater than 5 mm depth of invasion and/or the result o f lymphvascular metastasis. Survival studies are required to compare minima l stage II endometrial carcinoma patients and those with larger and/or lymp hvascular derived cervical tumor. Patients with minimal stage II and stage I patients should also be compared.