| Abstract: | The depth of myometrial invasion by endometrial carcinoma was evaluated using real-time sonography (US) in 20 patients with histologically proved adenocarcinoma of the endometrium. In 14 of 20 (70%) cases, US-based estimation of the depth of myometrial invasion was within 10% of the actual measurement in the gross specimen. The US-based estimation of tumor invasion was low in seven patients, high in four patients, and agreed with pathologic findings (+/- 5%) in nine patients. In four patients with polypoid intraluminal extension of tumor, a deeply invasive tumor was suspected on US but was not found on pathologic examination. In 12 superficially invasive tumors, the continuity of the demarcating subendometrial halo was intact in nine and incomplete in three. In six patients with deeply invasive tumors, this zone was partially disrupted in four, totally disrupted in one, and intact in one. Errors of estimation of the depth of myometrial invasion on US most frequently occurred when a tumor had a significant intraluminal polypoid extension. Demonstration of a subendometrial halo usually indicated superficial invasion, whereas the absence of a halo was frequently associated with deep invasion. |
