Optimal sampling of grossly normal omentum in staging of gynecologic malignancies

The pathologic detection of microscopic omental metastases is important for the staging and treatment of ovarian and endometrial cancer. The question of how to sample grossly negative omentectomy specimens has not been adequately answered. We reviewed our institutional experience by retrieving a series of gynecologic cancer cases from 1998 to 2013 in which the omentum was grossly negative, but microscopically positive. There were 21 patients with ovarian carcinoma, 7 with ovarian borderline tumors, and 16 with endometrial carcinoma (44 patients in total). Cases in which the omentum was grossly abnormal or suspicious were excluded. A mean of 5.2 blocks were submitted per case (range, 4-15), of which a mean of 2.7 were positive for metastatic disease (range, 1-5). The distribution of cases by percentage of blocks positive was bimodal: some cases showed only 1 or 2 foci of disease in the entire sampled omentum, whereas in other cases nearly every block was positive. Only 3 cases had been sampled with >5 blocks. We used a series of simulated cases, bootstrapped on the retrospective series, to determine the additional sensitivity conferred by submitting >5 blocks. This model indicated that 5 blocks will, in fact, be insufficient to capture microscopic metastases in some cases. Examination of 5 blocks has an estimated sensitivity of 82%, whereas submission of 10 blocks raises the sensitivity to 95%. These results suggest that submission of 10 blocks should be considered for grossly negative omentectomy specimens when other staging is negative.