Purpose: To determine the accuracy of spiral computed tomography (CT) in the depiction of peritoneal metastases by using surgical findings in patients with ovarian cancer as the standard of reference.
Materials and methods: Three independent readers reviewed the preoperative CT scans obtained in 64 patients who underwent primary surgery for ovarian cancer. Readers rated the likelihood of peritoneal metastases on a five-point scale and recorded the presence or absence of ascites, parietal peritoneal thickening or enhancement, and small-bowel wall thickening or distortion. Peritoneal metastases were identified as nodular, plaquelike, or infiltrative soft-tissue lesions in the peritoneal fat or on the peritoneal surface. Area under the receiver operating characteristic curve was calculated for each reader. Interreader agreement was evaluated with the kappa statistic. Descriptive statistical data were determined with dichotomized ratings (1-3 = absent; 4-5 = present).
Results: Areas under the receiver operating characteristic curves for the three readers were 0.95, 0.93, and 0.89. Paired kappa values ranged from 0.75 to 0.91. Reader sensitivity for metastases 1 cm or smaller in maximum diameter (25%-50%) was significantly (P <.05) lower than overall sensitivity (85%-93%). Ascites, parietal peritoneal thickening or enhancement, and small-bowel wall thickening or distortion demonstrated positive predictive values of 72%-93%, with kappa values of 0.12-0.80.
Conclusion: Spiral CT is accurate in the depiction of peritoneal metastases from ovarian cancer, although sensitivity is reduced in patients with tumor implants 1 cm or smaller. Ancillary signs of peritoneal malignancy are limited by low interobserver agreement.
Copyright RSNA, 2002