Accuracy of CT in detection of persistent or recurrent ovarian carcinoma: correlation with second-look laparotomy

Computed tomographic (CT) studies in 39 patients with advanced ovarian carcinoma were retrospectively evaluated to assess their accuracy in detection of persistent or recurrent disease as seen at second-look laparotomy (SLL). Twenty patients were studied before June 1983 (group 1), and 19 patients were studied afterward (group 2). There were 16 true-positive, ten true-negative, two false-positive, and 11 false-negative examinations. Five false-negative studies resulted from microscopic disease found at SLL. In group 1, there were eight false-negative studies. In five, macroscopic disease was not recognized. In group 2, there were three false-negative studies; in one, macroscopic disease was not recognized. Statistical analysis showed an observable improvement in the accuracy in group 2. The differences included use of faster scanners, routine use of thin sections for the pelvis, and air-contrast colonic opacification in group 2. These results suggest that carefully performed scanning on state-of-the-art equipment may have a higher accuracy in documenting persistent or recurrent macroscopic tumor.