Objective: To determine the sensitivity and specificity of CT-peritoneography (CT-P) in detecting peritoneal carcinomatosis in primary or recurrent ovarian cancer.
Methods: Twenty-five patients were submitted to a standard abdominal CT (CT) as well as a computed tomography after intraperitoneal infusion of contrast material (CT-P). Twenty patients had ovarian masses clinically suspected to be malignant. In five patients with ovarian cancer who underwent prior debulking surgery, recurrent disease was suspected.
Results: In 21/25 patients an ovarian malignancy was histologically confirmed. During surgery peritoneal spread was found in 13 patients; only in 5 cases CT correctly suggested peritoneal metastases (sensitivity 38%). However, in 10/13 women CT-P indicated peritoneal spread, increasing the overall sensitivity from 38% to 77%. Sensitivity varied substantially according to the different abdominal areas, and was lowest in the left subphrenic space (25%). The sensitivity of CT-P was 71% and 72% in the right paracolic gutter and the pelvis, respectively. Sensitivity of CT-P was not found to be size-dependent, but was mainly related to the morphology of the lesions, with 100% sensitivity in nodular lesions, and only 21% sensitivity for flat peritoneal lesions. CT-P did not improve detection of omental metastases, and was not advantageous when ascites was present. Previous surgery reduced overall specificity from 80% to 57%.
Conclusion: CT-P greatly improved the sensitivity of CT in the preoperative detection of peritoneal spread of ovarian malignancy. However, the technique failed to detect flat peritoneal metastases, and had a low specificity in patients with a history of prior abdominal surgery.