Thin section dual-phase multidetector-row computed tomography detection of peritoneal metastases in gynecologic cancers

J Comput Assist Tomogr. 2003 May-Jun;27(3):333-40. doi: 10.1097/00004728-200305000-00006.

Abstract

Purpose: To determine the sensitivity, specificity, and accuracy of multidetector-row computed tomography (CT) using thin sections and multiplanar reconstruction for the detection of peritoneal implants in patients with ovarian cancer.

Method: Seventeen thin section dual-phase multidetector-row CT scans were performed on 17 women with potential peritoneal metastases from ovarian cancer, which scans were then followed by surgery. Axial and multiplanar images from the CT scans were reviewed by 2 observers, and the results were compared with the operative and clinical notes.

Results: Peritoneal metastases were detected by both readers in all 7 patients presenting with ovarian cancer and disease at laparotomy. Metastases were detected in 5/6 patients with recurrent tumor by observer 1 and in 4/6 patients by observer 2. Sensitivity, specificity, and accuracy for detecting peritoneal metastases at individual sites in the abdomen and pelvis were improved when both axial and multiplanar images were reviewed. Sensitivities were highest for the paracolic gutters and infracolic omentum (>70%). Approximately 50% of liver and diaphragmatic lesions were detected. Specificities approached 100% for all sites and accuracies were >80% for most sites of disease.

Conclusion: The sensitivity, specificity, and accuracy of CT for peritoneal metastases in patients is high using thin slices and axial and multiplanar review of the data.

MeSH terms

  • Female
  • Humans
  • Imaging, Three-Dimensional
  • Middle Aged
  • Observer Variation
  • Ovarian Neoplasms / pathology
  • Ovarian Neoplasms / surgery
  • Peritoneal Neoplasms / diagnostic imaging*
  • Peritoneal Neoplasms / secondary*
  • Peritoneal Neoplasms / surgery
  • Retrospective Studies
  • Sensitivity and Specificity
  • Tomography, X-Ray Computed* / methods