Purpose: The purposes of this study were to assess whether multislice CT can identify tumors having a free or involved circumferential margin, to investigate the additional role of multislice CT as a "one-stop shopping" staging tool for staging nodal and distant metastases.
Methods: A total of 250 patients with adenocarcinoma of the rectum underwent multislice CT scans of the chest and abdomen before undergoing total mesorectal excision. The scans were scored by two teams. The main outcome was yes/no involvement of the mesorectal fascia. Histology was taken as the standard for determining the involvement.
Results: The overall sensitivity for predicting an involved mesorectal fascia was 74.2 percent and the overall specificity was 93.9 percent. The overall sensitivity for low tumors was 65.6 percent and the overall specificity was 81.5 percent. The overall sensitivity for mid-/high rectal tumors was 76.1 percent and the overall specificity was 96.3 percent. The interobserver agreement was substantial (kappa 0.695). The overall sensitivity for the prediction of liver metastases was 64.3 percent and the overall specificity was 94.4 percent with kappa 0.82. The accuracy in predicting lymph node metastases was low.
Conclusions: Multislice CT can be used for the assessment of mesorectal fascia involvement in primary rectal cancer, especially those located in the middle rectum and the high rectum; however, in the prediction of an involved margin of tumors located in the distal rectum, the accuracy of multislice CT falls short.