Comparison between magnetic resonance imaging and rigid rectoscopy in the preoperative identification of intra- and extraperitoneal rectal cancer.

Abstract

Accurate preoperative discrimination between extra- and intraperitoneal rectal cancer has important treatment implications. Our main objective was to compare the diagnostic performance of MRI with rigid rectoscopy (RRS) in assessing the location of rectal cancers above or below the peritoneal reflection (PR), using the findings obtained during abdominal surgery for treatment of the cancer as the reference standard. We also compared the accuracy of MRI and RRS in assessing the level of the lower border of the tumour from the anal verge. Patients with rectal carcinoma awaiting surgery underwent MRI and RRS. The MRI images were reviewed by two abdominal radiologists who determined the location of the inferior border of the tumour in relation to the PR. Receiver-operating characteristics (ROC) curve analysis was performed to determine the diagnostic performance of RRS at different cut-off values. The sensitivity and specificity were 98.15% and 100%, respectively, for MRI, and 100% and 76.92%, respectively, for RRS at a cut-off value of < 10 cm. The mean level of the lower border of the tumour from the anal verge was 68 ± 44.3 mm on RRS and 73.5 ± 42.4 mm on MRI (P = 0.25), with a trend towards overestimation with MRI. RRS is still the main means of assessing the level of a rectal tumour from the anal verge, but MRI has value in determining the level of the tumour in relation to the PR, which cannot be seen on endoscopy.