Magnetic resonance imaging (MRI) has been used as one of the diagnostic tools in the preoperative evaluation of rectal cancer. However, the usefulness of endorectal coil (ERC) compared with phased array coil (PA) MRI is still unknown.

Nineteen patients with rectal (17 patients) and anal (2 patients) tumors undergoing both ERC-MRI and PA-MRI preoperatively were included in the study. Seventeen patients had advanced stage tumors, and the remaining 2 had early stage tumors. The diagnostic accuracy of depth tumor invasion and lymph node metastasis were compared with reference to the histopathologic findings as the gold standard.

ERC-MRI was evaluable in only 10 (52.6%) of 19 patients, because of difficulties in ERC placement, whereas PA-MRI could be obtained in all patients (100%). In 10 patients examined by both modalities, the diagnostic accuracy of depth of tumor invasion was 80% by ERC-MRI and 80% by PA-MRI. In lymph node staging, ERC yielded sensitivity of 100%, specificity of 62.5%, and overall accuracy of 70.0%. The corresponding values for PA-MRI were 50%, 100% and 90.0%, respectively. These figures were not significantly different between the two modalities.

ERC-MRI and PA-MRI showed similar diagnostic accuracy. ERC-MRI may be abandoned in the preoperative staging of patients with locally advanced anorectal tumors because of its limited clinical utility.