Background: The use of surface coils for magnetic resonance imaging (MRI) allows enhanced image definition and so potentially more accurate staging of colorectal cancer. Endorectal coil imaging is invasive, operator dependent and impossible in a high proportion of patients due to rectal stricture. The phased-array pelvic coil, however, is non-invasive and applicable to all rectal tumours.
Methods: A pelvic phased-array coil was used for preoperative MRI staging of 38 primary rectal carcinomas. Results were expressed according to the Dukes and tumour nodes metastasis (TNM) classifications. After resection of the tumour, the stage predicted on MRI was compared with the pathological classification.
Results: The overall accuracy of preoperative staging with the pelvic phased-array coil was 55 per cent for both Dukes class and T stage. Assessment of nodal involvement gave an overall accuracy for MRI of 76 per cent with a sensitivity of 57 per cent and specificity of 88 per cent.
Conclusion: Use of a pelvic phased-array coil did not improve the staging accuracy of MRI to a clinically useful level.