Rectal tumour staging: MR imaging using pelvic phased-array and endorectal coils vs endoscopic ultrasonography

The aim of this study was to compare MR imaging and endoscopic ultrasonography (EUS) for the local staging of rectal tumours. Forty-nine patients were examined on a 1.5-T MR unit using either a pelvic phased-array coil (n = 37) alone or combined with an endorectal coil (n = 12). Sagittal and axial sequences with T2-weighted fast spin-echo and axial T1-weighted spin-echo techniques were employed. The EUS technique was performed using a flexible endosonoscope. The results were compared with findings at histopathological sectioning of the specimen. The T-stage on MR correlated with histopathology in 32 of 49 patients and on EUS in 29 of 49 patients. The N-stage on MR correlated with histopathology in 22 of 49 patients and on EUS in 26 of 49 patients. Tumour penetration of the rectal wall was predicted by MR with 86 % sensitivity and 65 % specificity, and by EUS with 89 % sensitivity and 33 % specificity. Preoperative radiotherapy was administered to 40 of the patients after the examinations which may explain some of the overstaging by MR and EUS. Three patients with surgically and histopathologically confirmed invasion of neighbouring organs in the pelvis were detected preoperatively on MR but none on EUS. Tumour penetration of the rectal wall and local lymph node metastases cannot accurately be predicted with MR or EUS. Magnetic resonance, however, seems to be more useful for preoperative identification of clinically occult advanced disease. Received: 18 February 1999; Revised: 17 September 1999; Accepted: 20 September 1999