Myocutaneous flap reconstruction of the perineum after radical abdominoperineal resection

Objective:  Accurate tumour re-assessment after down-sizing preoperative chemoradiotherapy (CRT) for locally advanced middle and lower rectal cancer provides valuable prognostic information prior to surgical resection. (1) We assessed the accuracy of post-CRT magnetic resonance imaging (MRI) in predicting circumferential resection margin status. (2) The accuracy of MRI in staging chemoradiation down-sized rectal tumours was also assessed.
Method:  Patients from a single tertiary referral centre who received preoperative CRT for biopsy-proven rectal cancer between January 2003 and February 2005 were identified from a database. All patients received long-course CRT (50 Gy over 5 weeks and 5-fluorouracil/leucovorin) prior to surgery. A comprehensive MRI reporting proforma was completed for each patient and results were compared directly with histological specimens.
Results:  Fifty patients were eligible for the study. The overall accuracy for predicting circumferential resection margin status was 76% (38/50, P  = 0.006, sensitivity = 96.9%, specificity = 73.8%, negative predictive value = 96.9%). For T staging, MRI was accurate in 52% (26/50) and for nodal staging MRI was correct in 60% (30/50). Venous invasion was predicted in 66% (33/50, P  = 0.04). A radiological rectal cancer regression grade was derived and this correlated with histological rectal cancer regression grade in 65% (32/49).
Conclusion:  This study demonstrates that post-CRT MRI is accurate in predicting the circumferential resection margin status, which provides useful prognostic information prior to surgery.