Purpose:To evaluate the diagnostic accuracy of MRI for predicting the circumferential resection margin (CRM), mesorectal fascia (MRF) invasion, and the tumor response to neoadjuvant chemoradiotherapy (CRT) for locally advanced rectal cancer.
Materials and Methods:Sixty-five consecutive patients with locally advanced rectal cancer (ՆT3 or lymph nodepositive) who underwent neoadjuvant CRT and subsequent surgery were enrolled in this retrospective study. Two blinded radiologists independently reviewed both the preand post-CRT rectal MR images and measured the post-CRT CRM; they recorded their confidence level with respect to the MRF invasion and tumor response using a 5-point scale. The diagnostic accuracy of each reviewer was calculated using receiver operating characteristic curve (ROC) analysis.
Results:The measured CRM was not significantly different from the reference standard (mean difference, Ϫ1.4 mm; 95% limits of agreement, Ϫ8.3-5.4 mm; interclass correlation coefficient, 0.82). The diagnostic accuracy (A z ) for determining MRF invasion was 0.890 for reviewer 1 (95% confidence interval [CI], 0.788 -0.954) and 0.829 for reviewer 2 (95% CI, 0.715-0.911). The A z for predicting complete or near-complete regression was 0.791 for reviewer 1 (95% CI, 0.672-0.882) and 0.735 for reviewer 2 (95% CI, 0.611-0.837).Conclusion: MRI provides accurate information regarding the CRM of locally advanced rectal cancer after neoadjuvant CRT; it also shows relatively high accuracy for predicting MRF invasion and moderate accuracy for assessing tumor response. MRI HAS BEEN an excellent diagnostic tool for predicting the circumferential resection margin (CRM) as well as mesorectal fascia (MRF) invasion in primary rectal cancer (1-3) since Beets-Tan et al demonstrated the accuracy of high-resolution MRI for predicting the tumor-free resection margin in rectal cancer surgery (4). As over half of rectal cancers are assumed to be locally advanced cancers (5,6), neoadjuvant chemoradiotherapy (CRT) has been accepted as the standard of care for treating locally advanced rectal cancer (T3-T4 or nodepositive) since several investigators demonstrated that it decreased the chance of local recurrence (7,8). With the recent advances in neoadjuvant CRT regimens, the pathologic complete response (pCR) rate has increased up to 30% (9 -16) in addition to the tumor down-staging rate of 46%-60% (17-19). Therefore, it is important for investigators who advocate less extensive surgery to select in advance favorable responders (pCR or nearcomplete regression) for patients who would benefit from less extensive surgical resection by using a noninvasive imaging modality.Furthermore, from a surgical perspective, assessment of the MRF involvement and tumor-free CRM is crucial for surgical planning that determines whether total mesorectal excision (TME) or extended TME should be performed (20 -22). However, to the best of our knowledge, the majority of studies using MRI have focused on the accuracy of T-staging, which is surgically less relevant than the ...