Evaluation of gadolinium-enhanced T1-weighted magnetic resonance imaging in the preoperative assessment of local staging in rectal cancer.

Eighty-eight patients with primary rectal cancer underwent preoperative MR imaging, followed by surgical resection. Two radiologists independently reviewed (i) T2-weighted MR images (T2WI); (ii) gadolinium-enhanced T1-weighted MR images (T1 + Gd); (iii) MR combined with T2WI and T1 + Gd for the prediction of tumour and nodal stage compared with histopathologic findings as the end point. Differences in the diagnostic performance of T2WI only, T1 + Gd image only and combined T2WI and T1 + Gd MR images were analyzed by comparing areas under receiver operating characteristic curves (Az) for each reader. Interobserver agreement was also calculated. Gadolinium-enhanced T1-weighted MRI did not increase the diagnostic yield for tumour and nodal staging, and may be omitted in the MR protocol for preoperative assessment of primary rectal cancer. The aim of this study was to determine whether gadolinium-enhanced T1-weighted magnetic resonance (MR) sequence is beneficial in the preoperative assessment of tumour and nodal staging in patients with primary rectal cancer. There was no significant difference in the Az values of T2WI only, T1 + Gd image only and combined T2WI and T1 + Gd images for the prediction of tumour staging (Az of T2WI, T1 + Gd and combined MR images for reader 1, 0.80, 0.76 and 0.85; reader 2, 0.83, 0.82 and 0.87) and nodal staging (Az for reader 1, 0.73, 0.73 and 0.81; reader 2, 0.79, 0.80 and 0.83). Interobserver agreement for the prediction of tumour staging was moderate to substantial, while only fair agreement was noted for the prediction of nodal staging.