Rectal cancer: comparison of accuracy of local-regional staging with two- and three-dimensional preoperative 3-T MR imaging.

PURPOSE: To compare the local-regional staging accuracy of the conventional two-dimensional (2D) T2-weighted imaging protocol and of the three-dimensional (3D) T2-weighted imaging protocol for preoperative magnetic resonance (MR) imaging in rectal cancer patients.
MATERIALS AND METHODS: This retrospective study was approved by the institutional review board, and a waiver of informed consent was obtained. A review was conducted of 109 preoperative 3-T MR images obtained with 2D and 3D T2-weighted imaging protocols in rectal cancer patients. Two radiologists independently assessed the radiologic findings for T and N category lesions, conspicuity of tumor margin, and image quality of 2D and 3D data. Interactive multiplanar reconstruction was performed for 3D data analysis. The linear weighted kappa values for T2-weighted imaging staging results (2D and 3D data) and histopathologic staging results were calculated and compared. Wilcoxon signed rank test was performed to compare tumoral conspicuity and overall image quality.
RESULTS: T category lesion staging accuracy values for 2D and 3D data, respectively, were 66.0% and 67.0% for reviewer 1 (P = .465) and 63.3% and 56.9% for reviewer 2 (P = .402). N category lesion staging accuracy values for 2D and 3D T2-weighted images, respectively, were 64.2% and 57.8% for reviewer 1 (P = .427) and 47.7% and 62.4% for reviewer 2 (P = .666). Tumor conspicuity was better for 2D T2-weighted imaging, but no significant difference in image quality was observed.
CONCLUSION: Preoperative MR imaging in rectal cancer patients for staging with conventional 2D and multiplanar reconstruction 3D T2-weighted imaging protocols showed no significant differences in accuracy of T and N category staging and overall image quality, as determined by degree of artifact. However, the 3D T2-weighted imaging protocol had limitations in regard to lesion conspicuity.