Accuracy of pelvic MRI in measuring tumor height in rectal cancer patients with or without preoperative chemoradiotherapy

Abstract

INTRODUCTION: In measuring tumor height for rectal cancer, rigid sigmoidoscopy (RS) is a standard modality, and the accuracy of magnetic resonance imaging(MRI) in patients with/without preoperative chemoradiotherapy (CRT) has not been fully investigated. The aim of this study was to investigate the accuracy of MRI for measuring tumor height.

MATERIALS AND METHODS: Among rectal cancer patients seen between July 2006 and May 2012, the initial group (RS and MRI available at initial diagnosis) and the post-CRT group (RS and MRI available after the completion of preoperative CRT) were selected. Intra-class correlation coefficient (ICC) comparison tests were performed between RS and MRI results for each group.

RESULTS: Ninety-nine and 29 patients were allocated into the initial group and the post-CRT group, respectively. The tumor heights measured by RS and MRI demonstrated a positive relationship in the scatter plot (linear regression; R2 = 0.898; p < 0.001 in the initial group, R2 = 0.696; p < 0.001 in the post-CRT group). With respect to difference of absolute value (DAV) between RS and MRI, the overall mean and standard deviation of DAV were 10.9 ± 10 mm in the initial group and 8 ± 6 mm in the post-CRT group. ICC comparison analysis revealed that inter-rater agreement of RS and MRI in the initial group was significantly better than that of the post-CRT group [ICC (95% CI) 0.946 (0.919-0.963) vs. 0.823 (0.621-0.917); p = 0.004)].

CONCLUSIONS: MRI can be used as a viable option to measure tumor height in rectal cancer even in patients who have undergone preoperative CRT.