Prognostic stratification by extramural depth of tumor invasion of primary rectal cancer based on the Radiological Society of North America proposal

The purpose of this study was to investigate the diagnostic performance of MRI in stratifying tumors stage III (T3) in patients with rectal cancer by measuring the extramural depth of tumor invasion based on the Radiologic Society of North America (RSNA) proposal and to validate its role as a prognostic indicator. From January 2006 to July 2009, 146 patients with surgically and pathologically confirmed T3 rectal adenocarcinoma who underwent preoperative MRI were enrolled. Two blinded radiologists measured the maximum extramural depth of tumor invasion on T2-weighted images. To evaluate the diagnostic performance of MRI for stratifying subgroups, the study population was assigned into three subgroups (T3a, T3b, and T3c) according to extramural depth of tumor invasion (< 5, 5-10, and > 10 mm).To validate the role of extramural depth of tumor invasion as a prognostic indicator, Cox regression analysis was used for estimation of independent risk factors for postoperative recurrence. Three-year recurrence-free survival was evaluated by the Kaplan-Meier method with a log-rank test. Histopathologic reports were used as the reference standard. The overall accuracy of MRI for stratifying subgroups was 71.2% (104/146) and 77.4% (113/146) for reviewers 1 and 2. Extramural depth of tumor invasion was an independent risk factor for 3-year recurrence-free survival (hazard ratio, 2.186; 95% CI, 1.336-3.577; p = 0.002). Kaplan-Meier curves revealed a significant difference in 3-year recurrence-free survival rates for each subgroup (86%, 69%, and 43% for T3a, T3b, and T3c; p < 0.03). MRI can be used for prognostic stratification according to extramural depth of tumor invasion based on the RSNA proposal for patients with T3 rectal cancer.