Purpose: Preoperative combined chemoradiotherapy is currently the main neoadjuvant therapy used to treat locally advanced middle and low rectal adenocarcinoma. A restaging work-up with magnetic resonance imaging was hoped to provide information about the effects related to combined chemoradiotherapy. The goal was to evaluate the correlation between pathologically verified tumor stages and clinical stages predicted by magnetic resonance imaging after combined chemoradiotherapy.
Methods: Between August 2000 and June 2003, 50 patients with biopsy-proven middle and lower rectal adenocarcinoma, with initial stage T3-T4 or N+, M0, were recruited in this series. Pelvic magnetic resonance imaging was used to stage the tumor before and after combined chemoradiotherapy. A protocol of the standard external radiation dose and oral combined uracil and 5-fluorouracil plus leucovorin was used. The results of magnetic resonance imaging restaging after combined chemoradiotherapy were correlated with the pathologic staging.
Results: The overall predictive accuracy in T stage was 52 percent, whereas overstaging and understaging occurred in 38 percent and 10 percent of patients, respectively. Most of the inaccurate T staging was a result of the overstaging of superficial tumors (T0-T2). In N stage, accurate staging was noted in 68 percent of all patients, whereas 24 percent were overstaged and 8 percent were understaged.
Conclusion: In restaging irradiated tumors, magnetic resonance imaging had the accuracy of 52 percent in T stage and 68 percent in N stage. Poor agreement between post-combined chemoradiotherapy magnetic resonance imaging and pathologic staging was observed in both T (k = 0.017) and N (k = 0.031) stages. Most of the inaccuracy in both T and N stages was caused by overstaging. The problem with magnetic resonance imaging was believed to be that it could not completely differentiate fibrosis from viable residual tumors.