EUS and magnetic resonance imaging in the staging of rectal cancer: a prospective and comparative study

Background: Accurate locoregional staging is crucial in rectal cancer for deciding patient management because the administration of neoadjuvant therapy depends on it. EUS and magnetic resonance imaging (MRI) are used indistinctly in the pretherapeutic workup of rectal cancer. Objective: To prospectively compare the performance of EUS and MRI in the locoregional staging of rectal cancer in a large series of patients. Design: Prospective and comparative study. Setting: Tertiary center. Patients: Patients with histologically proven rectal cancer. Interventions: EUS and MRI were performed in all patients by a different operator unaware of the results of the other procedure. Main Outcome Measurements: Epidemiological, clinical, radiological, and echographic variables were evaluated. Pathological examination of the surgical specimen was used as the criterion standard. Results: Ninety patients (54 men and 36 women with a mean age of 68 +/- 12 years; range 33-87 years) constitute the final sample of this study. Most of the tumors were stages T2-T3 (85%; 95% CI, 77%-92%). Twenty of them (22%; 95% CI, 14%-32%) were stenotic and 24 (27%; 95% CI, 18%-37%) had polypoid morphology. The accuracy of T staging was very similar for EUS and MRI for stage T2 (76%; 95% CI, 65%-84% and 77%; 95% CI, 67%-85%, respectively; P = not significant) and stage T3 (76%; 95% CI, 65%-84% and 83%, 95% CI, 73%-90%, respectively; P = not significant). MRI was not able to visualize any T1 tumor, whereas EUS understaged all T4 tumors. The univariate analysis showed that the polypoid morphology of the tumor inversely correlated with T staging on MRI. The accuracy of MRI for N staging was higher than that of EUS, although the difference did not reach statistical significance (79%; 95% CI, 65%-88% and 65%; 95% CI, 51%-78%, respectively). When performing the univariate analysis to assess the reasons for this difference, the presence of a stenotic tumor was the only parameter significantly related to a poorer performance of EUS in N staging. Limitations: The small number of early and locally advanced lesions. Conclusions: EUS and MRI have similar accuracy in the T and N staging in rectal cancer. The presence of stenosis and polypoid morphology is inversely associated with accuracy for either EUS or MRI. (Gastrointest Endosc 2011;74:347-54.)