CT Staging to Triage Selection of Patients With Poor-Prognosis Rectal Cancer for Neoadjuvant Treatment.

Abstract

OBJECTIVE. The purpose of this study was to evaluate CT as a screening tool for determining high risk of local recurrence of rectal tumors in a scenario of limited MRI availability. MATERIALS AND METHODS. Data were retrospectively analyzed for 180 consecutively registered patients with rectal adenocarcinoma and no previous treatment who underwent baseline CT and MRI staging within 30 days of each other. Two radiologists independently reviewed CT and MR images. CT scans were interpreted in multiplanar reformation. High risk of local recurrence was based on the MRI reference standard: T3cd (more than 5 mm of mesorectal fat infiltration) or T4 disease, N2 nodal status, mesorectal fascia involvement, extramural venous invasion, or positive pelvic sidewall nodes. The performance of CT for determination of high risk of local tumor recurrence was evaluated. RESULTS. Among the 180 patients 128 (71%) met MRI criteria for high risk of local recurrence. CT sensitivity was 84.4% (108/128) and specificity was 78.8% (41/52). The positive predictive value (PPV) of any high-risk CT feature was 90.7% (108/119). When T status was considered, the sensitivity of CT was 75.2% (79/105), specificity was 90.7% (68/75), and PPV was 91.9% (79/86). When tumors within 5.0 cm of the anal verge were excluded, sensitivity was 89.5% (51/57), specificity was 85.7% (24/28), and PPV was 92.7% (51/55). Using CT for disease staging could reduce MRI use by 66%. CONCLUSION. Tumors at high risk of local recurrence can be identified with CT without baseline MRI. Use of CT rather than MRI could markedly reduce costs of baseline staging and shorten time to initiation of neoadjuvant treatment.