Purpose: To evaluate the impact of chemoradiation on the reliability of MRI in assessing tumor involvement of the mesorectal fascia in patients with rectal cancer.
Materials and methods: Presurgical MRI was performed in 150 patients; among them 85 had received neoadjuvant long-course chemoradiation. A standardized imaging protocol (1.5 Tesla [T] system, image voxel size 0.6 × 0.4 × 3 mm(3) ), standardized surgery, and histopathological examination were applied for the entire patient population. Images were analyzed to identify potential tumor involvement of the mesorectal fascia (minimum tumor distance to fascia ≤1 mm) and compared with histopathology as the reference standard. Results of nonirradiated and irradiated patients were compared to define the impact of chemoradiation on imaging reliability.
Results: In nonirradiated patients, MRI was reliable in predicting or excluding tumor involvement of the mesorectal fascia, positive predictive value 80%, negative predictive value 89%. The frequency of overestimating tumor involvement was significantly higher in irradiated patients (P = 0.005, positive predictive value 42%).
Conclusion: Discussions about MRI assessment of tumor involvement of the mesorectal fascia as a basis for recommending neoadjuvant chemoradiation should focus on investigations that excluded irradiated patients, because MRI is less reliable after chemoradiation and tends to overestimate mesorectal tumor involvement.
Copyright © 2012 Wiley Periodicals, Inc.