Selection of patients with rectal cancer for neoadjuvant therapy using pre-therapeutic MRI-Results from OCUM trial

Purpose: No consensus is available on the appropriate criteria for neoadjuvant chemoradiotherapy selection of patients with rectal cancer. The purpose was to evaluate the accuracy of MRI staging and determine the risk of over-and undertreatment by comparing MRI findings and histopathology. Method: In 609 patients of a multicenter study clinical T-and N categories, clinical stage and minimal distance between the tumor and mesorectal fascia (mrMRF) were determined using MRI and compared with the histopathological categories in resected specimen. Accuracy, sensitivity, specificity, positive predictive, and negative predictive value (NPV) were calculated. Overstaging was defined as the MRI category being higher than the histopathological category. mrMRF and circumferential resection margin (CRM) were judged as tumor free at a minimal distance > 1 mm. The chi-squared test or Fisher's exact test were used. P < 0.05 was considered significant. Results: The T category was correct in 63.5% (386/608) of patients; cT was overstaged in 22.9% (139/608) and understaged in 13.5% (82/608). MRI accuracy for lymph node involvement was 56.5% (344/609); 22.2% (28/ 126) of patients with clinical stage II and 28.1% (89/317) with clinical stage III disease were diagnosed by histopathology as stage I. The accuracy for tumor free CRM was 86.5% (527/609) and the NPV was 98.1% (514/ 524). In 1.7% (9/524) mrMRF was false negative. Conclusion: MRI prediction of the tumor-free margin is more reliable than the prediction of tumor stage. MRF status as determined MRI should therefore be prioritized for decision making.