eP240 - Concordance Between MRI and Pathology in Patients with Rectal Cancer

Purpose/Background: As the care of rectal cancer has progressed, the importance of pre-operative staging has been magnified. Magnetic resonance imaging (MRI) is now standard of care for pre-operative locoregional staging of rectal cancer. While the explosion of use in MRI for this staging, the accuracy in community settings has not been as well established. Further, in the era of watchful waiting (WW), the accuracy of MRI in predicting complete clinical response in community settings has also not been fully elucidated.Methods/Interventions: This is a single center retrospective review of a prospectively maintained database at a National Accreditation Program for Rectal Cancer accredited community hospital including all patients who underwent surgery from January 2021 to June 2023. At multidisciplinary tumor board, all patients who had undergone preoperative treatment and had a post-treatment MRI had a formal restaging performed by our expert abdominal radiologists. The database was then queried to assess for radiologic to pathologic concordance in patients who had and had not undergone preoperative treatment. In cases of discordance, the appropriateness of treatment based on National Comprehensive Cancer Network standards were assessed based on the final pathology. It was additionally assessed if the MRI had been followed without factoring in additional patient and tumor related factors if appropriate care would have been performed.Results/Outcomes: Overall, 90 patients met inclusion criteria of which 37 went directly to surgery and 53 underwent preoperative treatment. The MRI to pathologic concordance was 54% in patients who did not undergo preoperative treatment compared with only 36% in patients who underwent preoperative treatment (p-value=0.09). Amongst patients who went directly to surgery, the MRI was unable to determine between T1/2 in 32% of cases. Overall, 78% of patients who went straight to surgery had appropriate treatment (28/36 patients). If the MRI was blindly followed, this rate would have dropped to 61%. The most common inappropriate treatment was patients staged as T2N0 who went straight to surgery who ended up with node positivity or a T3 tumor that could have benefited from neoadjuvant therapy (⅝ patients who had inappropriate treatment). Amongst patients who underwent preoperative treatment, appropriate care was performed 98% of the time (52/53 patients). Amongst the 15 patients who care could have been different, 10 of these patients had a complete pathologic response when the MRI suggested residual tumor and 5 of these patients had residual tumor when the MRI suggested a complete pathologic response.Conclusion/

Discussion: MRI is currently the standard of care to determine locoregional extent of rectal cancer. Our study shows that in patients who underwent preoperative treatment and in patients who went straight to surgery, MRI has significant limitations. It struggles to define T1 vs T2 tumors. In the era of WW, the MRI misreads the presence or absence of residual tumor 28% of the time. Despite these shortcomings, the correct decisions were made for patients most of the time. This highlights the importance of multidisciplinary approach and rectal cancer center accreditation wherein the whole picture including endoscopic findings, radiographic findings, and digital rectal exam are considered.