How reliable is restaging MRI after neoadjuvant therapy in rectal cancer?

Abstract

Despite the pivotal role of magnetic resonance imaging (MRI) in rectal cancer staging and evaluation, the reliability of restaging MRI after neoadjuvant therapy is still debatable. This study aimed to assess the accuracy of restaging MRI by comparing post-neoadjuvant MRI findings with those of the final pathology. This study was a retrospective review of the medical records of adult rectal cancer patients who had restaging MRI following neoadjuvant therapy and prior to rectal cancer resection in a NAPRC-certified rectal cancer centre between 2016 and 2021. The study compared findings of preoperative, post-neoadjuvant MRI with final pathology relative to T stage, N stage, tumour size, and circumferential resection margin (CRM) status. A total of 126 patients were included in the study. We found fair concordance (kappa -0.316) for T stage between restaging MRI and pathology report, and slight concordance for N stage and CRM status (kappa -0.11, kappa = 0.089, respectively). Concordance rates were lower for patients following total neoadjuvant treatment (TNT) or with a low rectal tumour. In total, 73% of patients with positive N pathology status had negative N status in the restaging MRI. Sensitivity and specificity regarding positive CRM in post-neoadjuvant treatment MRI were 45.45% and 70.4%, respectively. We found low concordance levels between restaging MRI and pathology regarding TN stage and CRM status. Concordance levels were even lower for patients after TNT regimen and with a low rectal tumour. In the era of TNT and watch-and-wait approach, we should not rely solely on restaging MRI to make post-neoadjuvant treatment decisions.