A Comparison of Various Modalities in Staging Rectal Cancer: DRE, EUS, and MRI

Abstract

Introduction: Accurate pre-treatment staging of rectal cancer is critical to determine the need for neoadjuvant chemoradiation therapy (CRT), which is offered for patients staged as T3 or TxN1 disease or worse. Along with digital rectal examination (DRE), both EUS and MRI are commonly used in clinical practice to evaluate tumor depth (T stage) and locoregional lymph node involvement. This retrospective review aims to evaluate the utility and clinical impact of EUS and MRI as adjuncts to DRE for the staging and management of newly diagnosed rectal cancer. Methods: Rectal cancer cases from January 2014 through June 2016 where MRI was performed as part of staging and/or where T-stage assessment by surgical DRE was documented prior to endoscopy were reviewed. All EUS procedures were performed by two experienced ultrasonographers. DRE was performed by one of two colorectal surgeons. Kappa statistics was used to measure agreement between different staging methods. This was done for both T-stage and treatment stage, defined as ≤T2, N0 versus ≥T3 or TxN1 as this dichotomy often dictates treatment with CRT. Results: T-stage by DRE was documented prior to EUS in 58 cases. T-stage agreement between DRE and EUS was moderate (K=0.46), with substantial agreement in treatment stage between the two (K=0.64). EUS differed with DRE on treatment stage in 3 cases (5%): 1 was upstaged from T2 to T3 and 2 were down staged from T3 to T2. T-stage by DRE was documented prior to MRI in 45 cases. Agreement between DRE and MRI in both T-stage and treatment stage was modest (K=0.09 and K=0.23 respectively). T-stage by EUS and MRI was documented in 67 cases. There was fair agreement for T-staging between EUS and MRI (K=0.29) with greater agreement in treatment stage between the two (K=0.51). EUS and MRI differed on treatment stage in 5 cases (7%): in 3 cases MRI provided a higher stage while in 2 cases EUS provided a higher stage. T-stage and treatment stage agreement across all three staging modalities was modest (0.24) and moderate (K=0.46) respectively (Table 1). Conclusion: DRE performed by an experienced colorectal surgeon correlates well with EUS in determining the stage of rectal cancer and the need for neoadjuvant therapy. The more modest correlation between DRE and MRI may be due to the smaller sample size. EUS and MRI generally agree on the treatment stage of rectal tumors. The use of both modalities rarely alters clinical decisions in the treatment of newly diagnosed rectal cancer and is likely redundant.Figure 1