Linear-array Endoscopic Ultrasound and Narrow-Banding Imaging Measure the Invasion Depth of Nonpedunculated Rectal Lesions with Comparable Accuracy based on a Randomized Controlled Trial.

ABSTRACT

BACKGROUND AND AIMS: Linear-array endoscopic ultrasound (EUS) and narrow-banding imaging (NBI) are both used to estimate the invasion depth of nonpedunculated rectal lesions (NPRLs). However, it is unclear which procedure is more accurate. This randomized controlled trial aimed to compare the diagnostic accuracy of linear EUS and NBI for estimating the invasion depth of NPRLs. METHODS: This study is a single-center, randomized, tandem trial. Eligible patients with NPRLs were randomly assigned to A group (Assessment with EUS followed by NBI) or B group (Assessment with NBI followed by EUS). The invasion depth of each lesion was independently measured by each procedure and categorized as mucosal to slight submucosal (M-SMs, invasion depth <1000 µm) or deep submucosal (SMd, invasion depth ≥1000 µm) invasion, with postoperative pathology as standard measurement. The primary outcome was diagnostic accuracy, and secondary outcomes included sensitivity, specificity, and procedure time. RESULTS: 86 patients with NPRLs were enrolled and 79 patients were finally analyzed, including 39 cases in the A group and 40 cases in the B group. Comparable diagnostic accuracies were observed between EUS and NBI (96.2% vs. 93.7%, P = 0.625). EUS identified lesions with deep submucosal invasion with 81.8% sensitivity, while that of NBI was 63.6% (P = 0.500). The specificity of both EUS and NBI was 98.5%. The procedure time was also similar between EUS and NBI (5.90 ± 3.44 vs. 6.4 ± 3.94 minutes, P = 0.450). Furthermore, the combined use of EUS and NBI did not improve diagnostic accuracy compared to EUS or NBI alone (94.9% vs 96.2% vs 93.7%, P = 0.333). CONCLUSIONS: Linear EUS and NBI measure the invasion depth of NPRLs with comparable accuracy. The combination of the two methods does not improve the diagnostic accuracy. Single NBI should be preferred, considering its simplicity and convenience in clinical practice.