Ultrathin endoscopy versus high-resolution endoscopy for diagnosing superficial gastric neoplasia

Background: Ultrathin endoscopy (UTE) is in acceptable and cost-effective alternative to EGD with the patient under sedation, although the diagnostic accuracy of UTE is not well established. Objective: To compare the diagnostic accuracy of UTE and high-resolution endoscopy (HRE) for superficial gastric neoplasia. Design: Prospective comparative study. Setting: Academic center. Patients and Interventions: Patients with or without superficial gastric neoplasia underwent peroral UTE and HRE, back-to-back in a random order while under standard sedation. The procedures were performed by 2 endoscopists who were blinded to the clinical information. Main Outcome Measurements: The rate of missed lesions and misdiagnosis, sensitivity, and specificity for the diagnosis of gastric neoplasia when using pathology as the reference standard. Results: In total, 126 lesions (41 Superficial gastric neoplasias, 85 nonneoplastic lesions) were recorded in 57 enrolled patients. For the diagnosis of gastric neoplasia, the sensitivity of UTE (58.5%) was significantly (P =.021) lower than that of HRE (78%), and the specificity of UTE (91.8%) was significantly (P =.014) lower than that of HRE (100%). The rate of missed lesions and misdiagnosis of gastric neoplasias when using UTE (41.5%) was significantly (P >.001) higher than that of HRE (22.0%). The corresponding rate of neoplasias at the proximal portion (fornix and corpus) when using UTE (29%) was significantly (P =.002) higher than that of FIRE (7.2%), although the rates of neoplasias at the distal portion (angulus and antrum) were comparable for UTE and HRE. Limitation: Small sample numbers in an enriched Population Conclusions: The diagnostic accuracy of UTE is significantly lower than that of HRE for superficial gastric neoplasia, and this difference is particularly striking for neoplasias in the proximal stomach. For UTE to be used as an alternative modality, improvements in optical quality and the incorporation of additional procedures, including close-range observations and chromoendoscopy, are required to enhance visualization. (Gastrointest Endosc 2009;70:240-5.)