Performance of Narrow Band Imaging and Magnification Endoscopy in the Prediction of Therapeutic Response in Patients With Gastroesophageal Reflux Disease Ping-Huei Tseng, MD,* Chien-Chuan Chen, MD,* Han-Mo Chiu, MD, PhD,* Wei-Chih Liao, MD, MSc,* Ming-Shiang Wu, MD, PhD,* Jaw-Town Lin, MD, PhD,*w Yi-Chia Lee, MD, PhD,*z and Hsiu-Po Wang, MD*y Background: Imaging-enhanced endoscopy enhances the contrast of the mucosal surface and helps in the diagnosis of gastroesopha- geal reflux disease. However, whether the increased detection of subtle erosive foci corresponds to the effect of acid suppression remains elusive. Goals: We aim to evaluate the utility of narrow band imaging with and without magnification endoscopy in the prediction of therapeutic response in patients with reflux. Study: Endoscopic evaluation with conventional white light, narrow band imaging, and narrow band imaging with magnifica- tion was performed sequentially in consecutive patients with reflux. All patients received proton pump inhibitor for 14 days. Their therapeutic responses were correlated with the baseline endoscopic findings, including mucosal breaks under standard endoscopy, mucosal brownish changes under narrow band imaging, and increased and/or dilated intrapapillary capillary loops or micro- erosions under narrow band imaging with magnification. Results: Of a total of 82 patients, 22 (26.8%) patients were diagnosed with erosive disease under standard endoscopy. Among the remaining 60 (73.2%) patients, 14 (23.3%) and 30 (50%) were considered erosive under narrow band imaging and narrow band imaging with magnification, respectively. Sixty-five (79.3%) pa- tients showed a positive therapeutic response. The sensitivity, specificity, positive predictive value, negative predictive value, and accuracy in predicting therapeutic response were 33.8%, 100%, 100%, 28.3%, and 47.6%, respectively, for standard endoscopy; 52.3%, 88.2%, 94.4%, 32.6%, and 59.8%, respectively, for narrow band imaging; and 70.8%, 64.7%, 88.4%, 36.6%, and 69.5%, respectively, for narrow band imaging with magnification. Conclusions: Narrow band imaging with and without magnification endoscopy substantially improve our ability to predict therapeutic response in patients with gastroesophageal reflux. Key Words: narrow band imaging, magnification endoscopy, gastroesophageal reflux disease, proton pump inhibitor (J Clin Gastroenterol 2011;45:501–506) G astroesophageal reflux disease (GERD) is associated with a broad spectrum of manifestations, which require individual tailored treatment for specific symptom. 1–3 On the basis of endoscopic evaluation, patients with GERD can be stratified into erosive and nonerosive counterparts. 4–9 Although acid suppressants with proton pump inhibitors (PPIs) are very effective to control reflux symptoms and heal mucosal damage in erosive patients, therapeutic response in patients with nonerosive reflux disease (NERD) is notori- ously unpredictable. This indicates that substantial hetero- geneity exists and complicates the clinical management of these patients. 5,10 Endoscopic detection with narrow band imaging (NBI) has enabled endoscopists to identify subtle mucosal damage, 11 and a significant number of patients can be diagnosed as having erosive GERD among those who were previously considered to have NERD. 12–14 However, in these patients clinicians put more emphasis on the treatment efficacy of symptom relief than on the morphological improvement of their minimally inflammatory process. 15 Therefore, even though imaging-enhanced endoscopy may have provided us the ability to visualize mucosal details, it remains unclear whether this advantage will help us understand the therapeu- tic outcome of patients, or, in contrast, whether it may lead to overdiagnosis. Therefore, the primary aim of this study was to evaluate the clinical utility of NBI and NBI with magnifica- tion in the prediction of therapeutic response in a hetero- geneous group of patients with GERD, especially in patients who were diagnosed with NERD under standard endoscopy. MATERIALS AND METHODS Study Population Patients with typical GERD symptoms were enrolled and received a baseline assessment of symptom frequency and severity with the Reflux Disease Questionnaire, which has been validated in both Asian and Western popula- tions. 16–19 Typical GERD symptoms were defined as at least 3 episodes per week of heartburn and/or regurgitation Copyright r 2011 by Lippincott Williams & Wilkins Received for publication March 4, 2010; accepted June 17, 2010. From the *Department of Internal Medicine, National Taiwan University Hospital and College of Medicine; zDivision of Biostatistics, Graduate Institute of Epidemiology, College of Public Health, National Taiwan University, Taipei; wDepartment of Internal Medicine, E-DA Hospital and I-Shou University, Kaoh- siung County; and yDepartment of Internal Medicine, National Taiwan University Hospital, Yun-Lin Branch, Yun-Lin County, Taiwan. Supported by research grants from the National Science Council (98- 2314-B-002-089-MY3), National Taiwan University Hospital (NTUH-N05), and the Taipei Institute of Pathology. The short- term treatment with a proton pump inhibitor (rabeprazole) was supported by the Eisai Co., Ltd. The authors report no conflict of interest. Reprints: Yi-Chia Lee, MD, PhD, Department of Internal Medicine, National Taiwan University Hospital, 7, Chung-Shan South Road, Taipei 100, Taiwan (e-mail: yichialee@ntu.edu.tw). ORIGINAL ARTICLE J Clin Gastroenterol  Volume 45, Number 6, July 2011 www.jcge.com | 501