Abstract
OBJECTIVES: There are no guidelines regarding the best practice for when Barrett's esophagus (BE) is suspected but not confirmed by histology. The aim of this study was to examine the value of endoscopic follow-up for individuals with endoscopic only BE at index endoscopy. METHODS: We performed a longitudinal study of patients diagnosed with suspected columnar lined esophagus (CLE) (suspected BE in the absence of histological confirmation of specialized intestinal metaplasia (IM)). We examined three possible outcomes (definite BE defined as CLE plus IM in targeted biopsies, suspected CLE, or no suspected CLE) on repeat endoscopy within 2 years after the index endoscopy and their predictors (clinical, demographic as well as endoscopists' identity). RESULTS: A total of 107 of 1,844 patients had suspected CLE (101 were <3 cm), and 80 underwent a repeat endoscopy within 2 years. Approximately, 71% (95% confidence interval (CI) 61.1-80.9%) had suspected CLE confirmed at repeat endoscopy and only 29% (95% CI 19.1-38.9%) had IM. The length of CLE on the index esophagogastroduodenoscopies was slightly longer among patients with definite BE on repeat endoscopy than those with suspected CLE and no IM or no CLE (1.6 cm (s.d. 1.3) vs. 1.5 cm (s.d. 1.4), and 1.4 cm (s.d. 1.2), respectively P>0.1). Patient demographics, body mass index, gastro-esophageal reflux disease symptoms, hiatal hernia, and endoscopists' identity were not significantly associated with the outcome on the repeat endoscopy. CONCLUSIONS: Most (71%) patients with suspected CLE remain negative for IM in the 2 years following the index endoscopy. The findings support withholding BE diagnosis for individuals with suspected CLE.