Purpose: Screening for Barrett's Esophagus (BE) has been hampered by the lack of an acceptable and widely applicable tool. The acceptability and findings of screening using less invasive techniques such as capsule endoscopy (CE) and unsedated transnasal endoscopy (uTNE) in the general population compared to sedated endoscopy (sEGD) for BE screening remain unknown. We AIMED to compare the acceptability and findings using sedated and unsedated techniques for esophageal screening in a population based cohort.
Methods: Subjects with BE or an EGD within 5 years were excluded and an age (>50), sex, and reflux-symptoms stratified random sample was created from a historical cohort (n=2152) with GER/GI symptom data, in Olmsted County, MN. The subjects were then RANDOMIZED to 1 of 3 groups: sEGD, uTNE, and CE. An inverse sampling design was used (number that needed to be contacted in order for 20 subjects per group to participate). Recruitment consisted of an invitation letter followed by a telephone call and subsequent screening using either sEGD or uTNE (using the Fujinon 4.3 mm transnasal endoscope) or CE performed at the Mayo CTSA. CE was performed using the PillCam ESO2. All subjects underwent biopsies from the midesophagus and the gastroesophageal junction. BE was defined as the presence of >1 cm visible columnar segment in the esophagus with intestinal metaplasia on histology (all histology was reviewed by an expert GI pathologist).
Results: Overall 127 subjects were offered screening in order to have 60 subjects undergo screening: 20 participants in each group (overall participation rate: 47%). Age and gender distributions of non-participants were similar to those of participants. Participation rates were numerically higher for unsedated techniques: CE (59%) and uTNE (50%) than sEGD (38%), but these rates were not significantly different. Findings on screening are shown in table 1. All three subjects with BE had short segment BE (segments 1-2 cm). Assuming the subjects contacted are a representative sample of the general population over age 50 and that those not completing a procedure would either not have the specific findings (lower bound) or have the same rates as those who completed a procedure (upper bound), the estimated proportions (%) in community residents over 50 years old would be 2.4% to 5.4% for BE, 7.1% to 16.1% for esophagitis and 31% to 74% for hiatal hernia.
Conclusion: Screening for BE is feasible using unsedated techniques in the population: this may enable obtaining an accurate estimate of the true population prevalence in a cost effective manner and assessment of additional novel risk factors for BE. This could ultimately allow improvements in the outcomes of subjects with esophageal adenocarcinoma.
Disclosure: This research was supported by an industry grant from Fujinon.