Celiac disease (CD) has an estimated prevalence of 1% worldwide, including North America. Barrett’s esophagus (BE) is a complication of chronic gastroesophageal reflux disease (GERD), which predisposes to development of dysplasia, which can transition to esophageal adenocarcinoma (EAC). GERD is commonly found in patients with CD and there is evidence that reflux symptoms improve after starting the gluten-free diet. However, the prevalence of CD in patients with BE, and its impact on progression to EAC have not been studied.
To assess the prevalence of CD in patients with BE and its impact on development of dysplasia and EAC.
We performed a retrospective review of medical records of patients attending the Digestive Diseases Clinic and the specialized Barrett’s Clinic at McMaster University Medical Centre. We collected information related to demographics, esophageal and duodenal biopsies, as well as CD serological markers. CD diagnosis was based on the presence of Marsh III lesions in duodenal biopsies and/or positive tissue transglutaminase IgA antibodies (TTG). Categorical data was presented as n/N and %, continuous data as mean ±SD. Logistic regression was performed to estimate the association between BE and CD. For dysplasia development, Kaplan Meyer Breslow survival curve was used.
CD was found in 1.8% of patients attending the Digestive Diseases Clinic (n=4800). Two hundred and sixteen patients with BE from the Digestive Diseases Clinic and the Barret’s Clinic (age 63 ±11.2, 70.3% males; follow-up period 7.0 ±5.5 years) were included in the analysis. Patients with or without CD were similar in demographic factors, concomitant diseases or previous surgeries, use of PPI, presence of symptoms and duration of treatment. The prevalence of CD was increased 5-fold in patients with BE (OR=6.0; 95% CI 2.1–17.4; p=0.006). BE patients with CD developed dysplasia earlier than those without CD (Long Rank 0.004; mean 6.0; 95% CI 3.0–4.97[PB1]). The risk of EAC was higher in BE patients with CD compared to those without CD (2/11 vs 2/205; OR=22.5; 95% CI 2.8–178.8; p= 0.003). The mean time for development of EAC from the time of BE diagnosis [PB2] was 1.5 years in patients with CD and 6.5 years in patients without CD.
The prevalence of CD is higher in patients with BE compared with the population attending a tertiary gastroenterology clinic, and with the general population. Patients with BE and concomitant CD are at increased risk of accelerated development of dysplasia and EAC, suggesting that both conditions may have a synergic detrimental effect.