Helicobacter pylori infection is associated with reduced risk of Barrett's esophagus: a meta-analysis and systematic review

Abstract

Background: Helicobacter pylori (Hp) is a class I carcinogen in gastric carcinogenesis, but its role in Barrett's esophagus (BE) is unknown. Therefore, we aimed to explore the possible relationship.

Methods: We reviewed observational studies published in English until October 2019. Summary odds ratios (ORs) and 95% confidence intervals (CIs) were calculated for included studies.

Results: 46 studies from 1505 potential citations were eligible for inclusion. A significant inverse relationship with considerable heterogeneity was found between Hp (OR = 0.70; 95% CI, 0.51-0.96; P = 0.03) and BE, especially the CagA-positive Hp strain (OR = 0.28; 95% CI, 0.15-0.54; P = 0.0002). However, Hp infection prevalence was not significantly different between patients with BE and the gastroesophageal reflux disease (GERD) control (OR = 0.99; 95% CI, 0.82-1.19; P = 0.92). Hp was negatively correlated with long-segment BE (OR = 0.47; 95% CI, 0.25-0.90; P = 0.02) and associated with a reduced risk of dysplasia. However, Hp had no correlated with short-segment BE (OR = 1.11; 95% CI, 0.78-1.56; P = 0.57). In the present infected subgroup, Hp infection prevalence in BE was significantly lower than that in controls (OR = 0.69; 95% CI, 0.54-0.89; P = 0.005); however, this disappeared in the infection history subgroup (OR = 0.88; 95% CI, 0.43-1.78; P = 0.73).

Conclusions: Hp, especially the CagA-positive Hp strain, and BE are inversely related with considerable heterogeneity, which is likely mediated by a decrease in GERD prevalence, although this is not observed in the absence of current Hp infection.

Keywords: Barrett’s esophagus; Gastroesophageal reflux disease; Helicobacter pylori.

Fig. 1

Flow chart of the study selection process

Fig. 2

Forest plot of the random effect analysis of the 36 studies. The weights and heterogeneities of studies are indicated too. OR: Odds ratio, CI: 95% confidence interval

Fig. 3

Funnel plot of the random effect analysis of the 36 studies

Fig. 4

Forest plot of subgroup analysis according to definition of control group

Fig. 5

Forest plot of subgroup analysis according to status of Hp infection. 5.1: Hp positive with rapid urease test, urea breath test, histology or culture; 5.2: Hp positive with serological detection, treatment history, or infection history; 5.3: not sure to status of Hp infection

Fig. 6

Forest plot of the correlation between the CagA-positive Hp strain and BE. The weights and heterogeneitie s of studies are also indicated. OR: Odds ratio, CI: 95% confidence interval