The Role of Helicobacter pylori in the Spectrum of Barrett's Carcinogenesis

The article recently published by Islami and Kamangar (1) raises the issue of the inverse association between Helicobacter pylori (Hp) gastric colonization and a (lower) risk of esophageal adenocarcinoma [i.e., Barrett's adenocarcinoma (BAc)]. The authors show that the declining rate of the bacterial infection among Western populations coincides with a rising incidence of BAc. The antisecretory activity of Hp may, at least partially, explain the pathophysiology of such a situation, but the real mechanism behind this inverse relationship is more complex and not entirely clear. Barrett's esophagus (BE), the precursor of BAc, has recently been associated with both a low prevalence of Hp colonization and multifocal atrophic gastritis (2, 3). The bacterium might therefore be listed among the factors “protecting” against the whole spectrum of histologic lesions occurring in Barrett's carcinogenesis. To further support this hypothesis, we ran a retrospective survey in our case series on the prevalence of Hp infection coexisting with both BE and BAc. Our test group included 210 consecutive BE patients (156 men; mean age, 59.9 ± 13.2; median, 62) and 23 cases of BAc (20 men; mean age, 61.2 ± 12.1; median, 63). A control population of 210 consecutive dyspeptic patients (156 men; mean age, 59.0 ± 14.5; median, 62) was also considered (2). Hp was histologically detected in 31 of 210 (14.8%) BE patients, 4 of 23 (17.4%) BAc patients, and 86 of 210 (41.0%) controls (P = 0.000, Fisher's exact test). The prevalence of severe atrophic gastritis (OLGA stages III-IV) was significantly lower among the BE patients than among controls (6 of 210 versus 14 of 210, respectively; P = 0.035, Fisher's exact test; ref. 4). Moreover, controls had a higher prevalence of noncardia gastric neoplasia (both invasive and noninvasive): 2 gastric cancers and 1 high-grade and 3 low-grade noninvasive cancers, as opposed to 1 gastric cancer and 1 low-grade noninvasive cancer among the BE patients. Expanding on the findings reported by Islami and Kamangar, our data add to the evidence that a peculiar phenotype of gastric mucosa does coexist with BE [with a low prevalence of both Hp infection and advanced atrophic gastritis (OLGA stages III-IV)]. We also found comparable rates of Hp colonization in BAc and BE patients, which further supports the divergent Hp role in the progression of Barrett's and gastric carcinogenesis (the higher the prevalence of BAc, the lower the prevalence of Hp, but a higher prevalence of gastric cancer coincides with a higher Hp prevalence). Future population-based studies should also focus on the concurrent role of other risk factors for BAc, such as gastroesophageal reflux, obesity, and smoking.