Barrett's Esophagus in an Area with an Exceptionally Low Prevalence of Helicobacter pylori Infection

Objective. This study was undertaken to gain an insight into the relationship between Helicobacter pylori (H. pylori) infection, Barrett's esophagus and reflux esophagitis in an area of exceptionally low prevalence of H. pylori infection. Methods. A total of 1895 consecutive upper endoscopies performed between January 2005 and July 2007 were reviewed. 120 cases of columnar-lined esophagus and endoscopic esophagitis were evaluated. H. pylori infection was determined using the urease test and/or histology. Results. The rate of endoscopic esophagitis was 5.49% (80 Malays, 24 non-Malays) while histological reflux esophagitis was found in 3.75% (56 Malays, 15 non-Malays). Barrett's esophagus was present in 0.79% (11 Malays, 4 non-Malays). H. pylori infection was present in 8/120 or 6.67% subjects. Conclusion. The low rate of Barrett's esophagus in this population does not support the hypothesis that the absence of H. pylori infection is more than a minor risk factor for Barrett's esophagus.


Introduction
Reflux esophagitis is thought to be uncommon among Asians, but recent studies have shown that erosive esophagitis is more common than what was originally thought [1,2]. Similarly, Barrett's esophagus, a precursor of esophagus adenocarcinoma, is thought to have a low prevalence among Asians. There are two reports from Peninsular Malaysia showing different rates of Barrett's esophagus in two different regions. Rajendra et al. [3] reported a 6% prevalence rate of Barrett's esophagus in the northwestern part of Peninsular Malaysia while Rosaida and Goh [4] in the Klang Valley of central Peninsular Malaysia reported a rate of 2%. The difference in rates may be related to differences in the ethnic composition of the population in these regions. The predominantly Malay population of northeastern Peninsular Malaysia has a naturally low Helicobacter pylori (H. pylori) prevalence rate of 4%-5%, a rate that is among the lowest reported in the world [5][6][7]. The H. pylori seroprevalence was also studied in the Aboriginal (Orang Asli) community from this region, and the prevalence rate was only 19% [8]. In a recent study, patients underwent upper endoscopy in northeastern Peninsular Malaysia; the rate of chronic atrophic gastritis was found to be 42%, and H. pylori infection was present in only 11% of the study sample [9]. The aim of this study was to determine the rate of Barrett's esophagus and histological oesophagitis in an endoscopybased sample of this unique population with a view to gaining an insight into the relationship between H. pylori infection and Barrett's esophagus. This is especially pertinent in the light of recent suggestions of an inverse relationship between H. pylori infection, gastroesophageal reflux disease (GERD), Barrett's and esophageal adenocarcinoma.   Endoscopic esophagitis was classified according to Los Angeles (LA) classification (Grade A to D) as described previously [10]. Barrett's esophagus was identified as an extension of columnar epithelium above the gastroesophageal junction. It was standard practice for all patients with esophagitis or suspected Barrett's esophagus to be biopsied at multiple levels at the squamocolumnar junction. H. pylori infection was routinely determined during endoscopy using the rapid urease test or CLO test (Ballard Medical Products, Utah, USA) and/or histology with or without Werthein stain from two tissue biopsies taken over the antrum and/or the body.

Patients and Methods
Reflux esophagitis was defined histologically by the presence of basal cell thickening, elongation of papillae of the epithelium, dilated vessels in the papillae of lamina propria, and distended pale squamous cells. The severity of reflux esophagitis was recorded as mild, moderate, and severe based on the presence of lymphoplasma cells infiltration in the tissue sample, and a visual grading of severity was given by the pathologist based on the volume of cell infiltration [11]. Barrett's esophagus was confirmed histologically by the presence of intestinal metaplasia epithelium within the columnar-lined esophagus which contains goblet cells and stains positively with Alcian blue [12]. Dysplasia was identified by the presence of definite cytologic and architectural abnormalities in the neoplastic epithelium. Cases were excluded if the diagnosis of dysplasia was confounded by the presence of marked inflammation [13].
Statistical analysis was performed using SPSS software version 12.0.1 (SPSS Inc, Chicago, Ill, USA). Chi-square test or Fischer's exact test was used for categorical variables and independent t test for numerical variables where appropriate. A P value of less than 0.05 was considered statistically significant.

Results
The rate of esophagitis seen endoscopically was 5.49% (104/1895) of which 71 patients were later confirmed histologically (Table 1). More males had endoscopic esophagitis than females, but it was not statistically significant (P = 0.97).  (Table 2). Compared to the demographics of the area and the ethnic distribution of all patients attending the hospital during study period, non-Malays were overrepresented among patients with reflux esophagitis and Barrett's esophagus. There was only one case of low-grade dysplasia seen, and the patient had background history of scleroderma. Normal histology was reported in 34 cases of which 31 cases were normal in erosive esophagitis seen endoscopically, and 3 cases were normal in suspected Barrett's esophagus.

Discussion
It is acknowledged at the outset that any prevalence study based on endoscopy is subject to selection bias. Observer variation among endoscopists in the recognition of esophagitis and Barrett's esophagus constitutes another limitation. Notwithstanding comparison with the data of other endoscoped series does permit useful inferences. The 5.49% prevalence rate of endoscopic esophagitis in this study is generally lower than that reported in most other endoscopy-based studies especially those from the Western world (Table 3). In concordance with this observation, the prevalence of Barrett's esophagus is also low at 0.79%. This is perhaps not surprising as Barrett's esophagus is considered to be a consequence of long-standing gastroesophageal reflux disease. The differences in the prevalence rates of esophagitis and Barrett's esophagus reported in the other two studies from Malaysia may relate to differences in ethnic composition and socioeconomic demographics. Northeastern Peninsular Malaysia has a larger proportion of ethnic Malays (90%) than Northwestern (51.4%) or Central (50%) Peninsular Malaysia. Rajendra et al. showed that Indians are more predisposed to gastroesophageal reflux than Chinese or Malays demonstrating the potential role of ethnic differences [3]. This is supported by an observation by the same author of a predominance of human leucocyte antigen (HLA) B7 subtype among Indians with Barrett's [20]. Indeed even in our study non-Malays were overrepresented among patients with reflux esophagitis and Barrett's esophagus.
There has been a suggestion that H. pylori infection rate is inversely related to the prevalence of GERD, Barrett's esophagus, and esophageal adenocarcinoma [21][22][23]. The low rate of esophagitis and Barrett's esophagus observed in this study despite the fact that H. pylori is almost absent does not support the notion that the absence of H. pylori infection plays more than a minor role as a risk factor for Barrett's esophagus and reflux esophagitis. Other factors relating to race, life style, and obesity are likely to be more important than H. pylori. The fact that non-Malays in this study were more likely to have esophagitis and Barrett's esophagus despite their well-documented higher prevalence of H. pylori infection also argues against the decline of H. pylori being an important factor in the etiopathogenesis of reflux disease. Indeed in northeastern region of Peninsular Malaysia, the age-standardised incidence of oesophageal adenocarcinoma a sequelae of Barrett's was found to be only 0.72/100,000 among the Malays [24]. Although progression from Barrett's to dysplasia is slow and rare (30% low-grade dysplasia and 5% high-grade dysplasia in six years) [25], it is notable that only one case of dysplasia was detected in this study, and even that was in a patient with scleroderma, a condition known to predispose to reflux esophagitis.
In conclusion, the low rate of Barrett's esophagus in this series from an area with an exceptionally low H. pylori prevalence does not support the hypothesis that the absence of H. pylori infection is more than a minor risk factor for 4 ISRN Gastroenterology Barrett's esophagus. This is corroborated by the observation that Non-malay subjects who have a higher prevalence of H. pylori were more likely to have Barrett's than the Malays.

Conflict of Interests
All authors did not report any conflict of interests.