| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
|
Interact CardioVasc Thorac Surg 2007;6:304-307. doi:10.1510/icvts.2006.143370 © 2007 European Association of Cardio-Thoracic Surgery
Gastric fluid bile concentrations and risk of Barrett's esophagus
a Department of Surgery, University of Washington, Seattle, WA, USA Received 3 September 2006; received in revised form 16 January 2007; accepted 22 January 2007
*Corresponding author. Department of Cardiothoracic Surgery, University of Wisconsin, Madison, 600 Highland Ave, H3/466 Madison, WI 53792-3236, USA. Tel.: +1-608-220-3401; fax: +1-608-263-0547.
Patients with Barrett's esophagus are at high risk of progression to adenocarcinoma. A growing, but conflicting body of evidence implicates bile reflux as a contributor to Barrett's esophagus. To investigate whether duodenogastric reflux was associated with an increased risk of Barrett's esophagus, a case-control study of incident Barrett's esophagus was performed. Cases (n=72) were identified by new histologically-confirmed diagnosis of specialized intestinal metaplasia (indicative of Barrett's esophagus) following upper endoscopy for refractory gastroesophageal reflux between October 1997 and September 2000. Cases were compared to gastroesophageal reflux patients without specialized intestinal metaplasia (controls; n=72). There was no difference in total bile acid concentrations between cases and controls. Risk of Barrett's esophagus did not significantly vary with increasing concentrations of total or free bile acids, respectively (OR 0.35 (95% CI 0.12, 1.02) and 0.60 (95% CI 0.22, 1.66)). Low gastric fluid pH (toxic range 3–5), was associated with a non-significant increase in the risk of Barrett's esophagus. In conclusion, no significant association between Barrett's esophagus and total or free bile acids in gastric refluxate was found. Patients with low gastric fluid pH (3–5) may represent a subset of patients at high risk of developing Barrett's esophagus.
Key Words: Barrett's esophagus; Metaplasia; Esophageal neoplasms; Gastroesophageal reflux; Bile acids and salts
The majority of esophageal adenocarcinoma (EC) arises in mucosa replaced by specialized intestinal-type glandular epithelium (SIM) [1]. Barrett's esophagus (BE) is defined as an endoscopically-detectable change in the epithelium of the tubular esophagus of any length with biopsy-proven SIM, excluding intestinal metaplasia of the cardia [2]. Patients with BE have an estimated 0.5–1.0% annual risk of developing EC [3]. In addition to chronic gastroesophageal reflux (GERD), obesity, male sex and cigarette smoking [4, 5] appear to be important in the etiology of BE, but much remains to be learned. Growing evidence suggests that bile reflux is important in the etiology of BE. In animal studies, bile acids, especially in acid environments, accumulate in esophageal mucosal cells and cause cell membrane and tight junction dissolution [6]. This process allows acid and activated pepsins access to the submucosal region, precipitating more severe injury. Duodenal fluid also increases the gastric fluid pH to 3–5, a range which promotes phenotypic differentiation of cardiac-type mucosa toward SIM [7]. Several human studies have attempted to quantify the association between mucosal injury and duodenogastric reflux (DGR). Fein and colleagues found that 80% of patients who had DGR, as measured by Bilitec probe monitoring, demonstrated esophageal mucosal injury compared to 54% with acid and 10% with bile reflux alone [8]. In addition, studies specifically investigating BE risk and DGR have reported a correlation between bile acid levels in refluxate and the presence of BE [9]. Identification of patients at risk for BE amongst all patients with GERD is an important step in halting the rising incidence of EC, increasing endoscopic accuracy and decreasing the number of screening endoscopies. DGR may represent an important measurable, and modifiable, etiologic factor. The purpose of this study was to determine whether DGR (measured as bile in fasting gastric fluid) is associated with an increased risk of BE among persons undergoing endoscopy to investigate GERD in a community practice setting.
Participants in this study were part of a larger case-control study of incident BE in patients aged 20–80 years who underwent upper endoscopy (EGD) for GERD symptoms between October 1, 1997 and September 30, 2000. GERD symptoms were defined as pyrosis, regurgitation, eructation, transient dysphagia or atypical chest pain. Participants with previously diagnosed BE were ineligible for the study. Informed consent was obtained prior to endoscopy. In addition to biopsies taken in the course of clinical management, four-quadrant biopsies were taken in the tubular esophagus just proximal to the squamocolumnar junction. Specimens were fixed in zinc formaldehyde and evaluated by pathologists (masked to endoscopy findings) for the presence or absence of SIM using Alcian blue staining [10]. BE cases were defined by the presence of SIM. GERD patients for whom the pathologic specimens were histologically negative for SIM were randomly selected as controls.
Gastric fluid samples ( At endoscopy, the physician recorded the presence or absence of visible columnar epithelium. Recently, the requirement for endoscopically visible columnar metaplasia has been challenged, with some authors describing ‘ultra-short-segment’ BE as histologically confirmed columnar epithelium which is not endoscopically-visible. Since some cases had SIM with grossly normal-appearing esophageal mucosa, results for all patients will refer to the histologic diagnosis of SIM. When analysis specifically addressed those participants whose EGD and histology meet criteria for BE, presentation of results will refer to the diagnosis of BE. To minimize bias from variation in analysis, gastric fluid was processed with equal numbers of cases and controls in batches. Analysis for bile acids was performed as follows: 4 ml of gastric juice was added to 16 ml of chloroform/methanol (2:1) in a 22-ml screw-cap tube and shaken thoroughly. After centrifugation, the upper layer (methanol/water) was removed and evaporated to dryness under nitrogen at 40 °C. The residue was dissolved in 3 ml of distilled water.A 300-mg solid-phase extraction cartridge (C18 Bond Elut, Varian Assoc., Walnut Creek, CA) was washed with 10 ml methanol, followed by 10 ml water. The sample was applied to the cartridge, which was then washed with 10 ml water and 3 ml hexane. Bile acids were then eluted with 3 ml methanol. The methanol was evaporated to dryness and the residue taken up in 1 ml methanol for HPLC analysis. HPLC analysis for conjugated and free bile acids was carried out with a three-pump Rainin Dynamax system (Rainin Division of Varian), using a modification of Scalia's method [11], in which ammonium acetate is substituted for the phosphate buffer in order to utilize the greater sensitivity of the ‘mass detector’. The sample was applied to a 5 micron, 4.6x250 mm Spherisorb ODS2 column (Phase Separations Ltd., Clwyd, UK) and eluted with a gradient from 65% methanol/35% 0.03 M ammonium acetate, pH 4.5–90% methanol/10% ammonium acetate buffer, followed by 100% methanol. Fourteen standard bile acids were quantitated using an evaporative light-scattering detector (ELSD Mk III, Alltech, Deerfield, IL). Statistical analysis was performed using STATA Corp 8.0 [12]. Geometric mean total and individual bile acid concentrations were calculated. Means were compared for differences between cases and controls using a two-tailed t-test (alpha=0.05). Estimates of the relative risk of specialized intestinal metaplasia associated with gastric fluid concentrations of both total and individual bile acids were approximated by the odds ratio, controlling for potentially confounding factors (age, tobacco use, body mass index and sex) using stratified analysis and logistic regression models. GERD controls were used as the referent population for quartile determination of bile acid levels.
Selected characteristics of cases and controls are presented in Table 1. Sex, race, body mass index (BMI) and waist-to-hip ratio (WHR) distributions were similar for cases and controls. Cases were older and more likely to be former smokers than controls. Male cases tended to have higher BMI and WHR than controls.
A mean volume of 3.6 ml of gastric aspirate was obtained from both cases and GERD controls. For the latter half of the study, gastric aspirate pH was measured and found to be 3.2±2.0 for cases and 3.1±2.0 for GERD controls. EGD findings are shown in Table 2.
There was no statistically significant association between presence of SIM and total or free bile acid levels, or any of the bile acid subspecies (Table 3a). This finding persisted in multivariate analysis (Table 3b: controlling for age, BMI, smoking status, WHR and sex; P for trend=0.25 and 0.14, respectively).
30/72 cases with SIM had visible columnar epithelium on EGD suggestive of Barrett's metaplasia. Comparing these cases with all GERD controls, the OR tended to be closer to unity. For example, the OR comparing extreme quartiles changed from 0.35 to 0.81 (95% CI 0.17, 3.75) and from 0.60 to 1.05 (95% CI 0.27, 4.10) for total and free bile acids, respectively. Comparing persons with gastric pH in the toxic range (six cases and four controls) to those whose gastric pH was either very acidic (pH 1–2) or close to neutral (pH 6–7), the OR was 1.68 (95% CI 0.43, 6.54). There was no evidence of a significant interaction between toxic pH and total or free bile acid levels (data not shown) To examine whether the level of total bile acids or bile acid subspecies was associated with the presence of esophagitis at endoscopy, the odds ratio for esophagitis relative to bile acid level was determined. As with SIM, no association was found between quartiles of total and free bile acids and the presence of esophagitis on EGD. When stratified by SIM cases and GERD controls, the odds ratios for esophagitis by total and free bile acids were 0.94 (95% CI 0.68, 1.29) and 1.06 (95% CI 0.79, 1.41), respectively, comparing the highest quartile to the lowest.
The findings of this study do not support the hypothesis that increased bile acid concentrations in gastric fluid aspirate increase the risk of SIM or BE in patients referred for evaluation of GERD. This lack of association persisted after controlling for potential risk factors for BE. When analyzed by pH level, there was a modest and non-significant increase in risk of SIM when gastric fluid pH was within the toxic range of pH 3–5. However, this was based on small numbers, and could easily have occurred by chance. These findings suggest that the presence of bile acids alone in the gastric fluid of patients with GERD is not a sufficient explanation for subsequent development of BE. The presence of endoscopically visible columnar metaplasia may be an important distinguishing characteristic of BE [2]. However, when analysis was limited to those cases with visible columnar metaplasia of the tubular esophagus, there continued to be no evidence for an increased risk of BE. The question of a toxic range for pH is especially important and is an area where bile acids may play a critical role. When bile acids in duodenal bicarbonate solution reflux into the stomach, esophageal refluxate pH increases. The level of elevation depends on the baseline pH of the patient. When pH is in the range of 3–5, bacteria propagate and deconjugate bile acids. Bile acids then become soluble, thus enhancing entry into the esophageal mucosal cells and phenotypic differentiation toward SIM [13]. Fein and colleagues studied the impact of DGR in surgically altered rats and found that the presence of bile in the refluxate was carcinogenic and that the administration of acidified water (pH 1.8) protected against the development of adenocarcinoma [14]. As stated, this study lacked sufficient numbers to fully elucidate the relationship between pH and incident cases of SIM, but did suggest an association when analysis was limited to those subjects with pH range of 3–5. This study has several strengths. First, it is one of the largest studies investigating the relationship between bile in the refluxate and the finding of incident BE. Secondly, standardization of techniques for tissue and data collection and case identification promoted reliable and replicable information. Finally, use of the HPLC technique for analysis of bile acid levels is an accurate and reproducible means of quantifying bile acid levels [15]. There are, however, several limitations to this study as well. First, these bile acid measurements are one-time measurements from fasting gastric aspirate. Endoscopy is typically performed following a night of fasting, an important consideration as bile acid synthesis decreases and storage increases during the night. In addition, bile acid secretion is stimulated by eating; therefore, fasting bile acid levels may be markedly different than post-prandial bile acid levels and may not be reflective of the average level of bile acids present in the stomach over time. In addition, gastric bile acid levels are being used as a surrogate for esophageal exposure. It is possible that actual esophageal exposure (duration, volume, total time of exposure) may be different between cases and controls due to potential differences in gastric or esophageal motility. For this study, physiologic measures such as esophageal pH or bile exposure (as measured by pH and bile probes), and esophageal clearance as measured by manometry, were not available. Finally, the scope of the analysis which took into account pH was limited by small numbers. In summary, this study did not find a significant association between concentrations of total or free bile acids in gastric refluxate and SIM with or without visible columnar epithelium in patients referred for upper endoscopy to evaluate GERD. There is, however, a 68% increase in the risk of SIM/BE in patients with gastric fluid pH in the toxic range (3–5) which, though not statistically significant, may merit further study. This may represent a subset of patients with GERD who are at particularly high risk of developing BE and subsequent EC.
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
Top
Abstract
1. Introduction
5 ml) were collected during endoscopy. Samples were frozen and stored at –70 °C until assayed. For each case, one control with gastric fluid was randomly selected for analysis of bile acid concentrations. One control was found to be ineligible after bile acid assays had been completed, leaving 72 cases and 71 controls for analysis.