Article Text

Download PDFPDF

Original article
The gastric acid pocket is attenuated in H. pylori infected subjects
  1. David R Mitchell1,
  2. Mohammad H Derakhshan1,
  3. Angela A Wirz1,
  4. Clare Orange2,
  5. Stuart A Ballantyne3,
  6. James J Going4,
  7. Kenneth E L McColl1
  1. 1Section of Gastroenterology, Institute of Cardiovascular and Medical Sciences, University of Glasgow, Glasgow, UK
  2. 2Department of Pathology, School of Medicine, University of Glasgow, Glasgow, UK
  3. 3Department of Clinical Radiology, NHS Greater Glasgow and Clyde, Glasgow, UK
  4. 4Institute of Cancer Sciences, University of Glasgow, Glasgow, UK
  1. Correspondence to Professor Kenneth E L McColl, Institute of Cardiovascular and Medical Sciences, University of Glasgow, 126 University Place, Glasgow G12 8TA, UK; kenneth.mccoll{at}glasgow.ac.uk

Abstract

Objective Gastric acid secretory capacity in different anatomical regions, including the postprandial acid pocket, was assessed in Helicobacter pylori positive and negative volunteers in a Western population.

Design We studied 31 H. pylori positive and 28 H. pylori negative volunteers, matched for age, gender and body mass index. Jumbo biopsies were taken at 11 predetermined locations from the gastro-oesophageal junction and stomach. Combined high-resolution pH metry (12 sensors) and manometry (36 sensors) was performed for 20 min fasted and 90 min postprandially. The squamocolumnar junction was marked with radio-opaque clips and visualised radiologically. Biopsies were scored for inflammation and density of parietal, chief and G cells immunohistochemically.

Results Under fasting conditions, the H. pylori positives had less intragastric acidity compared with negatives at all sensors >1.1 cm distal to the peak lower oesophageal sphincter (LES) pressure (p<0.01). Postprandially, intragastric acidity was less in H. pylori positives at sensors 2.2, 3.3 and 4.4 cm distal to the peak LES pressure (p<0.05), but there were no significant differences in more distal sensors. The postprandial acid pocket was thus attenuated in H. pylori positives. The H. pylori positives had a lower density of parietal and chief cells compared with H. pylori negatives in 10 of the 11 gastric locations (p<0.05). 17/31 of the H. pylori positives were CagA-seropositive and showed a more marked reduction in intragastric acidity and increased mucosal inflammation.

Conclusions In population volunteers, H. pylori positives have reduced intragastric acidity which most markedly affects the postprandial acid pocket.

  • HELICOBACTER PYLORI
  • GASTRIC ACID SECRETION
  • GASTRITIS
  • GASTRIC PARIETAL CELL
View Full Text

Statistics from Altmetric.com

Significance of this study

What is already known on this subject?

  • There is a negative association between Helicobacter pylori infection and both gastro-oesophageal reflux disease and oesophageal adenocarcinoma.

  • The mechanism of this negative association is unclear, but might be related to H. pylori reducing gastric acidity.

  • The gastric acid which refluxes into the oesophagus originates from the proximal gastric acid pocket.

What are the new findings?

  • In population volunteers, intragastric acidity was less in those with H. pylori infection, and this was most marked in the proximal stomach close to the gastro-oesophageal junction.

  • The density of parietal cells and chief cells was reduced in H. pylori positives compared with negatives.

  • The reduction in intragastric acidity and severity of inflammation were more marked in CagA-positive versus CagA-negative H. pylori infected subjects.

How might it impact on clinical practice in the foreseeable future?

  • The reduced intragastric acidity close to the gastro-oesophageal junction in the H. pylori infected subjects provides a mechanism for the negative association between the infection and reflux disease and its complications.

Introduction

Helicobacter pylori (H. pylori) is a common bacterial infection of the stomach present in the majority of the world's human population and resulting in varying degrees of inflammation of the underlying gastric mucosa. The infection is acquired in early childhood and usually persists indefinitely unless specifically eradicated.1

One of the major medical advances of the past 25 years has been the discovery that this common infection plays an important role in the aetiology of duodenal and gastric ulcers and also of gastric cancer.2 Eradicating the infection produces a long-term cure for the majority of patients with peptic ulcers unrelated to non-steroidal anti-inflammatory drug therapy. There is also increasing evidence that eradication of the infection reduces the risk of gastric cancer.3

An unexplained observation regarding the infection is its negative association with gastro-oesophageal reflux disease (GORD) and its complications of Barrett's oesophagus and oesophageal adenocarcinoma, with these disorders being less than half as common in infected subjects.4 ,5 It has been postulated that this negative association may represent the gastric infection protecting against these oesophageal disorders. If so, the falling incidence of the infection in the general population might explain the rising incidence of these oesophageal diseases.

One mechanism by which the infection might protect against oesophageal disease is by reducing the ability of the gastric mucosa to secrete acid and pepsin which are the constituents of gastric juice which can induce oesophageal damage. The infection is known to exert varying effects on gastric secretory function. In subjects with duodenal ulcers, the infection produces a non-atrophic gastritis with well-maintained gastric secretory cell mass which secretes increased amounts of acid due to the infection inhibiting the gastrin-mediated negative feedback control of acid secretion.6 In patients who develop gastric cancer, the infection induces an atrophic gastritis with loss of gastric secretory cells and thus reduced acid secretion. Only approximately 1 in 10 H. pylori infected subjects develop complicating ulcer disease or gastric cancer, and relatively little is known about the effects of the chronic infection on gastric secretory function in the 90% of infected subjects without these complications.7 If the degree of reduction in oesophageal disease in the H. pylori infected population is due to the infection reducing gastric acid secretion, then this suppression of acid secretion would need to be apparent in the majority of infected subjects.

Recent evidence indicates that it is the acidity of the gastric contents close to the gastro-oesophageal junction (GEJ), referred to as the acid pocket, which refluxes and causes oesophageal damage.8 It is also known that loss of gastric secretory cells due to H. pylori induced atrophic gastritis does not occur uniformly throughout the stomach but may be more marked at the periphery of the acid-secreting mucosa.9 In assessing any potential protective effect of the infection against oesophageal damage, it is important to examine the structure and secretory function of different anatomical regions of the stomach as well as its overall secretory capacity.

The aim of our study was to assess the gastric secretory status in different anatomical regions of the stomach and in subjects representative of the majority of the H. pylori infected population.

Methods and materials

Subjects

Study participants were volunteers from the general population of the West of Scotland. Subjects who were currently taking, or had recently taken, proton pump inhibitors (PPIs), were currently using H2 receptor antagonists or had ever received H. pylori eradication therapy were excluded. Recruitment was by general advertisement and from the NHS Scotland SHARE database.

Study design

Study day 1: clinical measurements and urea breath test

The presence and severity of any GI symptoms were assessed using the Short-Form Leeds Dyspepsia Questionnaire,10 and a medication history was recorded. Measurements of height, weight, waist and hip circumference were taken. Volunteers were tested for H. pylori infection by C14 urea breath test. Fasting serum and plasma samples were stored at −20°C and later tested for H. pylori CagA IgG, using ELISA (Genesis Diagnostics, Littleport, UK).

Study day 2: upper GI endoscopy

Volunteers attended after an overnight fast for upper GI endoscopy. They were offered topical lidocaine throat spray or conscious sedation with midazolam 1–3 mg. Biopsies were taken using large capacity biopsy forceps (Radial Jaw 4; Boston Scientific, Hemel Hempstead, UK) with a jaw span of 8 mm. Two junctional biopsies were taken perpendicular to the squamocolumnar junction (SCJ), one from the lesser and one from the greater curve, and targeted to include squamous mucosa at the proximal end. Three further junctional biopsies were taken longitudinally below the SCJ, aiming for end-to-end biopsies starting at 6, 12 and 18 mm distal to the SCJ down the lesser curve. In addition, six further gastric biopsies were taken from gastric fundus, midbody on greater curve, midbody on lesser curve, distal body on greater curve, incisura angularis and antrum. Finally, two small metal radio-opaque clips were attached to the SCJ using a single use rotatable clip fixing device (QuickClip 2; Olympus, Southend-on-Sea, UK).

Biopsy specimen processing

Biopsies were immediately placed onto non-adherent dental wax and oriented flat. More detailed information concerning the two-stage orientation method has been described elsewhere.11 The specimens were later embedded in agar on the filter paper without further manipulation. Staining was performed with conventional H&E, as well as monoclonal antibodies to H+/K+ATPase, pepsinogen I and gastrin.

Study day 3: combined manometry and pH study

The volunteers attended after an overnight fast for combined high-resolution manometry and pH studies. The combined probe was passed prenasally and positioned, so that the most proximal pH sensor was 5 cm above the lower oesophageal sphincter (LES), with the remaining 11 sensors lying across the sphincter and within the proximal stomach. The relative positions of the 12-sensor pH catheter, 36-sensor manometer and SCJ are shown in figure 1. Manometry and pH data were recorded concurrently for a 20-minute fasting period. Subjects then consumed a standardised meal over 10 min (400 g Waitrose spaghetti bolognese ready meal and 100 mL water (500 kcal; 55.2 g carbohydrate, 27.8 g protein, 17.6 g fat)). Following this, manometry and pH recordings were continued for a further 90 min. An X-ray was taken before and after the meal to visualise the metal clips at the SCJ.

Figure 1

Schematic diagram of the relative positions of the 12-sensor pH catheter, 36-sensor manometer and squamocolumnar junction (SCJ) (identified by attached metal clip).

Equipment

High-resolution pH metry

pH recordings were taken using a high-resolution pH catheter (Synectics Medical, Enfield, UK). This was a custom-made pH probe composed of 12 antimony pH electrodes, with the most distal electrode situated 5 mm from the tip of the catheter, with the other 11 electrodes 35, 46, 57, 68, 79, 90, 101, 112, 123, 134 and 169 mm proximal to the tip. The probe was calibrated prior to each study using pH buffer solution (Synmed, Enfield, UK) at pH 7.01 and pH 1.07. Recordings were captured using POLYGRAM NET software (Synectics Medical, Enfield, UK).

High-resolution manometry

Manometry was performed using a high-resolution solid-state catheter, with 7.5 mm spacing between the 36 circumferential sensors (Given Imaging, Hamburg, Germany). Calibration was performed prior to each study, and in vivo calibration was carried out on a weekly basis and applied to each study to compensate for thermal drift. Recordings were captured with ManoScan 360 high-resolution manometry system and analysed with ManoView ESO V.3.0.1 software (Given Imaging, Hamburg, Germany).

Combined probe

The manometry and pH catheters were combined using two thin strips of Leukoplast Sleek waterproof tape (BSN Medical, Pinetown, South Australia, USA), such that manometry sensor 25 was immediately adjacent to pH sensor 3.

Data analysis

Intragastric acid

The 90-minute postprandial period was split into three 30-minute periods for analysis. The median pH for each of the 12 pH sensors was calculated for the 20-minute fasting period and the three 30-minute postprandial periods. Acid exposure at the GEJ was also examined by calculating the percentage of time pH <4.

Manometry

Manometric characteristics were analysed in detail during fasting and the same three postprandial periods. For each 2-minute period, one inspiratory point and one expiratory point were chosen from the longest period, without interference from swallowing, coughing or transient lower oesophageal sphincter relaxations (TLESRs). The mean pressure in inspiration and expiration was calculated for each of the 36 sensors over the 20-minute fasting period and 30-minute postprandial periods. The peak LES pressure was taken as the sensor showing the highest mean pressure. The position of the SCJ was derived from the position of the metal clips relative to the combined manometry and pH sensors seen on X-ray.

Histopathological assessment

A. Studies using conventional H&E

Glandular height: The vertical height of epithelium starting from lamina propria to the tip of the gland was measured in three well-oriented and representative fields and expressed as ‘total thickness of epithelium’. To measure the ‘glandular height’, the same method was limited to the areas of gland containing secretory cells, but not superficial foveolar epithelial cells. All results were expressed as median (IQR), in mm.

Inflammatory scoring: The intensity of inflammatory infiltrate by polymorphonuclear (PMN) and mononuclear (MN) cells was scored semiquantitatively (0=none; 1=mild; 2=moderate; 3=severe) as recommended in the Updated Sydney Classification of Gastritis.12 A combined inflammatory score was calculated as the sum of these two scores. Intestinal metaplasia (IM) was scored by estimating the proportion of epithelial surface covered by goblet cells.

B. Immunohistochemistry

The oriented biopsies, double-embedded in agar and paraffin, were cut in standard 4-μm thickness and immunostained individually for parietal cells, chief cells and G cells. For parietal cells, we used a commercial mouse monoclonal anti-H+/K+ ATPase (Ab 2866, Abcam, Cambridge, UK) diluted at 1:20 000. For chief cells, a mouse monoclonal antipepsinogen 1 antibody (Ab 50123, Abcam, Cambridge, UK) was used at a dilution of 1:4000. The G cells were stained with antigastrin (Ab 16035, Abcam, Cambridge, UK) diluted at 1:200. A Thermo Quanto Detection kit (TL-125-OHD, Thermo Fisher, UK) was used as secondary antibody.

Quantification of secretory cells

To calculate the density of parietal cells, chief cells and G cells, absolute numbers of stained cells were counted at a magnification of ×125 in three well-oriented and representative fields (1 mm2 each) and expressed as mean cell number per 1 mm2 area in each patient. All selected areas must have had complete glands located in sagittal plane, in which the lamina propria was at the bottom and luminal side of epithelium was at the top.

Statistical analysis

All continuous data are expressed as medians and IQRs unless otherwise stated. Comparison of variables between groups was made using the Mann-Whitney U test. Biopsy inflammatory scores are presented as cross-tabulations and compared using Fisher's exact test. Significance for all statistical tests was set as p<0.05.

Ethics

The study protocol was approved by the West of Scotland Ethics Committee, and all volunteers provided informed written consent.

Results

Of the 137 subjects assessed for eligibility for the study, 49 were excluded due to current or recent use of PPI therapy (n=9) or history of previous H. pylori eradication therapy (n=8) or declining to participate following full explanation of the study protocol (n=32). Eighty-eight subjects proceeded to the urea breath test, of which 31 were H. pylori positive, and all of these went on to complete the full study protocol. Of the 57 testing H. pylori negative, 28 went on to complete the study due to 1 withdrawing consent after the endoscopy and 28 not being selected to proceed in order to maintain matching of the positive and negative groups with respect to age, gender and body mass index (BMI).

The 31 H. pylori positive and 28 H. pylori negative subjects who completed the study were well matched with respect to age (55 vs 56 years; p=0.95), gender (18/31 vs 18/28 males; p=0.84) and BMI (26.3 vs 26.8 kg/m2; p=0.72). There were seven current smokers in the H. pylori positive group compared with one current smoker in the H. pylori negative group (p=0.035).

The median dyspepsia score for H. pylori positives was 2.0 (range 0–9) compared with 0 (range 0–3) for the H. pylori negative subjects (p=0.002). Out of 31 H. pylori positives, 17 (54.8%) of them were taking no medication compared with 10/29 (35.7%) of the H. pylori negative subjects. The most frequent medications were antihypertensives, statins, antidepressants and inhalers for asthma. No subject was taking medications known to affect gastric secretion.

At endoscopy, four H. pylori positive subjects had a hiatus hernia (2–4 cm in length), one subject had Los Angeles (LA) grade A reflux oesophagitis and one subject had Barrett's mucosa of 3 cm. None of the H. pylori negatives had a hiatus hernia, although two subjects had reflux oesophagitis (LA grades A and B).

Gastro-oesophageal acidity

Under fasting conditions, the H. pylori positive subjects had less intragastric acidity compared with the H. pylori negatives at all sensors more than 1.1 cm distal to the peak LES pressure (table 1). The fall from neutral oesophageal pH to highly acidic intragastric pH also occurred more abruptly in the H. pylori negatives. At the sensor 3.3 cm distal to the peak LES pressure, the median pH in the H. pylori negatives had fallen to 2.27 compared with 6.13 in the positives (p<0.001). The radio-opaque clips indicated that this pH sensor was 1.8 cm distal to the SCJ.

Table 1

Median (IQR) pH values at sensors relative to peak LES pressure comparing Helicobacter pylori negative (HP−) (n=28) and positive (HP+) (n=31) groups during 20-minute fasting period and three 30-minute postprandial periods

Throughout the three postprandial periods, intragastric acidity was significantly less in the H. pylori positives at the pH sensors placed 2.2, 3.3 and 4.4 cm distal to the peak pressure of the LES, but no significant difference was detected by the more distal sensors placed at 5.5 and 6.6 cm distal to this reference point (table 1). These three sensors detecting a significant difference in gastric acidity between the two groups were those closest to the GEJ, with the most proximal of them being only 0.6 cm distal to the SCJ (figure 2).

Figure 2

Median pH for 0–30 min period after meal relative to lower oesophageal sphincter (LES) and squamocolumnar junction (SCJ) in Helicobacter pylori positive (HP+) and negative (HP−) subjects.

The percentage of time pH<4 for each of the three postprandial periods was significantly greater in the H. pylori negatives versus positive subjects for the electrodes extending 3 cm distal to the peak LES pressure, at the peak LES pressure and also extending 1.1 cm above the peak LES pressure (table 2).

Table 2

Median (IQR) percentage time pH<4 at sensors relative to peak LES pressure comparing Helicobacter pylori negative (HP−) (n=28) and positive (HP+) (n=31) groups during 20-minute fasting period and three 30-minute postprandial periods

Gastric histopathology

A. Conventional H&E staining

Inflammation

The H. pylori positives had a greater combined inflammatory cell infiltrate at each of the 11 biopsy sites compared with the H. pylori negatives (table 3). The increased combined inflammatory cell infiltrate in the H. pylori positives consisted of a mixture of PMN cells and MN cells and tended to be more intense, close to the SCJ, lesser curve, distal stomach, incisura and antrum, compared with the gastric fundus and midbody (p<0.05 for each). The H. pylori negatives had MN cell infiltrate limited to the SCJ and also to a lesser extent at the antrum and angularis incisura, but its intensity was less than that of the H. pylori positives at these sites. There was minimal evidence of PMN cell infiltrate at any location in the H. pylori negatives.

Table 3

Cross-tabulation table showing the number of subjects within the Helicobacter pylori negative (HP−) and positive (HP+) groups, with each combined inflammatory score (0–6) at the 11 different gastric biopsy locations

Intestinal metaplasia

IM was identified in 14 of the 31 H. pylori positive subjects. In seven of these, it was limited to one or more of the biopsies from midbody lesser curve, distal body greater curve, incisura angularis and antrum. In three of the subjects, it was present in at least one of the above sites and also in the biopsies close to the SCJ. In a further three, it was limited to the region close to the SCJ. In one subject, it was present in each biopsy, except for one of the biopsies from the SCJ.

IM was identified in only 4 of the 28 H. pylori negative subjects. In three of these, it was only seen in the biopsies across the SCJ, and in the fourth subject, it was only seen in the biopsy from the fundus.

Gastric gland height

The height of the gastric secretory glands was significantly reduced in the H. pylori positive versus negative subjects throughout the gastric mucosa, except for the biopsies taken across the SCJ (table 4).

Table 4

Median (IQR) of glandular thickness and densities of parietal and chief cells at each biopsy location comparing Helicobacter pylori negatives (n=28) and positives (n=31)

B. Immunohistochemistry

Parietal and chief cell density

The H. pylori positives had a significant reduction in density of both parietal and chief cells compared with H. pylori negatives, and this was seen at each of the 11 intragastric locations assessed except for the SCJ greater curve where the difference did not achieve statistical significance (table 4). The degree of reduction was similar for the two cell types.

The depletion of both cells in the H. pylori positives versus negatives was more marked in the biopsies taken from the distal gastric mucosa (ie, antrum, incisura angularis and distal body greater curve), being reduced by 67%–100% compared with that observed in the more central region of the oxyntic mucosa (fundus and midbody) at 26%–35% (figure 3). In addition, the length of mucosa extending distal to the SCJ which contained no detectable parietal cells was greater in the H. pylori positives versus negatives (1.5 vs 1.0 mm; p=0.013). However, the degree of reduction in specialised cell density in the biopsies taken 6 and 12 mm distal to the SCJ (38%–47%) was not dissimilar from that observed in the more central oxyntic mucosa (ie, fundus and midbody) (26%–35%) (figure 3).

Figure 3

Relative reduction in parietal and chief cell densities at different gastric locations in Helicobacter pylori infected versus non-infected. Note: At the GE junction and distal stomach, these cells are reduced by 80%, whereas in the midbody, reduction was about 30%. Biopsy locations: JG––across squamocolumnar junction (SCJ) above greater curve; JL1––across SCJ above lesser curve; JL2––6 mm distal SCJ; JL3––12 mm distal SCJ; JL4––18 mm distal SCJ; BG3––fundus; BL––midbody, lesser curve; BG2––midbody, greater curve; BG1––distal body, greater curve; IA––incisura angularis; Ant––antrum.

G cell density

The density of G cells was reduced in the antrum of the H. pylori positive versus negative subjects (48 (IQR: 31–86) vs 91 (64–129), p<0.001), but the converse was seen with respect to the biopsies taken from the distal body region (0 (IQR: 0–32) vs 0 (0–0), p=0.007).

Intragastric acidity and histology in CagA-Positive H. pylori subjects

Seventeen of the H. pylori positives were CagA-seropositive and fourteen CagA-seronegative. The associations with reduced intragastric acidity in comparison with H. pylori negatives were more apparent for the CagA positives: being significant in five of the six intragastric sites, both fasting and after the meal in the CagA positives, but in only two of the six intragastric sites for the CagA negatives, only under fasting conditions (see online supplementary table S1). There was a statistically significant difference between the CagA negatives and CagA positives in only two of the six sites during fasting and one of the six sites after the meal.

The CagA positives had a significantly greater combined inflammatory cell infiltrate evident at three of the eleven biopsy locations (6 and 18 mm distal SCJ, and distal body greater curve), compared with the CagA negatives (see online supplementary table S2). The reduction in parietal and chief cell density was significant at each intragastric location for both CagA-positive and CagA-negative subjects, with no apparent difference between these two groups.

Discussion

In our volunteers recruited from the general population of the West of Scotland, those with H. pylori infection had less intragastric acidity, both under fasting conditions and following a meal, compared with uninfected volunteers matched for age, gender and BMI. In addition, those with the infection had a reduced density of both acid-secreting parietal cells and pepsin-producing chief cells compared with those uninfected. These findings indicate that H. pylori infection within our Western population is associated with a less acidic and proteolytic intragastric environment.

The reduced intragastric acidity in the H. pylori positive subjects was apparent throughout the stomach under fasting conditions. After the meal, however, the reduced acidity in the H. pylori positives was evident within the first few centimetres distal to the GEJ, but no significant difference in acidity was apparent in the main body of the stomach. There was also evidence of increased acidity after the meal in the H. pylori negatives, right at the SCJ junction and extending 2 cm above it, indicating increased intrasphincteric acid reflux. We and others have previously reported that the proximal region of the stomach close to the GEJ largely escapes the buffering effect of ingested food and may remain highly acidic after a meal.13–15 This phenomenon has been called the acid pocket and is thought to be important in GORD-induced oesophageal damage after a meal when reflux is most common. It is therefore interesting that it is at this region close to the GEJ where the reduced acidity was most apparent in the H. pylori infected subjects.

What is the reason for the reduced acidity in the H. pylori positives after a meal, being most markedly close to the GEJ. There was no evidence that the depletion in parietal cell density in the H. pylori positives was more pronounced over the few centimetres close to the GEJ compared with other regions in the stomach. Inflammation may also inhibit gastric secretory function,16 and this was slightly increased close to the GEJ and also in the distal stomach compared with the midbody gastric mucosa. The elevation of intragastric pH following the meal in the H. pylori positives being most marked close to the GEJ may simply reflect the relative intragastric distribution of gastric juice and ingested food. Following a meal, the food occupies the centre of the stomach and the secreted gastric juice, the region close to the stomach wall which secretes it. Impaired acid secretion will elevate the pH of the gastric juice, and this will be most apparent close to the stomach wall. In contrast, the central region of the stomach will reflect the pH of the food and thus will be relatively unaffected by changes in the acidity of secreted juice. The effect of H. pylori on intragastric pH after the meal being most evident close to the GEJ may be due to this region being close to the wall of the stomach.

Whatever the explanation for the changes in acidity between H. pylori positives and negatives being most marked close to the GEJ, after the meal, the observation is likely to be important with respect to the propensity of gastro-oesophageal reflux producing oesophageal damage. It is well recognised that gastric juice which refluxes into the oesophagus is that present close to the GEJ, and also that reflux most commonly occurs during the postprandial period when TLESRs are most frequent.17

The reduction in parietal cell density observed in the H. pylori positive subjects was associated with a similar reduction in chief cell density. This is consistent with the infection and inflammation causing a loss in gastric glands and also with the previous literature showing that the development of parietal and chief cells is intimately linked.18 We did not measure the secretion of pepsin and other digestive enzymes produced by the chief cells, but their reduced density is likely to be associated with reduced secretory capacity after the meal. Reduction in gastric juice peptic activity has previously been reported in H. pylori infected subjects.19 The peptic activity of the gastric juice is as important, and arguably more important than its acidity, with respect to the ability to damage oesophageal mucosa, and therefore the reduction in both specialised cells is likely to represent a substantial reduction in the damaging capacity of reflux gastric juice in H. pylori infected subjects.20

There was a reduction in the density of G cells in the antrum of the H. pylori positives indicating a depletion of antral as well as oxyntic glands. In contrast, G cell density in the distal body mucosa of the H. pylori positives was higher than that in the H. pylori negative subjects. This can be explained by the distal acid-secreting body mucosa, which does not have G cells, being replaced by an antral-like mucosa that contains G cells (a process that has been called ‘antralisation’). This process can be associated with the development of pseudopyloric metaplasia, also called spasmolytic polypeptide expressing metaplasia.21–24 This is consistent with our observation that the reduction in parietal and chief cell densities in H. pylori positives was most pronounced in the distal body mucosa. Together, these findings are likely to represent the previously reported proximal progression of the junction between the antrum and body-type mucosa, leading to shrinkage in the surface area of the stomach covered by oxyntic mucosa in H. pylori atrophic gastritis.25

There are few previous studies assessing gastric secretory function in H. pylori infected healthy volunteers in the Western world. In a retrospective analysis of 95 healthy, young male volunteers (age 19–26 years), Smith et al26 reported that the 8 seropositive for H. pylori had similar intragastric acidity to the other 87. In a retrospective analysis of 136 healthy volunteers, Peterson et al27 reported reduced basal acid output, but no significant difference in gastrin-stimulated peak acid output or meal-stimulated acid output, assessed by intragastric titration in H. pylori seropositives. In a prospective study of 206 healthy volunteers, Feldman et al28 reported reduced gastrin-stimulated peak acid output and reduced basal pepsin output in those with H. pylori detected histologically in gastric biopsies. In 1998, our own group reported a reduced acid secretory response to gastrin stimulation in 20 H. pylori positive versus 24 H. pylori negative healthy volunteers.29 Several studies in the Japanese population have reported reduced gastric secretory function in H. pylori positive healthy volunteers.30 ,31

Our current study differs from previously published studies in a number of important respects. First, we aimed to study subjects representative of the general population infected with H. pylori rather than asymptomatic healthy volunteers. Second, by using intragastric pH sensors, we avoided the use of non-physiological gastric stimuli, gastric aspiration or intragastric titration which may not be representative of the subjects’ usual gastric functioning. Third, we focused on the middle-aged population rather than young students, as the former is the population in whom reflux disease manifests itself. Finally, and probably most critically, we employed a technique which allowed us to assess the acidity in different regions of the stomach, and in particular, close to the GEJ.

Our observation that gastric acidity was reduced most markedly close to the GEJ is interesting in the light of the previously reported but unexplained observations by Feldman et al.32 They observed that in healthy volunteers, eradication of H. pylori did not alter basal or meal-stimulated gastric acid secretion assessed by intragastric titration, but did result in a twofold to threefold increase in gastro-oesophageal acid reflux.32 In the light of our current study, the observed increase in gastro-oesophageal acid reflux may have been explained by the H. pylori infection reducing intragastric acidity close to the GEJ.

Is our finding of reduced gastric secretory function in the H. pylori infected population a peculiar feature of our West of Scotland population or relevant to the wider Western community? H. pylori induced atrophic gastritis and reduced acid secretory function is associated with gastric cancer, and the prevalence of the two correlates at a population level.33 The incidence of gastric cancer in Scotland is 9.7/10 000 py, similar to that of Western European and North American countries, and substantially lower than that of Eastern European and Far East countries.34 This would suggest that our findings of reduced acid secretory function is representative of what is happening in Western countries.

Though our study demonstrates that the H. pylori infected general adult population has less intragastric acidity than the uninfected population, this association does not necessarily indicate that the reduced intragastric acidity is caused by the infection. However, causal association seems highly likely as H. pylori gastritis is recognised to cause loss of gastric glands and impaired secretory function. In addition, the more marked changes in gastric secretory function in those with the more virulent CagA strain supports it being caused by the infection. Confirming causality by an intervention study has potential problems, as H. pylori induced loss of gastric glands is generally regarded as being irreversible.

In summary, our current study indicates that H. pylori infected population volunteers have reduced intragastric acidity compared with uninfected controls and that this is most marked close to the GEJ. This observation may explain the negative association between the infection and GEJ disease and its complications.

References

View Abstract

Footnotes

  • Contributors DRM: Clinical investigations, manometry, pH metry, data analysis and drafting manuscript. MHD: Histological assessment, biopsy orientation, data analysis and drafting manuscript. AAW: Recruitment of volunteers and assisting clinical investigations. CO: Technical assistance in histology and scanning of histological slides. SAB: Radiological assessment. JJG: Histological assessment and drafting manuscript. KELM: Conception of original idea, drafting manuscript and overall supervision.

  • Funding This study was funded by a grant from NHS Greater Glasgow and Clyde Endowment Fund.

  • Competing interests None declared.

  • Patient consent Obtained.

  • Ethics approval West of Scotland Ethics Committee.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.