An increased number of gamma/delta T-cells and gastric epithelial cell expression of the groEL stress-protein homologue in Helicobacter pylori-associated chronic gastritis of the antrum

Numerous studies have shown that the presence of Helicobacter pylori in the stomach is linked to the development of chronic gastritis most commonly seen in the antrum. However, the pathogenic mechanisms are unclear. In 23 of 31 patients, examined due to symptoms from the upper gastrointestinal tract, H. pylori-associated chronic gastritis of the antrum was diagnosed histologically and by growth of H. pylori. Immunoperoxidase staining on gastric biopsy specimens from these patients showed an increased number of gamma/delta T-cells within the epithelium. Furthermore, the Mab ML30 (raised against the 65 kDa heat shock protein of mycobacteria) demonstrated positive staining in the gastric epithelial cells in all H. pylori-positive but not in H. pylori-negative biopsy specimens. H. pylori also reacted with ML30, as detected by immunoperoxidase staining as well as by immunoblotting. Intraepithelial gamma/delta T cells may play a role in host defense against invading H. pylori, and the bacteria may trigger an autoimmune response to stress proteins expressed by the gastric epithelial cells.

Gastritis and its relation to gastric carcinogenesis

There have been rapid changes in understanding of chronic gastritis recently, fueled largely by the recognition of gastric infection by Helicobacter pylori. This report reviews the newer classifications of chronic gastritis, and describes the characteristics of H pylori-positive and -negative patterns of gastritis. The relevance of H pylori infection to gastric carcinogenesis is also discussed.

Helicobacter pylori and gastric histamine concentrations

An enzyme isotopic assay was used to determine the histamine concentration in the gastric mucosa of patients positive for Helicobacter pylori with (n = 11) and without duodenal ulceration (n = 9) and in negative controls (n = 7). A significant difference was observed when the histamine content of H pylori negative subjects was compared with that of positive patients. On the other hand, there was no significant difference in histamine concentration between H pylori positive patients with duodenal ulceration and those without duodenal ulceration. H pylori positive patients with and without duodenal ulceration had significantly lower gastric histamine concentrations than H pylori negative subjects. The lower gastric histamine concentration observed in H pylori positive patients might be due to increased histamine release which could in turn induce increased gastric acid secretion.

Long-term follow up of patients with gastritis associated with Helicobacter pylori infection

The aim of this study was to evaluate the long-term prognosis for patients suffering from gastritis associated with Helicobacter pylori infection, and in particular the proportion of cases progressing to peptic ulcer. The study was carried out in one urban general practice. One hundred and three patients who had presented with dyspepsia over the 1973-80 period and who were found to have a macroscopically normal endoscopy were reassessed between seven and 14 years later. Gastric antral biopsies had been taken routinely at endoscopy and were subsequently re-examined for the presence of H pylori. The patients' medical records were examined to establish their consulting rates over the follow-up period and whether they suffered from any other medical conditions. Patients were interviewed to assess the course of their dyspeptic symptoms. Comparison of patients who were unequivocally H pylori positive with those who were negative revealed no significant differences in the consultation rate for gastroenterological symptoms, in the proportion of patients referred to a hospital consultant or for further gastroenterological investigations or in the proportion reporting that their symptoms had improved. However, a statistically highly significant relationship was found between H pylori infection and hypertension. The results of this study have shown that there is a good prognosis for non-ulcer dyspepsia whether or not H pylori infection is present. The association between H pylori gastritis and hypertension clearly merits further investigation.

Sensitivity of the gastric mucosa to acid and duodenal contents in patients with nonulcer dyspepsia

Nonulcer dyspepsia is a common clinical syndrome whose etiology is unknown. The sensitivity of the gastric mucosa to acid and duodenal contents in 18 patients with nonulcer dyspepsia was studied. The patients had a normal upper gastrointestinal endoscopy and biopsy specimens were obtained for determination of Helicobacter pylori status. Fifteen of the 18 patients were infected with H. pylori. All patients underwent intubation with double-lumen tube and collection of cholecystokinin-stimulated pancretico-biliary secretions. Subsequently, normal saline, 0.1N hydrochloric acid, and autologous duodenal secretions were infused into the stomach in a randomized blinded fashion. A positive response was defined as the production of epigastric pain by acid and/or bile but not by saline. By this definition, only 6 patients (33%) had a positive response and none had reproduction of their usual symptoms. In patients with a negative response, only 4 remained asymptomatic during all infusions. The remaining 8 had symptoms during infusion of saline, 7 of whom also had symptoms during infusion of acid and/or duodenal secretions. Two of these patients had reproduction of their usual symptoms. In conclusion, the gastric mucosa in patients with nonulcer dyspepsia is not abnormally sensitive to acid or duodenal contents.

[Campylobacter pylori gastroduodenitis in children with juvenile rheumatoid arthritis]

Clinico-morphological study of the gastroduodenal system and also morphometric investigations of the degree of gastroduodenitis activity and the condition of the local immune system were made in 30 children aged from 4 to 15 years with juvenile rheumatoid arthritis (JRA). Microbiological, histological and electron microscopic methods for revealing Campylobacter pylori (CP) were used. Interrelationship between various clinical manifestations of JRA with the nature of gastroduodenal system affection and with the incidence of microbial colonization of SP was determined. All the patients with JRA had lesions in the gastric and duodenal mucosa, and some morphological features were revealed in them. SP was diagnosed in 86.7 per cent. There was a relation between the microbial clonization of SP and the degree of the rheumatoid process activity. It is suggested that SP does not play an etiological role in the development of gastroduodenitis in JRA but may serve as a pathogenetic factor in the development of an erosive-ulcerous lesion.

Some perspectives on peptic ulcer

In this article, the author comments on the following topics: whether peptic ulcer is a disease or just the sign of another process; whether the crater has a multifactorial origin; the ubiquity of Helicobacter pylori and the conclusion that a specific strain will doubtless prove to be the cause of some, but not all, ulcers; the inherent definitional problem of "nonulcer dyspepsia"; and the controversy over how much gastric acid needs to be reduced to speed the healing of peptic ulcer.

Non-ulcer dyspepsia and the dilemma posed to its management

Dyspepsia is one of the most common ailment that is managed by general practitioners and gastroenterologists worldwide and particularly so in Tropical Africa. Non-ulcer dyspepsia can be defined as the presence of classic ulcer symptoms or any combination of burning, bloating, indigestion or other abdominal symptoms which only sometimes have a relation to food intake with out any radiological or endoscopic evidence of peptic ulceration. Compared to peptic ulcer disease, the literature on Non-ulcer dyspepsia is relatively scanty and studies have been done mostly in populations non-negroid origin. The present study has reviewed Non-ulcer dyspepsia with particular reference to the Africans and has highlighted the intriguing areas of the disorder with the consequent dilemma posed to its management. Further research on this topic is indicated and would be fascinating but a challenging task.

Helicobacter pylori: aggressor or innocent bystander?

Helicobacter pylori seeks gastric mucosa, whether found in the stomach, duodenum, or Barrett's esophagus. Definitive diagnosis can be secured by appropriate stains of mucosal biopsies and culture, but the rapid urease test, breath isotope studies, and serologic testing are also useful. The frequency of colonization increases with advancing age, but infection occurs earlier in underdeveloped countries. Although the reservoir is uncertain, water or food transmission seems likely. There is sufficient evidence to assign an etiologic role to the bacteria in the causation of type B antral gastritis. H. pylori is found in areas of gastric metaplasia within the duodenum and is associated with duodenitis. Although acute infection leads to hypochlorhydria, chronic colonization has little effect on acid secretion. Studies have thus far failed to establish a convincing relationship between H. pylori and nonulcer dyspepsia, although the bacteria may play a role in selected patients. H. pylori is found in association with most idiopathic gastric and duodenal ulcers, but it is unclear as to whether the bacteria plays a causative or permissive role. The organism has a predilection for intercellular spaces and the mucous layer, thus affording relative isolation from luminally active antibiotics. Monotherapy with bismuth preparations transiently eliminates the bacteria, but recolonization is rapid, probably due to regrowth of sequestered organisms. A combination of metronidazole, bismuth, and tetracycline (or amoxicillin) affords the best eradication rate, but the potential side effects of this program should be considered. The present therapy of duodenal ulcer disease is effective and without significant risk. Treatment of H. pylori should be reserved for those patients who relapse on adequate maintenance therapy. If a safe and effective antibiotic becomes available, more frequent testing and earlier treatment intervention may become more attractive. H. pylori is probably an "innocent bystander" for most patients, but the bacteria may sufficiently impair the defenses of the antral and duodenal mucosa to facilitate the development and relapse of ulcer disease in subsets of patients.

Non-ulcer dyspepsia in Nigeria: a case-control study

Non-ulcer dyspepsia is common in Africa, yet there are few data on its possible cause. In this study 40 patients with strictly defined non-ulcer dyspepsia were matched with 40 asymptomatic volunteers. Both groups were questioned concerning their intake of alcohol, cola nut, non-steroidal anti-inflammatory drugs, and smoking. They then underwent upper gastrointestinal endoscopy with mucosal biopsy of the gastric mucosa: these biopsies were examined for gastritis and Helicobacter pylori. Ingestion of cola nut was the only statistically significant difference between the two groups. A high prevalence of gastritis and H. pylori infection was found in both groups, suggesting that these factors are not important in the aetiology of non-ulcer dyspepsia in northern Nigeria.

Peptic ulcer pathophysiology

Despite extensive research, the etiology of peptic ulcer disease remains unclear. Given the multiple processes that control acid and pepsin secretion and defense and repair of the gastroduodenal mucosa, it is likely that the cause of ulceration differs between individuals. Acid and pepsin appear to be necessary but not sufficient ingredients in the ulcerative process. It is clear that the majority of gastric ulcers and a substantial number of duodenal ulcers do not have increased gastric acid secretion. Recent research has focused more on protection and repair of the stomach and duodenum. NSAIDs cause a significant number of gastric and duodenal ulcers; this is probably due to inhibition of prostaglandin production with loss of its protective effects. In the absence of NSAIDs and gastrinoma, it appears that most gastric ulcers and all duodenal ulcers occur in the setting of H. pylori infection. Evidence is mounting in support of H. pylori as a necessary ingredient in the ulcerative process, similar to acid and pepsin. It is not known whether the bacteria or the accompanying inflammation is the more important factor in the pathophysiology. Although the pathophysiology of gastric ulcer and duodenal ulcer is similar, there are clearly differences between the two groups. Duodenal ulcer is typified by H. pylori infection and duodenitis and in many cases impaired duodenal bicarbonate secretion in the face of moderate increases in acid and peptic activity. These facts suggest the following process: increased peptic activity coupled with decreased duodenal buffering capacity may lead to increased mucosal injury and result in gastric metaplasia. In the presence of antral H. pylori, the gastric metaplasia can become colonized and inflamed. The inflammation or the infection itself then disrupts the process of mucosal defense or regeneration resulting in ulceration. A cycle of further injury and increased inflammation with loss of the framework for regeneration may then cause a chronic ulcer. Gastric ulcer often occurs with decreased acid-peptic activity, suggesting that mucosal defensive impairments are more important. The combination of inflammation, protective deficiencies, and moderate amounts of acid and pepsin may be enough to induce ulceration. Many questions remain in understanding the pathophysiology of peptic ulcer disease. The physiology and pathophysiology of mucosal regeneration and the mechanisms by which H. pylori and inflammation disrupt normal gastroduodenal function will be fruitful areas of future investigation.

Eradication of Helicobacter pylori abolishes 24-hour hypergastrinaemia: a prospective study in healthy subjects

In a prospective study, eight young healthy subjects (five with an active H. pylori infection in the antral mucosa) were treated with a course of tripotassium dicitrato bismuthate, amoxycillin and metronidazole. The triple therapy eradicated infection when assessed 20-24 weeks later by antral biopsy (urease, histology, and 13C urea breath test [4 out of 5 subjects]). Twenty-four hour intragastric acidity and plasma gastrin concentration were measured before treatment, and 4-6 weeks and 20-24 weeks post-treatment. Treatment did not affect acidity in either the H. pylori-positive or H. pylori-negative groups, nor did it affect the plasma gastrin profile in the H. pylori-negative group. Eradication of H. pylori infection in five subjects caused a drop of the median integrated 24-hour plasma gastrin concentration from 558 pmol.h/L before treatment to 307 and 289 pmol.h/L at 4-6 and 20-24 weeks post-treatment, respectively. It is concluded that H. pylori infection is associated with 24-hour hypergastrinaemia, and that in apparently healthy subjects normal gastric physiology can be restored by eradication of the infection.

Helicobacter pylori gastritis mimicking gastric carcinoma at CT evaluation

The abdominal computed tomographic (CT) scans from 61 patients with biopsy-proved Helicobacter (formerly Campylobacter) pylori gastritis were retrospectively reviewed. The CT scans were interpreted on the basis of the original report of the findings at CT examination and without knowledge of the results of biopsy. Of 19 patients (31%) with gastric abnormalities at CT, 14 (74%) had inflammatory changes initially reported as suspicious for gastric malignancy; malignancy was entertained as the primary diagnosis in four of those patients. In five of the 19 abnormal cases (26%), the diagnosis with CT was gastritis. The two major patterns of severe H pylori infection identified were (a) circumferential antral wall thickening and (b) thickening of the posterior gastric wall along the greater curvature, with or without evidence of ulceration. Thickening averaged 1.5-2.0 cm in cases suspicious for malignancy. The majority of abnormalities involved the gastric antrum (68%). No cases demonstrated significant adenopathy, obliteration of fat planes, or invasion of adjacent organs.

Gastroduodenal mucus gel thickness in patients with Helicobacter pylori: a method for assessment of biopsy specimens

The gastroduodenal mucus layer is considered the primary mucosal protective barrier, especially important in the maintenance of a mucosal pH gradient. Thus, the measurement of the mucus layer thickness in various disease states could advance our understanding of gastroduodenal pathophysiology. We present a novel method for measuring the mucus layer in endoscopic biopsy material and compare layer thickness in Helicobacter pylori (HP)-negative and HP-positive specimens. Endoscopic biopsies were obtained from 17 patients with gastroduodenal mucosa harboring HP and from 15 patients without current HP colonization. The thickness of the mucus layer was measured in fresh specimens by the phase-contrast dark-field microscopy technique. In patients with confirmed HP infection, the thickness of the mucus layer (mean +/- SD) was 0.093 +/- 0.033 mm in duodenal, 0.085 +/- 0.027 mm in antral, and 0.105 +/- 0.033 mm in corporal mucosa. In patients without concomitant HP colonization, the thickness of the mucus gel was 0.162 +/- 0.045 mm, 0.175 +/- 0.067 mm, and 0.161 +/- 0.064 mm in duodenum, antrum, and corpus, respectively. The differences between the means were statistically significant (p less than 0.001 for the duodenal, p less than 0.001 for antral, and p less than 0.01 for corporal mucosa). This study suggests that colonization of the gastroduodenal mucosa by HP impairs the mucus layer covering the surface epithelium. This mucus layer impairment may lead to mucosal injury with subsequent development of inflammation and, possibly, peptic ulcer disease.

Gamma delta T cell receptor-positive cells of the human gastrointestinal mucosa: occurrence and V region gene expression in Heliobacter pylori-associated gastritis, coeliac disease and inflammatory bowel disease

T cells expressing the gamma delta heterodimer of the T cell receptor (TCR) were studied with respect to their occurrence and expression of gamma delta TCR variable region (V) genes in the normal gastrointestinal mucosa and in a variety of inflammatory conditions. In controls, gamma delta TCR+ cells were a minority population confined to the epithelial compartment of stomach, small bowel and colonic mucosae. Unlike in the periphery, gastro-intestinal gamma delta TCR+ intraepithelial lymphocytes (IEL) were mainly V delta 1+ (89.98 +/- 17.70%); few were V delta 2+ (6.04 +/- 13.8%) or V gamma 9+ (11.38 +/- 10.73%). All gamma delta TCR+ IEL were CD5low; nearly half were CD8+ and the remainder were CD4-CD8- 'double negatives'. There was no significant change from normal in percentages of gamma delta TCR+ IEL in H. pylori-associated gastritis, Crohn's disease and ulcerative colitis. However, in coeliac disease, gamma delta TCR+ IEL were elevated from 2.54% (+/- 1.71) in controls to 29.6% (+/- 16.1) in untreated patients (P less than 0.001) and 18.5% (+/- 7.2) in treated patients (P less than 0.001) and more were CD4-CD8-. Otherwise, gamma delta TCR+ IEL phenotypes were little changed: the majority remained V delta 1+V delta 2-V gamma 9- and all were CD5low. These data suggest that increased gamma delta TCR+ IEL are not a generalized response to intestinal inflammation or to stress proteins, although the typical V delta 1+V delta 2-V gamma 9- CD5low phenotype is retained.

Effect of Helicobacter pylori infection on the severity of gastroduodenal mucosal injury after the acute administration of naproxen or aspirin to normal volunteers

This study asked whether Helicobactor pylori infection accentuated the severity of NSAID-induced mucosal injury of the stomach or duodenum. We evaluated the severity of acute mucosal injury and H. pylori status in 61 normal volunteers (ages 22-43 yr) receiving naproxen (1000 mg, n = 30) or aspirin (3900 mg, n = 31) daily for 7 days. NSAID-induced gastric and duodenal mucosa each were endoscopically graded separately for hemorrhages and erosions-ulcers on a scale of 0 to 4. H. pylori infection was identified by a sensitive and specific ELISA. Nine of the 30 subjects in the naproxen group and 12 of the 31 subjects in the aspirin group were H. pylori positive (p = NS). There was no statistically significant difference between the frequency of mucosal hemorrhage in those with and those without H. pylori infection (44% compared with 33% for those receiving naproxen and 90% of those receiving ASA, p = NS for each). There were also no differences in the frequency or severity of erosive mucosal injury seen, e.g., acute ulcers were found in 16.5% and 17.5% of infected and uninfected subjects, respectively. We conclude that the presence of H. pylori infection does not influence the degree or type of mucosal damage associated with the acute administration of naproxen or aspirin.

Quantification of Helicobacter pylori infection in gastritis and ulcer disease using a simple and rapid carbon-14-urea breath test

Gastric urease was studied isotopically in 230 patients with biopsy-proven normal mucosa or chronic gastritis, including 59 patients with ulcer disease. Carbon-14-urea was given in 25 ml of water without substrate carrier or nutrient-dense meal, and breath samples were collected over a 60-min period. The amount of 14CO2 excreted at 10 min was independent of the rate of gastric emptying and was not quantitatively influenced by the buccal urease activity. The 10-min 14CO2 values discriminated well between Helicobacter pylori positive and negative patients (94% sensitivity, 89% specificity) and correlated with the number of organisms assessed by histology. The test was a good predictor of chronic gastritis (95% sensitivity and 96% specificity), and a quantitative relationship was observed between 14CO2 values and the severity and activity of the gastritis. In H. pylori positive patients, breath 14CO2 was found to be similar in patients with and without ulcer disease, suggesting that the number of bacteria is not a determining factor for the onset of ulceration.

Estimation of prevalence of Helicobacter pylori infection in an asymptomatic elderly population comparing [14C] urea breath test and serology

A non-invasive serological assay devised in this laboratory had a sensitivity and specificity of 100% as determined by culture and confirmed by histology in a group of 47 patients who had undergone endoscopy. The correlation between serology and the non-invasive [14C] breath test was very good. Only one of 24 culture positive patients was, while all 23 culture negative patients were, breath test negative. In a group of 46 healthy elderly persons, however, significant anomalies between serology and breath test were observed. Only 83% of the breath test negative persons were seronegative, while only 68% of the breath test positive persons were seropositive. These results can be explained in terms of age related atrophic gastritis and immune incompetence, causing reduced colonisation and decreased antibody production, respectively. These investigations suggest that non-invasive tests for H pylori infection may not be reliable in the elderly.

[The experimental and clinical effect of ciprofloxacin on the microflora of the gastrointestinal tract]

The study of the influence of cyprofloxacin on the microflora of the gastrointestinal tract has been made under experimental and clinical conditions. As revealed in this study, cyprofloxacin produces a corrective effect on the intestinal microflora; the action of this preparation, in contrast to that of other antimicrobial preparations, is retained for a long time. In patients having duodenal ulcer with bacteriosis caused by Campylobacter pylori and with intestinal dysbacteriosis the combination of cyprofloxacin and cimetidine yields a higher therapeutic effect than the use of cimetidine alone.

Can diagnosis of Helicobacter pylori be rapid and yet sensitive?

Diagnostic accuracy of 5 tests viz., endoscopy, rapid urease, 24 hours urease, culture and histology, were evaluated in the detection of Helicobacter pylori (H. pylori) infection in 50 patients undergoing upper G.I. endoscopy. Endoscopic evidence of gastritis to predict H. pylori infection was 50% specific and 46% sensitive. Rapid and 24 hours urease test and culture were 100% specific when compared with histology and their sensitivity was 71%, 62% and 21% respectively. Of the three 100% specific tests, rapid urease test yields results within 15 minutes; therefore this test being easy, rapid and sensitive should be used for screening of H. pylori infection; followed by histology for further confirmation.

Helicobacter pylori in peptic ulcer disease in Kenya

Sixty six patients with dyspeptic symptoms underwent upper gastrointestinal endoscopy and biopsies for Helicobacter pylori culture. The number of H. pylori isolated increased with age reaching a peak at 51 to 60 years of age. Antral gastritis closely followed by duodenitis accounted for the highest number of H. pylori isolated, (87.5% and 85.7% respectively). In patients with duodenal ulceration only, 57% had H. pylori isolated from their antral biopsies, a result that was just slightly higher than that where no endoscopic diagnosis was made (50%).

Sampling efficiency in the diagnosis of Helicobacter pylori infection and chronic active gastritis

The methods and sampling procedures used in the diagnosis of Helicobacter pylori infection and chronic active gastritis were evaluated. Five biopsy specimens for bacteriological cultivation and three specimens for histological examination were obtained endoscopically from a defined area of the gastric antral mucosae of 83 patients. An increase in the number of biopsy specimens for cultivation from one to five revealed only one more H. pylori-infected patient. H. pylori was isolated from 31 of 83 patients. Three technically adequate samples for histological examination were obtained from each of 74 patients. Of these 74 patients, chronic active gastritis was diagnosed by demonstration of typical histological changes in all three specimens from each of 20 patients, in two of three specimens from each of 3 patients, and in one of three specimens from 1 patient. The results indicate that one biopsy specimen is sufficient for the isolation of H. pylori, whereas several specimens may be necessary for the histological diagnosis. Chronic active gastritis was found in four patients not infected with H. pylori; on the other hand, H. pylori was isolated from nine patients who showed no signs of chronic active gastritis in any of three samples.

Low peptic ulcer and high gastric cancer prevalence in a developing country with a high prevalence of infection by Helicobacter pylori

We compared the prevalence rates of peptic ulcer (duodenal and gastric) and gastric cancer in 1,796 dyspeptic Peruvian patients with those reported in 2,883 similar patients from developed countries. The prevalence of total peptic ulcer was significantly lower, and that of gastric cancer significantly higher, in the Peruvian patients. The prevalence of gastric ulcer was lower but not significantly so. We deduced that the significantly lower prevalence of total peptic ulcer was directly related to the low prevalence rate of duodenal ulcer. We hypothesize that the reason for these differences was probably a higher prevalence of Helicobacter pylori-associated chronic atrophic gastritis with hypochlorhydria in the Peruvian patients. Hypochlorhydria decreases the predisposition to peptic ulcer (especially duodenal ulcer), and chronic atrophic gastritis may predispose an individual to gastric cancer.

Helicobacter pylori in children

The clinical manifestations of Helicobacter pylori (formerly Campylobacter pylori) infection in children overlap with those encountered in adults but there are some important differences. The aim of this review is to describe the range of manifestations of this infection in children, together with means of diagnosis and treatment.

The prevalence of Helicobacter pylori infection in the Peoples Republic of China

Meta-analysis on data from 16 (two volunteer and 14 endoscopic) studies ahs been performed to investigate the prevalence of Helicobacter pylori infection in the stomachs of individuals within the Peoples Republic of China. In this survey covering 2216 cases (89 volunteers and 2127 endoscopy patients), the incidence of H. pylori infection in 15- to 22-yr-old healthy volunteers, and in 13- to 88-yr-old symptomatic patients undergoing gastroscopy was 49.4% and 64.5%, respectively. The frequency of H. pylori infection in chronic gastritis (63.6%), gastric ulcer (71.9%), and duodenal ulcer disease (73.1%) differed significantly from that in histologically normal individuals (8.2%), confirming the relationship of H. pylori infection with those disease states. We found no significant association between H. pylori infection and symptoms.

A 20-minute breath test for helicobacter pylori

In this study, we evaluated a simplified rapid 14C-urea breath test for the diagnosis of Helicobacter pylori. Fasting patients undergoing initial assessment for H. pylori drank 5 microCi of 14C-urea in 20 ml of water. Breath was collected at intervals for 30 min. Samples were counted in a beta-counter, and the results were expressed as counts per minute (cpm). In the same week, patients underwent endoscopy, and a blinded investigator examined biopsy samples of gastric mucosa by culture and histology for H. pylori. There were 49 H. pylori-negative (HP-) and 104 H. pylori-positive (HP+) patients in the study. HP+ patients expired a mean of 4398 cpm (SD 2468) per mmol CO2 in a sample taken 20 min after ingestion of the isotope. In contrast, HP--patients expired only 340 cpm (SD 196). If the mean +3 SD of HP- patients was used as a cutoff value, the 20-minute sample gave a sensitivity of 97% and a specificity of 100% for detecting H. pylori. The radiation exposure from this test is less than 1% of that received from an upper gastrointestinal series, and the short collection time makes it both convenient and cost effective.