SEARCH

SEARCH BY CITATION

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Study design
  6. Patients
  7. Study protocol
  8. Symptom assessment
  9. Biopsy, histology and 13C-urea breath test
  10. Statistical analysis
  11. Results
  12. Patient population and baseline characteristics
  13. Histological findings and topographical expression of chronic gastritis
  14. Eradication of H. pylori infection
  15. Effect on dyspeptic symptoms
  16. Sub-group analysis
  17. Reflux complaints
  18. Evaluation of safety
  19. Discussion
  20. Acknowledgements
  21. References

Aim : To assess whether the eradication of Helicobacter pylori leads to long-term relief of symptoms in functional dyspepsia.

Methods : Eight hundred patients with functional dyspepsia were randomized to receive double-blind treatment with twice-daily 30 mg lansoprazole, 1000 mg amoxicillin and 500 mg clarithromycin for 7 days (L30AC), twice-daily 15 mg lansoprazole, 1000 mg amoxicillin and 500 mg clarithromycin for 7 days (L15AC), or once-daily 15 mg lansoprazole for 14 days (LP). Dyspepsia and reflux symptoms were monitored for 12 months.

Results : In intention-to-treat analysis, the non-ulcer dyspepsia sum score showed a statistically significant benefit in terms of symptom relief in the L30AC group (P = 0.0068) compared with the LP group, but there was no significant difference between the L15AC and LP groups (P = 0.2). When all patients in the two eradication therapy arms were considered together, successful eradication had a significant benefit with regard to the complete absence of symptoms (P < 0.04). H. pylori eradication did not lead to an increase in reflux symptoms.

Conclusion : This study suggests that H. pylori infection causes dyspeptic symptoms in a subset of patients with functional dyspepsia, and that these patients may obtain long-term symptomatic benefit following H. pylori eradication.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Study design
  6. Patients
  7. Study protocol
  8. Symptom assessment
  9. Biopsy, histology and 13C-urea breath test
  10. Statistical analysis
  11. Results
  12. Patient population and baseline characteristics
  13. Histological findings and topographical expression of chronic gastritis
  14. Eradication of H. pylori infection
  15. Effect on dyspeptic symptoms
  16. Sub-group analysis
  17. Reflux complaints
  18. Evaluation of safety
  19. Discussion
  20. Acknowledgements
  21. References

It has been suggested that Helicobacter pylori-induced inflammation of the gastric mucosa may cause functional dyspepsia, at least in a subset of patients. However, so far, neither epidemiological studies nor a careful search for specific symptoms in patients with H. pylori-associated functional dyspepsia have supported this hypothesis.1 In several, mostly small, clinical trials, H. pylori eradication has not provided symptomatic short-term benefit in comparison with other medications providing symptomatic relief.1, 2

By contrast, a possible long-term effect of H. pylori eradication therapy in functional dyspepsia appears to be conceivable, because of the finding of a greater symptomatic benefit in patients 1 year after H. pylori eradication compared with those in whom the infection persisted.3 A subsequent meta-analysis by Laheij et al. supported a long-term therapeutic gain of about 10% over conventional symptomatic therapies.4 However, most of these studies and those undertaken subsequently had limitations in study design, particularly in terms of the small sample size.1, 2, 4, 5

As a consequence, the issue of H. pylori eradication in functional dyspepsia remains an important controversy in clinical practice. In the European guidelines developed during the 1996 Maastricht Conference6 and updated in 2000,7H. pylori therapy for functional dyspepsia was recommended as an advisable option on the basis of only limited scientific facts and circumstantial evidence. At present four large well-designed prospective trials have been published. Of these, three multi-centre trials failed to show a significant benefit,8–10 whilst one single-centre study showed the efficacy of H. pylori eradication in functional dyspepsia, compared with placebo or acid suppressant therapy alone.11 The discrepancy between these studies has been explained by differences in patient selection and in the methods used to evaluate the symptomatic benefit.12 Two very recent meta-analyses were conducted to further explore whether H. pylori eradication may confer benefit to patients with functional dyspepsia and a positive H. pylori status.13, 14 Again, the results were discordant and inconclusive.

To resolve this controversial topic, we designed a multi-centre, prospective, randomized controlled study to investigate the effect of H. pylori eradication in patients with functional dyspepsia. Special care was taken to select a representative and sufficiently large number of patients who lived in a similar environment, rather than recruiting patients from different countries with dissimilar cultural backgrounds,8, 9 and who displayed significant dyspeptic symptoms over a defined period of time. Further aims of the study were to search for factors that could positively impact on symptom outcome following H. pylori eradication, and to monitor whether reflux symptoms were affected during a 1-year follow up.

Study design

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Study design
  6. Patients
  7. Study protocol
  8. Symptom assessment
  9. Biopsy, histology and 13C-urea breath test
  10. Statistical analysis
  11. Results
  12. Patient population and baseline characteristics
  13. Histological findings and topographical expression of chronic gastritis
  14. Eradication of H. pylori infection
  15. Effect on dyspeptic symptoms
  16. Sub-group analysis
  17. Reflux complaints
  18. Evaluation of safety
  19. Discussion
  20. Acknowledgements
  21. References

This randomized, double-blind study (ELAN, Eradication vs. Lansoprazole in H. pylori-Associated Non-ulcer dyspepsia) was conducted at 85 centres in Germany between January 1997 and February 1999, according to the principles of Good Clinical Practice and in accordance with the revised Declaration of Helsinki and the current version of the German Drug Act. The study protocol and the patient information and consent form were approved by an independent ethics committee at each study centre, and written informed consent was obtained from the patients before enrolment.

Patients

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Study design
  6. Patients
  7. Study protocol
  8. Symptom assessment
  9. Biopsy, histology and 13C-urea breath test
  10. Statistical analysis
  11. Results
  12. Patient population and baseline characteristics
  13. Histological findings and topographical expression of chronic gastritis
  14. Eradication of H. pylori infection
  15. Effect on dyspeptic symptoms
  16. Sub-group analysis
  17. Reflux complaints
  18. Evaluation of safety
  19. Discussion
  20. Acknowledgements
  21. References

Patients between 18 and 65 years of age, seeking medical care for dyspeptic symptoms that had been present for at least four consecutive weeks, or had been chronically recurring for an extended period of time, were eligible for the study. At inclusion, endoscopy was performed to exclude the presence of gastric or duodenal ulcers or erosions, changes in the oesophageal mucosa and any signs of a pre-malignant or malignant condition. Patients with predominant heartburn or alarm symptoms suggesting serious diseases, such as unintentional weight loss or anaemia, and patients who had undergone gastric resection were excluded. Further exclusion criteria were known or suspected intolerance to one of the study drugs and treatment with H2-receptor antagonists or proton pump inhibitors 1 week prior to enrolment or antibiotics or bismuth compounds within the last 4 weeks before the start of the study. Patients with concomitant diseases which might be the cause of dyspeptic symptoms, such as gallstones, or who were taking any medication with a possible influence on dyspeptic symptoms were also excluded, as were patients with excessive nicotine and/or alcohol consumption.

Study protocol

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Study design
  6. Patients
  7. Study protocol
  8. Symptom assessment
  9. Biopsy, histology and 13C-urea breath test
  10. Statistical analysis
  11. Results
  12. Patient population and baseline characteristics
  13. Histological findings and topographical expression of chronic gastritis
  14. Eradication of H. pylori infection
  15. Effect on dyspeptic symptoms
  16. Sub-group analysis
  17. Reflux complaints
  18. Evaluation of safety
  19. Discussion
  20. Acknowledgements
  21. References

Patients with dyspeptic symptoms conforming to the inclusion criteria and H. pylori infection confirmed by the rapid urease test (CLO) during initial endoscopy were entered into the study. The severity of dyspepsia and reflux symptoms was assessed by the investigators, based on a validated dyspepsia and reflux scoring system. Patients with a dyspepsia score of ≥ 6 and a reflux score of < 3 (up to moderate reflux complaints) were randomly assigned to a 2-week treatment period as follows: L30AC (first week: 30 mg lansoprazole b.d., 1 g amoxicillin b.d. and 500 mg clarithromycin b.d.; second week: placebo for lansoprazole once daily); L15AC (first week: 15 mg lansoprazole b.d., 1 g amoxicillin b.d. and 500 mg clarithromycin b.d.; second week: placebo for lansoprazole once daily); LP (15 mg lansoprazole once daily, placebo for lansoprazole once daily, placebo for amoxicillin b.d. and placebo for clarithromycin b.d.). The randomization was performed in blocks, consisting of three consecutive numbers. As each study centre received only complete blocks of study medication, the balance of the treatment groups within ‘complete’ centres was ensured. This procedure was chosen to achieve a regular allocation of the patients to the treatment groups within one centre and between the centres.

Treatments were given orally in a double-blind fashion, the active medication and the corresponding placebo being identical in appearance. The patients' check-up visits were scheduled for the end of treatment, and 6 weeks, 6 months and 12 months after the end of treatment. Symptoms of dyspepsia and reflux were assessed at each visit. The 13C-urea breath test was used to confirm the eradication of H. pylori. Endoscopy with biopsy was performed on an optional basis at the 6-week, 6-month and 12-month visits, in addition to the baseline assessment.

Symptom assessment

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Study design
  6. Patients
  7. Study protocol
  8. Symptom assessment
  9. Biopsy, histology and 13C-urea breath test
  10. Statistical analysis
  11. Results
  12. Patient population and baseline characteristics
  13. Histological findings and topographical expression of chronic gastritis
  14. Eradication of H. pylori infection
  15. Effect on dyspeptic symptoms
  16. Sub-group analysis
  17. Reflux complaints
  18. Evaluation of safety
  19. Discussion
  20. Acknowledgements
  21. References

The dyspeptic symptoms included in the assessment were upper abdominal complaints, belching, nausea, fullness, flatulence, premature feeling of satiation, lack of appetite and bad breath. Daytime and night-time upper abdominal complaints were recorded separately. The frequency and intensity of symptoms reported by the patients were scored according to the validated Likert scale, from 0 to 3 for frequency (0, less than once per week; 1, one to three times per week; 2, daily or almost daily; 3, several episodes daily) and from 0 to 4 for intensity (0, not present; 1, mild complaints that can be ignored if the patient does not think of them; 2, moderate complaints that cannot be ignored, but that do not disturb daily activities or sleep at night; 3, severe complaints disturbing daily activities, as well as sleep at night; 4, most severe complaints representing a considerable disturbance and/or interruption of daily activities, as well as of sleep at night).

Reflux symptoms (heartburn, acid regurgitation, swallowing problems and retrosternal pain) were also assessed at each visit. The intensity of reflux symptoms was graded from 0 to 4 using the same scale as for the dyspeptic symptoms.

Biopsy, histology and 13C-urea breath test

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Study design
  6. Patients
  7. Study protocol
  8. Symptom assessment
  9. Biopsy, histology and 13C-urea breath test
  10. Statistical analysis
  11. Results
  12. Patient population and baseline characteristics
  13. Histological findings and topographical expression of chronic gastritis
  14. Eradication of H. pylori infection
  15. Effect on dyspeptic symptoms
  16. Sub-group analysis
  17. Reflux complaints
  18. Evaluation of safety
  19. Discussion
  20. Acknowledgements
  21. References

During the initial endoscopy, two biopsy specimens were taken from the corpus, antrum and angulus for histological analysis, and one additional specimen was taken from the antrum and the corpus for the CLO test.

Histological assessment of the biopsies was performed at a central histopathological institution and graded according to the Houston modification of the Sydney system.15 The 13C-urea breath test was performed centrally using a standardized, validated method.16

Patient enrolment in the study was based on a positive result of the CLO test at baseline. H. pylori infection was confirmed if at least one of the two further test results (13C-urea breath test or histology) was positive at baseline. Eradication was considered to be successful if all 13C-urea breath test results available at the 6- and 12-month visits were H. pylori negative. If one of the results was missing, the last available 13C-urea breath test result was considered.

Statistical analysis

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Study design
  6. Patients
  7. Study protocol
  8. Symptom assessment
  9. Biopsy, histology and 13C-urea breath test
  10. Statistical analysis
  11. Results
  12. Patient population and baseline characteristics
  13. Histological findings and topographical expression of chronic gastritis
  14. Eradication of H. pylori infection
  15. Effect on dyspeptic symptoms
  16. Sub-group analysis
  17. Reflux complaints
  18. Evaluation of safety
  19. Discussion
  20. Acknowledgements
  21. References

The statistical analysis was performed using SAS version 8.2 software. Confidence intervals for H. pylori eradication rates were calculated as exact confidence intervals according to the method of Clopper and Pearson.17 The non-ulcer dyspepsia sum score was calculated as the sum of the frequency and intensity scores for dyspeptic symptoms. No weighting was performed for the calculation.

Missing values were replaced for the calculation of the non-ulcer dyspepsia sum score only. For the frequency and intensity scores, single missing values were replaced by the rounded mean of the available score values for that patient. For the assessment of efficacy, an individual's last non-ulcer dyspepsia sum score was carried forward and used as the main end-point. The same procedure was applied for the reflux sum score.

The basis for the confirmatory statistical approach defined in the protocol was the change in the selected nine dyspeptic symptoms between baseline scores and the mean status at the 6- and 12-month visits. The status for each symptom was defined by the sum of the intensity score and the frequency score. The change in this combined score was then calculated in the form of an after/before ratio, i.e. the arithmetic mean of the score at 6 and 12 months was divided by the score at baseline.

Sensitivity analyses, including the comparison of results calculated on an intention-to-treat and per protocol basis, revealed that the statistical outcome of the study was robust with respect to the handling of missing values. All decisions with respect to the selection of the safety, intention-to-treat and per protocol populations were taken before unblinding the study. The intention-to-treat population comprised all patients who took the study medication at least once and had follow-up data for safety, efficacy or both. Patients with major protocol violations were excluded from the per protocol population. The details are given in Figure 1.

image

Figure 1. Overview of patients enrolled and analysed. *In some patients, more than one reason was present for exclusion from the per protocol population. †The numbers in parentheses indicate the three treatment groups (L30AC, L15AC, LP; see text for definition).

Download figure to PowerPoint

Patient population and baseline characteristics

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Study design
  6. Patients
  7. Study protocol
  8. Symptom assessment
  9. Biopsy, histology and 13C-urea breath test
  10. Statistical analysis
  11. Results
  12. Patient population and baseline characteristics
  13. Histological findings and topographical expression of chronic gastritis
  14. Eradication of H. pylori infection
  15. Effect on dyspeptic symptoms
  16. Sub-group analysis
  17. Reflux complaints
  18. Evaluation of safety
  19. Discussion
  20. Acknowledgements
  21. References

Of the 860 patients initially screened, 800 were randomly assigned to a treatment group and were eligible for inclusion in the intention-to-treat analysis; of these, 479 were included in the per protocol analysis. The procedure for allocation to the intention-to-treat and per protocol populations is given in Figure 1, together with the reasons for exclusion.

The baseline characteristics of the three treatment groups were similar (Table 1). At study entry, the average functional dyspepsia sum scores for dyspeptic symptoms, the total gastritis scores and the frequency and intensity of individual symptoms were comparable in the three treatment groups. The treatment groups were also comparable in terms of reflux symptoms. A history of previous ulcer was recorded in 33 patients; the baseline characteristics of these patients did not differ significantly from those of the rest of the study population.

Table 1.  Baseline characteristics (intention-to-treat population)
 L30ACL15ACLP
  1. NUD, non-ulcer dyspepsia; for definition of L30AC, L15AC and LP groups, see text.

  2. Data are presented as arithmetic mean ± standard deviation (median).

Sex (male)124 (45.9%)123 (46.6%)133 (50.0%)
Age (years)46.1 ± 12.8 (48.0)46.9 ± 12.0 (48.0)45.5 ± 12.6 (46.0)
Weight (kg)72.7 ± 12.6 (71.5)72.1 ± 11.1 (72.0)73.0 ± 11.6 (73.0)
Broca index1.05 ± 0.14 (1.05)1.04 ± 0.13 (1.04)1.04 ± 0.14 (1.05)
Risk factors (%)
 Regular alcohol consumption30 (11.1%)24 (9.1%)31 (11.7%)
 Regular smoking55 (20.4%)51 (19.3%)70 (26.3%)
 Regular coffee consumption178 (65.9%)155 (58.7%)174 (65.4%)
NUD sum score at baseline [range]19.1 ± 8.019.5 ± 7.319.4 ± 7.7
 [6–49] [7–46] [7–47]
Duration of acute episode of NUD (months)2.6 ± 4.6 (1.6)2.5 ± 4.5 (1.5)2.6 ± 5.4 (1.3)
Number of previous episodes of NUD7.9 ± 10.8 (5.0)8.3 ± 12.3 (5.0)7.1 ± 8.4 (4.0)
Diagnosis of NUD established since (years)7.5 ± 8.2 (4.5)7.5 ± 7.8 (5.2)7.7 ± 8.7 (5.1)
Pre-treatment of dyspeptic symptoms80 (29.6%)82 (31.1%)79 (29.7%)
Patients with reflux score ≠ 0139 (51.5%)147 (55.7%)129 (48.5%)
Reflux sum score at baseline [range]0.8 ± 1.0 (1.0)0.8 ± 0.9 (1.0)0.8 ± 1.0 (0.0)
 [0–6] [0–5] [0–6]
Gastritis score at baseline
 Antrum9.5 ± 2.9 (10.0)9.3 ± 3.0 (10.0)9.1 ± 3.2 (10.0)
 Corpus6.8 ± 2.6 (7.0)6.9 ± 2.9 (7.0)6.6 ± 2.9 (7.0)
 Angulus8.8 ± 2.9 (9.0)8.8 ± 3.3 (9.0)8.2 ± 3.4 (9.0)

The most frequently reported main dyspeptic symptom at baseline was ‘upper abdominal complaints during the day’, which was present in 187 patients (69.3%) in the L30AC treatment group, 171 patients (64.8%) in the L15AC treatment group and 194 patients (72.9%) in the LP treatment group. Therefore, ‘upper abdominal complaints by day’ was the dominating symptom with regard to the non-ulcer dyspepsia sum score and the main symptom assessment. The second most frequent main dyspeptic symptom at baseline was ‘upper abdominal complaints during the night’, which was reported in 29 (10.7%), 33 (12.5%) and 21 (7.9%) patients in the L30AC, L15AC and LP groups, respectively. No other symptom was reported as the main dyspeptic symptom in more than 8% of patients in any treatment group. At baseline, ‘upper abdominal complaints during the day’ was reported either as a main symptom or a lesser symptom in 264 (97.8%), 255 (96.6%) and 260 (97.7%) patients in the L30AC, L15AC and LP groups, respectively.

Of the reflux symptoms, ‘heartburn’ was reported in 70 (25.9%), 72 (27.3%) and 63 (23.7%) patients in the L30AC, L15AC and LP groups, respectively. ‘Acid regurgitation’ was present in 65 (24.1%), 86 (32.6%) and 74 (27.8%) patients, respectively. ‘Difficulties in swallowing’ and ‘retrosternal pain’ were only reported very rarely by the patients.

Histological findings and topographical expression of chronic gastritis

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Study design
  6. Patients
  7. Study protocol
  8. Symptom assessment
  9. Biopsy, histology and 13C-urea breath test
  10. Statistical analysis
  11. Results
  12. Patient population and baseline characteristics
  13. Histological findings and topographical expression of chronic gastritis
  14. Eradication of H. pylori infection
  15. Effect on dyspeptic symptoms
  16. Sub-group analysis
  17. Reflux complaints
  18. Evaluation of safety
  19. Discussion
  20. Acknowledgements
  21. References

H. pylori infection at baseline was assessed on the basis of the results of the CLO test, 13C-urea breath test and histological analysis. Infection was considered to be present if at least two of these tests were positive. The histological findings were considered to be positive if the sub-score for H. pylori colonization was ≥ 1 in at least one baseline sample (antrum, corpus and angulus); findings were considered to be negative if the score was equal to zero for all available samples.

The H. pylori status at baseline was positive in 740 patients (92.5%) in total, and in 253 (93.7%), 248 (93.9%) and 239 (89.8%) patients in the L30AC, L15AC and LP treatment groups, respectively. The rest of the patients had only a positive CLO test at baseline.

A total of 618 patients included in the intention-to-treat analysis were considered to be H. pylori positive on histological grounds; 498 had antrum-predominant gastritis, 75 had corpus-predominant gastritis and 47 had an equal degree of antrum and corpus involvement.

Eradication of H. pylori infection

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Study design
  6. Patients
  7. Study protocol
  8. Symptom assessment
  9. Biopsy, histology and 13C-urea breath test
  10. Statistical analysis
  11. Results
  12. Patient population and baseline characteristics
  13. Histological findings and topographical expression of chronic gastritis
  14. Eradication of H. pylori infection
  15. Effect on dyspeptic symptoms
  16. Sub-group analysis
  17. Reflux complaints
  18. Evaluation of safety
  19. Discussion
  20. Acknowledgements
  21. References

The eradication of H. pylori infection was defined as a negative 13C-urea breath test at all available tests from 6 weeks on, where at least two tests were available. In the case of at least one positive 13C-urea breath test or less than two available tests, non-eradication was assumed.

According to the intention-to-treat analysis, after 1 year, H. pylori infection had been eradicated in 65.6% of patients in the L30AC group, 62.1% of patients in the L15AC group and 4.5% of patients in the LP group. The corresponding values in the per protocol analysis were 78.8%, 79.3% and 6.3%, respectively. The eradication rates obtained for the two triple therapies were comparable and, at about 80%, in the range to be expected for eradication therapy. Therefore, it seems justified to combine the results of these two groups and compare them with the results of the group given lansoprazole alone.

Effect on dyspeptic symptoms

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Study design
  6. Patients
  7. Study protocol
  8. Symptom assessment
  9. Biopsy, histology and 13C-urea breath test
  10. Statistical analysis
  11. Results
  12. Patient population and baseline characteristics
  13. Histological findings and topographical expression of chronic gastritis
  14. Eradication of H. pylori infection
  15. Effect on dyspeptic symptoms
  16. Sub-group analysis
  17. Reflux complaints
  18. Evaluation of safety
  19. Discussion
  20. Acknowledgements
  21. References

According to the intention-to-treat analysis, the non-ulcer dyspepsia sum score decreased considerably for all three treatment groups between the baseline and 6-week visits. Whereas this improvement was maintained or even increased until 12 months in the two groups treated with the eradication regimens, no further symptomatic improvement was observed in the group treated with lansoprazole alone (Table 2). Between baseline and 12 months, the functional dyspepsia sum score decreased to 31.4% of the baseline values in the L30AC group, 36.4% of the baseline values in the L15AC group and 41.2% of the baseline values in the LP group.

Table 2.  Non-ulcer dyspepsia (NUD) sum score according to the intention-to-treat analysis
 NUD sum scoreP value (vs. LP)
L30AC (n = 270)L15AC (n = 264)LP (n = 266)
  1. LOCF, last NUD sum score carried forward; for definition of L30AC, L15AC and LP groups, see text.

  2. Data are presented as arithmetic mean ± standard deviation (median) [range] <95% confidence interval>.

  3. * Combined antibiotic treatment groups.

At baseline19.1 ± 8.0 (17)19.5 ± 7.3 (18)19.4 ± 7.7 (18) 
 [6–49] [7–46] [7–47] 
19.3 ± 7.6 (18)*   
 [6–49]*   
At 2 weeks7.0 ± 7.0 (6)7.0 ± 7.5 (5)7.2 ± 8.1 (4) 
 [0–32] [0–34] [0–42] 
At 8 weeks5.7 ± 7.6 (3)6.0 ± 7.3 (3)7.0 ± 8.7 (4) 
 [0–52] [0–36] [0–49] 
At 6 months5.7 ± 7.7 (3)5.3 ± 7.3 (2)5.6 ± 7.7 (3) 
 [0–52] [0–37] [0–37] 
At 12 months5.4 ± 8.0 (2)6.3 ± 8.5 (3)6.2 ± 7.4 (4) 
 [0–56] [0–41] [0–32] 
LOCF6.0 ± 7.9 (4)7.1 ± 9.0 (4)8.0 ± 9.2 (4) 
 [0–56] [0–41] [0–43] 
6.5 ± 8.5 (4)*   
 [0–56]*   
Difference (LOCF − baseline)− 13.2 ± 9.4 (− 12.5)− 12.4 ± 9.5 (− 13)− 11.4 ± 9.4 (− 12)0.0068 (L30AC)
<− 14.3 | − 12.3><− 13.3 | − 11.3><− 12.4 | − 10.4>0.2054 (L15AC)
− 12.8 ± 9.5 (13)*  0.0220*

As comparable eradication rates were seen in the two antibiotic treatment groups, the two eradication therapies were considered together for the confirmatory analyses.

An analysis of covariance (ancova), using the change from baseline to end-point (last non-ulcer dyspepsia sum score carried forward) as dependent variable, treatment group as factor and baseline value as a covariate, indicated a difference between the combined antibiotic treatment groups and the LP group (P = 0.0220) and between the L30AC and LP groups (P = 0.0068), whereas no significant difference was observed between the L15AC and LP groups (P = 0.2054; Table 2). The total benefit of eradication therapy was a change in sum score of 1.4, or 12.3%.

Pertinent analyses of non-ulcer dyspepsia symptoms in the sub-group of patients with positive H. pylori status at baseline revealed similar results to those in the total population.

At last observation carried forward (LOCF) the resolution of non-ulcer dyspepsia symptoms (defined as a non-ulcer dyspepsia sum score of ≤ 1) was achieved in 97 (35.9%), 99 (37.5%) and 89 (33.5%) patients in the L30AC, L15AC and LP groups, respectively. In total, non-ulcer dyspepsia symptoms resolved in 196 patients (36.7%) receiving antibiotic treatment. No statistically relevant pair-wise differences between the treatment groups, or between the combined antibiotic treatment group and the LP group, were detected using Fisher's exact test.

In addition to the treatment arm-related analysis, we compared the effect of eradication based on H. pylori status following treatment and independent of the type of treatment. According to the intention-to-treat analysis, further amelioration after the end of treatment was observed in patients with successful eradication: the non-ulcer dyspepsia sum score decreased from 34% of the baseline values at the end of treatment visit 2 to 30% at 12 months. For patients without eradication, the development of the non-ulcer dyspepsia sum score indicated a deterioration from 36% of the baseline value after the end of treatment to 39% at 12 months. Therefore, a difference of 9% in favour of the eradication group was found after 1 year of follow-up. If treatment success is defined as the complete absence of dyspeptic symptoms or the persistence of mild symptoms (non-ulcer dyspepsia sum score of ≤ 1) 1 year after treatment (visit 5), treatment success was achieved in 41.5% of patients in the combined triple therapy groups, compared with 33.2% of patients in the group given lansoprazole alone (P = 0.042).

Sub-group analysis

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Study design
  6. Patients
  7. Study protocol
  8. Symptom assessment
  9. Biopsy, histology and 13C-urea breath test
  10. Statistical analysis
  11. Results
  12. Patient population and baseline characteristics
  13. Histological findings and topographical expression of chronic gastritis
  14. Eradication of H. pylori infection
  15. Effect on dyspeptic symptoms
  16. Sub-group analysis
  17. Reflux complaints
  18. Evaluation of safety
  19. Discussion
  20. Acknowledgements
  21. References

In order to screen for factors influencing the improvement of non-ulcer dyspepsia symptoms, a set of a posteriori sub-group analyses was performed. With respect to the severity of the symptoms at baseline, the positive effect of eradication therapy was almost doubled in patients with more severe complaints (score of > 20 at baseline) than in patients with a baseline score below 15.

A second sub-group analysis was performed with respect to the age of the patients. Amongst patients aged 36–45 years, those with successful eradication showed a mean reduction of the non-ulcer dyspepsia sum score to 21% of the baseline value, whereas the non-ulcer dyspepsia sum score of patients in this age group without eradication was 46% of the baseline value. The treatment differences in patients in other age groups at 12 months were between 2% in patients aged ≤ 35 years and 8% in patients aged between 46 and 55 years. A correlation analysis revealed that this age effect could not be explained by the duration of symptoms.

A third sub-group analysis, involving patients with severe or most severe ‘upper abdominal complaints at night’, showed an advantage for patients with successful eradication, which was about twice as high as the effect seen for all patients. Further correlation analyses and activity were performed to examine whether any other symptom characteristics of the patients or pattern of gastritis at baseline could be used to predict a response to treatment. Histological parameters, non-ulcer dyspepsia and reflux symptoms at presentation and the duration of non-ulcer dyspepsia complaints since their first appearance were not shown to be predictive.

Reflux complaints

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Study design
  6. Patients
  7. Study protocol
  8. Symptom assessment
  9. Biopsy, histology and 13C-urea breath test
  10. Statistical analysis
  11. Results
  12. Patient population and baseline characteristics
  13. Histological findings and topographical expression of chronic gastritis
  14. Eradication of H. pylori infection
  15. Effect on dyspeptic symptoms
  16. Sub-group analysis
  17. Reflux complaints
  18. Evaluation of safety
  19. Discussion
  20. Acknowledgements
  21. References

The explorative ancova (defined as for the non-ulcer dyspepsia sum score) did not indicate any difference in reflux symptoms, either between the treatment groups or between the antibiotic treatment and LP groups (Table 3). Pertinent analyses of reflux symptoms in the sub-group of patients with positive H. pylori status at baseline revealed similar results to those in the total population. The reflux symptoms deteriorated (defined as a sum score greater than that at baseline) at the end-point in 39 (14.9%), 33 (13.2%) and 36 (13.7%) patients in the L30AC, L15AC and LP groups, respectively (Figure 2). No statistically relevant pair-wise difference between the treatment groups was detected using Fisher's exact test.

Table 3.  Reflux sum score according to the intention-to-treat analysis
 Reflux sum score
L30AC (n = 270)L15AC (n = 264)LP (n = 265)
  1. For definition of L30AC, L15AC and LP groups, see text.

  2. Data are presented as arithmetic mean ± standard deviation (median) [range].

  3. * Combined antibiotic treatment groups.

At baseline0.8 ± 1.0 (1)0.8 ± 0.9 (1)0.8 ± 1.0 (0)
 [0–6] [0–5] [0–6]
0.8 ± 0.9 (1)*  
 [0–6]*  
End-point0.6 ± 1.2 (0)0.5 ± 1.0 (0)0.6 ± 1.4 (0)
 [0–9] [0–5] [0–11]
0.5 ± 1.1 (0)*  
 [0–9]*  
image

Figure 2. Change in reflux symptoms after 1 year (intention-to-treat population) (for definition of L30AC, L15AC and LP groups, see text).

Download figure to PowerPoint

Evaluation of safety

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Study design
  6. Patients
  7. Study protocol
  8. Symptom assessment
  9. Biopsy, histology and 13C-urea breath test
  10. Statistical analysis
  11. Results
  12. Patient population and baseline characteristics
  13. Histological findings and topographical expression of chronic gastritis
  14. Eradication of H. pylori infection
  15. Effect on dyspeptic symptoms
  16. Sub-group analysis
  17. Reflux complaints
  18. Evaluation of safety
  19. Discussion
  20. Acknowledgements
  21. References

Safety data were available for 800 patients and are summarized in Table 4. In total, 46 patients suffered from serious adverse events during the study. None of these events was judged to be related to the study medication by the investigator. Four patients terminated the study due to serious adverse events. More adverse events in the two triple therapy groups (90 and 64 in the L30AC and L15AC groups, respectively) were judged to be related to the study medication compared with those in the lansoprazole group (20). Gastrointestinal system disorders occurred most frequently and, within this class, diarrhoea was the most common (42, 34 and six in the L30AC, L15AC and LP groups, respectively). Other adverse event classes, often judged to be caused by the antibiotic medication, were skin and appendage disorders and, within this class, drug eruption was the most common (12, six and three, respectively).

Table 4.  Safety data
 L30ACL15ACLP
  1. For definition of L30AC, L15AC and LP groups, see text.

Number of patients270264266
Severe adverse events all not related to medication18 (6.7%)13 (4.9%)15 (5.6%)
Minor adverse events107 (39.6%)72 (27.3%)29 (10.9%)
Withdrawals11 (4.1%)10 (3.8%)21 (7.9%)

Forty-two patients (11, 10 and 21, respectively) terminated the study prematurely. This included 22 patients withdrawn due to lack of efficacy (five, five and 12, respectively).

Twenty patients were withdrawn because of adverse events and, for five patients (two, one and two, respectively), these events were assessed to be related to the study medication.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Study design
  6. Patients
  7. Study protocol
  8. Symptom assessment
  9. Biopsy, histology and 13C-urea breath test
  10. Statistical analysis
  11. Results
  12. Patient population and baseline characteristics
  13. Histological findings and topographical expression of chronic gastritis
  14. Eradication of H. pylori infection
  15. Effect on dyspeptic symptoms
  16. Sub-group analysis
  17. Reflux complaints
  18. Evaluation of safety
  19. Discussion
  20. Acknowledgements
  21. References

This large multi-centre study demonstrates that a subset of patients with H. pylori-positive functional dyspepsia obtains a significant symptomatic benefit from successful cure of their H. pylori infection. The magnitude of the benefit is small, and is best expressed by a 12% reduction in the symptom score following H. pylori eradication therapy, compared with acid inhibition with lansoprazole alone, and by complete symptom resolution in 41.5% of the antibiotic treatment group vs. 33.2% in the control group. This results in a number needed to treat of 12.

The results from this study lie approximately in the middle of the range of benefits (from + 14% to − 2%) reported in previous large well-designed clinical trials of H. pylori eradication in functional dyspepsia.18 Three of these multi-centre trials reported no or small, non-significant benefits following successful eradication.8–10 In one of these studies, however, when the healing of gastritis was considered independently of the type of treatment used in the different patient groups, 15% more patients with healed gastritis experienced symptom relief, compared with patients with persistent gastritis.9 This supports the concept of chronic inflammation being a cause of symptoms in some patients with dyspepsia. The best substantiated evidence in favour of H. pylori eradication in patients with functional dyspepsia, however, comes from a single-centre study in Glasgow which reported complete symptom resolution in 21% of patients following H. pylori eradication vs. 7% in the placebo group.11

Several aspects are worthy of consideration when interpreting the different results in the various trials.

Whereas all studies conducted in Europe, including ours, found either a significant symptomatic benefit from H. pylori eradication or at least a trend towards a benefit, two studies from North America reported a negative outcome.9, 10 Could it be that the population of patients with functional dyspepsia is characterized differently in the USA and Europe? The functional dyspepsia population may simply be more pre-selected in the American10, 19 and international multi-centre studies involving many different study centres. The covert use of aspirin or non-steroidal anti-inflammatory drugs may be an additional factor responsible for the lack of symptom response following H. pylori eradication, especially for patients included in US trials. Furthermore, the lower background prevalence of ulcer disease could explain the failure to demonstrate any benefit in American and international multi-centre studies.8, 9 The prevalence of gastric and duodenal ulcer is particularly high in the Glasgow area, with 45% of H. pylori-positive patients presenting with dyspepsia.20 In the Glasgow study, as in ours, the majority of patients presented with significant epigastric pain associated with a varying cluster of other symptoms. Ulcer-like pain, regardless of whether it is associated with an ulcer or not, has repeatedly been shown to respond best to H. pylori eradication.3, 21 In our sub-group analysis, patients with epigastric pain by day and at night displayed the most marked relief of symptoms after H. pylori eradication. Whether these patients would eventually be found to have an ulcer is irrelevant in practical terms, as long as patients with epigastric symptoms in the absence of ulcer respond positively to eradication therapy.

The relationship between H. pylori infection and symptoms is certainly more complex than just being limited to an association with ulcers. In our study, a small group of patients with no detectable ulcer at the time of inclusion, but who had symptoms and a previous ulcer history, were excluded from the final analysis, although they did not differ in terms of symptom response from the rest of the study population. The benefit of H. pylori eradication in patients with functional dyspepsia is extended by the observation that patients with functional dyspepsia and no eradication are at higher risk of developing ulcers during follow-up. In our study, seven patients in the non-eradicated group and only one in the eradicated group developed an ulcer during follow-up. This is consistent with observations made in Scotland11 and Ireland.3 Hence, the aims of therapy in patients with functional dyspepsia are to heal gastritis, with a chance of reducing or eradicating symptoms, and to prevent ulcers.

The differences between multi-centre studies, including several European studies, and the single-centre study from Glasgow, as well as our study in Germany, require particular consideration. Multi-centre studies across nations and countries are likely to be biased by the selection of patients with different presentations of functional dyspepsia to which different dietary habits and cultural backgrounds contribute by influencing the expression of dyspeptic symptoms. The selection may also not reflect the usual patients with functional dyspepsia, bearing in mind that many of the participating centres included fewer than four patients. The reason for the low numbers recruited in many participating centres may be that patients with particularly resistant symptoms were included, and these patients may not reflect the majority of those with functional dyspepsia in clinical practice. In a recent short report, Forbes and Foster concluded that, from 1000 patients considered for a study, few would ultimately meet the criteria set out in the study design for therapy evaluation in functional dyspepsia.22 The functional dyspepsia patient in routine clinical practice may therefore be quite different from those included in multi-centre studies. The variety of symptoms and personal traits of patients with functional dyspepsia render it difficult to prove a benefit of any placebo-controlled treatment, both for current antisecretory therapy,23, 24 as well as eradication therapies. Future studies may not obtain any greater benefit than that achieved in this study, due to the great diversity of patients falling under the diagnosis of functional dyspepsia. A therapeutic gain of about 10% over symptomatic therapy is apparently all that can be achieved.

Currently, the critical question is whether the benefit shown for eradication therapy in H. pylori-positive patients with functional dyspepsia is sufficient to justify therapy in all patients, or whether this strategy should be limited to a pre-selected group. We identified some features, such as a higher symptom score, more severe upper abdominal complaints at night and an age between 36 and 45 years, that may predict greater benefit from H. pylori eradication. However, these criteria are not sufficient for making a decision in individual cases. Consequently, ultimate proof of whether the therapy helps can only be obtained by treating patients and observing the individual response to therapy. Criteria based on the endoscopic picture or histological pattern of gastritis are not helpful in predicting the effect of therapy on symptoms, as we did not find a distinct response of symptoms that was dependent on the topographic phenotype of gastritis and its activity.

Objections to eradication therapy due to the frequently cited induction of gastro-oesophageal reflux disease are not supported by our study.25–27 No increase in reflux symptoms or any deterioration of such symptoms was observed following successful eradication. This is in agreement with other recent studies.28, 29 The concern about increasing microbial resistance due to eradication therapies needs to be balanced against the long-lasting benefit of short-term antibiotic use for a chronic condition such as H. pylori-associated functional dyspepsia.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Study design
  6. Patients
  7. Study protocol
  8. Symptom assessment
  9. Biopsy, histology and 13C-urea breath test
  10. Statistical analysis
  11. Results
  12. Patient population and baseline characteristics
  13. Histological findings and topographical expression of chronic gastritis
  14. Eradication of H. pylori infection
  15. Effect on dyspeptic symptoms
  16. Sub-group analysis
  17. Reflux complaints
  18. Evaluation of safety
  19. Discussion
  20. Acknowledgements
  21. References

This study was funded by Takeda Pharma GmbH, Aachen, Germany.

We thank the following investigators who participated in this study:

M. Abdel-Qader (Winsen); Dr H. Berling (Attendorn); PD Dr Dr D. Bojanovski (Hannover); Dr W. Borngräber (Hamburg); Dr H. Brinkhoff (Stuhr); Dr B. Buck (Ertingen); Dr J. Christ (Worbis); Dr M. Dudek (Düsseldorf); Dr W. Dübel (Berlin); Professor Dr W. Fischbach (Aschaffenburg); Dr A. Fischer (Marienhafen); Dr M. Frevel (Bitburg); Drs Hebbeln/Martens (Itzehoe); Professor Dr W. Heldwein (München); Dr S. Kaspari (Lüneburg); Dr R. Kobes (Werdau); Dr B. Kostrewa (Flensburg); Drs Krey/Walther (Norderstedt); Dr R. Küchler (Freudenstadt); Professor Dr P. Malfertheiner (Magdeburg); Dr E. Meier (Amberg); Professor Dr J. Mössner (Leipzig); Dr W. Poeplau (Amberg); Dr P. Prause (Göttingen); Dr I. H. Rehmann (Lippstadt); Dr G. Rosprich (Saarbrücken); Dr J. Sauter (Wangen); Dr M. Schumacher (Wolmirstedt); Dr A. Sommer (Köln); Dr G. Tangerding (Wangen); Dr R. Vogt (Mannheim); Dr F. Weitendorff (Bad Soden); Dr K. Ziegler (Berlin); Dr F. Eller (Neuburg); Professor Dr W. Zoller (München); Professor Dr Layer/Dr Günther (Hamburg); Dr H. Koch (Gaggenau); Dr P. Priebe (Gaggenau); T. S. Kammermeier (Bogen); Drs Bretzke/Elsel (Zwickau); Drs Döppenschmidt/Westphal (Germersheim); Dr U. Groß (Neuwied); Dr E. Hommel (Stuttgart); Dr C. Klein (Künzing); Dr M. Neumeyer (Oldenburg); Dr K. Dietrich (Saarbrücken); Dr Zöllner (Dippoldiswalde); Drs Kocjan/Müser (Lüdenscheid); Dr A. Kühn (Cottbus); Dr Düker (Meiningen); Dr Hartl (Nittenau); Dr B. Marowski (Berlin); Dr W. Resch (Landshut); Dr J. Hagel (Schwabach); J. Kreutzer (Dülmen); Dr M. C. Cüppers (Duisburg); Dr G. Scholz (Offenbach); Dr K. Franke (Schloß Holte-Stukenbrock); Dr M. Kahl (Hochheim); Dr O. Mickisch (Mannheim); Dr D. Klein (Köln); Dr B. Jonas (Duisburg); Dr P. Krapp (Sigmaringen); Dr P. Levi (Ahaus); Dr S. Orlemann (Rödermark); Dr B. Stölzle (Lindau); Dr E. Schütz (Regensburg); Dr C. Bauknecht (Rottweil); Dr Mucha/Vintila (Schwarzenbek); Dr M. Schirmer (Oberursel); Dr W. Dillmann (Aschaffenburg); Dr A. Lütke (Koblenz); Dr G. P. Müller (Mönchengladbach); Dr G. Laudage (Gelsenkirchen); Dr V. Lachmann (Bayreuth); Dr V. Schmid (Fürstenzell); Dr H. Toluipur (Schiffweiler); Dr D. Kiss (Wachtendonk); Dr K. Hehemann (Beckum); PD Dr Stölzel (Döbeln); Dr E. Gozdowsky (Berlin); Dr D. Vogt (Frechen); Dr M. Zimmermann (Konradsreuth); Dr B. Dir (Hechingen); Dr Zeuss (Erlangen).

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Study design
  6. Patients
  7. Study protocol
  8. Symptom assessment
  9. Biopsy, histology and 13C-urea breath test
  10. Statistical analysis
  11. Results
  12. Patient population and baseline characteristics
  13. Histological findings and topographical expression of chronic gastritis
  14. Eradication of H. pylori infection
  15. Effect on dyspeptic symptoms
  16. Sub-group analysis
  17. Reflux complaints
  18. Evaluation of safety
  19. Discussion
  20. Acknowledgements
  21. References
  • 1
    Jakkimainen RL, Boyle E, Tudiver F. Is Helicobacter associated with non ulcer dyspepsia and will eradication improve symptoms? A meta-analysis. Br Med J 1999; 0: 10404.
  • 2
    Talley NJ. A critique of therapeutic trials in Helicobacter pylori positive dyspepsia. Gastroenterology 1994; 106: 117483.
  • 3
    Gilvarry J, Buckley MJ, Beattie S, Hamilton H, O'Morain C. Eradication of H. pylori affects symptoms in non-ulcer dyspepsia. Scand J Gastroenterol 1997; 32(0): 53540.
  • 4
    Laheij RJF, Jansen JBMJ, Vandelisdonk EH, Verbeek ALM. Symptom improvement through eradication of H. pylori with non-ulcer dyspepsia. Aliment Pharmacol Ther 1996; 10: 84350.
  • 5
    Soo S, Moayyedi P, Deeks J, et al. Helicobacter pylori eradication for non-ulcer dyspepsia (full Cochrane Review). Oxford Update Software 2000.
  • 6
    Malfertheiner P (for European Helicobacter Study Group). Current European concepts in the management of Helicobacter pylori infection — the Maastricht Consensus Report. Gut 1997; 41(1): 813.
  • 7
    Malfertheiner P, Megraud F, O'Morain C, et al. Current concepts in the management of H. pylori infection — the Maastricht 2-2000 Consensus Report. Aliment Pharmacol Ther 2002; 16: 16780.
  • 8
    Blum AL, Talley NJ, O'Morain C, et al. Lack of effect of treating H. pylori infection in patients with nonulcer dyspepsia. N Engl J Med 1998; 339: 187581.
  • 9
    Talley NJ, Janssens J, Lauritsen K, Tacz I, Bolling-Sternevald E. Eradication of H. pylori in functional dyspepsia: randomised double blind placebo controlled trial with 12 months' follow up. Br Med J 1999; 318: 8337.
  • 10
    Talley NJ, Vakil N, Ballard D, Fennerty MB. Absence of benefit of eradicating H. pylori in patients with non-ulcer dyspepsia. N Engl J Med 1999; 341: 110611.
  • 11
    McColl KEL, Murray L, El-Omar E, et al. Symptomatic benefit from eradicating H. pylori infection in patients with non-ulcer dyspepsia. N Engl J Med 1998; 339(26): 186974.
  • 12
    Danesh J, Pounder RE. Eradication of Helicobacter pylori and non-ulcer dyspepsia. Lancet 2000; 4(355): 7667.
  • 13
    Moayyedi P, Soo S, Deeks J, et al. on behalf of the Dyspepsia Review Group. Symptomatic review and economic evaluation of H. pylori eradication treatment for non-ulcer dyspepsia. Br Med J 2000; 321: 65964.
  • 14
    Laine L, Schoenfeldt P, Fennerty MB. Therapy for H. pylori in patients with non-ulcer dyspepsia. Ann Intern Med 2001; 134: 3619.
  • 15
    Dixon MF, Genta RM, Yardley JH, Correa P. Classification and grading of gastritis. The updated Sydney System. Am J Surg Pathol 1996; 20: 116181.
  • 16
    Leodolter A, Dominguez-Munoz E, Von Arnim U, Kahl S, Peitz U, Malfertheiner P. Validity of a modified 13C-urea breath test for pre- and posttreatment diagnosis of Helicobacter pylori infection in the routine clinical setting. Am J Gastroenterol 1999; 94(8): 21004.
    Direct Link:
  • 17
    Clopper CP, Pearson ES. The use of confidence or fiducial limits illustrated in the case of the binominal. Biometrika 1934; 26: 40413.
  • 18
    McColl K, Malfertheiner P. Helicobacter pylori and functional dyspepsia. Curr Opin Gastroenterol 2000; 16(Suppl.): S2932.
  • 19
    Greenberg PD, Cello JP. Lack of effect of treatment for H. pylori on symptoms of nonulcer dyspepsia. Arch Intern Med 1999; 159: 22838.
  • 20
    McColl K, El-Nujumi A, Murray L, et al. The Helicobacter pylori breath test: a surrogate marker for peptic ulcer disease in dyspeptic patients. Gut 1997; 40(3): 3026.
  • 21
    Trespi E, Broglia F, Villani L, Luinetti O, Fiocca R, Solcia E. Distinct profiles of gastritis in dyspepsia subgroups. Their different clinical responses to gastritis healing after H. pylori eradication. Scand J Gastroenterol 1994; 29(10): 8848.
  • 22
    Forbes GM, Foster NM. Patient recruitment in H. pylori drug trials. Lancet 1998; 352: 10712.
  • 23
    Talley NJ, Janssens J, Lauritsen K, Racz I, Bolling-Sternevald E. Eradication of H. pylori in functional dyspepsia; randomised double blind placebo controlled trial with 12 months' follow up. The Optimal Regimen Cures Helicobacter Induced Dyspepsia (ORCHID) Study Group. Br Med J 1999; 318: 8337.
  • 24
    Finnicy JS, Kinnersley N, Hughes M, et al. Metaanalysis of antisecretory and gastrokinetic compounds in functional dyspepsia. J Clin Gastroenterol 1998; 26: 31220.
  • 25
    Labenz J, Blutm AL, Bayerdörffer E, et al. Curing Helicobacter pylori infection in patients with duodenal ulcer disease may provoke reflux esophagitis. Gastroenterology 1997; 112: 14427.
  • 26
    Fallone CA, Barkun AN, Friedman G, et al. Is Helicobacter pylori eradication associated with gastroesophageal reflux disease? Am J Gastroenterol 2000; 95: 91420.
    Direct Link:
  • 27
    Hamada H, Haruma K, Mihara M, et al. High incidence of reflux oesophagitis after eradication therapy for Helicobacter pylori: impacts of hiatal hernia and corpus gastritis. Aliment Pharmacol Ther 2000; 14: 72935.
  • 28
    Malfertheiner P, Dent J, Zeijlon L, et al. Impact of Helicobacter pylori eradication on heartburn in patients with gastric or duodenal ulcer disease – results from a randomized trial programme. Aliment Pharmacol Ther 2002; 16: 143142.
  • 29
    Vakil N, Hahn B, McSorley D. Recurrent symptoms and gastro-oesophageal reflux disease in patients with duodenal ulcer treated for H. pylori infection. Aliment Pharmacol Ther 2000; 14(1): 4551.