Article Text


H pylori and functional dyspepsia
  1. University of Sydney
  2. Clinical Sciences Building
  3. Nepean Hospital, PO Box 63
  4. Penrith NSW 2751, Australia
  5. ntalley{at}
  1. K E L McCOLL
  1. Department of Medicine and Therapeutics
  2. Gardiner Institute, Western Infirmary, Glasgow, UK
  3. K.E.L.McColl{at}

Statistics from

Editor,—I read with interest the results reported by Blum et al(

) and the editorial which accompanied this article by McColl (

). Professor McColl argues that the greater response to acid suppression observed in theHelicobacter pylori positive compared with the negative subjects in the Blum trial resulted from a lack of blinding of the investigators and the presence of an incentive for categorising H pylori positive subjects as non-responders. In support of this argument, he cites pooled data from two large randomised controlled trials (BOND and OPERA)1comparing omeprazole 20 mg, 10 mg, and placebo where overall no significant relationship between H pyloristatus and symptom relief was observed. However, there may be other explanations for the results observed by Blum et al.

We have previously reported in pooled data from the BOND and OPERA trials that in the H pylori positive group, 43% on omeprazole 20 mg had complete relief of dyspeptic symptoms (by intention to treat) compared with 35% in H pylori negative patients (versus 34% and 24% on placebo, respectively).1 Hence there was a small therapeutic gain observed in H pylori positive patients receiving omeprazole 20 mg versus those who were H pylori negative although the differences were not statistically significant. In the pooled data from the BOND and OPERA studies, there was also an association between symptom relief and the a priori identification of dyspepsia subgroups based on the predominant symptom.1 Moreover, there were differences observed between H pylori infected and uninfected cases, albeit non-significant. Those with ulcer-like dyspepsia who wereH pylori positive for example, had complete symptom relief on omeprazole 20 mg in 47% of cases versus 31% on placebo. On the other hand, in those who were H pylori negative with ulcer-like dyspepsia, 35% had complete symptom relief on omeprazole 20 mg versus 24% on placebo. In contrast, no benefit of omeprazole versus placebo was identified in dysmotility-like dyspepsia.1 Dyspepsia subgroups were not considered in the Blum trial. Presumably the decision to conduct statistical analyses separately in H pyloripositive and negative patients was exploratory in the Blum trial, as randomisation was not performed stratifying for H pylori status. Although this is a valid approach, it is conceivable that the distribution of potential responders to acid suppression in the H pylori positive and negative arms (for example, ulcer-like dyspepsia) was quite dissimilar.

Professor McColl suggests that the same therapeutic gain achieved with potent acid suppression may be achieved more permanently by eradication of H pylori. However, this conclusion remains controversial. Mayoyedi et alrecently published a meta-analysis of 12 trials reporting that the risk reduction with H pylori eradication therapy from the pooled data was 9% (95% confidence interval (CI) 4–14%).2 However, these results are in conflict with a similarly high quality meta-analysis of 10 trials which failed to detect a significant benefit of H pylorieradication (odds ratio for success 1.3, 95% CI 0.89–1.189).3 The positive and negative trial results in non-ulcer dyspepsia may be reconcilable if H pylori eradication only benefits those with hidden peptic ulcer disease or those predisposed to subsequently developing peptic ulceration which in turn is likely to vary across different populations. If this hypothesis is correct, then eradication therapy in a subset of patients with non-ulcer dyspepsia would be indicated (as the benefits should then outweigh the risks). In the absence of predictors which usefully identify a subset who will achieve symptom relief with eradication therapy, the decision to treat the infection is by necessity on a case by case basis.


Conflict of interest: Professor Talley is a consultant for Astra Zeneca, Janssen, TAP, Pharmacia, Lederle, and GlaxoSmithKline.



Editor,—I am very confused by Professor Talley's comments regarding the symptomatic benefit from omeprazole inHelicobacter pylori positive versus negative subjects with functional dyspepsia. In his original paper,1-1 he states in the results section that “relief of dyspepsia in the active treatment arms was not significantly different in the H pylori infected and uninfected patients” and in the abstract that “symptom relief was similar inH pylori positive and negative cases”. I was therefore very surprised to read Professor Talley's letter in which referring to the same paper he states that “there was a small therapeutic gain observed in H pyloripositive patients versus those who were H pylori negative”. He goes on to state that “in theH pylori positive group, 43% on omeprazole 20 mg had complete resolution of dyspepsia symptoms compared with 35% in H pylori negative patients”. However,therapeutic gain compares the benefit of theactive treatment versus the placebo treatment. The therapeutic gain in the H pylori positive group on omeprazole is 9% and the therapeutic gain in the H pylori negative patients is 11% (fig 1-1). Therefore, contrary to the comments made in his above letter, Professor Talley's original study showed that the therapeutic gain for omeprazole is similar or less in H pylori positive versus negative patients with functional dyspepsia.

Figure 1-1

This figure is derived from Professor Talley's original paper to which he refers, comparing the effects of omeprazole 20 mg versus placebo in Helicobacter pylori positive and negative subjects with functional dyspepsia.1-1 It shows that the therapeutic gain from omeprazole is similar or less in H pylori positive versus negative patients. Contrary to Professor Talley's letter, the therapeutic gain observed in his original study is certainly not greater in H pylori positive subjects.

Professor Talley also challenges the therapeutic gain achieved with eradication of H pylori infection in functional dyspepsia. There are only two recent high quality meta-analyses of H pylori eradication in functional dyspepsia published as full papers.1-2 1-3 These both showed unequivocal benefit of the active treatment over placebo. The other meta-analysis Professor Talley refers to has only appeared in abstract form and it is therefore difficult to comment on its quality.1-4 In addition, it did not include the most recent studies. None the less, even this study showed a strong trend in favour of eradication therapy over placebo (odds ratio for success 1.2, 95% CI 0.98–1.6). Hence all of the studies and meta-analyses indicate that eradicating H pylori infection is superior to placebo in non-ulcer dyspepsia, producing a therapeutic gain of about 10%. Although the magnitude of the benefit may not be large, it is equivalent to the only other treatment which is effective for this common disorder, namely proton pump inhibitor therapy.H pylori eradication treatment however has the advantage of producing this benefit by a one off course of treatment rather than requiring long term maintenance therapy.


  1. 1-1.
  2. 1-2.
  3. 1-3.
  4. 1-4.
View Abstract

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.