We read with interest the paper by Mera and colleagues (Gut 2005;54:1536–40) on the effect of eradicating Helicobacter pylori infection on precancerous gastric lesions. However, we have concern regarding the extent to which the limited data provided in the paper support the authors' conclusions of regression of atrophy and intestinal metaplasia following H pylori eradication.
The main outcome reported was the average histological score. This is an arbitrary ordinal scale which is based on the presence of superficial gastritis, atrophic gastritis, intestinal metaplasia, dysplasia, or cancer as the most advanced lesion found in the four gastric biopsies performed at each time point. The average score was 3.77 at baseline, and following eradication of the infection decrease to 3.18 after 12 years. The extent to which resolution of the different components of the average histological score contributed to the fall in the score is not made clear in the paper. However, the authors do present data showing that the acute polymorphonuclear cell infiltration fully resolved and the chronic mononuclear cell infiltration partially resolved. The magnitude of the resolution of the inflammatory infiltrate would appear to be adequate to explain the fall in the average histological score without any associated resolution of the intestinal metaplasia or atrophy. The results section does not give information on the score for atrophy or intestinal metaplasia at baseline versus 12 years.
In the discussion section, comment is made to changes in atrophy and intestinal metaplasia. However, there are insufficient data to draw any meaningful conclusions or perform statistical analysis. In addition, there is clearly an error in the data provided is the discussion section as 70/182 is called 20%.
We would be grateful if the authors would provide individual metaplasia and atrophy scores at baseline and at 12 years in the two groups in order to allow independent analysis and interpretation of this important study. This is important as these are the lesions most strongly associated with cancer risk.
Our interpretation of the limited data made available in the current version of the paper is that there is evidence of early complete resolution of acute polymorphonuclear infiltration and partial resolution of chronic mononuclear cell infiltration but no convincing evidence presented of resolution of intestinal metaplasia or atrophy. The findings are therefore consistent with a number of previous studies showing no evidence of reversal of the important precancerous lesions. We also note that during the 12 years of follow up, five cancers occurred in the H pylori treatment group and four in the non‐treatment group. Contrary to the conclusions of the authors, the benefit of eradicating H pylori infection in patients with advanced gastritis as a means of preventing cancer remains far from clear.
Footnotes
Conflict of interest: None declared.