Article Text

H2 antagonists for gastric cancer: small numbers are not beautiful
  1. Queen Elizabeth Hospital, Edgbaston, Birmingham B15 2TH, UK

Statistics from

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.

See article on page 17

The study reported on page 17 illustrates some of the frustrations of clinical trials. The rationale for undertaking a prospective trial of the use of ranitidine in gastric cancer was appropriate. A previous study from Tonnesen et al 1 had demonstrated a survival benefit using cimetidine in patients with gastric cancer. This group of authors argued that cimetidine had an immunological effect which included inhibition of T suppresser activity and increased interleukin 2 production by lymphocytes.2 3

Ranitidine has a similar effect: the Yorkshire GI tumour group began this study in 1989 to evaluate the potential effect of this drug in a controlled trial treating all stages of gastric cancer. Between 1989 and 1995, 222 patients were recruited. This illustrates one of the frustrations of running trials in gastric cancer. This is a recruitment rate of less than 50 patients per year and is a feature many gastric cancer trials have in common.4 5 The patients were randomised to receive either ranitidine or placebo. The randomisation has worked satisfactorily: the number of patients distributed between the placebo and active treatment groups was equal. The most significant prognostic factor in gastric cancer is stage and the randomisation resulted in an higher proportion of patients in tumour stages I and II randomised to the treatment group compared with the control group (29%v 21%). It is the patients in stages I and II disease that have the best long term survival. This distribution has had the result of loading the placebo group with those of worse prognosis: stages III and IVa and b. The mean length of follow up was 185 days. There was no statistical benefit from treatment. The figures suggest that patients receiving ranitidine may be doing better than those in the placebo group. The authors honestly report that this is not a statistical benefit and this highlights the problem of a randomised controlled trial with insufficient numbers. This is acknowledged by the authors in their discussion.

Previous studies using cytotoxic chemotherapy have generally looked for big effects (15 to 20% improvements). It is felt that large benefits should be achieved as many cytotoxic regimes have significant side effects, or reduce the quality of life in patients with advanced disease. However, this is not the case with ranitidine. This study demonstrates it has been well tolerated and it would be in order to look for a small effect since the quality of life is unlikely to be diminished by taking H2 receptor antagonists: such a study would require 2000 patients. Other studies have been undertaken: the British Stomach Cancer Group have now recruited over 400 patients to a trial of cimetidine. However, even analysis of the combined studies is unlikely to have enough patients to be able to demonstrate a small benefit. The other factor that cannot be stressed too strongly is our inability to recruit significant numbers of patients with gastric cancer into trials.

Hopefully the reorganisation of cancer services and the identification of clinicians specifically interested in gastric cancer will allow valuable studies to be undertaken over a reasonable time period.


Linked Articles