Article Text

Download PDFPDF

Proton pump inhibitors and gastric neoplasia
  1. Helge L Waldum,
  2. Gunnar Qvigstad
  1. Department of Gastroenterology and Hepatology, St Olavs Hospital, Trondheim University Hospital, Trondheim, Norway
  1. Correspondence to:
    Professor H L Waldum
    Department of Gastroenterology and Hepatology, St Olavs Hospital, Trondheim University Hospital, Olav Kyrres gt., N-7006 Trondheim, Norway; helge.waldum{at}
  1. E J Kuipers

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.

We read the leading article by Kuipers (Gut 2006;55:1217–21) with great interest. The conclusion that proton pump inhibitor (PPI) treatment does not increase the risk of gastric neoplasia is reassuring. The overview by Kuipers, however, lacks a few important points:

  1. Hypergastrinaemia. The maximal trophic effect of gastrin is reached at concentrations lower than that realised by most clinicians, both in rats and in humans.1 A concentration of 400 ng/ml or 200 pM induces nearly maximal effect.

  2. Hypergastrinaemia and gastric carcinoids. Kuipers claims that prolonged hypergastrinaemia leads to enterochromaffin-like (ECL) cell carcinoids only in rats.2 This is not correct. Mice also develop gastric carcinoids when treated with insurmountable histamine 2 antagonists.3 PPIs have low potency in this species, which is the cause of the widespread misunderstanding that hypergastrinaemia in mice does not induce ECL cell carcinoids. Moreover, long term treatment of dogs with PPI was reported not to induce any changes in the gastric mucosa.4 Unfortunately, the dose used was not sufficient to induce hypoacidity and accompanying hypergastrinaemia. Even in humans gastric endocrine tumours have been described after PPI treatment,1 although, surprisingly, a more detailed report on these patients does not exist. It is, therefore, likely that not only rats, but every species, including humans, exposed to long-term hypergastrinaemia is at risk of developing ECL cell carcinoids.

  3. Role of the ECL cell in gastric cancer. Kuipers does not discuss the role of the ECL cell in gastric carcinogenesis. It is now apparent that a higher proportion of gastric carcinomas than previously realised are probably neuroendocrine carcinomas.1,5 Moreover, most of these neuroendocrine carcinomas develop from the predominant neuroendocrine cell in the stomach, the ECL cell.1 In addition, gastric ECL cell carcinoids secondary to hypergastrinaemia may progress to neuroendocrine carcinomas.1 The gastric carcinomas developing in patients with pernicious anaemia most often originate in the oxyntic area, indicating that the lack of acidity is not the most important factor in gastric carcinogenesis in these patients, and that the ECL cell is probably the cell of origin in these carcinomas.1 It should also be recalled that the carcinogenic effect of Helicobacter pylori infection may be related to hypergastrinaemia.1

To conclude, PPI treatment induces hypergastrinaemia. Long-term hypergastrinaemia predisposes to ECL cell tumours, such as ECL cell carcinoids and ECL cell neuroendocrine carcinomas. Long-term PPI treatment might, therefore, increase the risk of gastric cancer. It should be remembered that carcinogenesis is a long-term process often taking decades.


Author’s response

In theory, profound acid suppression can promote the development of gastric neoplasia via various pathways, including the trophic effect of hypergastrinaemia on the epithelial lining, stimulation of enterochromaffin-like (ECL) cells leading to formation of carcinoids and neuroendocrine carcinomas, promotion of bacterial overgrowth and induction of a chronic active Helicobacter pylori pan-gastritis pattern. Waldum et al focus on the theoretical implications of hypergastrinaemia and ECL cell stimulation, a relevant hypothesis that has often been discussed but remains speculative. Indeed, as Waldum points out, data suggest that the incidence of gastric carcinoids has been rising. This has been noted over the past 50 years, in particular since the early 1970s. This primarily correlates with the increase in gastric diagnostic procedures, firstly using x rays and later endoscopy with biopsy sampling, rather than the more recent introduction of proton pump inhibitors (PPIs).1 Nevertheless, gastric carcinoids remain rare, with an incidence of 0.1–0.2 per 100 000 population per year,2 without an obvious association with PPI use. I mentioned previously that carcinoids have never been described in long-term PPI users, but I was mistaken. Dr Manson from the UK had reported a patient who was diagnosed with a primary gastric carcinoid after having taken 40 mg omeprazole for 4 years for reflux disease.3 Waldum refers to another case; this was, however, not a long-term PPI user, but a patient with recurrent peptic ulcer who was diagnosed with a gastric carcinoid after intermittently having been treated with famotidine, omeprazole and lansoprazole for a total of 5 months over a 5-year period (ie, <10% of the time).4

ECL cells may also play a role in the formation of gastric adenocarcinoma. In fact, it has been suggested that about 10% of gastric adenocarcinomas, in particular of the diffuse type, may have a neuroendocrine component.5 In the US, a rise in the incidence of the diffuse type of gastric cancer between 1973 and 2000 has been noted, from 0.3 to 1.8 per 100 000 per annum.6 It is unknown whether acid-suppressive drugs have made any contribution to this process. As pointed out previously, data on gastric cancer formation in long-term PPI users have not shown any increased incidence of gastric cancer (diffuse and intestinal) at all within the first 18 years after the introduction of PPI.7

In conclusion, in contrast with the statement of Waldum et al, there is, at present, no evidence that promotion of ECL cell growth is a clinically important issue in long-term PPI users, for the formation of either gastric carcinoids or neuroendocrine adenocarcinomas. I fully agree though with the implicit proposition that monitoring of large cohorts of long-term PPI users as well as reporting of any specific cases such as that by Dawson et al3 remains relevant, because only in that way can we eventually settle issues such as raised in this discussion.



  • Competing interests: None.


  • Competing interests: In the past 5 years, EJK has received research funding from Altana Pharma, AstraZeneca and Janssen Pharmaceutics, and consulting fees from AstraZeneca.

Linked Articles

  • Leading article
    E J Kuipers
  • Digest
    Robin Spiller Alastair Watson