Proton Pump Inhibitor Use, Hypergastrinemia, and Gastric Carcinoids-What Is the Relationship?

Neuroendocrine tumors (NETs) throughout the body are the focus of much current interest. Most occur in the gastrointestinal tract and have shown a major increase in incidence over the past 30 years, roughly paralleling the world-wide increase in the use of proton pump inhibitor (PPI) drugs. The greatest rise has occurred in gastric carcinoids (g-NETs) arising from enterochromaffin-like (ECL) cells. These tumors are long known to occur in auto-immune chronic atrophic gastritis (CAG) and Zollinger-Ellison syndrome (ZES), with or without multiple endocrine neoplasia type-1 (MEN-1), but the incidences of these conditions do not appear to have increased over the same time period. Common to these disease states is persistent hypergastrinemia, generally accepted as causing g-NETs in CAG and ZES, and postulated as having similar tumorigenic effects in PPI users. In efforts to study the increase in their occurrence, g-NETs have been classified in a number of discussed ways into different grades that differ in their incidence and apparent pathogenesis. Based on a large amount of experimental data, tumorigenesis is mediated by gastrin's effects on the CCK2R-receptor on ECL-cells that in turn leads to hyperplasia, dysplasia, and finally neoplasia. However, in all three conditions, the extent of response of ECL-cells to gastrin is modified by a number of genetic influences and other underlying risk factors, and by the duration of exposure to the hormonal influence. Data relating to trophic effects of hypergastrinemia due to PPI use in humans are reviewed and, in an attached Appendix A, all 11 reports of g-NETs that occurred in long-term PPI users in the absence of CAG or ZES are summarized. Mention of additional suspected cases reported elsewhere are also listed. Furthermore, the risk in humans may be affected by the presence of underlying conditions or genetic factors, including their PPI-metabolizer phenotype, with slow metabolizers likely at increased risk. Other problems in estimating the true incidence of g-NETs are discussed, relating to non-reporting of small tumors and failure of the Surveillance, Epidemiology, and End Results Program (SEER) and other databases, to capture small tumors or those not accorded a T1 rating. Overall, it appears likely that the true incidence of g-NETs may be seriously underestimated: the possibility that hypergastrinemia also affects tumorigenesis in additional gastrointestinal sites or in tumors in other organ systems is briefly examined. Overall, the risk of developing a g-NET appears greatest in patients who are more than 10 years on drug and on higher doses: those affected by chronic H. pylori gastritis and/or consequent gastric atrophy may also be at increased risk. While the overall risk of g-NETs induced by PPI therapy is undoubtedly low, it is real: this necessitates caution in using PPI therapy for long periods of time, particularly when initiated in young subjects.

Potential risks associated with long-term PPI use.

Several adverse effects associated with long-term PPI therapy have been reported. These potential risks or side effects have included hypergastrinemia, carcinoid formation, devel- opment of gastric cancer, bacterial overgrowth, Clostridium difficile infection, pneumonia, bone fracture and malabsorption of vitamin B12 and iron. I explain each of these adverse events based on the guideline for GERD in both Japan and US, multiple meta-analyses and systematic review previously published.

Does the dose matter? Antiproliferative efficacy and toxicity of everolimus in patients with neuroendocrine tumors - Experiences from a tertiary referral center.

The mTOR-inhibitor everolimus has been approved for the treatment of advanced neuroendocrine tumors (NETs) but is associated with relevant toxicities in clinical practice. Hence, optimal treatment sequencing and the impact of dose reductions have yet to be clarified. This retrospective analysis assessed patients with advanced, well-differentiated NET treated with everolimus at the Medical University of Vienna. The primary objective was to evaluate the efficacy of everolimus in a real-world cohort. A total of 52 patients treated with everolimus for advanced NET grade 1 (G1) or G2 (or typical or atypical carcinoid) 2010-2021 were included in this analysis. The most common sites of origin were pancreas (44%) and lung (29%). The initial dose was decided by the treating physician based on clinical assessment and 25 patients (48%) each were started at 10 mg/day and 5 mg/day. Median progression-free survival (PFS) following everolimus in the overall cohort was 9.8 months (95% CI: 4.3-15.3), with a statistically significant PFS difference (p = .03) between NET G1/typical carcinoids (42.9 months) and NET G2/atypical carcinoids (8.9 months). PFS was numerically but not significantly shorter in patients treated with a reduced dose (7.5 months vs. 12.4 months, p = .359). Even in this mixed full/half dose cohort, 93% developed treatment-related side effects (mostly grade I, no grade IV), 63% had dose reductions or interruptions, and five stopped due to toxicity. Median survival following treatment was 40.9 months (95% CI: 21.5-60.3) and no difference with regard to dosing was observed (p = .517). These data from an unselected patient cohort show long-term outcomes similar to those reported in the pivotal studies. Comparing everolimus starting dose, median PFS did not significantly differ for patients treated at a lower dose. While this finding is limited by the sample size and warrants prospective verification, initiating therapy at a reduced dose might be practicable and safe in a distinct subset of patients.

Ultrastructure of ECL cells in Mastomys after long-term treatment with H2 receptor antagonist loxtidine.

Gastric ECL-cell hyperplasia and carcinoids (ECLoma) develop after 1 year in rats treated with omeprazole or 2 months in Mastomys treated with loxtidine. The aim of this study was to examine the ultrastructure of ECL cells in Mastomys after loxtidine treatment with an attempt to evaluate whether an impairment of autophagy was involved in the tumorigenesis. Mastomys were given loxtidine for 8 or 27 weeks. Morphological analysis of ECL cells showed that (1) cell size was not increased after 8 or 27 weeks; (2) secretory vesicles, a hallmark feature of welldifferentiated ECL cells, were unchanged after 8 weeks but reduced after 27 weeks; (3) granules were reduced after 8 or 27 weeks; (4) microvesicles were unchanged after the treatment; and (5) vacuoles and lipofuscin bodies were found occasionally after 8 weeks but not at 27 weeks. In addition, the appearance of ECL-cell ultrastructure differed between loxtidine-treated Mastomys and rats treated with omeprazole or subjected to antrectomy, but was similar between Mastomys treated with loxtidine for 27 weeks and mice deficient in CCK(2) receptor. We suggest that the ultrastructure of ECL cells in Mastomys after long-term treatment with loxtidine displayed an impaired formation of vacuoles and lipofuscin bodies, markers of the autophagic pathway.

Adverse effects of long-term proton pump inhibitor therapy.

Proton pump inhibitors have an excellent safety profile and have become one of the most commonly prescribed class of drugs in primary and specialty care. Long-term, sometimes lifetime, use is becoming increasingly common, often without appropriate indications. This paper is a detailed review of the current evidence on this important topic, focusing on the potential adverse effects of long-term proton pump inhibitor use that have generated the greatest concern: B12 deficiency; iron deficiency; hypomagnesemia; increased susceptibility to pneumonia, enteric infections, and fractures; hypergastrinemia and cancer; drug interactions; and birth defects. We explain the pathophysiological mechanisms that may underlie each of these relationships, review the existing evidence, and discuss implications for clinical management. The benefits of proton pump inhibitor use outweigh its risks in most patients. Elderly, malnourished, immune-compromised, chronically ill, and osteoporotic patients theoretically could be at increased risk from long-term therapy.

Atypical carcinoid tumor of the thymus with ectopic ACTH production developed during the course of etanercept treatment--case report.

Ever since the introduction of anti-tumor necrosis factor (TNF) agents, concerns have been raised regarding their potential for developing malignancy. We report the development of thymic atypical carcinoid tumor 9 months after the initiation of etanercept therapy in a patient having refractory spondylarthritis. This case indicates the need for following large cohorts of patients receiving anti-TNF agents to address the long-term effect of these agents on malignancies.

Enhanced effect of gastrin on rat stomach carcinogenesis induced by N-methyl-N'-nitro-N-nitrosoguanidine.

This study deals with the effects of gastrin on the incidence of gastric tumors in rats induced by N-methyl-N'-nitro-N-nitrosoguanidine. Inbred Basel-Wistar rats received N-methyl-N'-nitro-N-nitrosoguanidine in drinking water (50 micrograms/ml for 32 weeks) in order to produce gastric carcinoids. A treatment with s.c. injection of pentagastrin (300 micrograms/kg, once daily for 4 weeks) was started at the beginning of N-methyl-N'-nitro-N-nitrosoguanidine treatment simultaneously, on the 4th, 8th, 16th, and 32nd week after start of N-methyl-N'-nitro-N-nitrosoguanidine treatment, respectively. At autopsy, from the 55th to 60th week after start of the experiment, only in the eighth-week group of gastrin-treated rats was the incidence of gastric carcinoid significantly higher than in the gastrin-untreated group of rats receiving N-methyl-N'-nitro-N-nitrosoguanidine alone. The incidence of adenocarcinoma in the glandular stomach also was high only in the fourth-week group of gastrin-treated rats. However, these effects could not be seen in other gastrin-treated or untreated groups of rats. The data suggest that gastrin treatment in the early stage of rat stomach carcinogenesis by N-methyl-N'-nitro-N-nitrosoguanidine is effective in increasing the development of gastric tumors.

Growth hormone receptor (GH)-expressing carcinoid tumors after recombinant human GH therapy for human immunodeficiency virus-related lipodystrophy.

We describe a patient who had growth hormone receptor-expressing carcinoid tumors develop in the distal colon and rectum after he received recombinant human growth hormone therapy for human immunodeficiency virus-related lipodystrophy. This case report serves as a cautionary note regarding the use of potentially oncogenic recombinant human growth hormone therapy to treat human immunodeficiency virus-positive persons.

Pharmacological basis for the induction of gastric carcinoid tumours in the rat by loxtidine, an insurmountable histamine H2-receptor blocking drug.

The very late occurrence of gastric carcinoids in a life-span carcinogenicity study with loxtidine in the rat might have resulted from continuous achlorhydria induced by this long-acting unsurmountable histamine H2-antagonist. The nature of the anti-secretory activity of loxtidine was compared with that of ranitidine on histamine-induced acid secretion in the perfused stomach preparation of the rat and in the rat isolated gastric mucosa preparation. Ranitidine and loxtidine had qualitatively different inhibitory effects on acid secretion, ranitidine being a competitive antagonist of histamine even at high concentrations, whereas the effect of loxtidine on both preparations was unsurmountable at relatively low concentrations. These results support the hypothesis that the late formation of gastric carcinoids in rats receiving loxtidine is a consequence of persistent achlorhydria caused by unsurmountable blockade of parietal cell H2-receptors.

Review article: potential gastrointestinal effects of long-term acid suppression with proton pump inhibitors.

This review examines the evidence for the development of adverse effects due to prolonged gastric acid suppression with proton pump inhibitors. Potential areas of concern regarding long-term proton pump inhibitor use have included: carcinoid formation; development of gastric adenocarcinoma (especially in patients with Helicobacter pylori infection); bacterial overgrowth; enteric infections; and malabsorption of fat, minerals, and vitamins. Prolonged proton pump inhibitor use may lead to enterochromaffin-like cell hyperplasia, but has not been demonstrated to increase the risk of carcinoid formation. Long-term proton pump inhibitor treatment has not been documented to hasten the development or the progression of atrophic gastritis to intestinal metaplasia and gastric cancer, although long-term studies are required to allow definitive conclusions. At present, we do not recommend that patients be tested routinely for H. pylori infection when using proton pump inhibitors for prolonged periods. Gastric bacterial overgrowth does increase with acid suppression, but important clinical sequelae, such a higher rate of gastric adenocarcinoma, have not been seen. The risk of enteric infection may increase with acid suppression, although this does not seem to be a common clinical problem with prolonged proton pump inhibitor use. The absorption of fats and minerals does not appear to be significantly impaired with chronic acid suppression. However, vitamin B12 concentration may be decreased when gastric acid is markedly suppressed for prolonged periods (e.g. Zolllinger-Ellison syndrome), and vitamin B12 levels should probably be assessed in patients taking high-dose proton pump inhibitors for many years. Thus, current evidence suggests that prolonged gastric acid suppression with proton pump inhibitors rarely, if ever, produces adverse events. Nevertheless, continued follow-up of patients taking proton pump inhibitors for extended periods will provide greater experience regarding the potential gastrointestinal adverse effects of long-term acid suppression.

Proton pump inhibitors, enterochromaffin-like cell growth and Helicobacter pylori gastritis.

In both rodents and humans the development of gastrin-promoted gastric argyrophil enterochromaffin-like cell carcinoids requires the involvement of a genetic factor inherent to multiple endocrine neoplasia syndrome or of type A autoimmune chronic atrophic gastritis. Prolonged severe hypergastrinaemia acting on non-gastritic mucosa, as in Zollinger-Ellison syndrome patients, results in diffuse argyrophil enterochromaffin-like cell hyperplasia but, as a rule, does not produce tumours. Combination of chronic atrophic gastritis (mostly related to Helicobacter pylori infection) with hypergastrinaemia frequently causes linear and micronodular hyperplasia of argyrophil cells, whereas carcinoids are exceptional. No tumours or pre-neoplastic lesions have been observed in patients treated long-term with proton pump inhibitors, apart from rare cases in patients with combined Zollinger-Ellison and multiple endocrine neoplasia syndromes. A moderate increase in the incidence of argyrophil cell clustering, with or without hyperplasia, probably results from the parallel evolution of ulcer-associated Helicobacter gastritis into chronic atrophic gastritis. Eradication of H. pylori with a combination of proton pump inhibitors and antibiotics suppresses gastritis and prevents ulcer recurrence.

Induction of carcinoids in the glandular stomach of rats by N-methyl-N'-nitro-N-nitrosoguanidine.

A total of 30 inbred Wistar rats were orally administered 70 microgram/ml solution of N-methyl-N'-nitro-N-nitrosoguanidine (MNNG) for 35 weeks and then tap water for the following 20 to 30 weeks. Four of the 20 females and two of ten males developed carcinoids in the glandular stomach, but no metastasis could be found. Carcinoids developed most frequently in the fundic portion along the greater curvature. Histologically, these tumors were medullary anaplastic carcinomas containing two different endocrine cell populations. The first cell type was argentaffin having the electron-dense, somewhat pleomorphic secretory granules (437-810 nm) and the second type was argyrophil having round granules with a dense core and a pale halo (550 nm). None of these tumors showed endocrine immunoactivity for gastrin, somatostatin, insulin, glucagon, and enkephalin. One of these gastric tumors developed into scirrhous carcinoma, but differentiated adenocarcinoma could not be seen in the glandular stomach.

H2-receptor blockade induces peptide YY and enteroglucagon-secreting gastric carcinoids in mastomys.

Gastric carcinoid tumor formation has been reported with prolonged achlorhydria in both animals and human beings. The hypothesis in this study was that the ablation of parietal cell function in an animal (mastomys) genetically predisposed to gastric neuroendocrine neoplasia would promote and accelerate tumor formation. Loxtidine, an irreversible H2-receptor blocker, was administered at 1 mg/kg/day in drinking water for 4 months to young mastomys (n = 16). After 4 months of treatment, 14 of 16 animals had gastric carcinoids compared with 0 of 16 young control animals and 4 of 16 older control animals. Ultrastructurally, these tumors were characterized by the presence of neurosecretory granules. Serum gastrin levels were elevated (230 +/- 40 pmol/L) in loxtidine-treated animals compared with control animals (26 +/- 8 pmol/L) (p less than 0.05). In addition, both peptide YY (620 +/- 160 pmol/L) and enteroglucagon (500 +/- 147 pmol/L) were significantly elevated compared with control groups (p less than 0.05). Similarly, in tumor tissue, peptide YY (676 +/- 152 pmol/gm) and enteroglucagon (551 +/- 164 pmol/gm) were found in large quantities, whereas gastrin was undetectable. These observations provide substantial support for the possible pathophysiologic role of gut peptides, particularly gastrin, in the generation of endocrine neoplasia. The advent of endocrine tumors after inhibition of a gut secretory cell (parietal) may be of considerable significance in understanding the genesis of endocrine neoplasia. Whether the drug acts as a neoplastic promoter of enterochromaffin-like cells or the tumor development is related to elevation of peptides such as gastrin cannot be established in this study. Long-term H2-receptor blockade with new potent, irreversible agents as an alternative to surgery may have potential grave implications that require careful consideration.

A comparative overview of the adverse effects of antiulcer drugs.

During the past 2 decades, great advances have been made in the treatment of ulcer disease. This has involved the development of new drugs that are not only well tolerated, but are relatively inexpensive. The lack of significant adverse effects has revealed a degree of tolerability that, to write a review of the adverse effects, poses a difficult task. Most of the adverse effects are related to an excessive reaction to the relevant pharmacological characteristic that mediates the therapeutic response. The drug dosage can be reduced, freeing the patient of the adverse reaction, but leaving behind a background activity adequate to produce a therapeutically beneficial effect. The adverse effects of H2-antagonists fall into 2 groups. Firstly, there are poorly defined symptoms that have a prevalence similar to that in the community; these include headache, giddiness, dizziness, fatigue, constipation and diarrhoea. Secondly, they may delay the metabolism of drugs metabolised by the the cytochrome P450 system, and rarely be androgenic. Many antacids and the site-protective agent sucralfate contain aluminium, which can be absorbed, producing elevation of serum aluminium levels. In view of the possible association of aluminium with Alzheimer's disease, anxiety has arisen as to whether aluminium from these sources may, in those on prolonged treatment, cause Alzheimer's disease. However, the evidence so far indicates that aluminium is not a risk factor for Alzheimer's disease. The association of gastric cancer with achlorhydria has led to the fear that long term use of potent acid inhibitors may cause cancer. This fear has been accentuated by the observation that some rats, given omeprazole over their lifetime, developed carcinoid tumours of the stomach. However, enthusiastic research, both clinical and epidemiological, indicates that drug-induced achlorhydria is unlikely to be a problem in humans. Site protective agents have a role in certain conditions such as pregnancy where the systemic effect of a drug may produce adverse effects.

5-HTP induced diarrhea as a carcinoid syndrome model in mice?

Serotonin (5-HT) is present in the gastrointestinal tract and is probably one of the compounds responsible for diarrhea in patients presenting with carcinoid syndrome. Intraperitoneal administration of L-5-hydroxytryptophan (L-5-HTP) at doses of 25 to 100 mg/kg dramatically increase defecation in mice. In this new paradigm, counting fecal boli deposited is simple and the appraised or inhibition of diarrhea induced by ip 25 mg/kg of L-5-HIP is very clear, with a good reproducibility of scores. L-5-HTP needs to be metabolized into 5-HT to be active; benserazide, an inhibitor of decarboxylase, antagonized the diarrhea induced by 5-HT. Among the 5-HT antagonists used in interaction with 5-HT, only these of the 5-HT3 type (ondansetron, granisetron, tropisetron) and, to a lesser extent 5-HT2 type (ritanserin), decreased the diarrhea induced by 5-HTP. The 5-HT4 receptor agonists from the benzamide family (metoclopramide and zacopride) increased defecation in mice but the effect failed to reach statistical significance.

Pharmacology and toxicology of omeprazole--with special reference to the effects on the gastric mucosa.

Omeprazole is a long acting inhibitor of gastric acid secretion in different species including rat and dog. Due to the long duration of action, steady state inhibition at repeated once daily administration is reaches within 4-5 days in dogs and in about 3 days in rats. Daily dosing at high dose levels results in virtually complete 24-hour inhibition of acid secretion in experimental animals. The elimination of the inhibitory feedback effect of acid on gastrin secretion leads to hypergastrinaemia. Because gastrin has a trophic effect on the oxyntic mucosa, the hypergastrinaemia results in a reversible hypertrophy of the oxyntic mucosa and an increased capacity to produce acid following maximal stimulation with exogenous secretagogues after discontinuing treatment. Despite the increased capacity to produce acid, basal acid secretion seems to be unchanged. The pronounced hypergastrinaemia which occurs during long-term treatment with high doses rapidly normalizes after discontinuing treatment. The hyperplasia of the oxyntic endocrine ECL cells, and the eventual development of gastric ECL cell carcinoids after lifelong treatment of rats with high doses, can also be attributed to the hypergastrinaemia developing after almost complete elimination of gastric acid secretion in these animals.

Potential hazards of long-term acid suppression.

Neuroendocrine cell (carcinoid) tumours have been reported in the acid-secreting part of the stomach of rodents after long-term administration of a range of potent chemically diverse antisecretory agents. Although evidence shows a link between the sequence of acid suppression, hypergastrinaemia, and neuroendocrine cell hyperplasia, other factors are also thought to be involved in neoplastic transformation. Prolonged hypochlorhydria or achlorhydria resulting in bacterial colonization of the stomach may allow the generation of carcinogenic substances. Other as yet unidentified trophic factors may be involved in tumour formation. In view of the potential risks associated with these agents, there must be concern about the possible consequences in man of marked suppression of acid. It seems wise to limit the use of these more potent agents to situations in which conventional therapy has failed and to short-term treatment.

[Risk for developing tumors in therapy with the proton pump inhibitor omeprazole]. / Risiko der Entwicklung von Tumoren unter der Therapie mit dem Protonenpumpen-Inhibitor Omeprazol.

The experience received hitherto with the treatment of omeprazole proves this drug as absolutely save and poor of side effects. ECL-cell hyperplasia and carcinoids which occur during application of very high doses of omeprazole in rats are not caused by a direct effect of omeprazole. These changes are induced by extremely elevated serum gastrin levels due to achlorhydria. However, these high doses of omeprazole are not applicated in man. Therefore, the results from animal studies can not be applied on humans. During long-term treatment with omeprazole in most patients only slight increases of serum gastrin levels and ECL-cell densities occur. Only a very few patients exhibit a progressive and marked increase of serum gastrin levels. These patients should be carefully monitored by endoscopy in regularly intervals. There is no current evidence to support the contention that omeprazole is genotoxic. Results from Burlinson on the potential of omeprazole to induce DNA damage proved to be unsounded and of no clinical relevance. Omeprazole produced negative results in all well established genotoxicity tests. The incidence of neoplasias in the stomach is not increased after long-term acid suppression as indicated by 15 years application of H2-blockers and even a longer period of experience with vagotomy. Hitherto, there is also no evidence that acid inhibition induced by omeprazole causes an increased rate of carcinoma of the stomach. The risk for promoting carcinomas of the colon by the mild hypergastrinaemia during treatment with omeprazole seems to be inferior. In addition the correlation between development of carcinomas of the colon respective rectum and hypergastrinaemia is discussed controversly.

Immunohistochemical evidence for an impairment of autophagy in tumorigenesis of gastric carcinoids and adenocarcinomas in rodent models and patients.

BACKGROUND/AIM: Autophagy has dual roles in tumorigenesis: tumor-promoting or tumor-suppressing. The aim of the present study was to examine autophagy-related markers by immunohistochemistry in gastric carcinoids and adenocarcinomas in rodent models and patients. METHODS: Gastric carcinoids in Mastomys were induced by loxtidine treatment. Spontaneously developed gastric adenocarcinomas in Japanese cotton rats and INS-GAS transgenic mice were included. Patient tissue samples of gastric carcinoids or adenocarcinomas were collected. Immunohistochemistry was performed against autophagy-related gene protein-6 (ATG-6, also called beclin-1), ATG-5 and ATG-16. RESULTS: In tumor-free Mastomys, ATG-5, ATG-16 and beclin-1 were immunepositive in the gastric mucosa. In tumor-bearing Mastomys, ATG-5 and ATG-16 were negative in the tumors, whereas beclin-1 was positive in four of five animals. In carcinoid patients, ATG-5 was negative in six of ten, ATG-16 negative in nine of ten, and beclin-1 negative in three of ten patients. In cotton rats, ATG-5 and ATG-16 were negative in all tumors. Beclin-1 was negative in three of five rats. In INS-GAS mice, ATG-5 and beclin-1 were positive in the tumor area, but the numbers of immunopositive cells per gland were reduced by about 50% in comparison with wild-type mice. In adenocarcinoma patients, ATG-5 and ATG-16 were negative in eight of ten, and beclin-1 positive in all ten patients. CONCLUSIONS: An impaired autophagy took place at the stage of formation of ATG-5-ATG-12-ATG-16 complex in both gastric carcinoids and adenocarcinoma of both rodent models and patients. ATG-5 and ATG-16 might be better markers than beclin-1 in assessing autophagy in these lesions.

Gastric carcinoids after long-term use of a proton pump inhibitor.

BACKGROUND: Proton pump inhibitors (PPIs) are potent inhibitors of gastric acid secretion and give hypergastrinemia secondary to gastric hypoacidity. PPI treatment therefore induces enterochromaffin-like (ECL) cell hyperplasia. Long-term hypergastrinemia in rodents and man also leads to ECL cell neoplasia. Whether long-term PPI treatment will induce ECL cell neoplasia in man has been disputed. AIM: To describe gastric carcinoids in two patients with a history of long-term PPI use. RESULTS: Two patients had been taking PPI for 12-13 years due to gastro-oesophageal reflux disease. At routine upper gastrointestinal endoscopy a solitary tumour was found in the oxyntic mucosa of both patients. Histology from the tumours showed in both cases a well-differentiated neuroendocrine tumour. Biopsies from flat oxyntic mucosa showed no signs of atrophic gastritis and a normal presence of parietal cells in both cases, but hyperplasia of ECL cells. The tumour in patient 1 was resected endoscopically. After cessation of PPI treatment the tumour regressed in patient 2 and the ECL cell hyperplasia regressed in both patients. In patient 2 serum gastrin and chromogranin A were elevated during PPI treatment, and normalised after cessation of treatment. In patient 1, unfortunately, we had serum only after treatment, and at that time both parameters were normal. CONCLUSION: These cases show that hypergastrinemia secondary to proton pump inhibitors treatment, like other causes of hypergastrinemia, may induce enterochromaffin-like cell carcinoids in man.