Because gastrin secretion by the G cells of the antral mucosa is inhibited by an acid pH in the gastric lumen, hypergastrinemia may be induced by any treatment that decreases gastric acidity. For example, the administration of H₂-receptor antagonists causes an increase in plasma gastrin levels proportional to the dose [6]. Omeprazole administration causes a greater increase because it is a more potent inhibitor of gastric acid secretion [7]. Plasma gastrin levels increase within days of starting omeprazole therapy and peak at 2 to 4 months, after which they generally remain stable at 2 to 4 times the baseline levels [2, 6-9]. Gastrin levels remain within the normal range of up to 100 ng/L in most persons, but an occasional patient may have levels more than 10-fold the normal level. Levels greater than 500 ng/L were found in 10 of 91 (11%) patients by Klinkenberg-Knol and colleagues [2] and in 2 of 31 (6.5%) patients [10] and 4 of 74 (5.4%) patients [9] in similar long-term studies. All 10 patients described by Klinkenberg-Knol and colleagues had increased gastrin levels before omeprazole administration, presumably caused by previous treatment with high doses of H₂-receptor antagonists, and all had gastric retention of food, indicating the presence of a gastric motility disturbance. In some patients, the disturbance appeared to be related to previous gastric surgery. The gastric retention of food presumably enhanced gastrin release, thereby exacerbating the hypergastrinemia.

The concern with sustained hypergastrinemia relates largely to the trophic effect of gastrin on the oxyntic mucosa of the stomach, a phenomenon shown in rat gastric enterochromaffin-like cells. Rats given large doses of omeprazole [11] or ranitidine [12] throughout life developed hyperplasia of enterochromaffin-like cells that progressed in some rats to gastric carcinoid tumors. This effect is now known to be caused by hypergastrinemia and can be produced by nonpharmacologic methods that induce hypochlorhydria and hypergastrinemia, such as subtotal resection of the gastric fundus [13].

Gastric enterochromaffin-like cell carcinoid tumors have been reported in patients with massive hypergastrinemia because of pernicious anemia [14] and the Zollinger-Ellison syndrome [15] but so far not with omeprazole treatment maintained for up to 5 years [2, 8, 9]. The study by Klinkenberg-Knol and colleagues confirms previous reports of long-term treatment studies [8, 9] that found hyperplasia of enterochromaffin-like cells but no evidence of dysplasia, carcinoid tumors, or other neoplastic changes.

It does not necessarily follow that the hyperplasia in patients treated with omeprazole will progress to gastric carcinoid tumors, as seen in rats and in patients with pernicious anemia or the Zollinger-Ellison syndrome. Rats have an unusually high density of enterochromaffin-like cells and have an exaggerated gastrin response to achlorhydria [16]. Long-term omeprazole therapy given to species with lower densities of these cells (such as guinea pigs, hamsters, dogs, and mice) has not caused carcinoid tumors [16]. Patients with pernicious anemia generally have gastrin levels several times higher than those of patients receiving omeprazole [7], and those who developed carcinoid tumors had pernicious anemia for many years. In 123 patients with pernicious anemia, gastric carcinoid tumors were found by endoscopy in 5 (4%) who had a diagnosis of pernicious anemia for a mean duration of 19 years [14]. Thus, massive hypergastrinemia for many years appears to be necessary for tumors to develop in humans. Nearly all gastric carcinoid tumors in patients with the Zollinger-Ellison syndrome have developed in those with the multiple endocrine neoplasia I syndrome [17] despite the fact that gastrinomas are far more common in patients without the multiple endocrine neoplasia I syndrome. This observation has raised the possibility that the multiple endocrine neoplasia I gene itself is a cofactor along with massive hypergastrinemia for causing gastric carcinoid tumors in patients with the Zollinger-Ellison syndrome [17].

Data indicate that the enterochromaffin-like cell hyperplasia described by Klinkenberg-Knol and colleagues and others [2, 18, 19] may not be a drug effect. A unique study in Estonian patients with gastric ulcer disease who were not treated with antisecretory agents found that hyperplasia developed during a 4-year period [19]. The hyperplasia appeared to be a manifestation of selective preservation of enterochromaffin-like cells, whereas surrounding cells became atrophic during the progression of chronic gastritis to chronic atrophic gastritis, presumably because of H. pylori infection. Similar histologic changes were reported by Klinkenberg-Knol and colleagues, supporting the belief that the hyperplasia may not have been caused entirely by omeprazole-induced hypergastrinemia. The only way to be certain of this is to do a long-term controlled study of patients treated and not treated with omeprazole; such a study is unlikely to be done for ethical reasons, given the frequency of symptomatic relapses in patients with reflux esophagitis.

Given the diminishing concern about omeprazole-induced hypergastrinemia, should clinicians bother to monitor gastrin levels during continuous omeprazole therapy for long periods? Frequent monitoring would add cost but little value in view of the lack of therapeutic options for patients with high gastrin levels [20] and the infrequency of very high levels. Switching to standard doses of H₂-receptor antagonists, such as 150 mg of ranitidine twice daily, is an unsatisfactory option because it results in relapses in nearly 90% of patients within 1 year [4]. High doses of H₂-blockers should be effective in preventing relapses, but this has never been documented. In any case, this approach would be expected to cause increases in gastrin levels in proportion to the increased degree of acid suppression. Decreasing the omeprazole dosing regimen to 20 mg three times a week is also untenable because it is associated with a 50% increase in relapses within 6 months [21]. Antireflux surgery is appealing in young patients who otherwise face a lifetime of medical treatment, but it is important to select a surgeon with considerable experience and documentation of good results because successful antireflux surgery is highly operator-dependent.

Because the frequency of high gastrin values (greater than 500 ng/L) has been no higher than 11% in three studies [2, 9, 10] of long-term omeprazole use, a conservative and inexpensive compromise would be to measure gastrin levels after 1 year of continuous therapy. This is as likely to detect gastrin levels greater than 500 ng/L as are three separate measurements during the year [20]. If such levels are recorded, it could signify the development of chronic atrophic gastritis; this would diminish acid secretion, and the patients might no longer require omeprazole. As shown by Klinkenberg-Knol and colleagues [2], such patients could also have a gastric motility disturbance, which might respond to the addition of a prokinetic agent such as cisapride (Propulsid; Janssen Pharmaceutica, Titusville, New Jersey), although this remains to be determined. Thus, the specter of hypergastrinemia does not seem to be great enough to dissuade clinicians from using long-term omeprazole therapy in appropriately selected patients.