Omeprazole, one of the proton pump inhibitors (PPI) offers potent suppression of gastric acid secretion by direct inhibition of the proton pump, H + /K + –adenosine triphosphatase (ATPase), in the parietal cell (1). Omeprazole is superior to ranitidine in long-term treatment of reflux esophagitis and is widely used for severe erosive esophagitis in adults and children (2,3).
Long-term omeprazole treatment is associated with hypergastrinemia in humans and rats. Rats given high dosages of omeprazole for a long period develop severe hypergastrinemia and gastric mucosal changes, including endocrine cell hyperplasia (4,5).Gastric carcinoid tumors have been reported as a result of lifelong therapy with omeprazole in female rats (5). Hypergastrinemia observed in humans prescribed omeprazole therapy is moderate compared with that found in rats, but gastric mucosal changes and endocrine cell hyperplasia have been reported (6,7). Gastrin is a potent trophic factor in the proximal stomach and distal intestine. The consequence of sustained hypergastrinemia in humans caused by long-term omeprazole treatment is of concern because with severe hypergastrinemia, as seen in patients with pernicious anemia, there is an increased risk of gastric carcinoids (8).
Gastrin secretion by antral G cells is a complex process that is inhibited by somatostatin secreted by the adjacent D cells (9). The presence of gastric acid in the lumen inhibits gastrin secretion and stimulates release of somatostatin by D cells. In contrast, increased intragastric pH stimulates G cells and inhibits D cells (10). The effects of omeprazole therapy on G and D cells have been extensively studied in rats (4), but the data in humans are limited. The adult studies on this topic have the confounding variables of the presence of Helicobacter pylori infection and atrophic gastritis (7,11). These two factors cause changes in antral G and D cells (12–15), and therefore the results of these studies may not reflect the effect of PPI alone.
In Western societies, children have a lower prevalence of H. pylori infection and gastric disease than adults do. There are no data on the effects of long-term omeprazole treatment on antral G and D cells in children. Therefore, the current study was undertaken to evaluate the effect of long-term omeprazole treatment on antral G- and D-cell numbers in children. The authors elected to study children who received omeprazole for reflux esophagitis for 4 years or more, as in the previous study. There were no changes in G or D cells before 4 years of omeprazole therapy (7). Biopsy specimens from healthy, untreated children of similar age to that of children in the study group were included as controls to assess the effect of age on antral G- and D-cell numbers.
MATERIAL AND METHODS
Sixty-three children were treated with omeprazole for reflux esophagitis between 1990 and 1998. Review of hospital records identified seven children who received omeprazole for longer than 4 years for severe reflux esophagitis. Omeprazole treatment was started at the mean age of 7.5 years (range, 3–14 years) and was administered orally or through a gastrostomy tube. The dose of omeprazole was adjusted according to clinical or endoscopic assessment. Upper gastrointestinal endoscopy and biopsies for histology were performed before starting omeprazole (baseline), every 3 months until healing, and then at intervals of 1 or 2 years.
All children fasted overnight before endoscopy. At endoscopy, two to three biopsy specimens were taken from the gastric antrum in addition to esophageal, duodenal, and gastric body biopsy specimens. After fixation in Bouin solution, all antral biopsy specimens were dehydrated and embedded in paraffin wax. Care was taken to ensure the correct orientation of the tissue. Five-micron sections were cut and stained with hematoxylin and eosin to check the orientation of glands and inflammation. Steiner stain was used to determine H. pylori infection. Exclusion criteria for the study group were the presence of inflammatory changes or H. pylori infection in the antral biopsy specimens at any time while undergoing omeprazole therapy.
The control group was designed to match the ages of children undergoing omeprazole treatment. Four children (two boys and two girls) from each of the age groups of 3, 5, 8, 11, 14, and 17 years were randomly selected from the endoscopy database. These children had undergone upper gastrointestinal endoscopies for abdominal pain, vomiting, or failure to thrive and had not received any acid-suppressing medications. Gross endoscopic appearance was normal in all, and biopsy specimens of the esophagus, the duodenum, the gastric body, and the antrum showed normal histology without H. pylori infection.
The antral biopsy specimen obtained at baseline and at 1 year, 4 years, and 7 years while undergoing omeprazole therapy and from the control group were reviewed. Multiple 5-μm sections of these antral biopsy specimens were mounted on coded slides such that the immunocytochemical staining and morphometrics were performed in a blinded manner.
Antral sections were immunostained using gastrin and somatostatin antibodies separately, as described previously (9,16). The primary antibodies used were rabbit gastrin antibody (DAKO Diagnostics Inc, Mississauga, Ontario, Canada; 1 in 4,000 dilution) and mouse somatostatin antibody (CURE, Los Angeles, U.S.A.; 1 in 5,000 dilution). The procedure in brief was as follows: Antral sections were dewaxed in xylene, then cleaned in petroleum ether before immersion in 0.3% H 2 O 2 in methanol to block the endogenous peroxidase activity. The sections were then incubated with primary antibodies for 48 hours at 4°C. The sections were successively incubated with respective biotinylated secondary antibodies and with avidin–biotin–peroxidase complex (Vectastatin, Vector Laboratories, Burlington, CA, U.S.A.) for 60 minutes each. The bound antibodies were detected by immersing the sections in diaminobenzidine solution. The sections were then briefly counterstained with Meyer hematoxylin solution.
The immunoreactive G and D cells were quantitated using the method described previously (17). For quantification of immunostained cells, five to eight antral glands with clear visible lumens were randomly examined with a Zeiss Axiophot microscope (Zeiss, Mississauga, Ontario, Canada). The microscope was linked to a G3 Macintosh (Apple Computer, Inc., Cupertino, CA, U.S.A.) computer equipped with video input. National Institutes of Health image software (shareware) was used to calculate the area of the gland and to count cells in the gland. The border of the antral gland was delineated manually with the cursor on the screen, and the area of the gland was expressed in pixels. The stained cells in that gland were counted, and the number of cells per total glandular area was calculated. The G and D cells were expressed as total number per unit glandular area, which was defined as 100,000 pixels.
Paired Student t test was used to compare G- and D-cell numbers in children undergoing omeprazole treatment from baseline to 1, 4, and 7 years. The ratios of G to D cells were expressed as mean ± SEM. For comparison of G- to D-cell ratio in omeprazole-treated children, log transformation was used because of the skewed nature of the raw data, and paired t test was performed on the logged data. In the control group, analysis of variance was performed to compare cell numbers and G- to D-cell ratio in children of different ages. Student t test was used to compare the baseline results of children taking omeprazole and the whole control group (n = 24).
Of seven children undergoing long-term omeprazole treatment, in one child, antral inflammation was seen during review of the biopsy specimens, and therefore the child was excluded from the study. Therefore, the final study group consisted of six children (three boys, three girls) who received omeprazole for 4 years or longer for severe reflux esophagitis. All six children had an underlying condition that predisposed them to severe gastroesophageal reflux: three had neurologic impairment, two had corrected esophageal atresia, and one had hiatal hernia. All had severe erosive esophagitis, refractory to H 2 –receptor blockers and prokinetics. Four children were taking H 2 –receptor blockers for 1 to 3 years before omeprazole therapy. Omeprazole therapy was tolerated well without any clinically adverse effects. The average dosage was 1.9 mg/kg/d (range, 1.4–2.5 mg/kg/d; actual dose, 20–60 mg). The omeprazole dose/kg was stable throughout the period reviewed but was slightly increased in two children. The changes in G- and D-cell count for these two children were similar to that of the other four children. Omeprazole therapy was associated with clinical improvement and healing of esophagitis in all children. There was no relapse of esophagitis during maintenance omeprazole therapy. None of these children had evidence of H. pylori infection or gastritis at any time while taking omeprazole. One of six children had a polyp in the gastric body, and two other children had nodules in the gastric body that appeared during omeprazole therapy. These findings have been reported elsewhere (18).
From the study group of six children, one child was undergoing omeprazole therapy before referral and did not undergo endoscopy at baseline or at 7 years, whereas another child did not undergo endoscopy at 1 year; therefore, three biopsy sections were unavailable. The child who did not undergo baseline endoscopy was excluded from statistical analysis.
The biopsy specimens from the study group were collected at ages ranging from 3 years to 17 years. To determine the normal number and distribution of G and D cells at these ages, control biopsy specimens from age-matched children were evaluated. The data showed that the numbers of antral G and D cells did not change significantly with age in healthy children (Fig. 1). The mean G- to D-cell ratio in the control group was 2.5 ± 0.1 at 3 years; 2.4 ± 0.4 at 5 years; 2.7 ± 0.4 at 8 years; 2.5 ± 0.2 at 11 years; 2.5 ± 0.4 at 14 years; and 2.4 ± 0.1 at 17 years. Again, no significant change was observed with increasing age. The results from the control group were compared with the baseline results from the study group, and there were no significant differences in the G- and D-cell numbers or the G to D cell ratios in these two groups.
In the study group, the number of antral G cells increased over time during omeprazole therapy; the differences between the baseline and 4 years (P < 0.01) and baseline and 7 years (P < 0.01) were significant (Fig. 2). The increase in cell number was not the result of an increased spread of G cells throughout the antral glands; the G cells remained concentrated in the middle third of the mucosa (Fig. 3 A and B). The morphology of G cells remained unchanged over time. There was considerable variation in the number of D cells, but no statistically significant change was observed at any time during omeprazole therapy compared with the baseline levels (Fig. 2). The morphology and distribution of D cells within the antral mucosa was also unchanged.
Because of the marked increase in G-cell numbers, there was a significant alteration in the G- to D-cell ratio at 7 years while undergoing omeprazole therapy when compared with the baseline results of the same children. The ratio was 2.6 ± 0.3 at baseline, 3.4 ± 0.3 at 1 year, 4.0 ± 0.5 at 4 years, and 5.2 ± 0.8 at 7 years (P < 0.02 for the difference from baseline) during omeprazole therapy.
This study shows that omeprazole therapy in children is associated with an increased number of antral G cells without significant changes in antral D-cell numbers. These changes in G cells were observed at 4 years and at 7 years during omeprazole therapy, but not at 1 year. The result was a significant increase in the G- to D-cell ratio at 7 years during omeprazole therapy.
Omeprazole therapy was initiated at different ages in the individual patients recruited into the study. This meant that any changes in G- and D-cell numbers might be influenced by the normal development of these endocrine cells in the maturing antrum over the study period. There is no information available on the changes in G- and D-cell numbers with increasing age in humans. Therefore, biopsy specimens from the antrum of a group of healthy, untreated children covering the range of ages in the study group were evaluated to determine normal endocrine cell dynamics. The results demonstrated that the number of G and D cells in the healthy, untreated children were comparable with that at the initiation of treatment in the children undergoing omeprazole therapy. In addition, no significant change in G- and D-cell numbers or G- to D-cell ratio was observed in the children aged 3 to 17 years. These data indicate that the initial steady state of these endocrine cells within the antrum has developed by 3 years of age and is maintained at a constant level thereafter. In mice, a comparison of G- and D-cell numbers in a 1-month-old compared with a 3-month-old animal showed no change in G cells, with a decrease in D cells (19). This suggests that there may be species differences in the development of antral endocrine cell populations.
The lack of effect of age on antral G- and D-cell numbers in children indicates that the increase in G-cell numbers in the study group can be attributed to omeprazole therapy. The majority of the data on the effects of omeprazole therapy on antral endocrine cells comes from rat experiments. Rats receiving long-term (1 year) omeprazole had a twofold increase in G-cell numbers compared with controls, whereas D-cell numbers were reduced to half, resulting in an almost fourfold increase in G- to D-cell ratio (4).
The effect of long-term omeprazole or PPI therapy on gastric endocrine cell populations in adult humans has been evaluated in two major studies. An initial report by Lamberts et al. (6) showed no change in antral G or D cells in 17 adults receiving omeprazole therapy for 2 years. In a subsequent follow-up, 74 adults receiving omeprazole therapy longer than 4 years were studied (7). In the latter study, 28 patients had reflux esophagitis, whereas other patients had gastric or duodenal ulcer diseases. H. pylori infection was present in the gastric mucosa in approximately 50% of the patients, and atrophic gastritis increased from 1.8% to 20% of patients during the study period. The G-cell and D-cell volume densities per mucosal area were estimated in 29 adults receiving omeprazole. A marginal increase in G-cell volume density was noted after 4 years of receiving omeprazole therapy without any change in D cells. The increase in G- to D-cell ratio was not significant (7). Recently, Eissele et al. (11) studied 42 adult patients with reflux esophagitis and gastric and duodenal ulcerations who were receiving long-term treatment with lansoprazole. H. pylori infection was present in approximately one half of the patients, and a steady increase in G-cell numbers per mucosal area was observed. The increase in G-cell numbers was significant after 27 months; however, D cells were not studied (11).
The current study has some important differences when compared with the earlier studies. Patients in the previous studies were adults and had H. pylori infection and atrophic gastritis (7,11). The presence of H. pylori infection and atrophic gastritis is associated with changes in antral G and D cells (12–15). Eradication of H. pylori infection in adults was associated with a significant increase in D-cell numbers (12). In children, an increase in D-cell numbers, along with a decrease in G-cell numbers, was observed after eradication of H. pylori infection (13). Kamada et al. (14) reported decreased numbers of G and D cells in severe H. pylori gastritis as a result of the possible damaging effect of severe inflammatory infiltrate on G and D cells. Atrophic gastritis has also been associated with a decrease in G-cell numbers in adults, depending on the severity of gastritis (15). In contrast to the previous PPI studies, the patients in this study did not have gastritis or H. pylori infection, and therefore the observed changes in G-cell numbers and the G- to D-cell ratio reflect the effect of omeprazole alone. However, compared with the studies in adults, the current study has a small number of patients because the use of omeprazole in the pediatric population is much less than use in adults. Endoscopy monitoring did not follow a rigid schedule, and we did not have biopsy specimens taken between 1 and 4 years in all patients receiving omeprazole. Despite these differences, our results were similar to previous human studies regarding an increase in G-cell numbers over time during omeprazole therapy without any change in D cells. Unlike the results of a study by Lamberts et al. (7), an increase in G- to D-cell ratio was significant in the current study because of a marked increase in G cells.
Omeprazole causes potent suppression of gastric acid secretion and leads to persistently high intragastric pH. A study of human biopsy material from patients in different states of acid secretion suggested that the G- to D-cell ratio is governed by intragastric pH in humans (10). In conditions associated with persistently high intragastric pH, such as pernicious anemia, G-cell numbers and G- to D-ratio are significantly increased (10). Therefore, the observed increase in G-cell numbers and G- to D-cell ratio in the current study are likely caused by sustained, increased intragastric pH associated with omeprazole action rather than by the direct effect of omeprazole on G cells. The actual mechanism of G-cell hyperplasia is not known. Cadiot et al. (20) suggested a prolonged life span of G cells as a cause of increased G-cell density in rats receiving omeprazole treatment. Enhanced mitosis is another possibility proposed after experiments in rats that showed formation of a new G-cell population when rats were again fed after prolonged fasting (21).
A long-term follow-up study of 230 adults over a period of up to 11.3 years (mean, 6.5 years) recently showed the safety of omeprazole therapy for control of reflux esophagitis (22). Nevertheless, this study did not address changes in G and D cells, and therefore the clinical significance of G-cell hyperplasia remains unclear. Previous studies showed no correlation between serum gastrin level and G-cell numbers in adults receiving PPI therapy (7,11). Serum gastrin level increased within a few weeks during PPI therapy and stayed at a higher level in the long term, whereas G-cell changes were not observed until several months later (7,11). In rat experiments, enhanced secretion from G cells and an increased gastrin messenger RNA level were related to hypergastrinemia during short-term omeprazole therapy (23). A similar mechanism may be a more important factor than G-cell hyperplasia in causing hypergastrinemia in humans, at least in the initial period of PPI therapy. G-cell hyperplasia may contribute to persistent, increased gastrin secretion observed with long-term PPI therapy.
In conclusion, long-term omeprazole therapy leads to significant increase in antral G-cell numbers and G- to D-cell ratio in children. Further studies are required to assess the clinical significance of these changes.
1. Israel DM, Hassall E. Omeprazole
and other proton pump inhibitors: pharmacology, efficacy and safety, with special reference to use in children. J Pediatr Gastroenterol Nutr 1998; 27: 568–79.
2. Halerback B, Unge P, Carling L, et al. Omeprazole
or ranitidine in long term treatment of reflux esophagitis. Gastroenterology 1994; 107: 1305–11.
3. Hassall E, Israel DM, Shepherd R, et al. Omeprazole
for chronic erosive esophagitis in children: a multicenter study of dose requirement for healing [abstract]. Gastroenterology 1997; 112: A143.
4. Sundler F, Anderson K, Mattsson H. Administration of omeprazole
to rats for one year produces reciprocal effects on antral gastrin
and somatostatin cells and no effect on endocrine cells in the colon. Digestion 1995; 56: 194–8.
5. Havu N. Enterochromaffin-like cell carcinoids of gastric mucosa in rats after life-long inhibition of gastric secretion. Digestion 1986; 35: 42–55.
6. Lamberts R, Creutzfeldt W, Stockmann F, et al. Long term omeprazole
treatment in man: effects on gastric endocrine cell populations. Digestion 1988; 39: 126–35.
7. Lamberts R, Creutzfeldt W, Struber HG, et al. Long term omeprazole
therapy in peptic ulcer disease: gastrin
, endocrine growth and gastritis. Gastroenterology 1993; 104: 1356–70.
8. Kokkola A, Sjobolm SM, Haapiainen R, et al. The risk of gastric carcinoma and carcinoid tumors in patients with pernicious anemia. A prospective follow-up study. Scand J Gastroenterol 1998; 33: 88–92.
9. Campos RV, Buchan AMJ, Meloche RM, et al. Gastrin
secretion from human antral G cells in culture. Gastroenterology 1990; 99: 36–44.
10. Arnold R, Hulst MV, Neuhof CH, et al. Antral gastrin
-producing G-cells and somatostatin-producing D-cells in different states of gastric acid secretion. Gut 1982; 23: 285–91.
11. Eissele R, Brunner G, Simon B, et al. Gastric mucosa during treatment with lansoprazole:Helicobacter pylori
is a risk factor for argyrophil cell hyperplasia. Gastroenterology 1997: 112: 707–17.
12. Moss SF, Legon S, Bishop AE, et al. Effect of Helicobacter pylori
on gastric somatostatin in duodenal ulcer disease. Lancet 1992; 340: 930–2.
13. Queiroz DMM, Moura SB, Mendes EN, et al. Effect of Helicobacter pylori
eradication on G-cell and D-cell density in children. Lancet 1994; 343: 1191–3.
14. Kamada T, Haruma K, Kawaguchi H, et al. The association between antral G and D cells and mucosal inflammation, atrophy, and Helicobacter pylori
infection in subjects with normal mucosa, chronic gastritis, and duodenal ulcer. Am J Gastroenterol 1998; 93; 748–52.
15. Marotta F, Hayakawa K, Mikami Y, et al. Relationship between gastrin
cell number, serum, antral mucosa and luminal gastrin
concentration and gastric acidity in antral atrophic gastritis. Gut 1990: 31: 279–81.
16. Buchan AMJ, Sikora LKJ, Levy JG, et al. An immunocytochemical investigation with monoclonal antibodies to somatostatin. Histochemistry 1985; 83: 175–80.
17. Buchan AMJ, Seal AM, Debas HT. Increased antral G-cell number and gastrin
content in dogs after massive small bowel resection. Dig Dis Sci 1985; 30: 236–9.
18. Pashankar D, Israel DM. Endoscopic gastric changes in children on long-term omeprazole
therapy [abstract]. J Pediatr Gastroenterol Nutr 1998; 27: 479.
19. Sandstrom O, Mahdavi J, El-Sahy M. Age related changes in antral endocrine cells in mice. Histol Histopathol 1999; 14: 31–6.
20. Cadiot G, Lehy T, Bonfils S. Action of somatostatin analogue (SMS 201–995) on the growth-promoting effect resulting from sustained achlorhydria in rat gastric mucosa with special reference to endocrine cell behavior. Euro J Clin Invest 1988; 18: 360–8.
21. Bertrand P, Willems G. Induction of antral gastrin
cell proliferation by refeeding of rats after fasting. Gastroenterology 1980; 78: 918–24.
22. Klinkenberg-Knol EC, Nelis F, Dent J, et al. Long term omeprazole
treatment in resistant gastroesophageal reflux disease: efficacy, safety, and influence on gastric mucosa. Gastroenterology 2000; 118: 661–9.
23. Dockray GJ, Hamber C, Evans D, et al. The secretory kinetics of the G cell
-treated rats. Gastroenterology 1991; 100: 1187–94.