Omeprazole is approved only for short-term use in the United States but is approved in other countries, such as the Netherlands, United Kingdom, Belgium, and Brasil, for long-term treatment of reflux esophagitis. Most clinical experience has been gained outside rigorously controlled clinical trials. In the Netherlands, a protocol of omeprazole to treat patients with refractory reflux disease was started in 1985. We report the results of omeprazole to treat patients for as long as 64 months.

Methods

Top Methods Results Discussion Author & Article Info References

Patients

This open, compassionate-use study started in November 1985; our analysis includes all data collected to March 1991. It was conducted in 7 referral centers in the Netherlands. From November 1985 to January 1989, all patients with erosive reflux esophagitis diagnosed with endoscopy and no response to treatment with H₂-receptor antagonists (cimetidine, 800 to 2000 mg/d; ranitidine, 600 to 900 mg/d; or famotidine, 80 to 120 mg/d) for at least 3 months before the start of the study were selected for the study. All eligible patients (n = 91) participated. Pre-entry severity of esophagitis was classified endoscopically as grade II (noncircumferential erosions), grade III (circumferential erosions or ulceration), or grade IV (erosions, ulceration plus deep ulcer or stricture) according to the classification of Savary and Miller [23]. We classified patients with Barrett esophagus according to the grade of esophagitis proximal to the Barrett epithelium. We defined healing of reflux esophagitis as absence of mucosal defects. Thirty-two patients had Barrett esophagus in combination with esophagitis. Twenty-five patients previously had antireflux surgery (Nissen fundoplication or hiatoplasty). Nine patients had been hospitalized for esophageal bleeding. Twenty-eight patients needed repeated dilatations before and during the initial healing phase.

All patients were given 40 mg of omeprazole once daily for 4 to 16 weeks, until endoscopic examinations showed that the esophagitis was healed. In 5 patients, the healing dose of omeprazole was continued as a maintenance dose for reasons specified later. We did follow-up examinations every 3 months in the first year of maintenance therapy and thereafter at 6-month intervals. We interviewed patients to determine recurrence of symptoms (heartburn, regurgitation, and dysphagia) and adverse events. In addition, upper gastrointestinal endoscopy was done, including fundic biopsies, and a fasting blood sample was obtained for routine laboratory screening and serum gastrin tests.

In patients with endoscopic relapse, we increased the dose of omeprazole by 20 mg. From 1990, patients received endoscopic examinations and laboratory tests annually.

Serum gastrin was measured by a previously described sensitive and specific radioimmunoassay [24]. We considered 100 ng/L the upper limit of normal for fasting serum gastrin. At endoscopy, two or three biopsy specimens were obtained from the gastric corpus mucosa approximately 10 cm below the cardia along the greater curvature. The biopsy specimens were coded, fixed in neutral formalin, and embedded in paraffin. Deparaffinized 5- microns sections were stained with hematoxylin and eosin to assess the overall tissue structure. Adjacent sections were also stained with Sevier-Munger silver for argyrophil endocrine cells [25]. Of the six endocrine cell types in gastric corpus mucosa, 50% [26] or more [25] are composed of enterochromaffin-like cells, which are distinctly positive with Sevier-Munger silver stain. Serotonin-producing argentaffin enterochromaffin cells, composing a minor part (10% to 20%), and a minute number of somatostatin D₁ cells are also argyrophilic, whereas the remaining endocrine cell types are negative [26].

We classified chronic gastritis as subatrophic or atrophic when at least 30% gland atrophy was evident, corresponding to both preatrophic and atrophic gastritis [27]. We classified the remaining cases of chronic gastritis as nonatrophic, thereby including superficial and interstitial gastritis and also mild subatrophic or atrophic changes. The endocrine cell changes were evaluated according to the classification of Solcia and associates [26]. The qualitative analysis of all corpus biopsy specimens was done blindly (by N. Havu, MD, Department of Pathology, Safety Assessment, AB Astra, Sodertalje, Sweden).

Statistical Analysis

We analyzed differences in serum gastrin level between the various time intervals for statistical significance using the Wilcoxon signed-rank test [28]. To compare percentages, we used Fisher exact tests. To compare the degree of enterochromaffin-like cell hyperplasia and gastritis at first and last biopsy, we used the McNemar test. This test considers all patients with changes but disregards those with unchanged classification of hyperplasia and gastritis, respectively. We estimated percentages of patients still in remission using the life-table method. Results are expressed as medians and ranges unless otherwise indicated. The protocol was approved by the ethics committee of the Free University Hospital. All eligible patients gave informed consent.

Results

Top Methods Results Discussion Author & Article Info References

Of the 91 patients entering the study, esophagitis healed in 58 (64%) after 4 weeks of treatment with 40 mg of omeprazole; it healed in 16 (18%) after 8 weeks and in 12 (13%) after 12 weeks. Four patients (4%) were healed only after acute treatment was extended to 16 weeks. One patient only healed after the dose of omeprazole was increased to 60 mg daily. These 5 patients continued to receive 40 mg of omeprazole as a maintenance dose. The time to healing for patients with Barrett esophagus and esophagitis was the same as for other patients, although prestudy endoscopic examination showed that patients with Barrett esophagus had more severe reflux esophagitis (63% grade IV esophagitis compared with 27% in the other patients). The median observation time during maintenance treatment was 48 months (range, 24 to 64 months).

Drop-outs

Four patients dropped out of the study. Three patients died: One died of postsurgical complications after a hip fracture 48 months after the start of maintenance treatment; one with known multiple myeloma died of related infectious complications after 36 months of maintenance therapy; and one died of pancreatic carcinoma that was diagnosed at autopsy after 42 months of therapy with omeprazole. One patient was lost to follow-up because of emigration.

Clinical Evaluation

All 91 patients were followed for at least 36 months: 67 patients for 48 months and 27 for 60 months or more. Symptomatic and endoscopic relapse of esophagitis occurred in 47% (95% CI, 36% to 58%) of the 86 patients receiving maintenance treatment with 20 mg of omeprazole daily (Figure 1). All patients with relapses healed within 3 months of increasing the daily dose of omeprazole by 20 mg daily (Figure 2). Seven of those 40 patients (18%; CI, 7% to 33%), who had already had a relapse, had another relapse after a mean follow-up period of 24 months (range, 9 to 36 months) and were successfully treated with a further dose increment to 60 mg daily for a mean duration of 36 months (range, 6 to 39 months). Attempts to decrease the dose after rehealing resulted in rapid relapse, as evidenced by symptoms and endoscopic examination. During the maintenance period, peptic strictures did not occur. We diagnosed Barrett esophagus by pre-entry endoscopic examination in 32 patients. During the follow-up period, we observed no regression or disappearance of Barrett epithelium. However, Barrett epithelium did not progress during the entire observation period. The relapse rate in patients with previous antireflux surgery was 40% (16 of 25 patients), which was similar to that for the whole group. Although most patients remained asymptomatic during maintenance treatment, 7 of the 91 (8%) patients reported occasional mild heartburn.

View larger version (12K): [in this window] [in a new window]   Figure 1. Remission rates (%) in 86 patients with reflux esophagitis during maintenance treatment with 20 mg of omeprazole once daily during 60 months of follow-up (life-table method. * = number of persons at risk.

View larger version (47K): [in this window] [in a new window]   Figure 2. Cumulative remission rates with dose adjustments required to maintain remission in 91 patients with reflux esophagitis during maintenance treatment with omeprazole. Cumulative remission rate in 91 patients after 36 months: 100% (95% CI, 96% to 100%); in 67 patients after 48 months: 100% (CI, 95% to 100%); and in 27 patients after 60 months: 100% (CI, 87% to 100%). The 5 patients receiving 40 mg of omeprazole at 0 months follow-up were not healed after the acute treatment period of 12 weeks and continued to receive this dose. The other patients receiving 40 mg of omeprazole represent a first relapse, whereas the patients receiving 60 mg of omeprazole represent a second relapse.

Effect on Fasting Serum Gastrin Levels

In the initial healing phase, median fasting serum gastrin levels increased from 60 ng/L before study and during treatment with an H₂-receptor antagonist to 162 ng/L during treatment with omeprazole (P <0.01). Thereafter, gastrin levels did not change significantly during maintenance treatment (Figure 3). For example, the mean of the paired differences in gastrin levels in 63 patients between 12 and 36 months of treatment was –109.73 ng/L, which was not significant (P = 0.1; CI, –247.31 to 27.85 ng/L). We detected no relation between the dose of omeprazole in the maintenance period and the serum gastrin level.

View larger version (15K): [in this window] [in a new window]   Figure 3. Distribution of fasting serum gastrin levels (all measured individual values) during maintenance treatment with omeprazole. Horizontal line indicates normal values (< 100 ng/L). 40 mg = acute treatment period.

Ten of the 91 patients showed very high serum gastrin levels (at least two recordings of more than 500 ng/L during the follow-up period), with the highest recording being 3867 ng/L (Table 1). Demographic data did not differ from those of the remaining patients, except for the presence of Barrett esophagus (6 of 10 patients), peptic stricture (5 of 10 patients), and previous antireflux surgery (5 of 10 patients). In all 10 patients, fasting serum gastrin levels before treatment were significantly higher than in other patients studied (median, 168 ng/L compared with 60 ng/L; P < 0.01). Furthermore, all patients repeatedly had food retention discovered with endoscopic examination, despite fasting for at least 12 hours. Food retention was consistently present at endoscopic examination during periods in which patients had very high gastrin levels (>1000 ng/L). Food retention was never observed in the other patients. In 50% of patients (5 of the 10) with food retention and hypergastrinemia, we observed progression of gastritis to subatrophic or atrophic gastritis; such progression was observed in only 12% of patients (11 or 79) in the remaining group (P = 0.002)

Histopathologic Findings

We examined biopsy specimens from all but two patients and found no dysplasia or neoplasia. For patients to be included in the analyses of histopathologic data, at least 330 days had to pass between the first and last biopsies. A subgroup of patients (n = 24) did not have pre-entry biopsies but had a first biopsy within a few weeks after starting omeprazole therapy. Serial biopsy specimens were available in 25 patients treated for as long as 4 years (Tables 2 and 3). The incidence of micronodular hyperplasia increased from 2.5% to 20% (P = 0.001) Figure 4, and the incidence of subatrophic or atrophic gastritis increased from less than 1% to 25% between the first and last biopsies (P < 0.001) (Figure 5). We found a strong association between argyrophil cell hyperplasia and degree of corpus gastritis. Eight of the 79 patients (10%) with serum gastrin levels less than 500 ng/L showed micronodular enterochromaffin-like cell hyperplasia on one or more occasions during the follow-up period compared with 6 of the 10 patients (60%) with serum gastrin levels greater than 500 ng/L during the maintenance treatment period (P < 0.01).

View larger version (29K): [in this window] [in a new window]   Figure 4. Argyrophil cell hyperplasia during maintenance treatment with omeprazole: distribution in normal, diffuse hyperplasia, linear hyperplasia, and micronodular hyperplasia according to the classification of Solcia and colleagues [26] .

View larger version (48K): [in this window] [in a new window]   Figure 5. Degree of corpus gastritis during maintenance treatment with omeprazole. Distribution in normal, superficial gastritis, and subatrophic or atrophic gastritis.

Tolerance

Adverse events occurred in 19 of the 91 patients. Three elderly patients died of causes not related to omeprazole treatment. In 3 patients, adverse events were related to preexisting bronchopulmonary disease. Deep venous thrombosis developed in 1 patient's right leg after 4 months of omeprazole therapy. He received anticoagulant therapy for 3 months and had an uncomplicated outcome. One patient had a recurrence of hyperthyroidism 7 months after starting omeprazole therapy and responded well to therapy with a thyroistatic agent. Complete retinal ablation in the left eye was reported by one patient after 11 months of omeprazole use. She was treated unsuccessfully with laser therapy. Ulcerative colitis of the distal colon developed in one patient 25 months after inclusion. He was treated successfully with salazosulfopyridine. Four patients complained of mild, transient skin rashes. No patient required specific medical treatment. Finally, self-limiting diarrhea, mild headache, and bouts of dizziness were reported by one patient each.

Sixty-six patients used concomitant medication at some time during omeprazole treatment, often on a regular basis. Eighteen patients were treated at times with at least four other simultaneous drugs during the trial. No patient required changes in dosage for any drug, and we found no relevant deviations in routine laboratory tests during treatment.

Discussion

Top Methods Results Discussion Author & Article Info References

Our results show that omeprazole is very effective in healing reflux disease resistant to H₂-receptor antagonist therapy and in preventing relapse. Peptic esophagitis recurs frequently after therapy is stopped, particularly in patients who need profound acid inhibition for healing to occur [5, 6, 11]. In prospective studies from Scandinavia [29] and Australia [6], results of 12-month maintenance treatment with 20 mg of omeprazole daily were substantially better than with 150 mg of ranitidine twice daily in patients with ulcerative reflux esophagitis. The proportion of patients remaining in remission was 67% in the Scandinavian study and 89% in the Australian study. In an open German study [30] with a mean follow-up of 24 months, one third of the patients relapsed. Most relapses occurred in the first 6 months of the study. For our study, omeprazole effectively prevented relapses for at least 5 years. Most relapses (23%) occurred within the first 12 months after dose reduction following healing, but all patients rehealed after we increased the dose by 20 mg daily. Some patients had a second or third relapse, but all responded well to higher doses of the drug, and, in some patients, doses of 60 to 80 mg omeprazole daily were required. Our study showed no correlation between initial degree of severity and minimal effective dose, confirming the findings of Koop and Arnold [30]. During the 5-year-follow-up, esophageal strictures did not recur in patients who previously needed repeated dilatation. Although pre-entry examination showed grade IV reflux esophagitis in most patients with Barrett esophagus, neither the healing rate nor the recurrence rate differed from that of the other patients. During treatment, we observed neither regression nor progression of Barrett epithelium. More than 90% of patients were completely asymptomatic, which is further evidence of successful endoscopic remission.

Our follow-up period enabled us to study the long-term side effects of omeprazole therapy. Three patients died after more than 36 months of maintenance treatment. Two of these deaths appeared to be unrelated to omeprazole therapy. Pancreatic carcinoma might be a chance occurrence but could have been related to omeprazole therapy. Several experimental studies have shown that gastrin stimulates mucosal growth in the stomach [31] and colon [32], but no study has suggested that hypergastrinemia promotes the development and growth of human gastrointestinal tract adenocarcinomas. Levels of serum gastrin likely to be encountered in humans have not been shown to stimulate tumor growth [33].

As in other studies [34-36], fasting gastrin levels increased during the initial healing phase and remained stable during long-term treatment with omeprazole. We found very high gastrin levels (> 500 ng/L measured in two or more tests) in 10 patients. These extreme levels have also been reported by Koop and Arnold [30] (in 2 of 31 patients) and by Lamberts and colleagues [35] (in 4 of 74 patients). We repeatedly found food retention at endoscopy in this subgroup, probably because of delayed gastric emptying. Thus, antral distention may have contributed to the high gastrin levels. Furthermore, all had significantly higher fasting serum gastrin levels before treatment than did the other patients studied. One half of the patients had previous gastric surgery, suggesting that previous vagal damage may be a contributing factor. Although the clinical presentation did not suggest the Zollinger-Ellison syndrome, we must admit that in neither this study nor in the previously mentioned studies [30, 35] was this completely excluded by secretin testing. We studied the cause of the hypergastrinemia in depth. Results of this study will be reported separately.

We observed a substantial progression toward subatrophic and atrophic gastritis during long-term treatment with omeprazole. A marked progression in chronic gastritis, from simple inflammation without atrophy to severe corpus atrophy, was shown to occur in both untreated controls [37, 38] and in untreated patients with gastric ulcer within a 4-year period [39]. Serum gastrin levels have been correlated directly with the severity of nonantral gastritis [36, 40]. In fact, the 10 patients with very high gastrin levels showed greater progression of gastritis and a higher incidence of subatrophic or atrophic corpus gastritis than did the other 81 patients.

Development of atrophic gastritis paralleled an increased frequency of micronodular hyperplasia of the argyrophil cell population, possibly resulting from the trophic influence of gastrin on the enterochromaffin-like cells of the gastric oxyntic mucosa [41]. Indeed, the subgroup of patients with very high gastrin levels showed a substantially higher incidence of enterochromaffin-like cell hyperplasia than did the other patients. A close association between the degree of atrophy and the occurrence of focal hyperplasia of argyrophil cells was found in patients with peptic ulcer who did or did not receive long-term antisecretory therapy [35, 42]. During treatment with omeprazole, Helicobacter pylori migrates from the antrum to the corpus in the stomach [43] and may be related to the progression of gastritis. However, we did not assess H. pylori in the biopsy specimens, and, given the uncontrolled nature of the study, we cannot definitively conclude whether these histologic changes were induced or enhanced by omeprazole treatment.

We observed no dysplasia or neoplasia in any of the biopsy specimens obtained during the study. However, the significance of the observed histologic changes in the stomach is not known. Nor do we know if persons with very high serum gastrin levels are at risk for progression of hyperplasia to dysplasia of the argyrophil cell population and to gastric atrophy during long-term omeprazole treatment. Therefore, continued follow-up of patients receiving long-term omeprazole therapy is needed.