Lansoprazole and omeprazole in the prevention of relapse of reflux oesophagitis: a long-term comparative study


Carling Department of Medicine, Bollnäs- Söderhamns Hospital, 821 81 Bollnäs, Sweden. E-mail:



Proton pump inhibitors are superior to H2-receptor antagonists in the prevention of relapse of oesophagitis, but few data directly compare the relative efficacies of lansoprazole and omeprazole in preventing oesophagitis relapse over a prolonged period.


Patients with healed Grade II, III or IV oesophagitis were treated with lansoprazole 30 mg o.d. or omeprazole 20 mg o.d. for 48 weeks. Endoscopy and symptom assessment were performed after 12, 24 and 48 weeks of treatment and an additional symptom assessment 36 weeks after starting treatment.


Intention-to-treat analysis included 248 patients (lansoprazole n = 126, omeprazole n = 122). Comparison of time to endoscopic and/or symptomatic relapse revealed no difference between the treatments. There was no significant difference between treatments with respect to the proportion of patients in whom endoscopic and/or symptomatic relapse was reported (lansoprazole 12/126 (9.5%), omeprazole 11/122 (9.0%)). No difference between the treatments in either the number or severity of adverse events was reported.


Continuous treatment with either lansoprazole 30 mg or omeprazole 20 mg is effective in preventing the relapse of oesophagitis over a 48-week period in a majority of patients. Both treatments exhibit a similar side-effect profile.


Proton pump inhibitors, such as lansoprazole and omeprazole, have repeatedly been shown to be superior to H2-receptor antagonists in preventing relapse of oesophagitis.1[2][3][4]–5 This study was initiated in order to compare the effect of two of these proton pump inhibitors, lansoprazole and omeprazole, in the prevention of symptomatic and endoscopic relapse of oesophagitis over a 48-week period in patients who had previously undergone successful healing of their oesophagitis with lansoprazole for 8–12 weeks. The drugs were given at standard, recommended doses.


Study design

This was a multicentre, prospectively randomized, double-blind parallel group study. Twenty-three centres in Denmark, Finland and Sweden participated in the trial. The study protocol was reviewed and approved by the Ethics Committee of each of the centres. Written informed consent (Denmark and Finland) or witnessed verbal informed consent (Sweden) was obtained. The study was conducted in accordance with the Declaration of Helsinki.


Patients of both sexes, aged between 18 and 80 years, with oesophagitis Grade 0 or 1 (without erosions) after an initial Grade II, III or IV oesophagitis (Savary–Miller grading scale) treated with lansoprazole 30 mg o.d. for 8–12 weeks and who reported no severe symptoms of heartburn, regurgitation or dysphagia, were eligible to participate in the maintenance study.

Patients who were pregnant or lactating, had undergone previous upper gastrointestinal surgery (except simple perforation closure), had a malignancy of the upper gastrointestinal tract, or severe oesophageal disease (e.g. stricture needing dilation, Barrett’s oesophagus) or other gastrointestinal disease which was likely to complicate evaluation of oesophagitis, were all ineligible to enter the study. In addition, patients were not permitted to enter the study if they were taking H2-receptor antagonists, prokinetic agents, proton pump inhibitors, anticholinergics, sucralfate, colloidal bismuth, corticosteroids, anticoagulants or continuous NSAID therapy.


Patients were assigned to treatment with either lansoprazole 30 mg o.d. (Wyeth Lederle, Maidenhead, UK) or omeprazole 20 mg o.d. (Astra, Hässle, Sweden) according to a computer-generated randomization list. Although all patients received lansoprazole 30 mg o.d. for the initial phase of the study, for administrative reasons the randomization was carried out prior to patients entering the initial healing phase of the study. All patients received identical blue opaque capsules which were dispensed to them in treatment packs which were identical apart from the randomization number. Patients were given a 12-week supply of treatment at each assessment and were instructed to take one capsule every morning. The total duration of treatment for each patient was scheduled to be 48 weeks. However, if patients experienced endoscopically verified oesophagitis or symptomatic relapse before this time treatment was curtailed.


The severity of any symptoms of heartburn, dysphagia or regurgitation (graded as none, mild, moderate and severe) and details of adverse events were recorded 12, 24, 36 and 48 weeks after patients entered the maintenance phase of treatment. In addition, endoscopy was carried out after 12, 24 and 48 weeks and at any other time if patients reported symptomatic relapse or aggravation. If patients were observed to have single erosions or oesophagitis of Grade II or worse, regardless of symptoms, they were classified as having a relapse and were withdrawn from the study. Similarly, severe symptoms of heartburn, regurgitation or dysphagia resulted in patients being assigned relapse status and withdrawn from the study.

Any adverse events experienced during treatment were reported, and haematological and biochemical parameters were measured before and at the end of treatment. Although gastric biopsies were taken at the start of the initial healing phase of the study and also at the end of maintenance treatment, no formal histopathological assessment was carried out.

Statistical methods

With an assumption that lansoprazole and omeprazole would demonstrate equivalent relapse rates of 20%, the number of patients per treatment group to show equivalence was calculated to be 112 (80% power and a two-sided significance level of 0.05).

Relapse was defined as being endoscopic (Grade II oesophagitis or single erosions or worse) and/or symptomatic.

Efficacy analyses were carried out on two patient populations: intention-to-treat and per protocol. The intention-to-treat population included all patients who received maintenance treatment, regardless of their original eligibility to enter the study. The per protocol population included only those patients who fulfilled the inclusion criteria but none of the exclusion criteria.

The primary efficacy variable was the time to endoscopic and/or symptomatic relapse for the intention-to-treat population. This was assessed using the life-table extension of the Mantel–Haenszel statistics which examined the number of patients who relapsed during each 12-week time interval throughout the maintenance phase. A patient who withdrew (or was lost to follow-up) during an interval was considered to be at risk for half of that interval. Patients who had no maintenance phase endoscopic or symptomatic assessments were excluded from the effective sample size.

The proportion of patients who relapsed during the entire maintenance period was compared between treatments using Fisher’s exact test. The effect of age (above or below 65 years), sex, investigational centre, smoking, initial oesophagitis grade and severity of heartburn, regurgitation and dysphagia were all investigated as potential prognostic factors. Cochran–Mantel–Haenszel statistics were calculated to assess any impact on relapse and odds ratios and confidence intervals were generated for each of the potential prognostic factors.

A secondary evaluation of efficacy was performed on the severity of the symptoms of heartburn, regurgitation and dysphagia reported at each of the assessments in each treatment group, using the Wilcoxon rank sum test.

Comparison of the treatments with regard to the incidence and severity of adverse events was performed at three levels: individual adverse events recorded in six or more patients, body system categories in which six or more patients recorded an adverse event and an overall patient category of all adverse events. The Wilcoxon rank sum test was used to compare the treatments.


Two hundred and eighty-nine patients were treated with lansoprazole in the acute phase. Of these 245 had healed (Grade 0 or 1 without erosions) after 8 weeks of treatment and a further 17 patients had healed after 12 weeks. Thus, the total number of patients healed after 12 weeks of treatment and eligible for the study was 262 (90.7%). Out of these, 248 patients (lansoprazole n = 126, omeprazole n = 122) continued into the maintenance phase of treatment and were thus included in the intention-to-treat analysis. For various reasons 14 patients did not enter the maintenance phase.

The minimum number of patients entering the maintenance phase per investigational centre was one and the maximum number was 23.

Two hundred and twenty-six patients (lansoprazole n = 111, omeprazole n = 115) were included in the per protocol analysis. Thirty-three patients were excluded from this analysis due to: forbidden medication (lansoprazole n = 18, omeprazole n = 9); missing assessment in the maintenance phase (lansoprazole n = 1, omeprazole n = 1); unhealed oesophagitis after the acute phase (lansoprazole n = 1); and lack of compliance with the study medication (omeprazole n = 3).

Demographic and clinical characteristics of the patients ( Table 1) were similar for the two groups, although a higher proportion of patients in the lansoprazole group had Grade III oesophagitis (26.2%) compared with the proportion in the omeprazole group (18.0%).

Table 1.  . Demographic and clinical characteristics of patients entering maintenance phase (data are given as means ± s.d. when applicable) Thumbnail image of

Time to relapse (endoscopic and/or symptomatic)

The relapse status of patients at each of the four time periods is summarized in Table 2. There was no difference between the two treatments with respect to time to relapse (Mantel–Haenszel, P = 0.95).

Table 2.  . Relapse status at each assessment (all patients) Thumbnail image of

Proportion of patients who relapsed (endoscopic and/or symptomatic)

Only 12 (9.5%) patients in the lansoprazole group had a relapse (11 endoscopic and one symptomatic) and 11 (9.0%) patients in the omeprazole group (nine endoscopic, one symptomatic and one both endoscopic and symptomatic) relapsed (N.S.). Likewise, the per protocol analysis failed to show any significant difference between the treatments (lansoprazole 11/111 (9.9%), omeprazole 9/115 (7.8%), P = 0.645).

In order to improve precision, the effect of potential prognostic factors was assessed by investigating the odds of relapse (relapse: no relapse) after adjusting for the effects of treatment. None of the factors investigated was found to influence relapse in either the intention-to-treat or the per protocol population.

Severity of symptoms

Table 3 summarizes the symptom grades at the last patient visit for both treatments. The treatments were compared with regard to the severity of each of the symptoms at each assessment (i.e. not just the last patient visit), using the Wilcoxon rank sum test. There were no significant differences between the treatment either for the intention-to-treat or the per protocol populations.

Table 3. Symptom grade at last assessment in maintenance phase (number of patients (%))Thumbnail image of


All 248 patients who received maintenance treatment were included in the safety analyses. A total of 299 adverse events were reported by 134 (54.0%) patients during the 48 weeks of treatment. Sixty-six of the 126 (52.3%) patients who received treatment with lansoprazole reported 143 events and 68 of the 122 (55.7%) patients in the omeprazole group experienced 156 events. There was no difference between the treatments for any of the events. The same was true for the incidence of adverse events when categorized by body system. Nine patients (lansoprazole n = 5, omeprazole n = 4) withdrew from the maintenance phase as a result of adverse events. Table 4 describes these events and also shows the relationship to drug treatment as reported by the treating physician. No serious adverse events was reported.

Table 4.  . Adverse events resulting in withdrawal from the study Thumbnail image of


The prevention of relapse of oesophagitis over a long period has only become a realistic possibility following the introduction of proton pump inhibitors. Prior to the use of these drugs for maintenance treatment, H2-receptor antagonists and promotility agents were the only drugs available and the proportion of patients who relapsed was unacceptably high.1[2][3][4][5]–6 Several studies comparing lansoprazole and omeprazole individually with ranitidine have demonstrated the superiority of the proton pump inhibitors.1[2][3][4]–5 However, at the time of commencement of this study there was no direct comparison between the two proton pump inhibitors licensed for long-term therapy.

The present study has demonstrated that lansoprazole 30 mg and omeprazole 20 mg, both given once daily, prevented the majority of patients from relapsing over a 48-week period, and no difference between the drugs was noted. This is in agreement with a recent study, so far only reported in abstract form,7 in which the relative efficacies of lansoprazole 30 mg and 15 mg and omeprazole 20 mg were directly compared in the long-term prevention of oesophagitis. The proportion of patients in whom relapse (Grade I or worse) of oesophagitis was reported was 14% for patients treated with lansoprazole 30 mg and 13% for those treated with omeprazole 20 mg.

The proportion of lansoprazole treated patients in whom relapse did occur in the current study (9.5%) was similar to that reported by Robinson et al.,8 who compared lansoprazole with placebo. In both studies the definition of endoscopic relapse was oesophagitis of Grade II or worse. In contrast, Gough et al.1 reported a relapse in 20% of the patients treated with lansoprazole 30 mg daily for 1 year. However, in this study relapse was defined as oesophagitis Grade I or higher. This is also reflected in the relapse rate of 68% in the control patients treated with ranitidine.

Previous studies investigating the efficacy of omeprazole 20 mg o.d. reported a relapse rate ranging from 11 to 33%.3[4][5]–6 The definition of relapse in all four of these studies was similar to the present study, making comparison of results reasonable. Thus, the results from the current study may appear slightly better than previous reports.

The number of patients who are maintained free of symptoms is important in the clinical situation. Both treatments were effective at preventing the recurrence of all three symptoms. Heartburn was the most frequently reported symptom at the initial presentation and more than three-quarters of the patients in both treatment groups were free of this symptom at their last assessment in the study. The proportion of patients who were free of regurgitation and dysphagia was over 80% and 90%, respectively, for the two treatments. However, fewer patients initially suffered from these symptoms compared with heartburn, and therefore this higher proportion of symptom-free patients could be expected.

The number and overall severity of adverse events reported by patients in each of the treatment groups was similar. The most common complaint was diarrhoea. The number of patients who withdrew from the study as a result of adverse events was low (3.6%) and equally distributed between the two treatment groups.


Lansoprazole 30 mg o.d. and omeprazole 20 mg o.d. are equally effective in preventing the relapse of oesophagitis over a 48-week period. Over 90% of the patients were in remission in the two treatment groups. Both treatments exhibit a similar side-effect profile.


This study was supported by Wyeth Ayerst and Wyeth Lederle Nordiska AB.