SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendix

Aliment Pharmacol Ther 2011; 33: 213–224

Summary

Background  The efficacy of rabeprazole 5 mg/day for patients with non-erosive reflux disease (NERD) has not been reported in the literature.

Aim  To evaluate the efficacy of rabeprazole 5 mg and 10 mg/day in Japanese NERD patients. The influence of baseline characteristics as well as genetic background on efficacy was also analysed.

Methods  Subjects were grade M (minimal changes) NERD patients. Two hundred and eighty-eight of these subjects, who were nonresponders to open label antacid therapy, entered in a 4-week, double-blind treatment (placebo, rabeprazole 5 mg or 10 mg/day).

Results  Complete heartburn relief rates were 21% in placebo, 34% in rabeprazole 5 mg and 44% in rabeprazole 10 mg (5 mg vs. placebo = 0.074, 10 mg vs. placebo = 0.001). Rabeprazole 5 mg was significantly more effective than placebo in elderly patients and in patients with low heartburn frequency or without hiatal hernia. The efficacy of rabeprazole 10 mg was not influenced by age, BMI, hiatal hernia, Helicobacter pylori infection, frequency and severity of heartburn or CYP2C19 genotypes.

Conclusions  Rabeprazole 5 mg was effective in a subgroup of Japanese NERD patients. Rabeprazole 10 mg provided more potent heartburn relief than 5 mg and was less fragile to baseline characteristics.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendix

Gastro-oesophageal reflux disease (GERD) is defined as ‘a condition which develops when the reflux of stomach contents causes troublesome symptoms and/or complications’.1 Within the GERD population, a subset with troublesome reflux-associated symptoms in spite of the absence of mucosal breaks is classified as having non-erosive reflux disease (NERD).2–4

Non-erosive reflux disease constitutes approximately 50–70% of GERD population,3, 4 and there is considerable evidence that quality of life (QOL) is seriously impacted according to reports primarily from the United States and Europe.5 Similar reports have emerged recently in Japan. Mishima et al. reported a prevalence of NERD of 11% and of erosive oesophagitis (EO) of 9% in the investigation of 2760 subjects, clearly a higher prevalence of NERD relative to EO.6 A QOL assessment in Japanese NERD patients using SF-36 (Japanese version) by Joh et al. demonstrated a lower QOL relative to the Japanese general population in all of the eight domains.7

Patients with NERD differ in demographic characteristics from patients with EO. Although percentages of men, older age, obesity, hiatal hernia and negative Helicobacter pylori (H. pylori) are high in populations of EO patients, these percentages are lower in NERD.3, 4, 8 Gastric acid is the main cause of NERD.9 However non-acidic gastro-oesophageal reflux, including air; contractile abnormalities of the oesophagus and oesophageal hypersensitivity10–12 appear to have a role in symptom generation of NERD patients. In addition, heartburn symptoms are increased and/or more frequent during times of psychological and emotional stress.13 NERD is clearly a different clinical entity from EO hence the need for a different therapeutic approach to treatment.14

As NERD is essentially a symptomatic disease, it is uncommon to observe complications seen in EO like bleeding and stricture. Therefore, the main aim of therapy in NERD is to relieve symptoms and improve QOL. Current guidelines continue to suggest the use of proton pump inhibitors (PPIs) for the treatment of NERD as first-line agents.15–17

Among PPIs, rabeprazole (RPZ) provides fast and potent relief from heartburn and other symptoms of GERD patients.18–22 Also, RPZ metabolism is thought to be primarily non-enzymatic compared with omeprazole and lansoprazole, and RPZ appears to be the least affected by CYP2C19 genotypes.23–25 According to a double-blind, placebo-controlled, randomised study in the United States, efficacies of RPZ 10 mg/day and 20 mg/day in NERD were comparable, and a dose of 10 mg/day was sufficiently effective.26 In view of findings of lower gastric acid secretion and less frequent and severe heartburn in Japanese patients compared with Westerners,27–30 it could be hypothesised that lower doses of RPZ may be potentially efficacious. Hence, the rationale to investigate the efficacy of RPZ 10 mg (standard dosage for peptic ulcer and EO in Japan) and the lower dose of 5 mg/day for NERD treatment in Japanese patients. We also analysed how baseline characteristics may affect efficacy in this population.

We report here a multicentre, randomised, double-blind, placebo-controlled study to investigate the efficacy and safety of RPZ 10 mg/day and 5 mg/day for 4 weeks in Japanese patients with NERD.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendix

Study design

This Phase III study was a multicentre (35 centres, 36 departments in Japan; see Appendix 1), randomised, double-blind, placebo-controlled, comparative study to investigate the efficacy and safety of treatment with RPZ at 5 mg or 10 mg once daily for 4 weeks in patients with NERD, endoscopic grade M (defined as ‘Erythema without sharp demarcation, whitish turbidity, and/or invisibility of vessels due to these findings’) by modified-Los Angeles (LA) Grade.31, 32 The grade M NERD was selected as the target of this study as patients with grade M were shown to have slightly higher gastro-oesophageal acid reflux in our previous study.7 The study was conducted between September 2004 and October 2005. This study was performed in compliance with the ethical principles based on the ‘Declaration of Helsinki’ and subsequent revisions and with good clinical practice (GCP), and it was registered with clinicaltrials.gov under the number NCT00165646. Prior to commencement of this study, ethical and scientific validity were reviewed and approved by the institutional review board of each study centre.

This study consisted of a 1- or 2-week observation period (open label administration of antacid) and a 4-week double-blind treatment period (Figure 1). Third-party organisation (Bellsystem24, Inc., Tokyo, Japan.) randomly created key code of study drug (six patients/set) and kept the code until the public key to maintain the blindness.

image

Figure 1.  Study design. The study is divided into open label observational period and double-blind treatment period. † In consideration of the circumstances involved in the patient coming to the hospital, up to 2 weeks was permitted. ‡ Patients whose heartburn improved from antacid administration during the observation period (less than 2 days of heartburn during the 7 days immediately before the start of treatment) could not be entered into the treatment period.

Download figure to PowerPoint

Patients’ enrolment

Subjects were out-patients ≥20 years old who signed written informed consent. Inclusion criteria for the entry of observational period were male or female patients who had ≥2 days/week episodes of ‘heartburn’ for three consecutive weeks immediately before the screening visit. Heartburn was defined as a burning sensation arising from the stomach or lower chest and tending to appear frequently or becoming aggravated after eating, when bending forward and/or when pressing on the abdomen. Endoscopy was performed within 14 days prior to beginning of the observation period without any medication that will influence reflux symptom including PPI and antidepressant or anxiolytic agent. Patients not responding to antacids during the observation period entered the treatment phase; i.e. patients who had ‘heartburn’≥2 days/week in the 7 days immediately before the treatment period were enrolled.

Patients meeting any of the following criteria were excluded; (i) major complaints of ‘feelings of heavy stomach’ and/or ‘abdominal distension’, (ii) with complication or history of psychiatric/psychosomatic disease (e.g. manic-depressive psychosis, obsessive-compulsive neurosis, or others) or receiving antidepressant or anxiolytic pharmacotherapy (except when used as a hypnotic), (iii) having undergone H. pylori eradication therapy, and being treated for H. pylori eradication therapy within 6 months, (iv) with active gastric ulcer, active duodenal ulcer or acute gastritis, (v) with a history of any surgical intervention affecting gastric acid secretion (e.g. upper gastrointestinal tract resection and/or vagotomy), (vi) with Barrett’s oesophagus, oesophageal stenosis, pyloric stenosis or scleroderma, (vii) with a history or presence of angina pectoris, (viii) working at night, (ix) receiving PPIs within 3 weeks prior to screening visit, (x) needing nonsteroidal anti-inflammatory drugs (NSAID) (excluding topical preparations), steroids (excluding topical preparations) and/or aspirin treatment every day, (xi) with serious complications such as cardiovascular disease (e.g. myocardial infarction), haematological disorder (e.g. aplastic anaemia), renal disease (e.g. acute or chronic renal failure), hepatic disease (e.g. cirrhosis) or malignant tumour, (xii) with known hypersensitivity to antacids or PPIs, (xiii) being pregnant, child-bearing potential, wishing to become pregnant or lactating during the study period, (xiv) receiving another investigational drug or who had received another investigational drug within 6 months prior to screening visit and (xv) being judged to be ineligible for the study entry by the investigators or subinvestigators.

Study protocol

At the first screening visit, baseline information was obtained. Eligibility was determined for all patients. Eligible patients entered an observation period and maintained a daily diary of heartburn symptoms. During the observation period, patients were orally administered with antacid 1.2 g (dried aluminium hydroxide gel – magnesium hydroxide ‘Maaredge suspension’; Towa Pharmaceutical, Osaka, Japan) three times/day after each meal. Patients were requested to record the occurrence of heartburn during the daytime (from the time of waking up in the morning to the time of going to bed) and nighttime (from the time of going to bed to the time of waking up in the morning) in the diary. The following patient-defined 4-point scale was used to assess the severity of heartburn symptoms: No heartburn. Mild: Heartburn was painful but bearable (e.g. Heartburn was uncomfortable but did not markedly interfere with daily activities, although it did slightly delay the time of falling asleep). Moderate: Heartburn was painful (e.g. Heartburn was uncomfortable, interfering with daily activities and awaking the subject from sleep). Severe: Heartburn was very painful (e.g. Heartburn was very uncomfortable, significantly interfering with daily activities and making it difficult to sleep).

At the second visit (at the start of the treatment period), patients who clear inclusion/exclusion criteria were assigned to double-blind treatment with RPZ 5 mg, RPZ 10 mg, or placebo. Baseline assessments (including measurement of serum anti-H. pylori antibodies and determination of CYP2C19 genotype) were performed. H. pylori status was assessed using a serum antibody test (SRL Medisearch, Tokyo, Japan). CYP2C19 genotypes were tested by PCR-RFLP at the start of the treatment period (SRL Medisearch). Results were recorded as ‘homozygous Extensive Metabolizer (EM)’; ‘heterozygous EM’ or ‘Poor Metabolizer (PM)’.

All treatments were administered once daily after breakfast for 4 weeks. Patients were given study medication and requested to continue completing the heartburn diary continuously as described above, and then returned twice after double-blind randomisation for efficacy and safety assessments (at week 2 and 4).

The following drugs (including over-the-counter drugs) were not allowed to be used throughout the study: (i) Prohibited concomitant drugs that might affect evaluation of the treatment effects of RPZ, including other PPIs, H2RAs, prokinetics, mucosal protective agents, antacids, sodium alginate, protease inhibitors, anticholinergic agents, cholinergic drugs, local anaesthetics for the mucosal membrane of the digestive tract, antigastrin agents, prostaglandin preparations, M1-receptor antagonists, antiemetics, antidepressants and anxiolytics (can be used as hypnotic agents). Helicobacter pylori eradication therapy was not allowed; (ii) Drugs contraindicated because of known interactions with the study drugs, including digitalis preparations and antacids containing aluminium hydroxide or magnesium hydroxide gels. Medications for complications were allowed based on the judgment of the investigators/subinvestigators, but in principle, the dosage and administration method were not allowed to be changed during the study.

Assessments of adverse events were performed during treatment. Adverse drug reactions were defined as adverse events for which a causal relationship could not be ruled out. Laboratory testing was taken at the first screening visit, at the second visit, and at the end of the treatment period.

Endpoints

Primary endpoint of this trial was the complete heartburn relief rate at the final evaluation. This was defined as ‘no episodes of heartburn for 7 days immediately before evaluation’ and the final evaluation showed ‘after 4 weeks or when treatment was discontinued’. Secondary endpoints were (i) complete heartburn relief rate at 2 and 4 weeks after initiation of treatment, (ii) satisfactory heartburn relief rate at 2 and 4 weeks after initiation of treatment and the final evaluation defined as ‘no more than 1 period (the daytime or nighttime) of heartburn during 7 days (14 periods) immediately before evaluation’, (iii) percentage of heartburn-free days, (iv) time (in days) to first 24-h heartburn-free interval (no heartburn for two consecutive periods) after initiation of treatment. We also performed subgroup analysis of background factors (baseline characteristics) affecting the complete heartburn relief rate, including age (<40 years old/≥40 years old), BMI (kg/m2) (<25/≥25), hiatal hernia (presence/absence), H. pylori (positive/negative), frequency and severity of heartburn at initiation of the observational period and CYP2C19 genotypes. Finally, the daily rates of patients who achieved ‘real-time heartburn resolution’ were provided by a line graph to show the change of patients without heartburn in three consecutive days. As patients who had ≥2 days/week episodes of ‘heartburn’ were enrolled in this study, it was considered that no heartburn for three consecutive days was the clinically meaningful indicator as the sign of better control, or at least the steady state, of symptom than the time of study enrolment.

Statistical analysis

According to the hypothesis that the complete heartburn relief rate in RPZ 10 mg group is 25% higher than the rate of placebo group (10–30%), sample size at the treatment period of 300 patients (100 patients/group) will have 90% power to give RPZ 10 mg superiority to placebo using Steel test with two-sided 5% level of significance. The main population for analysis was the full analysis set (FAS). For patients’ background, summary statistics were calculated for continuous variables, and composition ratios and frequency were provided for classified variables and ordinal variable. Uniformity among groups was confirmed by one-way analysis of variance, Chi-squared test and Kruskal–Wallis test. The primary endpoint was the comparison between the placebo group and study drug (RPZ) groups using Steel test. Subgroup analysis to compare the placebo group and RPZ groups based on patient background factors (baseline characteristics) were conducted as well. For secondary endpoints, following analysis were conducted while descriptive statistics was conducted; Steel test for periodical complete heartburn relief rate and satisfactory heartburn relief rate, Dunnett’s test for Percentage of heartburn-free periods, logrank test with contrast ‘−1 0 1’ and ‘−1 1 0’ for time to achieve first 24-h heartburn-free interval, Dunnett’s test was also used for rates of patients with real-time heartburn resolution (no heartburn on evaluation day and ±1 days). For safety analysis, frequency and incidence were calculated, and comparison between the placebo group and RPZ groups was performed using Fisher’s exact test.

Significance levels in the tests were set as follows: two-sided 15% for uniformity between groups, two-sided 5% for intergroup and intragroup comparisons. All analyses were performed with the statistical software sas release 8.2 (SAS Institute Inc., Cary, NC, USA).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendix

Patient disposition and demographics

As shown in Figure 2, number of patients registered during the observational period was 332. Among those registered patients, 44 patients were discontinued before randomisation, and 288 patients were enrolled in the treatment period. Among those 288 patients, 285 patients constituted the FAS for efficacy.

image

Figure 2.  Randomisation protocol and patient disposition. *Reasons to be excluded from analysis: (i) Not administered a study drug, no heartburn data (= 2). (ii) Inappropriate subject (= 1). (iii) Violation of exclusion criteria (= 1). (iv) Violation of concomitant drugs (= 2). (v) Violation of usage and dosage (noncompliance included) (= 10).

Download figure to PowerPoint

The demographics of patients (FAS) are summarised in Table 1. No statistically significant differences (P < 0.15) were observed in placebo (= 91), RPZ 5 mg (= 93) and RPZ 10 mg (= 101) groups.

Table 1.   Demographic and clinical characteristics of patients enrolled (full analysis set)
ItemPlacebo (= 91)RPZ 5 mg (= 93)RPZ 10 mg (= 101)P-value
  1. RPZ, rabeprazole; s.d., standard deviation; min, minimum; max, maximum; H. pylori, Helicobacter pylori; EM, extensive metabolizer; PM, poor metabolizer.

  2. † Chi-squared test.

  3. ‡ One-way analysis of variance.

  4. § Kruskal–Wallis test.

Gender
 Male40 (44%)38 (41%)50 (50%)0.469†
 Female51 (56%)55 (59%)51 (51%) 
Mean Age (years) ± s.d. (min–max)49.7 ± 17.0 (23–84)46.3 ± 15.3 (22–86)46.8 ± 15.7 (21–77)0.285‡
Mean BMI (kg/m2) ± s.d. (min–max)23.2 ± 3.5 (17.2–35.7)22.4 ± 3.2 (16.4–31.1)23.2 ± 3.4 (17.0–34.8)0.176‡
Frequency of heartburn at initiation of the observation period (mean days/week ± s.d.)4.8 ± 2.04.5 ± 2.04.5 ± 1.90.490‡
 2–3 days37 (41%)48 (52%)50 (50%)0.358§
 4–5 days14 (15%)9 (10%)15 (15%) 
 6–7 days40 (44%)36 (39%)36 (36%) 
Severity of heartburn at initiation of the observation period
 Mild45 (50%)51 (55%)49 (49%)0.699§
 Moderate43 (47%)38 (41%)48 (48%) 
 Severe3 (3%)4 (4%)4 (4%) 
Anti-H. pylori IgG antibodies (EIA method)
 (−)45 (50%)52 (56%)55 (55%)0.491§
 (±)2 (2%)6 (7%)4 (4%) 
 (+)44 (48%)35 (38%)42 (42%) 
CYP2C19 genotypes
 Homo EM29 (32%)34 (37%)32 (32%)0.942§
 Hetero EM50 (55%)40 (43%)52 (52%) 
 PM12 (13%)19 (20%)17 (17%) 

Efficacy analysis

The results of efficacy analysis (FAS) are summarised below.

Complete relief of heartburn at the final evaluation.  The complete heartburn relief rates at the end of the treatment period (FAS) were 21% in the placebo group, 34% in the RPZ 5 mg and 44% in the RPZ 10 mg group, respectively (Figure 3). The complete heartburn relief rate of the RPZ 10 mg group was significantly higher than the placebo group, but no statistical significant difference was seen between the RPZ 5 mg group and the placebo group. In the PPS, the results of the placebo, RPZ 5 mg and RPZ 10 mg groups were 21%, 35% and 43%, respectively, very similar to the results in the FAS. Superiority of the RPZ 10 mg group to the placebo group was achieved (= 0.003).

image

Figure 3.  Final complete and satisfactory heartburn relief by the administration of rabeprazole. Data are obtained from the full analysis set.

Download figure to PowerPoint

Complete heartburn relief rate (at week 2 and 4 during the treatment period).  Complete heartburn relief rate at week 2 and 4 during the treatment period were similar to the results at the final evaluation. The efficacy of RPZ 10 mg was already significantly superior to placebo at week 2, and continuous improvement was observed at week 4. On the other hand, no significant differences were observed between the placebo group and the RPZ 5 mg group (Table 2).

Table 2.   Summary of efficacy results (full analysis set)
ParameterPlacebo (= 91)RPZ 5 mg (= 93)RPZ 10 mg (= 101)P-value Placebo vs. RPZ 5 mg Placebo vs. RPZ 10 mg
  1. RPZ, rabeprazole; s.d., standard deviation; S.E., standard error.

  2. < 0.05.

  3. † Steel test.

  4. ‡ Dunnett’s test.

  5. § Log rank test.

  6. ¶ Change from 7 days immediately before the start of treatment period (mean ± S.E.).

2 weeks complete heartburn relief10% (9/89)20% (18/91)28% (28/100)= 0.124† = 0.003*†
4 weeks complete heartburn relief21% (19/89)35% (31/88)44% (43/98)= 0.074† = 0.002*†
2 weeks satisfactory heartburn relief24% (21/89)33% (30/91)44% (44/100)= 0.276† = 0.006*†
4 weeks satisfactory heartburn relief35% (31/89)50% (44/88)56% (55/98)= 0.076† = 0.006*†
Rate of heartburn free period (mean ± s.d.)45 ± 30% (19 ± 3%¶)57 ± 31% (32 ± 3%¶)61 ± 30% (34 ± 3%¶)= 0.004*‡ < 0.001*‡
Median time to first 24-h heartburn free interval (95% CI)3.0 days (1.0–5.0)1.0 day (1.0–3.0)1.0 day (0.0–1.0)= 0.035*§ = 0.004*§

Satisfactory heartburn relief rate (at the final evaluation, week 2 and 4).  Satisfactory heartburn relief rates at the final evaluation were 34% in the placebo group, 48% in the RPZ 5 mg and 55% in the RPZ 10 mg group, respectively (Figure 3). Satisfactory heartburn relief rate of the RPZ 10 mg was significantly higher than the rate of the placebo group. No significant difference between the RPZ 5 mg group and the placebo group was shown.

Similar to the results of the final evaluation, RPZ 10 mg showed significantly better efficacy in satisfactory heartburn relief rates at week 2 and week 4 compared with placebo. The improvement was continuous after week 2. On the other hand, no significant differences were observed between the placebo group and the RPZ 5 mg group (Table 2).

Percentage of heartburn-free periods.  The mean rates of heartburn-free period between day 1 and the end of the treatment period were 45% in the placebo group, 57% in the RPZ 5 mg group and 61% in the RPZ 10 mg group, respectively. The mean variations were 19%, 32% and 34%, respectively. Both RPZ 5 mg and 10 mg were superior to placebo (Table 2).

Time to first 24-h heartburn-free interval.  Median time to achieve first 24-h heartburn-free interval were 3.0 in the placebo group, 1.0 in the RPZ 5 mg group and 1.0 in the RPZ 10 mg group, respectively. The interval periods to heartburn free status for the RPZ 5 mg and 10 mg groups were significantly shorter than the placebo group (Table 2).

Subgroup analysis of background (baseline) factors influencing the complete heartburn relief rate at the final evaluation.  The complete heartburn relief rates at the final evaluation (FAS) based of the baseline characteristics were summarised in Table 3.

Table 3.   Final complete heartburn relief rate based on patient background (full analysis set)
CovariateClassificationPlacebo (= 91)RPZ 5 mg (= 93)RPZ 10 mg (= 101)P-value (Steel test) Placebo vs. RPZ 5 mg Placebo vs. RPZ 10 mg
  1. RPZ, rabeprazole.

  2. < 0.05.

Age (years)<409% (3/34)11% (4/36)33% (14/42)= 0.927 = 0.021*
 ≥4028% (16/57)49% (28/57)51% (30/59)= 0.040* = 0.023*
BMI (kg/m2)<2519% (13/68)32% (23/73)40% (30/76)= 0.162 = 0.015*
 ≥2526% (6/23)45% (9/20)56% (14/25)= 0.333 = 0.069
Hiatal Hernia(−)20% (13/66)39% (24/61)45% (32/71)= 0.028* = 0.003*
 (+)24% (6/25)25% (8/32)40% (12/30)= 0.994 = 0.345
Anti-H. pylori IgG antibodies (EIA method)(−)(±)17% (8/47)35% (20/58)36% (21/59)= 0.080 = 0.061
 (+)25% (11/44)34% (12/35)55% (23/42)= 0.575 = 0.009*
Frequency of heartburn at initiation of the observation period (days/week)2–4 days21% (10/48)42% (22/52)44% (27/62)= 0.040* = 0.023*
 5–7 days21% (9/43)24% (10/41)44% (17/39)= 0.902 = 0.053
Severity of heartburn at initiation of the observation periodMild22% (10/45)33% (17/51)47% (23/49)= 0.374 = 0.023*
 Moderate/severe20% (9/46)36% (15/42)40% (21/52)= 0.160 = 0.048*
CYP2C19 genotypesHomo EM24% (7/29)32% (11/34)44% (14/32)= 0.692 = 0.190
 Hetero EM24% (12/50)38% (15/40)44% (23/52)= 0.285 = 0.059
 PM0% (0/12)32% (6/19)41% (7/17)= 0.058 = 0.022*

The complete heartburn relief rate of RPZ 10 mg was not impacted by age (<40 years old/≥40 years old) or severity (mild, moderate and severe) of heartburn, and was significantly superior to placebo. When stratified by other background factors, i.e. BMI (<25/≥25), hiatal hernia (presence/absence), H. pylori (positive/negative), frequency (2–4 days/week, 5–7 days/week) of heartburn, CYP2C19 genotypes (homozygous EM, heterozygous EM, PM), the complete heartburn relief rate of RPZ 10 mg was not significant in all sub-groups. However, the therapeutic gain was 15% or higher in each sub-group compared with placebo group.

On the other hand, the complete heartburn relief rate of RPZ 5 mg was significantly superior to placebo only in a partial sub-group, i.e. the ≥40 years old, the group negative for hiatal hernia and frequency (2–4 days/week) of heartburn. In the all background factors, the complete heartburn relief rate of RPZ 5 mg was lower than that of RPZ 10 mg.

Real-time heartburn resolution.  Rates of patients with real-time heartburn resolution (FAS) were recorded daily (Figure 4).

image

Figure 4.  Daily rates of patients achieving real-time heartburn resolution. Real-time heartburn resolution was defined as ‘no heartburn on evaluation day and ±1 days’.

Download figure to PowerPoint

The rates of the placebo, RPZ 5 mg and RPZ 10 mg groups on day 8 (day 7–day 9 evaluation) of the treatment period were 21%, 33% and 42%, respectively. The rates on day 15 (day 14–day 16 evaluation) were 32%, 48% and 49%, respectively. The rates on the final evaluation were 43%, 53% and 65%, respectively. Efficacy of RPZ 10 mg was superior to placebo at every evaluation time point described above (= 0.005, = 0.028 and = 0.007). RPZ 5 mg was superior to placebo on day 15 evaluation, but did not show significant difference at day 8 and the final evaluation (= 0.156, = 0.040 and = 0.334).

Safety analysis

Incidence (after administration of study drug in the treatment period) of adverse events in 286 patients in safety analysis was 35% (32/91, 57 events) in the placebo group, 34% (32/93, 56 events) in the RPZ 5 mg and 36% (37/102, 69 events) in the RPZ 10 mg group, respectively. Incidence of adverse drug reactions were 11% (10/91, 14 events) in the placebo group, 12% (11/93, 14 events) in the RPZ 5 mg and 12% (12/102, 15 events) in the RPZ 10 mg group, respectively. No significant differences were observed in the incidence of adverse events or adverse drug reactions.

Adverse drug reactions with incidences ≥2% were diarrhoea (two patients, two events, 2%) and increase in lactate dehydrogenase (two patients, two events, 2%) in the placebo group, abdominal bloating (two patients, two events, 2%) and dry mouth (two patients, two events, 2%) in the RPZ 5 mg group, and constipation (three patients, three events, 3%) and abdominal bloating (two patients, two events, 2%) in the RPZ 10 mg group.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendix

The pathogenesis of NERD is believed to be multifactorial. Causes such as hypersensitivity of the oesophagus, abnormal contractility of the oesophagus, psychological factors are thought to be contributory;10–13 however, acid reflux into the oesophagus appears to be the main causative factor like in EO.9 Therefore, PPIs, currently the most potent acid inhibitors with known sustainability of efficacy, are used as the first-line agents.15–17

Rabeprazole was previously approved for the symptomatic GERD relief indication outside of Japan, and the doses of 20 mg/day in the United States and 10 mg/day in Europe are widely used.33 In Japan, the gastric acid secretion level of the indigenous population is lower than the American and European populations.27–29 The frequency and severity of heartburn are also milder in the Japanese population.30 Therefore, a lower than the standard dose of RPZ 10 mg for EO in Japan would be expected to be effective for the NERD indication.

Thus, we assessed the efficacy of RPZ 5 mg and 10 mg for Japanese patients with NERD in a multicentre, randomised, double-blind, placebo-controlled, comparative study. In Japan, a modified-LA Classification31, 32 is widely used as reflux oesophagitis classification system. Grade M is defined as ‘erythema without sharp demarcation, whitish turbidity, and/or invisibility of vessels due to these findings’, whereas the absence of changes in the oesophageal mucosa is classified as grade N. To increase the likelihood of enrolling patients with symptoms related to increased acid secretion, the present study was conducted with patients only with grade M with slightly higher gastro-oesophageal acid reflux as shown in our previous publication.7

In the present 4-week study of RPZ in Japanese NERD patients, complete heartburn relief rate of RPZ 10 mg was significantly higher than placebo (44% vs. 21%). The therapeutic gain in the present study was 23%, which is similar to the one observed in the meta-analysis reported by Dean et al. (37% vs. 10%; therapeutic gain, 27%) in the non-Japanese population.34 On the other hand, the therapeutic gain of RPZ 5 mg was 14%. However, the present study had a high placebo effect. This relatively high placebo effect was possibly due to the different study design between the present study and the study reported by Miner et al.26 In the study conducted by Miner et al., placebo was given by the single-blind method during the observation period, and patients responding to placebo were excluded. On the other hand, in our study, antacid was given by the open-label method during the observation period and patients with mild NERD were excluded. Therefore, in this study, more patients responding to placebo were carried forward to the treatment phase as compared with the case in the studies from the west, possibly resulting in the higher placebo effect. The lower frequency and milder severity of heartburn in the NERD patients enrolled in the study might also have some impact on the placebo effect. In addition, infection rate of H. pylori in the present study (43%) was slightly higher than the rate in the Western nations (30–33%).35

In this study, 332 patients were enrolled during the observation period. However, 44 patients could not be carried forward to the treatment phase on account of various criteria, including improvement of the symptoms by antacid use. Therefore, only 285 patients, short of the target number of 300, could be included in the FAS. However, number of patients in withdrawn from the study after being carried forward to the treatment phase was five in the placebo group, five in the RPZ 5 mg group and four in the RPZ 10 mg group, with no significant differences in the number among the three groups. In addition, considering the drug efficacy obtained in the RPZ 5 mg and placebo groups (complete relief rate 34% and 21%, respectively), the smaller number of patients in the FAS than the target of 300 patients may not have significantly influenced the conclusion that only RPZ 10 mg was significantly superior to placebo.

As previously reported by Adachi et al.,19 Robinson et al.20 and Holtmann et al.,21 a quick onset of heartburn relief is one of the main advantages of RPZ. A recent study reports that fast onset of efficacy is important for patient satisfaction.36 In this manner, early management of disease is very important thing in treatment of NERD. Median time duration to achieve first 24-h heartburn-free interval by RPZ 10 mg in the present study were 1.0 day and real-time heartburn resolution confirms the rapid onset of effective relief in Japanese NERD patients. An on-demand approach to NERD with 10 mg of RPZ has been documented as superior to placebo. Bytzer et al. reported that 94% of patients could be controlled by RPZ 10 mg on-demand therapy in the 6-month observational study in patients whose heartburn was relieved by an initial treatment.37

Although no significant differences in the complete heartburn relief rates were confirmed between the RPZ 5 mg group and the placebo group, superiority of RPZ 5 mg was achieved in the secondary endpoints nominally in percentage of heartburn-free periods and median days to first 24-h heartburn. Subgroup analyses also revealed that RPZ 5 mg was superior to placebo for certain subpopulation such as ≥40 years old, low frequency (2–4 days/week) of heartburn and absence of hiatal hernia. This indicates the possibility of early onset and stable efficacy by low dose of RPZ 5 mg/day when patients are selected appropriately.

In conclusion, low dose 5 mg of RPZ showed some trend towards heartburn relief in a subgroup of Japanese patients with NERD. Meanwhile, RPZ 10 mg suggested more potent heartburn relief than 5 mg and was less fragile to baseline characteristics. The results from this study demonstrate that RPZ 10 mg is an appropriate dosage for Japanese NERD patients.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendix

Declaration of personal interests: Yoshikazu Kinoshita has served as a consultant for Eisai Co., Ltd., Astellas Pharma Inc., AstraZeneca Pharmaceuticals, Taiho and Takeda Pharmaceutical Company Ltd. Kiyoshi Ashida has served as a consultant for Eisai Co., Ltd., AstraZeneca Japan, and Takeda Pharmaceutical Co., Ltd. Michio Hongo has served as a consultant for Eisai Co., Ltd., Abbott Japan Co., Ltd., Astellas Pharma Inc., AstraZeneca Japan, Asubio Pharma Co., Ltd., Dainippon-Sumitomo Pharma Co., Ltd., Kyowa Hakko Co. Ltd, Sucampo Pharmaceuticals, Inc., Takeda Pharmaceutical Co., Ltd., and Zeria Pharmaceutical Co., Ltd. Valuable advice was given by Suminobu Ito, National Hospital Organization, Tokyo Japan, regarding the plan and operation of the present clinical trial and by Chikuma Hamada, Tokyo University of Science, Tokyo, Japan, regarding medical statistics. Authors also thank the sponsor, Clinical Research Center of Eisai Co., Ltd., Tokyo, Japan and the project members of EPS Co., Ltd., Tokyo, Japan for their contribution in the study monitoring. Declaration of funding interests: This study was funded by Eisai Co., Ltd., Tokyo, Japan. The writing and preparation of this paper were funded in part by Clinical Research Center of Eisai Co., Ltd., Tokyo, Japan. Writing support was provided by Eisai Global Clinical, Woodcliff Lake, NJ, USA. Data analyses were undertaken by Clinical Research Center of Eisai Co., Ltd., Tokyo, Japan.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendix

Appendix

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Appendix

Appendix 1

Members of the Japan rabeprazole study group for NERD

Hiroshi Takeda, Mototsugu Kato, Yuichi Shimizu and Soichi Nakagawa from Hokkaido University Hospital, Sapporo, Japan; Ken Nishi from Sapporo Medical Center NTT EC, Sapporo, Japan; Masao Yamauchi from Yamauchi Clinic, Sapporo, Japan; Yoshiki Eda from Sendai City Hospital, Sendai, Japan; Shuichi Ohara, Tomoyuki Koike, Katsunori Iijima and Yasuhiko Abe from Tohoku University Hospital, Sendai, Japan; Motoyasu Kusano, Osamu Kawamura and Yasuyuki Shimoyama from Gunma University Hospital, Maebashi, Japan; Chuichi Sekine from Saiseikai Kawaguchi General Hospital, Kawaguchi, Japan; Jun Miwa from Toshiba General Hospital, Tokyo, Japan; Naomi Uemura and Junichi Akiyama from International Medical Center of Japan, Tokyo, Japan; Takahisa Furuta and Naohito Shirai from Hamamatsu University School of Medicine, Hamamatsu, Japan; Yasumasa Niwa from Nagoya University Graduate School of Medicine, Nagoya, Japan; Takashi Joh, Hiromi Kataoka, Makoto Sasaki, Tsuneya Wada, Satoshi Tanida and Tadayuki Oshima from Nagoya City University Graduate School of Medical Sciences, Nagoya, Japan; Kunio Kasugai, Toshihiro Konagaya and Yoshifumi Tokura from Aichi Medical University School of Medicine, Nagakute, Japan; Shoji Mitsufuji from Graduate School of Medical Science, Kyoto Prefectural University of Medicine, Kyoto, Japan; Hiroshi Senoo from Graduate School of Medicine, Kyoto University, Kyoto, Japan; Yasuhiro Fujiwara from Osaka City University Graduate School of Medicine, Osaka, Japan; Kiyoshi Ashida, Takumi Fukuchi from Saiseikai Nakatsu Hospital, Osaka, Japan; Yasuki Habu from Saiseikai Noe Hospital, Osaka, Japan; Masahiro Sakaguchi from Moriguchi Keijinkai Hospital, Moriguchi, Japan; Kazuhiko Inoue from Matsue Red Cross Hospital, Matsue, Japan; Kyoichi Adachi and Kenji Furuta from Shimane University Faculty of Medicine, Izumo, Japan; Akira Tari, Hiroshi Tani and Yoshiki Ishii from Hiroshima Red Cross Hospital & Atomic-bomb Survivors Hospital, Hiroshima, Japan; Tomohiko Shimatani and Masaki Inoue from Department of General Medicine, Hiroshima University Hospital, Hiroshima, Japan; Shinji Tanaka, Yasuhiko Kitadai, Masanori Ito, Toru Hiyama, Shigeto Yoshida, Shiro Oka, Noriaki Manabe and Yoshitaka Ueno from Department of Endoscopy, Hiroshima University Hospital, Hiroshima, Japan; Tomoharu Yoshida, Shomei Ryozawa and Toshiya Harada from Yamaguchi University Graduate School of Medicine, Ube, Japan; Tadasu Fuji from Fuji GI Clinic, Ube, Japan; Kazuo Harima from Harima Gastroenterologic Medicine, Sanyo Onoda, Japan; Atsushi Murakami from Murakami Clinic, Ube, Japan; Shigeo Nishimura from Nishimura Internal Medicine Clinic, Ube, Japan; Susumu Kawamura from Kawamura Internal Medicine, Sanyo Onoda, Japan; Nobutoshi Kuniyoshi and Kazushige Kuniyoshi from Kuniyoshi Hospital, Kochi, Japan; Toshiyuki Matsui from Fukuoka University Chikushi Hospital, Chikushino, Japan; Takemi Noda from Kasumigaoka Clinic, Fukuoka, Japan; Toru Umezu from Umezu Clinic, Chikushino, Japan; Fumitaka Koganemaru from Koganemaru Clinic, Kitakyushu, Japan; Kazuma Fujimoto, Ryuichi Iwakiri from Saga Medical School, Saga, Japan.