This was an active-controlled, double-blinded, double-dummy, randomized, single-center study with a parallel group designed to enroll 48 EE patients. Forty-eight outpatients (M/F: 38/10, age: 54.1±17.8 years), who sought medical care because of typical GERD symptoms for at least 1 mo, were consecutively enrolled in the study. They all received an endoscopy to confirm the EE diagnosis according to Los Angeles (LA) grading system[18], and met an inclusion criterion for an 8-wk treatment period with either esomeprazole tablet 40 mg or omeprazole capsule 20 mg (AstraZeneca, Gothenburg, Sweden). While endoscopy specimens were simultaneously obtained from stomach antrum and body for a rapid urease test to determine Helicobacter pylori (H pylori) infection. Those subjects with the following conditions were excluded: coexistence of healed or active peptic ulcer, gastrointestinal malignancy, esophago-gastric surgical history, esophagitis obviously resulted from systemic diseases, infections, drugs, burn, radiotherapy or physical deformity, severe esophageal stricture requiring dilatation at first endoscopy or expectation of requiring dilatation during the study, recent PPI treatment within 8 d prior to endoscopy, or using PPIs for more than 5 d in the last 28 d prior to endoscopy, with H pylori eradication therapy within the last 28 d prior to randomization or at any time during the study, using other antisecretory or prokinetic agents between endoscopy and randomization or at any time during the study, or with any investigational (non-approved) drug during the last 30 d prior to randomization. In addition, those with severe concurrent diseases judged by the investigators to complicate the evaluation of the trial, pregnancy, lactation or child-bearing potential without adequate contraception (contraceptive pill or intrauterine device), chronic alcoholism, drug abuse or any other conditions associated with poor patient compliance including expected non-co-operation, previous randomization in the study, and patients who needed continuously concomitant therapy with anticholinergics, cisapride, prostaglandin analogs, non-steroidal anti-inflammatory drugs (including COX-II) and aspirin (excluding low-dose aspirin e.g., 100 mg, as anti-platelet) were also excluded from this study. This study was approved by the Ethical Committee of Taipei Veterans General Hospital and carried out in accordance with the World Medical Association Helsinki Declaration. Written informed consent was obtained before any study-related procedures were performed.

The GERD symptoms such as heartburn, acid regurgitation, epigastric/chest pain, belching, nausea, vomiting and global well-being were assessed based on a standard visual analog scaled (VAS) questionnaire. Study medication was administered only to those subjects included in this study, following the procedures set out in the clinical study protocol. A sequence of patient numbers was assigned to the study center. All subjects entering the study received a patient number. This patient number was printed on the case report form and was used to identify the subject throughout the study. A randomization schedule was generated by the AstraZeneca using a validated system that automated the random assignment of treatment groups according to the randomization numbers. This schedule linked sequential numbers to treatment codes allocated at random. The schedule was prepared with a 1:1 randomization ratio in block size of 4. The study medication was labeled with the randomization numbers (medication numbers). At the end of baseline visit, eligible patients were randomized to the study medication in accordance with the randomization schedule. The next eligible subject received the study medication with the lowest available randomization number. Each subject was given only the study medication carrying his/her randomization number. The investigator documented the randomization number by sticking the label provided on the appropriate case report form. Subjects who permanently discontinued from the study were to retain their subject number and their randomization number, if already given. New subjects were always allotted a new subject number and, if applicable, a new randomization number.

Patients were asked to come back to the study office on three occasions (baseline, wk 4 and 8, respectively) during the trial. All doses were taken by mouth once daily in the morning before breakfast. During the study period, the patients were instructed to take one tablet of esomeprazole or matching placebo and one capsule of omeprazole or matching placebo in the morning with a glass of water. The first dose of study medication was taken on the morning after randomization. This was considered as d 1 of treatment. After the whole course of treatment, they were asked to come back on the last day of medication. At that time, they received the 2^nd endoscopy to assess the EE status again. Meanwhile GERD symptoms based on VAS after treatment were scored again. All the endoscopic EE diagnoses and their follow-up according to LA grading were initially performed by an experienced endoscopist (Chang), while all the endoscopic findings including EE were recorded by Polaroid films. When all the studies were completed, the recorded films assigned to their study codes were reviewed independently by another two endoscopists (Lu and Chen) who were blind to the order and code of endoscopy for each patient. If three readings in each film were dissimilar, this film was discussed by the above three investigators together to obtain the final endoscopic assessment. Healing was defined as no EE evidence.

The primary endpoint of this trial was the percentage of enrolled patients whose EE was healed by wk-8 visit. The secondary efficacy variables included for the first time the relief of heartburn symptoms, changes of reflux symptom scores (based on VAS score) and overall therapeutic effects judged via either subjective (patient) or objective (investigator) assessment.

The safety assessments included observed and reported adverse events (AE) and clinical laboratory evaluations (hematology and serum chemistry).

The study took place between 28 March, 2001 and 26 October, 2001.

For the primary efficacy parameter, the healing rate for each group was calculated with 95%CI. The difference between the two treatment groups and the corresponding 95%CI were also provided and compared by using Fisher’s exact test. The cumulative percentage of patients who exhibited the first relief of their diary-recorded symptom of heartburn was compared using Fisher’s exact test on d 1, 7, and 28, respectively. The median first time to relieve heartburn symptom between the two groups were compared using log-rank test. For reflux symptoms, Wilcoxon rank sum test was used to assess patient VAS score, and Fisher’s exact test was used to assess investigator scores as well as the overall therapeutic effect.

AE were summarized according to coding symbols for thesaurus of adverse reaction terms. The number of patients who reported a particular event and the number of events were summarized. Comparative incidence of AE was evaluated using Fisher’s exact test. Each laboratory parameter was listed with values outside the reference range identified. For each laboratory parameter, changes in abnormality/normality status from pre- to post-treatment were summarized in shift tables and assessed using McNemar test.

No interim analysis was performed in this study.

Finally, 48 eligible EE patients were enrolled and randomized according to the protocol. Their demographics and baseline characteristics are summarized in Table 1. Of them, 25 patients were distributed into esomeprazole group whereas 23 patients were on omeprazole treatment. Table 1 illustrates that both groups were comparable in their demographic characteristics and basal clinical manifestations except the subjects of esomeprazole group had a higher chance of belching (P<0.05). Among the 48 randomized patients, only 44 (esomeprazole: 24; omeprazole: 20) completed the whole study course. The reasons why four of them did not finish the trial were as follows: two lost their follow-up and another two discontinued the study medication. In addition, 2 (all were esomeprazole) of 44 who finished the study refused endoscopy follow-up at wk-8 visit were excluded from per protocol analysis.

The EE healing rates of esomeprazole and omeprazole treatment judged at the end of 8-wk trial [per-protocol (PP)] were 72.7% (16/22, 95%CI: 49.8-89.3%) and 50.0% (10/20, 27.2-72.8%) respectively [intent-to-treat (ITT): 64% (16/25, 95%CI: 44.3-83.8%) vs 45.5% (10/22, 95%CI: 22.7-68.3%), P = 0.2481], while the odds ratio was 2.667 (PP: 95%CI: 0.739-9.63, P = 0.2040) for esomeprazole over omeprazole.

In order to understand whether esomeprazole was effective in reducing LA-based EE grading, we further analyzed their extent of changed grading for both groups, e.g., 0 (no change), -1 (A to healed, B to A, C to B and D to C), -2 (B to healed, C to A and D to B) and -3 (C to healed, D to A), respectively (Figure 1). Although esomeprazole showed a better healing ability, however, no significant difference was found.

Open in New Tab   Full Size Figure   Download Figure

Figure 1 Histogram showing endoscopic assessment for the improved LA grading in erosive esophagitis patients after esomeprazole or omeprazole treatment. Grade of improvement: -3 means D to A or C to healed, -2 means D to B, C to A, or B to healed, -1 means D to C, C to B, B to A, or A to healed.

The median time to the first relief of heartburn between esomeprazole and omeprazole groups was similar on d 1 after treatment. On d 1, 77.3% and 65% of EE patients recorded the first relief of heartburn, respectively (Figure 2). Table 2 denotes that esomeprazole treatment was not significantly different from omeprazole in any of improved GERD symptom scores evaluated by patients themselves based on VAS except belching improvement was marked in patients undergoing esomeprazole treatment (P<0.05). Table 3 illustrates that the therapeutic symptomatic response scores of both treatments were similar (NS).

Open in New Tab   Full Size Figure   Download Figure

Figure 2 No cumulative percentage difference in relieving heartburn between erosive esophagitis patients after esomeprazole or omeprazole treatment illustrated by scattering plot.

In general, EE patients receiving esomeprazole (28.0%) and omeprazole (26.1%) treatment reported AE at least once during the trial (NS). Among them, constipation, dry skin sensation, diarrhea, headache, somnolence, etc., were the recorded AE in both groups (Table 4). Their distributions were also not different. Only one patient in omeprazole group reported a serious adverse event of cellulites during the treatment period, this causal relationship to treatment was judged to be unrelated. In addition, there were no clinically meaningful differences between treatment groups in terms of changed laboratory values or physical examinations.