Effect of lansoprazole versus roxatidine on prevention of bleeding and promotion of ulcer healing after endoscopic submucosal dissection for superficial gastric neoplasia
- First Online:
- Cite this article as:
- Imaeda, H., Hosoe, N., Suzuki, H. et al. J Gastroenterol (2011) 46: 1267. doi:10.1007/s00535-011-0447-1
- 311 Views
Proton pump inhibitors have been reported to be more useful than histamine-2 receptor antagonists for the prevention of bleeding after endoscopic submucosal dissection (ESD) for superficial gastric neoplasia. The aim of this study was to assess the effects of the proton pump inhibitor lansoprazole and the histamine-2 receptor antagonist roxatidine for the prevention of bleeding and the promotion of ulcer healing after ESD and to compare the cost-effectiveness of these two drugs.
The study subjects were 129 patients who underwent ESD for superficial gastric neoplasia. The patients were randomly assigned to the lansoprazole group (L group) or the roxatidine group (R group). Either drug was administered intravenously from the morning of the ESD day to the day after the ESD, followed by oral treatment for an additional 8 weeks. A second-look endoscopy was performed on the day after the ESD, and a repeat endoscopy was performed at 8 weeks after the ESD. The incidence of bleeding and the ulcer-healing rate at 8 weeks after the ESD were analyzed, as well as the total cost of treatment with these antisecretory agents.
Three patients in each group were excluded from the analysis, leaving 62 patients in L group and 61 in R group. Two of the 62 patients (3.2%) in L group and three of the 61 patients (4.9%) in R group showed bleeding after ESD ; there was no significant difference between the two groups (P = 0.68). The ulcer-healing rate was 93.5% (58/62) in L group and 93.4% (57/61) in R group (P = 1). The total cost of treatment with the antisecretory agent from the day of the ESD to day 56 after the ESD was Yen 13,212 for lansoprazole and Yen 5,841 for roxatidine.
Roxatidine appears to have high cost-effectiveness in the prevention of bleeding and in the promotion of ulcer healing after ESD for superficial gastric neoplasia.
KeywordsBleedingUlcer healingEndoscopic submucosal dissectionGastric neoplasiaRoxatidineLansoprazoleH2-receptor antagonistProton pump inhibitor
Endoscopic mucosal resection (EMR) has been established as a minimally invasive treatment for early-stage gastric cancer (EGC) [1, 2]. Bleeding after EMR has been reported to occur in 1.2–11.6% of patients . Following conventional EMR, proton pump inhibitors (PPIs) and histamine-2 receptor antagonists (H2RAs) have been reported to show equivalent efficacy in preventing bleeding from the ulcer and promoting ulcer healing . Of note, en-bloc resection is often not achieved by EMR, such as for a strip biopsy or EMR using a cap. Recently, endoscopic submucosal dissection (ESD) has been developed for EGC, and this procedure enables larger lesions to be resected, thereby yielding improved rates of successful en-bloc resection . Although ESD has been established as a standard treatment for EGC, it is associated with a higher incidence of complications, such as bleeding and perforation, than EMR . PPIs or H2RAs have been administered for the prevention of bleeding after ESD as well as after EMR. PPI therapy has been reported to be more useful than H2RA treatment for the prevention of delayed bleeding from the ulcer after ESD [7, 8].
Roxatidine is a second-generation six-membered-ring H2RA that not only suppresses gastric acid secretion, but also increases mucus secretion from the gastric mucosa [9, 10]. It is as effective and safe as the other H2RAs with a five-membered ring (cimetidine, ranitidine, famotidine, and nizatidine) in the treatment of gastric and duodenal ulcers, and may also be useful for the treatment of five-membered-ring H2RA-resistant ulcers, especially duodenal ulcers .
The first aim of this pilot study was to evaluate the efficacy of lansoprazole, as compared with that of the second-generation H2RA, roxatidine, for the prevention of bleeding and promotion of ulcer healing after ESD for superficial gastric neoplasia. The second aim of the study was to compare the cost-effectiveness of these two drugs.
Patients and methods
This prospective pilot study was performed at Keio University Hospital. Written informed consent was obtained from all patients, and patient enrollment started from March 2008 and finished in March 2010. The inclusion criteria were patients with EGC, gastric adenoma, or suspected gastric neoplasia who were referred to our hospital for the purpose of ESD. Patients who had received acid-suppressant treatment within 1 week prior to the procedure, or those who had a history of gastrectomy, major organ failure, or drug allergy were excluded. Patients were admitted on the day before the ESD and remained hospitalized for 4 days with a defined clinical path. All patients were randomly assigned to either the lansoprazole group (L group) or the roxatidine group (R group). A randomization table was generated using Excel 2007. Patients in the L group received lansoprazole 30 mg twice a day intravenously for 2 days from the morning of the ESD procedure, and then 30 mg once a day orally for 8 weeks. Patients in the R group received roxatidine 75 mg twice a day intravenously for 2 days from the morning of the ESD procedure, and then 75 mg twice a day orally for 8 weeks. The doses of both drugs were standard according to the health insurance system in Japan. A second-look endoscopy was performed on the day after the ESD. If there was bleeding or a visible vessel in the ulcer bed, argon plasma coagulation (ICC-200; ERBE, Tübingen, Germany) and/or endoscopic hemoclipping (HX-610-090S; Olympus Medical Systems, Tokyo, Japan) was performed. Oral intake of meals was started on the evening of the second-look endoscopy, except in patients with massive bleeding and patients who were discharged on the same day as the second-look endoscopy.
Any anticoagulant or antiplatelet drugs that the patients were being treated with were stopped at least 7 days before the ESD and restarted 7 days after the ESD, and any mucosal protectant agents and antacids were discontinued during the treatment period. Venous blood samples were analyzed for IgG Helicobacter pylori antibodies by the E-plate test (Eiken Kagaku, Tokyo, Japan) using an enzyme-linked immunosorbent assay (ELISA) kit. If the test for IgG H. pylori antibodies was negative, the urea breath test was performed (Otsuka Pharmaceutical, Tokyo, Japan).
Major bleeding was defined as hematemesis, melena, and/or decrease of blood hemoglobin by more than 2 g/dL necessitating endoscopic hemostasis. The incidence of bleeding after the ESD was evaluated in the two groups. Minor bleeding was defined as type 1b according to the Forrest classification, i.e., bleeding that necessitated endoscopic hemostasis at the second-look endoscopy, except for major bleeding .
The endoscopists who performed the ESD were unaware of the group allocation of the subjects. This study was approved by the ethics committee at Keio University Hospital, and was registered through the registries approved by the International Committee of Medical Journal Editors (UMIN000001069).
The indications for ESD for EGC included intramucosal differentiated-type EGC of any size without ulceration, intramucosal differentiated-type EGC measuring <3 cm in diameter with ulceration or scar formation, submucosal differentiated-type EGC showing invasion up to a depth of no more than 500 μm and <3 cm in diameter, and gastric adenoma of any size suspected by preoperative diagnostic examinations. The absence of lymph node and distant metastases was confirmed by contrast-enhanced computed tomography (CT) of the thorax and abdomen and/or abdominal ultrasonography. The ESD procedure was performed using a flex knife (Olympus Medical Systems). An ICC-200 (ERBE) or VIO300D (ERBE) was used as the electrosurgical unit. The ulcer created after the resection was carefully examined, and any visible vessels were coagulated by the flex knife in the soft coagulation mode with 80-W current or treated by hemoclipping. The resected specimen was immediately pinned flat to a rubber plate for measurement of its size, and then it was sent to the department of pathology for gross and histopathological examinations.
Follow-up and monitoring of adverse events
During the patients’ periods of hospitalization, the occurrence of any adverse events was evaluated daily by history-taking and physical examination. A complete blood cell count was assessed on the day after the ESD. Follow-up endoscopy was scheduled for the day after the ESD, for the evaluation of bleeding, and a repeat endoscopy was performed at 8 weeks after the ESD for the evaluation of ulcer healing. When the ulcer appeared to be closed, it was defined as having healed fully. All adverse events after discharge were verified by history-taking at the time of consultation. Patients were asked to bring all remaining tablets to the outpatient clinic at the end of the 8-week study period, and if they did not exceed 10% of the prescribed number, the patients were regarded as showing good compliance with the treatment. The rate of ulcer healing at 8 weeks after the ESD was also evaluated in the two groups.
Sample size calculation
The primary variable was the major bleeding rate. First, we hypothesized that bleeding rates in the two groups would not be significantly different. The sample size calculation was based on non-inferiority of the H2RA with a significance level of 0.05, a power level of 0.9, and an inferiority margin of 10%, and bleeding rates of 6.0% with a PPI and 17.2% with a first-generation-H2RA . The required sample size was more than a hundred thousand. Of note, there were no available data on the bleeding rate with the use of a second-generation H2RA. Uedo et al.  reported that a PPI was superior to an H2RA in preventing bleeding after ESD in 130 patients. Taking into account the feasibility of the sample size, we conducted a pilot study to compare a second-generation-H2RA with a PPI by using 65 samples in each group, for a total of 130, the same as the sample size reported by Uedo et al.
All statistical analyses were performed on a per-protocol basis. Statistical comparisons were performed using the χ2 test or Fisher’s exact test for categorical data, and Student’s t-test for numerical data. The cumulative non-bleeding rates in the two groups were estimated using the Kaplan–Meier method, and the log-rank test was used for analysis of the differences between the two curves. Multivariate analysis was performed using Cox’s proportional hazard model. Candidate covariates included the stratification factors, age, sex, smoking status, presence/absence of H. pylori infection, history of use of anticoagulant or antiplatelet drugs, identity of the endoscopists, location of the tumor, presence/absence of scar formation in the tumor, and the tumor size and depth. Computer software (SAS version 9.1; SAS Institute, Cary, NC, USA) was used for the data analysis. P values of <0.05 were considered to denote statistical significance.
Baseline data of the treatment groups
No. of patients
Mean age (years ± SD)
68.4 ± 8.0
67.6 ± 8.5
Mean tumor size (mm, ±SD)
13.7 ± 9.0
16.3 ± 11.7
Mean artificial ulcer size (mm, ±SD)
32.8 ± 10.3
34.2 ± 12.7
Depth of tumors
Location of tumors
Tumor with scar
Incidence of bleeding and ulcer healing after endoscopic submucosal dissection (ESD) for superficial gastric neoplasia
No. of patients
Major bleeding after ESD
Minor bleeding after ESD
The ulcer-healing rate at 8 weeks after the ESD was 93.5% (58/62) in the L group and 93.4% (57/61) in the R group; there was no significant difference in the rate between the two groups (P = 1). After adjustments for the potential factors influencing the risk of ulcer healing, treatment with roxatidine was not found to have any significant effect of decreasing the ulcer-healing rate (adjusted odds ratio 0.61, 95% CI 0.10–3.83, P = 0.60).
No bleeding or perforation happened after ESD in any of the three patients with no neoplasm excluded from the R group. Ulcer healing occurred at 8 weeks after ESD in all of these patients.
The total cost of treatment with the antisecretory agents from the day of the ESD to day 56 (8 weeks) after the ESD was Yen 13,212 for lansoprazole and Yen 5,841 for roxatidine in our study, demonstrating a significantly higher cost-benefit ratio for roxatidine. During the follow-up period, no side-effects induced by drug administration were found in either treatment group.
In ESD, a relatively large area of the gastric mucosa can be resected by electrocautery. However, bleeding occurs frequently during the procedure, and postoperative bleeding occurs in approximately 5% of all patients who undergo gastric ESD [6, 13, 14]. The size of the resected tumor has been reported as a significant risk factor for postoperative bleeding after ESD [7, 15]. Gastric pH affects the efficiency of blood coagulation and platelet aggregation at the site of bleeding. Pepsin is active below pH 5.0 and accelerates clot digestion. In vitro, platelet functions are severely impaired at low pH . Thus, a reduction in gastric acidity towards neutral would stabilize the clotting mechanism and prevent bleeding. PPIs are known to be more potent at elevating the intragastric pH than H2RAs. Uedo et al.  reported that the intragastric pH was significantly higher in patients administered a PPI than in those administered an H2RA on the day before the ESD. PPIs have also been reported to be superior to H2RAs in preventing bleeding after ESD [7, 8]. According to the report by Uedo et al. , bleeding after ESD was seen in four of 66 patients (6.1%) treated with rabeprazole and 11 of 64 patients (17.2%) treated with cimetidine. Jeong et al.  reported that the incidence of bleeding after ESD was significantly lower in a group treated with pantoprazole than in a group treated with famotidine (3.5 vs. 12.7%, p < 0.05).
In the present pilot study, there was no significant difference between the lansoprazole and roxatidine groups in the prevention of bleeding after ESD. Roxatidine is a second-generation-H2RA and has been reported to not only suppress gastric acid secretion, but also to activate gastric mucosal defense mechanisms [9, 10]. Although the exact mechanisms involved in the mucosal defense system are unknown, at least one or more of the naturally occurring gastric mucosal defense factors such as increase of the gastric blood flow and promotion of bicarbonate secretion and mucin metabolism are involved. Roxatidine has been reported to prevent the formation of gastric mucosal lesions induced by necrotizing agents in rats , and it has also been shown to be as effective and safe as the other H2RAs for the treatment of gastric ulcers and duodenal ulcers. Particularly in duodenal ulcers, switching from H2RAs with a five-membered ring to roxatidine may be useful in the treatment of ulcers resistant to treatment with five-membered-ring agents . Therefore, roxatidine might be as efficacious as lansoprazole in preventing bleeding after ESD.
The rate of major bleeding in the present study was 4.1% (5/123) in all patients, which seemed to be lower than that reported from many previous studies. In patients undergoing ESD, the risk of delayed bleeding is highest within 24 h of the procedure . We always perform a second-look endoscopy on the day after the ESD, and we also carry out endoscopic hemostasis for any minor bleeding, so as to prevent major bleeding from the ulcer after ESD. Ono et al.  also reported a very low rate of major bleeding (one of 155 patients), likely attributable to the second-look endoscopy performed to prevent bleeding after ESD.
In our study, the ulcer-healing rate at 8 weeks after ESD was similar in the lansoprazole and roxatidine groups. In the study reported by Uedo et al.  also, the ulcer-healing rate in the PPI group (83%) was similar to that in the H2RA group (89%). However, Ye et al.  reported that fewer patients in their omeprazole group than in their famotidine group showed active ulcers at 28 days after ESD. Kato et al.  reported that combined PPI plus rebamipide treatment yielded a higher gastric ulcer healing rate at 4 weeks after ESD (68%) as compared with PPI treatment alone (36%). PPIs in combination with mucosal protectant drugs have been reported to be more useful than PPIs alone for the healing of ulcers after ESD [20, 21]. Mucosal protectant drugs may have an important role in promoting the healing of ulcers after ESD. In our study, we checked on ulcer healing only at 8 weeks after ESD; therefore, further studies are required to determine both the optimum duration of PPI therapy and the usefulness of follow-up endoscopy. Kakushima et al. [22, 23] reported that ulcer healing after ESD was delayed in patients with tumors showing ulceration or scar formation, while the presence/absence of H. pylori infection had no effect on ulcer healing. In our study, there was no significant difference between the two treatment groups in the number of tumors showing ulceration or scar formation, or in the prevalence of H. pylori infection. After adjustments for the potential contributors to ulcer healing, treatment with lansoprazole was not identified as a significant factor in increasing the ulcer-healing rate. Our study was, however, a single-center study and the sample size was so small that the difference between two groups might not have become apparent. We therefore hope to perform a multi-center trial with a larger study population in the future.
With respect to the costs and benefits, the total cost of treatment with the antisecretory agents from the day of the ESD to day 56 (8 weeks) after the ESD was Yen 13,212 for lansoprazole and Yen 5,841 for roxatidine. The cost-benefit ratio was thus significantly higher for roxatidine as compared with that for lansoprazole, and successful ulcer healing could be obtained in the roxatidine group at only 44.2% of the cost incurred in the lansoprazole group.
In conclusion, there was no significant difference in the incidence of bleeding or in the ulcer-healing rate after ESD for superficial gastric neoplasia between lansoprazole and roxatidine. Roxatidine appears to have high cost-effectiveness in the prevention of bleeding and promotion of ulcer healing after ESD for superficial gastric neoplasia.
Conflict of interest
The authors declare that they have no conflict of interest.