Rebamipide with Proton Pump Inhibitors (PPIs) versus PPIs Alone for the Treatment of Endoscopic Submucosal Dissection-Induced Ulcers: A Meta-analysis

Objective To contrast the effect of rebamipide with proton pump inhibitors (PPIs) versus PPIs alone for the treatment of endoscopic submucosal dissection (ESD-) induced ulcers. Methods PubMed, Embase, the Cochrane library, the WanFang database, and China National Knowledge Infrastructure (CNKI) were searched to identify studies that met the inclusion criteria. Results Nine randomized controlled trials (RCTs) were recognized, including 1170 patients. In general, rebamipide plus PPIs acted better than PPIs alone against ESD-induced ulcers at four weeks (RR = 1.42, 95% CI: 1.13-1.78, P = 0.003) but showed no significant differences at eight weeks (RR = 1.03, 95% CI: 0.97-1.10, P = 0.315). The use of rebamipide plus PPIs was superior to PPIs alone for ESD-induced ulcers greater than 20 mm in size (20-40 mm: RR = 1.98, 95% CI: 1.22-3.23, P = 0.006; >40 mm: RR = 5.14, 95% CI: 1.49-17.74, P = 0.010). In addition, rebamipide plus PPI therapy was discovered to be significantly more effective than PPIs alone for lower ESD-induced ulcers (RR = 1.82, 95% CI: 1.04-3.20, P = 0.037). There were no significant differences between the treatment groups with the ulcer reduction rate. Conclusion Evidences now available show rebamipide plus PPIs is practical for protecting against ESD-induced ulcers at four weeks but not at eight weeks, especially large ulcers (>20 mm). However, we still need more high-quality RCTs in the future to supplement our conclusions.


Introduction
Endoscopic submucosal dissection (ESD), first evolved in the late 1990s and early 2000s, is presently a diffusely adopted endoscopic resection technique for early gastric cancers (EGC) and some gastric adenomas [1]. ESD is minimally invasive for the patients, and the en bloc rate is higher than the endoscopic mucosal resection, no matter the injury size [2]. However, the use of ESD may lead to deep and large gastrointestinal ulcers and complications such as perforation, bleeding, indigestion, and abdominal pain [3]. Proton pump inhibitors (PPIs) are currently the main drugs for the treatment of peptic ulcer, usually used for ESD-induced ulcer [4]. However, some studies have indicated that the healing effect of PPIs alone is not sufficient for the ESD-induced ulcers within the duration of treatment [5].
Rebamipide, as a mucosal protective drug, can not only increase the production of endogenous prostaglandins but also has the cytoprotective antiulcer effects [6]. The treatment of peptic ulcers by using rebamipide is valid, and rebamipide can reduce the recurrence ratio, without impacting the Helicobacter pylori infection condition of the patients in the former researches [7]. Moreover, experimental trials have confirmed the protective effect of rebamipide against nonsteroidal anti-inflammatory drugs (NSAIDs-) induced gastrointestinal mucosal lesion [8]. Some studies have shown that the effect of rebamipide plus PPIs is the same as or better than PPIs alone [9]. Therefore, we conducted a meta-analysis to evaluate the efficacy of rebamipide plus PPIs in the treatment of ulcers after ESD compared with PPIs alone.

Materials and Methods
This meta-analysis was conducted based on the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement [10].

Search Strategy.
To ascertain all studies comparing the efficacy and safety of rebamipide plus PPIs with PPIs alone for the treatment of ulcers after ESD, we searched on the PubMed, Embase, the Cochrane Library database, China National Knowledge Infrastructure (CNKI), and the Wan-Fang database for all articles published up to June 2020. Search terms included: "Endoscopic submucosal dissection," "ESD," "proton pump inhibitors," "PPIs," and "rebamipide." There were no language constraints. Study citations and abstracts were assembled, and full papers were searched to filtrate for possibly relevant literature. The abovementioned process of literature search and selection was independently accomplished by two researchers. All disagreements were resolved by consensus.

Study
Selection. The inclusion criteria are as follows: (1) all studies contained the comparison of rebamipide plus PPIs versus PPIs alone for the treatment of ulcers after ESD and (2) the relevant data of the outcome measures of the two groups can be successfully extracted and analyzed.
The exclusion criteria are as follows: (1) non-RCTs, qualitative studies, or studies without withdrawable data; (2) the study population or trial size was not clear; and (3) case reports, editorials, comments, and reviews, or just abstract alone were ruled out.

Data
Extraction and Quality Assessment. The withdrawable data were the following: first author, year of publication, country, study type, age, gender, number of participants (rebamipide plus PPIs : PPIs), dose, medication duration, lesion size, tumor location, and endpoints. We used the sixstage Sakita and Fukutomi table (active (A1, A2), healing (H1, H2), and scarring (S1, S2)) to classify the grade of healing of the ulcers [11]. Two researchers independently extracted the data. The divergences in the data extraction process were resolved by discussion.
A quality assessment of each included RCT was executed by two researchers with the Cochrane Handbook for Systematic Reviews [12]. The risk ratio (RR) and 95% confidence interval (CI) was used to analyze dichotomous data, such as the reduction rate and healing rate of different duration of treatment, location of resection, and specimen size. The weighted mean differences (WMD) were used to analyze continuous data. The heterogeneity was investigated by using the Q test and the I 2 test. The values of I 2 25%-50%, 50%-75%, and >75% were considered as low, moderate, and high heterogeneity, respectively [13]. We perform the random-effects model when I 2 > 50% and P < 0:1. Otherwise, the fixed-effects model is executed. P < 0:05 was regarded as statistically significant in all tests. Begg's test was performed to assess potential publication bias [14].

Selection of Studies.
A sum of 139 records was confirmed by searching with the keywords and free words. After inspecting the titles and abstracts, 113 records were excluded because of duplication or irrelevance; 17 records were excluded as a result of insufficient data from the residual articles by full review. Eventually, 9 RCTs were involved to analyze [9,11,[15][16][17][18][19][20][21]. Figure 1 shows the flow of study inclusion.  3.3. Risk of Bias. Random sequence generation was found in five studies [11,15,17,19,21]. Blinding of participants and personnel was found in nine studies [9,11,[15][16][17][18][19][20][21]. Blinding of outcome assessment was found in one study [18]. The results of the quality assessment in this meta-analysis are shown in Figures 2 and Figure 3.

Reduction
Rate. The ulcer reduction rate was assessed between the rebamipide plus PPI therapy group and PPIs alone group in three studies [9,19,21]. As shown in Figure 7, there were no statistically significant differences in the ulcer reduction rate between the rebamipide plus PPIs group and PPIs alone group (RR = 1:03, 95% CI: 0.99-1.07, P = 0:204, I 2 = 0%).

Adverse
Events. Fujiwara et al. [15] reported that one patient underwent bleeding from the iatrogenic ulcers after ESD. Nakamura et al. [19] reported that bleeding occurred in three patients in the rebamipide plus PPIs group. There were no complications involved in the drugs used after ESD in any of the participants.

Publication
Bias. Using Begg's funnel plot, the potential publication bias in the included studies was evaluated ( Figure 8). No publication bias was detected by Begg's test due to the P value >0.05 for the healing effect of rebamipide plus PPI therapy compared with PPIs alone.

Discussion
ESD is the major class of endoscopic resection technique currently used to treat the superficial gastrointestinal neoplasms or lesions, regardless of the size or location. There are larger   BioMed Research International artificial gastric ulcers or delayed bleeding after the use of ESD. PPIs are the standard treatment for the healing of ulcers after ESD. However, Oh et al. [22] described that the premier ulcer size influenced ulcer healing by PPIs at 4 weeks after ESD, and Kakushima et al. [23] showed that 4 weeks of PPI treatment was not enough for ESD-induced ulcers of large size to heal and that 8 weeks was required. Therefore, it seems that the management of PPIs alone may not be sufficient for the ESD-induced ulcers to heal; the combination therapy is needed. Rebamipide, as a mucosal protective agent, can enhance the production of endogenous prostaglandins, inhibit the reduction of mucosal blood flow, suppress increases in permeability, scavenge free radicals, and has an anti-inflammatory effect [24][25][26][27].
We conducted this meta-analysis to explore the effects of treatment with rebamipide plus PPIs versus PPIs alone for the ulcers after ESD. Previous studies indicated that PPIs alone for the ulcers after ESD was not sufficient [11]. In this meta-analysis, the results showed that rebamipide plus PPIs is superior to PPIs alone for healing the ESD-induced ulcers at four weeks; however, there were no significant differences between the rebamipide plus PPIs and PPIs alone with a treatment of eight weeks. In general, the healing rate in the combination therapy group was higher than that in the PPIs alone group. Some studies reported that the healing of ESDinduced ulcers was associated with both the location of resection and the specimen size [11]. In this meta-analysis, we evaluated the degree of ulcer healing with respect to the location of resection and specimen size, and we detected significantly higher healing rate after ESD in the rebamipide plus PPIs group than in the PPIs alone group with the specimen size 20-40 mm, particularly with the specimen size >40 mm. We also found a significantly higher healing rate of ulcers after ESD in the rebamipide plus PPIs group than in the PPIs alone group with lower location of resection. Three studies evaluated the reduction rate of ESD-induced ulcers, but there were no significant differences between the treatment groups. In addition, serious complications connected with the drugs used after post-ESD were not noticed in the two groups.

Limitations.
Although increasing studies recommend rebamipide as a valid mucosal protective agent, the current data are not plenty for strictly confirming the effectiveness of rebamipide. Therefore, this meta-analysis has some limitations to be addressed. First, the number of inclusive studies was quite limited, and the quality of the containing RCTs was relatively low. Second, all the included studies were in Asia; the racial differences might lead to different reactions to drugs. Third, the various types and different use methods of PPIs might produce a bias. Fourth, some studies lack extractable data, random sequence generation methods, double-blind or triple-blind details, and uniform follow-up time.

Conclusion
In general, this meta-analysis has presented the crucial profit of rebamipide plus PPIs in managing ESD-induced ulcers.
The results of the pooled analysis mainly show that rebamipide plus PPIs is practical for protecting against ESDinduced ulcers at four weeks but not at eight weeks, especially large ulcers (>20 mm). However, more studies with designs of large scale and high quality are still required to further determine the efficacy of rebamipide plus PPIs.

Data Availability
The data supporting this meta-analysis is from previously reported studies and datasets, which have been cited.