Vonoprazan Versus Lansoprazole in the Treatment of Articial Gastric Ulcers After Endoscopic Submucosal Dissection: a Randomized Controlled Trial

Background: Vonoprazan is more potent and longer acting than traditional proton pump inhibitor. Although vonoprazan is expected to be superior to proton pump inhibitor, its ecacy in the treatment of gastric ulcers following endoscopic submucosal dissection (ESD) is not fully understood. The aim of this study was to evaluate the effectiveness of vonoprazan in articial ulcer healing following ESD. Methods: Patients with gastric tumors were randomly assigned to the vonoprazan group (group V) or lansoprazole group (group L) after ESD. Patients received intravenous lansoprazole (30mg) twice on the day of ESD. Thereafter, patients were treated with vonoprazan (20 mg/day) in group V or lansoprazole (30 mg/day) in group L. Esophagogastroduodenoscopy was performed 4 and 8 weeks after the ESD. Results: A total of 168 patients were analyzed. The 4-week healing rate for articial ulcer was not signicantly higher in group V versus group L (17/85, 20.0% vs. 14/83, 16.9%, respectively). In addition, there were no signicant differences between the 4-week shrinkage rates between the two groups. Postoperative bleeding occurred in none of the patients in group V and three in group L. One patient in group V presented delayed perforation 2 days after ESD. Conclusions: Vonoprazan might not be superior to lansoprazole in the healing of articial gastric ulcer after ESD.

Background Endoscopic submucosal dissection (ESD) for early gastric cancer has been widely accepted and is a wellestablished procedure in Eastern and Western countries [1][2][3][4]. ESD provides higher rate of en bloc and R0 resection rate, but is occasionally associated with some complications, such as bleeding and perforation [5]. Conventional proton pump inhibitors (PPIs) have been widely used for treating ESD-induced gastric ulcers. Although PPIs support to heal such lesions, some ulcers fail to heal; therefore, a more effective therapy is warranted.
Vonoprazan is a novel suppressant of gastric acid secretion and an active potassium-competitive acid blocker (P-CAB) [6]. Similar to PPIs, P-CABs inhibit gastric H+/K+-ATPase. Unlike PPIs, P-CABs inhibit the enzyme in a K+-competitive and reversible manner. The inhibitory effect of vonoprazan on gastric acid secretion is largely unaffected by ambient pH. Therefore, vonoprazan is more potent with a more longlasting effect than that of PPIs [7,8]. Vonoprazan is expected to be superior to PPIs, and lead to earlier healing of ESD-induced gastric ulcers versus conventional PPI-based therapy. However, its e cacy in treating these ulcers remains unclear. Hence, in this study, we aimed to evaluate the effectiveness of vonoprazan in healing arti cial ulcers after ESD.

Methods
This study was a prospective, single-center, randomized controlled trial (RCT). The study protocol was approved by the local ethics committee of Tsuyama Chuo Hospital, Tsuyama, Japan and registered with the University hospital Medical Information Network (URL: https://www.umin.ac.jp/ctr/index-j.htm; registration number: UMIN000016642). Written informed consent was provided by each patient.
Patients with gastric tumors were enrolled between April 2015 and December 2017. The exclusion criteria were: remnant stomach, administration of antithrombotic agents, non-steroidal anti-in ammatory drugs and steroids; occurrence of complication during ESD; allergy to lansoprazole or vonoprazan; and unwillingness to participate in the study. Following ESD, patients were randomly assigned to the vonoprazan group (group V) or the lansoprazole group (group L) through the minimization method using Kullback-Leibler divergence [9]. The ESD ulcer index was used to balance continuous variables. On the day of ESD, patients received 30 mg of lansoprazole twice intravenously. From postoperative day 2, patients in groups V and L received 20 mg/day of vonoprazan and 30 mg/day of lansoprazole for 8 weeks, respectively.
Esophagogastroduodenoscopy was performed 4 and 8 weeks after the ESD (Fig. 1). During the follow-up endoscopy, the arti cial ulcer was evaluated using a gastric ulcer stage system [10], and the length and width of the arti cial ulcer were evaluated with measure forceps (M2-2C, M2-3U or M2-4K, Olympus Co, Japan) ( Fig. 2A). The ulcer healing was de ned as scarring at S1 or S2. The ESD ulcer index was calculated by multiplying the length by the width of the resected specimen. The 4-and 8-week ulcer indices were also calculated by multiplying the length by the width of the arti cial ulcer at 4 and 8 weeks after ESD, respectively (Fig. 2B). The shrinking rate was de ned as [1-(the ulcer index)/(the ESD ulcer index)] × 100 (%).
The primary endpoint was the healing rate of the arti cial ulcer at 4 weeks after ESD. The secondary endpoints were: the healing rate at 8 weeks; shrinkage rates of the arti cial ulcers at 4 and 8 weeks; and complications, such as post-operative bleeding and delayed perforation.
The healing rate of ulcers after the 4-week administration of PPIs was reported as 11-35% [11,12].
According to our previous data, the healing rate after a 4-week administration of PPIs was 24%. A 20% improvement observed following the administration of vonoprazan was considered clinically effective. It was estimated that 82 cases were required in each group to have a power of 80% for detection of a difference at an α = 0.05 level of signi cance using Fisher's exact test. Assuming a study drop-out rate of approximately 10%, 90 cases were enrolled in each group.

Results
Of the 263 patients who underwent gastric ESD from April 2015 to December 2017, 182 patients (90 in group V and 92 in group L) were eligible to participate in the study. Fourteen patients were excluded during follow-up mainly because of complications or the need for additional surgery. Finally, 85 and 83 patients were allocated to group V and group L, respectively (Fig. 3). The age, gender, status of Helicobacter pylori infection, tumor location, and ESD ulcer index of the two groups did not differ signi cantly (Table 1). The 4-week healing rate of arti cial ulcers was not signi cantly higher in group V versus group L (17/85, 20.0% vs. 14/83, 16.9%, respectively). Furthermore, there was no signi cant difference between the 4-week shrinkage rates noted in groups V and L. Postoperative bleeding was observed in three patients of group L and none of the patients in group V. One patient in group V presented delayed perforation 2 days after ESD (Table 2).

Discussion
In this RCT, we attempted to prove the superiority of vonoprazan to lansoprazole in the healing rate after gastric ESD in patients strati ed according to the ESD ulcer index. Unfortunately, vonoprazan was not superior to lansoprazole in terms of the healing rates and shrinkage rates of arti cial gastric ulcers at 4 and 8 weeks after ESD. Furthermore, there were no signi cant differences in the occurrence of postoperative bleeding and delayed perforation between the two groups.
Some RCTs suggested that vonoprazan was as effective as PPIs in the treatment of ESD-induced ulcer [13][14][15]. Our results were consistent with those previously reported, in which the healing rates at 4 weeks in the vonoprazan groups ranged 7.4-20.9%. Our hypothesis was that the strong and rapid inhibition of gastric acid secretion by vonoprazan may enhance the healing of arti cial ulcers. However, the effect of ulcer shrinkage in the vonoprazan group was similar to that noted in the PPI group. There may be two reasons for this result. Firstly, acid suppression by both vonoprazan and lansoprazole are excellent at shrinking arti cial ulcers. Secondly, other factors than acid suppression are involved in the rapid resolution of ulcer. These factors include the existence of ulcer scar, ulcer area, ulcer site, blood coagulation status, Helicobacter pylori infection, and other comorbidities.
On the other hand, some studies concluded that vonoprazan was superior to PPIs for healing ESDinduced ulcers [16][17][18][19]. However, there were few prospective, randomized controlled studies conducted. Tsuchiya et al. reported that the vonoprazan group had a signi cantly superior shrinkage rate at 8 weeks; however, there was no signi cant difference observed in the rate of postoperative bleeding [16]. In this prospective study, the shrinkage rates until 6 weeks were not signi cantly different. However, the 8-week shrinkage rate was signi cantly higher in the vonoprazan group versus the PPI group. In a systematic review and network meta-analysis, the effect of vonoprazan at 8 weeks was superior to that of PPIs for the treatment of arti cial ulcers following ESD [20]. However, another meta-analysis reported that the healing rate at 8 weeks was signi cantly higher in the PPI group versus the vonoprazan group [21]. Thus, the effect of vonoprazan remains controversial and further RCTs are warranted. Although the primary endpoint did not meet, we believe the ndings of our study are meaningful.
In this investigation, only three patients in the PPI group developed postoperative bleeding. In the 5 RCTs (including this trial) conducted thus far, the rate of postoperative bleeding in the vonoprazan group range 0-5.4%. Although the backgrounds of the study differ, the postoperative bleeding rate was equal or lower in the vonoprazan group versus the PPI group in all studies. Hamada et al. showed that vonoprazan e caciously reduced the delayed bleeding rate in patients with an ESD-induced gastric ulcer in comparison with the threshold rate recorded using binomial testing [22]. A larger-scale study with postoperative bleeding as its primary endpoint or meta-analysis using RCTs may prove the e cacy of vonoprazan against this complication.
There were some limitations in this study. Firstly, this study was conducted in a single center in Japan and sample size was small. Secondly, patients receiving antithrombotic agents were excluded. As postoperative bleeding was associated with administration of antithrombotic agents, its rate may be underestimated in this study.

Conclusions
Vonoprazan was not superior to lansoprazole in the healing of arti cial gastric ulcers after ESD. However, our data are meaningful because the effect of vonoprazan remains controversial and a larger-scale RCTs are required to verify the present ndings. The study was approved by the institutional review board of our hospital (Tsuyama Jihukai Ethics Committee) on March 3, 2015 (reference number 220) and was conducted in accordance with the Declaration of Helsinki. Written informed consent was provided by all patients.

Consent for publication:
Not applicable Availability of data and materials All data generated or analyzed during this study are available from the corresponding author on reasonable request.
Competing interests: The authors declare that they have no competing interests.

Funding
There was no nancial support provided for this study.
Authors' contributions: Concept and design: DK and RT; data collection: DK, RT, MI, SO, TG, YK, KT, HT; analysis and interpretation of data: DK, RT and MI; drafting of the article: DK and RT; critical revision of the article: SF. All authors read and approved the manuscript.