Preoperative H. pylori Eradication Therapy Facilitates Precise Delineation in Early Gastric Cancer with Current H. pylori Infection

Early gastric cancer (EGC) is defined as gastric adenocarcinoma limited to the mucosa or submucosa of the stomach, regardless of lymph node status [1]. Endoscopic submucosal dissection (ESD) has been proven to perform curative resection for EGC in cases with a low risk of lymph node metastasis, achieving equivalent effectiveness with structural integrity compared to surgery [2‒5]. In this process, complete resection is essential for the operation’s curability, including negative horizontal resection margins [6]. As the first step of ESD, it is pivotal to determine the precise scopes of the EGC lesion, with a recommendation of marking dots at 5 mm outside its boundary [7]. Common endoscopic approaches for cancerous area delineation include white-light endoscopy (WLE), chromoendoscopy (CE), and magnifying endoscopy with narrow-band imaging (ME-NBI) by discovering EGC prominent appearances [8‒10], including the following: (1) irregularity in color or surface patterns such as superficial elevation (0-IIa) or superficial depression (0-IIc) under WLE or CE; and (2) a demarcation line with irregular micro-vascular or micro-surface patterns under ME-NBI [10, 11].

Helicobacter pylori (H. pylori) is categorized as a group I carcinogen for gastric cancer [12], and eradication therapy for it can effectively prevent carcinogenesis [13]. It is currently believed that the intestinal type of gastric adenocarcinoma develops through a multistep histopathological process from chronic gastritis, atrophy, intestinal metaplasia, and intraepithelial neoplasia to carcinoma [14]. There is solid evidence that H. pylori eradication therapy (Hp-ET) after endoscopic resection of EGC can reduce metachronous gastric cancer (MGC) risk by nearly 50% and improve the grade of atrophy and intestinal metaplasia at the histological level at the corpus lesser curvature [15].

Notably, up to 75% of patients diagnosed with gastric cancer have concurrent H. pylori infection, including EGC with concurrent H. pylori infection (HpC-EGC), in the real world [16], and simultaneous operation and Hp-ET are required. However, researchers about the sequential order between Hp-ET and ESD are scarce and contrary. Some scholars have retrospectively found that lower inflammation levels after Hp-ET delivered higher precision of EGC boundary delineation than lesions without Hp-ET [17]. In contrast, the endoscopic characteristics of gastric cancer after eradication therapy (GCAE) tend to be smaller than 20 mm [18], have flatter appearances [19], or even be covered with nontumorous epithelium (also called gastritis-like appearances) [20]. And those features are significantly adverse to the delineation of lesions’ boundary and may result in the positive horizontal resection margin or even missing lesions. Nevertheless, the above experience could not reasonably apply to HpC-EGC patients since GCAE was emphasized as neoplasms developing in the follow-up period, usually for years after Hp-ET [21‒23]. There is a lack of evidence on the short-term impact of Hp-ET on delineating EGC boundaries and the curability of ESD for HpC-EGC to guide the optimal sequential strategy.

Thus, we conducted this prospective randomized controlled study to demonstrate Hp-ETs short-term effect on the above concerns. This study will provide evidence for H. pylori eradication timing versus endoscopic resection for HpC-EGC patients.

Between November 2018 and November 2021, patients who met the following inclusion criteria were prospectively enrolled at the department of gastroenterology, Peking Union Medical College Hospital (PUMCH). Inclusion criteria were as follows: (1) age 18–80 years; (2) diagnosis of high-grade intraepithelial neoplasia or EGC with a differentiated type by esophagogastroduodenoscopy (EGD) and corresponding biopsy histopathology; (3) meeting absolute or expanded indications for ESD according to the Japan Gastroenterological Endoscopy Society (JGES) guidelines (2016) [24]; and (4) current H. pylori infection confirmed by at least one positive result from the ¹³C-urea breath test (¹³C-UBT) or rapid urease test (RUT). Exclusion criteria were as follows: (1) previous Hp-ET within 3 months before recruitment; (2) past history of gastric tumor or gastric surgery; (3) allergy or intolerance to any elements of Hp-ET; (4) poor general conditions, pregnancy, or lactation; (5) contraindications for EGD- or ESD-related examinations; and (6) inability to provide informed consent. The demographic characteristics, endoscopic information, histopathological findings, and clinical data of enrolled patients were collected.

Enrolled HpC-EGC patients were randomly assigned to two groups by computer-generated random numbers: the eradication group received preoperative Hp-ET before ESD operation, and the control group underwent ESD operation followed by Hp-ET. According to the latest Chinese national consensus on H. pylori management [25], 14-day quadruple Hp-ET regimens were used in this study, consisting of the following: (1) proton pump inhibitors, including omeprazole (20 mg) or esomeprazole (20 mg), twice daily; (2) bismuth potassium citrate (220 mg), twice daily; and (3) two of the three antibiotics, including amoxicillin, clarithromycin, and levofloxacin. After completing Hp-ET, all patients in the two groups were required to re-evaluate H. pylori infection status (¹³C-UBT or RUT) to determine the efficacy of Hp-ET at a minimum interval of 4 weeks.

The Chinese Academy of Medical Sciences Ethics Committee approved the study protocol (ZS-1759), and this trial was registered on the Chinese Clinical Trial Registry (ChiCTR) at www.chictr.org.cn (ChiCTR2200055169). This study was strictly conducted in accordance with the guidance of Good Clinical Practice and the Declaration of Helsinki for research involving human studies. Written informed consent was obtained from all individual participants included in the study before enrollment.

ESD indications were referred to in the JGES guidelines (2016) [24]. Endoscopic curability (eCura) of resection was defined according to the Japanese Gastric Cancer Association (JGCA) guidelines and classified into eCura A, B, and C, in which eCura A was considered curative resection [26]. Complete resection was defined as an en bloc resection with negative horizontal and vertical margins and without lymphovascular invasion microscopically [27].

According to the standard ESD process, scopes of the operation were labeled with coagulation dots of 5 mm marked outside the endoscopic demarcation line of the targeted lesion (equivalent to its boundary) before circumferential incision, whereas the pathological demarcation line (PDL) was referred to as the neoplastic margin under microscopy with the definition of microscopic junctions between cancerous and noncancerous epithelia. The distance between marking dots and the PDL (referred to as D) in all slices of one en bloc resected lesion was standardly measured under microscopy and the shortest one was recorded as D_min. Notably, when the PDL was inside marking dots, D was recorded as a positive value; when the PDL extended on marking dots or even outside, D was recorded as 0 mm. Theoretically, the endoscopic and PDLs mentioned above should overlap with a D_min of 5 mm (shown in Fig. 1). Thus, considering that the subtext of specimens with a D_min >5 mm had absolutely negative horizontal margins, it was defined as precise delineation. Gastritis-like appearances were referred to as nonneoplastic epithelium covering the neoplastic lesion and dichotomized as presence or absence in this study [20], as shown in Figure 2.

In the study, the acute massive bleeding was referred to as active sputtering or oozing, requiring repeated hemostasis via needle knives, coagulation graspers, or endoclips. Synchronous gastric cancer and MGC were defined as gastric neoplastic lesions that occurred within and over 1 year after ESD, respectively [28].

All lesions of enrolled HpC-EGC patients were observed by GIF-H260Z endoscopy^® (Olympus, Tokyo, Japan) and combined CE with indigo carmine when necessary. At the beginning of the ESD operation, an independent endoscopist, who was blinded to the grouping information, delineated the targeted lesion. In detail, the endoscopist identified the targeted lesion under WLE and then delineated its margin mainly by NBI, ME-NBI, or CE, marking dots at 5 mm outside by electric coagulation. Subsequently, mucosal incision and submucosal dissection were performed by another experienced endoscopist (A.Y., X.W., T.G., Q.W.J., and Q.W.) with GIF-Q260J^® (Olympus, Tokyo, Japan), plus a standard electrosurgical generator ICC 200^® (ERBE Elektromedizin Gmbh, Tübingen, Germany) and a DualKnife^® (Olympus, Tokyo, Japan) following guideline specifications [29]. The two endoscopists mentioned above were sequentially present in the operation room instead of simultaneously, and there was no communication related to the patient during the whole procedure. Standardized information on the ESD operation was timely recorded as follows: (1) size, location, and macroscopic type of EGC lesions; (2) operation process and time; and (3) intraoperative complications. Postoperative complications were obtained from the hospital information system [30, 31]. All involved endoscopists’ experience was greater than 6 years in ME-NBI and ESD operations.

Resected specimens were independently examined by two experienced gastrointestinal pathologists (W.Z. and L.Z.), who were also blinded to group information. For controversial cases, two pathologists discussed to reach a consensus. In addition to general items, histopathology results included the following standardized information: (1) presence or absence of gastritis-like appearances; (2) neoplastic infiltration status at resection margins and marking dots; and (3) D and D_min. Endoscopic curability was assessed simultaneously.

H. pylori infection status of all enrolled HpC-EGC patients was re-evaluated by the ¹³C-UBT or RUT 3 months after the ESD operation. If any test were positive, the second line of Hp-ET would be prescribed. In addition, EGD with CE or ME-NBI was repeated at 3 months, 6 months, and 12 months after the ESD operation. After that, H. pylori infection tests, EGD, and abdominal CT scans were repeated annually.

The primary outcome of this study was to compare the precise delineation rate of HpC-EGC patients treated with preoperative and postoperative Hp-ET; the secondary outcomes comprised the D_min and negative marking dot specimen rate, plus the complete resection rate, the curative resection rate, and the presence of gastritis-like appearances. In detail, the precise delineation rate was calculated as the proportion of specimens with a D_min of >5 mm (the number of specimens with a D_min of >5 mm/the number of resected lesions). The negative marking dot specimen rate was calculated as the proportion of specimens in which all marking dots were negative (the number of specimens with all marking dots negative/the number of resected lesions). Similarly, the complete resection rate was referred to as the proportion of specimens that met complete resection, and the curative resection rate was referred to as the proportion of specimens with eCura A. Detailed data on ESD, including operation time, intraoperative complications (acute massive bleeding and perforation), postoperative complications (delayed bleeding and perforation), tumor recurrence (synchronous gastric cancer and MGC), hospitalization stay, and expenses, were also collected and analyzed.

This study was designed as a superiority trial. Considering the lack of previous studies directly associated with the primary outcome of our study, we took the research data on the diagnostic precision of demarcation using ME-NBI for undifferentiated-type EGC with Hp-ET and without Hp-ET for reference and an odds ratio of 8.5 was used to calculate the anticipated sample size [17]. With a two-sided confidence level of 95%, power of 80%, a ratio of unexposed to exposed of 1.0, and a percentage of unexposed with an outcome of 10%, the sample size of each group was estimated as 26 based on the Fleiss with continuity correction. In total, 52 HpC-EGC patients were needed for this trial.

All variables were calculated and demonstrated as the means and standard deviation or as numbers and percentages. Quantitative variables were compared by t test or Wilcoxon rank-sum test based on the Shapiro-Wilk test results. Where appropriate, categorical variables were compared by Pearson’s χ² test or Fisher’s exact test. Ordinal data were analyzed by the Wilcoxon rank-sum test. A p value of <0.05 was considered statistically significant. All statistical analyses were performed using SPSS (version 26.0, IBM Corporation, Armonk, NY, USA).

In the case of multiple neoplastic lesions in various resected specimens resected at one session of ESD operation, each lesion was included in the analysis as an independent lesion. Regarding multiple neoplastic lesions in one resected specimen, only the most severe and advanced lesion was analyzed as a representative.

During the study period, 57 consecutive HpC-EGC patients were enrolled, of whom 5 patients dropped out: 1 patient chose surgery and the other 4 patients received endoscopic resection in other hospitals. After randomization, 3 patients were excluded: in 1 patient, the targeted lesion was not recognizable under endoscopy after completing Hp-ET, and the subsequent ESD procedure was abandoned; in the other 2 patients, histopathological results of both resected specimens were low-grade intraepithelial neoplasia and mucosal inflammation. And all the above 3 patients were followed up and no neoplastic lesions were discovered. Finally, 49 HpC-EGC patients with 52 EGC lesions were included for analysis: 23 patients with 25 lesions in the eradication group and 26 patients with 27 lesions in the control group (shown in Fig. 3). None of them had an autoimmune gastritis history.

Demographic characteristics and endoscopic and histopathological findings are shown in Table 1. The average interval from the first day of administrating Hp-ET to the day of the ESD procedure in the eradication group was 34 days. Eradication success rates were 95.6% in the eradication group and 84.6% in the control group (p = 0.35). The average days of follow-up in the two groups were similar and greater than 1 year (p = 0.55). There were 1 and 4 patients with unsuccessful eradication at the first Hp-ET in the eradication and control groups, respectively, and they were given the second line of medication regimen, and all following re-evaluations were negative.

The histopathological characteristics of excised specimens in the two groups are shown in Table 2. For the primary outcome, according to the above definition, specimens with a D_min of >5 mm can be categorized as “precise delineation,” in which 10 lesions (40.0%) in the eradication group met this criterion compared to only 2 lesions in the control group (7.4%) (relative risk = 5.40, 95% CI 1.31–22.28, p = 0.008). Regarding other outcomes associated with resection margins, D_min in the control group was significantly smaller than in the eradication group (2.67 ± 2.30 mm vs. 4.17 ± 2.52 mm, p = 0.029). Positive marking dots were present in 4 of 27 lesions (14.8%) in the control group (shown in Fig. 4), while no lesion (0.0%) in the eradication group had it (p = 0.11).

The complete resection rate (96.0% vs. 100%) and curative resection (eCura A) rate (84.0% vs. 81.5%) were quite similar between the eradication and control groups without statistical significance; meanwhile, the rate of both negative horizontal and vertical margins was almost identical between the two groups, with a p value of >0.05. The presence of gastritis-like appearance in the neoplastic lesion of the eradication and control groups was 56.0% versus 55.6%, without significant difference. Of note, comparing the ratio of neoplastic lesion area to resected specimen area (lesion area/specimen area), the eradication group resembled the control group (0.15 ± 0.16 vs. 0.18 ± 0.18, p = 0.45).

The clinical data, such as ESD complications and long-term prognosis, etc., are summarized in online supplementary Table S1 (for all online suppl. material, see https://doi.org/10.1159/000534332). There were no significant differences between the two groups in those parameters. ESD operation time was 82.78 ± 58.91 min versus 70.32 ± 55.08 min in the eradication and control groups. Both 2 patients in each group experienced tumor recurrence. Hospitalization stays and expenses were 9.12 ± 3.26 days and 23,767.46 ± 9,747.76 CNY in the eradication group, compared with 9.46 ± 3.99 days and 23,643.97 ± 7,764.72 CNY in the control group. Intriguingly, with similarity in both intraoperative and postoperative complications in the two groups, more patients underwent acute massive bleeding in the eradication group (7 patients, 30.4%) than in the control group (4 patients, 15.4%) with a p value of 0.31.

H. pylori eradication has been proven to benefit EGC patients in the long term, but its impact on ESD treatment in the near term is still unclear. As a single-blind prospective controlled design, the present study showed that preoperative eradication therapy facilitates lesion boundary delineation and potentially reduces the risk of positive lateral margins. To the best of our knowledge, this is the first randomized controlled study addressing this issue to date. Given the high proportion of gastric cancer patients with current H. pylori infection [16], this study provides direct evidence for the optimal choice of eradication timing in these individuals facing endoscopic resection.

The more precise endoscopic delineation of boundaries after medicine administration of eradicated cases in this study conflicts with the results of some studies on GCAE. The following reasons may exist: (1) H. pylori infection causes mucosal inflammatory cell infiltration and makes the intercrypt distance wide [32, 33], resulting in blurring the boundary of the cancerous foci from the surrounding background mucosa [17]. And eradication therapy can reverse those morphological changes in the short term [34‒37]. This was also observed and verified in this study from histopathological perspectives that the inflammation of background mucosa significantly improved after Hp-ET in the eradication group compared to the control group. As shown in online supplementary Figure S1, moderate chronic and mild active inflammation of the representative patient from the eradication group has gotten relieved into mild chronic inflammation under Hp-ET, whereas another patient with the same grade of inflammation from the control group solidly remained throughout the trial; (2) eradication therapy does not increase gastritis-like appearances rate in the short term according to our data. This is consistent with the results from Kobayashi et al. [38] study, in which 6 months after eradication therapy is the lower limit needed for appearing gastritis-like appearances; and (3) all lesions in this study were delineated in combination with ME-NBI, which is more accurate for detecting micro or flat lesions than WLE alone [39, 40].

During endoscopic resection, imprecise delineation or even positive-marking dots immensely carry the risk of a positive lateral margin. All resected specimens in this study had negative lateral margins that were lower than previous data from our center [41], and the positive marking dot specimen rate was also lower than that reported in the previous study from Horiuchi et al. [17]. We believe that the main reason for the low rate of “precise delineation” was due to the stringent judgment criteria in the present study. According to the ESD procedure, the distance between marking dots and the endoscopic demarcation line of the targeted lesion was set as 5 mm, and the actual resection line was further 5 mm away from marking dots, implying a total distance of about 1 cm between the lateral resection edge and the tumor margin. However, this is too lenient for evaluating lesions of HpC-EGC patients with an average diameter <2 cm. In this case, we adopted “precise delineation” in this study, which required all marking dots to be located at a distance of 5 mm or more from the microscopical tumor margin, implying that the endoscopic boundaries almost overlapped with the microscopic pathological boundaries. Although the whole ESD operation procedure could not be strictly performed in a blinded manner, the blinded strategy of marking operators was designed to reduce the observer bias to some extent. By contrast, in the study by Horiuchi et al. [17], the observer performed markings in a non-blinded state and only two dots were set on the oral and anal sides, which may have reduced the marking precision. And there was no significant difference in the ratio of neoplastic lesion area to resected specimen area between the two groups, suggesting that even if there was investigator bias in our study, the effect on the results was slight.

The proportion of submucosal invasion was higher in the eradication group than in the control group (12.0% vs. 7.4%) at baseline, which may lead to morphological changes of the mucosa in the eradication group favoring delineation. However, EGC lesion sizes were similar between the two groups, suggesting comparable gross appearances at enrollment and not biases.

There was no statistically significant difference between the two groups in terms of ESD complications, long-term prognosis, hospitalization stay, and expenses, whereas the incidence of intraoperative acute massive bleeding and the percentage of patients with eCura C were higher in the eradication group. Although there was a higher proportion of lesions with locations in the upper and middle third parts of the stomach, where there may be rich in vessels, we had no cases that had uncontrollable bleeding during the operation or required termination of the operation [42‒44].

There are several limitations to this study. First, the single-center study with a small sample size limited the power of the conclusion. The fact was the relatively slow recruitment of this trial. As a tertiary medical center, most patients were referred to our hospital. During the study period, 97 of 328 EGC patients (29.6%) were diagnosed with current H. pylori infection, of whom up to 29 (30.0%) were excluded since preceding Hp-ET. Second, five endoscopists performed ESD operations, in which the inter-operator variability may influence outcomes. Last, experienced endoscopists possess the ability to infer H. pylori infection status from the mucosal appearance, making it difficult to rule out investigator bias completely, despite the blinded design.

In summary, for HpC-EGC patients, administrating eradication medication before ESD is beneficial for the precise delineation of lesions and reducing the risk of positive horizontal resection margins. Consequently, we recommend receiving standard quadruple Hp-ET before ESD for HpC-EGC.

We thank all the patients who participated in the study and all physicians, surgeons, pathologists, radiologists, and medical staff who helped recruit patients, collected data, provided professional advice and invaluable support, and gave intellectual comments.

This study protocol was reviewed and approved by the Chinese Academy of Medical Sciences Ethics Committee (ZS-1759), and this trial was registered on the Chinese Clinical Trial Registry (ChiCTR) at www.chictr.org.cn (ChiCTR2200055169). This study was strictly conducted under the guidance of Good Clinical Practice and in accordance with the Declaration of Helsinki for research involving human studies. Written informed consent was obtained from participants to participate in the study before enrollment.

All the authors have no conflicts of interest to declare.

This study was supported by the National High-Level Hospital Clinical Research Funding (2022-PUMCH-C-063 and 2022-PUMCH-B-024) and the CAMS Innovation Fund for Medical Sciences (CIFMS, 2022-I2M-C&T-B-014).

Conceptualization and methodology: Xi Wu, Weixun Zhou, Aiming Yang, Qiang Wang, and Shengyu Zhang. Acquisition and curation of data: Zhiyu Yan, Long Zou, Yuhao Jiao, Dingkun Xiong, Qingwei Jiang, Tao Guo, Yunlu Feng, Dongsheng Wu, Yamin Lai, Xuemin Yan, and Tao Xu. Formal analysis and interpretation of data: Zhiyu Yan, Long Zou, Qiang Wang, and Shengyu Zhang. Writing – original draft: Zhiyu Yan. Writing – review and editing: Shengyu Zhang, Weigang Fang, Xi Wu, and Weixun Zhou. Visualization: Zhiyu Yan and Long Zou. Study supervision: Xi Wu, Weixun Zhou, Aiming Yang, and Qiang Wang.

Zhiyu Yan and Long Zou contributed equally to this work.

All data generated or analyzed during this study are included in this article and its supplementary material files. Further inquiries can be directed to the corresponding author.