This single-institution study included patients who underwent curative surgery for GC at a tertiary hospital between April 2003 and March 2018. A retrospective review of medical records was performed to investigate H. pylori infection status, eradication therapy, and success rates according to the surgical method. This study was approved by the Institutional Review Board (IRB) of Seoul National University Bundang Hospital (IRB B-1902-523-107). Informed consent was waived by the IRB, due to the retrospective nature of this study.

H. pylori infection was diagnosed based on preoperative test results using three methods: urea breath test, rapid urease test, and histology. Among these, preoperative endoscopic biopsies (separate biopsies of the antrum and body of the stomach, in addition to the cancerous lesion) were performed in most patients. For assessment after eradication therapy, either the rapid urease test or histological examination was primarily used (approximately 90%) because of the low diagnostic accuracy of the urea breath test after gastrectomy [8], especially in patients who underwent follow-up endoscopic examinations after surgery. For the rapid urease test, tissue samples were collected from both the lesser and greater curvatures of the remnant stomach and tested using a single kit. For the histological analysis, tissue samples were also obtained from both the lesser and greater curvatures of the stomach for examination. Patients with clear preoperative and postoperative H. pylori test results and a documented history of eradication therapy, who had been followed up for at least one year, were included in the study. Cases without H. pylori test results, those in which the clinical diagnosis of infection status was uncertain, or those lost to follow-up were excluded from the analysis. Either a triple regimen or a sequential regimen was used as the first-line treatment, a bismuth quadruple regimen was used as the second-line treatment, whereas the third-line treatment was a quinolone-based regimen, according to the guidelines [9]. Failure of eradication therapy was defined as cases in which treatment using the above regimens either failed or was discontinued.

Patients were classified into four groups based on the reconstruction method after partial gastrectomy: 1) distal gastrectomy with gastroduodenostomy (Billroth-I reconstruction), 2) distal gastrectomy with gastrojejunostomy (Billroth-II or Roux-en-Y reconstruction), 3) proximal gastrectomy (with double-tract anastomosis), and 4) partial gastrectomy (including pylorus-preserving gastrectomy or wedge resection).

A total of 2827 patients underwent gastrotomy during the study period. Of these, 1510 patients were H. pylori-positive perioperatively, whereas 1121 were H. pylori-negative (including those with a history of eradication before GC diagnosis). Among the H. pylori-positive patients, 765 underwent eradication therapy, of which 729 achieved successful eradication and treatment failure was observed in 36 patients. Additionally, H. pylori spontaneously disappeared during follow-up in 206 patients (Table 1 and Fig. 1). There were differences in the proportion of patients who received eradication therapy and those who showed spontaneous negative conversion depending on the surgical method; the proportion of patients who underwent eradication therapy was higher in the pylorus-preserving gastrectomy group, whereas the rate of spontaneous negative conversion was higher in the proximal gastrectomy (with double-tract anastomosis) (21/41, 51.2%) and pyloruspreserving gastrectomy groups (9/19, 47.3%) (p<0.001). However, no significant differences were observed in the failure rate of eradication therapy based on the surgical method (p=0.875) (Table 2). In the regression and correlation analyses, the surgical method was not identified as a significant risk factor for treatment failure.

In patients who underwent gastrectomy for GC, the volume of the remnant stomach decreased, thereby reducing the amount of gastric mucosa in which metachronous cancer could develop. Additionally, the physiology of the remnant stomach changes after subtotal gastrectomy from that of the normal stomach because bile reflux can suppress or clear H. pylori from the remnant stomach [10], resulting in spontaneous negative conversion [11]. Earlier studies showed that almost 40% of patients had spontaneous regression postoperatively [12,13], and some studies suggests that spontaneous H. pylori clearance was related to the type of gastric reconstruction procedure due to the bile reflux rate, and that the duration of duodenogastric bile reflux was longer in the group of patients with H. pylori spontaneous negative conversion after surgery, compared to the group with H. pylori persistence [14]. Histological metaplasia, which is thought to be caused by severe bile reflux, was involved in the spontaneous regression of H. pylori [11,15]. This was potentially influenced by the limitations of H. pylori tests in surgically treated patients, especially in cases of atrophy and metaplasia. For example, the sensitivity and specificity of the rapid urease test decreases because the number of H. pylori decreases due to bile reflux and hypochlorhydria in the remnant stomach after surgery [16]. Furthermore, the diagnostic accuracy of the urea breath test also decreases as the gastric anatomy is altered, and the tested urea might be expected to pass through the stomach faster, or the results of the urea breath test could be influenced by bile acid reflux [17]. In summary, changes such as a reduction in gastric mucosal surface and a decrease in H. pylori colonization density, changes with gastric acidity and inflammation due to bile reflux, and alterations in gastric motility and emptying caused by resection of the pyloric ring and injury of vagus nerve occur after gastrectomy [8,18]. In this study, the rate of spontaneous negative conversion varied from 6.7% to 14.9% depending on the reconstruction method, which was similar to previous reports. This difference was thought to be due to the previously mentioned factors, which vary according to the reconstruction method used.

To date, well-designed double-blind randomized controlled trials on the beneficial effects of H. pylori eradication therapy after gastrectomy are limited. The probability of metachronous cancer after gastrectomy is low [19,20], and various mechanisms besides H. pylori are known to be involved [18]. The presence of reports suggesting that antibiotic susceptibility differs between the antrum and corpus of the stomach [21], has raised concerns that the treatment failure rate may vary depending on the extent of gastrectomy and the reconstruction method. However, an increasing number of studies have reported the benefits of postoperative eradication therapy. Notably, findings of the present study showed no significant differences in treatment failure rates between gastrectomy and non-gastrectomy patients, which could help build a consensus on more proactive treatment considerations.

This study is a preliminary report with several limitations. As a retrospective study, it lacks information on drug compliance and adverse effects and carries a risk of selection bias. Second, this study included patients who underwent first-, second-, and third-line eradication therapies, resulting in a relatively higher eradication success rate than that reported in other studies. The inability to compare the success rates of each eradication regimen is also a limitation of this study. Therefore, we plan to compare the success rates of first-, second-, and third-line eradication regimens according to the recon-struction methods performed in future studies. Additionally, the limitations of various tests for diagnosing H. pylori infection after gastrectomy are considerable; however, no standard diagnostic method has yet been established. Therefore, in this study, we aimed to diagnose H. pylori primarily through endoscopic examination using the rapid urease test and histology, whenever possible. Further studies are required to establish a standardized diagnostic method for H. pylori infection after gastrectomy. Finally, critical factors influencing treatment failure, such as the eradication regimens used and presence of antibiotic-resistant H. pylori strains, were not assessed in all patients.

Nevertheless, very few studies have compared the spontaneous negative conversion rate of H. pylori and the eradication failure rate according to the reconstruction method. This study found that the spontaneous negative conversion rate of H. pylori was 27.7% in non-treated patients (206 of 745), which was lower than previous studies that reported rates of 42.9% (60 of 140) [11] and 38.6% (27 of 70) [13]. The rates differed among the groups, with the Billroth-I group, Billroth-II group, proximal gastrectomy with double-tract reconstruction group, and pylorus-preserving gastrectomy group showing negative conversion rates of 24.8% 24.8%, 51.2%, and 47.3%, respectively. However, the differences were not as pronounced as those reported in previous studies, which found rates of 56.7%, 33.3%, 3.3%, and 6.7%, respectively [10]. Additionally, the failure rate of eradication therapy did not vary significantly depending on the reconstruction method used. Therefore, we plan to conduct a more refined analysis, including an evaluation of the eradication regimens used and a comparison with the nongastrectomy group.