Feasibility of endoscopic evaluation of Helicobacter pylori infection status by using the Kyoto classification of gastritis in the population‐based gastric cancer screening program: A prospective cohort study

Abstract Background and aims We have started a new population‐based endoscopic gastric cancer screening program in Kurashiki city with consideration of Helicobacter pylori infection status based on endoscopic features. We aimed to verify the feasibility of this attempt in a prospective case‐registration study (UMIN000028629). Methods Data were collected from 1784 subjects without past eradication of H. pylori and who underwent endoscopic gastric cancer screening in Kurashiki Central Hospital Preventive Healthcare Plaza from September 2017 to June 2018. Endoscopic judgment of H. pylori infection status was made according to the Kyoto classification of gastritis. For comparison, a combination serum test of anti‐H. pylori antibody and pepsinogen I and II, the ABC method, was used. Results The endoscopic diagnoses were nongastritis, 1215 (68.1%); active or inactive gastritis, 469 (26.3%); and undefined, 23 (1.3%). With the ABC method as a reference standard, the false‐negative rate of the endoscopic judgment for H. pylori infection was 16.3% (95% confidence interval: 13.1%‐20.0%). Most false‐negative cases were of Group B in the ABC method, which is considered gastritis with mild mucosal atrophy. Antibody titers in this population were mostly in the weak‐positive range but clinically significant elevation of the antibody suggesting current infection was observed in some cases. Conclusions Endoscopic diagnosis of H. pylori infection status in a population‐based gastric cancer screening program is mostly reliable, but false‐negative results may occur, especially in patients with mild gastric atrophy. To avoid this limitation, we recommend adding H. pylori antibody test to the program.


| INTRODUCTION
Gastric cancer is decreasing in incidence, but it is still one of the major causes of cancer death in Japan. Helicobacter pylori is believed to be the pathogen responsible for developing gastric cancer. 1,2 Eradication of H. pylori reduces the gastric cancer risk [3][4][5][6][7] and mortality, 8 but the risk still remains even in the second decade after eradication. 9 The persuasive evidence of H. pylori in causation of gastric cancer notwithstanding, the population-based gastric cancer screening program in Japan has operated without considering patients' H. pylori infection status, that is, radiographic screening every year for everyone above 40 years of age. In 2015, endoscopic screening was reported to reduce gastric cancer mortality by 67% compared with radiographic screening, 10 and esophagogastroduodenoscopy (EGD) has become an option for population-based gastric cancer screening programs since 2016.
However, gastric cancer is still being surveilled without taking H. pylori infection into account, that is, EGD every 2 years in all people above 50 years of age; any tests for H. pylori infection, including serum anti-H. pylori antibody measurement, are not supported financially in the government-directed screening program.
In 2014, the Kyoto classification of gastritis was announced, and evaluating the status of H. pylori infection according to endoscopic findings of gastritis has been found reliable. 11,12 Thus, in 2017, we began our population-based endoscopic gastric cancer screening program in Kurashiki city, in which evaluation of H. pylori infection according to the Kyoto classification of gastritis was conducted as part. If this attempt works, gastric cancer screening with consideration of H. pylori infection can be achieved without further expense or change of the program. A major concern about this approach was that endoscopists in the screening program had various levels of skills and knowledge for judging endoscopic findings of gastritis according to the Kyoto classification.
Another method with links to gastric atrophy and H. pylori infection status in assessment of gastric cancer risk is the ABC method. It is a combination of a serum test for H. pylori antibody and serum concentrations of pepsinogen (PG). In subjects with severe atrophic gastritis after long-time H. pylori infection, checking H. pylori antibody titers alone is often insufficient to detect past, or even current, H. pylori infection, which can be detected by adding the PG test, a marker for gastric mucosal atrophy. 13 The ABC method has been touted as an effective nonendoscopic mass-screening method for gastric cancer 14 and was shown useful also for evaluation of H. pylori infection status, 15

| Serum test for H. pylori infection
For serum diagnosis of H. pylori infection, we used the ABC method 14 : Serum antibody to H. pylori was measured by an enzyme immunoassay method with E-plate (Eiken Chemical, Tokyo, Japan); a cutoff value of ≥3 U/mL was used, which is a value recommended for the gastric cancer screening program to reduce the frequency of falsenegative results 18,19 (an original cutoff value for clinical practice was ≥10 U/mL 14 ). PGI and II were measured by the chemiluminescent enzyme immunoassay method, Fujirebio, Tokyo, Japan); results were considered positive for gastric atrophy when PGI and PG I/II ratios were ≤70 ng/mL and ≤3.0, respectively. 20

| Statistical analysis
We calculated 95% CIs of the false-negative or false-positive rates using the Clopper-Pearson method. To assess the factors associated with the false-negative results of the endoscopic judgment, we used logistic regression analysis including factors which were likely to affect endoscopic findings. Odds ratios and 95% CIs were calculated, and P < .05 was considered statistically significant. For the statistical analysis, we used EZR (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (the R Foundation for Statistical Computing, Vienna, Austria).  13.1%-20.0%), and the false-positive rate was 6.8% (89/1312, 95% CI: 5.5%-8.3%). The false-negative rate of the ABC method, with endoscopic diagnosis of gastritis as a reference, was 19.0% (89/469, 95% CI: 15.5%-22.8%), and the false-positive rate was 6.0%
In cases with discrepancy between the endoscopic judgment and the ABC method, the recorded endoscopic images were carefully reexamined by two experienced endoscopists (MM, RH). The endoscopic diagnosis of gastritis after the re-evaluation was nongastritis in 1315 (73.7%) and active or inactive gastritis in 446 (25.0%) ( Table 3).  Table 4. Fifty of them were Group B of the ABC method, supposed to have gastritis with mild mucosal atrophy. In 39 of the 50 cases, the antibody titers were 3 $ <10, a weak-positive range, leaving the possibility that endoscopic judgment of nongastritis was correct in these cases. However, clinically significant elevation of the antibody to 10 or more, suggesting current H. pylori infection, was present in the other 11 cases. In the multivariate logistic regression analysis, weakly positive antibody titers (odds ratio = 15.7, 95% CI: 8.45-29.0, P < .01) and negative PG test (5.45, 2.14-13.9, P < .01) were identified as independent factors associated with the false-negative endoscopic judgment (Table 5). Typical endoscopic images of one such case is presented in   Figure 3.

| DISCUSSION
In this prospective study, we tried to address the major drawback of the population-based gastric cancer screening program in Japan, that

FUNDING
None.

CONFLICT OF INTEREST
There is no conflict of interest in this study. The work has not been published previously.

TRANSPARENCY STATEMENT
The lead author affirms that this manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned (and, if relevant, registered) have been explained.

DATA AVAILABILITY STATEMENT
All relevant data and materials are provided within the manuscript.