Helicobacter pylori infection with atrophic gastritis: An independent risk factor for colorectal adenomas

Qin-Fen Chen, Xiao-Dong Zhou, Dan-Hong Fang, En-Guang Zhang, Chun-Jing Lin, Xiao-Zhen Feng, Na Wang, Jian Sheng'Wu, Dan Wang, Wei-Hong Lin

Abstract

Key Words: Helicobacter pylori; Gastritis; Atrophy; Adenomas; Colorectal; Health-check

INTRODUCTION

Colorectal cancer is one of the most common human malignancies worldwide, and the fifth common cause of cancer death in China[1]. Due to genetic mutations, colorectal adenomas may develop into carcinoma[2,3]. Common risk factors, such as age, male gender, nonalcoholic fatty liver disease, metabolic syndrome, family history, smoking, alcohol consumption, diet and lifestyle, contribute to the development of colorectal neoplasms[4,5].

Helicobacter pylori(H. pylori) is a gram-negative, microaerophilic bacterium generally found in the stomach[6].H. pyloriinfection is associated with the development of gastric cancer[7]. In addition to its well-known association with gastric adenocarcinoma,H. pyloriis associated with numerous extragastric malignancies[8,9]. Inconsistent conclusions of the relationship betweenH. pyloriinfection and colorectal neoplasia were presented in previous studies. In the early years,H. pyloriinfection had been confirmed as a risk factor for colorectal neoplasm[10-14]. However, the association betweenH. pyloriinfection and development of colorectal neoplasia remains unclear in recent studies[15].

Gastric mucosal atrophy is a typical symptom of atrophic gastritis (AG). AG in 8.1% of patients per year results from a chronicH. pyloriinfection with a ten-fold increased risk[16,17]. It is well established that gastric cancer and/or adenomas are associated with higher rates of colorectal cancer. In addition, precancerous lesions such as dysplasia or AG are important risk factors for gastric adenomas and gastric cancer[18,19]. However, only limited studies have investigated the association between AG and colorectal neoplasia. One study reported that intestinal metaplasia, often accompanied by AG, was closely related to any type of colorectal neoplasia[13].

In contrast, another study showed that the presence of AG has insignificantly increased the risk of colon cancer[20]. In addition, a recent study showed a significant association between colorectal neoplasm and AG, which was diagnosed by Kimura and Takemoto criteria. However, this study did not have the criteria for a histologic diagnosis[21]. The relationship between AG and colorectal neoplasia, especially that betweenH. pylori-related AG and colorectal neoplasia, is still controversial.

Thus, the aim was to assess the relationship between colorectal adenomas andH. pylori-related AG based on the histologic diagnosis.

MATERIALS AND METHODS

Eligible subjects

This retrospective study analyzed records between August 2014 and August 2017 that were extracted from the Medical and Health Care Center at The First Affiliated Hospital of Wenzhou Medical University. Relevant information was obtainedviaa survey, utilizing a standard relevant questionnaire. Out of these 13400 individuals, 6086 individuals aged 30 years and older underwent a gastroscopy, colonoscopy,13Curea breath test and related pathological examination. Exclusion criteria were: A previous history ofH. pylorieradication therapy; incomplete colonoscopy; polyp resection; inflammatory bowel disease; and gastrointestinal cancers. Finally, the data of 6018 individuals were included in our analysis. The investigation conforms to the principles outlined in the Declaration of Helsinki. The study was approved by the ethical committee of The First Affiliated Hospital of Wenzhou Medical University Ethical Committee

Data collection

Baseline characteristics, including age, gender, smoking, alcohol consumption, previous medical history and family history, were obtained from the standard questionnaires. Physical parameters and laboratory assays, including body mass index (BMI), systolic blood pressure (SBP), diastolic blood pressure (DBP), fasting blood glucose (FBG), total cholesterol (TC), triglyceride (TG), low-density lipoprotein (LDL), and high-density lipoprotein (HDL) and were collected and recorded from reports of physical examination. All blood samples were drawn from antecubital vein sampling following an overnight fast. The tests for physical parameter measurements were operated by trained nurses.

Diagnostic criteria

H. pylori(HP) infection was diagnosed by the13C-urea breath test or a histological diagnosis of biopsied stomach specimens. All enrolled subjects were divided into HP (+) group and HP (-) group depending on the above check mentions. Also, subjects were divided into AG (+) group and AG (-) group depending on the histopathological results of the gastric mucosa. For further subgroup analysis, subjects were divided into the nonpolyp group, the nonadenomatous polyp group (including inflammatory polyps and hyperplastic polyps) and the adenoma group based on the results from colorectal biopsies. Advanced colorectal adenoma was diagnosed by an adenoma with a diameter of ≥ 10 mm, a significant villous component, high-grade dysplasia or any combination thereof[21]. Additionally, the size of the polyps was divided into two groups: 0-9 mm and 10 mm +. While the number of polyps was divided into two groups: One and two or more. Following full bowel preparation, GIF-H260 gastroscopy and CF-H260AI colonoscopy (OLYMPUS, Tokyo, Japan) were performed in all eligible subjects. The surgeries were performed by experienced gastroenterologists with standard protocol followed. All examinations were performed in 2 d.

Statistical analysis

SPSS software (SPSS version 23.0 for Windows) was used for analysis. Continuous variables for nonadenomatous polyps, adenoma and advanced adenoma were presented as mean ± standard deviation. Pearsonχ2tests for categorical variables and one-way analysis of variance or Kruskal–Wallis test for continuous variables were used to compare the baseline of the study population among the previously described groups. Associations of the risk factors with nonadenomatous polyps, adenoma and advanced adenoma were tested using univariate logistic regression and multivariate analysis. A two-sidedPvalue of < 0.05 was considered statistically significant.

RESULTS

Baseline characteristics of eligible subject

As shown in Table 1, a summary of the characteristics stratified by nonpolyp, adenoma, nonadenomatous polyp and advanced adenoma groups are presented. Of 6018 subjects studied, 2035 (33.8%) presented with colorectal polyps, 1012 (16.8%) with adenomas and 1023 (17.0%) with nonadenomatous polyps. Out of 1012 subjects in the adenoma group, there were 143 cases of advanced adenomas. The prevalence ofH. pyloriinfection in the nonpolyp group, adenoma group, nonadenomatous polyp group and advanced adenoma group were 48.6% (1936/3983), 53.0% (536/1012), 49.8% (509/1023) and 54.5% (78/143), respectively. The prevalence of AG in the nonpolyp group, adenoma group, nonadenomatous polyp group and advanced adenoma group were 8.7% (347/3983), 13.7% (139/1012), 11.3% (116/1023) and 14.7% (21/143), respectively. Overall, subjects with adenoma were older, had higher values of BMI, SBP, DBP, FBG, TC, TG, LDL, and lower values of HDL-cholesterol.

Association between H. pylori infection and adenoma

Based on the status of theH. pyloriinfection, all 6018 subjects were divided into HP (+) (2981, 49.5%) and HP (-) (3037, 50.5%). As reported in Table 2, the prevalence of adenoma in the HP (+) group was significantly higher than that of HP (-) group [unadjusted odds ratio (OR) = 1.1919, 95% confidence interval (CI): 1.037-1.367,P= 0.013; adjusted OR = 1.220, 95%CI: 1.053-1.413,P= 0.008, Table 3]. The mean age was not significantly different between the HP (+) and HP (-) groups. Compared to the HP (-) group, individuals in the HP (+) group had a higher proportion of men (P= 0.027, Table 2) and a higher prevalence of multiple colorectal polyps (P= 0.045). But the prevalence of nonadenomatous polyp, advanced adenoma, villous adenoma, adenoma size of ≥ 10 mm, single polyps, polyp size andH. pyloriinfection were similar (P> 0.05).

Association between AG and adenoma

Based on the AG status of all the 6018 subjects, we divided our cohort into two groups, the AG (+) group (602, 10.0%) and the AG (-) group (5416, 90.0%). Compared with the AG (-) group, subjects in the AG (+) group were older (P< 0.001, Table 4). The prevalence of adenoma in the AG (+) group was higher than that in the AG (-) group (unadjusted OR = 1.668, 95%CI: 1.352-2.059,P< 0.001, Table 4; adjusted OR = 1.237, 95%CI: 0.988-1.549,P= 0.064; Table 3). The prevalence of nonadenomatous polyps in the AG (+) group and AG (-) group was 19.3% and 16.7%, respectively (unadjusted OR = 1.340, 95%CI: 1.073-1.674,P= 0.010; adjusted OR = 1.103, 95%CI: 0.872-1.394,P= 0.413, Table 3). In addition, the prevalence of advanced adenoma in the AG (+) group and AG (-) group was 3.49% and 2.25%, respectively (unadjusted OR = 1.804 (95%CI: 1.121-2.903,P= 0.015; adjusted OR = 1.320, 95%CI: 0.805-2.165,P= 0.271, Table 3). The association of polyps with AG (+) was highest for individuals with more than one polyp (OR = 1.608, 95%CI: 1.302-1.985,P= 0.003). In patients with a polyp size of 0-9 mm, there existed a significant association between the prevalence of polyps and AG status (OR = 1.519, 95%CI: 1.275-1.809,P< 0.001).

Presence of both H. pylori infection and AG may increase the risk for adenoma significantly

According to the different statuses ofH. pyloriinfection and AG, the individuals in our study were divided into HP (-) AG (-) group, HP (-) AG (+) group, HP (+) AG (-) group and HP (+) AG (+) group to understand whetherH. pyloriinfection with AG increased the risk of adenoma. As reported in Table 5 and Table 6, the HP (+) AG (+) group had an approximately 1.5-fold risk for colorectal adenomas in comparison with that in the HP (-) AG (-) group (unadjusted OR = 1.964, 95%CI: 1.477-2.610,P< 0.001; adjusted OR = 1.491, 95%CI: 1.103-2.015,P= 0.009).

Presence of both H. pylori infection and AG also increase the risk for advanced adenoma

In subgroup analysis, the risk of colorectal adenomas was similar in either the HP (-) AG (-) group or HP (-) AG (+) group (unadjusted OR = 1.377, 95%CI: 0.618-3.064,P= 0.434), or between the HP (-) AG (-) group and HP (+) AG (-) group (unadjusted OR = 1.184, 95%CI: 0.825-1.699,P= 0.360). However, the presence ofH. pylori-related AG was related to a significant increased risk for advanced adenomas (unadjusted OR = 2.496, 95%CI: 1.366-4.562,P= 0.003; adjusted OR = 1.910, 95%CI: 1.022-3.572,P=0.043).

Table 1 Baseline characteristics of 6018 subjects

DISCUSSION

In this study, the potential roles ofH. pyloriinfection, AG andH. pylori-related AG in the progress of colorectal adenomas and advanced adenoma were investigated. According to previous research, the association betweenH. pyloriand colorectal adenomas remains unclear[20,22-26]. In our study,H. pyloriinfection was an independent risk factor for colorectal adenomas. The finding is consistent with current studies that indicate a positive correlation was revealed between colorectal adenomas andH. pylori. Additionally, HP (+) AG (-) may indicate a higher risk of colorectal adenomas. However, it was not associated with an increased risk of advanced adenomas. In our study,H. pyloriinfection was diagnosed by the results from the13C-urea breath test or a histological diagnosis of biopsied gastric specimen serology test that can accurately reflect a currentH. pyloriinfection[21]. With the development of detection technologies ofH. pyloriinfection, the role ofH. pyloriin the colorectal carcinogenesis may be revealed.

No significant association between AG and colorectal adenomas was observed in our cohort. Moreover, HP (-) AG (+) was not an independent risk factor for colorectal adenomas. Some subjects with HP (-) AG (+) may be affected with severe AG following a long-term infection withH. pylori. Theoretically, these patients may present with hypergastrinemia and have a higher risk of colorectal adenomas. However, our study did not indicate any correlation based on this hypothetical reasoning. In the multivariate analysis, the relatively small number (n= 279) of the HP (-) AG (+) group may have concealed the possible effects on colorectal adenomas.

After controlling all confounding factors, the ORs for colorectal adenomas in eligible individuals withH. pylori-related AG were higher than those in individuals of the HP (-) AG (-) group (adjust OR = 1.491, 95%CI: 1.103-2.015,P= 0.009). HP (+) AG (+) is independently associated with colorectal adenomas. Additionally, HP (+) AG (+) is significantly associated with an increased risk of advanced adenomas. However, no such association was observed in the HP (-) AG (+) or the HP (+) AG (-) group. Thisfinding is consistent with that of a recent study that indicated thatH. pyloriinfection along with AG increased the risk of both overall and advanced colorectal neoplasm[21]. ChronicH. pyloriinfection can lead to the occurrence of gastric mucosal atrophy[27]. In our study, the mean age in HP (+) AG (+) group was higher than in the HP (+) AG (-) group (52.3 yearsvs47.5 years). This can be explained as the individuals in the HP (+) AG (+) group may haveH. pyloriinfection for a longer period.

Table 2 Correlation between Helicobacter pylori infection and colorectal neoplasm

Table 3 Logistic regression model of the association between Helicobacter pylori infection, atrophic gastritis and colorectal neoplasm after adjustments for confounding factors

The presence of theH. pyloriinfection and AG increases the risk of colorectal adenoma. This may occurviavarious mechanisms. The cholecystokinin type B/gastrin receptor and gastrin are present in human colorectal polyps, and they are activated in the early stages of the adenoma-carcinoma sequence[28,29]. Persistent exposure toH. pyloriinfection directly induces the atrophic changes of the gastric body mucosa and increases the gastrin secretion. This has a nutritional effect on the growth and proliferation of epithelial cells and ultimately contributes to colorectal carcinogenesis[30,31]. In addition, hypochlorhydria caused byH. pylori-related AG may hamper protein assimilation, leading to an increase of some unabsorbed nutrients and metabolites[32]. Hypochlorhydria generates bacterial overgrowth and colorectal disorders, resulting in colorectal carcinogenesis[33].

Table 4 Correlation between atrophic gastritis and colorectal neoplasm

Table 5 Association between Helicobacter pylori infection, atrophic gastritis and colorectal neoplasm

Table 6 Logistic regression model of the association between Helicobacter pylori infection, atrophic gastritis and colorectal neoplasm after adjustments for confounding factors

Generalizability of findings in this study is limited by several factors. First, based on general health check-ups, a potential selection bias may have existed. In addition, the data affecting the changes of gastric mucosa, viz. dietary habit, was insufficient. Second, serum gastrin level, as a key mechanism in the progress of colorectal carcinogenesis, was not included in our analysis. Third, biopsy samples accounted for only 74% of the data. This may have potentially lowered the rate of gastric disease detection. Finally, our analyzable data were derived from a single center and local region in Chinese people, thereby limiting the ability to generalize our finding. Therefore, further multicenter research should be established to determine the potential association of individuals with other nations and ethnic groups. Despite these limitations, it is a novel study as we not only analyzed the relationship betweenH. pyloriinfection and colorectal adenomas but also further investigated the role of AG in colorectal carcinogenesis.

CONCLUSION

In summary, our study clearly demonstrated that subjects withH. pylori-related AG did have an increased risk for colorectal adenoma. Due to the high prevalence ofH. pyloriinfection and colorectal cancer in the Chinese population, strict colonoscopy screening and surveillance are necessary for patients withH. pyloriinfection, especially for those withH. pylori-related AG.

ARTICLE HIGHLIGHTS

ACKNOWLEDGEMENTS

The authors thank all the staff at the Medical and Health Care Center of The First Affiliated Hospital of Wenzhou Medical University for their assistance.