Association of Chronic Periodontitis with Helicobacter pylori Infection in Stomach or Mouth: A Systematic Review and Meta-Analysis

Abstract Helicobacter pylori ( H. pylori ) infection and periodontitis are both inflammatory conditions associated with systemic diseases. Researchers have attempted to investigate the correlation between them. This systematic review and meta-analyses were conducted to investigate the association of H. pylori infection in the stomach and/or in subgingival plaque and gingival crevicular fluid with chronic periodontitis. The protocol was created according to the Preferred Reporting Items for Systematic review and Meta-Analysis Protocols (PRISMA-P) statement. The study was designed according to the Cochrane criteria. A comprehensive literature search was performed in MEDLINE, Scopus, and CENTRAL, combined with hand-searching and assessment of gray literature. The meta-analysis of the included studies was made by the Review Manager (RevMan) 5.4 software. The effect measure of the outcome was odds ratios with 95% confidence intervals. Heterogeneity was assessed by chi-square and I 2 . Four observational studies involving 818 subjects were included in this meta-analysis. The odds of oral H. pylori presence were higher in patients with chronic periodontitis, compared to healthy controls, with an odds ratio of 1.87 (95% confidence interval 0.85–4.10; p = 0.12). The odds of the presence of H. pylori in the stomach also were higher in patients with chronic periodontitis, with an odds ratio of 1.80 (95% confidence interval 0.82–3.95; p  = 0.15). There is no evidence for an association between chronic periodontitis and the prevalence of H. pylori , detected either in subgingival plaque and gingival crevicular fluid or in the stomach.


Introduction
Helicobacter pylori (H. pylori) is a gram-negative, spiral (Sshaped), microaerophilic organism that colonizes the gastric mucosa. 1 It is the leading cause of gastritis, peptic ulcer and gastric cancer, 1,2 mainly transmitted through the oral-oral 3 or fecal-oral routes. 3,4 Although the global prevalence of H. pylori infection is more than 50%, 5,6 higher rates are observed in developing countries (51%) compared with developed ones (35%). 7 H. pylori infection has been associated with several systemic diseases, such as iron deficiency anemia, 8 cardiovascular disease, [9][10][11][12] type 2 diabetes, 13,14 and pregnancy complications. 15 The diagnosis of H. pylori infection is set by the urea breath test (UBT), stool antigen test (SAT), serology, endoscopy, rapid urease test (RUT), histology, and polymerase chain reaction (PCR). Each of these methods carries advantages and disadvantages. 16 Periodontitis is an inflammatory disease of the supporting dental tissues whose manifestation and development are determined by the nature of the immune response to bacterial biofilms. The latter are typically composed of gram-negative microorganisms adhering to the hard dental surfaces, known as dental plaque. [17][18][19] In the advanced form of the disease, destruction of the alveolar bone is caused, which leads to the formation of periodontal pockets and retraction of the gums. 17,20 The prevalence of periodontitis is high, ranging from 20 to 50% worldwide. 21 Periodontal disease has been associated with a variety of chronic diseases, such as cardiovascular disease, 20,21 diabetes, 22 and pregnancy complications. 23 Research efforts focus on understanding the mechanisms of periodontal diseases. Traditional detection methods are insufficient in detecting nonculturable microbial species. On the contrary, metagenomic technology, as it is not based on microbial cultivation but on analysis of the functional genes of the microbial communities, interprets the microbial diversity, the synthesis of metabolic pathways, and the interaction between microorganisms and the environment. 24,25 Metagenomics studies microbial genetic material directly from environmental samples by sequence analysis. 26 This approach might lead to the detection of new and specific periopathogenic bacterial species and clarify the differences between symbiotic and dysbiotic biofilm. The latter is important for understanding the molecular mechanisms of the onset and progression of periodontitis and for providing targeted treatment. 25 The common features of H. pylori infection and periodontitis (inflammatory response, association with chronic diseases), 27 as well as the transmission of H. pylori through the oral route, led the researchers to investigate colonies in areas within the oral cavity in patients with chronic periodontitis. At the same time, an association has been established between periodontitis and H. pylori infection, suggesting that the oral cavity is a potential reservoir of H. pylori. 28,29 In interventional studies, successful eradication of gastric H. pylori resulted in improved periodontal disease. 30,31 As the studies published so far have been focused on the supragingival plaque or patients with periodontal diseases in general (including gingivitis), the present systematic review and meta-analysis aimed to investigate the association of H. pylori infection in the stomach and/or in specific oral cavity areas (subgingival plaque, and gingival crevicular fluid) with chronic periodontitis.

Protocol and Registration
The protocol was created according to the Preferred Reporting Items for Systematic review and Meta-Analysis Protocols (PRISMA-P) statement and registered to the International prospective register of systematic reviews (PROSPERO) database (Record ID: CRD 42021229036).

Data Sources
A comprehensive search was performed in three electronic databases (MEDLINE/PubMed, Scopus, Cochrane Controlled register of Trials) from conception until January 1st, 2021. Manual searching was performed on Google and Google Scholar. Gray literature was assessed via opengrey.eu, applying the search terms "chronic periodontitis" and "H. pylori." The search strategy in MEDLINE is presented in ►Table 1. 32

Inclusion and Exclusion Criteria
The studies were considered eligible if they (i) were randomized controlled trials and of observational type (cohort, cross-sectional, case-control) studies, (ii) were approved by ethics committees, (iii) were written in English, (iv) reported relevant data on two study arms [(a) patients with chronic periodontitis, (b) healthy controls], and (v) had adopted specific criteria for the definition of chronic periodontitis.
The diagnosis of chronic periodontitis had to be based on clinical or/and radiographic criteria, according to the 1999 classification system 33 or the 1989 classification system. 34 The studies were excluded if they (i) were of a low level of evidence (case-reports, case-series), (ii) included non-adult populations, and (iii) referred to specific conditions, namely pregnancy, orthodontic treatment, systemic diseases, malignancies, diabetes mellitus, auto-immune diseases, chronic use of non-steroidal anti-inflammatory drugs, antibiotics, proton pump inhibitors and bismuth salts use during the last two months, periodontal treatment (scaling, root planning) during the last six months, history of H. pylori eradication, gastrectomy, and less than 20 remaining teeth.

Study Records
Citations exported by the electronic databases in compatible file versions were imported to the Mendeley platform for managing study records. After removing the duplications, the records were exported to the Rayyan platform. 35 After reading the title and abstract, two reviewers (AG, IT) decided independently about the study eligibility. In relevant studies, the full text was assessed by two reviewers (AG, IT) independently. Conflicts were solved by a third reviewer (AT).

Data Extraction
A Microsoft Excel sheet was used for data extraction. Study identification data (name of the first author, year of publication, country) and population data (age, gender, sample size) were recorded. Regarding chronic periodontitis, the number of cases and controls were recorded. Regarding H. pylori infection, the number of positive and negative subjects (among total sample and cases with chronic periodontitis), diagnostic methods (histology, culture, rapid urease test [RUT], urea breath test, enzyme-linked immunosorbent assay, polymerase chain reaction [PCR], stool antigen test), and areas in which H. pylori was assessed (stomach, gingival crevicular fluid, subgingival plaque, periodontal pocket) were recorded. Data were extracted by two reviewers (AG, IT) independently. Conflicts were solved by a third reviewer (AT).

Outcomes
The outcome of the systematic review was the prevalence of H. pylori in chronic periodontitis and healthy control arms. The prevalence of H. pylori in the stomach and/or in specific oral cavity areas (gingival crevicular fluid, subgingival plaque) was recorded where available.

Bias Assessment and Confidence
The Newcastle-Ottawa Scale (NOS) was applied to assess the quality of observational studies. 36 Based on the collected quality stars, selection, comparability, and exposure (casecontrol studies)/outcome (cohort and cross-sectional studies) bias were evaluated as "low", "high" or "unclear" by two reviewers (AG, IT) independently. Conflicts were solved by a third reviewer (AT).
The Grading of recommendations, assessment, development, and evaluations (GRADE) tool was applied to assess the strength of the evidence. 37 Two reviewers (AG, IT) independently evaluated the evidence of the included studies as "high," "moderate," "low," or "very low." Conflicts were solved by a third reviewer (AT).

Statistical Analysis
The meta-analysis of the included studies was made by the Review Manager (RevMan) 5.4 software. The effect measure of the outcome (presence of H. pylori-binary) was odds ratios (OR) with 95% confidence intervals (CI). For the quantitative synthesis, a random-effects model (inverse variance) was applied. Heterogeneity was assessed by chisquare and I 2 . Subgroup analyses were performed based on the diagnostic method of H. pylori and the oral cavity area of H. pylori infection.

Results
The literature search located 1723 studies. After duplicate removal, 1600 studies were assessed based on the title and abstract. Of them, 66 studies were examined as full-text articles, and 13 were included in the qualitative synthesis (PRISMA flowchart-►Fig. 1). The reasons for exclusion are presented in ►Table 2. Four studies [38][39][40][41] were included in the quantitative synthesis (meta-analysis), as nine were excluded for an unclear definition of chronic periodontitis or violated the rule of independent observations in samples.
The summary of the characteristics of studies included in the meta-analysis is presented in ►Table 3. The characteristics of excluded studies are presented in ►Tables 4, 5, and 6.

Risk of Bias Assessment
The quality of the included studies was assessed by NOS. According to NOS, the risk of bias was low (►Fig. 2). A detailed graph of bias items for each included study is presented in ►Fig. 3.

Association between H. pylori and Chronic Periodontitis
The odds of presence of oral H. pylori in patients with chronic periodontitis were higher compared with healthy controls for oral (OR ¼ 1.87, p ¼ 0.12-►Fig. 4) and stomach (OR  (generalized periodontitis) OR (chronic periodontal inflammation)) OR (periodontitis)) OR (chronic periodontitis)) OR (mild periodontal disease)) OR (moderate periodontal disease)) OR (advanced periodontal disease)) OR (severe periodontal disease)) OR (periodontal disease)) OR (

Subgroup Analyses
Subgroup analysis was performed based on the detection method of oral H. pylori. When PCR was applied, the odds of the presence of oral H. pylori in patients with chronic periodontitis were lower compared with healthy controls (OR ¼ 0.71, p ¼ 0.47-►Fig. 6). When RUT was applied, the odds were higher (OR ¼ 2.88, p ¼ 0.01-►Fig. 6).

Sensitivity Analyses
The study results were not changed after excluding the Salehi et al 40

Evaluation for Publication Bias
Publication bias could not be assessed as the meta-analysis included only four studies.

Strength of the Evidence
The GRADE tool was used to assess the strength of the evidence. As all included studies were observational, their initial rating was low. Based on the predefined GRADE criteria, the overall strength of the evidence was low (►Table 7).

Discussion
The role of chronic periodontitis in the recurrence of H. pylori infection and/or the resistance to gastric H. pylori eradication has been demonstrated by several studies. 42 A two-way association between these two disease entities has been suggested. 29 The present meta-analysis provided evidence for an association between the presence of H. pylori in the subgingival plaque and chronic periodontal disease, as H. pylori was detected at a higher rate in the subgingival plaque of patients with periodontitis compared with healthy controls. This finding is consistent with a recent meta-analysis, which concluded that periodontitis is associated with oral H. pylori infection due to the presence of the bacterium in saliva and plaque in general. 43 Furthermore, original studies 44,45 using the PCR method arrived at the same conclusion by demonstrating the subgingival plaque as a supply reservoir of H. pylori infection in patients with periodontitis. However, other studies did not detect H. pylori in the subgingival plaque of patients with chronic periodontitis using the same method. [46][47][48] The reason for this divergence may be the differences in methodological procedures, population samples, 49,50 PCR primers, 51,52 sampling methods, and protocols. 51 Even the collection of the subgingival sample by paper cones differs from the use of periodontal curettes, as the cones can carry a smaller and, therefore, undetectable microbial load. 44 This fact may be the reason why, in the present meta-analysis, the significant association between subgingival H. pylori and periodontitis is lost when the sample includes Gingival Crevicular Fluid (GCF). Another reason for the divergence could be the transient presence of H. pylori in the oral cavity. Some authors argue that H. pylori exists in the oral cavity only as a transient organism, as other competing species colonize and predominate. 53 H. pylori infection may be indirectly related to periodontitis via periopathogenic oral cavity microbes that can compete and bind H. pylori strains. This binding of H. pylori by periodontal disease bacteria may lead to a crossantigenicity of H. pylori and periopathogens through heat shock proteins, resulting in an increased inflammatory immune response. 53,54 Furthermore, the transient presence of        H. pylori in the oral cavity may be due to its contamination by gastric fluid that reflux from the stomach. 47,55 The present study concluded that gastric H. pylori infection is not associated with periodontal disease, consistent with part 56,57 but not all of the literature. 38,58 Studies have supported the correlation between the H. pylori presence in the stomach and periodontitis, concluding that periodontal treatment contributes to the most effective and long-lasting eradication of gastric H. pylori. 30,59 However, the possibility of different H. pylori genotypes in the oral cavity and stomach of the same individual 60,61 may be the reason for the additional diagnostic difficulty. Cześnikiewicz-Guzik et al 62 did not find an association between the occurrence of H. pylori in the stomach and the oral cavity. This finding suggests that other factors, such as susceptibility to infection due to the acidic environment in the stomach, are the main cause of gastric infection with the bacterium. At the same time, the oral cavity can only serve as a means of transient food-related H. pylori contamination.
In the present meta-analysis, the correlation between subgingival H. pylori and periodontitis was significant only when H. pylori was detected by RUT, while this was not the case with PCR. RUT sensitivity ranges from 77 to more than 90%, and its specificity from 98 to 100%. [63][64][65][66] Song et al 60 concluded that the oral cavity may be a permanent H. pylori reservoir that can host multiple strains of the bacterium. The different sensitivity of the methods to different H. pylori strains could explain why RUT detected a higher percentage of H. pylori, as in the PCR method, depending on used primers amplificated specific strains. However, false-positive results of the RUT method are possible under certain conditions, as microorganisms, such as Klebsiella pneumoniae, Staphylococcus aureus, Proteus mirabilis, Enterobacter cloacae, and Citrobacter freundii, which colonize the oral cavity and/or stomach, have urease activity. 16 On the other hand, one possible reason that PCR detected H. pylori more frequently in controls could be the method's main disadvantage, which is the detection of non-living bacteria. 67 Two of this study's strengths are the comprehensive literature search and the assessment of the gray literature to restrict publication bias. Detecting H. pylori in both subgingival plaque and gingival crevicular fluid provides a better understanding of the association between the presence of H. pylori and chronic periodontitis, given the limited evidence from the literature. One additional strength of this review is the focus on chronic periodontitis, whereas most       studies have assessed the presence of H. pylori in periodontal diseases in general, including gingivitis. A couple of limitations are also observed in this study. The number of selected studies was low, restricting authors from conducting additional analyses, such as funnel plots. In each of these studies, a different method for detecting gastric H. pylori was performed, which can be explained by the absence of a gold standard detection method. In addition, an alternative of zero-count correction was performed by adding one event in each of the cells of study results by Silva et al. Although, in some meta-analysis tools, this procedure is made automatically by adding 0.5 in each of the cells, no difference was observed in the results by either including or excluding the study mentioned above, leading authors to make this amendment.
Although the term chronic periodontitis has been sufficiently described in previous classification systems, all subjects with periodontal pockets being more than 3 mm were considered periodontitis cases. In addition, it was not feasible to spot any studies in which H. pylori was detected in periodontal pockets, as it was designed in the protocol.
Future studies should be more specific regarding the level of periodontal destruction to investigate in detail whether there is a dose-response association between the presence of H. pylori and the stages of chronic periodontitis. There is also a need for more studies assessing H. pylori in gingival crevicular fluid, as the current evidence is limited.
In summary, there is no evidence of an association between chronic periodontitis and the prevalence of H. pylori, when the latter is detected either in specific oral cavity areas or in the stomach. The detection method of oral H. pylori can play an important role in affecting this association.