We examined whether colonization of selected oral pathogens is associated with gastric precancerous lesions in a cross-sectional study. A total of 119 participants were included, of which 37 were cases of chronic atrophic gastritis, intestinal metaplasia, or dysplasia. An oral examination was performed to measure periodontal indices. Plaque and saliva samples were tested with real-time quantitative PCR for DNA levels of pathogens related to periodontal disease (Porphyromonas gingivalis, Tannerella forsythensis, Treponema denticola, Actinobacillus actinomycetemcomitans) and dental caries (Streptococcus mutans and S. sobrinus). There were no consistent associations between DNA levels of selected bacterial species and gastric precancerous lesions, although an elevated but non-significant odds ratio (OR) for gastric precancerous lesions was observed in relation to increasing colonization of A. actinomycetemcomitans (OR = 1.36 for one standard deviation increase, 95% Confidence Interval = 0.87–2.12), P. gingivalis (OR = 1.12, 0.67–1.88) and T. denticola (OR = 1.34, 0.83–2.12) measured in plaque. To assess the influence of specific long-term infection, stratified analyses by levels of periodontal indices were conducted. A. actinomycetemcomitans was significantly associated with gastric precancerous lesions (OR = 2.51, 1.13–5.56) among those with ≥ median of percent tooth sites with PD≥3 mm, compared with no association among those below the median (OR = 0.86, 0.43–1.72). A significantly stronger relationship was observed between the cumulative bacterial burden score of periodontal disease-related pathogens and gastric precancerous lesions among those with higher versus lower levels of periodontal disease indices (p-values for interactions: 0.03–0.06). Among individuals with periodontal disease, high levels of colonization of periodontal pathogens are associated with an increased risk of gastric precancerous lesions.
Citation: Salazar CR, Sun J, Li Y, Francois F, Corby P, Perez-Perez G, et al. (2013) Association between Selected Oral Pathogens and Gastric Precancerous Lesions. PLoS ONE 8(1): e51604. https://doi.org/10.1371/journal.pone.0051604
Editor: Songtao Shi, University of Southern California, United States of America
Received: August 1, 2012; Accepted: November 2, 2012; Published: January 7, 2013
Copyright: © 2013 Salazar et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This research was supported by U.S. National Institute grants R21DE018438, P30CA016087, ES000260, K23CA107123, and T32-DE007255. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Gastric cancer is the fourth most common cancer and second highest in mortality worldwide . In 2008, there were approximately 989,000 new gastric cancers diagnosed globally and 738,000 deaths. While environmental and lifestyle factors including H. pylori colonization, cigarette smoking, dietary intake of salt and preserved foods, and low intake of Vitamin C and vegetables, all contribute to the etiology of the malignancy –, these risk factors could only explain less than 60% of gastric cancer incidence .
For the most common “intestinal” subtype of gastric adenocarcinomas, which account for approximately 70% of gastric cancer cases , , a pre-neoplastic sequence has been defined from chronic superficial gastritis through atrophic gastritis, intestinal metaplasia, and ultimately to dysplasia and adenocarcinoma , . Studies of gastric precancerous lesions can be used to identify risk factors that play a role in the development of gastric cancer. Several case-control studies of gastric precancerous lesions, predominantly atrophic gastritis and intestinal metaplasia, have identified common risk factors for gastric cancer such as cigarette smoking, old age, and low levels of fruit or vitamin C intake –.
The association of self-reported tooth loss or poor oral hygiene with increased risk of gastric cancer has been reported in several epidemiologic studies with diverse populations –. We have previously reported a significant positive association between active gingival bleeding and presence of gastric precancerous lesions in a cross-sectional study of an endoscopic population , consistent with the hypothesis that periodontal disease may increase the risk of gastric cancer. However, these studies lack information on specific oral pathogens necessary to elucidate the role of oral health in gastric cancer development. The investigation of specific oral pathogens related to gastric cancer or gastric precancerous lesions may help identify the underlying mechanisms and generate knowledge that may lead to improved clinical treatment and interventions. In this study, we examined the association between selected oral pathogens and presence of gastric precancerous lesions. We considered microbes that are etiologically linked with periodontal diseases , including Porphyromonas gingivalis, a gram-negative oral anaerobe commonly associated with adult periodontal disease ; Tannerella forsythensis, an anaerobic gram-negative member of the Cytophaga-Bacteroides family implicated in periodontal diseases, Treponema denticola, a motile and highly proteolytic bacterium associated with the incidence and severity of human periodontal disease, and Actinobacillus actinomycetemcomitans, a gram-negative facultative anaerobe, which is not only a typical cause of periodontitis but also associated with systemic infections , . We also measured levels of Streptococcus mutans and S. sobrinus, gram-positive bacteria, which are significantly associated with tooth decay , .
A detailed description of the study has been presented elsewhere . Briefly, subjects ≥40 years of age undergoing an upper endoscopy at Bellevue Hospital Center in New York City were invited to participate in the study. Information about demographic and lifestyle factors was collected using structured questionnaires administered by a trained interviewer.
Serum samples were collected at the time of endoscopy and were tested for H. pylori using ELISA for IgG antibodies to H. pylori whole cell antigens in duplicate and in parallel with known positive controls (cutoff for positivity was an optical density ratio >1.0 ). IgG antibodies to cagA were measured by ELISA (the cutoff for positivity was an optical density ratio >0.35 ).
Endoscopic evaluation was performed using Olympus GIF-XQ140 or GIFQ-160 videoendoscopes following existing protocol . A total of 8 biopsies of approximately 2×2×2 mm in dimensions were obtained from the stomach (4 from the antrum, and 2 each from the corpus and fundus) and 2 from the distal esophagus 2 cm above the gastroesophogeal junction. Biopsy specimens were processed according to pre-established standardized protocols and reviewed by an experienced gastrointestinal histopathologist (Z.P.) designated to the study . The pathologist reviewed all biopsy specimens and was blinded to questionnaire data and findings from the oral examination . Parameters evaluated for gastric biopsies in cluded (a) degree of acute inflammation (b) atrophy, and (c) intestinal metaplasia. The degree of gastritis was scored using the updated Sydney System . Participants were categorized according to the most advanced lesion found. Results from biopsies were classified as cases and non-cases: 1) cases had chronic atrophic gastritis, intestinal metaplasia, or dysplasia; and 2) non-cases had either chronic superficial gastritis or no evidence of gastric lesions.
The Institutional Review Boards of New York University School of Medicine and Bellevue Hospital Center approved all procedures involving human subjects. Written informed consent was obtained from all subjects at time of enrollment.
Overall, 283 eligible participants were approached, and 162 were recruited between April 2009 and April 2011. One hundred thirty one (131) subjects underwent an oral health clinical assessment and provided saliva and plaque samples for microbiological evaluation. Participants and non-participants were relatively similar in demographic distribution; roughly two thirds of non-participants were of Asian or Hispanic race/ethnicity, 60% were female, and their average age was 58 years. The top three reasons for declining participation (80%) included: not having time for the oral exam, not interested in the present research, and not wanting to participate in research studies in general. As of April 2012, complete data on selected oral pathogens was available from 119 participants (37 cases and 82 non-cases).
Oral examination and collection of saliva and plaque sample
Details of the oral examination have been previously presented elsewhere . Within 1–4 weeks following upper endoscopy, a dental hygienist calibrated in all assessments performed a comprehensive oral examination on all participants, and was blinded to biopsy results of gastroenterology. Periodontal assessments were standardized in a calibration session performed prior to the study and followed National Institute of Dental and Craniofacial Research diagnostic criteria, used previously in the Oral Health Survey of U.S. Adults . Periodontal examinations involved assessments at six tooth sites (mesiobuccal, buccal, distobuccal, distolingual, lingual and mesiolingual) of all teeth present, using a manual North Carolina 15 periodontal probe . Levels of clinical attachment loss (CAL), defined as the distance from the cemento-enamel junction to the free gingival margin in millimeters, and periodontal probing depth (PD), defined as the distance from the free gingival margin to the bottom of the pocket to the nearest whole millimeter, were recorded for each of the six tooth surfaces assessed. Bleeding on probing (BOP) was recorded dichotomously for each tooth surface and deemed positive if it occurred within 15 seconds after the assessment of probing depth.
All subjects were asked to chew a piece of paraffin wax and then to gently expectorate 2–5 ml of saliva directly into a sample collection tube, on ice. A total of 6 plaque samples were collected from each subject with sterile curettes; 4 supragingival plaque samples from the 1st molar or most posterior tooth in each quadrant and 2 subgingival plaque samples from teeth exhibiting the deepest periodontal pockets. The plaque samples were immediately transferred to a pre-labeled sterile sample vial containing 200 µl TE buffer. Both saliva and plaques samples were vortex-mixed thoroughly for 30 sec, immediately placed into a container containing ice, and transferred to the laboratory within 1 h for further processing.
Quantitative real-time PCR assays
Quantitative real-time PCR was performed to detect the presence/absence and to quantify targeted bacterial DNA in the saliva and plaque samples. Bacterial genomic DNA was isolated using the MasterPure DNA purification kit (Epicentre, Madison, WI) following previously established procedures . After DNA isolation, the total genomic DNA concentration of each sample was quantified using a UV spectrophotometer at 260 nm and 280 nm (NanoDrop 2000C, Thermo Scientific, Wilmington, DE). The final concentration of each DNA sample was adjusted to 10 ng/µl for real-time qPCR. The qPCR reaction mixture contained a total volume of 25 µl, which consisted of 10× buffer, 2.5 mM/ea dNTPs, 10 pmole/µl primers, 5 U/µl Taq, 50 mm MgCl2, and 20 ng/µl DNA template. The amplifying reaction of DNA in a volume of 25 µl was composed of 1× QuantiTect SYBR Green PCR Master Mix (Qiagen), 10∼100 ng of total genomic DNA and 0.4 µM of each primer. The species-specific PCR primers used in this study are listed in Table S1. All PCR programs used for the amplifying reactions are listed in Table S2. Serial dilutions (10-fold) of standard DNA concentrations of P. gingivalis (ATCC33277), T. forsythensis (ATCC43037), T. denticola (ATCC35404), A. actinomycetemcomitans (ATCC29522), S. mutans (UA159), and S. sobrinus (OMZ176), were used in each reaction as external standards for absolute quantification of the targeted bacteria. DNA levels for all 6 bacterial species were measured in saliva samples, and 4 bacterial species relevant to periodontal disease (A. actinomycetemcomitans, P. gingivalis, T. denticola, T. forsythia) were measured in each plaque sample. We tested for the colonization of S. sobrinus and S. mutans in saliva rather than in plaque because the plaque samples were collected from specific tooth sites exhibiting the deepest periodontal pockets and may not truly reflect the overall prevalence of these cariogenic microorganisms in the oral cavity. Use of our real-time PCR method involving stimulated whole saliva has been a standard and well-documented procedure for evaluating prevalence of S. sobrinus and S. mutans , . The analytical specificity of the fluorescence signal was determined on the basis of subsequent melting curve analysis. All reactions were carried out in duplicate, and the final analyses were based on the mean of the two reactions. Output data were analyzed using Opticon Monitor 2 software (MJ Research).
We first conducted descriptive analyses comparing participants with gastric precancerous lesions and those without gastric precancerous lesions in terms of socio-demographic, lifestyle characteristics, and periodontal disease indices, including CAL, BOP, and PD. We estimated Odds Ratios (ORs) for gastric precancerous lesions in relation to bacterial DNA levels in saliva samples and plaque samples by entering each measure as a continuous variable in serially adjusted models. DNA levels for each bacterial species measured in plaque were averaged for each individual (6 samples per person). Log-transformations of bacterial DNA levels were considered to postulate that disease risk is proportional to a power of dose in bacterial DNA levels (a log-linear relation between log odds and exposure), a relationship known to occur frequently from both human and animal studies . To facilitate interpretation, we estimated ORs for gastric precancerous lesions in relation to a one standard deviation (SD) increase in log-transformed bacterial DNA levels of each of the bacterial species.
We also created a cumulative summary score of bacterial burden for bacterial species etiologically related to periodontal disease (A. actinomycetemcomitans, P. gingivalis, T. denticola, T. forsythia , ) and dental caries (S. mutans and S. sobrinus , ). The average colonization level of each species was natural log-transformed and standardized by dividing each person-specific value by the log-transformed population standard deviation (SD). One SD on the natural log scale was hence equivalent across microbes. The standardized values were then summed and grouped accordingly. We estimated ORs for gastric precancerous lesions in relation to the cumulative summary score entered as a continuous variable. ORs were adjusted for major risk factors of gastric cancer that may be related to oral health, and these factors included age, sex, ethnic background, smoking status, body mass index, H. pylori status, and educational attainment.
Since clinical periodontal disease is a manifestation of long-term infection of periodontal pathogens, we conducted stratified analyses to assess whether the association between periodontal disease-related pathogens and gastric precancerous lesions differs by periodontal indices, including CAL, BOP, and PD. The median values for percentage of tooth sites with BOP, CAL ≥3 mm, and PD≥3 mm were considered cut-points in stratified analyses. Significance of the interaction was tested using a cross-product term between a given periodontal disease index and a given periodontal disease bacterial species (or the cumulative burden of all four species), all entered as continuous variables in the model. Sensitivity analyses were performed using different cut-points for periodontal indices (e.g., percentage of sites with CAL≥5 mm and PD≥4 mm). Exploratory analyses were conducted to assess whether the exposure-disease association differed by H. pylori status.
We performed all statistical analyses using SAS 9.2 (SAS Institute, Inc., Cary, North Carolina). All tests were two sided and P<0.05 was considered significant.
Participants were largely of Hispanic ethnicity (45.4%), female (63.0%) and had a mean age of 57 years. Cases had a significantly higher prevalence of smoking (p = 0.02) and were more likely to have tested positive for H. pylori colonization (particularly with cagA strain, p = 0.07) than non-cases (Table 1). When we compared mean levels of periodontal indices by case group, we found a higher mean percentage of gingival bleeding sites (p = 0.04) among cases than non-cases. There were no other significant differences in periodontal disease indices by case status. As expected, we observed that the means of log-transformed periodontal pathogen DNA levels in plaque increased across tertiles of periodontal indices (Table S3). The DNA levels of periodontal pathogens were generally higher among those who did not floss as compared to those who did floss, although most did not reach statistical significance (data not shown).
Age and gender adjusted mean bacterial DNA levels in either saliva or plaque samples were not significantly different by case status (Table 2). Results from multivariate logistic regression models using log-transformed bacterial DNA levels appear generally similar. In saliva, there was no evidence that levels of caries-related bacteria were related to gastric precancerous lesions. For periodontal disease-related bacteria, DNA levels for T. forsythia were significantly inversely associated with gastric precancerous lesions in a model adjusted by age and gender (OR = 0.59, 0.40–0.88) and in the fully adjusted model (OR = 0.61, 0.38–0.98). There was no evidence that levels of other periodontal disease-related bacteria in saliva were related to gastric precancerous lesions. In plaque, however, there was no evidence of an association between DNA levels for T. forsythia. A. actinomycetemcomitans and T. denticola in plaque samples were positively associated with gastric precancerous lesions (ORs for one SD increase in log-transformed bacterial levels: 1.36, p = 0.17 and 1.34, p = 0.23, respectively; model 2). However, these associations did not reach statistical significance. Using the summary measure of bacterial burden in plaque, we similarly found a positive albeit non-significant association between one SD increase in log-transformed bacterial burden level and presence of gastric precancerous lesions in a fully adjusted multivariable model (OR = 1.22, 95% confidence interval: 0.77–1.92). Sensitivity analyses were performed using different cut-points for periodontal indices (e.g., percentage of sites with CAL≥5 mm and PD≥4 mm). However, the results were similar and are not shown here.
To assess the effect of more specific long-term infection, we further investigated whether the association between periodontal pathogens in plaque and gastric precancerous lesions differed by periodontal disease indices (Table 3). The odds ratio for gastric precancerous lesions associated with one SD increase in levels of A. actinomycetemcomitans was greater (OR: 2.51, 95% CI = 1.13–5.56) among participants with a higher percentage of tooth sites (≥ the median) with PD≥3 mm, after controlling for gender, age, race, smoking status, educational attainment, BMI, and H. pylori status. By contrast, this effect was considerably attenuated among those with a lower percentage of sites (< the median) with PD≥3 mm (OR = 0.86, 95% CI = 0.43–1.72). The association between A. actinomycetemcomitans and gastric precancerous lesions was also elevated among individuals with a higher percentage of sites with CAL≥3 mm (OR = 1.92, 95% CI = 0.96–3.85) and a higher percentage of BOP (OR = 1.38, 95% CI = 0.72–2.65). Similar patterns of associations were evident for other periodontal disease-related pathogens. When the cumulative summary score of all four bacterial species was used in the analyses, we observed that the association between one SD increase in the summary score and gastric precancerous lesions among participants with higher levels of periodontal disease (i.e., percentage of sites with BOP, PD≥3 mm, or CAL≥3 mm) was significantly or marginally significantly stronger than those with lower levels of periodontal disease (p-values for interaction ranged from 0.03–0.06). Lastly, in exploratory analyses, we investigated whether the associations between oral pathogens and gastric precancerous lesions differed by H. pylori. The associations were not appreciably different by H. pylori status and the results are therefore not shown.
Previous studies have reported positive associations between poor oral health and tooth loss with gastric cancer –. Ours is the first to investigate associations between DNA levels of oral pathogens and presence of gastric precancerous lesions. In the present study, there was no consistent association between DNA levels of selected bacterial species etiologically related to periodontal disease and presence of gastric precancerous lesions. Among those with higher levels of periodontal disease, periodontal pathogens measured in plaque were positively related to presence of gastric precancerous lesions, independent of gender, age, race, smoking status, educational attainment, BMI, and H. pylori status. The findings support the role of specific manifestations of periodontal disease and periodontal infection in the development of gastric cancer.
It has been proposed that bacterial infections are linked with gastric cancer through the induction of systemic or gastric chronic inflammation . In the classic example of H. pylori, infection with one pathogen alone triggers chronic inflammation and leads to gastric non-cardia adenocarcinoma –. Periodontal pathogen A. actinomycetemcomitans has been associated with increased secretion of interleukins 1α and 1β and tumor necrosis factor α, cytokines that are involved in the inflammatory response , . High serum titre to P. gingivalis and the presence of periodontal disease are independently related to high C-reactive protein levels . Recently, polymorphisms in the gene of Interleukin 6, a pro-inflammatory cytokine, were associated with A. actinomycetemcomitans, P. gingivalis, T. denticola, and T. forsythia . Periodontal disease is a set of inflammatory conditions that leads to destruction of the supporting alveolar bone, connective tissue, and eventual tooth loss . In this study, DNA levels for T. forsythia in saliva were significantly inversely associated with gastric precancerous lesions. However, there was no evidence for this in plaque. A non-significant positive association was observed between other periodontal disease-related bacteria measured in plaque and gastric precancerous lesions. Among individuals with more severe symptoms of periodontal disease, we observed consistent and positive associations between periodontal pathogens and gastric precancerous lesions. It is possible that higher levels of periodontal pathogens in individuals with a higher percentage of BOP and deeper periodontal pockets are indicative of more specific long-term infection that has a more pronounced effect on inflammation. There is also some evidence to show that higher levels of periodontal pathogens are associated with more aggressive forms of periodontitis , . Importantly, the positive associations among those with periodontal disease were independent of H. pylori and other established risk factors. If results are confirmed in future studies, the combined status of periodontal disease indices and their corresponding pathogens may be considered risk factors for gastric precancerous lesions and gastric cancer. Since bacterial profiles and oral health conditions are modifiable, identification of bacterial risk factors of malignancy could have important implications for cancer prevention.
In the present study, the positive associations between selected pathogens or their cumulative burden score and gastric precancerous lesions were similar across bacterial species among participants with higher levels of periodontal disease indices. The findings suggest a potential role of multiple periodontal disease-related bacteria. An emerging concept in microbial pathogenesis is the notion that the community functions as a pathogen and contributes to disease development collectively, rather than through the action of one specific member alone , . Future studies are needed to comprehensively investigate the role of other oral pathogens or oral microbiomes related to inflammation and periodontal disease in gastric cancer development.
There were several strengths of the present study. First, clinical indices of periodontal disease were measured objectively in a consistent fashion for all participants. Second, the availability of data on important risk factors of gastric cancer and potential confounders enhanced the ability to control for potential confounding and strengthened the validity of the study findings. Third, the bacterial specificity of our findings lends further support to the role of periodontal disease in the development of gastric cancer. In our study, no apparent relationship was evident between caries-related bacteria and gastric precancerous lesions. This is consistent with our previous study, which found no association between any clinical measure of caries (number of decayed, missing, or filled teeth/tooth surfaces) and gastric precancerous lesions20. In addition, the association between selected periodontal disease bacteria and gastric precancerous lesions were more evident for bacterial levels measured in pooled plaque compared with levels measured in whole saliva. Levels of microorganisms in subgingival plaque may be less diluted than in saliva and therefore more sensitive to detection by real-time qPCR.
Several potential limitations, however, should also be noted. First, the study had a small sample size and therefore we would have limited power to assess associations between oral pathogens and gastric precancerous lesions if oral pathogens were treated as categorical variables. We also investigated a number of comparisons in the study. Since p-values from the various models were not independent and the work was hypothesis-oriented, we adjusted p-values using the Benjamini–Hochberg false discovery rate (FDR) method to account for multiple comparisons . After adjustment for FDR, the adjusted p-values for the interaction effects of % CAL≥3 mm with Pg and Tf remained marginally significant (p = 0.08), while the significance of the other interactions did not (FDR-adjusted p-value >0.15). Moreover, the inverse association of T. forsythia with gastric precancerous lesions was no longer significant (FDR-adjusted p-value = 0.25). Larger studies are needed to confirm our findings and to further investigate the full dose-response relationships. Second, the study is cross-sectional and precludes any definitive statement regarding a causal association. However, DNA levels of oral pathogens and periodontal disease indices were measured objectively and were blinded to case status. Although we cannot also exclude the possibility of potential selection bias, it is not likely that cases with severe periodontal disease preferentially volunteered to be included in the study, since the hypotheses were not known to participants and were unaware of their lesion status at the time of recruitment for the oral health examination.
In conclusion, we found an overall non-significant positive association between DNA levels of selected oral pathogens related to periodontal disease, and the associations were stronger among individuals with higher levels of periodontal disease indices. Our findings highlight the role of periodontal infection in the development of gastric cancer.
Species-specific primers that were utilized for real-time quantitative PCR (qPCR).
Programs that were utilized for real-time quantitative PCR (qPCR).
Conceived and designed the experiments: YC YL FF AD. Performed the experiments: YC YL FF GPP PC. Analyzed the data: YC JS CRS YL. Contributed reagents/materials/analysis tools: YL ZP GPP. Wrote the paper: CRS YC.
- 1. Ferlay J, Shin HR, Bray F, Forman D, Mathers C, et al. (2010) Estimates of worldwide burden of cancer in 2008: GLOBOCAN 2008. Int J Cancer 127: 2893–2917.
- 2. Terry MB, Gaudet MM, Gammon MD (2002) The epidemiology of gastric cancer. Semin Radiat Oncol 12: 111–127.
- 3. Crew KD, Neugut AI (2006) Epidemiology of gastric cancer. World J Gastroenterol 12: 354–362.
- 4. Correa P, Piazuelo MB, Camargo MC (2004) The future of gastric cancer prevention. Gastric Cancer 7: 9–16.
- 5. Engel LS, Chow WH, Vaughan TL, Gammon MD, Risch HA, et al. (2003) Population attributable risks of esophageal and gastric cancers. J Natl Cancer Inst 95: 1404–13.
- 6. Lauren P (1965) The two histological main types of gastric carcinoma: diffise and so-called intestinal-type carcinoma. An attempt at a histo-clinical classification. Acta Pathol Microbiol Scand 64: 31–49.
- 7. Correa P (1988) A human model of gastric carcinogenesis. Cancer Res 48: 3554–3560.
- 8. Correa P, Haenszel W, Cuello C, Tannenbaum S, Archer M (1975) A model for gastric cancer epidemiology. Lancet 2: 58–60.
- 9. Fontham ET, Ruiz B, Perez A, Hunter F, Correa P (1995) Determinants of Helicobacter pylori infection and chronic gastritis. Am J Gastroenterol 90: 1094–1101.
- 10. Fontham E, Zavala D, Correa P, Rodriguez E, Hunter F, et al. (1986) Diet and chronic atrophic gastritis: a case-control study. J Natl Cancer Inst 76: 621–627.
- 11. Farinati F, Cardin R, Libera GD, Rugge M, Herszènyi L, et al. (1995) Determinants for the development of chronic atrophic gastritis and intestinal metaplasia in the stomach. Eur J Cancer Prev 4: 181–186.
- 12. Kato I, Vivas J, Plummer M, Lopez G, Peraza S, et al. (2004) Environmental factors in Helicobacter pylori-related gastric precancerous lesions in Venezuela. Cancer Epidemiol Biomarkers Prev 13: 468–476.
- 13. Weck MN, Brenner H (2006) Prevalence of chronic atrophic gastritis in different parts of the world. Cancer Epidemiol Biomarkers Prev 15: 1083–1094.
- 14. Peleteiro B, Lunet N, Figueiredo C, Carneiro F, David L, et al. (2007) Smoking, Helicobacter pylori virulence, and type of intestinal metaplasia in Portuguese males. Cancer Epidemiol Biomarkers Prev 16: 322–326.
- 15. Abnet CC, Kamangar F, Dawsey SM, Stolzenberg-Solomon RZ, Albanes D, et al. (2005) Tooth loss is associated with increased risk of gastric non-cardia adenocarcinoma in a cohort of Finnish smokers. Scand J Gastroenterol 40: 681–687.
- 16. Abnet CC, Qiao YL, Mark SD, Dong ZW, Taylor PR, et al. (2001) Prospective study of tooth loss and incident esophageal and gastric cancers in China. Cancer Causes Control 12: 847–854.
- 17. Watabe K, Nishi M, Miyake H, Hirata K (1998) Lifestyle and gastric cancer: a case-control study. Oncol Rep 5: 1191–1194.
- 18. Demirer T, Icli F, Uzunalimoglu O, Kucuk O (1990) Diet and stomach cancer incidence. A case-control study in Turkey. Cancer 65: 2344–2348.
- 19. Michaud DS, Liu Y, Meyer M, Giovannucci E, Joshipura K (2008) Periodontal disease, tooth loss, and cancer risk in male health professionals: a prospective cohort study. Lancet Oncol 9: 550–558.
- 20. Salazar CR, Francois F, Li Y, Corby P, Hays R, et al. (2012) Association between oral health and gastric precancerous lesions. Carcinogenesis 33: 399–403.
- 21. Socransky SS, Haffajee AD, Cugini MA, Smith C, Kent RL Jr (1998) Microbial complexes in subgingival plaque. J Clin Periodontol 25: 134–144.
- 22. Kigure T, Saito A, Seida K, Yamada S, Ishihara K, et al. (1995) Distribution of Porphyromonas gingivalis and Treponema denticola in human subgingival plaque at different periodontal pocket depths examined by immunohistochemical methods. J Periodontal Res 30: 332–341.
- 23. Buchmann R, Muller RF, Heinecke A, Lange DE (2000) Actinobacillus actinomycetemcomitans in destructive periodontal disease. Three-year follow-up results. J Periodontol 71: 444–453.
- 24. Slots J, Ting M (1999) Actinobacillus actinomycetemcomitans and Porphyromonas gingivalis in human periodontal disease: occurrence and treatment. Periodontol 2000 20: 82–121.
- 25. Hamada S, Slade HD (1980) Biology, immunology, and cariogenicity of Streptococcus mutans. Microbiol Rev 44: 331–384.
- 26. Whiley RA, Beighton D (1998) Current classification of the oral streptococci. Oral Microbiol Immunol 13: 195–216.
- 27. Perez-Perez GI, Dworkin BM, Chodos JE, Blaser MJ (1988) Campylobacter pylori antibodies in humans. Ann Intern Med 109: 11–17.
- 28. Blaser MJ, Perez-Perez GI, Kleanthous H, Cover TL, Peek RM, et al. (1995) Infection with Helicobacter pylori strains possessing cagA is associated with an increased risk of developing adenocarcinoma of the stomach. Cancer Res 55: 2111–2115.
- 29. Francois F, Roper J, Goodman AJ, Pei Z, Ghumman M, et al. (2008) The association of gastric leptin with oesophageal inflammation and metaplasia. Gut 57: 16–24.
- 30. Dixon MF, Genta RM, Yardley JH, Correa P (1996) Classification and grading of gastritis. The updated Sydney System. International Workshop on the Histopathology of Gastritis, Houston 1994. Am J Surg Pathol 20: 1161–1181.
- 31. National Institute of Dental Research (1987) Epidemiology and Oral Disease Prevention Program. Oral health of United States adults: the National Survey of Oral Health in U.S. Employed Adults and Seniors, 1985–1986: national findings. Bethesda, Md.: U.S. Dept. of Health and Human Services, Public Service, National Institutes of Health.
- 32. Loe H (1967) The Gingival Index, the Plaque Index and the Retention Index Systems. J Periodontol 38: Suppl: 610–616.
- 33. Li Y, Saxena D, Barnes VM, Trivedi HM, Ge Y, et al. (2006) Polymerase chain reaction-based denaturing gradient gel electrophoresis in the evaluation of oral microbiota. Oral Microbiol Immunol 21: 333–339.
- 34. Psoter WJ, Ge Y, Russell SL, Chen Z, Katz RV, et al. (2011) PCR detection of Streptococcus mutans and Aggregatibacter actinomycetemcomitans in dental plaque samples from Haitian adolescents. Clin Oral Investig 15: 461–469.
- 35. Yano A, Kaneko N, Ida H, Yamaguchi T, Hanada N (2002) Real-time PCR for quantification of Streptococcus mutans. FEMS Microbiol Lett 217: 23–30.
- 36. Breslow N, Day NE (1980) Statistical Methods in Cancer Research, Vol 1—The Analysis of Case-Control Studies. In. WHO Lyon: IARC Scientific Publication No. 32: 227–38.
- 37. Consensus report (1996) Periodontal diseases: pathogenesis and microbial factors. Ann Periodontol 1: 926–932.
- 38. Loesche WJ (1986) Role of Streptococcus mutans in human dental decay. Microbiol Rev 50: 353–380.
- 39. Kuper H, Adami HO, Trichopoulos D (2000) Infections as a major preventable cause of human cancer. J Intern Med 248: 171–183.
- 40. Pritchard DM, Crabtree JE (2006) Helicobacter pylori and gastric cancer. Curr Opin Gastroenterol 22: 620–625.
- 41. Talley NJ, Zinsmeister AR, Weaver A, DiMagno EP, Carpenter HA, et al. (1991) Gastric adenocarcinoma and Helicobacter pylori infection. J Natl Cancer Inst 83: 1734–1739.
- 42. Parsonnet J, Friedman GD, Vandersteen DP, Chang Y, Vogelman JH, et al. (1991) Helicobacter pylori infection and the risk of gastric carcinoma. N Engl J Med 325: 1127–1131.
- 43. Fives-Taylor PM, Meyer DH, Mintz KP, Brissette C (1999) Virulence factors of Actinobacillus actinomycetemcomitans. Periodontol 2000 20: 136–167.
- 44. Zadeh HH, Nichols FC, Miyasaki KT (1999) The role of the cell-mediated immune response to Actinobacillus actinomycetemcomitans and Porphyromonas gingivalis in periodontitis. Periodontol 2000 20: 239–288.
- 45. Dye BA, Choudhary K, Shea S, Papapanou PN (2005) Serum antibodies to periodontal pathogens and markers of systemic inflammation. J Clin Periodontol 32: 1189–1199.
- 46. Nibali L, Ready DR, Parkar M, Brett PM, Wilson M, et al. (2007) Gene polymorphisms and the prevalence of key periodontal pathogens. J Dent Res 86: 416–420.
- 47. Pihlstrom BL, Michalowicz BS, Johnson NW (2005) Periodontal diseases. Lancet 366: 1809–1820.
- 48. Casarin RC, Ribeiro Edel P, Mariano FS, Nociti FH Jr, Casati MZ, et al. (2010) Levels of Aggregatibacter actinomycetemcomitans, Porphyromonas gingivalis, inflammatory cytokines and species-specific immunoglobulin G in generalized aggressive and chronic periodontitis. J Periodontal Res 45: 635–642.
- 49. da Silva-Boghossian CM, do Souto RM, Luiz RR, Colombo AP (2011) Association of red complex, A. actinomycetemcomitans and non-oral bacteria with periodontal diseases. Arch Oral Biol 56: 899–906.
- 50. Blaser MJ (2010) Harnessing the power of the human microbiome. Proc Natl Acad Sci U S A 107: 6125–6126.
- 51. Foxman B, Goldberg D, Murdock C, Xi C, Gilsdorf JR (2008) Conceptualizing human microbiota: from multicelled organ to ecological community. Interdiscip Perspect Infect Dis 2008: 1–5.
- 52. Benjamini Y (1995) Controlling the false discovery rate: a practical and powerful approach to multiple testing. Journal of the Royal Statistical Society Series B (Methodological) 57: 289–300.