Jon H. Higham and Frank Scannapieco
Epidemiological studies have established an association between malignancy and exposure to chronic infection and aberrant immune response to microbiota. Several microbial species have been associated with a variety of cancers. These include viruses, bacteria, parasites, and fungi1. Evidence is building to suggest that the association between chronic infection and malignancy is at least in part influenced by chronic inflammation induced by infections1–3.
The association between chronic inflammation and malignancy was first postulated in 1863 by Rudolf Virchow, who suggested that cancer originates at sites of chronic inflammation. An improved understanding of tumour microenvironments, particularly with regard to inflammation, as well as epidemiological evidence linking chronic inflammation with malignancy, has resulted in a renewed interest in inflammation as a contributor to malignancy1,2. Indeed, it has now been estimated that infection and inflammation are linked to 25% of all cancer cases4. Well-recognised examples of infection and associated inflammation preceding cancer development include inflammatory bowel disease and colon cancer, hepatitis B and C viruses and hepatocellular carcinoma, as well as Helicobacter pylori and gastric cancer and mucosa-associated lymphoid tissue (MALT) lymphoma3.
Periodontitis is a chronic inflammatory disease with a polymicrobial aetiology. Periodontitis is initiated by dental biofilm bacteria that drive the host immune response. This process is modified by genetic and environmental factors. It has been argued that a hyper-responsive inflammatory trait associated with an impaired host immune response accounts for the tissue destruction seen in periodontitis and the increased systemic inflammation seen in individuals with periodontitis5. For periodontitis to occur, bacteria as well as a susceptible host are required. The immune response within the gingival and periodontal tissues in response to the chronic presence of plaque bacteria drives the destruction of structural components of the periodontium. This inflammation leads to the clinical signs of periodontitis, and ultimately to tooth loss, if left untreated. The host response is determined by genetic, environmental and acquired factors and is essentially protective in nature. However, a hyper-responsive inflammatory trait or an impaired host immune response can result in enhanced tissue destruction. The potential association between periodontitis and malignancy is the focus of this chapter.
8.2.1 Total cancer risk
A positive association between periodontitis and overall cancer risk has been reported in five large cohort studies6–10 and one updated analysis11. All five studies reported a small but statistically significant increase in total cancer risk in individuals with periodontitis, and this association was confirmed by Michaud et al11 in their updated analysis of the Health Professionals Follow-up Study. All of these studies have the strength of large population-based designs. Interestingly, Hwang et al8 report that the overall risk of developing cancer was significantly lower in individuals with periodontitis who underwent periodontal treatment compared to those who remained untreated, and these data suggest a causal relationship between periodontitis and cancer. This study is of particular interest as it is the only study that investigates the effect of periodontal treatment on the risk of developing cancer, albeit retrospectively, and therefore provides strong evidence for an association between periodontitis and cancer. However, the authors did not adjust for confounding factors such as tobacco use and alcohol consumption, and this was also the case in the studies by Chung et al7 and Hwang et al8.
Conversely, four cohort studies did not support the association between total cancer risk and periodontitis12–15. Mai et al14 found that periodontitis was associated with total cancer risk in ever- but not never-smoking postmenopausal women, which suggests that the reported relationship might simply be due to unadjusted confounding from smoking. Tu et al13 investigated the association of periodontitis and later life cardiovascular disease and cancer mortality in the Glasgow Alumni Cohort and did not find an association between periodontitis and all-cause mortality or cancer mortality. Similar findings were reported by Cabrera et al12, Aida et al15 and Huang et al15a.. Contrarily, a Finnish health register-based cohort study with a mean follow-up period of 10 years and including 68,273 adults reported a mildly increased risk (risk ratio [RR] 1.33, 95% confidence interval [CI] 1.10 to 1.58) of overall cancer mortality in periodontitis patients. However, this study did not adjust for smoking and alcohol consumption as confounders, and periodontal status was assessed by means of recorded dental treatment codes only, presenting a risk of over- or underestimating periodontitis severity16.
Other studies provide conflicting evidence. Hiraki et al17, in their large case-control study, identified an association between periodontitis and head and neck, oesophageal and lung cancers but found no association between gastric, colon, liver, pancreas, breast, uterine, prostate, bladder or thyroid cancers. Similarly, Wen et al18, in another large cohort study, report a slightly elevated total cancer risk in individuals with periodontitis. Sub-analysis for cancer type identified increased hazard ratio (HR) for oral cancer and not cancers of the pharynx, oesophagus, stomach, colon, pancreas, larynx, lung, breast, uterus, prostate, bladder, kidney, thyroid or lymphoreticular system18. Studies from the Buffalo OsteoPerio study cohort did not identify an association between the presence of individual periodontal pathogens or red-complex pathogens with total or site-specific cancer risk, but did report borderline associations between the presence of orange-complex pathogens and increased total and lung cancer risk19.
Finally, a well-conducted recent meta-analysis (2018) of ten case-control and cohort studies (out of 450 initially retrieved papers) reported a substantial lack of studies with standardised and comparable methods, making it difficult to assess the association between periodontitis and cancer. From those papers included, the authors calculated an HR of 1.14 (95% CI 1.04 to 1.24) for total cancer risk in periodontitis patients, indicating a low but statistically significant association. However, this result is hampered by a limited statistical power due to the low number of included studies20. A well-designed prospective cohort study published in the same year by Michaud et al21 and not included in the above meta-analysis, included 7466 participants who were edentulous or dentate. All dentate subjects underwent periodontal examination, case definitions for periodontitis were used and a variety of confounders was taken into account. After a median follow-up period of 15 years, the authors determined HRs for developing cancer in the different periodontal severity and the edentulous groups. An increased risk of total cancer (adjusted HR 1.24, 95% CI 1.07 to 1.44) was observed for severe periodontitis (> 30% of sites with attachment loss > 3 mm) compared with no/mild periodontitis (< 10% of sites with attachment loss > 3 mm), along with the finding of cancer site–specific and racial differences21. Similar risks were reported in a recent study conducted by Lu et al22. This study was aimed at statistically accounting for a large number of possible confounding factors related to lifestyle, socio-economic status, oral and general health. The authors used clinical and demographic data from 7466 individuals recorded in the Atherosclerosis Risk in Communities (ARIC) study. They concluded that complex sources of confounding are contributing to, but not fully accountable for the positive associations between periodontal disease and total cancer risk (HR 1.23, 95% CI 1.05 to 1.44)22.
188.8.131.52 Studies investigating the association between periodontitis and specific cancers
In this section only those studies that specifically investigate individual cancers or groups of cancers are discussed. Some of the studies investigating the association of periodontitis and total cancer risk (Section 8.2.1 ‘Total cancer risk’) are stratified for specific types of cancer and therefore provide evidence for the association between periodontitis and individual cancers. These associations are summarised in Table 8-1 and the overall outcomes of the studies are summarised in Table 8-2. Some of these studies identified associations between periodontitis and genitourinary (bladder and kidney) cancer. Skin cancers were also investigated in several of these studies, but no association was found8,10,11,14. There are currently no studies specifically investigating these cancers and these will therefore be discussed here.
Table 8-1 Total cancer risk studies stratified by cancer type
|Type of malignancy||Total cancer risk studies|
|Soder et al6||Chung et al7||Hwang et al8||Soder et al9||Michaud et al10||Michaud et al11||Cabrera et al12||Tu et al13||Mai et al14||Aida et al15||Hiraki et al17||Wen et al18||Mai et al19|
|Total cancer risk||✔||✔||✔||✔||✔||✔||?||?||?||?||✔/?||✔/?||?|
|Oral and oropharyngeal cancers||Data not stratified for individual cancers||✔||?||Data not stratified for individual cancers||?||✔||Data not stratified for individual cancers||Data not stratified for individual cancers||–||Data not stratified for individual cancers||✔||✔||–|
|Upper GI cancers: oesophageal and stomach cancers||✔
(includes all upper and lower GI cancers)
|Upper GI cancers: pancreatic cancers||✔
(includes all upper and lower GI cancers)
|Upper GI cancers: Liver and pancreatobiliary tract cancers||✔
(includes all upper and lower GI cancers)
|Lower GI cancers||✔
(includes all upper and lower GI cancers)
(Attenuate when adjusted for smoking status)
(includes in GU cancers)
|✔||? (inverse)||?||?||? (inverse)||?||–|
|Renal and GU cancers||✔||?||✔||✔||–||?||?||–|
|Skin cancers||✔||?||? (inverse)||?||?||–||–||–|
✔ = positive association; ? = no or inverse (specified) association; − = not assessed; ✔/? = equivocal findings; GI = gastrointestinal; GU = genitourinary.
Table 8-2 In vivo human studies investigating the association of periodontitis and malignancy
|Type of malignancy||Overall outcome (summary)||Studies||Study design|
|Total cancer risk||There is conflicting evidence to support an association between periodontitis and total cancer risk. The majority of the available evidence is from cohort studies, five of which support the association, four of which do not. Additionally, one case-control study and one cohort study made equivocal findings. Furthermore, the results of those studies that identified a positive association need to be interpreted with caution as there is potential for unadjusted confounding, particularly from smoking.
Overall, there is equivocal evidence to support an association between periodontitis and total cancer risk.
|Corbella et al20||Meta- analysis|
|Cabrera et al12, Tu et al13, Lopez-Galindo et al106, Soder et al6,9, Michaud et al10,11,21, Aida et al15, Wen et al18, Hwang et al8, Chung et al7, Mai et al14,19, Dizdar et al109, Heikkilä et al16||Cohort|
|Hiraki et al17||Case-control|
|Oral and oropharyngeal cancers||The majority of available evidence is from case-control studies. There are fewer studies with the highest level of evidence (meta-analyses and cohort studies). Most studies support an association between periodontitis and OOSCC. However, this association needs to be interpreted with caution. Firstly, even in those studies that adjusted for confounders, confounding particularly from smoking may still have occurred. Secondly, there is marked heterogeneity in the assessment of periodontitis between studies. However, only three case-control studies did not specifically identify an association between periodontitis and OOSCC.
Overall, the balance of available evidence supports an association between periodontitis and OOSCC.
|Zeng et al51, Wang et al52, Yao et al53, Ye et al54, Xu et al55, Javed and Warnakulasuriya30 , Gopinath et al30a||Meta-analysis/systematic review|
|Wen et al18, Chung et al7 Michaud et al11||Cohort|
|Zheng et al31, Marshall et al32, Bundgaard et al33, Talamini et al34, Garrote et al35, Rosenquist et al36,37, Tezal et al38,40,47,48, Guha et al39, Hiraki et al17, Divaris et al41, Moergel et al42, Ansai et al43, Eliot et al44, Moraes et al45, Laprise et al46, Shin et al49 , Chen et al49a||Case-control|
|Tezal et al50||Cross-sectional|
|Upper gastrointestinal tract cancers: oesophageal and stomach cancer||There are some inconsistencies between and within studies due to the effect of confounding factors, particularly smoking, that may not have been adequately adjusted for.
Overall, the balance of available evidence supports an association between periodontitis and oesophageal and stomach cancer.
|Chen et al65,66, Yin et al67, Zhang et al68||Meta- analysis|
|Abnet et al58–60, Ansai et al43, Wen et al18, Lee et al64, Hwang et al8, Chung et al7, Michaud et al11||Cohort|
|Hiraki et al17, Dar et al63, Abnet et al62, Shakeri et al61||Case-control|
|Upper gastrointestinal tract cancers: pancreatic cancer||The available studies are relatively few in number; however, the findings are generally consistent between studies.
On balance, the currently available evidence supports an association between periodontitis and pancreatic cancer.
|Maisonneuve et al76, Zhang et al68||Meta- analysis|
|Stolzenberg-Solomon et al72, Michaud et al10,11,73, Wen et al18, Hwang et al8, Chang et al75, Huang et al74||Cohort|
|Hiraki et al17, Chrysanthakopoulos and Chrysanthakopoulos107||Case-control|
|Upper gastrointestinal tract cancers: liver and pancreatobiliary tract cancer||There are currently very few studies investigating the association between periodontitis and liver cancer, but the majority of these support an association.
In summary, too few studies have been conducted; therefore, there is currently no evidence to support an association between periodontitis and liver and pancreatobiliary tract cancers.
|Wen et al18, Hwang et al8, Michaud et al11, Yang et al80||Cohort|
|Hiraki et al17||Case-control|
|Maruyama et al81||Cross-sectional|
|Lower gastrointestinal tract cancers||The available studies show divergent results. Additionally, a recent meta-analysis did not support an association. Therefore, there is currently no evidence to support an association between periodontitis and lower gastrointestinal tract cancers.||Ren et al87, Zhang et al68, Xuan et al68a||Meta- analysis|
|Wen et al18, Hwang et al8, Chung et al7, Michaud et al11, Mai et al14, Momen-Heravi et al84, Hu et al85||Cohort|
|Hiraki et al17||Case-control|
|Yen et al86, Kim et al108||Cross- sectional|
|Lung cancer||The majority of available evidence is from large cohort studies. Studies support an association between periodontitis and lung cancers. However, this association needs to be interpreted with caution. Firstly, even in those studies that adjusted for confounders, confounding particularly from smoking may still have occurred, indeed one study identified a positive association, but this was attenuated when adjusted for smoking. As in other cancers there is also marked heterogeneity in the assessment of periodontitis between studies. Of the available studies only three cohort studies did not identify an association between periodontitis and lung cancer.
Overall, the balance of available evidence supports an association between periodontitis and lung cancer.
|Zeng et al96, Wang et al96a||Meta- analysis|
|Soder et al6, Chung et al7, Hwang et al8, Soder et al9, Michaud et al10,11, Mai et al14,19,97, Wen et al18, Hujoel et al94||Cohort|
|Chrysanthakopoulos95, Hiraki et al17||Case-control|
|Prostate cancer||In summary, too few studies have been conducted and these show divergent results. Therefore, there is no evidence to support an association between periodontitis and prostate cancer.||Wen et al18, Hwang et al8, Michaud et al11, Lee et al98, Wei et al98a||Cohort|
|Hiraki et al17||Case-control|
|Haematological malignancies||There is conflicting evidence for an association between periodontitis and haematological malignancy in general. Additionally, the evidence for an association between periodontitis and non-Hodgkin lymphoma specifically is drawn from a single cohort. In summary, too few studies have been conducted and these show divergent results. Therefore, there is no evidence to support an association between periodontitis and haematological malignancies.||Michaud et al10, Wen et al18, Hwang et al8, Mai et al14, Bertrand et al100||Cohort|
|Hiraki et al17||Case-control|
|Breast cancer||In summary, too few studies have been conducted and these show divergent results. Therefore, there is no evidence to support an association between periodontitis and breast cancer.||Shao et al103, Shi et al104||Meta- analysis|
|Soder et al101, Wen et al18, Hwang et al8, Chung et al7, Freudenheim et al102||Cohort|
|Hiraki et al17, Sfreddo et al110||Case-control|
|Gynaecological cancers||In summary, very few studies have investigated ovarian and uterine cancer in patients with periodontitis and these show divergent results. Based on the available evidence there is no association between periodontitis and gynaecological cancers.||Hwang et al8, Babic et al105||Cohort|
|Hiraki et al17||Case-control|
|Genitourinary cancers||The available evidence is drawn from studies investigating total cancer risk. Additionally, two of three studies that describe a positive association investigated the same cohort at different stages in follow-up.
In summary, too few studies have been conducted and these show divergent results. Therefore, there is no evidence to support an association between periodontitis and genitourinary cancers.
|Xie et al23||Meta- analysis|
|Michaud et al10, Wen et al18, Hwang et al8, Michaud et al11||Cohort|
Two of the studies discussed in Section 8.2.1 identified a strong association between periodontitis and genitourinary cancers; however, these data were derived from the same cohort at different stages during follow-up10,11. Similarly, Chung et al7 describe a positive association, but as has already been mentioned this study did not adjust for confounding factors7. Conversely, Wen et al18, in another large cohort study, report a slightly elevated total cancer risk in individuals with periodontitis; however, after multivariate analysis this association was maintained only in oral cancers and not in other cancers (including kidney cancers). Hwang et al8 found that unlike other cancers investigated in their study, the association between periodontitis and kidney cancers was not significant8. There are currently no studies specifically investigating the association between periodontitis and genitourinary cancers; however, Xie et al23 conducted a recent meta-analysis on five studies, which assessed bladder cancer, amongst other cancers. No significant association between periodontitis and bladder cancer risk was found by the authors23. It should be mentioned that in both periodontitis and malignancy, older age and smoking are major risk factors; however, the majority of studies adjusted for these and other confounding factors. Nevertheless, there remains the possibility that unadjusted confounding, particularly from smoking, may still have occurred. It should also be kept in mind that tooth loss, which has often been used as a surrogate marker of oral health status in association studies investigating periodontitis and cancer, might in itself have implications on cancer initiation and progression17.
184.108.40.206 Oral and oropharyngeal cancers
Oral and oropharyngeal cancers grouped together are the sixth most common cancer in the world, with a combined annual estimated incidence of approximately 405,30024. The majority of malignancies occurring in the upper aerodigestive tract are squamous cell carcinoma (approx. 90%)24. Several studies have reported increasing incidence of oral and oropharyngeal squamous cell carcinoma (OOSCC) (Fig 8-1) since the early 1970s to the present day, which has been in part attributed to the increasing incidence of human papilloma virus (HPV) infection25–28. The major risk factors for the development of OOSCC are tobacco (smoking and smokeless tobacco use), excessive consumption of alcohol, betel quid usage, and HPV infection24,29. An emerging relationship between OOSCC and periodontitis has been reported and will be the focus of this section. However, it is important to acknowledge that periodontitis and OOSCC share risk factors, particularly tobacco use, which is increased in those with underprivileged socio-economic status30.
Several studies have investigated an association between OOSCC and periodontitis. A positive association between periodontitis and OOSCC was reported in 17 case-control studies31–49, one cross-sectional study50, five meta-analyses51–55 and one systematic review30. Primary studies (case-control and cross-sectional) reported a 2- to 5-fold increase in the risk of OOSCC in patients with periodontitis31–33,36–38,40,42–46,50; existing meta-analyses suggest a slightly lower but increased risk of 1- to 3.5-fold51–54. Three case control studies did not specifically identify an association between periodontitis and OOSCC34,39,41; however, of these studies Guha et al39 did identify an increased risk for other head and neck cancers in patients with periodontitis39, and Talamini et al34 identified a 4.5-fold increased frequency of poor general oral condition among patients with oral cancer. For the majority of studies, an association between periodontitis and OOSCC persisted after adjustment for confounders such as tobacco use and alcohol consumption. In most studies, the assessment of periodontitis was made by determining the number of missing teeth, but it should be acknowledged that this is not an accurate means of assessing periodontitis as teeth can be lost for other reasons, including caries and trauma. Some studies used additional or alternative methods including clinical evaluation of oral hygiene, dental and periodontal status31,36,37,39,43,45,46,50, self-reported by means of questionnaires31,33,35–37,39,41–44, and radiographic assessment of bone loss31,36–38,40,42,47–49.
There are further studies that also suggest an association between periodontitis and OOSCC. Meisel et al56 identified a positive association between oral leukoplakia, a premalignant lesion of oral squamous cell carcinoma (OSCC), and periodontitis assessed by clinical attachment loss and bleeding on probing. This association was maintained after adjustment for confounding factors56. Katz et al57 used immunohistochemical staining to investigate the presence of Porphyromonas gingivalis, a common periodontal pathogen, in normal gingivae and gingival squamous cell carcinoma, with markedly higher levels being found in gingival carcinoma; however, a causal relationship between P. gingivalis and gingival squamous cell carcinoma cannot be inferred from this study. Additionally, an association between periodontal inflammation and HPV in OOSCC has been described40,47, and the authors suggest that periodontal pockets may act as a reservoir for HPV, which might explain this association.
220.127.116.11 Upper gastrointestinal tract cancers
Upper digestive tract cancers include oesophageal cancer, stomach cancer, pancreatic cancer, liver cancer and gallbladder cancer (Fig 8-2). These have been grouped to aid discussion as follows: oesophageal and stomach cancers, pancreatic cancer, and liver and pancreatobiliary tract cancers.
Oesophageal and stomach cancers
Some studies utilising a population-based design showed an association between periodontitis and a range of oesophageal and stomach cancers, including oesophageal squamous cell carcinoma, gastric cardia adenocarcinoma, and gastric non-cardia adenocarcinoma43,58–64. Furthermore, Lee et al64, in a large population-based cohort study, investigated confirmed cases of oesophageal cancer and stratified subjects by presence of periodontitis and whether they had undergone dental prophylaxis or intensive periodontal treatments including root surface debridement and surgical interventions. Interestingly, they found a reduced risk of oesophageal cancer in those patients with periodontitis receiving dental prophylaxis intensive periodontal treatments compared to all patients with periodontitis and patients without periodontitis among males but not females64. A positive association is also supported by two recent meta-analyses65,66. A third recent meta-analysis found that tooth loss is a potential marker of gastric cancer; however, the authors state that they could not conclude that tooth loss is a risk factor for gastric cancer67. This was due to significant heterogeneity among the included studies regarding the definition and assessment of periodontitis and mixed results from case-control studies and cohort studies. A fourth and very recent meta-analysis by Zhang et al68 investigated the association between periodontitis and gastrointestinal cancer mortality and stratified their analysis by cancer type. The results suggest that periodontitis does not correlate with gastric cancer or oesophageal cancer mortality68.
Other evidence suggesting an association between periodontitis and upper aerodigestive cancers includes data from the NHANES III cohort, from which Ahn et al69 describe a positive association between periodontitis and serum antibody levels against P. gingivalis with orodigestive cancer mortality, and between periodontitis and orodigestive cancer69. Salazar et al70 investigated the association between selected oral pathogens and gastric precancerous lesions and found that among individuals with periodontal disease, high levels of colonisation of periodontal pathogens (Aggregatibacter actinomycetemcomitans, P. gingivalis, and Treponema denticola) were associated with an increased risk of gastric precancerous lesions. However, this study was limited by its small sample size and cross-sectional design, thus these results need to be interpreted with caution70. The same group looked at the association between oral health and gastric precancerous lesions and found that specific oral health conditions and behaviours, including gingival bleeding and no tooth flossing, were associated with gastric precancerous lesions71.
Four large population-based cohort studies and two meta-analyses have specifically investigated the association between periodontitis and pancreatic cancer. Of the cohort studies, Stolzenberg-Solomon et al72 found that tooth loss was positively associated with pancreatic cancer, which, in turn, was not significantly associated with the independent risk factor H. pylori seropositivity. Michaud et al73 reported that a history of periodontitis was associated with increased risk of pancreatic cancer but that the baseline number of teeth and cumulative tooth loss during follow-up were not strongly associated with pancreatic cancer. Huang et al74 found that subjects with fewer teeth at baseline tended to have an increased risk for pancreatic cancer but this was not statistically significant. However, they identified a positive association between the presence of dental plaque and pancreatic cancer and between candida-related and denture-related oral mucosal lesions and pancreatic cancer74. This suggests that generally poor oral hygiene and/or oral state may be an additional risk factor for cancer. Chang et al75 describe an association between periodontitis and pancreatic cancer amongst those aged 65 years or older, but not in younger patients. However, although they adjusted for confounders, they used chronic obstructive pulmonary disease as a proxy for cigarette smoking, and alcoholic-related conditions, such as liver disease, as an indicator for alcohol intake, which are unlikely to have fully accounted for these confounders75. Additionally, a positive association between periodontitis and pancreatic cancer is described by Maisonneuve et al76 and by Zhang et al68 in their recent meta-analyses.
The NHANES III cohort identified a weak association between periodontitis and P. gingivalis serum antibody levels and pancreatic cancer mortality after adjustment for confounding factors66. Michaud et al77, in their cohort study, investigated plasma antibodies to the oral bacteria P. gingivalis and A. actinomycetemcomitans and identified a positive association between the presence of these antibodies and risk of pancreatic cancer. Similarly, Fan et al78, in their population-based nested case-control study, describe a positive association between pancreatic cancer and the presence of periodontal bacteria including P. gingivalis and A. actinomycetemcomitans, but interestingly found that fusobacteria were associated with decreased pancreatic cancer risk. Conversely Mitsuhashi et al79, in their database of 283 patients with pancreatic cancer, detected Fusobacterium species in 8.8% of tumours, and although Fusobacterium status was not associated with any clinical or molecular features of these tumours it was associated with significantly higher cancer-specific mortality rates.
Liver and pancreatobiliary tract cancer
Only one study has looked at the association between liver cancer and periodontitis. Yang et al80 found that increasing tooth loss was associated with higher risk of primary liver cancer, and the association persisted after adjustment for confounders including smoking, hepatitis B and C virus infection, and H. pylori seropositivity80. There are currently no studies that have investigated whether there is an association between periodontitis and gall bladder cancer, but Maruyama et al81 found that low body mass index (BMI), raised C-reactive protein and severe periodontitis were prognostic factors for the survival rate in pancreatobiliary tract cancer. They postulated that periodontitis indirectly affected the prognosis of pancreatobiliary tract cancer by increasing systemic infammation81. In another study that might suggest an association between periodontitis and liver cancer, Tamaki et al82 found that periodontitis was associated with more advanced stages of hepatocellular carcinoma and higher circulating reactive oxygen species than in patients with hepatocellular carcinoma without periodontitis. Nagao et al83 describe an association between periodontitis and liver fibrosis in patients with hepatitis B and C virus infection and suggest that periodontitis is associated with the progression of viral liver disease. This implies that periodontitis is an indirect risk factor for hepatocellular carcinoma, which is a common outcome in patients with advanced hepatitis B and C virus infection.
18.104.22.168 Lower gastrointestinal tract cancers
A number of studies have investigated the association between periodontitis and colorectal cancer. Momen-Heravi et al84, in their cohort study using data from the Nurses’ Health Study, found that women with fewer teeth and moderate or severe periodontitis have a modest increased risk of developing colorectal cancer; however these results were not adjusted for smoking status, body mass index or alcohol consumption. Hu et al85 found that periodontitis severity was associated with increased risk of colorectal cancer in their recent cohort study utilising a large population-based design. Yen et al86, in their cross-sectional study investigating Community Periodontal Index (CPI) and faecal haemoglobin concentration, an indicator of colorectal neoplasm, found an association between increased periodontal probing depths and faecal haemoglobin concentration in individuals with colorectal neoplasm; however, the mechanism that might account for this association is not understood86. Additionally, data from the NHANES III cohort support an association between periodontitis and colorectal cancer. On the other hand, three recent meta-analyses of cohort studies did not support an association between periodontitis and colorectal cancer87,68,68a, and a large-scale retrospective cohort study of patients from the Taiwanese National Health Insurance Research Database confirmed this finding88. The latter study, however, assessed the incidence of general gastrointestinal tract cancers (upper and lower as well as pancreatic) and found no association between periodontitis and any gastrointestinal cancer88. As a limitation, this study relied on recorded periodontal data measured by different health care professionals and compared mild (n = 25,485) with severe periodontitis (n = 25,485) cases, with no periodontally healthy controls88.
Ahn et al69 investigated presence of periodontitis, P. gingivalis serum antibody levels and orodigestive cancer mortality and found that subjects with periodontitis had excess mortality due to colorectal cancer after adjustment for age, sex, smoking, education, ethnicity, and BMI. They also identified an association between greater serum P. gingivalis antibody levels and overall increased orodigestive cancer mortality69. Interestingly, a number of studies have identified Fusobacterium nucleatum, a Gram-negative anaerobic periodontal pathogen, in colorectal carcinoma tissues89–93. Although this is an interesting finding, identification of F. nucleatum in colorectal tumours does not necessarily imply a causal relationship. The detection of periodontal pathogens in colorectal tumours is intriguing and might suggest a potential mechanism that could account for an association between periodontitis and colorectal cancer, should one exist.
22.214.171.124 Lung cancers
Hujoel et al94, in their cohort study, identified an association between periodontitis and lung cancer mortality after adjustment for known risk factors for lung cancer. Similarly, in a recent case-control study Chrysanthakopoulos95 concluded that pocket probing depth as an index for periodontitis severity was significantly associated with the risk of developing lung cancer. Zeng et al96 and Wang et al96a also found evidence to support this association in their meta-analysis. Conversely Mai et al97, in their cohort study, did not find that periodontitis was associated with lung cancer in non-smoking postmenopausal women, but that smoking in combination with periodontitis increased lung cancer risk.
126.96.36.199 Prostate cancer
One large population-based cohort study and a recent meta-analysis investigated the association between periodontitis and prostate cancer and identified a significant association after adjustment for confounders98,98a. Additionally, Joshi et al99 assessed periodontitis and prostate-specific antigen (PSA) levels in patients with chronic prostatitis. They identified increased PSA levels in patients with periodontitis and suggested that increased cytokine production in periodontitis might increase inflammation in prostatitis99. However, it remains unclear whether increased prostatitis might influence the development of prostate cancer; indeed, prostatitis as a risk factor for prostate cancer is in itself controversial.
188.8.131.52 Haematological malignancies
In a follow-up study to an earlier prospective cohort analysis within the Health Professionals Follow-Up Study, Bertrand et al100 assessed the association between periodontitis and non-Hodgkin lymphoma (NHL), including chronic lymphocytic leukaemia and small lymphocytic lymphoma. They identified a positive association between periodontitis and NHL after adjustment for confounders including smoking10. Interestingly, after adjusting for periodontitis, tooth loss due to other causes than periodontitis was inversely associated with NHL, i.e., there was less NHL occurrence in these patients100. This suggests that periodontitis-associated tooth loss has implications for NHL aetiology. This finding may also account for the lack of an association described in other studies, as tooth loss from other causes may have masked a positive association if it existed.
184.108.40.206 Breast cancer
Soder et al101 identified an association between periodontitis and missing molars in the mandible with breast cancer, which persisted after adjustment for confounders. Freudenheim et al102 found that periodontitis was associated with increased risk of postmenopausal breast cancer, particularly among former smokers who had stopped smoking within the past 20 years, and this association persisted after adjustment for confounders. Two recent (2018) meta-analyses, which included eight and eleven studies, found a mildly increased risk of breast cancer in periodontitis patients (RR 1.22, 95% CI 1.06 to 1.40; and RR 1.18, 95% CI 1.11 to 1.26)103,104. However, the included studies showed large variations in the definition of periodontitis (tooth mobility as a surrogate marker, self-reported periodontitis, medical records, radiographic measurements, clinical periodontal measurements). Furthermore, these studies did not consistently report, or adjust for, confounders such as smoking or history of periodontal therapy. Therefore, the results of these meta-analyses are likely to be influenced by this marked heterogeneity and need to be interpreted with care103,104.
220.127.116.11 Gynaecological cancers
Only one study investigated the association between periodontitis and gynaecological cancers. In the Nurses’ Health Study, Babic et al105 found that among women younger than 69 years, periodontal bone loss was associated with decreased ovarian cancer risk, while there was no association in women older than 69.
Epidemiological studies have identified a range of malignancies associated with periodontitis; however, in the majority of cancers the available evidence does not allow a clear conclusion as to whether a true association exists and therefore in these cancers we conclude that no association exists based on the current evidence (Fig 8-2). However, in the case of head and neck/upper aerodigestive tract, lung, stomach and oesophageal cancer, and pancreatic cancers the available evidence supports an association. Many of the studies used different methods to assess periodontitis, some of which are potentially inaccurate such as tooth loss as a surrogate for periodontitis. Inaccuracy in the assessment of periodontitis would therefore have a major impact on the validity of the data presented. In both periodontitis and malignancy older age and smoking are major risk factors; however, the majority of studies adjusted for these.
8.3 Cellular and molecular mechanisms
8.3.1 Local and systemic inflammation
Periodontitis is a chronic inflammatory disease. The association between chronic inflammation and malignancy has long been postulated. The exaggerated host response to the polymicrobial infection seen in periodontitis directly results in upregulation of proinflammatory cytokines111 that in turn induce an acute-phase inflammatory response including elevation in C-reactive protein112, as well as release of free radicals113. Additionally, bacterial products such as lipopolysaccharide (LPS; endotoxin) independently induce both local and systemic inflammation114. The periodontium consists of a diverse cellular community including epithelial cells, fibroblasts, endothelial cells and cells of the innate and adaptive immune systems (see Chapter 11 ‘Periodontitis and autoimmunity’). The innate and adaptive immune systems are thought to be tumour protective during the early stages of tumorigenesis, but tumour-associated immune cells have been shown to promote tumour growth in established lesions115. Pro-inflammatory changes in the periodontium are mediated by cell surface receptors, growth factors, cytokines, chemokines, enzymes and transcription factors, all of which interact through a complex interplay of autocrine, intracrine, juxtacrine and paracrine signalling. Analogous interactions occur in tumours, therefore, signalling pathways in periodontal inflammation may also impact on the tumour microenvironment116, as illustrated in Fig 8-3.
18.104.22.168 Cytokines, chemokines and growth factors
Pro-inflammatory cytokines, such as interleukins (IL-1, IL-6, IL-8, IL-13) and tumour necrosis factor-α (TNF-α), are upregulated in periodontitis. These cytokines stimulate cellular proliferation, remodel the extracellular matrix, and activate immune cells111,117. The pro-tumorigenic function of TNF-α and IL-6 is well established and current evidence suggests that this is primarily mediated through the transcription factor nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) resulting in stimulation of cell proliferation and survival118,119. NF-κB also controls the expression of many genes linked with inflammation and therefore engages in a feed-forward loop by driving inflammation that then results in further NF-κB activation115. It is a key mediator in any chronic inflammation, including periodontitis120, as well as in oncogenesis. Sharma et al121 found elevated levels of IL-6 in patients with leukoplakia and coexisting periodontitis and in periodontitis patients when compared to healthy controls, and salivary IL6 levels are increased in patients with oral epithelial dysplasia and OSCC122.
Chemokines are soluble factors that regulate the directional migration of leukocytes in inflammation123, are increased in periodontitis and have been implicated in its pathogenesis124. They also recruit immune cells including macrophages to tumours, thereby altering the tumour microenvironment. Inflammatory cells of a developing neoplasm can have diverse effects including regulation of tumour growth and angiogenesis2. Growth factors such as fibroblast growth factor (FGF), platelet-derived growth factor (PDGF), insulin-like growth factor (IGF), and transforming growth factor-β (TGF-β) function to promote cell migration and proliferation, which are important in wound healing but can also result in upregulation of inflammation in the presence of pro-inflammatory mediators. They are increased in periodontitis111,125, and dysregulation in expression of growth factors by cancer cells has been implicated in cancer progression126.
Cell surface receptors such as Toll-like receptors (TLRs) play an important role in periodontitis through recognition of periodontal pathogens and their products. TLR signalling results in the upregulation of pro-inflammatory cytokines and activation of the adaptive immune response127,128. Interestingly, TLR expression is increased in the presence of inflammation, representing a potential feedback mechanism129,130. Kauppila et al131 investigated TLR5 in oral tongue squamous carcinoma and found that its expression levels predicted patient survival and recurrence; the authors suggested that this may represent a link between bacteria and oral oncogenesis. Hsu et al128 found that LPS induced TLR4 signalling in colorectal cancer cells and suggested that this might increase metastasis of colorectal cancer.
Enzymes such as matrix metalloproteinases (MMPs) degrade the extracellular matrix but also act as regulators of extracellular tissue signalling networks132. MMPs are upregulated in periodontitis and this, at least partially, is mediated through cytokine expression. MMPs are implicated in the tissue damage seen in periodontitis133 and in tumour tissue remodelling, thereby enabling tumour invasion and metastasis. MMP8 and other MMPs are upregulated in periodontitis133 and are expressed in head and neck squamous carcinoma134 and other cancers135. Tissue inhibitors of metalloproteinases (TIMPs) inhibit the activity of MMPs, and TIMP concentrations are reduced in cancers135. They also act as mitogenic growth factors and inhibit apoptosis136. All human cancers show TIMP dysregulation resulting in alteration in pericellular proteolysis, thereby altering tumour architecture and cell signalling137.
22.214.171.124 Free radicals
Reactive oxygen species (ROS) and reactive nitrogen species (RNS) are types of free radicals and are a product of normal cellular metabolism acting as a protective mechanism against bacterial pathogens by direct effects and through recruitment of innate immune response cells138,139. In high concentration, free radicals induce tissue damage through DNA damage and lipid and protein peroxidation140. Abnormal host responses to the polymicrobial infection found in periodontitis result in excess ROS and RNS production in periodontal tissues141,142, and this is at least in part mediated by pro-inflammatory cytokines113. As free radical concentration increases in the periodontium, ROS and RNS diffuse into the bloodstream, resulting in increased oxidative and nitrosative stress and ultimately distant tissue/organ damage139,140. Generation of ROS and RNS leads to a signalling cascade that triggers the production of pro-inflammatory cytokines and chemokines, thereby upregulating systemic inflammation143. Oxidative stress from other causes of inflammation, including smoking, toxins, diabetes and obesity, also impact on the progression of periodontitis144. Free radicals contribute to carcinogenesis by the upregulation of systemic inflammation, but also by direct DNA damage, which can result in a number of events associated with carcinogenesis including transcriptional arrest, replication errors and genomic instability145.
Additionally, mitochondria control a variety of cellular process including redox status, energy production and apoptosis; dysregulation of these events has been implicated in carcinogenesis146 and it has been shown that free radicals can cause oxidative damage to mitochondria147. Furthermore, protein damage caused by free radicals can result in oxidation of regulatory proteins such as DNA repair enzymes and P53139. Finally, ROS at very low and transient levels have been reported to increase cell proliferation, cell survival and cellular migration and may therefore impact on tumour progression148. Periodontitis has been associated with a number of immune-inflammatory diseases, some of which are independently associated with increased risk of malignancy, most notably diabetes mellitus type 25,149. The question remains as to whether such diseases may have acted as confounders in the epidemiological studies associating periodontitis with cancer, as hardly any studies accounted for comorbid illness in their analysis.
8.3.2 The oral microbiome and specific oral pathogens
The microbiome includes all Bacteria, Archaea, Fungi, protists and viruses found in and on the human body. Microbiome dysbiosis is increasingly identified in human disease and has a significant impact on human health. Several studies have identified altered microbial diversity in patients with a variety of cancers, including gastric cancers150, oesophageal cancers151, pancreatic cancers77,78, lung cancer152, colorectal cancers153 and breast cancer154. Additionally, associations between specific bacterial genera or species and cancers have been identified and of these the most well recognised is that of gastric cancer associated with H. pylori, which has become the model of bacterial-associated cancer155. The oral microbiome is a complex community composed of commensal bacteria as well as representatives of other microbiological domains. The association of periodontitis with malignancy therefore cannot be discussed without considering polymicrobial interactions.
With regards to periodontal species, a number of studies have investigated the oral microbiome and its relation to cancer incidence. Faecal carriage of Fusobacterium species and Porphyromonas species has been described in colorectal cancer153. Oral carriage of Neisseria elongata156, Streptococcus mitis156, P. gingivalis77,78, A. actinomycetemcomitans78, and Alloprevotella species78 has been detected in pancreatic cancer and one study found that increased oral carriage of fusobacteria and closely related Leptotrichia was associated with a decreased risk of pancreatic cancer78. Salazar et al70 found that in those with periodontitis, high levels of oral colonisation by periodontal pathogens was associated with an increased risk of gastric precancerous lesions. Additionally, a recent comprehensive review of oral health and the oral microbiome in pancreatic cancer identified convincing epidemiological evidence to support an association between periodontitis and pancreatic cancer157. However, the author found that although those studies that investigated the association between oral microbiome profiles and/or circulating antibodies to oral bacteria with pancreatic cancer showed generally consistent results, most of these studies were not statistically significant. The author therefore suggests that the oral microbiome influences pancreatic cancer risk through immune function and inflammation rather than through specific effects of periodontal bacteria157. Although studies investigating the oral microbiome and head and neck cancer are few in number, associations between carriage of specific bacterial species and head and neck cancers have been reported, including P. gingivalis, F. nucleatum, Prevotella intermedia and melaninogenica, S. mitis and Capnocytophaga gingivalis 69,158–160. Another group identified P. gingivalis, Tannerella forsythia and P. intermedia DNA in the submandibular and submental lymph nodes of patients who had undergone resection of head and neck cancers161. These findings, however, must be interpreted with caution. It is currently not clear whether carcinoma causes changes in oral microbiome diversity, or if alterations in the microbiome promote carcinogenesis. Figure 8-4 depicts the potential influence of F. nucleatum and P. gingivalis on tumorigenesis and tumour progression, and this is described in further detail below.
The red complex periodontal pathogens include P. gingivalis, T. denticola and T. forsythia, and they are associated with periodontitis162. Of these, P. gingivalis is the most investigated periodontal pathogen and it has been postulated as a keystone pathogen in orchestrating inflammatory disease by remodelling the oral microbiome163. It possesses the potent virulence factors fimbriae, LPSs and proteases, including gingipains164–167. Its ability to attach to and invade epithelial and other cell types168,169 and to persist in these170, to promote cell survival by restraining apoptosis171, and to induce local and systemic inflammation172 might contribute to carcinogenesis. Katz et al57 identified higher levels of P. gingivalis in gingival squamous cell carcinoma; however, this does not imply causation and may simply represent opportunistic colonisation of malignant tissues by P. gingivalis. Ahn et al69 describe an association between orodigestive cancer mortality and periodontitis and P. gingivalis independently of periodontitis. The authors proposed that P. gingivalis can be considered a biomarker for microbial-associated orodigestive cancer risk69. Dissemination of oral bacteria to other parts of the body such as heart, liver, kidney and spleen is well documented173.
A recent in vitro study investigated the effect of repeated and prolonged exposure of oral cancer cells to P. gingivalis and found that this increased cancer aggressiveness by promoting morphological changes that ultimately resulted in the acquisition of cancer stem cell properties174. In an in vitro study by Groeger et al175, P. gingivalis induced expression of B7-H1 and B7-DC receptors in squamous cell carcinoma cells. Both of these receptors contribute to supressing the immune system in cancer175. P. gingivalis gingipains induce pro-MMP9 expression in an OSCC cell line, thereby promoting cellular invasion176. Lin et al177 describe the effect of P. gingivalis heat shock protein 60 (GroEL) on the neovasculogenesis of tumour cells, which resulted in accelerated tumour growth both in vitro and in vivo. Conversely, P. gingivalis LPS had no effect on cancer progression of an OSCC cell line178. In summary, P. gingivalis displays a range of behaviours and possesses a variety of virulence factors, which might be pro-tumorigenic and may account for the association between periodontitis and cancer.
Orange complex periodontal pathogens such as F. nucleatum are bridging bacteria that allow for colonisation of other members of the more pathogenic red complex species within subgingival biofilms179. Interestingly, F. nucleatum has been implicated in colorectal cancer (CRC) in several studies. One study identified an overabundance of Fusobacterium in CRC samples compared to normal colorectal tissue controls and also observed a positive association with lymph node metastasis of colorectal carcinoma91. Yu et al180 identified an abundance of Fusobacterium in CRC tissues of patients with recurrence post chemotherapy and promoted CRC resistance to chemotherapy through a variety of mechanisms. Similarly, faecal abundance of Fusobacterium including F. nucleatum has been strongly associated with CRC but not with advanced or unadvanced adenomas, a precursor lesion of CRC; the authors of this study suggest that Fusobacterium is a passenger species rather than a causative agent in CRC93. Conversely, McCoy et al181 identified significantly higher abundance of Fusobacterium species in adenoma subjects compared with controls, and this was correlated with increased gene expression of both pro-inflammatory (TNF-α) and anti-inflammatory (IL-10) cytokines. Rubinstein et al182, in their in vitro study, present further evidence that Fusobacterium may play a causal role in CRC. They showed that F. nucleatum binds to E-cadherin through its FadA adhesin, thereby invading CRC cells and inducing oncogenic and inflammatory responses that stimulate the growth of these cells182.
It appears that the association of Fusobacterium with malignancy is not limited to CRC. One study investigating Fusobacterium in pancreatic cancer reported a low detection rate of 8.8% and did not identify an association with clinical and molecular features of pancreatic cancer in F. nucleatum-positive patients79. However, they did find that tumour Fusobacterium species status was independently associated with a worse prognosis of pancreatic cancer79. Similarly, Yamamura et al183 identified significantly higher F. nucleatum DNA in oesophageal cancer tissues than in matched controls, and F. nucleatum positivity was significantly associated with tumour stage and cancer-specific survival. Binder Gallimidi et al184 found that chronic bacterial infection promoted OSCC in a murine model of periodontitis-associated oral tumorigenesis and that P. gingivalis and F. nucleatum stimulated tumorigenesis via direct interaction with human oral epithelial cells through Toll-like receptors.
H. pylori has been identified in subgingival plaque of periodontitis patients185. One study compared individuals with periodontitis and healthy controls and found significantly higher H. pylori carriage in those individuals with periodontitis186. Indeed, periodontitis may favour the colonisation of H. pylori, as the periodontal pocket offers a range of binding sites and nutrients, and in particular Fusobacterium was shown to co-aggregate with H. pylori 187. There is also evidence to suggest that the oral cavity acts as a reservoir for H. pylori in transmission and re-infection188. Hence, it is possible that oral carriage of H. pylori in patients with periodontitis, at least in part, might account for the reported association between periodontitis and gastric cancers.
Interestingly, some studies have identified a synergy between periodontitis and human papilloma virus (HPV), a virus which is known to contribute to the pathogenesis of oropharyngeal and possibly oral squamous cell carcinoma25–27. Tezal et al47 found that HPV-positive oropharyngeal squamous cell carcinoma was associated with increasing alveolar bone loss as a measure of periodontitis. They postulated that local inflammation as a result of periodontitis and the direct effects of periodontal bacteria may create a favourable environment for HPV, analogous to the well-established role of local inflammation in promoting cervical epithelium infection by HPV47. The same group identified similar findings in patients with base of tongue cancers40, and there is evidence to suggest that periodontal pockets act as a reservoir of HPV189,190.
With regard to other oncogenic viruses, cytomegalovirus (CMV) has been identified in several types of malignancy including colon cancer, malignant glioma, prostate carcinoma and breast cancer. However, the role of CMV in oncogenesis remains unclear191. Epstein-Barr virus (EBV) is clearly oncogenic and is a causative agent in lymphoproliferative disease192, gastric cancers193 and nasopharyngeal carcinoma194. Both CMV and EBV exist in periodontal pockets and saliva195 and are increased in patients with periodontitis196, and there is evidence to suggest that periodontal lesions are the main source of salivary cytomegalovirus197. Additionally, one study investigated periodontal pathogens in Kaposi sarcoma and found that short-chain fatty acids from periodontal pathogens stimulated Kaposi sarcoma herpesvirus (human herpesvirus 8) replication by promoting epigenetic changes198. Changes in the microbiome in periodontitis resulting in increased carriage of oncogenic viral species may be contributory to the development of some malignancies, particularly head and neck cancers, hematologic malignancies and gastric cancers.
Candida species, particularly C. albicans, have long been associated with the development of epithelial cancers and particularly head and neck squamous cell carcinomas. Candidal leucoplakias such as chronic hyperplastic candidiasis are considered to be premalignant in the oral cavity199,200 and Candida species have been identified in oral cancers159. Candida species possess pro-carcinogenic traits, including the production of nitrosamines201 and acetaldehyde202, but apart from the increased risk of OSCC in chronic hyperplastic candidiasis there is currently no epidemiological evidence supporting an association between Candida and malignancy in immunocompetent patients. The role that Candida species might play in carcinogenesis therefore remains controversial203. Candida species co-aggregate with bacteria in subgingival biofilms and form a part of the microbial community of periodontal pockets204. As discussed in Section 126.96.36.199, ‘Upper gastrointestinal tract cancers’, Huang et al74 identified an association between periodontitis, candida-related and denture-related oral mucosal lesions and pancreatic cancer. Although the exact role of Candida species in carcinogenesis is unclear, oral candidiasis as a result of poor oral hygiene and associated with periodontitis may have an impact on the development of malignancy in some individuals and additionally may represent a potential unidentified confounder in epidemiological studies investigating periodontitis and cancer.
8.3.3 Bacterial metabolic products
Periodontal bacteria engage in a variety of metabolic processes, producing diverse chemical products that can influence tumorigenesis and tumour progression (Fig 8-5). Nitrosamines have been implicated in the development of human cancers for over 40 years and this is thought to be mediated through direct damage to DNA205. Tobacco-specific nitrosamines are considered to be a potent risk factor among tobacco-related carcinogens in lung cancer progression206. Dietary sources of nitrosamines such as cured and preserved foods have been implicated in nasopharyngeal, oesophageal, gastric, pancreatic and colorectal cancers although the data supporting this association remain inconclusive207. In humans, some oral bacteria produce nitrosoamino acids208, but whether these can directly cause cancer remains elusive and poses an interesting topic of investigation.
Acetaldehyde is a volatile compound that is the main oxidation product of ethanol. In vitro and in vivo studies have shown acetaldehyde to have mutagenic and carcinogenic effects through its induction of a variety of genetic events, such as inducing point mutations and chromosomal aberrations and interfering with DNA repair. These effects are mediated directly by covalent binding to DNA and indirectly through catabolisation of folate, with diminished folate levels leading to DNA hypomethylation, an event that has been observed in many human cancers. Epidemiological, genetic and biochemical studies suggest that acetaldehyde is a local carcinogen in humans209,210. Oral bacteria including Neisseria, Streptococcus and Prevotella species are capable of metabolising ethanol to acetaldehyde through their alcohol dehydrogenases211–213.
Hydrogen sulphide (H2S) is a gas that in low concentrations protects against tissue injury, reduces inflammation and is involved in tissue repair214. However, there is evidence that in high concentrations H2S increases inflammation through stimulation of oxidative stress and ultimately induction of apoptosis and DNA damage215,216. H2S is produced endogenously by mammalian tissues, but is also a metabolic product of sulphate-reducing bacteria, which are increased in periodontitis217 and inflammatory bowel disease218. A potential role for H2S in the pathogenesis of periodontitis215, inflammatory bowel disease and colorectal cancer218 mediated through its pro-inflammatory effects has been suggested. Moreover, several studies suggest that endogenous H2S production is important for the growth and proliferation of colon219,220 and ovarian cancer220,221, but that the converse is true in melanoma220,222 and glioma220,223. The role of H2S produced by sulphate-reducing bacteria in the initiation and/or progression of cancer has not been specifically investigated, but H2S produced by sulphate-reducing periodontal bacteria might represent a potential mechanism accounting for the reported association between periodontitis and malignancy.
8.3.4 Host epigenetic changes encountered in periodontitis
Epigenetic mechanisms (those not coded for via DNA sequence that influence the development of an organism) are involved in normal development. However, disruption of epigenetic processes has been demonstrated to play a major role in oncogenesis. Epigenetic changes documented in cancer include DNA methylation, histone modifications, nucleosome positioning, and non-coding RNAs224. Epigenetics in periodontitis has not been extensively investigated, but there is some evidence that epigenetic changes play a role in the development of periodontitis225. Loo et al226 identified an increase in gene hypermethylation of E-cadherin and cyclooxygenase-2, both of which are implicated in tumour growth and metastasis227, in periodontitis patients when compared to controls, but to a lesser extent than in breast cancer patients. Similarly, Benakanakere et al228 demonstrated that P. gingivalis can induce DNA methylation in normal gingival epithelial cells and corroborated these findings in mice infected with P. gingivalis and in human periodontal tissues obtained from individuals with periodontitis. Yu et al198, in their study investigating periodontal pathogens in Kaposi sarcoma, demonstrated that P. gingivalis and F. nucleatum short-chain fatty acids inhibit various components of host epigenetic regulatory machinery including class-1/2 histone deacetylases, and ultimately lead to increased histone acetylation. Similarly, Martins et al229 describe histone acetylation in both periodontitis and on in vitro exposure of oral epithelial cells to LPS.
Peptidylarginine deiminase enzymes (PADs) are involved in the post-translational modification of protein to produce citrulline and have been implicated in the pathogenesis of rheumatoid arthritis (see Chapter 5: ‘Periodontitis and rheumatoid arthritis’). More recently, PADs expressed by periodontal bacteria, particularly P. gingivalis, have emerged as a potential mechanism to help explain the association between periodontitis and rheumatoid arthritis. There is growing evidence to suggest that PADs play an important role in tumorigenesis and tumour progression by histone citrullination resulting in altered gene expression230,231. The role of PADs from periodontal bacteria in cancer has not yet been investigated but this might be a promising area for future research.
8.3.5 Host genetics associated with both periodontitis and malignancy
Cancer has been traditionally considered a genetic disease and is thought to be driven by the accumulation of genetic mutations232. This view is changing as our understanding of epigenetics and tumour microenvironments evolves, but abnormal gene expression is still considered the major hallmark of cancer233. There is evidence that Grade C periodontitis is inheritable234,235, and host genetic polymorphisms have long been associated with periodontitis and appear to play a significant role in the aberrant host immune-inflammatory response to the polymicrobial infection seen in Grade C periodontitis236,237. In a study of direct relevance to this chapter, Arora et al238 explored the possibility of shared genetic risk factors between periodontitis and cancers in their prospective co-twin study and found that dizygotic twins with periodontitis had a 50% increase in total cancer risk, whilst this association was attenuated in monozygotic twins. Although these results might suggest that shared genetic risk factors may at least partially explain the association between periodontitis and cancers, the fact that this association was attenuated in monozygotic twins and that the assessment of periodontitis was unreliable (self-reported tooth mobility) strongly limits the impact of this study.
Cytochrome P450 enzyme isoforms are important in metabolising tobacco-derived substances. Specifically, polymorphisms of genes encoding P450 isoforms are associated with an increased risk of periodontitis and several cancers, including OSCC239–241. One study identified an increased risk of periodontitis in individuals with polymorphisms in the gene CYP1A1 encoding a member of the cytochrome P450 family239. Polymorphisms in the same gene have been implicated in a number of cancers. Other genes of relevance to this section are those encoding proteins involved in inflammation and its regulation. IL-1 polymorphisms are the most studied genetic association with periodontitis235, suggesting that IL-1 gene polymorphisms are associated with an increased risk of periodontitis242. Interestingly, IL-1 gene polymorphisms have been implicated in gastric cancer243,244; however, this association may be accounted for by increased susceptibility to H. pylori infection in these patients244. Polymorphisms in genes encoding other pro-inflammatory cytokines and receptors have been implicated in periodontitis and are associated with increased risk of certain malignancies: IL-1β and IL-10 in gastric cancer245, IL-8, IL-10 and vascular endothelial growth factor (VEGF) in prostate cancer246, cyclooxygenase-2 in colorectal cancer, gastric cancer, pancreatic cancer and hepatocellular carcinoma247,248, and several matrix metalloproteinases in lung cancer249, colorectal cancer250 and gastric cancer251. Additionally, polymorphisms in the high-affinity fMLP receptor (formyl peptide receptor 1 [FPR1]) of phagocytic cells have been implicated in both aggressive (Grade C) periodontitis252 and stomach cancer. One study found that the risk allele for stomach cancer pointed in the same direction as periodontitis253. Nucleotide-binding oligomerisation domain (NLRP) 2 and 3 genes encode proteins that perform complex functions as upstream activators of nuclear factor-κB signalling in inflammation, immune response and apoptosis254,255. Miskiewicz et al256 found that mutations in NLRP2 and NLRP3 were associated with periodontitis, pancreatic cancer and chronic pancreatitis. Lastly, a recent bioinformatic analysis of available datasets implicated several directly and indirectly interacting cross-talk genes as well as micro-RNAs present in both periodontitis and OSCC; however, whether these are implicated in either disease development remains unclear. Furthermore, the authors point out that due to the different types of tissues typically affected by periodontitis and OSCC, the identified biomarkers shared by these two diseases cannot be considered to be involved in the progression from chronic periodontal inflammation to oral cancer, and thus cannot be regarded as playing critical roles in the early stage of oral cancer257.
● The potential mechanisms supported by the strongest evidence include local and systemic inflammation, oral microbiome dysbiosis and the direct effects of specific periodontal pathogens, and the effects of the metabolic products of periodontal bacteria (Fig 8-6).
● Inflammation and associated oxidative/nitrosative stress have been shown to contribute to tumorigenesis and tumour proliferation in multiple ways.
● Should an association between cancer and periodontitis exist then periodontally induced local and systemic inflammation and free radical production would appear to be leading candidates to account for this association.
● Although there is currently no direct evidence for a role in tumorigenesis or tumour progression of bacterial metabolic products such as nitrosamines, acetaldehyde and H2S, a possible role can be inferred as other exogenous as well as endogenous sources of these molecules have been implicated in carcinogenesis.
● The potential causal mechanisms that are supported by less evidence, either direct or indirect, but which remain of potential interest include epigenetic changes and shared genetic risk factors, such as aldehyde dehydrogenase (ALDH), CYP1A1 or cytokine gene polymorphisms.