4.1 Introduction

The term inflammatory bowel disease (IBD) is used to describe two major types of disease: Crohn disease (CD) and ulcerative colitis (UC)1. IBD is an immunological disorder of the gastrointestinal tract (Fig 4-1). While CD affects any part of the intestine and can affect the mouth, oesophagus, stomach and anus, UC primarily affects the rectum and may extend continuously to involve the colon (large intestine only). From a histological point of view, the main difference between the two subtypes of IBD is the number of bowel wall layers affected. In UC patients, only the innermost layer is involved, whereas all layers are affected in CD patients. An increasing prevalence worldwide over the last 30 years has been observed in both developed and developing countries2. A changing global lifestyle marked by the diffusion of a western diet, smoking habits, increased body hygiene, pollution and the increasing use of medications are the factors suspected in the sudden increase in IBD burden. In the USA, UK and France, approximately 1.6 million, 300,000 and 200,000 persons, respectively, are currently diagnosed with either CD or UC. A peak of incidence is observed during the second and third decades of life. Male sex is associated with a higher incidence rate of UC, and younger age is associated with a higher incidence rate of CD3^–5. Important protective factors in CD as well as UC are being breastfed as a child, prior Helicobacter pylori infection, and sufficient levels of vitamin D6. Important risk factors for both diseases are increased hygiene habits, experiencing a bacterial gastroenteritis in the past, urban environment, air pollution and the use of antibiotics, nonsteroidal anti-inflammatory drugs and oral contraceptives7. Smoking seems to have a contradictory effect on CD and UC. A meta-analysis of 15 studies, comprising more than 20,000 patients with IBD, found that current smoking was associated with an increased risk for CD (odds ratio [OR] 1.76; 95% confidence interval [CI] 1.40 to 2.22), whereas it was a protective factor for UC (OR 0.58, 95% CI 0.45 to 0.75)8. This apparent contradictory effect can only be speculated on, but might be related to smoking having an organ-specific, rather than a disease-specific, effect9. The anti-inflammatory effect of nicotine could have a positive influence on UC, whereas the negative effect of smoking on CD could be mediated by non-nicotine substances.

IBD pathogenesis involves the dysregulated interactions between the gut microbiota and the gut-associated mucosal immune system that take place when genetically susceptible individuals are exposed to detrimental environmental triggers10^,11. Different experimental animal models led to the establishment of four major principles in IBD pathogenesis12:

● different causes of inflammation, either induced or genetically based, give rise to few common pathways of immune response in the gut

● commensal resident gut microbiota can drive intestinal inflammation per se

● loss of oral tolerance, which is defined as the suppression of immune responses to food and commensal microbial antigens after they have been previously administered orally, and disruption of the epithelial barrier contribute to the development of intestinal inflammation

● polarised T helper cell (Th) responses (Th1 versus Th2), as well as defects in innate immunity, such as those initiated by Toll-like receptors (TLRs) and nucleotide-binding oligomerisation domain-like receptor (NLR) P3 inflammasome stimulation, mediate inflammatory chronic disease in the gut.

From a genetic point of view, one of the most frequent gene polymorphisms associated with IBD concerns the CARD15/NOD2, CARD 9, interleukin (IL)-10, and IL-23 receptors that function in the immune response against microorganisms13^,14. IBD, particularly CD, is generally thought to be driven by T lymphocytes, particularly Th1. The disease has classically been characterised by serum increases in the pro-inflammatory cytokines tumour necrosis factor (TNF), interferon-gamma (IFNγ), IL-17 and IL-23. Other specialised T cells, known as regulatory T cells (Treg), act to counterbalance these pro-inflammatory responses by producing the anti-inflammatory cytokine IL-10. The equilibrium between Th1 and Treg results in the control of intestinal inflammation11. From a microbiological point of view, a shift from symbiosis to dysbiosis (increased number of pathobiontic and decreased number of symbiotic microorganisms) characterises the gut microbiota in IBD patients and drives the immunological disequilibrium. An increased abundance in bacteria in the orders of Enterobacteriales, Pasteurellales, Veillonellales, and Fusobacteriales, and decreased abundance in Erysipelotrichales, Bacteroidales, and Clostridiales was strongly correlated with CD status in a large US cohort of children and adolescents15. Microbiome comparisons between CD patients with and without antibiotic exposure indicate that antibiotic use amplifies the microbial dysbiosis associated with CD. The biodiversity of the commensal gut microbiota is a time-dependent phenomenon that begins from birth and stabilises during childhood to ensure digestive (fermentation, hydrolysis), metabolic (vitamins, fatty acids), and immune (antimicrobial barrier) functions. Recent data collectively suggest that intestinal microbiota interact with the immune system to direct the lineage differentiation of both pro- and anti-inflammatory T-cell populations to promote either health or IBD11. Consequently, a quantitative or qualitative disequilibrium between pro- and anti-inflammatory microorganisms leads to a dysfunctional gut microbiota. In IBD patients, the gut microbiota diversity is clearly reduced with lower levels of the phyla Firmicutes, Bacteroidetes and of the bacterium Faecalibacterium prausnitzii than in healthy subjects16^,17. A promising intervention strategy has been recently studied in UC patients: the faecal microbiota transplantation18. In this multi-centric randomised clinical trial, remission of UC patients was associated with an increased and persistent gut microbial diversity after the faecal microbiota transplantation. This innovative treatment result suggests that resolution of chronic inflammatory diseases could be achieved through microbiota modulation19.

The diagnosis of IBD is clinically based on abdominal pain, painful defecation (diarrhoea), fever, fistulae and weight loss, and is usually confirmed with imaging such as ileocolonoscopy, enteroscopy or capsule endoscopy20. Inflammatory biomarkers such as C-reactive protein, lactoferrin and faecal calprotectin (released by leucocytes into intestinal lumen and stool during inflammation) are helpful in IBD patient screening21. A definitive diagnosis of CD is sometimes established due to the identification of multinucleated giant cells in mucosal biopsy, especially in gingivae22. Activity indices are used for clinical and research purposes to assess the severity of symptoms and the efficacy of various treatments (drugs) both in CD and UC (Tables 4-1 and 4-2). The Harvey-Bradshaw Index (HBI) is a useful and simplified version of the Crohn Disease Activity Index, which is considered the gold-standard tool24^,25. However, those scoring systems are criticised because they do not take into account fatigue, a common symptom of IBD, quality of life and endoscopic findings. For UC, several indices are used, such as the Truelove and Witts score26, the Mayo Score, and, more often, the Simple Clinical Colitis Activity Index (SCCAI)23. The thresholds for determining activity status in IBD patients are usually set at HBI ≥ 10 for CD and SCCAI ≥ 5 for UC. It should be noticed that the terms ‘activity of the disease’ and ‘severity of the symptoms’ are usually confused in diagnostic and therapeutic research and practice.

Table 4-1 The Harvey Bradshaw Index (HBI) for Crohn disease activity scoring*

*The clinical signification is as follows: remission < 5, mild disease 5–7, moderate disease 8–16, severe disease > 16.

Table 4-2 The Simple Clinical Colitis Activity Index (SCCAI) for ulcerative colitis activity scoring*23

*The clinical signification is as follows: active disease > 5.

The therapeutic strategy of IBD includes immunosuppressants (5-aminosalicylic acid/mesalami- ne), biotherapies such as TNF inhibitors (infliximab) and anti-integrin antibody (vedolizumab), thiopurines, or methotrexate and surgeries, depending on disease severity. The natural history of IBD is characterised by alternating periods of reactivation and long remission. Overall, life expectancy of patients is normal in UC and slightly reduced in CD. Nevertheless, the severe chronic symptoms may have a significant impact on quality of life. Aside from intestinal inflammation, musculoskeletal, dermatological, ocular, renal, pulmonary and cardiovascular complications, and colorectal cancers characterising disease severity, the oral-specific and non-specific mucosal lesions have long been considered as extra-intestinal manifestations of IBD4. Among multiple studies, tag-like and cobblestone lesions, aphthous-like ulcerations, cheilitis, caries and periodontitis (Fig 4-2) were found to occur in 20% to 50% of IBD patients27^,28. Recently, there has been increasing interest in the association between periodontitis and IBD, as reflected by a number of narrative and systematic reviews on the topic, from both medical and dental researchers5^,27^,29^–35. IBD is understood to be an inflammatory systemic disorder that can have an impact on periodontal tissue loss by influencing periodontal inflammation36.

4.2 Clinical evidence

Since the first case report of a young male with CD exhibiting periodontitis with severe neutrophil dysfunction37, evidence has accumulated for an association between IBD and periodontitis in both children and adult patients (Table 4-3). The relationship between periodontitis and IBD has been investigated largely through cross-sectional and case-control studies that show slight differences for periodontal parameters among IBD subtypes (CD and UC) across countries after adjustments for age and dental plaque. No randomised clinical trials have been conducted so far.

4.2.1 Association between IBD and periodontitis

Prevalence and severity of periodontitis have been shown to be higher in different IBD populations, coming from Greece, Switzerland, Jordan, Brazil and USA40^,42^,45^,47^,55. Recently, three meta-analyses including cross-sectional and case-control studies, reported signifcant associations between periodontitis and IBD, and presented similar odds ratios and confidence intervals34^,34a^,34b. One of these is detailed in Table 4-3. In terms of treatment needs, a recent study in 110 children and adolescents with and without IBD aged 4 to 18 years indicated that none of the IBD patients had a healthy periodontium, whereas 40% of controls had a healthy periodontium (P < 0.001). No difference in the amount of plaque between the two groups was shown, so IBD patients were more susceptible to periodontal inflammation than healthy individuals40.

Table 4-3 Human studies investigating the association of inflammatory bowel disease and periodontal parameters (clinical and/or microbiological). The majority of the evidence is based on case-control studies. Those studies show consistent periodontal outcomes. However, in each study, there are confounding factors to notice.

BOP = bleeding on probing; CAL = clinical attachment level; CD = Crohn disease; CPITN = Community Periodontal Index of Treatment Need; ELISA = enzyme-linked immunosorbent assay; GF = gingival fluid; HBI = Harvey Bradshaw Index; IBD = irritable bowel disease; LA-PPD = loss of attachment at sites with maximal periodontal pocket depth; GI = Gingival Index; OR = odds ratio; PBI = papilla bleeding index; PCR = plaque control record; PI = Plaque Index; PPD = probing pocket depth; TNF = tumour necrosis factor; UC = ulcerative colitis.

*Same study population.

In US adults, an early cross-sectional study in 107 patients found more frequent pocket depths ≥ 4 mm (13.9% more subjects with at least 1 pocket depth ≥ 4 mm) and more frequent attachment loss (11.9% more sites exhibited attachment loss ≥ 2 mm), but less bleeding on probing (2.2% less bleeding sites) and less mean attachment loss (0.6 mm less on average) in IBD patients than in controls (P < 0.01)55. IBD patients appeared to be at higher risk of mild to moderate periodontitis than non-IBD subjects. Nevertheless, this early study was based on a partial-mouth recording of clinical periodontal parameters, which represents a risk of underestimation bias in prevalence studies. No confounding factors such as smoking, other systemic diseases, medications or diet were reported. Consequently, these contradictory results should be carefully considered. A few years later, a case-control study in 101 German individuals failed to show a statistically significant difference in clinical attachment loss ≥ 4 mm in IBD patients. Once again, the study protocol included a partial-mouth recording of the clinical periodontal parameters. However, patients were matched for age, sex and smoking status with controls48. A recent Taiwanese cohort study revealed a mildly increased risk of developing UC in patients affected by periodontitis after adjusting for confounding factors such as age, gender and comorbidities. However, this study obtained its data from a national database; therefore, the methods by which periodontal health was assessed and defined are unknown39.

4.2.1.1 Effect of the IBD subtype (CD and UC)

The subtype of IBD (CD or UC) has a distinct effect on periodontal health, with UC patients being more severely affected compared to CD patients. An additional increase in periodontitis risk by 42 per 1000 patients was found; however, the underlying pathophysiological mechanisms remain unclear34. A Brazilian age-matched case-control study based on full-mouth examination reported a higher percentage of sites with clinical attachment loss ≥ 3 mm in UC patients than in CD47. When the study sample was stratified on smoking status, the difference between UC and CD were lost; underlying smoking was an important modifier of the association between IBD subtypes and periodontitis.

Dietary habits and medication use vary across IBD subtypes and activity/remission stages of the disease. Those factors need to be evaluated in further studies to confirm or not an independent relationship between IBD subtype and periodontitis.

4.2.1.2 Effect of the contributing factors

Common risk factors are shared by IBD and periodontitis, such as age, genetic predisposition, smoking, diet and socio-economic status34. Recently, a case-control study in 226 Swiss individuals explored whether the IBD activity or other disease characteristics, such as extra-intestinal manifestations, anti-inflammatory therapy, immunosuppressive therapy, probiotics, oral health practices, smoking and alcohol, were associated with periodontitis42. The thresholds for active versus inactive disease were set at HBI = 10 for CD patients and SCCAI = 5 for UC patients (Tables 4-1 and 4-2). ‘Brushing teeth at least twice daily’ was a protective factor for gingivitis risk in IBD patients (OR 0.2, 95% CI 0.05 to 0.75, P = 0.02). Interestingly, immunosuppressive therapy did not indicate any protective effect. The analysis showed that the periodontitis risk (defined by attachment loss ≥ 5 mm on the tooth with the deepest pocket) was increased with the existence of perianal manifestations (OR 2.7, 95% CI 1.9 to 6.67, P = 0.03) and by proctitis i.e. inflammation of the anus and the lining of the rectum (OR 8.7, 95% CI 1.05 to 72.28, P = 0.05) in UC patients. In patients with CD, the two significant risk factors for periodontitis were having a smoking history (OR 3.71, 95% CI 1.14 to 12.05, P = 0.03) and having an HBI ≥10 (OR 7.23, 95% CI 1.36 to 36.38, P = 0.02). Altogether, protective and risk factors for periodontitis differed in CD and UC patients. While no association was observed between immunosuppressive medications and periodontitis in adult IBD patients, an association was observed in children. Young patients with IBD immunosuppressive medications including azathioprine/6-mercaptopurine, methotrexate and cyclosporine had more plaque and gingival inflammation, and a higher periodontal treatment need than patients with IBD anti-inflammatory (aminosalicylates, corticosteroids) or anti-TNF medications40. Future well-controlled studies reporting food avoidance, medication use, subtype of IBD, duration of IBD, activity and severity (intestinal and extra-intestinal complications) should further investigate these inconsistent results.

4.2.2 Strength of association and quality of evidence

In this chapter, the overall evaluation of the strength of association between IBD and periodontal diseases (gingivitis, periodontitis) is based on the severity of the reported periodontal parameters, the number of published studies and the consistency of the results between studies. A significant association of IBD and periodontitis exists since more than 10 observational studies across different populations consistently reported periodontal breakdown, and three meta-analyses confirmed this association34^,34a^,34b. However, the overall quality of the evidence, based on the study design (cross-sectional and case-control) and the risk of bias, is moderate.

Future clinical studies should systematically take into account not only the psychosocial stress and smoking status, which are the classic confounding factors in periodontitis, but should also consider the patient’s diet and medications. Indeed, the potential impact of IBD-induced nutritional alterations and IBD-associated medications on the periodontium are highly relevant (see Section 4.3.2.3 ‘Nutrition and medication’).

4.3 Cellular and molecular mechanisms

4.3.1 Animal studies

4.3.1.1 Spontaneous colitis and periodontal breakdown (natural history of the disease)

Spontaneous periodontal breakdown has also been observed in chemically induced and genetically based animal models of IBD. These models can be suitable for studying the cellular and molecular aspects of the relationship between periodontitis and IBD.

Chemically-induced IBD in rodent models

Dextran sulphate sodium (DSS)-induced colitis is colitis due to loss of the epithelial barrier function and entry of luminal microorganisms or their products into the lamina propria. Such entry results in stimulation of innate and adaptive lymphoid immune responses and secretion of pro-inflammatory cytokines and chemokines. In addition, it results in the influx of cells with cytotoxic potential such as neutrophils and inflammatory macrophages. A polarised Th1 response has been observed in acute DSS colitis, whereas a mixed Th1/Th2 response was found in a chronic form of DSS colitis, achieved by repeated cycles of DSS administration. It should be noted that the DSS colitis model is not a mimic of human IBD, as extensive epithelial cell damage and invasion by microbes is unlikely to be the primary mechanism of human disease, both in the gut and in the mouth.

Another chemically induced mouse model of colitis is based on the intra-rectal administration of trinitrobenzene sulphonic acid (TNBS), which functions as a hapten (a small molecule that elicits an immune response only when attached to a larger protein) modifying autologous colonic proteins that subsequently induce a T cell-mediated response (mainly Th1). This results in autoimmune-like inflammatory responses in the intestinal mucosa12. The TNBS colitis model has contributed to the understanding of the relationship between oral tolerance mechanisms, a physiological process that allows the host to tolerate food and microbial antigens, and the resulting intestinal inflammation.

Histomorphometric studies have evaluated whether DSS and TNBS can also elicit alveolar bone destruction56^,57. Mice were treated by oral application of TNBS twice per week or by oral DSS delivery in the diet over a period of 18 weeks. Only animals treated with DSS developed severe systemic symptoms (intestinal inflammation), whereas alveolar bone loss (ABL) in both IBD models was significantly higher than in controls, at 7 weeks and beyond in a time-dependent manner. To explore the requirement for an intact host immune system in this model of ABL, severe combined immunodeficient (SCID) mice lacking both T and B cell responses were examined. No significant ABL was shown in those TNBS-induced colitis animals compared with controls. ABL in this model is modulated by T and B cell-dependent responses, supporting that exaggerated ABL cannot progress in the absence of host responses in this model. These findings suggest that the chemically damaged oral mucosa in mice results in inflammatory-induced ABL in response to commensal oral microbiota, similar to the way that gut mucosa reacts with the luminal commensal gut microbiota.

Genetically induced IBD in rodent models

Three main genetically induced mouse models have been used to study the relationship between periodontitis and IBD. The IL-10 knockout (KO) mouse model (i.e. mouse lacking IL-10, a major cytokine of the immune response) is one of the earliest models of chronic enterocolitis. In a seminal study in these mutant mice, alterations in the intestine were extensive, whereas animals kept under specific pathogen-free conditions developed only localised inflammation limited to the proximal colon58. This KO model led to the understanding that uncontrolled immune responses in IBD were stimulated by enteric antigens and that IL-10 was an essential immunoregulator in the intestinal tract. In addition, genetic polymorphisms at the IL-10 locus confer a higher risk of IBD in humans14. In IL-10 KO mice examined for naturally occurring ABL and for their humoral immune responses to the three periodontopathogens Prevotella intermedia, Bacteroides fragilis and Bacteroides vulgatus, and two enteric bacteria, periodontal destruction was 39% higher than in IL-10-positive controls, with a late onset and strong serum-reactivity against these three periodontopathogens59^,60. Secondly, the human leucocyte antigen (HLA)-B27 transgenic rat model has been associated with increased risk of developing the multi-organ system diseases, termed spondyloarthropathies, which include chronic gastrointestinal inflammation61. HLA-B27 transgenic animals express the human gene for HLA-B27, a human class I major histocompatibility molecule involved in antigen presentation. In these animals, periodontal destruction was 42% to 250% higher than in wild type controls62. In addition, the accelerated ABL found in HLA-B27 transgenic rats has been shown to be an adult-onset, age-dependent and sex-independent process63^,64.

Lastly, the senescence-accelerated mouse (SAM) has been shown to be highly susceptible to gut inflammation65. The SAMP1/Yit strain was created by selective breading of SAMP1 mice exhibiting spontaneous skin ulcerations. Spontaneous inflammation of the terminal ileum and caecum, which are the primary locations of CD lesions in humans, occurred by 20 weeks. Inflammation originates from the epithelial barrier defect, and is driven, at least in the early stages of the disease, by a Th1 response12. A histomorphometric study investigated the occurrence and progression of periodontitis in SAMP1/Yit mice, in the absence of any exogenous pro-inflammatory stimuli, and the temporal correlation between the onset and progression of periodontitis and the onset of CD-like ileitis. The study found that SAMP1/Yit mice exhibited greater ABL compared with parental control mice and a strong positive correlation (r² = 0.432; P = 0.002) between CD-like ileitis severity and ABL, independent of animal age66.

4.3.1.2 Bidirectional effect of the oral and gut dysbiotic microbiota

Dysbiosis is a condition in which the normal microbiome structure is disturbed, through burdens such as systemic diseases, environmental factors or medications67. On the one hand, the impact of IBD-associated gut dysbiosis on the oral microbiota was explored in two mouse colitis models: either DSS-induced colitis or bacterially induced colitis caused by Citrobacter rodentium infection. Both colitis models changed their oral microbiome, with different levels according the different sites sampled in the oral cavity: 1.8% to 5% on the tongue and buccal mucosa, and 7.2% to 35% in saliva. These findings indicate that IBD affects the overall oral microbiome composition, but in a higher proportion for the salivary microbiome than for the mucosal tissue microbiome that seemed to be more resistant to systemically induced dysbiosis. Significant changes were observed in Bacteroidetes and Gammaproteobacteria levels in oral and stool samples after colitis induction in mice68.

On the other hand, the impact of the oral microbiota on gut dysbiosis has been also investigated in two well-designed studies. As a background, it is noteworthy that a person produces and swallows 500 ml to 1.5 l of saliva per day, containing approximately one gram of bacteria (1011)69. It is therefore plausible that salivary microbiology would affect gastrointestinal microbiology70. Oral bacteria, from symbionts to pathobionts, poorly colonise the healthy intestine, but have been shown to colonise the diseased intestine. The use of gnotobiotic (germ-free animal born and living in sterile conditions) and specific pathogen-free animal models provide surrogate measures to assess the function of specific microbial communities or individual strains in the context of controlled environments, genetics, diets and diseases. Therefore, a study with gnotobiotic and specific pathogen-free mice demonstrated that Klebsiella species, aerotolerant bacteria isolated from the salivary microbiota in CD patients, are strong inducers of Th1 cells when they colonise the gut, leading to severe gut inflammation, especially in an IBD-genetically susceptible mice model (IL-10 KO)71. It must be highlighted that Klebsiella species are not periodontopathogens but commensal members of the oral microbiome. However, this bacterium was selected based on its intestinal colonisation, immune stimulation and antibiotic resistance capacities demonstrated in gnotobiotic mice.

A metagenomics study also revealed that the intestinal bacteria belonging to the Bacteroidales families were significantly elevated in Porphyromonas gingivalis administered mice. Of interest, the proportions of Porphyromonadaceae were still very low in these animals, suggesting that the pathobiont P. gingivalis may alter the gut microbiome not by outgrowing in the gut but by indirectly inducing endotoxemia. This P. gingivalis-induced alteration of gut microbiota coincided with increases in insulin resistance, endotoxemia and a decrease in tight junction gene expression in the gut, leading to mucosal intestinal barrier permeability72. Recent data have shown in animal models that P. gingivalis72 and oral Klebsiella species71 drive aberrant immune system responses in the gut resulting in elevated systemic inflammation. However, the impact of the ‘real life’ oral dysbiosis associated with periodontitis on gut dysbiosis and the immune dysregulation is still unexplored. Taken together, the pre-clinical studies suggest that periodontitis and IBD share common immunological, microbiological and (epi)-genetic mechanisms.

4.3.2 Human studies

4.3.2.1 Genetic polymorphisms

Polymorphisms of the NOD2/CARD15 gene, which encodes an intracellular pattern recognition receptor in macrophages, neutrophils, dendritic cells and Paneth cells, are well investigated73. NOD2 stimulation in Paneth cells of the epithelial intestinal barrier leads to antimicrobial peptides secretion; these participate in the gut innate immune response. In a cross-sectional study of 147 CD patients, no difference in periodontal clinical and microbiological parameters was shown amongst patients with and without a CARD15 polymorphism, except for the decrease of P. intermedia levels in CD patients with a CARD15 mutation (69.7% versus 87.7%; P = 0.008)53. While CD patients with NOD2/CARD15 mutation exhibit a higher risk of IBD14, there is no evidence to support a higher risk of periodontitis.

Similarly, the TNF-α polymorphism that is associated with a higher risk of IBD14 was investigated in a second cross-sectional study from the same group. A positive association between TNF-α polymorphism and periodontal breakdown (measured by probing pocket depth and clinical attachment loss) was observed in 142 CD patients, independently of age, sex, medications and smoking52. Recently, a global transcriptomic case-control study analysed the gene expression profile of periodontitis and chronic inflammatory diseases. The pleckstrin (PLEK) gene was the only upregulated gene shared by periodontitis and the comorbidities of UC, rheumatoid arthritis and cardiovascular diseases with fold-changes of 1.6, 1.8, 4.1 and 1.5, respectively74. PLEK is a protein kinase C substrate expressed by macrophages. It is an important intermediate in the secretion and activation of the pro-inflammatory cytokines TNF and IL-1β, and is induced in macrophages in response to the lipopolysaccharide of Gram-negative bacteria. There is weak evidence for a genetic background modulating effect in the association between IBD and periodontitis.

4.3.2.2 Dysbiosis of the oral and gut human microbiota

The human microbiome is a complete ecosystem with a trophic organisation of species-rich bacterial, virus and fungal communities showing extreme heterogeneity, with a relatively small number of highly connected nodes (phyla). Microbial communities are highly resistant to endogenous and exogenous perturbations (hormonal fluctuations, antibiotics, etc.). However, a loss of keystone species may have cascade effects, with secondary species extinctions. Therefore, a high microbial biodiversity reduces the risk of pathogen selections, disease-associated dysbiosis and inflammatory responses75.

Impact of IBD on the oral microbiota

Since the classic work of Van Dyke et al49 in 1986, oral and subgingival microbiota particular to IBD have been suggested. In the study, bacterial cultures of the subgingival plaque of 20 patients with IBD (CD and UC) identified the predominance of the genus Wolinella, a Gram-negative bacterium, irrespective of the periodontal status49. Streptococcus mutans, lactobacilli and yeasts in saliva were then associated with the activity of CD in a cross-sectional study in 53 patients. Despite high acidogenic bacteria and yeast counts that were probably due to high-sugar diets associated with CD activity, a plausible contributing role for these microorganisms in IBD pathogenesis could not be ruled out54. In a case-control study of subgingival plaque, 15 CD patients and 15 UC patients with untreated periodontitis presented higher counts of Prevotella melaninogenica, Staphylococcus aureus, Streptococcus anginosus, S mutans and Peptostreptococcus anaerobius, compared to 15 controls with untreated periodontitis43. Consequently, the periodontal microbiota associated with IBD can modulate periodontitis and dental caries initiation and progression76. These classic studies were limited by the a priori approach using bacterial culture or DNA-DNA hybridisation. A step forward in the field was recently taken with the metagenomics analysis based on pyrosequencing of the bacterial 16S rRNA gene.

The oral microbiome has been studied in saliva of both IBD adults and children without periodontitis. In adults, a significant increase in Bacteroidetes with a concurrent decrease in Proteobacteria was observed41. The dominant genera, Streptococcus, Prevotella, Neisseria, Haemophilus, Veillonella and Gemella, were found to contribute greatly to the oral dysbiosis observed in the salivary microbiota of IBD patients. These changes in microbiota were also strongly correlated with changes in inflammatory responses. While Prevotella, Actinomyces, Veillonella, and Lachnospiraceae were positively associated with elevated IL-1β and IL-8 in saliva of IBD patients, Streptococcus, Rothia, Neisseria, Haemophilus, and Gemella were negatively associated with those pro-inflammatory cytokines41. In a similar case-control study including 71 children with IBD without periodontitis, the overall microbial diversity in oral mucosal brushing samples was significantly lower compared with 43 healthy controls with significant differences were reported in Fusobacteria and Firmicutes44. No differences were noted in the microbial composition of 31 UC patients. These findings imply that IBD as a systemic disease has an influence on the oral microbiome dysbiosis that could predispose patients to oral and periodontal diseases.

The subgingival plaque microbiome has also been investigated in a longitudinal cohort study in CD children without periodontitis, at the time and 8 weeks after infliximab initiation (anti-TNF biotherapy)38. Capnocytophaga, Rothia and Candidatus Saccharibacteria (previously TM7) were more abundant in CD patients relative to controls. Although these taxa are part of the commensal oral microbiome, it should be noted that Candidatus Saccharibacteria has been associated with periodontitis77^–79 and could predispose susceptible CD patients to periodontal breakdown. Moreover, in CD patients after 8 weeks, the infliximab therapy did not restore a periodontal microbiome similar to the controls. Thus, disease per se (and not anti-TNF medication) seems to alter the oral microbiota in young CD patients.

Impact of periodontitis on the gut microbiota

Continued development in sequencing technologies provides a platform to analyse oral (saliva samples and mucosa tissue biopsies), periodontal (subgingival plaque samples), and faecal (stool samples) microbiomes of both periodontally healthy and diseased individuals. The active molecular mechanisms that link dysbiosis of the oral and periodontal microbiota to intestinal inflammation should be systematically explored in animal models as well as human studies because it remains to be demonstrated whether oral-associated dysbiosis could be a contributory (or even causative) factor or is a consequence of IBD-associated intestinal dysbiosis (Fig 4-3). The intestinal microbiome has been explored in a case-control study conducted in 44 patients, 23 with periodontitis, 14 with gingivitis and 7 controls80. The metagenomics analysis of the stool samples revealed that the Firmicutes, Proteobacteria, Verrucomicrobia and Euryarchaeota were increased, whereas the Bacteroidetes were decreased in abundance in patients with periodontitis compared with controls. There was a significant correlation between the periodontal status (measured by clinical attachment loss) and the periodontal genera identified in the gut microbiome. Taken together, these preliminary results support the periodontal-induced intestinal dysbiosis.

4.3.2.3 Nutrition and medication

The frequency and importance of diarrhoea and abdominal pain can lead to nutritional deficiency due to malabsorption and unbalanced diet. Particularly, high sugar intake, anaemia and deficiencies in vitamins A, C, D and in zinc are frequent complications in IBD patients48^,81. Supplementation with iron, zinc, magnesium and vitamins is often necessary, either orally or intravenously. Recent data strengthen the association between high intake of refined carbohydrates, vitamin deficiencies and periodontitis82^,83. Moreover, nutrients can modulate the inflammatory reaction leading to periodontal breakdown both directly and indirectly by altering the diversity of the gut microbiome (Fig 4-4). Studies have shown that the individual’s diet contributes to shaping the microbial communities that take up residency in the oral and the gastrointestinal tract microbiomes84^,85. A rural diet was associated with gut microbiome enrichment in Prevotella, whereas a western diet led to a Bacteroidetes-enriched microbiome84. Another interventional study demonstrated that a strict vegetarian diet corrected the Firmicutes/Bacteroidetes ratio in obese patients85. At the micronutrient level, insufficient or excessive iron intake can lead to microbial dysbiosis through the growth of periodontal pathogens such as P. gingivalis86^–88. In this context, a large retrospective study in the US found that 79.2% of subjects with CD and 85.1% of those with UC had iron-deficiency anaemia measured by ferritin plasma levels89. A recent clinical study has found greater gingival inflammation, pocket depths and clinical attachment ≥ 6 mm in periodontal patients with iron-deficiency anaemia than in controls90.

Considering medications, all of the drug classes that are applied to the treatment of IBD can lead to periodontal alterations due to the indirect toxic effects of these drugs on soft and hard tissues, as well as direct immunosuppressive effects5^,27^,81. The roles of IBD treatments, such as corticosteroids, immunomodulators (sulphasalazine, azathioprine, 6-mercaptopurine, methotrexate, cyclosporine and tacrolimus) and biotherapies (anti-TNF-α and anti-integrin) on the pathogenesis of periodontitis are unknown.

4.3.2.4 Immune response

Due to the immune-inflammatory pathogenesis shared by periodontitis and IBD (Fig 4-4), several human studies have aimed to investigate alterations in the innate and adaptive immune response in IBD patients with periodontitis31^,33. A historical case report in a CD patient indicated an impaired neutrophil function associated with a rapidly progressive severe periodontitis37. Subsequently, a defect in neutrophil chemotaxis was assessed in 10 IBD patients with periodontitis compared with 10 IBD patients without periodontitis49. This neutrophil dysfunction was driven by the subgingival plaque extract in vitro, though neutrophil phagocytosis was not affected. The flow-cytometric profile of a woman with CD and severe periodontitis indicated a paucity of B lymphocytes and an elevated proportion of T lymphocytes in comparison with periodontally healthy controls91. These early results suggested both innate and adaptive immune dysfunction in the IBD and periodontitis association. Currently, an imbalance between cytokines of the adaptive and innate immune systems is thought to be crucial for the onset and progression of chronic inflammatory diseases.

The aim of a pilot case-control study was to characterise the pattern of cytokine concentrations in the gingival fluid and in serum from periodontitis patients with and without IBD46. The expression pattern of cytokines in gingival fluid was not different in periodontitis patients with and without IBD. The study results indicate that IBD has a relatively small effect on the local expression of the cytokines. Another study from the same population used a cross-sectional design to investigate the pattern of cytokines directly in gingiva and intestinal mucosa51. The results showed different clusters of cytokines between gingiva and intestinal mucosa in IBD patients. In addition, when those clusters were used to assess a site-specific disease activity score, the results showed a good correlation, indicating that activity of IBD influenced the cytokine expression pattern in gingival tissue50. In summary, few human studies suggest an altered innate and adaptive immune function in IBD and periodontitis, providing a possible biological mechanism driving both diseases.

Stay updated, free dental videos. Join our Telegram channel

Join