Periodontitis and rheumatoid arthritis
Isabel Lopez-Oliva and Mark Bartold
Rheumatoid arthritis (RA) and periodontitis are both chronic inflammatory diseases characterised by an exacerbated immune-inflammatory reaction that leads to destruction of bone and other connective tissue. Due to these similarities, the relationship between these two diseases has been investigated for many years. RA is a chronic immune- mediated inflammatory disease characterised by destruction of cartilage and bone and its cause remains unknown1. The prevalence of RA in the first world is believed to be increasing, currently affecting about 1% to 2% of the world population in a male/female ratio of 1:32. Although the cause of RA remains unknown, understanding of the pathogenesis of RA has increased substantially over the last decade. The inflammatory process that occurs in the synovium of RA patients is similar to the one occurring in the tissues adjacent to the periodontal pocket of periodontitis patients, characterised by an elevated production of pro-inflammatory cytokines including tumour necrosis factor (TNF), interleukins (IL)-1, -6, -15, -17 and granulocyte-macrophage colony-stimulating factor (GM-CSF). In the presence of these cytokines, T cells and synoviocytes activate osteoclast maturation, which leads to bone degradation in the joint (Fig 5-1). In addition to the inflammation of the synovium, the pannus (an anomalous fibrovascular coating) encroaches into and destroys the joint surface3,4.
This constant state of inflammation and degradation of the joint structure causes chronic pain, deformations, functional impairment and disability. In addition, the mortality of RA patients is increased due to associated comorbidities, such as cardiovascular disease5. The typical signs and symptoms of RA are general pain and swelling of joints (often symmetrically), morning stiffness and movement limitations that last more than one hour and that can be reduced by gentle movements. These clinical signs help the differential diagnosis with osteoarthritis (OA), which is the most common form of joint disease caused by cartilage degeneration and mechanical wear (Table 5-1). OA is not an autoimmune inflammatory disorder, does not normally present with swelling, the morning stiffness lasts less than an hour and the joint movement can worsen the pain instead of helping it6. Interestingly, OA has not been found to be associated with chronic periodontitis and it is frequently used as control group in studies that investigate the relationship between RA and periodontitis7–9.
Table 5-1 Different signs of symptoms characterising rheumatoid arthritis (RA) and osteoarthritis (OA)
|Auto-immune inflammatory destruction of joint tissues||Mechanical wear of joint tissues|
|Morning stiffness > 1 h||Morning stiffness < 1 h|
|Swelling of joints||Little or no swelling|
|Joint movement can help pain||Joint movement can worsen pain|
|Fatigue, generally unwell, rapid progression||No general symptoms, develops over years|
The 2010 American College of Rheumatology (ACR)/European League Against Rheumatism (EULAR) devised new classification criteria that employ a score-based algorithm considering type and number of joints involved, serological parameters (rheumatoid factor and anti-citrullinated protein antibody), acute-phase reactants (C-reactive protein and erythrocyte sedimentation rate) and duration of the symptoms (Table 5-2). A score of 6 out of 10 is needed to diagnose the patient with definitive RA10. One of the new additions to these classification criteria was the detection of specific antibodies as a diagnostic tool. Rheumatoid factor (RF) is a family of antibodies against the fragment crystallisable region (Fc) of immunoglobulin G (IgG) and these antibodies are found to be elevated in RA and other connective tissue diseases, infectious diseases and liver disease. Anti-citrullinated protein antibodies (ACPA) are found in 80% of RA patients and have 99% specificity, being the most disease-specific antibodies in RA. Both RF and ACPA appear years before clinical signs11 and are independent risk factors for bone erosion12. The process of protein citrullination and ACPA is outlined in Section 5.3 ‘Cellular and molecular mechanisms’.
Table 5-2 American College of Rheumatology (ACR)/European League Against Rheumatism (EULAR) classification criteria of rheumatoid arthritis (based on Aletaha et al10)
|A. Joint involvement||1 Large joint||0|
|2–10 large joints||1|
|1–3 small joints (with or without involvement of large joints)||2|
|4–10 small joints (with or without involvement of large joints)||3|
|> 10 joints (at least 1 small joint)||5|
|B. Serology (at least 1 is needed for classification)||Negative RF and negative ACPA||0|
|Low-positive RF or low-positive ACPA||2|
|High-positive RF or high-positive ACPA||3|
|C. Acute-phase reactants (at least 1 is needed for classification)||Normal CRP and normal ESR||0|
|Abnormal CRP or abnormal ESR||1|
|D. Duration of symptoms||< 6 wk||0|
|≥ 6 wk||1|
ACPA = anti-citrullinated protein antibodies; CRP = C-reactive protein; ESR = erythrocyte sedimentation rate; RF = rheumatoid factor.
The pharmacological treatment of RA aims to relieve pain, reduce inflammation and prevent destruction of cartilage and bone (Table 5-3). Rheumatologists often use a number of different drugs to achieve this, such as non-steroidal anti-inflammatory drugs (NSAIDs), corticosteroids (orally or by injection) and conventional disease-modifying anti-rheumatic drugs (DMARDs). This last group of drugs acts on the immune system to retard the inflammatory progression of RA, with methotrexate being the most common DMARD prescribed. Biological therapies, a subcategory of DMARDs, target specific steps of the inflammatory process and have proved to be very effective in patients with active RA who have not responded to other treatment. A very common biological therapy, anti-TNF treatment, in combination with conventional DMARDs provides clinical benefits and halts the progression of joint damage13.
Table 5-3 Drugs most commonly used for the treatment of rheumatoid arthritis
|Type of drug||Most commonly used|
|NSAIDs, analgesics||Paracetamol, ibuprofen|
|Corticosteroids||Pills or injections|
|DMARDs||Methotrexate, azathioprine, hydroxychloroquine, leflunomide|
|Biological DMARDs||Abatacept, rituximab, tocilizumab|
DMARDS = disease-modifying anti-rheumatic drugs; NSAIDs = non-steroidal anti-inflammatory drugs.
5.2.1 Epidemiological association between periodontitis and RA
Over the last 20 years many studies have identified an association between periodontitis, tooth loss and RA (Table 5-4). While two of these studies reported negative results, the majority observed a higher risk and prevalence of periodontitis in RA patients and vice versa. Cross-sectional studies have shown that patients with RA have a significantly increased prevalence of periodontitis compared to systemically healthy controls, with odds ratios (ORs) ranging between 1.8217 and 8.118. Furthermore, patients with periodontitis have a higher prevalence of RA, with ORs ranging between 1.1621 and 2.0520. Some of these studies, after adjusting for confounding factors, showed that this relationship appears independent of smoking21, oral hygiene (plaque)15,18 and genetic factors16. A well-conducted 2019 study with the aim to assess whether periodontitis is associated with RA and whether periodontitis severity is linked to RA disease activity included 187 patients diagnosed with RA and 157 age-matched controls with OA or soft tissue rheumatic diseases. The authors measured DAS28 (disease activity score), Simplified Disease Activity Index, Clinical Disease Activity Index, RF, ACPA titres, erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), presence of extra-articular manifestations and type of RA therapy. Periodontal status was determined by masked examiners though Plaque Index, bleeding on probing, probing pocket depth and clinical attachment levels. Sociodemographic variables and comorbidities were evaluated as confounding variables in this study. A significant association was demonstrated between periodontitis and RA with an adjusted OR of 20.57 (confidence interval [CI] 6.02 to 70.27) and an association between periodontal severity and RA disease activity with an adjusted OR of 2.66 (CI 1.24 to 5.74)31.
Table 5-4 Clinical studies published since 2000 (with study population ≥ 60) studying the association between rheumatoid arthritis and periodontitis
|Study||Study population||Outcomes||Study design|
|Mercado et al14||1412 patients attending dental hospital||Periodontitis: higher prevalence of RA (3.95%)||Cross-sectional|
|Mercado et al15||65 patients (RA vs. non-RA)||RA: higher number of missing teeth, deeper pockets. No difference in bleeding or PI||Case-control|
|Marotte et al16||147 patients (RA)||Association between periodontal bone loss and wrist bone destruction (χ2 = 11.82) and shared epitope HLA-DR||Cross-sectional|
|De Pablo et al17||4461 patients (NHANES III)||RA: higher prevalence of periodontitis (OR 1.82), edentulous (OR 2.27), less decay (P < 0.001)||Cross-sectional|
|Pischon et al18||109 patients (57 RA, 52 non-RA)||RA: higher prevalence of periodontitis (OR 8.05), statistically significant after adjusting to confounding factors (PI, GI)||Case-control|
|Dissick et al8||69 RA patients vs. 35 osteoarthritis (controls) patients||RA: higher prevalence of periodontitis and more severe RA patients with periodontitis associated with RF positive and anti-CCP||Case-control|
|Arkema et al19||81,132 patients (Nurses’ Health Study prospective cohort)||No evidence of higher incidence of RA in periodontitis||Cohort study|
|Demmer et al20||9702 patients (NHANES I)||Periodontitis: higher prevalence of RA (OR 2.05)||Cross-sectional|
|Potikuri et al21||91 RA (DMARD naive, non-smokers) vs. healthy controls||RA: higher prevalence of periodontitis (OR 4.28)||Case-control|
|Smit et al22||95 RA, 420 matched controls||RA: higher risk of periodontitis (RR 3.7)||Cross-sectional|
|Chen et al23||13,779 newly diagnosed RA, 137,790 non-RA||Periodontitis: higher prevalence of RA (OR 1.16)||Cohort study|
|Monsarrat et al24||74 RA patients||94% of RA had periodontitis (48% moderate and 46% severe)||Cross-sectional|
|Eriksson et al25||2740 RA cases and 3942 non-RA||No difference in periodontitis prevalence between groups||Case-control|
|Äyräväinen et al26||53 patients with ERA, 28 patients with CRA, 43 age- and gender-matched controls||Moderate (Grade B) periodontitis present in 67.3% of patients with ERA, 64.3% of patients with CRA versus 39.5% of control participants. No association between anti-rheumatic treatment and periodontal parameters||Cohort study|
|Schmickler et al27||168 RA cases and 168 age- and gender-matched controls||RA patients: significantly higher probing depths and attachment loss than controls. Moderate to severe periodontitis (Grade B–C) present in 98% of RA patients vs. 82% in controls||Cross-sectional|
|Hashimoto et al28||89 Japanese RA patients||No significant relationship between RA severity and periodontal status||Cross-sectional|
|Fathi et al29||60 RA patients and 30 controls||Periodontitis present in 71.7% of RA patients vs. 46.7% in controls||Case-control|
|Ziebolz et al30||168 RA patients, divided into groups according to type of medication||Patients taking methotrexate + TNF-α antagonists had higher bleeding scores than those taking leflunomide or methotrexate + rituximab||Cross-sectional|
|Rodríguez-Lozano et al31||187 RA patients and 157 controls with osteoarthritis and soft tissue rheumatic disease||Significant association between periodontitis and RA (OR 20.57, 95% CI 6.02–70.27) and between periodontal severity and RA disease activity (OR 2.66, 95% CI 1.24–5.74)||Case-control|
|Zhao et al32||128 Chinese RA patients and 109 healthy controls||RA patients had hig-er prevalence of periodontitis||Cross-sectional|
anti-CCP = anti-cyclic citrullinated peptide; CI = confidence interval; CRA = chronic RA; ERA = early DMARD-naive RA; GI = Gingival Index; NHANES = National Health and Nutrition Examination Survey; OR = odds ratio; PI = Plaque Index; RA = rheumatoid arthritis; RF = rheumatoid factor; RR = risk ratio.
In the National Health and Nutrition Examination Survey (NHANES) data reported by De Pablo et al17, patients with RA had a higher prevalence of periodontitis (10% to 16%) compared to non-RA, and 50% of subjects with RA were identified as edentulous. The authors concluded that these patients may be representative of severe cases of periodontitis and although the cause of tooth loss was not known, RA patients had a lower frequency of fillings and caries17. Similarly, in one of the largest cohort studies investigating the Taiwanese National Health Insurance Research Database (NHIRD), the authors found a significant and independent association between RA and history of periodontitis23. Three other studies reported that out of the RA patients examined, between 72% and 98% suffered from moderate to severe periodontitis24,27,29. In addition, a recent prospective study also assessed whether anti-rheumatic drugs have an influence on periodontal status after 16 months. The authors reported that although periodontitis was more prevalent in RA patients, medication did not have an effect on the periodontal parameters periodontal probing depth and clinical attachment loss26. In a recent retrospective study on 54 RA patients, Yamashita et al33 found an association between periodontal inflamed surface area (PISA) and RA clinical parameters. Moreover, patients with a low PISA exhibited greater improvement after 6 months of biological DMARD therapy compared to patients with high PISA33.
In contrast, there are also studies that did not report such associations. In a large prospective cohort study, using data from 91,132 nurses followed over 12 years, there was no evidence of a higher risk of developing RA among patients with periodontitis19; however, in this study periodontitis was defined based on history of periodontal surgery through a self-report questionnaire, and no clinical examinations were conducted to diagnose periodontitis. More recently, no difference in the prevalence of periodontitis was observed in a Swedish case-control study including 2740 RA patients. However, again in this study periodontitis was investigated using a self-reporting questionnaire25. Thus, these two studies may be biased due to the recruitment methodology, as no clinical examination was used to diagnose periodontitis and therefore their conclusions may be misleading.
The most recent systematic reviews and meta-analyses exploring the epidemiological evidence, predominantly from case-control studies, concluded that there is a strong and significant association between RA and periodontal disease. However, the authors highlight the need for further rigorous studies using consistent case definitions for periodontitis and a more defined population to avoid bias due to invalid control groups34,34a,34b,34c.
● Numerous studies exist stating a positive association between RA and periodontitis.
● Those studies with negative results have risks of bias.
● Systematic reviews conclude that there is a strong and significant association between RA and periodontitis.
● Further large rigorous studies are needed with appropriate case definitions of periodontitis.
5.2.2 Effects of periodontal treatment on RA
For a long time, researchers have believed that treating oral health problems improved RA. The first records date to the time of Hippocrates, who suggested that the cure for arthritis included tooth extraction. Many years later, in the 19th century, Benjamin Rush35 proposed that rheumatism could be cured by removal of infected teeth. However, later studies by Cecil and Angevine36 failed to replicate these findings. During the 1950s and 1960s, a ‘total dental clearance’ was advocated for the treatment of RA37.
A few small clinical studies evaluating the effect of non-surgical periodontal therapy on RA suggest that treatment of periodontitis may have a significant positive effect on RA severity (Table 5-5)38,43. Due to its high prevalence, periodontitis may represent an important modifiable risk factor for RA incidence and severity. If proven, then the treatment of periodontitis could offer a relatively inexpensive and safe non-pharmacological treatment with direct benefit for patients with RA. Although results are promising, some authors have failed to observe this effect. In a case report of one RA patient with periodontal treatment for 15 years, no beneficial effect was reported47. In a randomised clinical trial, Pinho et al41 did not observe a reduction in acute phase reactants following periodontal therapy in RA, and these systemic markers of inflammation did not correlate with observed improvements in periodontal health. Also in a very recent randomised controlled trial of 22 patients, the authors failed to observe an improvement after periodontal therapy24. However, in the workshop between the European Federation of Periodontology and the American Academy of Periodontology in 2013, it was concluded that more rigorous controlled clinical trials and research were needed in the field48.
Table 5-5 Interventional studies evaluating the effect of periodontal therapy in rheumatoid arthritis
|Study||Duration||Patient number||Parameters evaluated||Results|
|Ribeiro et al38||3 mo||42 RA + PD patients: 16 periodontal treatment, 26 oral hygiene and supragingival cleaning||RF, ESR, HAQ||ESR significantly reduced|
|Al-Katma et al39||8 wk||29 RA + PD: 17 periodontal treatment, 12 no treatment||DAS 28, ESR||VAS, DAS and ESR reduced|
|Ortiz et al40||8 wk||40 RA + PD patients: 10 periodontal treatment and DMARDs only; 10 periodontal treatments and DMARDs with anti-TNF drugs; 10 no periodontal therapy, DMARDs only; 10 no periodontal treatment, DMARDs with anti-TNF drugs||ESR, TNF-α, signs and symptoms||VAS and DAS improved in treatment groups; ESR not significantly reduced; anti-TNF drugs improved PPD and CAL|
|Pinho et al41||6 mo||75 patients: 15 RA + PD with periodontal treatment; 15 RA + PD no periodontal treatment; 15 PD with periodontal treatment; 15 PD no periodontal treatment||DAS 28, CRP, ESR, AAG||No clear relationship; AAG, ESR and CRP not significantly reduced with periodontal therapy|
|Okada et al42||8 wk||55 RA + PD patients: 26 supragingival cleaning; 29 no treatment||DAS 28, CRP, anti-CCP, RF, TNF-α and levels of IgG to P. gingivalis||Reduction of DAS 28 and levels of IgG to P. gingivalis and citrulline|
|Erciyas et al43||3 mo||60 RA and PD patients: 30 RA moderate–severe disease activity; 30 RA low disease activity||ESR, CRP, TNF-α, DAS28||Significant reduction of ESR, CRP, TNF-α, DAS28|
|Cosgarea et al44||3 and 6 mo||18 RA and PD patients, 18 systemically healthy PD patients – both groups received periodontal treatment||DAS28, CRP, ESR||Significant decrease of serum CRP at 3 mo only|
|Monsarrat et al45||3 mo||22 RA and PD patients: 11 periodontal treatment with systemic antibiotics; 11 no periodontal treatment||DAS28 based on ESR||No improvement of DAS-ESR|
|Kaushal et al46||8 wk||40 RA and PD patients: 20 periodontal treatment, 20 no periodontal treatment||CRP, anti-CCP, RF, SDAI||Significant reduction of SDAI, no reduction of CRP, anti-CCP and RF|
AAG = alpha-1 acid glycoprotein; CAL = clinical attachment level; CRP = C-reactive protein; DAS = Disease Activity Score; ESR = erythrocyte sedimentation rate; HAQ = Health Assessment Questionnaire; IgG = immunoglobulin G; PD = periodontitis; PPD = pocket probing depth; RA = rheumatoid arthritis; RF = rheumatoid factor; SDAI = Simplified Disease Activity Index; VAS = visual analogue scale; TNF = tumour necrosis factor.
To date, only a few randomised clinical trials have investigated the effect of periodontal therapy on RA and these studies have a short follow-up period (ranging between 8 weeks and 6 months) and a small sample size (< 75 patients). Furthermore, each study employed different definitions of periodontitis and used different parameters to measure RA status, with the DAS28 being the most widely reported, based on subjective measures, such as the visual analogue scale (VAS) and number of tender joints. Due to the heterogeneity in all the studies connecting RA and periodontitis, in the first systematic review considering the effects of treatment of periodontitis on RA outcomes, only a few parameters could be included for meta-analysis; ESR being the only parameter found to be significantly reduced after periodontal treatment49. More recently, a systematic review considering those same four studies concluded that although DAS28 seemed to be constantly improved after periodontal therapy in the four trials, further RCTs are needed9.
Therefore, although the current evidence suggests that there is an improvement in RA parameters after periodontal therapy, investigators in the field agree that more long-term randomised controlled trials are necessary to assertively determine the effects of periodontal treatment on RA.
For a more rigorous study, a large population would be required (> 100 patients) to reduce the effect of the patient variability introduced by the nature of RA (flare-ups and medications). Moreover, it is necessary to use the most established case-definition of mild, moderate and severe periodontitis (i.e. Eke et al50) and evaluate the most commonly studied RA parameters (i.e. DAS 28, ESR, CRP, anti-cyclic citrullinated peptide [CCP]), in order to standardise study outcome measures and be able to perform a meaningful meta-analysis. Furthermore, longer review times (> 6 months) might be necessary to estimate the effect of periodontal therapy on RA, although this raises ethical concerns for the control group. Lastly, periodontal health must be achieved through consistent periodontal review and maintenance. With treatment performed to defined periodontal endpoints. Clearly, poor periodontal outcomes are unlikely to impact positively upon RA or RA disease activity outcomes. If periodontal healing is not achieved locally, the systemic effects would consequently not exist.
● Very few studies on the effects of periodontal treatment on RA exist.
● Different case definitions for periodontitis and different oral and RA parameters have been studied, making comparisons of these studies difficult.
● Systematic reviews conclude there is an improvement of RA following periodontal therapy.
● Larger and more rigorous studies are needed.
5.3 Cellular and molecular mechanisms
The notable similarities between RA and periodontitis have been the object of study for many years51. The biological mechanisms that explain the interrelations between the two conditions are not known, and different theories have been proposed.
5.3.1 Shared risk factors
RA and periodontitis share numerous risk factors:
● host-mediated destruction
● a role for plasma cells in active disease
● polymorphonuclear neutrophil infiltration = substantial
● oxidative stress = major feature
● female sex hormones play a role
● symptoms respond to anti-inflammatory drugs.
Gene polymorphisms have been proposed as shared risk factors in several studies. These are discussed in further detail in Section 5.3.7 ‘Genetic background’. Oral hygiene has been proposed as a link between RA and periodontitis. A common misconception is that RA patients have impaired dexterity and accordingly they are likely to have poorer oral hygiene than healthy control subjects. Although, due to the joint and bone deformities typically affecting the hands of RA patients, it seems logical that these patients would have an impediment to maintain a good oral hygiene (Fig 5-2), studies have consistently failed to show a difference in plaque control that could explain the association with periodontitis15,16,18,28,52. Smoking is also a shared risk factor for both diseases. Patients with established RA who are smokers have a higher prevalence of periodontitis but patients with new-onset RA exhibit a high prevalence of periodontitis at disease onset, despite their low smoking history (16% smokers) and young age (mean age 42.2 years old)53. Although periodontitis and RA share genetic and lifestyle (smoking) risk factors and similar inflammatory pathways, these are not believed to be sufficient to explain this connection16,21,54.
Citrullination or ‘deimination’ is a post-translational modification of proteins in which the amino acid arginine is transformed into citrulline, catalysed by the peptidyle arginine deiminase enzyme (PAD). In the presence of calcium, PAD substitutes the ketamine group (=NH) with a ketone group (=O), converting arginine into citrulline and neutralising the previous positive charge of the amino acid55. This makes the protein hydrophobic, resulting in a change of the protein folding (Fig 5-3). Citrullination occurs as a physiological phenomenon in healthy individuals and it is essential for skin physiology, gene regulation and correct functioning of the immune system56. For instance, PAD2 is essential for the induction of macrophage apoptosis through the citrullination of vimentin or to maintain the normal skin function by citrullinating filaggrin in the process of protein degradation in epidermis cells. Also, when DNA is damaged, histone citrullination by PAD4 suppresses gene expression56,57.
In RA the expression of PAD4 and PAD2 in the synovium is correlated with intensity of inflammation58. These enzymes, during oxidative stress or apoptosis, become active and citrullinate proteins in the joint. While elevation of citrullination has been found in other joint diseases (such as psoriasis, ankylosing spondylitis and OA59,60), the production of auto-antibodies against these epitopes is rather specific to RA. ACPA are found in 80% of patients with RA and are synthesised by plasma cells of synovial tissue to target citrulline residues. These auto-antibodies are highly specific for RA (98%) and have been found in the serum of patients years before they develop any clinical signs, being the earliest and most specific autoantibodies in RA61,62. This unique production of auto-antibodies in RA suggests the necessity for an external factor to trigger this auto-immune reaction. It has also been shown that PAD2 and PAD4 are expressed in inflamed periodontal tissues and this could be one important source of protein citrullination years before RA develops63.
188.8.131.52 Citrullination by Porphyromonas gingivalis
P. gingivalis is one of the bacteria strongly associated with periodontitis and it is considered the key to disruption of the host-microbial homeostasis that causes this disease64–66. A unique characteristic of P. gingivalis is the expression of peptidyl arginine deiminase (PPAD), capable of citrullinating both host and bacterial peptides67. It has been observed in some in vitro studies that PPAD can only citrullinate C-terminal arginine and not internal arginine, in contrast to its homologue human PAD2 and PAD4, creating citrullinated peptides that would not normally occur in the absence of P. gingivalis. To that end, P. gingivalis uses its gingipain enzyme to cleave peptides and generate arginine residues that PPAD then citrullinates68. Also, unlike human PAD, PPAD is not calcium dependant and this enzyme is capable of auto-citrullination, becoming a citrullinated bacterial protein itself69. This process, although it has not been proven to occur in vivo, provides a plausible pathway in which, by presenting neoepitopes to the immune system, P. gingivalis (in susceptible individuals, such HLA-DRB1 [human leucocyte antigen class II histocompatibility antigen, DRB1 beta chain] carriers) would break the tolerance to citrullinated proteins and lead to the subsequent antigen response characteristic in RA patients (Fig 5-4). It is believed that P. gingivalis uses these mechanisms for its survival in the gingival pocket. As a result of protein citrullination, ammonia is generated, which neutralises acidity and energy is released for P. gingivalis to use during anaerobic growth70. This could also act as a strategy to suppress the inflammatory activity of complement component 5a (C5a) by citrullinating it, which reduces its chemottractant nature for neutrophils71.
Numerous researchers have investigated the role of P. gingivalis in RA. The first to suggest that P. gingivalis and PPAD could break the tolerance and trigger an antibody response against citrullinated antigens were Rosenstein et al72. Later, Martinez-Martinez et al73 found DNA from P. gingivalis in synovial fluid of patients with RA, and established that periodontal pathogens could be the trigger for the autoimmune response in RA. Wegner et al67 demonstrated that P. gingivalis citrullinates fibrinogen and alpha-enolase and other human and bacterial peptides. Furthermore, Hitchon et al74 reported an association between an immune responses to P. gingivalis and the presence of ACPA in a population with a high background prevalence of RA predisposing human leucocyte antigen (HLA) alleles. They suggested that this gene-environment interaction might be the cause of the break of tolerance of the immune system to citrullinated proteins, leading to RA.
A study by Mikuls et al75, showed that high levels of anti-P. gingivalis antibodies in RA subjects correlate with levels of anti-CCP antibodies, suggesting that this bacteria plays a role in the risk for, and severity of RA. Interestingly, it has been observed that the periodontium of periodontitis patients with no signs of RA express citrullinated proteins76 and the serum of these patients contains higher levels of antibodies to citrullinated and non-citrullinated human peptides compared to healthy controls77. De Pablo et al77 suggested that the greater citrullination that occurs in periodontitis leads to a loss of tolerance to citrullinated and uncitrullinated peptides that may evolve in a cross-reaction against the citrullinated proteins in the joint of RA patients. Based on these findings, researchers propose that bacterial and human PADs could be a powerful target for therapy in RA78. Moreover, it has been hypothesised that periodontal therapy could decrease the load of P. gingivalis and PPAD in the periodontal pocket and that by reducing gingival inflammation, the expression of human PADs could be reduced, weakening the autoimmune response in RA. Yet, it is still unclear whether there is a causative role of antibodies to citrullinated proteins in the evolution of RA and there is no strong evidence that periodontal therapy reduces citrullination systemically.
5.3.3 Citrullinated human proteins targeted in RA
Anti-CCP assays are regularly used by rheumatologists to measure antibodies against citrullinated proteins and to diagnose patients as ACPA+ and ACPA−, which can be used to anticipate and monitor disease severity and treatment response62. This test uses a synthetic citrullinated cyclic peptide as an antigen to capture antibodies against any citrullinated proteins present in patient sera. Moreover, investigating the proteins that are citrullinated in RA can help us understand the development of the autoimmune response and development of the disease. To date, numerous studies have reported that proteins including fibrinogen, vimentin, enolase and tenascins are targeted in RA and antibodies against these citrullinated proteins are associated with severity of the disease62,79,80.
Fibrinogen is an acute-phase reactant protein, synthesised in the liver and involved in clot formation by its conversion to fibrin by thrombin. Fibrinogen is elevated in numerous inflammatory diseases, such as periodontitis and RA, and antibodies against citrullinated fibrinogen are elevated and correlated to clinical and laboratory parameters in RA81,82. Vimentin is a cytoskeletal component responsible for maintaining mesenchymal cell (mucoid connective tissue) structure. In RA, antibodies against citrullinated vimentin and mutated vimentin (MCV) are used as a diagnostic tool, especially in RF-negative patients83. These antibodies are found before the onset of RA and are also used to estimate bone destruction84. Similarly, in systemically healthy periodontitis patients, antibodies to citrullinated vimentin are elevated compared to non-periodontitis controls77.
Enolase 1, commonly known as α-enolase, is a glycolytic enzyme present in most tissues that acts as a cell surface receptor, tumour suppressor, heat shock protein and transcription co-factor. Antibodies against enolase 1 are elevated in numerous autoimmune diseases such as RA, ulcerative colitis and Crohn disease85. In the joints of patients with RA, enolase is upregulated in response to inflammation. While high levels of anti-enolase antibodies are associated with a variety of autoimmune diseases, antibodies against citrullinated enolase are only found in RA (67% of RA patients, only 3% of healthy individuals)86 and are correlated with disease severity87. Elevated antibodies against citrullinated enolase 1 are higher in CCP-positive and RA patients and those who smoke88 and are also present in early-RA patients89.
Tenascin-C (TNC) is a large extracellular glycoprotein that stimulates inflammation activating toll-like receptor 4 (TLR4). This protein is absent in most healthy tissues and upregulated during inflammation, such as in the RA joint, maintaining tissue damage and preventing resolution of inflammation90. Antibodies against an epitope of citrullinated TNC (cTNC5) have been detected in 40% to 50% of RA patients and are correlated with bone erosion and associated with RA development in early stages of the disease90,91. These results show that both RA and periodontitis induce citrullination, which may be the link between these two diseases. Recently, it has been found that the production of citrullinated proteins in the periodontium is associated with gingival inflammation63, occurring in 80% of periodontitis-affected stroma76. This is supported by the discovery of higher serum antibodies against citrullinated proteins in periodontitis patients compared to controls77. Furthermore, in a 2019 study on first-degree relatives of RA patients, it was found that all ACPA-positive subjects had periodontitis (91.2% diagnosed with moderate to severe periodontitis)92. For this reason, investigators believe that the gingiva of periodontitis patients might be an extra-articular site of citrullinated proteins, which can lead to ACPA production contributing to RA progression.
184.108.40.206 P. gingivalis antigens targeted in RA
It has been proposed that proteins from P. gingivalis could be recognised as an antigen by the host and initiate an immune response that could break tolerance to human antigens such as citrullinated proteins. For this reason, researchers have recently investigated antibodies against some P. gingivalis– antigens and their role in RA. Antibodies against citrullinated PPAD (CPP3 and CPP5) are increased in ACPA+ RA patients as well as in periodontitis patients (CPP5)69. In RA and in patients in the pre-clinical period of the disease (often referred to as “pre-RA” patients), anti-CPP3 antibodies are elevated93, although other authors have been unable to find an association between anti-PPAD antibodies and early RA94. Although only a few studies have investigated these antibodies and more research is needed, considering these results the hypothesis that P. gingivalis-antigens might break tolerance in RA requires further investigation and, if proven, these epitopes could be targeted in RA patients95.
5.3.4 The role of the oral microbiome in RA
The human microbiome is defined as the genes carried by the particular community of microorganisms that live in and on the human body. Using DNA-based technologies to investigate bacteria has opened a new horizon in our understanding of the microbes that live in and on us96, as cultivation techniques are not applicable for most of the complex ecosystems in our body, such as the oral cavity. Currently, five next-generation sequencing studies have investigated the oral microbiome in RA53,97–100. In the metagenome study of Zhang et al97, the authors found that subgingival, oral and gut microbiome are altered and certain species were identified as enriched in RA and therefore possible diagnostic tools (for example, Lactobacillus salivarius and Cryptobacterium curtum). This is supported by the more recent study from Eriksson et al100, who found a significantly different oral microbial composition in RA patients compared to controls. In the study by Scher et al53, in early RA patients some oral bacterial taxa were identified to be significantly different in abundance, but overall measures of diversity were not different to systemically healthy controls.
However, in these studies more than 75% of the new-onset RA and RA patients studied had moderate to severe periodontitis53 or periodontal status was not considered97. Since the oral microbiome is known to be altered in periodontitis, the contradictory results from these two studies might be confounded by the presence of periodontitis. A new study investigating the periodontally healthy oral microbiome in RA found that compared to controls, RA patients have a different microbiome with a significantly higher percentage of anaerobes, including C. curtum98. This is supported by another recent study, which reported that the oral microbiome in patients with RA shows an increased load of periodontal pathogens compared to controls, and that this microbiome shift is associated with worse RA status99. It has also been proposed that, in RA patients, the use of DMARDs and biological therapies restores the dysbiotic oral microbiome97,101 or may at least influence the frequency of certain periodontal pathogens in the subgingival biofilm30. However, the effect of periodontal therapy on the oral microbiome has not been fully investigated using next-generation sequencing techniques and the effect in RA patients is not known.
5.3.5 Cytokine imbalance in RA and periodontitis
In RA, there is an imbalance in the pro-/anti-inflammatory systems102 in a similar manner to that in periodontitis patients103. Understanding the cytokines involved in the pathobiology of RA has led to a revolution in the therapeutics of this disease in the last decade104 and could be one of the key components of the biological link between RA and periodontitis. The human microbiota lives in cross-talk with the human immune system, and bacteria can trigger an immune response to their benefit. In periodontitis, cytokines such as IL-1β can help the growth of some species such as Aggregatibacter actinomycetemcomitans, which in the presence of this cytokine decreases its metabolism and increases its biomass105. Investigating the cytokine profile and correlating it with the microbial communities in RA and periodontitis could help to understand the complex mechanisms that link these two diseases.
The imbalance between pro-inflammatory and anti-inflammatory cytokines and growth factors is comparable in both diseases: In RA and periodontitis high levels of IL-6 and TNF-α are found systemically and locally, and these are known to activate the inflammatory response, and lower levels of IL-10 and transforming growth factor beta are found, which are known to have anti-inflammatory effect106. Of special importance is the pro-inflammatory cytokine TNF, as it plays a key role in the immune response for both RA and periodontitis. TNF increases the inflammatory response by binding to the receptor p55 (TNF receptor type 1; CD120a), whereas binding to receptor p75 (TNF receptor type 2; CD120b) attenuates the inflammatory response107. In periodontitis, TNF has been associated with the breakdown of connective tissue attachment and bone loss108. RA patients suffering with periodontitis have higher levels of TNF compared to non-periodontitis RA patients and these levels are positively correlated with RA severity106.
To date, several studies have evaluated the effect of TNF inhibitors in periodontitis and RA40,109–112. Although the authors concluded that anti-TNF therapy could benefit periodontal status, they suggest that it is very difficult to predict the harm that may result from targeting cytokines. Side effects range from rashes and headache to more severe symptoms such as tuberculosis, heart failure and lymphoma113. In the study by Pers et al109, anti-TNF therapy increased gingival inflammation in RA patients, although it decreased clinical attachment loss. The authors believe that this dichotomy supports the theory that inflammation and bone destruction are two interrelated but separate concepts in periodontitis and in RA. Therefore, it is not clear if anti-TNF therapy is beneficial and safe to treat periodontitis, and more studies are needed in this area. Anti-TNF drugs increase the risk of infection114; therefore, an initial course of periodontal therapy to reduce bacterial load and inflammation followed by anti-TNF therapy could offer improved outcomes in comparison with the use of the drug alone.
Interestingly, periodontal therapy has been shown to reduce serum TNF levels115,116, although other studies did not find this reduction117. For this reason, it is important to investigate periodontal therapy as a possible safe and non-pharmacological treatment that could reduce inflammation in systemic inflammatory conditions such as RA. There are similar arguments for IL-6 in RA and periodontitis. IL-6 is a pro-inflammatory cytokine with numerous functions and anti-IL-6 therapy was shown to reduce periodontal parameters such as bleeding on probing, Gingival Index and clinical attachment loss118. A variety of studies have shown that periodontal treatment reduces serum levels of cytokines like IL-6119–121. In RA patients, IL-6 was reduced in the gingival crevicular fluid after periodontal therapy122. These findings suggest that periodontal therapy could be a non-pharmacological method to reduce the pro-inflammatory cytokines locally and systemically in patients with RA.
5.3.6 Neutrophils and neutrophil extracellular traps
Neutrophils are the first white blood cell type to arrive at the site of infection and their main functions are the recognition and phagocytosis of microorganisms, and activation of humoral immunity. When activated by inflammatory signals such as IL-8, neutrophils release their enzymatic content to the exterior of the cell, together with its DNA, forming a web-like structure with the aim of trapping microbes called neutrophil extracellular traps (NETs). Although effective, the release of these contents, along with reactive oxygen species (ROS) provokes collateral damage to the surrounding tissues, exacerbates the inflammatory response and exposes possible autoantigens123,124.
Aberrant NET release has been observed in RA synovial fluid and serum compared to healthy controls and osteoarthritis, and NET release was correlated with high ACPA and RF levels125. Interestingly, in RA patients, IL-17, TNF-α and ACPA such as anti-citrullinated vimentin induced NET formation. In turn, NETs stimulated the expression of pro-inflammatory genes in fibroblast-like cells, which may amplify inflammation in the synovium of RA patients125.
It has been shown that PAD4 activation is necessary for NET production. In Pad4-deficient mice, neutrophils did not undergo NET formation, compared to the control group126. Also in this study, it was noted that that histone citrullination by PAD4 leads to chromatin decondensation, facilitating the production of NETs126. These results suggest that in RA, NETs contribute to the maintaining of the inflammatory response and become a source of citrullinated antigens in RA. Recent work has demonstrated a significantly reduced ability of plasma from periodontitis patients to degrade NETs, due to lower levels of DNAses within serum/plasma of periodontitis patients; this may offer an explanation based upon increased NET retention in periodontitis patients, which may in turn impact on the development of a ‘cirtullinome’ within periodontal tissues127.
Elevated NET formation also occurs in periodontitis patients. Although not yet proven, researchers believe that this might be due to chronic exposure to periodontal pathogens and their components/metabolites128. Moreover, NETs and neutrophils were identified recently in synovial fluid from RA patients, along with active PAD enzymes, which appeared both free and NET-bound129. The authors demonstrated a clear correlation between levels of DNA in the joints, serum ACPA level and also neutrophil counts, as well as a clear mapping of citrullinated proteins to neutrophils within cell pellets prepared from inflamed joints. Taken together, these results suggest that, in periodontitis, a chronic exposure of gingival tissues to PAD4 and citrullination may lead to breakdown of tolerance against citrullinated peptides in a susceptible individual, leading or contributing to RA pathogenesis. NET generation could not only contribute to periodontal and RA pathogenesis, but also provide a potential causal link between these two conditions (Fig 5-5)127.
5.3.7 Genetic background
RA and periodontitis have certain common genetic factors impacting upon the hosts’ immune response that translate into, for example, higher levels of pro-inflammatory cytokine production. The HLA-DR4 epitope is located on the surface of leucocytes and has been found to be associated with both RA and periodontitis130. In common with several other immune-mediated diseases, numerous studies report an association between HLA genes and periodontitis. The most recent systematic review reported a protective association with HLA-A2 and B5 and an increased susceptibility for periodontal disease with HLA-A9 and B15 genotypes131. HLA phenotypes and their association with periodontitis and autoimmune diseases are explored in Section 220.127.116.11 ‘HLA phenotypes and major histocompatibility complex’. Whilst classic twin studies have shown that periodontitis exhibits a genetic component132, genome-wide association studies (GWAS) have been unable to identify significant associations with chronic periodontitis133. A recent systematic review of 43 studies concluded that there was no evidence for an interaction between any genetic variants (including IL1) and the subgingival microflora134. Therefore, whilst genetics plays a clear role in aggressive forms of periodontitis (Grade C periodontitis), its importance in chronic periodontitis remains unclear.
In RA, robust data are available from genetic association studies, suggesting HLA SE (shared epitope) alleles as the main candidate gene in ACPA-positive patients135. In addition to these alleles, GWAS have revealed another 32 loci associated with RA136. There is also a clear dissimilarity between genetic risk factors for ACPA-positive and ACPA-negative RA, thus more research is required to understand the genetic differences between these two groups137. There are currently few studies investigating a genetic link between RA and periodontitis138–140 and no GWAS studies. One recent 2018 study demonstrated that a polymorphism of the KCNQ1 gene, which is required for potassium channel assembly, was significantly associated with comorbidity of RA and periodontitis in a Japanese cohort after adjustment for age, sex and smoking status141. The carriers of the allele among patients with RA and periodontitis showed significantly higher DAS28 scores based on C-reactive protein values than the non-carriers141. In a 2019 study comparing periodontitis patients with or without RA, it was reported that interferon gamma gene polymorphism may be marker of RA and periodontitis comorbidities142. In another 2019 study, authors used text-mining through PubMed to search for genes associated with RA and periodontitis. They found more than 500 common genes and concluded that more research is needed in this area143. Overall, however, the importance of genetic polymorphisms as a possible link between these two diseases remains largely unknown.
5.3.8 ‘Two-hit’ model
Golub et al144 were the first to describe a theory in which the association between periodontitis and systemic diseases could be explained by a ‘two-hit’ model. In this theory, a first ‘hit’ occurs within the periodontum, initiated by an infection that activates a destructive inflammatory cascade in the periodontal tissues. In susceptible patients, a second systemic ‘hit’ then occurs, characterised by increased serum levels of pro-inflammatory cytokines that amplifies the inflammatory cascade, with production of local and systemic pro-inflammatory mediators (cytokines and prostaglandins) being released. These cytokines, particularly IL-1 and IL-6, activate the expression of receptor activator of nuclear factor-kappa B ligand (RANKL) in osteoblasts, which binds to RANK from osteoclasts, activating their differentiation and survival, and leading to bone resorption144. Interestingly, peripheral blood neutrophils from periodontitis patients have recently been shown to be hyper-reactive in respect of cytokine release (IL-1β, IL-8, TNF-α)145.
Six years later, a similar ‘two hit’ model was described to explain the breakdown of tolerance to citrullinated proteins in RA patients triggered by smoking. Prolonged cigarette smoking leads to a chronic inflammatory response in the lungs that upregulates pulmonary PAD expression, increasing citrullination. Loss of immune tolerance in susceptible individuals would result in ACPA production in an extra-articular ‘first hit’. In a second hit that occurs some years later, ACPA are released systemically, targeting the joint structure as the synovium is especially sensitive to inflammatory stimuli146. However, it is still not clear why this autoimmune reaction to citrullinated peptides occurs in RA specifically, and not in other conditions.
Similarly to smoking, periodontitis could represent a first hit of citrullinated peptides. In a first extra-articular hit, chronic periodontitis could break the tolerance to citrullinated proteins, due to abnormal and bacterial citrullination by P. gingivalis. Through epitope spreading (a process in which the immune response does not remain fixed towards a specific epitope, but extends to include other epitopes on the same protein or other proteins in the same tissue), this local autoimmune response to citrullinated proteins could lead to the production of ACPA systemically. These ACPA react against citrullinated proteins in the joint since, as explained in Section 5.3.2 ‘Citrullination’, in joint diseases there is an increase of protein citrullination in the synovium. This model describes a primary ‘hit’ of ACPA production due to chronic periodontitis followed by a secondary ‘hit’ in the joint, that could induce RA52,147 (Fig 5-6).
● Different theories connecting immunological and microbiological aspects of RA and periodontal disease exist. These have not yet elucidated the relationship and require more research.
● The presence of certain HLA antigen polymorphisms may be an influencing factor in both RA and periodontitis; however, their importance as a link between these two diseases remains unknown.
● It is plausible that widespread hyper-citrullination due to NET release within periodontal tissues acts as a first hit in a ‘two-hit’ model of RA pathogenesis, alongside P. gingivalis PAD citrullination.