Periodontitis and periodontopathic bacteria as risk factors for rheumatoid arthritis: A review of the last 10 years

Abstract

Rheumatoid arthritis (RA) is characterized by chronic inflammatory destruction of joint tissue and is caused by an abnormal autoimmune response triggered by interactions between genetics, environmental factors, and epigenetic and posttranslational modifications. RA has been suggested to be interrelated with periodontitis, a serious form or stage of chronic inflammatory periodontal disease associated with periodontopathic bacterial infections, genetic predisposition, environmental factors, and epigenetic influences. Over the last decade, a number of animal and clinical studies have been conducted to assess whether or not periodontitis and associated periodontopathic bacteria constitute risk factors for RA. The present review introduces recent accumulating evidence to support the associations of periodontitis and periodontopathic bacteria with the risk of RA or the outcome of RA pharmacological treatment with disease-modifying antirheumatic drugs. In addition, the results from intervention studies have suggested an improvement in RA clinical parameters after nonsurgical periodontal treatment. Furthermore, the potential causal mechanisms underlying the link between periodontitis and periodontopathic bacteria and RA are summarized.

1. Introduction

Rheumatoid arthritis (RA) is a systemic autoimmune disease characterized by chronic inflammatory destruction of joint tissue and subsequent functional limitation [1]. The global prevalence of RA was estimated to be approximately 0.5% in a recent systematic review with a meta-analysis [2]. The etiology of RA has not been fully elucidated, although it is a pathologically heterogeneous disorder caused by an abnormal autoimmune response triggered by the interactions between genetic and environmental factors and epigenetic and posttranslational modifications, which can lead to loss of immune tolerance [1], [3].

For many years, RA has been suggested to be clinically and pathologically interrelated with periodontitis [4], [5], a serious form or stage of chronic inflammatory periodontal disease associated with periodontopathic bacterial infections and accompanied by irreversible inflammation-mediated destruction of periodontal tissue [6], [7]. Notwithstanding these early observations, a workshop jointly held in 2012 by the European Federation of Periodontology (EFP) and American Academy of Periodontology (AAP) concluded that there was insufficient evidence to infer an association between RA and periodontitis [8]. Given the record of discussions at this workshop, an increasing number of studies have been conducted to further elucidate the relationship between these two inflammatory diseases, with the findings partially summarized by comprehensive reviews over the 2012–2017 and 2018–2019 period [5], [9]. It is now generally accepted that there is sufficient evidence to conclude that there is a significant and clinically relevant association between RA and periodontitis.

Several plausible causal mechanisms have also been proposed to account for the underlying relationship between periodontitis and RA [4], [5], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23], [24], [25], [26], [27], [28], [29], [30], [31], [32], [33], [34], [35], [36], including the notion that periodontitis and periodontopathic bacteria can constitute risk factors for RA, that RA may predispose patients to periodontitis, and that both diseases share common risk factors [37]. Of these, much attention has been paid to the role of periodontitis and periodontopathic bacteria in the pathogenesis of RA. Porphyromonas gingivalis and periodontitis-affected gingival tissue have been suggested to trigger a citrulline-specific autoimmune response characterized by an antibody response to citrullinated proteins. These citrullinated proteins and their related antibodies have been detected in the blood and joint of RA patients as well as in the inflamed gingiva of periodontitis patients [4], [5], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23], [24], [25], [26], [27], [28], [29], [30], [31], [32], [33], [34], [35], [36].

Therefore, the present review focuses on the evidence that has been published over the last 10 years, including the associations of periodontitis and periodontopathic bacteria with either the risk of RA or the outcome of RA pharmacological treatment, as well as the effect of periodontal treatment on RA. Furthermore, the potential causal mechanisms underlying the association between periodontitis and periodontopathic bacteria and RA are summarized.

2. Publications 2013–2022

Since 2013, a total of 161 papers have been published to assess the association of periodontitis and periodontopathic bacteria with RA, including 35 reviews and 126 original research articles. These research articles are 22 animal pathogenesis studies excluding in vitro studies, 13 clinical cohort and case-control studies, 63 clinical pathogenesis studies, 8 intervention studies assessing the effect of periodontitis and periodontopathic bacteria on the response to RA pharmacological treatment, and 20 intervention studies evaluating the effect of periodontal treatment on RA. The papers published between 2013 and 2022 covered in the present review are summarized in Table 1.

Table 1.

Summary of publications published between 2013 and 2022 evaluating the effect of periodontitis and periodontopathic bacteria on RA.

Publication type-research topic	Number of publications
Review articles	35
Original research articles	126
-Animal pathogenesis studies	22
-Clinical cohort and case-control studies	13
-Clinical pathogenesis studies	63
-Effect of periodontitis and periodontopathic bacteria on RA pharmacological treatment	8
-Effect of periodontal treatment on RA	20

2.1. Animal pathogenesis studies

Twenty-two animal studies were conducted between 2013 and 2022 to study the role of periodontitis and periodontopathic bacteria in the pathogenesis of RA in a reasonably well-controlled environment using DBA/1, C57BL/6, BALB/c, B10.RIII, SKG, F1, and B6. DR1 mice or DA, Wistar, Lewis, and Sprague-Dawley rats [38], [39], [40], [41], [42], [43], [44], [45], [46], [47], [48], [49], [50], [51], [52], [53], [54], [55], [56], [57], [58], [59]. Most of these animal studies demonstrated the involvement of P. gingivalis infection in the induction and progression of experimental arthritis. This appeared to be related to increased levels of protein citrullination through peptidylarginine deiminase (PAD) from P. gingivalis (PPAD) [38], [41], [45], [49], [53], [57] and increased levels of T helper 17 (Th17) cell differentiation and subsequent inflammatory responses and induction of osteoclastogenesis [39], [40], [42], [43], [44], [48]. In addition, an alteration of gut barrier, microbiota, and immune profile toward Th17 was noted [47], [56], [58]. The association between P. gingivalis infection and experimental arthritis was also explained by fimbrilin (FimA)-mediated adhesion to immune cells, activation of complement C5a generated by arginine-specific gingipain (Rgp), followed by an overexpression of proinflammatory cytokines as well as a bacterial migration to the joints [50], [52]. This association was further confirmed by the observation that the severity of experimental arthritis was reduced after preimmunization against FimA and Rgp type A (RgpA) [50], [59] or use of a mouthwash containing chlorhexidine (CHX) and metronidazole [55]. The severity of P. gingivalis-induced experimental arthritis was also shown to be influenced by coinfection with other periodontopathic bacteria such as, Aggregatibacter actinomycetemcomitans and Fusobacterium nucleatum, suggesting interplay among these bacteria [51]. However, these observations obtained in the animal model should be interpreted with caution, as they do not completely reflect the complexities of human diseases for RA and periodontitis.

2.2. Clinical cohort and case-control studies

For the period 2013–2022 thirteen clinical cohort and case-control studies investigating whether periodontitis is associated with the risk of onset and progression of RA in patients were published [60], [61], [62], [63], [64], [65], [66], [67], [68], [69], [70], [71], [72]. These studies were conducted in a very wide geographic spread of countries, including Finland, Germany, Indonesia, Iran, Korea, Spain, Sweden, Taiwan, and Thailand, and included cohort sizes ranging from 44 to 797,470 individuals. Of the 13 studies, 8 indicated that the history, presence, or severity of periodontitis was associated with an increased risk of RA in patients [62], [63], [65], [66], [68], [70], [71], [72], and 1 study reported similar results in nonsmoking patients [60]. Although four studies failed to show an association between periodontitis and the risk of RA [61], [64], [67], [69], the majority (69%) of these studies confirmed the statistical significance in this association.

Since 2013, three systematic reviews with meta-analyses have been published analysing the clinical cohort and case-control studies. All of these concluded that periodontitis represents a risk factor for RA [10], [25], [30]. The first review of 19 studies published up to June 2012 showed that an increased clinical attachment level (CAL) and tooth loss were significantly associated with RA (odds ratio [OR]: 1.17 for CAL; 2.38 for tooth loss) [10]. Similar results were also reported (OR of 1.97) in another systematic review with a meta-analysis of 3 articles, which should be interpreted with consideration of the mediators shared between both diseases [25]. The most recent systematic review and meta-analysis study included 706,611 periodontitis patients and 349,983 control subjects in 13 articles published up to September 2019 and demonstrated that periodontitis patients had a 69% greater risk for RA than control individuals, with an OR of 1.69 [30]. These observations suggest that there is now a need to clarify the causal mechanisms linking periodontitis and RA, rather than a need for further clinical cohort and case-control studies.

2.3. Clinical pathogenesis studies

Sixty-three clinical studies were conducted between 2013 and 2022 to assess the role of periodontitis and/or periodontopathic bacteria in the pathogenesis of RA [73], [74], [75], [76], [77], [78], [79], [80], [81], [82], [83], [84], [85], [86], [87], [88], [89], [90], [91], [92], [93], [94], [95], [96], [97], [98], [99], [100], [101], [102], [103], [104], [105], [106], [107], [108], [109], [110], [111], [112], [113], [114], [115], [116], [117], [118], [119], [120], [121], [122], [123], [124], [125], [126], [127], [128], [129], [130], [131], [132], [133], [134], [135]. Of these studies, 30 considered periodontitis and of these, 28 reported a significantly positive association between periodontitis and RA [73], [76], [79], [81], [83], [84], [85], [88], [90], [94], [98], [103], [104], [107], [109], [110], [113], [115], [116], [118], [119], [120], [121], [122], [124], [126], [129], [135] but not in 2 [89], [132]. Most studies showing the positive association were related to increased levels of anti-citrullinated protein antibodies (ACPAs) through PPAD [79], [94], [98] or endogenous PAD induced by periodontal inflammation [73], [94], [116] and neutrophil extracellular traps (NETs) [88], [104], [110]. Other posttranslational protein modifications, carbamylated protein [85], [107], [110], [126], and malondialdehyde-acetaldehyde adduct formation [107] were observed in human inflamed gingiva but not in healthy or noninflamed gingiva, which may partially explain how periodontitis affects the development of RA. The significant association between the two diseases may also be partially explained by increased levels of host inflammatory mediators, including interleukin (IL)-6, matrix metalloproteinase-8, receptor activator of nuclear factor-kappa B ligand, and a proliferation-inducing ligand [109], [113], [115], [119], [121], [135], and of autoantibodies to agalactosyl immunoglobulin G (IgG) [129].

Of the 63 clinical pathogenesis studies, 38 confirmed a significantly positive association between periodontopathic bacterial infection and the clinical and biochemical measures of RA in the patients [74], [75], [77], [78], [79], [82], [86], [87], [91], [93], [94], [95], [96], [97], [98], [99], [100], [101], [102], [103], [105], [106], [108], [111], [112], [114], [117], [118], [121], [122], [123], [125], [127], [128], [130], [133], [134], [135], while 8 studies found no evidence or equivocal results [80], [84], [89], [90], [92], [116], [119], [131]. Most of these studies showed a significant role of P. gingivalis infection in the pathogenesis of RA, which includes the detection of P. gingivalis DNA in the sera and synovial fluid and tissue [74], [75], [106], the increased levels of ACPAs mediated by PPAD [77], [79], [91], [93], [94], [98], [122], [123], [125], [133] in combination with Rgp in patients [96], [100] or in nonsmoking patients [82], and by the molecular mimicry underlying the homology between P. gingivalis enolase and human α-enolase [95], [123]. Another explanation for the significant effect of P. gingivalis infection on RA may be partially related to the increased prevalence of highly toxic fimA genotype II [114], to the impaired P. gingivalis-induced tumor necrosis factor (TNF) production by dendritic cells and subsequent prolonged bacterial survival [97], or to the alteration of the gut microbiota and immune system [111], [134].

In addition to P. gingivalis, the following periodontopathic or periodontitis-related bacteria have been implicated in the association with RA: A. actinomycetemcomitans [101]; Prevotella intermedia [87], [99], [106], [108]; F. nucleatum [102], [118], [135]; Capnocytophaga ochracea [99]; Tannerella forsythia [105], [106], [127]; Cryptobacterium curtum [112]; Leptotrichia spp. [127]; Megasphaera spp. [127], [135]; Anaeroglobus geminatus [127]; Veillonella spp, [130], [135]; Granulicatella spp. [135]. In particular, A. actinomycetemcomitans has been shown to induce cellular hypercitrullination in neutrophils by activation of PAD-4 through its pore-forming toxin leukotoxin-A (LtxA) [101], while C. curtum has been reported to have the ability to produce citrulline via the PAD pathway [112]. However, most other studies only described the antibody responses to these bacteria in relation to RA and not the causal mechanism underlying the association.

Since 2013, four systematic reviews have been published concerning the role of periodontopathic bacteria in RA, all of which concluded that P. gingivalis infection constitutes a risk factor for RA [19], [24], [32], [35]. Of these, one systematic review with a meta-analysis of 13 studies published up to December 2015 revealed that RA patients exhibited a significantly higher antibody response to P. gingivalis than systemically healthy individuals [19]. Another more recent systematic review with a meta-analysis of 28 studies published up to April 2022 showed a significant increase in the risk of RA in individuals with P. gingivalis exposure (OR of 1.86) [35]. These observations suggest that determination of the serum antibody response to P. gingivalis and subsequent antimicrobial periodontal treatment may help prevent the development of RA. However, there is a need to further elucidate the mechanism linking periodontal inflammation, P. gingivalis infection, and RA in a larger-scale patient cohort.

2.4. Effect of periodontitis and periodontopathic bacteria on RA pharmacological treatment

Pharmacological treatment for RA includes the use of conventional synthetic disease-modifying antirheumatic drugs (csDMARDs), biological DMARDs (bDMARDs), targeted synthetic DMARDs (Janus kinase inhibitors), or corticosteroids [1], [136]. In particular, the therapeutic beneficial effects of bDMARDs, including inhibitors of TNF and IL-6 receptor (IL-6R), have been demonstrated in RA patients who are refractory to treatments with csDMARDs and corticosteroids [1], [136]. Periodontitis and periodontopathic bacteria have been suggested to trigger an autoimmune inflammatory response in relation to RA [4], [5], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23], [24], [25], [26], [27], [28], [29], [30], [31], [32], [33], [34], [35], [36], which may be partially mediated by autoantibodies to citrullinated proteins through PPAD [77], [79], [91], [93], [94], [98], [122], [123], [125], [133] and endogenous PAD [73], [88], [94], [104], [110], [116] in periodontitis patients. These observations led to the hypothesis that periodontal inflammation and associated periodontopathic bacterial infection may affect the responses of RA patients to pharmacological treatment. To date, special attention has been focused on identifying periodontitis-related predictors of the response in patients before bDMARD therapy due to the risk of adverse effects and considerable costs.

Since 2013, eight original intervention studies have been conducted to assess the effect of periodontitis and periodontopathic bacterial infection on the response to pharmacological treatment with DMARDs in RA patients [78], [137], [138], [139], [140], [141], [142], [143]. A summary of the study method and outcome in each study is presented in Table 2. In these studies, the number of patients treated ranged from 18 to 111, while the follow-up period ranged from 3 to 12 months. Likewise, the RA treatment tested varied from study to study: bDMARDs; csDMARDs; or a combination of csDMARDs and bDMARDs. As an outcome of RA treatment, the Disease Activity Score in 28 joints (DAS28) [144] or Clinical Disease Activity Index (CDAI) [145] was reported in 5 and 3 studies, respectively, as both the C-reactive protein (CRP) level and erythrocyte sedimentation rate (ESR) dropped to negative values in RA patients treated with IL-6R inhibitors [146].

Table 2.

Summary of effect of periodontitis and periodontopathic bacteria on response to RA pharmacological treatment.

Predictors	Patients	RA treatment	Follow-up	Outcome	Study
Periodontal disease	8 with/10 without periodontal disease	bDMARDs (anti-TNF agents)	6 months	Significant improvement in DAS28, ESR, and CRP in RA patients without periodontal disease	Savioli et al.[137]
Periodontitis	84 with/27 without periodontitis	bDMARDs	3 months	Failure to demonstrate a significant association between periodontitis and inadequate disease control in RA patients	Chen et al.[138]
PISA	27 high/27 low PISA	bDMARDs (TNF and IL-6R inhibitors)	6 months	Significantly greater improvement in CDAI in RA patients with low PISA	Yamashita et al.[139]
Periodontitis severity	22 moderate to severe/28 no or mild periodontitis	bDMARDs (TNF and IL-6R inhibitors)	1 year	Significantly greater improvement in CDAI in RA patients with no or mild periodontitis	Kobayashi et al.[140]
Serum anti-P. gingivalis IgG	16 positive/22 negative anti-P. gingivalis IgG	DMARDs ( MTX, or TNF inhibitors)	1 year	Significantly higher in ESR and DAS28-ESR in RA patients with positive anti-P. gingivalis IgG	Arvikar et al.[78]
Serum anti-P. gingivalis IgG	25 high/25 low anti-P. gingivalis IgG	bDMARDs (TNF and IL-6R inhibitors)	1 year	Significantly greater improvement in CDAI in RA patients with low anti-P. gingivalis IgG	Kobayashi et al.[141]
Serum anti-P. gingivalis IgG	37 good/47 poor DAS28-ESR responder 43 good/41 poor DAS28-CRP responder	csDMARDs	3 months	Significant association between anti-P. gingivalis IgG and treatment responses defined by DAS28-ESR and -CRP	Takeuchi-Hatanaka et al.[142]
Serum anti-PPAD IgG	30 high/30 low anti-PPAD IgG	bDMARDs (TNF and IL-6R inhibitors)	6 months	Significantly greater improvement in DAS28-CRP and anti-CCP IgG in RA patients with low anti-PPAD IgG	Kobayashi et al.[143]

As shown in Table 2, the first comparative study evaluated the effect of 6-months of treatment with anti-TNF agents on the rheumatologic changes in 8 and 10 RA patients with and without periodontal disease, respectively. The anti-TNF treatment consisted of 15 cases of infliximab (IFX; a chimeric mouse/human anti-TNF-α monoclonal antibody), 2 cases of adalimumab (ADA; a fully humanized anti-TNF-α monoclonal antibody), and 1 case of etanercept (ETN; a recombinant fusion protein linked to human type II TNF receptor-Fc portion). The results showed a significant improvement in the DAS28, ESR, and CRP level in RA patients without periodontal disease, but not in those with periodontal disease [137]. However, another study failed to demonstrate a significant association between periodontitis and the risk of inadequate disease control in 111 RA patients [138]. The periodontal inflamed surface area (PISA), which is a measure of the inflammatory burden posed by periodontitis [147], was studied for its effect on CDAI changes in 54 RA patients who received 6-months of treatment with bDMARDs. The results showed that 27 patients with a low PISA (low level inflammation) had significantly greater decreases in the CDAI than 27 patients with a high PISA (high level inflammation). The baseline PISA score was also significantly positively correlated with changes in the CDAI [139]. A retrospective cohort study was further conducted to assess whether or not periodontitis severity as classified by the Centers for Disease Control Prevention (CDC)/ AAP case definitions [148] affected the response to 1-year treatment with bDMARDs in 50 RA patients. The results showed significantly greater decreases in the CDAI in 28 patients with no or mild periodontitis than in 22 patients with moderate and severe periodontitis and a significantly positive association between the baseline CDC/AAP definitions and changes in the CDAI [140]. These results suggest that RA patients who have no or mild periodontitis and a low PISA may exhibit a better clinical response to bDMARD therapy than those with a moderate to severe periodontitis and a high PISA.

As another periodontitis-related predictor, serum antibody responses to P. gingivalis have been studied to assess their effects on rheumatologic changes after treatment with DMARDs in RA patients. The rheumatologic changes were compared between early RA patients with and without serum anti-P. gingivalis IgG after 12-months of treatment with DMARDs, usually csDMARDs (such as methotrexate) and in some cases bDMARDs. The results indicated a significantly higher ESR and a trend toward a higher DAS28-ESR in 16 RA patients positive for anti-P. gingivalis IgG than in 22 RA patients without anti-P. gingivalis IgG [78]. Similar results were obtained in recent studies [141], [142], and both concluded that baseline serum anti-P. gingivalis IgG, but not anti-A. actinomycetemcomitans IgG, was predictive of the response to treatment with bDMARDs and csDMARDs. However, these are serum IgG responses to sonicated preparations of P. gingivalis, and their target antigens are unclear. Thus far, there has been one study reporting a significantly greater improvement in DAS28-CRP and anti-cyclic citrullinated peptide (CCP) IgG levels in RA patients with low anti-PPAD IgG than in those with high anti-PPAD IgG after 6-months of treatment with bDMARDs [143]. Taken together, these results suggest that low baseline anti-P. gingivalis IgG responses are associated with a better clinical response to treatment with bDMARDs in RA patients. However, further studies are needed to elucidate the mechanism linking baseline serum immunity against P. gingivalis and the clinical response to bDMARDs in randomized clinical trials with a larger-scale patient cohort.

2.5. Effect of periodontal treatment on RA

Nonsurgical periodontal treatment is a major common procedure for periodontal therapy and is performed by mechanical debridement of the microbial biofilm to reduce the microbial burden and periodontal inflammation. This treatment includes oral hygiene instruction (OHI), scaling and root planing (SRP), and if necessary, adjunctive antimicrobials via a systemic or local route [149]. Given the increasing evidence that periodontitis and periodontopathic bacteria trigger the autoimmune response, it is plausible that periodontal treatment may have a beneficial effect on RA.

Since 2013, a total of 20 original intervention studies have been conducted to assess the effect of nonsurgical periodontal treatment on RA measures in patients [150], [151], [152], [153], [154], [155], [156], [157], [158], [159], [160], [161], [162], [163], [164], [165], [166], [167], [168], [169]. A summary of the study method and outcome in each RA measure is presented in Table 3. In these studies, the number of patients treated ranged from 10 to 60, while the follow-up period ranged from 1 to 12 months. The periodontal treatment tested also varied from study to study: a combination of OHI plus SRP; OHI plus scaling; or OHI plus SRP with use of CHX and essential oil (EO) mouthwash or with administration of antimicrobials amoxicillin (AMX). As a clinical measure of RA, the DAS28 or CDAI was reported in 17 studies. Among these studies, nine showed a significant reduction, as indicated by “↓” [150], [151], [152], [155], [157], [159], [162], [165], [166] and six indicated a trend toward a reduction, as indicated by “↘” [158], [161], [163], [164], [168], [169] in the DAS28 or CDAI following nonsurgical periodontal treatment. In addition, as biochemical measures of RA, the CRP level and ESR were evaluated in 16 and 14 studies, respectively. Among these studies, five showed a significant reduction [150], [155], [157], [160], [168] and six showed a trend toward a reduction [151], [153], [154], [158], [159], [169] in the CRP level, while four showed a significant reduction [150], [155], [157], [160] and four indicated a trend toward a reduction [151], [153], [166], [169] in the ESR. Other RA biochemical measures, including rheumatoid factor (RF) and ACPA were assessed in nine and six studies, respectively. Of these, one study showed a significant reduction [160] and one indicated a trend toward a reduction [151] in RF, while two studies showed a significant reduction [157], [160] and two indicated a trend toward a reduction [166], [168] in ACPA.

Table 3.

Summary of effect of periodontal treatment on RA measures.

Study	RA patients	Periodontal treatment	Follow-up	DAS28 or SDAI change	CRP change	ESR change	RF change	ACPA change
Erciyas et al.[150] Bıyıkoğlu et al.[151] Okada et al.[152] Roman-Torres et al.[153] Kurgan et al.[154] Khare et al.[155] Kurgan et al.[156] Zhao et al.[157] Cosgarea et al.[158] Kaushal et al.[159] Anusha et al.[160] Mariette et al.[161] Białowąs et al.[162] Buwembo et al.[163] Monsarrat et al.[164] Moura et al.[165] Nguyen et al.[166] Elsadek et al.[167] Ding et al.[168] De Pablo et al.[169]	30 mild/30 moderate-high activity 10 test 26 test, 29 control 12 test 14 gingivitis/13 periodontitis 30 test, 30 control 15 test 18 test 15 test 20 test, 20 control 15/15/15 test 41 test, 32 control 22 test 32 test, 26 control 11 test, 11 control 24 test, 23 control 38 test, 38 control 25 test 28 test 23 test, 26 control	OHI + SRP OHI + SRP OHI + Scaling OHI + SRP OHI + Scaling/SRP OHI + SRP OHI + SRP OHI + SRP OHI + SRP OHI + SRP OHI + SRP + CHX/EO/no OHI + Scaling OHI + SRP OHI + SRP OHI + SRP + AMX OHI + SRP OHI + SRP OHI + SRP SRP OHI + Scaling	3 months 6 months 8 weeks 3 months 3 months 3 months 3 months 1 month 6 months 8 weeks 6 weeks 1 year 4–6 weeks 6 months 3 months 45 days 6 months 12 weeks 6 weeks 6 months	↓ ↓ ↓ Not reported → ↓ → ↓ ↘ ↓ Not reported ↘ ↓ ↘ ↘ ↓ ↓ Not reported ↘ ↘	↓ ↘ → ↘ ↘ ↓ → ↓ ↘ ↘ ↓ Not reported → Not reported → Not reported → Not reported ↓ ↘	↓ ↘ Not reported ↘ → ↓ → ↓ → Not reported ↓ Not reported → Not reported → Not reported ↘ Not reported → ↘	Not reported ↘ → Not reported → Not reported Not reported Not reported → → ↓ Not reported Not reported Not reported Not reported Not reported → → → Not reported	Not reported Not reported → Not reported Not reported Not reported Not reported ↓ Not reported → ↓ Not reported Not reported Not reported Not reported Not reported ↘ Not reported ↘ Not reported

Five systematic reviews covering the effect of periodontal treatment on RA measures have been conducted since 2013. These studies have generally concluded that nonsurgical periodontal treatment might improve the rheumatologic condition in patients with RA and periodontitis [170], [171], [172], [173], [174]. Of these, a systematic review with a meta-analysis of five studies demonstrated that nonsurgical periodontal treatment led to a significant reduction in the ESR and a trend toward a reduction in the DAS28 [170]. Similar results were also reported in another systematic review with a meta-analysis [172]. These observations suggest the importance of periodontal health in the control of rheumatologic conditions in RA patients. However, there are two studies showing no impact of periodontal treatment on RA disease activity [154], [156]. These inconsistent results might be explained by the fact that all studies in the systematic reviews had a relatively low number of patients, a relatively short follow-up period, and a limited number of randomized clinical trials in relation to selection bias, requiring the data to be interpreted with caution. Therefore, it is necessary to conduct randomized clinical trials with a large number of patients and a longer follow-up period to fully elucidate the effect of nonsurgical periodontal treatment on RA, with consideration of RA medication, as proposed previously [175].

3. Causal mechanisms of periodontitis and periodontopathic bacteria-induced RA

After analysis of the results of animal and human clinical pathogenesis studies, proposed mechanisms by which periodontitis and periodontopathic bacteria may contribute to the development of RA are summarized and illustrated in Fig. 1. The most widely regarded mechanism is protein citrullination and subsequent ACPA production accelerated by the increased expression of NETs [88], [104], [110], PAD-2, and PAD-4 [73], [94], [116] in the periodontium of periodontitis patients. Protein citrullination can also be induced by LtxA from A. actinomycetemcomitans [101], by bacterial PAD from P. gingivalis [38], [41], [45], [49], [53], [57], [77], [79], [91], [93], [94], [98], [122], [123], [125], [133] and possibly C. curtum [112], or by the molecular mimicry underlying the homology between P. gingivalis enolase and human α-enolase [95], [123]. These circulating ACPAs spread from the inflamed periodontium to the synovial joint and react with citrullinated protein mediated by overexpressed NETs [176], [177], PAD-2 and PAD-4 [178], leading to joint inflammation and the development of RA. Periodontitis-induced NETs also produce carbamylated protein (CarP) and anti-CarP antibody [85], [107], [110], [126], reacting with CarP, followed by joint inflammation and the development of RA. P. gingivalis infection promotes Th17 cell development and the Th17-driven response of cytokines and chemokines [39], [40], [42], [43], [44], [48], generates active complement C5a by Rgp [52], and causes FimA-mediated adhesion to immune cells and subsequent bacterial translocation to the joints [50] and lipopolysaccharide (LPS)-induced cytokine synthesis [111]. Oral infection with P. gingivalis showing the influx of LPS into the gut can cause intestinal dysbiosis, leading to a shift in the immune profile toward Th17 [47], [56], [58], [111], [134].

Fig. 1.

Proposed causal mechanisms of periodontitis and periodontopathic bacteria-induced RA. Protein citrullination and subsequent ACPA production are accelerated by periodontitis-induced NETs and endogenous PAD-2 and PAD-4, by A. actinomycetemcomitans LtxA and P. gingivalis (and possibly C. curtum) PAD, or by the molecular mimicry underlying the homology between P. gingivalis and human enolase. These circulating ACPAs react with citrullinated proteins in the joints, leading to joint inflammation. Periodontitis-induced NETs also produce CarP and anti-CarP antibodies, reacting with CarP followed by joint inflammation. P. gingivalis infection promotes Rgp-, FimA-, or LPS-induced synthesis of cytokines, chemokines, or active complement and causes gut dysbiosis in relation to Th17 dominance.

4. Conclusions

The present review has focused on the literature published between 2013 and 2022 assessing whether or not periodontitis and periodontopathic bacteria are risk factors for RA. The results of most clinical studies indicate that individuals with periodontitis and P. gingivalis infection are at risk of RA, as confirmed by observations in animal studies. The presence and severity of periodontitis and P. gingivalis infection seem to reduce the efficacy of pharmacological treatment with DMARDs in RA patients. In addition, nonsurgical periodontal treatment appears to be effective in controlling RA conditions through the reduction in the ESR and possibly in DAS28, despite a few reports of equivocal results. These observations suggest the importance of periodontal health and the need for implementing the prevention and early treatment of periodontitis in RA patients. However, how periodontitis and P. gingivalis and whether or not other periodontopathic bacteria contribute to the pathogenesis of RA remains unclear. There is also a need to clarify the mechanisms by which nonsurgical periodontal treatment improves the RA condition. It is thus necessary to conduct randomized clinical trials with a large number of patients and a longer follow-up period to fully elucidate the causal mechanism underlying the association of periodontitis, P. gingivalis infection, and periodontal treatment with RA measures in patients.

Conflict of interest

The authors have no conflict of interests related to the present article.