Kulveer Mankia

Identifying arthralgia suspicious for progression to rheumatoid arthritis

We read with interest the article by van Steenbergen et al 1 in which a definition for arthralgia suspicious for progression to rheumatoid arthritis (RA) was proposed. The authors used a three-phase Delphi exercise to crystallise the concept of clinically suspect arthralgia (CSA), which is inherently subjective, into a core set of definable parameters. We agree that this set of characteristics should provide a useful secondary care framework for identifying homogeneous at-risk populations for future clinical studies. Recent data suggest that rheumatologists can use symptoms and signs to identify which patients with arthralgia referred to them will imminently develop RA.2 ,3 In the current cohort, up to 20% of individuals identified as CSA by their rheumatologist developed RA during follow-up, the majority doing so …

Periodontal disease and periodontal bacteria as triggers for rheumatoid arthritis

There is an epidemiological association between periodontitis and rheumatoid arthritis (RA), which is hypothesised to lead to enhanced generation of RA-related autoantibodies that can be detected years before the onset of RA symptoms. Periodontitis is a common dysbiotic disease; tissue damage occurs because the immune system fails to limit both the resident microbial community and the associated local immune response. Certain periodontal bacteria, including Porphyromonas gingivalis and Aggregatibacter actinomycetemcomitans, may contribute to RA autoantibody production through direct post-translational modification of proteins or, indirectly, by influencing neutrophil-mediated neo-epitope generation. Oral bacteria that invade the blood may also contribute to chronic inflammatory responses and generation of autoantibodies. The putative association between periodontitis and the development of RA raises the potential of finding novel predictive markers of disease and disease progression and for periodontitis treatment to be included in the future as an adjunct to conventional RA immunotherapy or as part of a preventive strategy.

A new window of opportunity in rheumatoid arthritis: targeting at-risk individuals.

Purpose of reviewProgress in our understanding of the preclinical events in rheumatoid arthritis (RA) has provided important insights into disease pathogenesis. Studying prospective cohorts of individuals at risk for RA development offers the opportunity to accurately characterize the sequence of events in preclinical disease as well as quantify the risk of different preclinical phenotypes. These data may provide the basis for preventive strategies in RA. Recent findingsRA-related systemic autoimmunity and inflammation occur long before clinical arthritis. There is growing evidence that initiating events may occur at mucosal surfaces including the periodontium, lung and gut and may be influenced by the local microbiome. For potential preventive strategies to be feasible, it is important that individuals at high risk for RA development can be readily identified from the general population. To this end, studying multiple biomarkers in prospective cohorts of at-risk individuals enables risk prediction in different at-risk phenotypes. RA prevention using immunomodulation is currently being investigated in individuals at high risk of RA development. SummaryThe prospective study of at-risk individuals can provide invaluable aetiological insights as well as facilitating accurate risk prediction data. In this way, high-risk individuals may be identified for preventive interventions.

Anti–Porphyromonas gingivalis Antibodies in Rheumatoid Arthritis: Comment on the Article by Seror et al

To the Editor: We read with interest the article by Seror et al (1), in which they reported that anti–Porphyromonas gingivalis antibody titers did not differ between patients with early rheumatoid arthritis (RA) and control subjects. Based on their observations in a large cohort of patients with early RA, the authors concluded that the well-documented association between periodontal disease and RA could be linked to bacterial species other than P gingivalis or to a mechanism other than citrullination. Several studies have investigated the relationship between anti–P gingivalis antibodies and RA. Although some demonstrated higher anti–P gingivalis antibody levels in individuals at risk of RA compared with controls (2–4), others (including the study by Seror et al) showed no differences in these antibody levels between patients with established RA and controls (5,6). As noted by Seror et al, methods for measuring anti–P gingivalis antibodies are not yet standardized, and different assays detect antibodies to different protein targets. Comparing the results of antibody assays across different studies must therefore be done with some caution. Although we acknowledge that these data suggest anti–P gingivalis serology cannot distinguish patients with RA from control subjects, we do not agree that this undermines an etiologic model involving P gingivalis and citrullination in RA. Porphyromonas gingivalis is a key component of the “red complex” bacteria that are synonymous with periodontal disease (7). It expresses a bacterial peptidylarginine deaminase (PPAD) that is able to citrullinate arginine residues in a distinct manner. First, P gingivalis expresses arginine-specific gingipains that cleave host proteins, exposing C-terminal arginines that are then citrullinated by PPAD (8). In contrast, endogenous human PADs cannot generate C-terminal citrullines. Second, there is evidence that PPAD undergoes autocitrullination in patients with RA (9). Both of these features highlight the unique ability of P gingivalis to generate novel citrullinated peptides that could potentially break immune tolerance at the oral mucosa. A previous study using 16S ribosomal RNA pyrosequencing in patients with RA confirmed that the oral microbiome is distinct in patients with periodontal disease compared with controls, with P gingivalis being significantly more prevalent and abundant in patients with periodontal disease (5). Consistent with the results described by Seror et al, there was no difference in anti–P gingivalis antibody levels between these groups. Because P gingivalis is a ubiquitous oral microorganism, it is not surprising that antibodies, a measure of past exposure, are found in individuals regardless of their periodontal disease state. Instead, the abundance and prevalence of P gingivalis in the oral mucosa are heightened in periodontal disease. Seror et al used DNA–DNA hybridization to detect P gingivalis in the periodontal tissue of their periodontitis control subjects. They observed that those with detectable P gingivalis had higher levels of anti–P gingivalis antibodies and concluded that a higher titer of anti–P gingivalis is a good marker for chronic P gingivalis infection. However, this implies that the prevalence of active P gingivalis in the healthy control group, of whom only 16% were ever smokers, was similar to that in the periodontitis patients. This would be inconsistent with previous studies (7,10) and is difficult to accept, particularly because a similar analysis in the healthy controls and RA patients was not performed. When considering the etiologic role of P gingivalis in RA, anti–P gingivalis antibodies alone are unlikely to be informative. Simple exposure to this ubiquitous bacterium is unlikely to be sufficient to trigger autoimmunity. Instead, an increased abundance of P gingivalis in the context of active periodontal inflammation is likely to be required to generate the local citrullination that breaks immune tolerance. As such, periodontal examination and microbiome characterization must be important inclusions in future prospective work in this area.

Imminent rheumatoid arthritis can be identified in primary care

We thank Vega-Morales et al 1 for their interest in our proposed approach for identifying individuals at risk of rheumatoid arthritis (RA) in a primary care setting.2 ,3 The authors agree that primary care is usually the first point of contact for patients with RA when they initially develop symptoms. For this reason, they agree that general practitioners (GPs) are well placed to be involved in screening strategies to identify individuals at risk of progressing to RA.1 Vega-Morales et al advocate the squeeze test (Gaenslens compression test) as a screening tool to aid GPs in identifying at-risk individuals for onward referral to …

FRI0102 Ultrasound and mri during flares of palindromic rheumatism reveal a distinct phenotype even in imminent ra

Background Palindromic rheumatism (PR) is a recurrent, self-abortive arthritis and/or peri-arthritis which progresses to RA in up to 50% of patients, especially those that are anti-CCP positive (1). Whether PR is truly a prodrome of RA or a distinct syndrome is unclear; the pathological phenotype of PR flare, and whether this changes when RA is imminent, remains unknown Objectives To describe the clinical and imaging (US and MRI) phenotype during PR flares and to determine whether this changes in PR patients with imminent RA. We hypothesised PR patients with imminent RA would have a RA phenotype during flare Methods Patients were recruited from a prospective PR cohort. PR flares were defined as ≥2 of pain, swelling, erythema in or around ≥1 joint, that later normalised. Clinical details were recorded during flares. Blinded US assessment (wrists, MCPs, PIPs, elbows, knees, MTPs, ECUs and 2nd-5th finger flexor tendons) was performed during and between flares. Synovitis, tenosynovitis, subcutaneous oedema, peri-articular inflammation and peri-tendinous oedema were reported at each joint. Where possible, MRI was also performed on the most symptomatic region during flare. Patients were followed for progression to RA Results US was performed in 22 patients during flare. 19 patients also had non-flare US. Mean age was 49 yrs. 16/22 (73%) were anti-CCP+ and 6/22 (27%) anti-CCP-. Six (27%) patients developed RA (mean 23 weeks post flare); these patients had higher frequency of flares and absence of flare trigger (figure) but no difference in distribution of flaring joints. The flare US showed grey scale (GS) synovitis in 13/22 (59%) patients, power Doppler (PD) synovitis in 4/22 (18%) and no erosions. 12/22 (55%) patients had peri-articular inflammation and/or subcutaneous oedema; in 6 patients this was without synovitis/tenosynovitis. Tenosynovitis and/or peri-tendinous oedema were present in 6/22 (27%) patients. Non-flare US demonstrated fewer abnormalities with improvement post flare. In 5 patients multiple flares were imaged with variable US abnormalities identified. The US flare phenotype did not differ (from non-progressors) in patients who progressed to RA, with PD synovitis present in only 17% (table 1) contrasting with our early RA cohort where PD synovitis occurred in 73% of patients (2). MRI was performed on 8 patients (2 flares imaged in 1 patient) and detected more pathology than US. Bone marrow oedema (BME) was found in only 1 patient. No erosions were seen.Table 1 US abnormality Progressed to RA (n=6) Not progressed to RA (n=16) PD synovitis 1 (17%) 3 (19%) GS synovitis 2 (33%) 10 (63%) Tenosynovitis 1 (17%) 4 (25%) Peri-tendinous oedema 0 (0%) 2 (13%) Peri-articular inflammation 1 (17%) 7 (44%) Subcutaneous oedema 3 (50%) 6 (38%) Conclusions PR flares have a distinct imaging phenotype characterised by peri-articular inflammation and subcutaneous oedema, often without synovitis. The low prevalence of PD synovitis, BME and erosions distinguishes PR from RA. PR patients with imminent RA have no triggers and more frequent flares, but retain the distinct PR phenotype. This suggests distinct pathological mechanisms in PR and should be of value for potential therapeutic interventions. References Russell AS et al.J Rheumatol 2006.33(7):1240–2. Horton SC et al.Rheumatol 2016;55:1177–87. Disclosure of Interest None declared

Comorbidities in Anti-Cyclic Citrullinated Peptide Positive At-Risk Individuals Do Not Differ from Those Patients with Early Inflammatory Arthritis

Objectives To compare comorbidities in a cohort of cyclic citrullinated peptide (CCP) antibody positive patients without or prior to onset of inflammatory arthritis (IA) to those in patients with early IA. Methods Baseline data from two established cohorts were used. The first recruited people at risk of IA: CCP antibody positive cases without IA (CCP Cohort, n = 296). The second cohort [the Inflammatory Arthritis CONtinuum study (IACON)] recruited patients with early IA (n = 725). Proportions of patients with given comorbidities were compared between cohorts and then logistic regression was used to determine odds ratios (OR) for the CCP cohort having specific comorbidities, compared to IACON patients. Analyses adjusted for gender, age, smoking status, and body mass index. Results Patients from the CCP cohort were younger (mean age 50, compared to 53 years). The proportion of patients with at least one comorbidity was higher in the IACON than the CCP cohort: (40% compared to 24%, respectively). Results of logistic regression analyses suggested the odds of hypertension, taking a lipid-lowering agent, ischemic heart disease, cerebrovascular disease, lung disease, and diabetes were not increased in either cohort. However, patients in the CCP cohort were more likely to be taking an antidepressant (OR = 1.62, 95% CI 1.03, 2.56, p = 0.037). Conclusion There was no significant difference in comorbidities among people with CCP antibodies but without IA, compared to those of patients with established IA.

Identification of a distinct imaging phenotype may improve the management of palindromic rheumatism

Objectives To use high-resolution imaging to characterise palindromic rheumatism (PR) and to compare the imaging pattern observed to that seen in new-onset rheumatoid arthritis (NORA). Methods Ultrasound (US) assessment of synovitis, tenosynovitis and non-synovial extracapsular inflammation (ECI) was performed during and between flares in a prospective treatment-naive PR cohort. MRI of the flaring region was performed where possible. For comparison, the same US assessment was also performed in anticyclic citrullinated peptide (CCP) positive individuals with musculoskeletal symptoms (CCP+ at risk) and patients with NORA. Results Thirty-one of 79 patients with PR recruited were assessed during a flare. A high frequency of ECI was identified on US; 19/31 (61%) of patients had ECI including 12/19 (63%) in whom ECI was identified in the absence of synovitis. Only 7/31 (23%) patients with PR had synovitis (greyscale ≥1 and power Doppler ≥1) during flare. In the hands/wrists, ECI was more prevalent in PR compared with NORA and CCP+ at risk (65% vs 29 % vs 6%, p<0.05). Furthermore, ECI without synovitis was specific for PR (42% PR vs 4% NORA (p=0.003) and 6% CCP+ at risk (p=0.0012)). Eleven PR flares were captured by MRI, which was more sensitive than US for synovitis and ECI. 8/31 (26%) patients with PR developed RA and had a similar US phenotype to NORA at progression. Conclusion PR has a distinct US pattern characterised by reversible ECI, often without synovitis. In patients presenting with new joint swelling, US may refine management by distinguishing relapsing from persistent arthritis.

OP0352 An increased prevalence of periodontal disease, porphyromonas gingivalis and aggregatibacter actinomycetemcomitans in anti-ccp positive individuals at-risk of inflammatory arthritis

Background The prevalence of periodontal disease is increased in RA, and periodontitis is associated with the bacterium Porphyromonas gingivalis (Pg), which can citrullinate arginine residues.1 These observations suggest periodontitis may be a key initiator of RA-related autoimmunity. Importantly, clinical periodontal disease, and the relative abundance of periodontal bacteria have not been described in seropositive individuals at risk of developing RA who do not have synovitis. Objectives To investigate the prevalence of periodontal disease and the relative abundance of key periodontal bacteria in anti-CCP positive at-risk individuals without synovitis. Methods Anti-CCP positive individuals with musculoskeletal symptoms but no clinical synovitis (CCP+), early RA (RA) patients and healthy controls (HC) were recruited. CCP +underwent a 38 joint ultrasound (US) assessment. Periodontal examination was performed by a dentist; six sites per tooth were assessed for clinical attachment level (CAL), pocket depth (PD) and bleeding on probing (BOP). Periodontal disease sites (PDD) were defined as CAL≥2 mm and PD ≥4 mm. A clinical consensus was agreed for each case by three dentists. DNA, isolated from subgingival plaque from diseased and healthy periodontal sites, was pair-end sequenced (Illumina HiSeq3000). Taxonomic and functional profiles were obtained from MG-Rast and differences between groups studied using DESeq2. Mann-Whitney U tests were used to compare groups and Spearman Rho used for correlations. For metagenomic data, Wald test was used to compare relative abundance.Abstract OP0352 – Figure 1 Relative abundance of Aggregatibacter actinomycetemcomitans (Aa) and Porphyromonas gingivalis (Pg) according to RA status and extent of periodontal disease Results 48 CCP+, 26 RA and 32 HC were recruited. Groups were balanced for age, sex and smoking. All but 2 (96%) CCP +had no US synovitis (greyscale ≥1 and power Doppler≥1). Dentists classified 73% CCP+, 38% HC (p=0.02) and 54% RA as having clinical periodontitis. The percentage of periodontal sites with CAL ≥2 mm, PD ≥4 mm, BOP, PDD and active PDD (PDD +BOP) were all greater in CCP +compared to HC (p<0.05) and similar to RA. In non-smokers, PDD and active PDD were more prevalent in CCP +compared to HC. Metagenomic data indicated CCP +had increased relative abundance of both Pg and Aggregatibacter actinomycetemcomitans (Aa) compared to HC (p<0.001) and RA (p<0.01). However, clinical periodontitis was only associated with increased relative abundance of Pg (p<0.001) but not Aa. Furthermore, the relative abundance of Pg was associated with the percentage of sites with active PDD in CCP+ (p=0.05) and HC (p=0.04) but this was not seen for Aa (figure 1). Conclusions We report an increased prevalence of periodontal disease, Pg and Aa in anti-CCP positive at-risk individuals without synovitis. Interestingly, relative abundance of Pg, but not Aa, was associated with periodontitis, suggesting potential mechanistic differences that require further exploration. These data support the concept that periodontal inflammation and periodontopathic bacteria may both be important in the initiation of RA-related autoimmunity. Reference [1] Lundberg K. Nat Rev Rheumatol2010. Disclosure of Interest None declared

FRI0047 Mri interosseous tendon inflammation occurs in anti-ccp positive at-risk individuals and may precede the development of synovitis

Background Tenosynovitis (TSV) occurs in individuals at-risk of developing RA and could explain pain and stiffness in the absence of synovitis. TSV of the wrist and finger flexor tendons has been described in at-risk individuals but involvement of other hand tendons has not been well investigated. The hand interossei are crucial to hand function and can become inflamed in RA(.1 Whether the interosseous tendons (IT) are sites of inflammation in at-risk individuals, and how this relates to joint inflammation and clinical features is unknown. Objectives To describe the anatomy, prevalence, pattern and clinical associations of IT inflammation in anti-CCP positive at-risk individuals. Methods Anti-CCP positive individuals with no synovitis (CCP+), healthy controls (HC), DMARD-naïve early RA patients (ERA) and treated ‘late’ RA patients (LRA) were recruited. All subjects underwent clinical and MRI assessment. 1.5T or 3T unilateral hand MRI scans were consensus scored for RAMRIS, TSV and IT inflammation by two radiologists. IT inflammation was defined as enhancing tissue around the tendon evident in two planes. For RAMRIS and tenosynovitis, scores were adjusted for 193 age-matched controls(.2 To understand the anatomical basis for MRI IT inflammation, a cadaveric study was performed on 20 fresh hand specimens; coloured dyes were injected along the first dorsal IT and into the adjacent second MCP joint and specimens were frozen and sectioned. Results 93 CCP+, 20 HC, 47 ERA and 28 LRA were recruited. Frequency of swollen and tender joints, MRI inflammation (synovitis, BME, erosions, TSV) and CRP level increased along the RA continuum with increasing disease duration. The proportion of patients with IT inflammation increased along the RA continuum. No HC, 18/93 (19%) CCP+, 23/47 (49%) ERA and 16/28 (57%) LRA patients had inflammation of ≥1 IT (p<0.001). The number of affected ITs increased along the RA continuum (p<0.001) and tendons associated with MCPJs 2 and 5 were most commonly affected. IT inflammation and MRI synovitis were associated with MCPJ swelling [OR 2.7 (0.9, 8.1) and OR 3.1 (1.0, 9.8) respectively] but IT inflammation was the only feature independently associated with MCPJ tenderness [OR 3.1 (1.4, 6.8) p=0.004]. In CCP+, 99/372 (27%) MCPJs had only one MRI abnormality; in 68% of these the abnormality was extra-capsular (57% TSV and 11% IT inflammation). No IT sheath was identified in the cadaveric specimens suggesting the MRI findings represent peri-tendonitis rather than TSV. Dye studies indicated no clear communication between the IT and the adjacent joint (figure 1).Abstract FRI0047 – Figure 1 Axial MRI showing inflammation of the 3rd palmar interosseous tendon and frozen section showing green dye around the interosseous tendon and blue dye within the MCP joint Conclusions IT inflammation represents a peri-tendonitis and is present in anti-CCP +at risk individuals and RA patients where it is associated with MCPJ swelling and tenderness. IT inflammation can occur as the lone MRI abnormality in CCP +at risk individuals suggesting the interossei may be an early extra-capsular target in the development of RA. References [1] Rowbotham, et al. Eur Radiol2015. [2] Mangnus, et al. Arthritis Rheum2016. Acknowledgements D Glinatsi, M Ostergaard, P Bird Disclosure of Interest None declared