Is glucocorticoid bridging therapy associated with later use of glucocorticoids and biological DMARDs during the disease course of patients with rheumatoid arthritis in daily practice? A real-world data analysis

OBJECTIVE

To evaluate if initially starting glucocorticoid (GC) bridging leads to a higher probability of long-term GC and biological (b)DMARD use in rheumatoid arthritis (RA)-patients.

METHODS

Electronical health records data from newly diagnosed RA-patients from the Leiden University Medical Center were used. Patients who started GC as part of initial treatment (iGC group) and who did not (niGC group) were compared in terms of GC and bDMARD use later in the disease course. Multivariable adjustment was performed to account for confounding by indication.

RESULTS

465/932 newly diagnosed RA-patients (50 %) were treated with GC as initial treatment step. Patients in the iGC group were older, included fewer females, had a higher disease activity at baseline compared to the niGC group plus a more rapid decrease in DAS28 in the first 6 months. During follow-up, 42 % of the iGC group started a second course of GC and 17 % started a bDMARD, compared to 34 % and 13 % In the niGC group. The hazard to start a bDMARD later in the disease course was not significantly different between the two groups in two time periods (0.34 95 %CI(0.09;1.21) resp. 1.48 95 %CI (0.98;2.22)), but the hazard to (re)start GC later on was higher for the iGC group (aHR 1.37 95 %CI(1.09;1.73)).

CONCLUSION

In this daily practice cohort of newly diagnosed RA patients, patients in the iGC group had a more rapid DAS28 decrease and an increased probability of starting GC later on compared to the niGC group. The probability of bDMARD use was not significantly increased.

OP0091 EXPANSION OF MYELOPEROXIDASE (MPO)-SPECIFIC IgM B CELLS IN ANTI-NEUTROPHIL CYTOPLASMIC ANTIBODY (ANCA)-ASSOCIATED VASCULITIS

Background

ANCA-associated vasculitides (AAV) are characterized by recurrent, chronic small vessel inflammation and deleterious organ damage. The main targets of ANCA are myeloperoxidase (MPO) and proteinase 3 (PR3). ANCA, B cells and the complement system are crucial to AAV pathogenesis, as evidenced by the clinical benefit of B cell depletion with rituximab and, more recently, the C5a receptor antagonist avacopan. While ANCA in serum have been studied extensively, phenotypic and functional characteristics of the underlying B cell responses remain largely unknown.

Objectives

To develop a flow cytometry-based technique for identifying MPO-specific B cells in the circulation of MPO-positive AAV patients in order to characterize this B cell response and its potential contribution to disease pathogenesis.

Methods

Human neutrophil-derived MPO was conjugated to two different fluorochromes and used to identify MPO-specific B cells by flow cytometry. An antigen-specific staining protocol was developed and validated using MPO- and PR3-specific hybridoma cells. MPO-specific B cells were phenotypically characterized and isolated from the peripheral blood of AAV patients by fluorescence activated cell sorting (FACS) and cultured as single cells. MPO-specificity was confirmed by ELISA on culture supernatants. B cell receptor (BCR) sequences were obtained from MPO-positive clones by full length ARTISAN-PCR and Sanger sequencing. MPO-specific IgG and IgM monoclonal antibodies (mAb) were produced to validate specificity and to examine their ability to activate complement. Finally, MPO-positive AAV patient plasma and plasma depleted of IgG or IgM was tested in in vitro complement assays.

Results

The newly developed, differential antigen labelling approach successfully identified MPO-specific but not PR3-specific hybridoma cells. Subsequently, we detected MPO-specific B cells in the circulation of MPO-positive AAV patients at a frequency of up to 1:1000 B cells. FACS sorting and single cell culture yielded an enrichment of MPO-specificity of ~80%. Notably, the majority of isolated, MPO-specific B cells (60-95%) displayed an IgM memory phenotype, which corresponded to the presence of anti-MPO IgM in plasma. The remainder of the MPO-specific cells were mainly IgG memory B cells and few naive cells. BCR sequencing revealed a polyclonal IgM response with diverse V-gene usage, consisting of both germline and highly mutated clones. Generation of mAb (n=5) confirmed MPO specificity by inhibition ELISA for both germline and somatically mutated clones. Interestingly, anti-MPO IgM mAb showed a high capacity for complement factor deposition upon MPO binding. MPO-specific complement assays with IgG- and IgM-depleted patient plasma showed that anti-MPO IgM activated complement much more efficiently than anti-MPO IgG.

Conclusion

We demonstrate the direct ex-vivo identification, isolation and characterization of MPO-specific B cells in human AAV. Intriguingly, we observed a remarkable expansion of MPO-specific, IgM-expressing memory B cells in patients. This so far unrecognized, active IgM-compartment may be clinically relevant, as both mAb and plasma-derived polyclonal MPO-specific IgM strongly activated complement, a pathway thought to play a central role in AAV. Next to these novel insights into autoreactive B cell biology in AAV, our findings now provide new opportunities for studying auto-reactive B cell responses in different clinical phases of AAV, amongst which active disease, remission and (imminent) flares.

Disclosure of Interests

None declared.

OP0272 PREDNISONE USE AND THE INCIDENCE OF HYPERGLYCEMIA OR DIABETES IN PATIENTS WITH RHEUMATOID ARTHRITIS; A 10-YEAR SUB ANALYSIS OF THE BeSt STUDY

Background

Use of prednisone in rheumatoid arthritis has been questioned because it may trigger side effects such as hyperglycemia and diabetes.

Objectives

To assess whether in RA the use of prednisone is associated with the development of hyperglycemia and diabetes.

Methods

The BeSt study is a multicenter, assessor-blinded randomized controlled 10-years follow-up trial in 508 non-diabetic early RA patients. Patients were randomised to 4 dynamic DMARD treatment strategy groups: 1) sequential monotherapy, 2) step-up combination therapy, 3) initial combination therapy including prednisone (60 mg/day, tapered to 7.5 mg/day in 7 weeks) and 4) initial combination therapy with infliximab. In groups 1, 2 and 4, prednisone had a maximum dose of 7.5 mg/day by protocol. Treatment was steered at disease activity score (DAS) ≤2.4. We performed a GEE over time to assess whether current prednisone use or cumulative prednisone dose were associated with hyperglycemia (glucose levels ≥7.8) and cox regression analyses to investigate the relationship between cumulative prednisone dose, previous prednisone use and diabetes (defined as either use of anti-diabetic medication or two instances of a glucose ≥ 11.1), assessed at 3-monthly visits. All analyses were adjusted for potential confounders.

Results

In total, 33/508 patients (6.5%) developed diabetes during the trial; 12 of these (36%) had received prior treatment with prednisone (any dose). Median (IQR) duration of prednisone use in all 508 patients was 9 (15) months and cumulative doses ranged from 0 to 27942 mg. The mean cumulative dose ranged from 55.5 mg in group 1 to 6170.0 mg in group 3. Previous prednisone use nor cumulative prednisone dose was associated with hyperglycemia or diabetes, with effect sizes ranging from a hazard ratio of 0.588 (95% CI 0.285; 1.21) for the association between any prednisone dose and diabetes to an odds ratio of 1.04 (95% CI 0.978; 1.13) for the association between cumulative prednisone dose and diabetes (Table 1). To identify potential causes for these results, we investigated the relationship between DAS and the same outcomes. We found a higher DAS was significantly associated with development of diabetes, but not with hyperglycemia.

Table 1.

The relationship between prednisone dose, DAS and glucose levels, hyperglycemia and diabetes

GEEhyperglycemia*OR95% CIAny prednisone dose10.9490.805; 1.12Cumulative dose11.04**0.978; 1.13DAS21.240.842; 1.85Cox Regressiondiabetes (any of the definitions)HR95% CIAny prednisone dose30.5880.285; 1.21Cumulative dose30.996**0.960; 1.03DAS21.601.13; 2.26

CI: confidence interval, GEE: Generalized Estimating Equations, OR: odds ratio, HR: hazard ratio, DAS: disease activity* hyperglycemia: glucose level above 7.8 mmol/L; diabetes: random glucose level above 11.1 mmol/L at at least two time points** odds ratio per 500mg cumulative prednisone increase: 1adjusted for DAS, age, diabetes and BMI: 2adjusted for cumulative prednisone dose, age, gender and BM: 3adjusted for DAS, age and BMI

Conclusion

In early RA patients, cumulative dose nor any previous prednisone use was associated with the risk of hyperglycemia or diabetes. A higher DAS was significantly associated with increased risk of developing diabetes. Potential risks of prednisone may have been mitigated by suppression of DAS.

Disclosure of Interests

Joy van der Pol: None declared, Sytske Anne Bergstra Grant/research support from: Pfizer, Thomas Huizinga: None declared, Cornelia Allaart Grant/research support from: The BeSt study was supported by a government grant from the Dutch Insurance Companies, with additional funding from Schering-Plough B.V. and Janssen B.V.

POS0865 THE EFFECT OF SILVER FIBER GLOVES ON RAYNAUD’S PHENOMENON IN PATIENTS WITH SYSTEMIC SCLEROSIS: A DOUBLE-BLIND RANDOMIZED CROSS-OVER TRIAL

Background

Over 90% of patients with systemic sclerosis (SSc) experience Raynaud’s phenomenon (RP), which strongly influences quality of life. Therapeutic options of RP include drug treatment and general lifestyle measures such as smoking cessation and avoiding cold by wearing warm clothes and gloves including electrically heated gloves or silver fiber gloves. Clinical observations suggest an additional benefit of silver fiber gloves compared to normal gloves. Silver is thought to help by reflecting heat back into the hands allowing less heat to escape and has an antimicrobial effect. Despite its generalized use among SSc patients, no objective evidence regarding its superiority for RP over normal gloves exists.

Objectives

To evaluate the added value of 8% silver fiber gloves compared to normal gloves in the treatment of patients with RP secondary to SSc.

Methods

This was a multicenter double-blind randomized cross-over trial in which 85 SSc patients were randomized in two sequences: 8% silver fiber gloves in period 1 and normal gloves in period 2 or vice versa; each period lasted six weeks. To reduce bias of interindividual differences and external factors (e.g. temperature), a cross-over design was performed in the Netherlands during the winter months. The primary outcome was the triweekly Raynaud Condition Score (RCS), a scale from 1 (no symptoms) to 10 (extreme symptoms). A linear mixed model was used with RCS as dependent and type of gloves as independent variable, adjusted for baseline RCS. Secondary outcome measures included number of RP attacks, RP attack duration, Health Assessment Questionnaire (HAQ-DI) and vascular complications. Secondary outcomes were also analyzed with linear mixed models. All analyses were performed and interpreted before unblinding.

Results

The 85 included SSc patients had a mean age of 60 (SD:12), 80% were female, 60% had limited cutaneous SSc and 67% used vasoactive medication. Ten patients prematurely ended the study due to various reasons, most notable: allergic reaction to gloves (n=2). At baseline, mean RCS was 6.43 (SD 1.6), with silver fiber gloves the mean RCS decreased to 3.91 (SD 2.3) and with normal gloves to 3.90 (SD 2.3) (Figure 1). No statistically significant difference in RCS during follow-up was observed between the silver fiber gloves and normal gloves (β 0.067, 95% CI -0.006 to 0.19), meaning that on the 1-10 scale, silver fibre gloves gave only a 0.067 higher RCS compared to normal gloves (Table 1). For all other secondary outcome measures, we did not find a statistically significant difference between silver fiber gloves and normal gloves, except for the HAQ (β 0.036, 95% CI 0.026 to 0.046; Table 1), which is not clinically relevant. One vascular complication occurred in the silver fiber gloves, compared to three vascular in the normal gloves, which was not statistically significant different (OR:3.2, 95% CI 0.32 to 31.1).

Table 1.

Primary and secondary efficacy outcomes

β95% confidence intervalPrimary outcomeRaynaud Condition Score0.067-0.0059; 0.194Secondary outcomesRaynaud attacks frequency-0.480-1.215; 0.255Raynaud attacks duration39.80-36.051; 115.654VAS warmth hands-0.086-0.212; 0.041Impact Raynaud0.088-0.035; 0.211HAQ_DI0.0360.026; 0.046

VAS: visual analogue scale; HAQ: Health Assessment Questionnaire

The reference category was Normal gloves.

Linear mixed models were performed with the primary and secondary outcomes as dependent variables, the type of gloves as independent variable, adjusted for baseline Raynaud Condition Score.

Figure 1.

Raynaud Condition Score during the study period

Conclusion

This trial shows that wearing any type of glove decreases the RP burden in SSc patients, but no additional benefit from gloves containing 8% silver fibers compared to normal gloves could be demonstrated. Potentially, less vascular complications may arise in SSc patients wearing silver fiber gloves. Further confirmation of this potential benefit is necessary.

Acknowledgements

The authors would like to thank all participants of this study and Skafit for providing the gloves.

Disclosure of Interests

None declared

POS0103 IN RHEUMATOID ARTHRITIS, THE ASSOCIATION BETWEEN ANTI-MODIFIED PROTEIN ANTIBODIES AND LONG-TERM OUTCOMES IS DOMINATED BY THE EFFECT OF ANTI-CITRULLINATED PROTEIN ANTIBODIES

Background

In rheumatoid arthritis (RA) many patients harbor autoantibodies against post translationally modified proteins (anti-modified protein antibodies (AMPA)): anti-citrullinated protein antibodies (ACPA), anti-carbamylated protein antibodies (anti-CarP) and anti-acetylated protein antibodies (AAPA). All three AMPA have been described to be associated with an increase in radiographic joint damage and ACPA also with a lower chance of obtaining sustained drugfree remission (SDFR). However, it is unclear whether the effects of these autoantibodies are independent of each other, and whether testing for all three AMPA offers additional information regarding these clinical outcomes.

Objectives

To investigate the individual and combined associations of AMPA with radiological progression and SDFR in RA.

Methods

In 612 RA patients from the Leiden Early Arthritis Clinic, we measured ACPA IgG with anti-CCP2-assays, anti-CarP IgG using homocitrullinated and native (as a control) fetal calf serum, and AAPA IgG by means of novel assays consisting of cyclic peptides with either acetylated or norleucine residues (as a control). Sharp-van der Heijde scores (SHS) were determined in 2685 sets of radiographs with yearly intervals for 7 years. The association of AMPA with SHS was assessed with a multivariate normal regression model. SDFR was defined as the absence of clinical synovitis after discontinuation of DMARD treatment. The association of SDFR with autoantibody-status was assessed with Kaplan-Meier curves. The very small number of patients single-positive for AAPA (n=4, 1%) precluded analyses in this subgroup.

Results

The prevalence of autoantibodies is shown in Table 1. A higher SHS was found in patients with autoantibodies compared to patients without antibodies, and in triple-positive versus single-positive patients (Figure 1A, p=0.04). To investigate if the higher SHS was due to the number of autoantibodies or a specific autoantibody, analyses were stratified for all three AMPAs. Interestingly, no difference was found in SHS between all ACPA-positive strata (Figure 1B). In the ACPA-negative stratum, a significant difference was found between patients with zero antibodies and solely anti-CarP (p=0.02).

Table 1.

Prevalence of autoantibodies in 612 rheumatoid arthritis patients in the EAC cohort

Autoantibody statusn (%)ACPA-positive316 (52)Anti-CarP-positive270 (44)AAPA-positive208 (34)ACPA-AAPA-CarP-255 (42)ACPA+AAPA-CarP-46 (7)ACPA-AAPA+CarP-4 (1)ACPA-AAPA-CarP+37 (6)ACPA+AAPA+CarP-37 (6)ACPA+AAPA-CarP+66 (11)ACPA-AAPA+CarP+- (0)ACPA+AAPA+CarP+167 (27)

EAC: early arthritis clinic, ACPA: anti-citrullinated protein antibodies, anti-CarP/CarP: anti-carbamylated protein antibodies, AAPA: anti-acetylated protein antibodies

The chance to obtain SDFR was lower in all antibody-positive groups compared to the autoantibody negative group. Moreover less triple-antibody-positive patients than single-antibody-positive patients achieved SDFR (P<0.001), Figure 1C. After stratification for AMPA, no difference was found in the percentage of SDFR in all ACPA-positive strata, or in the ACPA-negative stratum, Figure 1D.

Conclusion

In ACPA-positive patients, the presence of other AMPA influences neither radiographic progression, nor the chance of SDFR. Thus, long-term clinical phenotype in RA is particularly dependent on the presence of ACPA, and less on the presence of other AMPA. Therefore, there appears to be no added value of testing the other AMPA for predicting clinical outcome in ACPA-positive patients.

Disclosure of Interests

None declared

AB0400 A SYSTEMATIC LITERATURE REVIEW AND META-ANALYSIS INTO THE SUCCESS RATE OF GLUCOCORTICOID DISCONTINUATION AFTER THEIR USE AS INITIAL BRIDGING THERAPY IN RHEUMATOID ARTHRITIS PATIENTS IN OBSERVATIONAL COHORTS AND CLINICAL TRIALS

Background

Glucocorticoids (GC) are widely used for the initial treatment of rheumatoid arthritis (RA), to induce rapid suppression of inflammation and clinical symptoms and thereby limit radiographic damage progression. There are concerns that GC use in the long term is associated with a dose and duration dependent risk of serious side effects. Therefore, international guidelines have recommended to start GC when initiating a csDMARD, but to discontinue GC as rapidly as clinically feasible, preferably within 3 months (bridging therapy). In contrast, due to the concerns of GC side effects, the ACR guidelines published in 2021 conditionally recommend to start csDMARD monotherapy without GC bridging therapy.

Objectives

We aim to evaluate the success rate of GC discontinuation after using temporary GC as part of initial therapy (‘bridging’) both in observational cohorts and clinical trials in newly diagnosed RA patients.

Methods

Systematic literature searches were conducted to identify observational cohorts (scoping search) and clinical trials (in-depth search) that included RA patients who were treated with initial GC bridging therapy. GC bridging was defined as oral or intramuscular GC treatment that was discontinued within one year, alongside conventional DMARD therapy. Patient percentages still or again using GC were considered to represent the reverse of successful discontinuation. Random-effects meta-analyses were performed stratified by time point.

Results

The literature search on observational cohort studies could not identify any study answering the research question, since it remained unclear which patients had received GC as part of the initial treatment. The literature search for clinical trials identified 7160 abstracts, resulting in 10 included studies, with varying type and dose of GC and varying tapering schedules (Table 1). Of these included studies, 4 reported sufficient data on GC discontinuation or GC use after the bridging phase. The pooled proportion of patients who were still using GC was 22% (95% Confidence Interval (CI) 8; 37, based on 4 trials) at 12 months and 10% at 24 months (95% CI -1; 22, based on 2 trials) (Figure 1). Thus, the vast majority had stopped GC. Heterogeneity was substantial (I2 ≥ 65%).

Table 1.

Overview of included clinical trials.

Study (publication year)Tapering schedule (mg/day)COBRA (1997)In 7 weeks to 7.5. Stop after 28 weeks.*BeSt (2005)In 7 weeks to 7.5. Stop in 8 weeks after week 28 if DAS persistently ≤2.4IDEA (2014)N.A.COBRA-light (2015)arm 1: in 7 weeks to 7.5 arm 2: in 9 weeks to 7.5 Stop after 32 weeks if DAS<1.6.IMPROVED (2014)In 7 weeks to 7.5. Stop after 20 weeks if DAS <1.6 at 4 months.ARCTIC (2016)In 7 weeks to 0 if DAS <1.6 and no swollen joints present.tREACH (2013)In 10 weeks to 0.*CareRA (2017)- in 7 weeks to 7.5, further tapered from week 28, stop after 34 weeks.- Classic- in 6 weeks to 5, further tapered from week 28, stop after 34 weeks.- Slim- in 6 weeks to 5, further tapered from week 28, stop after 34 weeks.- Avant gardeAll if DAS28(CRP) ≤3.2.Hua et al. (2020)Tapering after 4 months to 5, stop after 6 months.*NORD-STAR (2020) - arm 1A (oral prednisolone)In 9 weeks to 5. Stop after 9 months.*

DAS=disease activity score; mg=milligram; N.A.=not applicable.

*GC tapered and stopped according to protocol, not depending on disease activity score.

Conclusion

The success rate of GC discontinuation after bridging as part of initial treatment of RA has been described in a limited number of studies. Reports on observational cohorts did not answer the research question and in clinical trials reports, GC (dis)continuation data were also scarce. However, the available data show that GC can be discontinued successfully in a large majority of patients. The paucity of data also reveals that more efforts are needed to provide data towards identifying the optimal GC bridging and discontinuation strategy, combining Treatment to Target with Starting to Stop.

Acknowledgements

We would like to thank J.W. Schoones for his help and expertise in the systematic literature search.

Disclosure of Interests

Lotte van Ouwerkerk: None declared, Andriko Palmowski: None declared, Isabell Nevins: None declared, Frank Buttgereit Consultant of: Consultant of AstraZeneca, AbbVie, Grünenthal, Horizon Pharma, Pfizer, and Roche., Grant/research support from: Grant/research support from AbbVie, Horizon Pharma, Pfizer, and Roche., Patrick Verschueren Consultant of: Was consultant for ABBVIE, BMS, Celltrion, Eli Lilly, Galapagos, Gilead, Nordic Pharma, Pfizer and UCB., Employee of: Holds the Pfizer Chair Early Rheumatoid Arthritis Management at KU Leuven., Josef Smolen: None declared, Robert B.M. Landewé Shareholder of: Shareholder of: Director of Rheumatology Consultancy BV., Consultant of: Consultant of: Honoraria from AbbVie, AstraZeneca, BMS, Boehringer Ingelheim, Celgene, Galapagos, Gilead, Glaxo-Smith-Kline, Janssen, Eli-Lilly, Novartis, Pfizer, UCB Pharma., Hans Bijlsma Consultant of: Consultant for Galapagos, Lilly and Sun., Grant/research support from: Received study grants from AbbVie and Roche., Andreas Kerschbaumer: None declared, Rene Westhovens Consultant of: Was consultant for Celltrion, Galapagos and Gilead., Thomas Huizinga: None declared, Cornelia Allaart Grant/research support from: Received study grants for BeSt and IMPROVED from Centocor Inc. (now Janssen) and AbbVie, respectively., Sytske Anne Bergstra Grant/research support from: Received an ASPIRE grant from Pfizer.

POS0916 A 10-YEAR JOURNEY OF CARING FOR PATIENTS WITH SYSTEMIC SCLEROSIS: FOLLOW-UP DATA ON DISEASE DURATION OF THE LEIDEN CCISS COHORT

Background

Combined Care in Systemic Sclerosis (CCISS) is a prospective cohort of patients referred to Leiden University Medical Center for Raynaud’s Phenomenon (RP), a suspicion of systemic sclerosis (SSc) or a connective tissue disease. This cohort is characterized by its standardized and extensive annual follow-up. Since initiation of the cohort in 2009, diagnostic criteria for SSc have been updated leading to a higher sensitivity for early SSc (ACR 2013 criteria). A recent Dutch study showed that there is a gap to time of diagnosis between men and women, despite of overall increase of SSc awareness (PMID: 31539063).

Objectives

To determine whether 1. time between first disease symptoms and diagnosis has changed over time, and 2. disease characteristics of SSc at first presentation in an expert clinic have changed over time for the total cohort, and between male and female patients.

Methods

Patients included in the CCISS cohort undergo annual evaluation and clinical, laboratory, and imaging variables are systematically recorded. For this study, patients fulfilling the ACR/EULAR 2013 SSc criteria were included, and categorized into three groups based on the cohort entrance year: 1) 2010 – 2013, 2) 2014 – 2017, and 3) 2018 – 2021. SSc patients with a baseline visit in 2009 (n=65) were excluded as these patients were often not newly referred. Disease duration was defined by months since first RP, since first non-RP symptom and months between first date of diagnosis by a physician and first non-RP symptom. Disease characteristics included presence of interstitial lung disease (ILD), pulmonary arterial hypertension, digital ulcers (DU), diffuse cutaneous SSc, anti-topoisomerase and anticentromere antibodies. At baseline, disease duration and disease characteristics were compared between the three groups using appropriate tests. In addition, disease duration was compared between males and females in the three groups.

Results

In total, 643 SSc patients were included of whom 229 (36%) had their baseline visit from 2010 until 2013, 207 (32%) from 2014 until 2017, and 207 (32%) from 2018 until 2021.

The proportion of female patients was significantly higher in the 2010 – 2013 group compared to the 2014 – 2017 and 2018 – 2021 group (Table 1). Over time, disease duration defined by RP duration and non-RP duration decreased as well as time between diagnosis and first non-RP symptom (Table 1). The proportion of patients presenting with ILD and DU was highest in the first group (Table 1).

Table 1.2010 - 2013 N=2292014- 2017 N=2072018 – 2021 N=207P-valueBASELINEAge, mean (SD)53 (15)57 (14)55 (14)0.003Female, %8676750.010RP duration, months (IQR)122 (46 – 240)93 (20 – 202)67 (20 – 210)0.003Non RP duration, months (IQR)43 (16 – 227)20 (5 – 112)17 (6 – 54)<0.001Diagnosis duration, months (IQR)116 (80 - 177)65 (45 – 105)25 (5 – 45)<0.001ΔRP and Non-RP, months (IQR)24 (0 – 99)18 (0 – 118)22 (0 – 120)0.337Anti-centromere antibodies, %3843490.092Anti-topoisomerase antibodies, %2424180.259Diffuse cutaneous SSc, %1923160.073Interstitial lung disease, %433131<0.001Pulmonary arterial hypertension, %3240.746Digital ulcers, %2013110.041

In both male and female SSc patients, disease duration and time between diagnosis and first non-RP decreased over time with a longer time in females for all durations which was significantly different for time between first RP and non-RP in 2014-2017 and 2018-2021 (Figure 1). For the 2018 – 2021 group, duration since diagnosis for female was 26 (4 - 46) and male 17 months (7 – 39; p=0.355), and time between RP and non-RP for female 24 (0 - 168) and male 12 months (0 – 48; p=0.029).

Figure 1.

Conclusion

Over time, we observe a decrease in disease duration and in SSc patients presenting with ILD or DU at cohort entrance. Our results indicate increased awareness of early SSc and identification of SSc patients before severe complications have occurred. At the same time our results show the urge for specific attention to improve timely diagnosis in female SSc patients.

Disclosure of Interests

None declared

POS0529 LONG-TERM LOCAL JOINT DAMAGE PROGRESSION IN RHEUMATOID ARTHRITIS IS RELATED TO CUMULATIVE LOCAL CLINICAL JOINT INFLAMMATION

Background

Previously we showed that joint inflammation in rheumatoid arthritis (RA) tends to recur in the same joint, suggesting local factors contributing to joint inflammation. In the same study population, we now investigated whether cumulative local joint inflammation is associated with local radiographic damage progression.

Objectives

To investigate whether there is an association between long-term cumulative joint swelling and progression of radiographic damage in treated-to-target patients with RA.

Methods

Data from the BeSt study were used, in which newly diagnosed patients with RA (ACR 1987 criteria) were treated-to-target (DAS ≤2.4) during 10 years. Local joint swelling (yes/no) was determined by clinical evaluation by trained nurses of all hand and foot joints at 3-monthly study visits. Yearly radiographs of hand and feet were scored for radiographic joint damage (Sharp-van der Heijde method) in random order by two independent readers who were blind for clinical results. Per joint, damage was expressed as the percentage of the maximum possible damage score, to account for differences in maximal scores per joint. Missing values were imputed using the last observation carried forward method. A generalized linear mixed model was used to assess the association between local joint swelling over time (i.e., percentage of study visits with observed local joint swelling) and degree of joint damage at the end of follow-up. Joints were clustered within patients. The model was adjusted for baseline damage and follow-up duration. To test the association between cumulative local joint swelling and joint damage as a local or a general inflammation effect, we did two analyses. First, we additionally adjusted the primary analysis for the mean disease activity score (DAS) over time. Second, we did a permutation test to study whether joint damage progression was better predicted by joint swelling in the joint itself than by joint swelling in randomly selected other joints, which is indicated by a p-value of <0.05.

Results

Of the 16,150 joints of 475 patients with at least one year follow-up with both radiographic and joint swelling assessment available, 16% (2,564) had radiographic joint damage (damage score ≥ 0.5) at the end of follow-up. Median (IQR) follow-up time was 10 (6-10) years. Of the joints with damage at the end of follow-up, 46% (1,163) was swollen at baseline, versus 36% (4,818) of the joints without damage. The median (IQR) percentage of visits at which joint swelling was observed was 6 (0-17) and 3 (0-8) for joints with and without joint damage respectively.

We found a β of 0.13 (95% CI 0.12 to 0.14) for the association between cumulative local joint swelling and local progression, that is, with each 1% increase in the number of visits with local joint swelling, local radiographic joint damage progression on average increased with 0.13 percent. In an analysis with 10-years completers only (both baseline and year 10 damage score available, n = 9,520) we also found an association between cumulative local joint swelling and local radiographic damage (β 0.24, 95% CI 0.22 to 0.26). The association was also found in a subset of joints that were swollen at least once (β 0.20, 95% CI 0.18 to 0.22), indicating that joint damage is not only associated with ever-occurrence but also with the frequency of joint swelling.

This association was found for both erosions (β 0.07, 95% CI 0.07 to 0.08) and joint space narrowing (β 0.21, 95% CI 0.19 to 0.22). The results of the primary analysis did not change after adjustment for DAS over time. The permutation test showed that local joint damage progression was better predicted by the frequency of joint swelling of that joint, than by joint swelling frequency of other joints (p<0.001). These results indicate a local, rather than a general, inflammation effect.

Conclusion

Cumulative local joint swelling over time is associated with joint damage progression in the same joint in treated-to-target (DAS ≤2.4) patients with RA. Our results indicate that this is a local effect rather than an effect of general disease activity.

Disclosure of Interests

Sascha Heckert: None declared, Sytske Anne Bergstra: None declared, Yvonne Goekoop-Ruiterman: None declared, Melek Güler-Yüksel: None declared, WIllem Lems: None declared, M. van Oosterhout: None declared, Thomas Huizinga: None declared, Cornelia Allaart Grant/research support from: The original BeSt study was funded by a research grant from the Dutch College of Health Insurances with additional funding from Schering-Plough BV and Centocor Inc.

POS0579 ABSENCE OF ASSOCIATION BETWEEN ABATACEPT EXPOSURE LEVELS AND INITIAL INFECTION IN PATIENTS WITH RA: A POST HOC ANALYSIS OF THE RANDOMIZED, PLACEBO-CONTROLLED AVERT-2 STUDY

Background

Infections are the most commonly reported AE observed in patients with RA treated with immunosuppressive therapies and can be clinically significant. A recent review reported differences in the risk of infection for some biologics such as tocilizumab and TNF inhibitors.1 Abatacept selectively modulates T-cell co-stimulation and is approved for the treatment of RA. In patients with polyarticular-course juvenile idiopathic arthritis, no association was found between higher serum abatacept exposure and the incidence of infection.2 This has not been evaluated for adult patients with RA.

Objectives

To determine if higher serum abatacept exposure during treatment with SC abatacept was associated with increased risk of infection in adult patients with RA.

Methods

AVERT-2 (Assessing Very Early Rheumatoid arthritis Treatment-2) was a randomized, placebo-controlled study of SC abatacept + MTX vs abatacept placebo + MTX in MTX-naive, anti-citrullinated protein antibody–positive patients with early, active RA.3 A post hoc population pharmacokinetic (PK) analysis was performed using PK-evaluable patient data from the induction period (year 1) of AVERT-2. Association between steady-state abatacept exposure (min plasma concentration [Cmin], max plasma concentration [Cmax], and average plasma concentration [Cavg]) and first infection was evaluated using Kaplan–Meier plots of probability vs time on treatment by abatacept exposure quartiles and Cox proportional-hazards models.

Results

PK of SC abatacept was defined as a linear 2-compartment model with first-order absorption and first-order elimination. The findings of the updated PK analysis were consistent with those reported in prior population analyses of abatacept PK in adults with RA. The final model included effects of baseline body weight, estimated glomerular filtration rate, sex, age, albumin, MTX use, NSAID use, SJC, and race on abatacept clearance. The only covariate with a clinically relevant effect was higher body weight, which caused an increase in clearance and volume. Infections occurred in a total of 330/693 (47.6%; serious, 1.6%) patients treated with abatacept, and 134/301 (44.5%; serious, 1.3%) with placebo during the first year of AVERT-2. In patients taking abatacept, the mean (SD) study exposure to abatacept was 376 (60) days, while mean (SD) prednisone equivalent dose was 6.7 (3.8) mg/day and mean (SD) MTX dose was 9.6 (3.0) mg/week. No exposure–response relationship was observed between the probability of first infection and steady-state abatacept exposure quartiles (Cavg, Cmin, and Cmax), or compared with placebo (Figure 1A–C). Kaplan–Meier assessment also showed no increase in risk of infection with concomitant use of MTX and glucocorticoids.

Conclusion

No association was found between initial infection and steady-state abatacept exposure (Cavg, Cmin, Cmax) or MTX and glucocorticoid use in patients with RA treated with SC abatacept.

References

[1]Jani M, et al. Curr Opin Rheumatol 2019;31:285–92.

[2]Ruperto N, et al. J Rheumatol 2021;48:1073–81.

[3]Emery P, et al. Arthritis Rheumatol 2019;71(suppl 10):L11.

Acknowledgements

This study was sponsored by Bristol Myers Squibb. Writing and editorial assistance were provided by Fiona Boswell, PhD, of Caudex, and was funded by Bristol Myers Squibb. Support was provided by Sandra Overfield as Protocol Manager, and Prema Sukumar and Renfang Hwang as Data Science Leads.

Disclosure of Interests

Paul Emery Consultant of: AbbVie, AstraZeneca, Boehringer Ingelheim, Bristol Myers Squibb, Eli Lilly, Galapagos, Gilead, Janssen, Merck Sharp & Dohme, Novartis, Pfizer, Roche, Samsung, Grant/research support from: AbbVie, Bristol Myers Squibb, Eli Lilly, Novartis, Pfizer, Roche, Samsung, Roy Fleischmann Consultant of: Amgen, AbbVie, Bristol Myers Squibb, Gilead, GlaxoSmithKline, Novartis, Pfizer, Grant/research support from: Amgen, AbbVie, Arthrosi, Biosplice, Bristol Myers Squibb, Gilead, GlaxoSmithKline, Horizon, Novartis, Pfizer, Regeneron, TEVA, UCB, Robert Wong Shareholder of: Bristol Myers Squibb, Employee of: Bristol Myers Squibb, Karissa Lozenski Shareholder of: Bristol Myers Squibb, Employee of: Bristol Myers Squibb, Yoshiya Tanaka Speakers bureau: AbbVie, Amgen, Astellas, AstraZeneca, Boehringer Ingelheim, Bristol Myers Squibb, Chugai, Eisai, Eli Lilly, Gilead, Mitsubishi Tanabe, YL Biologics, Consultant of: AbbVie, Ayumi, Daiichi Sankyo, Eli Lilly, GlaxoSmithKline, Taisho, Sanofi, Grant/research support from: AbbVie, Asahi Kasei, Boehringer Ingelheim, Chugai, Corrona, Daiichi Sankyo, Eisai, Kowa, Mitsubishi Tanabe, Takeda, Vivian Bykerk Consultant of: Amgen, Bristol Myers Squibb, Genzyme Corporation, Gilead, Regeneron, UCB, Grant/research support from: Amgen, Bristol Myers Squibb, Genzyme Corporation, Pfizer, Regeneron, Sanofi Aventis, UCB, Clifton Bingham Consultant of: AbbVie, Bristol Myers Squibb, Eli Lilly, Janssen, Pfizer, Sanofi, Grant/research support from: Bristol Myers Squibb, Thomas Huizinga Speakers bureau: Abblynx, Abbott, Biotest AG, Bristol Myers Squibb, Crescendo Bioscience, Eli Lilly, Epirus, Galapagos, Janssen, Merck, Novartis, Pfizer, Roche, Sanofi- Aventis, UCB, Consultant of: Abblynx, Abbott, Biotest AG, Bristol Myers Squibb, Crescendo Bioscience, Eli Lilly, Epirus, Galapagos, Janssen, Merck, Novartis, Pfizer, Roche, Sanofi- Aventis, UCB, Grant/research support from: Abblynx, Abbott, Biotest AG, Bristol Myers Squibb, Crescendo Bioscience, Eli Lilly, Epirus, Galapagos, Janssen, Merck, Novartis, Pfizer, Roche, Sanofi- Aventis, UCB, Gustavo Citera Speakers bureau: AbbVie, Amgen, Bristol Myers Squibb, Eli Lilly, Janssen, Pfizer, Sandoz, Consultant of: AbbVie, Amgen, Bristol Myers Squibb, Pfizer, Grant/research support from: Pfizer, Yedid Elbez Consultant of: Bristol Myers Squibb, Employee of: Signifience, Vidya Perera Shareholder of: Bristol Myers Squibb, Employee of: Bristol Myers Squibb, Bindu Murthy Shareholder of: Bristol Myers Squibb, Employee of: Bristol Myers Squibb, Kelly Maxwell Consultant of: Bristol Myers Squibb, Employee of: Cognigen Corporation, Julie Passarell Consultant of: Bristol Myers Squibb, Employee of: Cognigen Corporation, William Hedrich: None declared, Daphne Williams Consultant of: Black Diamond Network, Joule, Syneos, Employee of: Bristol Myers Squibb.

POS0570 UNSUPERVISED DEEP-LEARNING IDENTIFIES SIX CLINICAL SUBTYPES OF RHEUMATOID ARTHRITIS AT BASELINE THAT ASSOCIATE WITH METHOTREXATE FAILURE DURING FOLLOW-UP

Background

Currently methotrexate (MTX) is the prevailing baseline treatment for Rheumatoid Arthritis (RA). Treatment response varies since RA is a highly heterogeneous disease. A tantalizing idea is that the factors causing clinical heterogeneity can already be elucidated at baseline.

Objectives

Disentangle clinical heterogeneity of RA patients at baseline to identify likely MTX failure during follow-up.

Methods

We constructed patient-specific profiles, featuring baseline clinical measurements which we split into three layers: 1) joint counts, 2) numerical hematology work up, and 3) categorical features (binary serological markers (aCCP/RF) and localization of joint inflammation and tenderness). We applied Z-score scaling on the numerical data and one hot encoding on the categorical features. To identify hidden structure across these layers we used Maui (Multi-omics Autoencoder Integration) [1], and Phenograph [2] for subsequent clustering of patients within the extracted latent space. We examined the most discriminatory features post-hoc with SHAP. With Kaplan Meier curves we assessed MTX efficacy using treatment switch as proxy for failure. We calculated hazard ratios (HR) with univariate Cox-regression.

Results

We had 944 RA patients with baseline health record data. MAUI identified 23 latent factors from 335 baseline variables. Phenograph showed 6 RA-subgroups (Table 1 & Figure 1).

Table 1.

Baseline characteristics of the different clusters

C1C2C3C4C5C6N22417917116211692Sex (F)*131 (59)122 (68)125 (73)112 (69)71 (61)57 (62)RF*95 (42)89 (50)106 (62)107 (66)68 (59)55 (60)aCCP*82 (36.6)89 (50)110 (64)105 (65)65 (56)54 (59)DAS44(3)3.6 (2.7-4.2)2.7 (2.2-3.1)2.4 (1.9-2.9)2.1 (1.7-2.6)2.2 (1.7-2.6)2.8 (2.4-3.2)SJC15 (11-20)6 (3-8)9.0 (6-12)2 (1-5)4 (2-6)9 (6-12)TJC19 (14-27)9 (6-12)12.0 (9-18)4 (2-6)3 (2-6)11 (7-13)ESR (mm/hr)33 (14-53)33 (14-48)19 (9-35)28 (14-39)23 (11-38)25 (9-36)Age (yr)63 (14)60 (13)53 (16)59 (15)63 (13)58 (16)MTX prescription*192 (85)146 (81)138 (80)131 (80)88 (75)78 (85)Follow up (days)1308 (743-2060)1458 (880-2567)1821 (982-2566)1590 (1022-2245)1566 (787-2000)1468 (832-2211)Symptom duration (days)124 (52-334)155 (46-537)155 (62-365)217 (77-775)186 (62-548)155 (62-365)

Presented are *binary variables as n(%) and continuous as median (Q1-Q3) or mean (SD).

Figure 1.

Overview of the distinct RA-clusters: A) 2D UMAP, B) Kaplan Meier plot of MTX-probability across 8.6 years (defined by cluster with shortest follow up), C) SHAP plot of most discriminatory features per cluster.

The baseline clusters (C) are characterized by a different joint involvement or lab value: C1 had a low aCCP-positivity (37%) and high median ESR of 33. C1 had the most affected joints (primarily the small joints) with a swollen- (SJC) and tender joint count (TJC) of 15 and 19 respectively. C2 had mediocre aCCP-positivity (50%) and low median SJC=6 and TJC=9. C3 had MTP involvement, high aCCP-positivity (64%), and low ESR=19 but relatively high joint counts; SJC=9 and TJC=12. C4 had no wrists, high aCCP-positivity (65%), high ESR=28 and low joint counts; SJC=2, TJC=4. C5 had low lymphocyte numbers and a low median ESR=23, SJC=4 and TJC=3. C6 had MCP1 involvement, was mostly aCCP-positive (59%) and had a slightly higher median ESR=25, SJC=9 and TJC=11.

Clusters differed in MTX failure: 40%, 53%, 69%, 54%, 48% and 64% (for cluster 1-6, P=3.2e-2). Examining the local differences, we observed the biggest difference between C1 and C3 (HR 0.5 (95% CI 0.36-0.7), P=4.4e-5).

Conclusion

Using baseline data, we identified 6 putative novel RA subtypes which were associated with differences in MTX failure. Our study demonstrates the applicability of unsupervised deep learning and cluster analysis to elucidate hidden structure in the multi-modal EHR.

References

[1]Ronen J. doi:10.26508/lsa.201900517

[2]Levine JH. doi:10.1016/j.cell.2015.05.047

Disclosure of Interests

Tjardo Maarseveen: None declared, Marc Maurits: None declared, Thomas Huizinga: None declared, Marcel Reinders: None declared, Erik van den Akker: None declared, Rachel Knevel Grant/research support from: Rachel received a grant from Pfizer.

OP0086 ANTIBODIES AGAINST ADVANCED GLYCATION END-PRODUCTS (ANTI-AGE) DISTINGUISH PATIENTS WITH A MORE INFLAMMATORY PROFILE AND WORSE OUTCOME IN SERONEGATIVE RHEUMATOID ARTHRITIS

Background

In rheumatoid arthritis (RA) around two-thirds of patients are autoantibody-positive for rheumatoid factor, anti-citrullinated protein antibodies (ACPA) and/or anti-carbamylated protein antibodies (anti-CarP). The remaining seronegative subgroup of RA is clinically heterogeneous and thus far, biomarkers predicting the disease course in these patients are lacking. Therefore, we set out to investigate the value of a new/different autoantibody in rheumatoid arthritis directed against advanced glycation end-product (AGE) modified proteins (anti-AGE). AGEs are a marker of oxidative stress, and anti-AGE have been described in multiple diseases including diabetes, and hypertension.

Objectives

To investigate the prevalence of anti-AGE in RA and non-RA arthritis patients, as well as their association with clinical parameters and disease outcome in RA.

Methods

In 648 RA patients and 538 non-RA arthritis patients from the Leiden Early Arthritis Clinic anti-AGE IgG antibody levels were measured using an in-house fetal calf serum (FCS) ELISA based assay using native FCS as control. The cutoff for positivity was set as the mean optical density plus two times the standard deviation of 80 healthy controls. Radiological progression was measured with the with Sharp van der Heijde score (SHS) on yearly basis and the association with anti-AGE was assessed with a multivariate normal regression model.

Results

Anti-AGE was found in 299 (46%) of RA patients versus 163 (30%) of non-RA arthritis patients. Interestingly, 67 (34%) of completely seronegative (RF-, ACPA- and anti-CarP-negative) RA patients were positive for anti-AGE. Within RA, anti-AGE-positive patients had significantly higher ESR (median anti-AGE-positive: 38, anti-AGE-negative: 32, p<0.001) and CRP (median anti-AGE-positive: 19, anti-AGE-negative: 17, p<0.001), indicating an more inflammatory profile in these patients. Radiographic progression, was significantly higher in anti-AGE+ patients (Figure 1A, B=1.05, p<0.001). Since ACPA and anti-CarP (in ACPA-patients) are associated with radiological progression, the analysis was first stratified for ACPA-status. In the anti-AGE+ACPA- patients a significant association with SHS was found (Figure 1B: B=1.04, P<0.001), indicating that anti-AGE is associated with radiological progression in ACPA-negative patients. Next, the ACPA-negative stratum was also stratified for anti-CarP. Interestingly, SHS was significantly higher in all anti-AGE+ groups compared to the autoantibody-negative group.

Conclusion

Almost half of the RA patients are anti-AGE positive including a substantial part of otherwise completely seronegative RA patients. Anti-AGE antibodies are associated with inflammatory parameters and radiologic progression in seronegative RA patients. Therefore, although these autoantibodies are not specific for RA, anti-AGE could potentially identify patients with a more inflammatory phenotype and more severe disease outcome in “classic” autoantibody-negative RA patients.

Disclosure of Interests

None declared.

POS0104 ACPA-NEGATIVE AND ACPA-POSITIVE RA-PATIENTS ACHIEVING DISEASE RESOLUTION DEMONSTRATE DISTINCT PATTERNS OF MRI-DETECTED JOINT-INFLAMMATION

Background

Sustained DMARD-free remission (SDFR), the sustained absence of clinical-synovitis after DMARD-discontinuation, is increasingly achievable in RA. However, prevalence differs significantly between ACPA-negative (40%) and ACPA-positive RA (5-10%). In addition, early-DAS-remission (DAS4months<1.6) associates with SDFR in ACPA-negative RA but not in ACPA-positive RA.1 Based on these differences, we hypothesized that longitudinal patterns of local tissue-inflammation (synovitis/tenosynovitis/osteitis) might also differ between ACPA-negative and ACPA-positive RA-patients achieving SDFR.

Objectives

With the ultimate aim to increase understanding of disease-resolution in RA, we studied MRI-detected joint-inflammation over time in relation to SDFR-development in ACPA-negative RA and ACPA-positive RA.

Methods

198 RA-patients (94 ACPA-negative and 104 ACPA-positive) underwent repeated MRIs (0/4/12/24-months) and were followed on SDFR-development. The course of MRI-detected total-inflammation, and synovitis/tenosynovitis/osteitis individually, were compared between RA-patients who did and did not achieve SDFR, using Poisson-mixed-models. 170 ACPA-positive RA-patients from the AVERT-1 were studied as ACPA-positive validation-population.

Results

In ACPA-negative RA, patients achieving SDFR had similar baseline total inflammation-levels, which declined 2.0-times stronger in the first-year of DMARD-treatment (IRR 0.50, 95%CI;0.32-0.77, p<0.01) compared to patients without SDFR. This stronger decline was seen in tenosynovitis/synovitis/osteitis. In contrast, ACPA-positive RA-patients achieving SDFR, had already lower inflammation-levels (especially synovitis/osteitis) at disease-presentation (IRR 0.45, 95%CI;0.24-0.86, p=0.02) compared to non-SDFR patients, and remained lower during follow-up (p=0.02). Similar results were found in the ACPA-positive validation-population.

Conclusion

Compared to RA-patients without disease-resolution, ACPA-positive RA-patients achieving SDFR have less severe joint-inflammation from diagnosis onwards, whilst ACPA-negative RA-patients present with similar inflammation-levels but demonstrate a stronger decline in the first year of DMARD-therapy. These different trajectories suggest that mechanisms underlying resolution of RA-chronicity in both RA-subsets might be different and indicates the relevance of the total inflammatory-load in ACPA-positive-RA.

References

[1]Verstappen M, Niemantsverdriet E, Matthijssen XME, le Cessie S, van der Helm-van Mil AHM. Early DAS response after DMARD-start increases probability of achieving sustained DMARD-free remission in rheumatoid arthritis. Arthritis Res Ther. 2020 Nov 23;22(1):276.

Figure 1.

Patterns of MRI-detected joint-inflammation in RA-patients achieving SDFR compared to those who did not, stratified for ACPA-status

Disclosure of Interests

Marloes Verstappen: None declared, Xanthe Matthijssen: None declared, Sean Connolly Shareholder of: Dr. Sean E. Connolly, Ph.D. is a shareholder of Bristol Myers Squibb, Employee of: Dr. Sean E. Connolly, Ph.D. is an employee of Bristol Myers Squibb, Michael A Maldonado Shareholder of: Dr. M. Maldonado, Ph.D. is a shareholder of Bristol Myers Squibb, Employee of: Dr. M. Maldonado, Ph.D. is an employee of Bristol Myers Squibb, Thomas Huizinga Speakers bureau: Tom WJ Huizinga/the department of rheumatology LUMC has received research support/lecture fees/consultancy fees from Merck, UCB, Bristol Myers Squibb, Janssen, Pfizer, Novartis, Sanofi-Aventis, Galapagos, Boeringher and Eli Lilly, Consultant of: Tom WJ Huizinga/the department of rheumatology LUMC has received research support/lecture fees/consultancy fees from Merck, UCB, Bristol Myers Squibb, Janssen, Pfizer, Novartis, Sanofi-Aventis, Galapagos, Boeringher and Eli Lilly, Grant/research support from: Tom WJ Huizinga/the department of rheumatology LUMC has received research support/lecture fees/consultancy fees from Merck, UCB, Bristol Myers Squibb, Janssen, Pfizer, Novartis, Sanofi-Aventis, Galapagos, Boeringher and Eli Lilly, Annette van der Helm-van Mil: None declared

POS1407 THE MANAGEMENT OF CONTEMPORARY EARLY UNDIFFERENTIATED ARTHRITIS: DATA ON EULAR’S RECOMMENDATION ON THE RISK OF PERSISTENT DISEASE

Background

EULAR recommendations for early arthritis recommend assessing the following risk factors for persistent disease in Undifferentiated Arthritis (UA): polyarthritis, acute phase reactants (i.e. c-reactive protein(CRP), erythrocyte sedimentation rate(ESR)), rheumatoid factor (RF), ACPA and imaging findings/erosions. However, these recommendations are based on an outdated definition of UA: not fulfilling 1987 RA criteria and having no alternative diagnosis (‘conventional UA’). Since the introduction of the 2010 RA criteria the characterization of UA has changed. The contemporary definition of UA is: not fulfilling 1987- nor 2010 criteria, and having no alternative diagnosis. Therefore, predictors for persistent disease in conventional UA may not be applicable in contemporary UA.

Objectives

Our objective was to assess which risk factors for persistent disease, as mentioned in the EULAR recommendations, are applicable in contemporary UA.

Methods

Patients consecutively included in the Leiden Early Arthritis Clinic cohort with contemporary UA between 2006-2019, when DMARD start in UA was recommended, were studied. The outcome was sustained remission, defined as absence of clinical synovitis without DMARD use (including corticosteroids) for the entire follow-up (at least one year). Cox regression was used to test the association between the risk factors as mentioned in EULAR-guidelines and sustained remission. For comparison, patients with conventional UA (not fulfilling the 1987 criteria and no other diagnosis) were studied using similar analyses.

Results

Contemporary UA patients (n=710) were mostly ACPA negative (95%) and had a median of 2 swollen joints. Radiographic erosions were rare (1.8%). 60% of contemporary UA patients achieved sustained remission after median 1.5 years (IQR 1-3). After achieving remission patients were followed for 5.5 years without clinical arthritis, demonstrating sustainability. Univariably, CRP, ESR, ACPA and RF were associated with time to sustained remission, while polyarthritis was not. In multivariable analysis, only ACPA and CRP were independently associated with sustained remission (HR 0.10 (95%CI:0.03-0.32) and HR 0.67(0.50-0.91), respectively). 67% of contemporary UA patients had none of these risk factors. In contrast, only 2% had both CRP and ACPA. For comparison, multivariable analysis in conventional UA patients showed that ACPA, RF, CRP and polyarthritis were all independently associated with sustained remission.

Conclusion

The contemporary UA population is different from conventional UA and risk factors for disease persistence are partly dissimilar. ACPA and CRP remain to be predictive in contemporary UA. Other factors included in the current EULAR recommendation were uninformative (RF, ESR, polyarthritis) or rare (erosions). Therefore, risk factors recommended in future EULAR recommendations may require alterations.

Disclosure of Interests

None declared

AB0675 Health related quality of life in patients with mixed connective tissue disease: A comparison with matched systemic sclerosis patients

Background

Mixed connective tissue disease (MCTD) is a systemic auto-immune disorder, being probably the least common among the connective tissue diseases. Symptoms can be severe and could affect health-related quality of life (HRQoL). Identification of the burden of MCTD patients is of key importance to provide appropriate pharmacological and non-pharmacological care. No reports on HRQoL have been published in adult patients with MCTD.

Objectives

To perform an explorative study to evaluate HRQoL and its main determinants in MCTD patients, and compare HRQoL between MCTD and matched systemic sclerosis (SSc) patients.

Methods

MCTD patients fulfilling the Kahn criteria and participating in the MCTD prospective follow-up cohort of the Leiden University Medical Center were included. In addition, SSc patients matched for age, gender and disease duration were included for comparison. Data on disease characteristics, functional disability and HRQoL were collected annually for both disease groups. HRQoL was evaluated using the 36-Item Short Form Health Survey (SF36) and EuroQol (EQ5D). At baseline, HRQoL, as reflected by SF36 mental component score (MCS), SF 36 physical component score (PCS) and EQ5D were compared between MCTD and SSc patients. For MCTD patients, factors associated with HRQoL at baseline were identified using linear regression and change in HRQoL over 3 years was evaluated using linear mixed models. In addition, characteristics of MCTD patients who showed worsening of MCS and/or had PCS superior to the minimal clinical important difference of three points were identified.

Results

Thirty-four MCTD patients (121 visits; 82% female, mean age 42 years, median disease duration 45 months) and 102 SSc patients (424 visits; 82% female, mean age 45 years, median disease duration 49 months) were included. At baseline, MCTD-patients more often had ILD (47% vs. 34%, p=0.027), cardiac involvement (30% vs. 2%, p<0.001), synovitis (26% vs. 11%, p=0.004) and myositis (15% vs. 1%, p=0.001) compared to SSc patients, whereas SSc patients more often used immunosuppressive treatments except for hydroxychloroquine (MCTD:18% vs. SSc:7%, p=0.007).

Baseline HRQoL in MCTD was comparable to HRQoL in SSc, with mean SF36-PCS of 40.2 (SD:9.1) and mean SF36-MCS of 44.9 (SD:9.9), which is (nearly) one standard deviation lower than the general Dutch population. The SF36 subscore “general health perception” was the most impacted in both groups (MCTD: 38.5 [SD:7.0], SSc: 39.9 [SD:8.9]). The median EQ5DNL was 0.38 (IQR:0.14 – 0.54) and comparable between SSc and MCTD.

At baseline, in MCTD, ILD was significantly associated with SF36-PCS (β:6.98, 95% CI: 1.10 to 12.86) and SF36-MCS (β:-8.10, 95% CI:-14.93 to -1.26). Sclerodactyly was significantly associated with EQ5DNL (β:0.006; 95% CI:0.002 to 0.010) and SF36-PCS (β:0.12, 95% CI:0.03 to 0.21). No other significant associations were identified.

Over time, in MCTD, both the SF36-MCS and SF36-PCS improved significantly (MCS: β:2.35/year [95% CI:0.58 to 4.13], PCS: β:1.34/year [95% CI:0.03 to 2.65), whereas EQ5DNL was stable. Explorative analyses did not reveal a specific clinical characteristic with significant impact on the change of HRQoL over time. With an MCID of 3 points on the MCS and PCS, 7 MCTD-patients worsened on the MCS and 3 on the PCS. Patients who showed worsening of MCS over time tended to be older, more often had ILD, sclerodactyly and GI complaints, and had worse exercise tolerance. All these differences did not reach statistical significance. The patients who decreased PCS more often had ILD (100% vs. 41%, p=0.015), and used glucocorticoids more often (33% vs. 0%, p=0.046), were slightly older and had a worse exercise tolerance as compared to those who showed a stable/improving PCS over time.

Conclusion

Like in SSc, HRQoL is significantly impaired in MCTD, especially the general health perception of patients. Cardiac involvement, ILD, age and worse functional disability might specifically impact HRQoL in MCTD. However, these associations need further evaluations in larger cohorts.

Disclosure of Interests

None declared

POS0505 NO EXCESS MORTALITY IN CONTEMPORARY UNDIFFERENTIATED ARTHRITIS, IN CONTRAST TO RHEUMATOID ARTHRITIS; A STUDY WITH A FOLLOW-UP OF AT LEAST 10 YEARS

Background

Undifferentiated arthritis (UA) is traditionally considered an early phase of rheumatoid arthritis (RA),1 and management strategies of RA are often transferred to UA-patients assuming outcomes are comparable. Importantly, the UA population changed with the introduction of 2010 RA criteria as part of the conventional UA population became classified as RA. Consequently, contemporary UA-patients (not fulfilling the 1987 nor 2010 RA-criteria) are largely autoantibody negative and present with relatively mild disease.2 Subsequently, contemporary UA might no longer represent a group of patients in a pre-phase of RA but a population with intrinsically different characteristics than RA. We therefore hypothesized that mortality, a long term outcome which is increased in RA, is not increased in contemporary UA.

Objectives

To asses mortality rates in contemporary UA-patients, and to compare this to RA and conventional UA during at least 10 years of follow up.

Methods

We studied 860 conventional UA-patients (no 1987 RA-criteria or other diagnosis), 561 contemporary UA-patients (no 1987/2010 RA-criteria or other diagnosis) and 762 RA-patients according to the 1987 criteria and 826 patients fulfilling the 2010 criteria, included in the Leiden Early Arthritis Clinic (EAC) between 1993-2008 to ascertain 10-years of follow-up. Mortality data were obtained from the civic registries (June 2018). Standardized-mortality-ratios (SMRs) were adjusted for age, gender and calendar-year, and additionally stratified for ACPA-positivity (anti-CCP2, cut-off>10 mg/L) and DMARD-treatment.

Results

The contemporary UA-population was for 93% ACPA-negative and the SJC was 1 (median) and TJC 2. This presentation was milder than that of patients with conventional UA. In RA (both 1987-RA and 2010-RA) we observed excess mortality compared to the general population (Figure 1). In conventional UA, a trend towards excess mortality was observed: SMR 1.11 (95%CI:0.96-1.27). However, part of this UA population is currently classified as 2010-RA. Subsequently, in contemporary UA, which does not include these 2010 RA-patients, no excess mortality was observed: SMR 1.05 (95%CI:0.87-1.26). ACPA-stratification suggested excess mortality in ACPA-positive contemporary UA, but not in ACPA-negative contemporary UA (Figure 1). Mortality rates in contemporary UA-patients who did and did not receive DMARD-treatment were comparable (SMR 0.93 (95%CI:0.66-1.24) and SMR 1.08 (95%CI:0.82-1.36)).

Figure 1.

SMRs in UA and RA-patients with minimally 10-years of follow-up. Additionally, mortality was stratified for ACPA-status. In RA, SMRs were also stratified for early treat-to-target DMARD-therapy, which has been shown relevant for mortality analysis in RA.3ACPA: anticitrullinated autoantibody protein, UA: undifferentiated arthritis, RA: rheumatoid arthritis, SMR: Standardized mortality rate, IT: intensive treatment.

Conclusion

Contemporary UA has no excess mortality, in contrast to RA. So, besides milder disease at presentation within contemporary UA, also the long-term outcome mortality is more favorable and comparable to the general population. This supports the notion that contemporary UA-patients have intrinsically different characteristics than RA-patients, rather than representing an early stage of RA. Future studies are warranted to determine whether contemporary UA should be treated as RA; our results suggest that this population may deserve separate guidelines.

References

[1]T.W.J. Huizinga et al. Criteria for early rheumatoid arthritis: From Bayes’ law revisited to new thoughts on pathogenesis. Arthritis & Rheumatism 2002.

[2]Verstappen et al. Undifferentiated arthritis: a changing population who did not benefit from enhanced DMARD-strategies – results from a 25-years longitudinal inception cohort. Oxford rheumatology 2021.

[3]Matthijssen et al. Early intensive treatment normalizes excess mortality in ACPA-negative RA, but not in ACPA-positive RA. Ann Rheum Dis 2020:

Disclosure of Interests

None declared.

AB0071 ANTI-ACETYLATED PROTEIN IgM-ANTIBODIES AS THE STARTING POINT OF AUTOANTIBODY FORMATION IN RHEUMATOID ARTHRITIS?

Background

Anti-modified protein antibodies (AMPA) are an important hallmark of rheumatoid arthritis (RA) and the fact that they have consistently been found to develop years before disease onset, has provided important insights into the immunopathology underlying RA. In (auto)antibody development, IgM is the first isotype to be produced. However, it is unclear whether IgM-autoimmunity differs between AMPA targeting different post-translational modifications (citrulline, homocitrulline and acetylated residues), which could provide clues about the starting point of the AMPA response.

Objectives

We therefore investigated IgM-levels of anti-citrullinated protein antibodies (ACPA), anti-carbamylated protein antibodies (anti-CarP) and anti-acetylated protein antibodies (AAPA) in healthy individuals, non-RA and RA patients.

Methods

Autoantibodies were investigated in 2 cohorts:

1) a Japanese cohort of healthy individuals (community based Nagasaki Island study) known to be ACPA-positive (n=65) or ACPA-negative (n=197) were compared to Dutch healthy donors (n=30) and ACPA-positive RA patients (n=29). ACPA, anti-CarP and AAPA IgG were measured by ELISAs using CCP4, CHcitP4 and CAcetylP4 peptides with sequences similar to the commercial CCP2 antigen and native control peptides.

2) early arthritis patients from the Leiden Early Arthritis Clinic who had RA (n=648) or another form of arthritis (non-RA, n=555) and healthy controls (n=80). ACPA and AAPA were determined by ELISAs using CCP2 and cACP2 peptides and their native control peptides, while anti-CarP was measured on homocitrullinated versus native fetal calf serum. Mann-Whitney U-tests were performed for statistical comparisons.

Results

ACPA IgM reactivity was mainly present in established ACPA-positive RA patients (Figure 1A) and to a lesser extent in ACPA-positive healthy Japanese individuals, and non-RA arthritis patients (Figure 1D). A similar picture was observed for anti-CarP IgM reactivity, for which again highest levels were found in established RA patients (1B and 1E) and ACPA-positive compared to ACPA-negative healthy Japanese individuals (1B). Intriguingly, AAPA IgM reactivity-levels displayed a different pattern as these were comparable between healthy individuals and ACPA-positive RA patients (1C and 1F). Likewise, AAPA IgM reactivity-levels were also not increased in ACPA-positive healthy Japanese individuals, who instead had lower levels compared to their ACPA-negative counterparts. Furthermore, the AAPA IgM-reactivity levels did not differ between non-RA arthritis patients, healthy controls and RA patients (1F). AAPA IgG-levels on the other hand were clearly elevated in RA patients compared to healthy controls and non-RA arthritis patients (1G).

Conclusion

AAPA are exceptional compared to other AMPA because IgM AAPA-levels are similar among healthy individuals, non-RA arthritis and RA patients. This suggests that AAPA IgM is part of the “normal” immune repertoire and could constitute the starting point for RA-associated AMPA responses, with isotype switching and epitope spreading to other post-translational modifications leading to the typical RA-associated mature AMPA response.

Disclosure of Interests

None declared

AB0077 IN RHEUMATOID ARTHRITIS PATIENTS, TOTAL IgA1 AND IgA2 LEVELS ARE ELEVATED: IMPLICATIONS FOR THE MUCOSAL ORIGIN HYPOTHESIS

Background

Mucosal surfaces may be involved in the pathophysiology of rheumatoid arthritis (RA) (1). IgA is the most abundant class of immunoglobulin at mucosal sites. Therefore, it is worthwhile to study this isotype in RA patients in more detail. Humans have two IgA subclasses, IgA1 and IgA2, which are not evenly distributed. IgA1 is dominant in serum, whereas IgA1 and IgA2 are more balanced at mucosal surfaces (2). Besides these differences in location, IgA2 has also been ascribed pro-inflammatory properties (3).

Objectives

As IgA subclasses might provide new insights into mucosal involvement and potential pro-inflammatory mechanisms, we investigated total and autoantibody-specific IgA subclasses responses in sera of rheumatoid arthritis patients.

Methods

Sera from two cohorts of RA patients, the IMPROVED (baseline visit) and the EAC (1-year visit), were selected based on previous autoantibody measurements. All patients fulfilled the 1987 (EAC) or 2010 (IMPROVED) ACR criteria for RA. Total IgA subclasses and IgA subclasses of rheumatoid factor (RF) and anti-citrullinated protein antibodies (ACPA) were measured using (in-house) ELISA’s, and compared to healthy donors. Associations between these IgA subclass levels and markers of inflammation (CRP and disease activity score (DAS)) were investigated using Spearman’s rank correlation. Mann–Whitney U tests were performed to investigate the association between IgA1 and IgA2 levels and smoking, a proxy for chronic mucosal inflammation. To correct for confounders, a multivariate linear model including age, gender, CRP and smoking was used.

Results

Total IgA1 and IgA2 levels were increased in RA patients compared to healthy donors in both cohorts (Figure 1A-C, data IMPROVED). This increase was more pronounced in seropositive RA versus seronegative RA. Both total IgA subclasses were raised to the same extent, since the percentage of IgA2 of total IgA in serum did not differ between patients and healthy donors. In seropositive patients, RF and anti-CCP2 IgA1 and IgA2 could be detected, but measurements of anti-CCP2 IgA2 levels proved challenging due to interference of RF IgA. Although IgA2 has been postulated to be more proinflammatory, no correlations were found between total, RF and ACPA IgA subclass levels and DAS. An association between CRP and RF IgA2 was observed, but the effect size was small and did not remain significant after correction for multiple testing in the EAC. In smoking seropositive RA patients, a trend towards a selective increase in total IgA2 and RF IgA1 and IgA2 was observed (Figure 1D, data IMPROVED seropositive RA).

Figure 1.

Conclusion

Seropositive RA patients have raised IgA1 and IgA2 levels and can also harbor RF and ACPA IgA subclasses. No shift towards IgA2 was observed, indicating that the increase in total IgA is not due to translocation of mucosal IgA into the bloodstream. However, chronic mucosal inflammation might be one of the mechanisms involved in the raise in IgA(2) levels in RA, given the association between smoking and total IgA2 levels. Despite its’ pro-inflammatory properties, no strong associations between IgA2 and markers of inflammation were found, which suggests that IgA2 does not play a essential role in the ongoing pro-inflammatory processes in RA patients.

References

[1]Holers VM, Demoruelle MK, Kuhn KA, et al. Rheumatoid arthritis and the mucosal origins hypothesis: protection turns to destruction. Nature Reviews Rheumatology. 2018;14(9):542-57.

[2]Woof JM, Russell MW. Structure and function relationships in IgA. Mucosal Immunol. 2011;4(6):590-7.

[3]Steffen U, Koeleman CA, Sokolova MV, et al. IgA subclasses have different effector functions associated with distinct glycosylation profiles. Nat Commun. 2020;11(1):120.

Disclosure of Interests

None declared

AB0063 AMPA CANNOT BE DETECTED IN FAECES OF SEROPOSITIVE RA PATIENTS: NO EVIDENCE FOR LOCAL AMPA PRODUCTION IN THE LOWER GASTRO-INTESTINAL TRACT

Background

Rheumatoid arthritis (RA) patients harbor antibodies against several post-translational modifications (AMPA), for example anti-citrullinated protein antibodies (ACPA) and anti-acetylated protein antibodies (AAPA). The exact mechanism underlying the development of these autoantibodies is currently unclear, but the mucosal immune system has been hypothesized to play a role. ACPA IgA have been detected in sputum and saliva of seropositive RA patients (1-2), suggesting local production of autoantibodies in the lung and oral mucosa. This raises the question whether ACPA can also be produced at other mucosal sites. The intestines are an interesting candidate, as microbiome dysbiosis has been described in RA patients and intestinal content may represent a source of post-translationally modified antigens. However, it is currently unknown whether AMPA are also produced locally in the intestine.

Objectives

To determine whether AMPA are produced locally in the intestine, we set out to detect ACPA and AAPA in faeces samples of seropositive RA patients. These findings were compared to the ACPA/AAPA levels found in paired serum and saliva samples of the same patients.

Methods

Paired faeces, saliva and serum samples of 16 established seropositive RA patients and 16 healthy volunteers were collected. All patients fulfilled the ACR/EULAR 2010 criteria. Saliva was collected using the passive drooling method. Faeces was self-collected by participants at home and immediately frozen. Faeces samples were homogenized in PBS containing protease inhibitors and supernatants were used in ELISA. Saliva was diluted 1:4. Total IgA and anti-E. Coli IgA were measured using commercial and in-house ELISA, respectively. ACPA/AAPA IgA (serum, saliva) and ACPA/AAPA Ig (faeces) were detected using in-house ELISAs, coated with citrullinated and acetylated CCP2 or the unmodified arginine and lysine variants as control. Saliva samples were considered AMPA positive when the value was above the cut-off (mean + 2 times standard deviation of healthy donors) and the delta OD between the modified and unmodified peptide was larger than 0.1.

Results

9/16 ACPA IgG-positive RA patients were also ACPA IgA positive in serum. In the faecal supernatants total IgA was clearly detectable by ELISA: the supernatants contained on average 85 μg/ml total IgA, only slightly less compared to the mean 117 μg/ml total IgA found in the pre-diluted saliva. However, in none of the faecal supernatants ACPA Ig could be detected. Differences in optical density (OD) between the CCP2 and unmodified arginine peptide were close to 0 (Figure 1A). On the contrary, 8 RA patients were positive for ACPA IgA in saliva (Figure 1B). Similar results were found for AAPA (Figure 1C-D). To control whether the lack of an AMPA signal is explained by technical difficulties associated with measuring IgA in an antigen-specific manner in faeces, we determined anti-E. Coli IgA in the same samples. Anti-E. Coli IgA was clearly detectable in 11/32 faeces samples, indicating intact antigen-specific IgA was present in the faecal samples (Figure 1E).

Figure 1.

Conclusion

No ACPA or AAPA Ig antibodies could be detected in faeces of RA patients, even though these patients were positive for ACPA/AAPA IgA in both serum and saliva. The fact that it was possible to detect other antibodies in faeces, indicates that the absence of an AMPA signal is not due to inherent methodological issues. These findings indicate that the lower gastro-intestinal tract does not contain measurable levels of AMPA IgA and suggest that it is not a main site of ACPA production in RA patients.

References

[1]Willis et al. Sputum autoantibodies in patients with established rheumatoid arthritis and subjects at risk of future clinically apparent disease, Arthritis rheum 2013;65(10):2545-54.

[2]Svärd et al. Salivary IgA antibodies to cyclic citrullinated peptides (CCP) in rheumatoid arthritis. Immunobiol 2013;218(2):232-7.

Acknowledgements

We thank M. Yoganathan for his help with the experimental work during his internship

Disclosure of Interests

None declared

OP0070 INTERVENTION WITH METHOTREXATE IN ARTHRALGIA AT RISK FOR RHEUMATOID ARTHRITIS TO REDUCE THE DEVELOPMENT OF PERSISTENT ARTHRITIS AND ITS DISEASE BURDEN (TREAT EARLIER): A DOUBLE-BLIND, RANDOMISED, PLACEBO-CONTROLLED TRIAL

Background

Rheumatoid arthritis (RA) is the most common autoimmune disease, and requires long-term treatment to suppress inflammation. Currently, methotrexate is initiated as first-line treatment when arthritis becomes clinically apparent with joint swelling. However, disease processes begin long before and become clinically recognizable when patients develop symptoms. We hypothesized that the ‘at risk phase’ of symptoms and subclinical joint-inflammation is a therapeutic window to permanently modify the disease course.

Objectives

We studied if intervention in the pre-arthritis phase of arthralgia and subclinical joint inflammation prevents the development of clinical arthritis or reduces the burden of disease.

Methods

In this randomised, double-blind, 2-year proof-of-concept trial, adults with arthralgia clinically suspected of progressing to RA and MRI-detected subclinical joint-inflammation, recruited from all rheumatology outpatient-clinics in the southwest-Netherlands, were randomly assigned (1:1) to a single intramuscular glucocorticoid injection (120 mg) and a one-year course of oral methotrexate (up to 25 mg/week), or placebo injection and placebo tablets. Subsequently, participants were followed for another year without study medication. The primary endpoint was the development of clinically detectable arthritis (fulfilling the 2010 RA-criteria or involving ≥2 joints) that persisted for at least 2 weeks. Patient reported physical functioning, along with symptoms and workability, were key secondary endpoints and measured 4-monthly. Additionally, the course of MRI-detected inflammation was studied (the sum of tenosynovitis, synovitis, osteitis, scored with the RA-MRI Scoring (RAMRIS) method). All participants entered the intention-to-treat analysis. We performed two prespecified subgroup analyses. Firstly, analyses were restricted in participants with high risk of clinical arthritis development (PPV ≥70%). Secondly, analyses were stratified for ACPA-status. The trial is registered with the Netherlands Trials Registry (NTR4853 trial NL4599).

Results

From April 16th, 2015 to September 11th, 2019, we randomly assigned 236 participants to treatment (n=119) or placebo (n=117). After 24 months, arthritis free survival was similar in both groups (80% versus 82%, HR 0.81 (95%CI 0.45, 1.48)). Physical functioning improved more in the treatment-group during the first months and remained better (mean between-group difference over two-years HAQ -0·1(-0·2,-0·03;p=0·004). Similarly, pain (-9 on scale 0-100: (95%CI -12,-4; p<0·001), morning stiffness (-12 (95%CI -16,-8;p<0·001), presenteeism (-8% (95%CI -13%,-3%;p=0·001) showed sustained improvement compared to placebo. MRI-detected joint-inflammation was also persistently improved (mean difference over 2 years -1·4 points (95%CI -2·0,-0.9;p<0·001). High-risk participants in the treatment group showed a delay in clinical arthritis development: they developed the endpoint less often during treatment, but frequencies became similar at 24 months (67% in both groups). A similar delaying effect was observed in ACPA-positive participants, where 48% and 52% had developed persistent clinical arthritis at 24 months. The number of serious adverse events was equal between the groups; adverse events were as expected from methotrexate.

Conclusion

Methotrexate, the cornerstone treatment of RA, initiated at the pre-arthritis stage of joint symptoms and subclinical inflammation, did not prevent the development of clinical arthritis, but modified the disease course as measured by sustained improvement in MRI-detected inflammation, related symptoms and impairments. These findings of sustained disease modification may open up a new treatment landscape in a pre-arthritis phase of RA, where limitations can be just as severe as at the onset of clinical arthritis.

Figure 1.

Acknowledgements

We thank Prof. dr. R. ten Cate, prof. dr. S. le Cessie and dr. A.M.J. Langers for their role in the Data Safety and Monitoring Board. We thank all participants, and all rheumatologist of the following hospitals: Albert Schweitzer Hospital, Alrijne Hospital, Erasmus Medical Center, Haven-policlinic Rotterdam, IJselland Hospital, Ikazia Hospital, Franciscus Gasthuis & Vlietland Hospital, Groene Hart Hospital, Haaglanden Medical Center (all locations), Haga Hospital, Langeland Hospital, Meander Medical Center, Maasstad, Hospital, Reinier de Graaf Gasthuis, Reumazorg Zuid-West Nederland and Spaarne Gasthuis. We acknowledge the team of treating rheumatologists and research nurses of the LUMC, in particular Dr F.J. van der Giesen. Our gratitude also goes to the PhD students who scored MRIs for trial screening, in particular dr. H.W. van Steenbergen, dr. W. Nieuwenhuis, dr. R.M. ten Brink, dr. D.M. Boeters, dr. L. Mangnus, X.M.E. Matthijssen and F. Wouters. We thank dr. M. Reijnierse, prof. dr. S.C. Cannegieter and prof. dr. D. van der Heijde for their advice, and dr. J. Schoones for his help with the systematic literature search. We acknowledge the funder of the study: NWO ZonMW grant (program ‘translationeel onderzoek’, project number 95104004).

Disclosure of Interests

None declared.

Administration of an adeno-associated viral vector expressing interferon-β in patients with inflammatory hand arthritis, results of a phase I/II study

OBJECTIVE

Inflammatory hand arthritis (IHA) results in impaired function. Local gene therapy with ART-I02, a recombinant adeno-associated virus (AAV) serotype 5 vector expressing interferon (IFN)-β, under the transcriptional control of nuclear factor κ-B responsive promoter, was preclinically shown to have favorable effects. This study aimed to investigate the safety and tolerability of local gene therapy with ART-I02 in patients with IHA.

METHODS

In this first-in-human, dose-escalating, cohort study, 12 IHA patients were to receive a single intra-articular (IA) injection of ART-I02 ranging 0.3 × 10¹²-1.2 × 10¹³ genome copies in an affected hand joint. Adverse events (AEs), routine safety laboratory and the clinical course of disease were periodically evaluated. Baseline- and follow-up contrast enhanced magnetic resonance images (MRIs), shedding of viral vectors in bodily fluids, and AAV5 and IFN-β immune responses were evaluated. A data review committee provided safety recommendations.

RESULTS

Four patients were enrolled. Long-lasting local AEs were observed in 3 patients upon IA injection of ART-I02. The AEs were moderate in severity and could be treated conservative. Given the duration of the AEs and their possible or probable relation to ART-I02, no additional patients were enrolled. No systemic treatment emergent AEs were observed. The MRIs reflected the AEs by (peri)arthritis. No T-cell response against AAV5 or IFN-β, nor IFN-β antibodies could be detected. Neutralizing antibody titers against AAV5 raised post-dose.

CONCLUSION

Single IA doses of 0.6 × 10¹² or 1.2 × 10¹² ART-I02 vector genomes were administered without systemic side effects or serious AEs. However, local tolerability was insufficient for continuation.

TRIAL REGISTRATION

NCT02727764.

Evolution of anti-modified protein antibody responses can be driven by consecutive exposure to different post-translational modifications

Background

Besides anti-citrullinated protein antibodies (ACPA), rheumatoid arthritis patients (RA) often display autoantibody reactivities against other post-translationally modified (PTM) proteins, more specifically carbamylated and acetylated proteins. Immunizing mice with one particular PTM results in an anti-modified protein antibody (AMPA) response recognizing different PTM-antigens. Furthermore, human AMPA, isolated based on their reactivity to one PTM, cross-react with other PTMs. However, it is unclear whether the AMPA-reactivity profile is “fixed” in time or whether consecutive exposure to different PTMs can shape the evolving AMPA response towards a particular PTM.

Methods

Longitudinally collected serum samples of 8 human individuals at risk of RA and 5 with early RA were tested with ELISA, and titers were analyzed to investigate the evolution of the AMPA responses over time. Mice (13 per immunization group in total) were immunized with acetylated (or carbamylated) protein (ovalbumin) twice or cross-immunized with an acetylated and then a carbamylated protein (or vice versa) and their serum was analyzed for AMPA responses.

Results

Human data illustrated dynamic changes in AMPA-reactivity profiles in both individuals at risk of RA and in early RA patients. Mice immunized with either solely acetylated or carbamylated ovalbumin (AcOVA or CaOVA) developed reactivity against both acetylated and carbamylated antigens. Irrespective of the PTM-antigen used for the first immunization, a booster immunization with an antigen bearing the other PTM resulted in increased titers to the second/booster PTM. Furthermore, cross-immunization skewed the overall AMPA-response profile towards a relatively higher reactivity against the “booster” PTM.

Conclusions

The relationship between different reactivities within the AMPA response is dynamic. The initial exposure to a PTM-antigen induces cross-reactive responses that can be boosted by an antigen bearing this or other PTMs, indicating the formation of cross-reactive immunological memory. Upon subsequent exposure to an antigen bearing another type of PTM, the overall reactivity pattern can be skewed towards better recognition of the later encountered PTM. These data might explain temporal differences in the AMPA-response profile and point to the possibility that the PTM responsible for the initiation of the AMPA response may differ from the PTM predominantly recognized later in time.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13075-021-02687-5.

Handwork vs machine: a comparison of rheumatoid arthritis patient populations as identified from EHR free-text by diagnosis extraction through machine-learning or traditional criteria-based chart review

Background

Electronic health records (EHRs) offer a wealth of observational data. Machine-learning (ML) methods are efficient at data extraction, capable of processing the information-rich free-text physician notes in EHRs. The clinical diagnosis contained therein represents physician expert opinion and is more consistently recorded than classification criteria components.

Objectives

To investigate the overlap and differences between rheumatoid arthritis patients as identified either from EHR free-text through the extraction of the rheumatologist diagnosis using machine-learning (ML) or through manual chart-review applying the 1987 and 2010 RA classification criteria.

Methods

Since EHR initiation, 17,662 patients have visited the Leiden rheumatology outpatient clinic. For ML, we used a support vector machine (SVM) model to identify those who were diagnosed with RA by their rheumatologist. We trained and validated the model on a random selection of 2000 patients, balancing PPV and sensitivity to define a cutoff, and assessed performance on a separate 1000 patients. We then deployed the model on our entire patient selection (including the 3000). Of those, 1127 patients had both a 1987 and 2010 EULAR/ACR criteria status at 1 year after inclusion into the local prospective arthritis cohort. In these 1127 patients, we compared the patient characteristics of RA cases identified with ML and those fulfilling the classification criteria.

Results

The ML model performed very well in the independent test set (sensitivity=0.85, specificity=0.99, PPV=0.86, NPV=0.99). In our selection of patients with both EHR and classification information, 373 were recognized as RA by ML and 357 and 426 fulfilled the 1987 or 2010 criteria, respectively. Eighty percent of the ML-identified cases fulfilled at least one of the criteria sets.

Both demographic and clinical parameters did not differ between the ML extracted cases and those identified with EULAR/ACR classification criteria.

Conclusions

With ML methods, we enable fast patient extraction from the huge EHR resource. Our ML algorithm accurately identifies patients diagnosed with RA by their rheumatologist. This resulting group of RA patients had a strong overlap with patients identified using the 1987 or 2010 classification criteria and the baseline (disease) characteristics were comparable. ML-assisted case labeling enables high-throughput creation of inclusive patient selections for research purposes.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13075-021-02553-4.

AB0022 AUTOREACTIVE B CELLS IN RHEUMATOID ARTHRITIS DISPLAY AN ACTIVATED PHENOTYPE OF RECENT ANTIGEN EXPOSURE

Background:

Rheumatoid arthritis, in particular ACPA+ RA, is characterized by frequent disease flares and poor chances to achieve DMARD-free sustained remission. Recently, we have shown that ACPA-expressing memory B cells (MBC) remain in a persistently activated state throughout disease, even in patients in DMARD-induced clinical remission.(1) The reasons why the ACPA B cell response is continuously activated are unknown, as well as why the response does not revert to a resting, ‘quiescent’ state. We hypothesized that continuous antigen exposure in germinal centres drives ACPA B cell activation, leading to a ‘recent germinal centre emigrant’ phenotype of these cells in the circulation.

Objectives:

To understand whether the activated phenotype of ACPA-expressing B cells could be induced by recent antigen exposure, to thereby discern the processes of immune activation that remain active in patients even in clinical remission and to argue whether these processes could be targets for therapeutic intervention.

Methods:

ACPA-expressing B cells were identified in peripheral blood of RA patients by flow cytometry during different stages of disease and characterized by a panel of activation- and germinal centre related markers (CD80, CD86, CD32, CD95, Ki-67). In addition, three healthy donors received a TT booster vaccination. TT-specific MBC were identified in blood at different timepoints (before vaccination and up to 22 weeks after vaccination) and analysed phenotypically over time.

Results:

The majority of ACPA-expressing B cells strongly expressed CD95 and the co-stimulatory marker CD80. A part was also positive for the proliferation marker Ki-67 (on average 30%), and most cells downregulated the inhibitory marker CD32. TT-specific MBC adopted a comparable phenotype after booster vaccination, but most markers returned to the pre-vaccination expression level gradually over time. These effects were antigen-dependent because the phenotype of TT-negative B cells remained unchanged. The phenotypic composition of the proliferating ACPA-positive B cell pool most closely corresponded to a stimulation history of 1-2 weeks after antigen exposure. Notably, none of the Ki-67 negative ACPA-specific MBC showed phenotypic quiescence, indicating either a short life-time (in circulation) after antigen encounter or persistent additional factors of activation.

Figure 1.

Ki-67 expression on ACPA-specific MBC in RA (A) and on TT-specific MBC in 3 healthy donors before and after booster vaccination (B).

Conclusion:

ACPA-expressing MBC phenotypically resemble TT-specific MBC after recent (1-2 weeks) booster vaccination, reflecting the phenotype of recent germinal centre emigrants, and remain activated, whereas TT-specific MBC lose this marker profile over time. These observations suggest that ACPA-expressing MBC either home to tissue or survive shortly in the circulation, or that additional factors drive or program these cells to persistent activation. Transcriptomic profiling and analysis of the homing marker profile may help to answer these questions. Furthermore, it will be important to understand the association of persistent activation of ACPA-expressing B cells in clinical remission and the risk for disease flares upon treatment discontinuation.

References:

[1]Kristyanto H, Blomberg NJ, Slot LM, van der Voort EIH, Kerkman PF, Bakker A, et al. Persistently activated, proliferative memory autoreactive B cells promote inflammation in rheumatoid arthritis. Sci Transl Med. 2020;12(570).

Disclosure of Interests:

Nienke Blomberg: None declared, Hendy Kristyanto: None declared, Thomas Huizinga Grant/research support from: Gilead, Rene Toes: None declared, Hans Ulrich Scherer Grant/research support from: Pfizer, Lilly, Sanofi, BMS

POS0348 GENETIC SUSCEPTIBILITY VARIANTS FOR RHEUMATOID ARTHRITIS ARE NOT ASSOCIATED WITH EARLY REMISSION; A MULTI-COHORT STUDY

Background:

Patients who achieve remission promptly could have a specific genetic risk profile that supports regaining immune tolerance. The identification of these genes could provide novel drug targets.

Objectives:

To test the association between RA genetic risk variants with achieving remission at 6 months.

Methods:

We computed genetic risk scores (GRS) comprising of the RA susceptibility variants1 and HLA-SE status separately in 4425 patients across eight datasets from inception cohorts. Remission was defined as DAS28CRP<2.6 at 6 months. Missing DAS28CRP values in patients were imputed using predictive mean matching by MICE. We first tested whether baseline DAS28CRP changed with increasing GRS using linear regression. Next, we calculated odds ratios for GRS and HLA-SE on remission using logistic regression. Heterogeneity of the outcome between datasets was mitigated by running inverse variance meta-analysis.

Results:

Evaluation of the complete dataset, baseline clinical variables did not differ between patients achieving remission and those who did not (Table 1). Distribution of GRS was consistent between datasets. Neither GRS nor HLA-SE was associated with baseline DAS2DAS (OR1.01; 95% CI 0.99-1.04). A fixed effect meta-analysis (Figure 1.) showed no significant effect of the GRS (OR 0.99; 95% CI 0.94-1.03) or HLA-SE (OR 0.8CRP87; 95% CI 0.75-1.01) on remission at 6 months.

Table 1.

Summary of the data separated by disease activity after 6 months.

allRemission at 6 monthsNo remission at 6 monthsN4425*15582430Age, mean (sd)55.38 (13.87)5517 (14.09)55.62 (13.59)Female %68.98%65.43%70.73%ACPA+ %61.94%63.53%61.67%Baseline DAS28, mean (sd)4.76 (1.22)4.47 (1.23)5.1 (1.15)

*not all patients had 6 months data

Conclusion:

In these combined cohorts, RA genetics risk variants are not associated with early disease remission. At baseline there was no difference in genetic risk between patients achieving remission or not. Studies encompassing other genetic variants are needed to elucidate the genetics of RA remission.

References:

[1]Knevel R et al. Sci Transl Med. 2020;12(545):eaay1548.

Acknowledgements:

This project has received funding from the Innovative Medicines Initiative 2 Joint Undertaking under grant agreement No 777357, RTCure.

This project has received funding from Pfizer Inc.

Disclosure of Interests:

Samantha Jurado Zapata: None declared, Marc Maurits: None declared, Yann Abraham Employee of: Pfizer, Erik van den Akker: None declared, Anne Barton: None declared, Philip Brown: None declared, Andrew Cope: None declared, Isidoro González-Álvaro: None declared, Carl Goodyear: None declared, Annette van der Helm - van Mil: None declared, Xinli Hu Employee of: Pfizer, Thomas Huizinga: None declared, Martina Johannesson: None declared, Lars Klareskog: None declared, Dennis Lendrem: None declared, Iain McInnes: None declared, Fraser Morton: None declared, Caron Paterson: None declared, Duncan Porter: None declared, Arthur Pratt: None declared, Luis Rodriguez Rodriguez: None declared, Daniela Sieghart: None declared, Paul Studenic: None declared, Suzanne Verstappen: None declared, Leonid Padyukov: None declared, Aaron Winkler Employee of: Pfizer, John D Isaacs: None declared, Rachel Knevel Grant/research support from: Pfizer

OP0147 RHEUMATIC? - A DIGITAL DIAGNOSTIC DECISION SUPPORT TOOL FOR INDIVIDUALS SUSPECTING RHEUMATIC DISEASES: A MULTICENTER VALIDATION STUDY

Background:

Digital diagnostic decision support tools promise to accelerate diagnosis and increase health care efficiency in rheumatology. Rheumatic? is an online tool developed by specialists in rheumatology and general medicine together with patients and patient organizations for individuals suspecting a rheumatic disease.1,2 The tool can be used by people suspicious for rheumatic diseases resulting in individual advise on eventually seeking further health care.

Objectives:

We tested Rheumatic? for its ability to differentiate symptoms from immune-mediated diseases from other rheumatic and musculoskeletal complaints and disorders in patients visiting rheumatology clinics.

Methods:

The performance of Rheumatic? was tested using data from 175 patients from three university rheumatology centers covering two different settings:

A.Risk-RA phase setting. Here, we tested whether Rheumatic? could predict the development of arthritis in 50 at risk-individuals with musculoskeletal complaints and anti-citrullinated protein antibody positivity from the KI (Karolinska Institutet)

B.Early arthritis setting. Here, we tested whether Rheumatic? could predict the development of an immune-mediated rheumatic disease in i) EUMC (Erlangen) n=52 patients and ii) LUMC (Leiden) n=73 patients.

In each setting, we examined the discriminative power of the total score with the Wilcoxon rank test and the area-under-the-receiver-operating-characteristic curve (AUC-ROC).

Results:

In setting A, the total test score clearly differentiated between individuals developing arthritis or not, median 245 versus 163, P < 0.0001, AUC-ROC = 75.3 (Figure 1). Also within patients with arthritis the Rheumatic? total score was significantly higher in patients developing an immune-mediated arthritic disease versus those who did not: median score EUMC 191 versus 107, P < 0.0001, AUC-ROC = 79.0, and LUMC 262 versus 212, P < 0.0001, AUC-ROC = 53.6.

Figure 1.

(Area under) the receiver operating curve for the total Rheumatic? score

Conclusion:

Rheumatic? is a web-based patient-centered multilingual diagnostic tool capable of differentiating immune-mediated rheumatic conditions from other musculoskeletal problems. A following subject of research is how the tool performs in a population-wide setting.

References:

[1]Knitza J. et al. Mobile Health in Rheumatology: A Patient Survey Study Exploring Usage, Preferences, Barriers and eHealth Literacy. JMIR mHealth and uHealth. 2020.

[2]https://rheumatic.elsa.science/en/

Acknowledgements:

This project has received funding from EIT Health. EIT Health is supported by the European Institute of Innovation and Technology (EIT), a body of the European Union that receives support from the European Union’s Horizon 2020 Research and Innovation program.

This project has received funding from the Innovative Medicines Initiative 2 Joint Undertaking under grant agreement No 777357, RTCure.

Disclosure of Interests:

Rachel Knevel: None declared, Johannes Knitza: None declared, Aase Hensvold: None declared, Alexandra Circiumaru: None declared, Tor Bruce Employee of: Ocean Observations, Sebastian Evans Employee of: Elsa Science, Tjardo Maarseveen: None declared, Marc Maurits: None declared, Liesbeth Beaart- van de Voorde: None declared, David Simon: None declared, Arnd Kleyer: None declared, Martina Johannesson: None declared, Georg Schett: None declared, Thomas Huizinga: None declared, Sofia Svanteson Employee of: Elsa Science, Alexandra Lindfors Employee of: Ocean Observations, Lars Klareskog: None declared, Anca Catrina: None declared

POS0709 LUPUS FOG IS NOT DISSOCIATIVE FOG

Background:

The presence of a ‘fog’ is frequently reported by patients with systemic lupus erythematosus (SLE). However, little is known about this lupus fog: it is thought to be a result of cognitive dysfunction, but fogs can also be the result of dissociation. The Dissociative Experience Scale-II (DES) is a standardized tool to study dissociation. In the general adult population, scores range from 4.4-14.1-3

Objectives:

We aimed to study the prevalence of dissociative symptoms including dissociative fog in patients with SLE and neuropsychiatric symptoms.

Methods:

Patients visiting the tertiary referral center for neuropsychiatric systemic lupus erythematosus (NPSLE) of the LUMC between 2007-2019 were included. All patients underwent a standardized multidisciplinary assessment. Patients were classified as NPSLE if neuropsychiatric symptoms were attributed to SLE and immunosuppressive or anticoagulant therapy was initiated, otherwise patients were classified as minor/non-NPSLE. Dissociation was studied using the DES. The DES separates different types of dissociative symptoms: amnesia, absorption/imagination and derealization/depersonalization. It also contains one question regarding evaluating the presence of a dissociative fog: “Some people sometimes feel as if they are looking at the world through a fog, so that people and objects appear far away or unclear”. All statements (n = 28) regarding dissociative symptoms are rated from ‘none of the time’ to ‘all of the time’ (0-100%); scores >25 are considered abnormal. A multiple regression analysis (MRA) were performed to compare dissociation in patients with and without NPSLE. DES results are presented as median (range) and MRA as odds ratio (OR) and 95% confidence interval (CI).

Results:

DES questionnaires were available for 337 patients, of which 97 had the diagnosis NPSLE (29%). Mean age in patients with NPSLE was 41 ± 13 years and 84% was female. In minor/non-NPSLE, median age was 44 ± 14 years and 87% was female.

Median dissociation was 7 (0-75) and did not differ between patients with minor/non-NPSLE and NPSLE (OR: 1.0 (95% CI: -0.9; 1.1)). The most common type of dissociation was absorption/imagination (median: 12, range 0-75) and depersonalization/derealization was infrequent (median: 1, range 0-84). 43 patients (13%) had an abnormal score (>25) on the dissociative fog question.

Conclusion:

Dissociative symptoms are within normal range in patients with SLE and neuropsychiatric symptoms, regardless of underlying etiology. Dissociative fog seems uncommon and therefore lupus fog is most likely not the result of dissociation.

References:

[1]Bernstein EM and Putnam FW. Development, reliability, and validity of a dissociation scale. J Nerv Ment Dis 1986; 174: 727-735. 1986/12/01. DOI: 10.1097/00005053-198612000-00004.

[2]Maaranen P, Tanskanen A, Honkalampi K, et al. Factors associated with pathological dissociation in the general population. Aust N Z J Psychiatry 2005; 39: 387-394. 2005/04/30. DOI: 10.1080/j.1440-1614.2005.01586.x.

[3]van IJzendoorn MH and Schuengel C. The measurement of dissociation in normal and clinical populations: Meta-analytic validation of the Dissociative Experiences Scale (DES). Clinical Psychology Review 1996; 16: 365-382. DOI: 10.1016/0272-7358(96)00006-2.

Table 1.

Presence of dissociation in patients with SLE and neuropsychiatric symptoms

Total cohort(n = 337)Minor/non-NPSLE(n = 240)NPSLE(n = 97)DES (median, range)Total questionnaire7 (0 - 75)8 (0 - 66)6 (0 – 75)Amnesia5 (0 - 76)5 (0 - 68)4 (0 - 76)Absorption/imagination12 (0 – 75)13 (0 – 75)10 (0 – 73)Depersonalization/derealization1 (0 – 84)1 (0 – 73)1 (0 – 84)Dissociative fog* 0 (0-100)0 (0-100)0 (0-100)

DES = Dissociative Experience Scale NPSLE = neuropsychiatric systemic lupus erythematosus.

*Dissociative fog is question 28 of the DES-II

Disclosure of Interests:

None declared

POS0708 PSYCHIATRIC DISORDERS IN PATIENTS WITH DIFFERENT PHENOTYPES OF NEUROPSYCHIATRIC SYSTEMIC LUPUS ERYTHEMATOSUS (NPSLE)

Background:

Patients with systemic lupus erythematosus (SLE) may present with psychiatric disorders. These are important to recognize, as they influence quality of life and treatment outcomes and strategies.

Objectives:

We aimed to study the frequency of psychiatric morbidity as classified by the Diagnostic and Statistical Manual of Mental Disorders (DSM-V) in patients with SLE and neuropsychiatric symptoms of different origins.

Methods:

In the neuropsychiatric SLE (NPSLE) clinic of the Leiden University Medical Center, patients undergo a standardized multidisciplinary assessment by a neurologist, neuropsychologist, vascular internal medicine, rheumatologist, physician assistant and psychiatrist. After two weeks, a multidisciplinary consensus meeting takes place, in which the symptoms are attributed to SLE requiring treatment (major NPSLE) or to minor involvement of SLE or other causes (minor/non-NPSLE). Consecutive patients visiting the NPSLE clinic between 2007-2019 were included. Data of psychiatric evaluation and current medication use were extracted from medical records. The presence of cognitive dysfunction was established during formal neuropsychological assessment.

Results:

371 consecutive SLE patients were included, of which 110 patients had major NPSLE (30%). Mean age was 44 ± 14 years and 87% was female.

The most frequently diagnosed psychiatric disorders in the total group were cognitive dysfunction (42%) and depression (23%), as shown in Table 1. Furthermore, anxiety was present in 5% and psychotic disorders in 4% of patients. In patients with minor/non-NPSLE, especially depression (26% vs 15%) and anxiety (6% vs 2%) were more common than in major NPSLE. Cognitive dysfunction (54% vs 36%) and psychotic disorders (6% vs 4%) were more common in patients with major NPSLE than minor/non-NPSLE.

Psychiatric medication was used in 33% of patients, of which antidepressants and benzodiazepines the most frequently (both: 18% in both subgroups). Antipsychotics were more often used in patients with NPSLE (10% vs 7%) and benzodiazepines more often in minor/non-NPSLE (20% vs 14%).

In addition, 17 patients (5%) had a history of suicide attempt, which was more common in patients with minor/non-NPSLE than major NPSLE (6% vs 2%).

Conclusion:

Psychiatric morbidity, especially cognitive dysfunction and depression, are common in patients with lupus and differ between underlying cause of the neuropsychiatric symptoms (minor/non-NPSLE vs major NPSLE).

Table 1.

Presence of psychiatric diagnoses in patients with SLE and neuropsychiatric symptoms

All patients(n = 371)Minor/non-NPSLE(n = 261)Major NPSLE(n = 110)DSM V diagnosis, n (%)Neurodevelopmental disorder5 (1)2 (1)3 (2)Schizophrenia Spectrum and Other Psychotic Disorders16 (4)10 (4)6 (6)Bipolar and related disorders7 (2)5 (2)2 (2)Depressive disorders84 (23)68 (26)16 (15)Anxiety disorders17 (5)15 (6)2 (2)Obsessive-Compulsive and Related Disorders1 (0)1 (0)0 (0)Trauma- and Stressor-Related Disorders16 (4)12 (5)4 (3)Dissociative Disorders2 (1)2 (1)0 (0)Somatic Symptom and Related Disorders1 (0)1 (0)0 (0)Feeding and Eating Disorders0 (0)1 (0)0 (0)Elimination Disorders0 (0)0 (0)0 (0)Sleep-wake disorders2 (1)2 (1)0 (0)Sexual dysfunctions0 (0)0 (0)0 (0)Gender dysphoria0 (0)0 (0)0 (0)Disruptive, Impulse-Control, and Conduct Disorder0 (0)0 (0)0 (0)Substance-related and addictive disorders9 (2)8 (3)1 (1)Cognitive dysfunction154 (42)95 (36)59 (54)Personality disorders10 (3)9 (3)1 (1)Paraphilic disorders0 (0)0 (0)0 (0)Other mental disorders12 (3)7 (3)5 (5)Medication-Induced Movement Disorders and Other Adverse Effects of Medication0 (0)0 (0)0 (0)Unknown3 (1)3 (1)0 (0)

NPSLE = neuropsychiatric systemic lupus erythematosus.

Disclosure of Interests:

None declared

POS0479 THE IMPACT OF FLARES ON PATIENT REPORTED OUTCOMES IN RHEUMATOID AND UNDIFFERENTIATED ARTHRITIS PATIENTS – A SUB-ANALYSIS OF THE IMPROVED STUDY

Background:

In rheumatoid arthritis (RA) patients in low disease activity, flares are associated with symptom deterioration. Patients in clinical remission may flare but still have low disease activity. How does this affect patient reported outcomes?

Objectives:

To evaluate the prevalence of disease flares in patients treated to target drug free remission, and to study the impact of disease flares on patient-reported outcomes (PROs) for flares with different impact on disease activity.

Methods:

In the IMPROVED study 610 patients with rheumatoid arthritis (RA) or undifferentiated arthritis (UA) were treated to target drug free remission (DAS <1.6) for 5 years. As soon as DAS was <1.6, treatment was tapered to discontinuation. Patients with at least 8 months follow-up were selected. A flare was defined according to three definitions; A) DAS ≥1.6 and ≥0.6 increase from the previous visit regardless of the previous DAS; B) minor flare from remission; a DAS ≥1.6 with <0.6 increase and previous DAS <1.6; C) major flare from remission; a DAS ≥1.6 with ≥0.6 increase and previous DAS <1.6. Linear mixed models were used to compare functional ability, measured by the health assessment questionnaire (HAQ), at visits where a flare occurred versus visits without a flare. Fisher’s exact test were used to compare percentages with ≥0.22 HAQ increases between groups with and without LDA at the moment of flare. A generalized linear mixed model was used to calculate the odds ratio for a deterioration of ≥20 mm in VAS of PROs global health (GH), disease activity, pain and morning stiffness (from the preceding visit) during a flare.

Results:

Of the 585 patients with sufficient follow-up, 75% experienced a flare A, 26% a flare B, and 68% a flare C, at least once. Most flares were observed after t=8 and t=12 months. In 55%, 100%, and 69% of visits with a flare A, B or C, the patients were still in LDA (DAS=<2.4). In 55% of the visits where a flare was associated with a DAS increase ≥0.6 (flare A & C) there was also clinically relevant increase in HAQ of ≥0.22. The mean difference in HAQ was 0.27 with flare A (p<0.01), 0.03 with flare B (p=0.72) and 0.18 with flare C (p<0.01). If was DAS >2.4 (LDA) at the moment of flare, HAQ increased ≥0.22 in 68% of all flares A, and 77% of all flares C (p-values <0.01, compared to flares where patients were still in LDA, DAS=<2.4). The odds ratios of a >20 mm deterioration in VAS global health, VAS disease activity, VAS pain and VAS morning stiffness was significant ≥1 for flares with a ≥0.6 increase in DAS (flares A and C), and ≤1 for minor flares (B) (table 1).

Conclusion:

In early arthritis patients, during 5 years treated to target drug free DAS-remission, disease flares with loss of DAS-remission were common. Although the majority of patients who flared were still in LDA, most reported more pain, morning stiffness, increased disease activity and a diminished global health. On average, deterioration in HAQ only exceeded the minimum clinically important difference (delta HAQ >=0.22) in case of a ≥0.6 increase in DAS, independent of the previous DAS. Depending on the definition of flare, up to 45% of patients lost DAS LDA, and in this group the functional deterioration significantly more often exceeded the MCID as compared to the patients that flared but were still in LDA. More research is needed to find out which patients are most at risk for clinically relevant flares, and to evaluate the impact of flares in patients with remission on long term outcomes.

Table 1.

Odds Ratios and 95% confidence intervals for > 20 mm increase in PROs on 100mm visual analogue scales

Flare AFlare B (minor)Flare C (major)Prevalence ≥20 mmaOR(95% CI)Prevalence ≥20 mmaOR(95% CI)Prevalence ≥20 mmaOR(95% CI)Global health62%2.1 (1.5; 2.8)45%0.5 (0.4; 0.7)62%1.4 (1.1; 1.8)Disease activity62%2.5 (1.7; 3.8)45%0.4 (0.3; 0.6)62%2.1 (1.4; 3.0)Pain87%2.0 (1.3; 3.1)78%0.5 (0.3; 0.8)87%1.8 (1.2; 2.5)Morning stiffness84%1.7 (1.1; 2.6)77%0.6 (0.4; 0.9)86%2.1 (1.5; 2.9)

a The prevalence of >20 mm deterioration in VAS PRO’s during a visit with a flare.

Acknowledgements:

We would like to thank all patients for their contribution as well as the rheumatologists who participated in the IMPROVED-study group. We would also like to thank all other rheumatologists and trainee rheumatologists who enrolled patients in these studies, and all research nurses for their contributions.

Disclosure of Interests:

Johanna M. Maassen: None declared, Sytske Anne Bergstra: None declared, Petronella DM de Buck: None declared, M. van Oosterhout: None declared, Thomas Huizinga: None declared, Cornelia Allaart Grant/research support from: the IMPROVED study was designed by the investigators and financially supported by AbbVie in the first year.

POS0386 EVOLUTION OF ANTI-MODIFIED PROTEIN ANTIBODY RESPONSES CAN BE DRIVEN BY CONSECUTIVE EXPOSURE TO DIFFERENT POST-TRANSLATIONAL MODIFICATIONS

Background:

Besides anti-citrullinated protein antibodies (ACPA), rheumatoid arthritis patients (RA) often display autoantibody reactivities against other post-translationally modified (PTM) proteins, more specifically carbamylated and acetylated proteins. Immunizing mice with one PTM results in an anti-modified protein antibody (AMPA) response recognizing multiple PTMs. Furthermore, human AMPA, isolated based on their reactivity to one PTM, cross-react with other PTMs at the monoclonal and polyclonal level. However, it is unclear whether the AMPA reactivity profile is “fixed” in time, or whether consecutive exposure to different PTMs can shape the evolving AMPA-response.

Objectives:

To investigate the evolution of the AMPA response in mice with controlled exposure to PTMs as well as in AMPA-positive humans.

Methods:

Mice were immunized with acetylated (or carbamylated) protein (ovalbumin) twice or cross-immunized with an acetylated and then a carbamylated protein (or vice versa) and their serum was analyzed for AMPA responses with ELISA using a different backbone protein (fibrinogen) bearing the same modifications. Longitudinally collected serum samples of human individuals at risk of RA and with early RA were tested to investigate the evolution of the AMPA responses in humans.

Results:

Mice immunized twice with either solely acetylated or solely carbamylated ovalbumin (AcOVA or CaOVA) developed reactivity against both acetylated and carbamylated antigens. Irrespective of the PTM used for the first immunization, a booster immunization with the other PTM resulted in increased titers to the second/booster PTM (Figure 1A), suggesting that immunization with a defined PTM-antigen leads to the generation of anti-PTM memory B cells able to cross-recognize other PTMs. Furthermore, immunizing with CaOVA and boosting with AcOVA (or vice versa) skewed the overall AMPA-response profile towards a relatively higher reactivity against the “booster” PTM (Figure 1B). Human data also illustrated dynamic changes in AMPA reactivity profiles in both individuals at risk of RA and in early RA patients (not shown).

Conclusion:

The relationship between different reactivities within the AMPA response is dynamic. The initial exposure to a PTM antigen induces cross-reactive response that can be boosted by the same or other PTMs. The overall reactivity pattern can be skewed by subsequent exposure to other PTMs. These data might explain temporal changes in the reactivity profile of the AMPA response and point to the possibility that the PTM responsible for the initiation of the AMPA response may differ from the PTM predominantly recognized later in time.

Disclosure of Interests:

None declared

POS0470 GLUCOCORTICOID DISCONTINUATION IN EARLY RHEUMATOID AND UNDIFFERENTIATED ARTHRITIS PATIENTS. A SUB-ANALYSIS OF THE BeSt AND IMPROVED STUDIES

Background:

Discontinuation of glucocorticoids (GC) as bridging therapy in rheumatoid arthritis (RA) treatment is recommended as rapidly as clinically feasible. Little is known about the rate of, and possible characteristics associated with successful GC discontinuation.

Objectives:

To evaluate the success rate of GC discontinuation, and to study which factors are associated with successful GC discontinuation.

Methods:

Data from two treat-to-target studies; BeSt (target DAS ≤2.4, LDA), and IMPROVED (target DAS <1.6, remission) were evaluated for all patients initially treated with a tapered high-dose of oral GC with conventional DMARD(s). GCs were discontinued when DAS ≤2.4 was maintained for 28 weeks in BeSt, and as soon as DAS was <1.6 in IMPROVED. GC discontinuation could be attempted twice: first attempt called ‘primary’, second attempt after GC restart called ‘secondary’. Discontinuation was considered successful if the treatment target was maintained at the next visit. Univariable, and exploratory forward and backward multivariable logistic regression analyses were performed to identify potential characteristics associated with successful discontinuation (p-value <0.2 included in final model).

Results:

From the 131 patients initiating combination therapy with prednisone in the BeSt study, 93 attempted discontinuation. Primary discontinuation was successful in 60% (56/93) and secondary discontinuation in 54% (19/35). A lower DAS at the visit prior to GC discontinuation and ACPA negativity were associated with successful discontinuation (table 1). Of the 610 patients in the IMPROVED, 400 attempted discontinuation. Primary discontinuation was successful in 61% (242/400), and secondary in 51% (71/139). A lower DAS both at baseline and at the visit prior to GC discontinuation were associated with successful discontinuation (table 1).

Conclusion:

Primary GCs discontinuation was successful in approximately 60% and secondary in 50% of patients, independent of the treatment target and associated threshold for GC discontinuation. Most baseline characteristics were not predictive of successful GC discontinuation, but ACPA negativity (only in BeSt), baseline DAS (only in IMPROVED) and in both studies DAS prior to GC discontinuation were predictive for successful discontinuation. Based on this data it seems that ‘standard’ baseline characteristics are insufficient to ‘personalize’ the duration of temporary GC bridging but the DAS at the moment of GC discontinuation might give guidance.

Table 1.

Results logistic regression analyses

UnivariableMultivariableaR2= 0.173BeStOR (95% CI)p-valueOR (95% CI)p-valueAge, year1.00 (0.98; 1.03)0.98Gender, female0.51 (0.24; 1.09)0.08Symptom duration BL, weeks1.00 (0.99; 1.01)0.61DAS at BL0.92 (0.61; 1.40)0.70DAS prior to discontinuation0.13 (0.05; 0.33)<0.010.11 (0.04; 0.30)<0.01RF, positive1.28 (0.62; 2.69)0.502.24 (0.81; 6,17)0.12ACPA, positive0.70 (0.34; 1.43)0.320.32 (0.12; 0.86)0.02Erosions, present at BL0.65 (0.28; 1.49)0.31UnivariableMultivariableaR2= 0.065IMPROVEDOR (95% CI)p-valueOR (95% CI)p-valueAge, year1.00 (0.99; 1.02)0.64Gender, female0.62 (0.43; 0.89)0.010.75 (0.51; 1.11)0.15Symptom duration BL, weeks1.00 (0.99; 1.00)0.430.99 (0.99; 1.00)0.08DAS at BL0.80 (0.65; 0.98)0.030.78 (0.62; 0.98)0.03DAS prior to discontinuation0.24 (0.15; 0.38)<0.010.24 (0.14; 0.40)<0.01RF, positive0.82 (0.57; 1.17)0.27ACPA, positive0.95 (0.66; 1.35)0.76Erosions, present at BL0.80 (0.49; 1.29)0.35

ACPA: anti-citrullinated protein antibodies; BL: baseline; DAS: disease activity score; RF: rheumatoid factor. α The final multivariable logistic regression model was based on stepwise forward and backward selection of predictors, both resulting in the same final model.

Acknowledgements:

We would like to thank all patients for their contribution as well as the rheumatologists who participated in the BeSt study group and in the IMPROVED-study group. We would also like to thank all other rheumatologists and trainee rheumatologists who enrolled patients in these studies, and all research nurses for their contributions.

Disclosure of Interests:

Johanna M. Maassen: None declared, Raquel Dos-Santos: None declared, Sytske Anne Bergstra: None declared, Robbert Goekoop: None declared, Thomas Huizinga: None declared, Cornelia Allaart Grant/research support from: The original BeSt study was realized with a government grant from the Dutch College of Health Insurance Companies, with additional funding from Schering-Plough and Janssen. the IMPROVED study was financially supported by AbbVie in the first year.

AB0667 A PROSPECTIVE STUDY INTO COVID-19 LIKE SYMPTOMS IN PATIENTS WITH AND WITHOUT IMMUNE MEDIATED INFLAMMATORY DISEASES OR IMMUNOMODULATING DRUGS

Background:

Patients with an immune mediated inflammatory disorder or post solid organ transplantation (IMIDT), are at risk for infectious complications especially if they are treated with immunosuppressive drugs (imeds). There is still great uncertainty whether these IMIDT patients are more susceptible to COVID-19 than controls, and/or should be advised to avoid taking their immunosuppressive treatment.

Objectives:

To evaluate whether patients with IMIDT are more at risk for CLS than controls.

Methods:

The IENIMINI study is a prospective cohort study in patients with IMIDT and controls (healthy or no IMIDT) who were identified based on the registration database of the Leiden University Medical Center. Over time, participants registered COVID-like symptoms (CLS) as they occurred, and filled in additional questionnaires. Univariate and multivariate regression analyses were done to identify variables associated with having CLS.

Results:

Of the 8670 individuals approached, 2110 with IMIDT and 1067 controls agreed to participate. In March and April, 454 (22%) of IMIDT patients and 242 (23%) of controls recorded to have CLS, mostly mild with a median (IQR) duration of seven (3-14) days in the IMIDT group and six days (4-11) in the control group. Eleven (5%) of the IMIDT patients with immunosuppressive medication (imed), 6 (3%) of IMIDTs without imed and 2 (1%) of controls were hospitalized with CLS (p=0.04). In May and June, fewer episodes overall were recorded. Being female (OR 1.45 95%CI 1.15;1.82), having a lung disease (OR 1.50 95%CI 1.20;1.88) and wearing a face mask (OR 1.42 95%CI 1.13-1.77) were independently associated with a higher risk, while higher age (OR 0.96 95%CI 0.96;0.97) and having an IMIDT with immunosuppressive medication use (OR 0.68 95%CI 0.51;0.91) were independently associated with a lower risk (see Table 1). Similar results were found after data imputation.

Table 1.

Univariate & multivariate analysis of variables associated with having CLS or not (OR with 95% CI)

n0UnivariateMultivariate*Sex, female25461.89 (1.58;2.25)1.45 (1.15;1.82)BMI23910.99 (0.97;1.01)1.00 (0.98;1.03)Age25460.97 (0.96;0.97)0.96 (0.96;0.97)IMIDT without imed†25461.00 (0.82;1.23)0.94 (0.72;1.24)IMIDT with imed †25460.79 (0.65;0.97)0.68 (0.51;0.91)Smoking (current)24631.35 (1.02;1.78)1.05 (0.74;1.50)Physical contact with family**22201.47 (1.22;1.78)1.22 (0.98;1.53)Visiting other people (not family)22051.26 (1.05;1.51)0.96 (0.77;1.20)Wearing a face mask21961.46 (1.20;1.76)1.42 (1.13;1.77)Close contact (at work)21801.65 (1.34;2.03)1.27 (0.97;1.66)Self-reported Diabetes Mellitus23810.69 (0.50;0.96)0.89 (0.58;1.36)Self-reported lung disease23961.30 (1.09;1.54)1.50 (1.20;1.88)Self-reported heart disease23990.85 (0.69;1.04)1.09 (0.83;1.43)Daily alcohol use24160.84 (0.71;1.00)1.20 (0.96;1.50)Influenza vaccination***24150.71 (0.60;0.84)0.96 (0.76;1.21)Solid organ transplantation25460.74 (0.54;1.03)0.79 (0.47;1.35)Good adherence to lockdown rules22451.17 (0.41;3.29)2.46 (0.65;9.38)Use of oral corticosteroids25460.84 (0.66;1.06)1.44 (0.95;2.20)Working outside the house24351.39 (1.16;1.68)0.92 (0.71;1.20)

Abbreviations: BMI=body mass index; CI= confidence intervals; CLS=Covid like symptoms; IMIDT= with immune mediated inflammatory disorders or transplant organ; n0=number of observations; OR=odds ratio.

* number of observations: 1835

** physical contact specified as ‘holding/shaking hands, hugging etcetera’

*** in autumn 2019

† control group = reference group

Conclusion:

Between March and July 2020, IMIDT patients, whether or not taking imeds, did not show an increased risk of reported COVID-like symptoms compared to controls. Continuing immunosuppressant drugs as long as not ill, while following the Dutch COVID rules, appears to be safe.

Disclosure of Interests:

Lotte van Ouwerkerk: None declared., Andrea van der Meulen Speakers bureau: Dr. van der Meulen reports personal fees from Janssen, grants and personal fees from Takeda, personal fees from Galapogos, grants from Nestle, grants from Norgine, outside the submitted work., Grant/research support from: Dr. van der Meulen reports personal fees from Janssen, grants and personal fees from Takeda, personal fees from Galapogos, grants from Nestle, grants from Norgine, outside the submitted work., Maarten Ninaber: None declared., Y.K. Onno Teng: None declared., Thomas Huizinga: None declared., Cornelia Allaart: None declared.

POS0395 ANTI-ACETYLATED PROTEIN ANTIBODIES IN RHEUMATOID ARTHRITIS (RA): CLUES FOR THE STARTING POINT OF AUTOANTIBODY RESPONSES IN RA

Background:

Rheumatoid arthritis (RA) is characterized by autoantibodies such as rheumatoid factor (RF) and anti-modified protein autoantibodies (AMPAs) like anti-citrullinated protein antibodies (ACPA) and anti-carbamylated protein antibodies (anti-CarP). Recently, another AMPA: anti-acetylated protein antibodies (AAPA) have been found in RA patients [1]. The prevalence of AAPA antibodies and their isotypes have yet to be determined. Since isotype profiles reflect the breadth of an immune response, the prevalence of AAPA isotypes in arthritis patients with and without RA can help to understand the relevance of this autoantibody response in RA.

Objectives:

To describe the prevalence of AAPA isotypes in arthritis patients with and without RA.

Methods:

In 650 RA patients fulfilling the 1987 RA criteria and 555 non-RA arthritis patients from the Leiden Early Arthritis Cohort, baseline serum samples were screened by ELISA for IgG, IgM and IgA to an acetylated- and control peptide that was based upon the CCP-2 backbone. The cutoff for positivity was based on 80 controls (mean + 2SD). A sample was considered positive if it was above the cutoff and was 0.1 optical density higher on the acetylated peptide than on the control peptide.

Results:

AAPA IgG was found in 36% of RA patients versus 6.7% of non-RA arthritis patients (figure 1a). Within RA patients, AAPA IgG antibodies were mostly present in the ACPA-(CCP-2) positive group (64% in ACPA-positive, compared to 5% in ACPA-negative). Levels of AAPA IgG and IgA were higher in RA patients than in healthy controls and non-RA arthritis patients (figure 1b), however, surprisingly, no difference in levels was found for IgM.

When isotype profiles in AAPA- positive arthritis patients were compared, patients with RA were more often positive for two or more isotypes then patients without RA, and thus displayed considerably more overlap in AAPA isotypes compared to non-RA patients (table 1). Intriguingly, IgM AAPA was the most prevalent isotype in non-RA patients, versus IgG in RA patients.

Table 1.

Anti-acetylated protein antibody (AAPA) isotype overlap in AAPA positive patients.

AAPA isotypeRA patients (=310) n (%)Non-RA arthritis patients (n=106) n (%)IgG+IgM-IgA-115 (37.1)28 (5.1)IgG-IgM+IgA-52 (16.8)48 (8.7)IgG-IgM-IgA+14 (4.5)13 (2.3)IgG+IgM+IgA-24 (7.7)3 (0.5)IgG+IgM-IgA+37 (11.9)4 (0.7)IgG-IgM+IgA+9 (2.9)8 (1.4)IgG+IgM+IgA+59 (19.0)2 (0.4)

AAPA: anti-acetylated protein antibodies, RA: rheumatoid arthritis

Conclusion:

AAPA are detected in one third of RA patients, and mainly in the ACPA-positive subgroup. The predominance of IgM AAPA in non-RA arthritis patients and healthy controls suggests that healthy persons can develop AAPA IgM without the development of RA. These results also suggest that in healthy individuals, AAPA responses can occur, but do not mature past the IgM-stage, while in RA patients, the AAPA-response does mature and might form a “starting point” for development of other AMPA leading to the concurrent present of several AMPA in disease.

References:

[1]Juarez, M., et al., Identification of novel antiacetylated vimentin antibodies in patients with early inflammatory arthritis. Ann Rheum Dis, 2016. 75(6): p. 1099-107.

Disclosure of Interests:

None declared

POS0356 THE RELATIONSHIP OF GENETICS AND CLINICALLY SUSPECT ARTHRALGIA IN RA DEVELOPMENT ASSESSED USING HC, CSA AND RA PATIENTS

Background:

The identification of a pre-RA stage of patients with clinically suspect arthralgia (CSA) has proven to be beneficial in the early detection of Rheumatoid disease. Similarly, genetic susceptibility studies have identified important genetic risk factors for the development of (CCP positive) RA.1 The question that arises is whether these findings represent independent etiological pathways and could therefore be complimentary in the early diagnosis of RA.

Objectives:

To corroborate the knowledge of genetic differences between HC and RA patients and extend it to include the CSA stage of disease.

Methods:

We used three datasets sampled from the same region in the Netherlands: 1,085 healthy controls (HC), 530 CSA and 1,277 RA patients. CSA patients were monitored for a median of 2 years for conversion into clinically apparent inflammatory arthritis (CSAc) or not (CSAnc).2 We assessed the association between HLA SE and disease stage using logistic regression. The analysis was repeated in the CCP positive and CCP negative strata of both the CSA and the RA populations.

Results:

Consistent with previous studies, HLA SE was significantly enriched in RA patients compared to HC (OR 2.28) (Figure 1). HLA SE also differentiated HC vs CSAc (OR 1.69), CSAnc vs CSAc (OR 1.74), and CSAnc vs RA (0R 2.35). No difference was found in HC vs CSAnc and CSAc vs RA.

Conclusion:

HLA SE is more prevalent in patients who developed (rheumatoid) arthritis than in both healthy controls and CSA patients who do not progress to arthritis. The results presented here seem to indicate a clear distinction between CSA patients who develop arthritis and those who do not. We therefore believe that known RA genetics play a role in the development of arthritis rather than the CSA symptoms. While this relationship varies by CCP status, an independent effect remains. Studies into the broader role of genetics beyond HLA SE are currently underway.

Figure 1.

Distinguishing ability of HLA SE across HC, CSAc, CSAnc and RA in the full populations as well as in the CCP positive and negative stratifications. The arrowhead indicates the “case” status in each logistic regression. OR’s (95% CI) derived from regression coefficients indicate the change in odds ratio attributable to HLA SE positivity.

References:

[1]van der Helm-van Mil, A. H., et al. Arthritis and rheumatism, 2006. 54(4): p. 1117–1121.

[2]van Steenbergen, H.W., et al. Ann Rheum Dis, 2017. 76(3): p. 491-496.

Disclosure of Interests:

None declared

POS0001 PHENOTYPE AND FUNCTIONAL CHARACTERISTICS OF ANTIGEN-SPECIFIC, AUTO-REACTIVE B CELL RESPONSES REVEAL DIFFERENTIAL IMMUNOLOGICAL ACTIVITY IN PATIENTS WITH SYSTEMIC SCLEROSIS

Background:

Systemic Sclerosis (SSc) is a systemic autoimmune disease that carries the highest mortality burden among the rheumatic diseases. Disease risk and course are difficult to predict in individual patients, and anti-inflammatory and B-cell depleting therapies show varying results. >95% of SSc patients harbor autoantibodies. Among those, anti-topoisomerase antibodies (ATA) and anti-centromere antibodies (ACA) are most prevalent, mutually exclusive in individual patients and associate with distinct disease phenotypes. Despite these associations, the clinical value of both ATA and ACA for patient stratification within these phenotypes is limited. Here, we hypothesized that phenotypic and functional characteristics of the underlying autoreactive B cell responses could allow insights in differential ‘immunological disease activity’ in individual patients, thereby providing indications as to potential drivers of these responses as well as granularity as to which patients may benefit from targeted interventions.

Objectives:

To assess phenotypic and functional characteristics of anti-topoisomerase and anticentromere specific B cell responses in individual patients with SSc.

Methods:

Peripheral blood mononuclear cells (PBMC) from ATA- and ACA-positive SSc patients were cultured without stimulation or in the presence of CD40L-expressing fibroblasts, IL-21 and BAFF. Following culture, ATA- and ACA-IgG and -IgA were measured in culture supernatants by ELISA. In addition, PBMC were depleted of circulating plasmablasts by fluorescence activated cell sorting (FACS), and isolated plasmablasts were cultured separately. Furthermore, the presence of antigen-specific plasmablasts was confirmed by ELISPOT. Finally, the degree of spontaneous ATA secretion was correlated to the presence or absence of interstitial lung disease (ILD; based on high-resolution computed tomography). Healthy donors and patients with rheumatoid arthritis served as controls.

Results:

We observed that individual ATA- and ACA-positive SSc patients harbored circulating B cells that secrete either ATA-IgG or ACA-IgG upon stimulation, depending on their serotype. In addition, we noted spontaneous secretion of ATA-IgG and, more remarkably, extensive secretion of ATA-IgA in ATA-positive patients. This degree of spontaneous, antigen-specific IgA secretion was specific for the ATA response, while spontaneous ACA-IgA secretion was undetectable in patients harboring ACA. FACS experiments and ELISPOT showed that the spontaneous ATA-IgA and -IgG secretion was attributable to circulating plasmablasts. Of note, the degree of spontaneous ATA-IgG secretion was remarkably higher in patients with ILD than in those without.

Conclusion:

Our findings demonstrate that individual ATA-positive SSc patients harbor activated ATA-IgG and ATA-IgA B cell responses, as indicated by the spontaneous secretion of both ATA isotypes by circulating plasmablasts. Importantly, by taking the presence of plasmablasts as a proxy for recent B cell activation, our data suggest a link between the activity of the antigen-specific B cell response and the presence of ILD. In contrast, the ACA B cell response was far less active and lacked the active IgA component, which suggests a difference in the triggers driving these autoreactive B cell responses in patients. In fact, the remarkable ATA-IgA secretion points towards a potential mucosal trigger of the ATA response, which may be continuously active in individual patients.

Disclosure of Interests:

None declared.

POS0097 JOINT INFLAMMATION TENDS TO RECUR IN THE SAME JOINTS DURING THE RHEUMATOID ARTHRITIS DISEASE COURSE

Background:

It is unknown whether in the disease course of rheumatoid arthritis (RA), inflammation recurs in the same joints over time or is more variable in joint locations. Joint involvement patterns over time might provide clues about the underlying mechanisms causing local joint inflammation.

Objectives:

The aim of this study is to assess if local joint inflammation at presentation of RA tends to recur or persist in the same joints.

Methods:

Data from the BeSt study were used, a treat-to-target (DAS≤2.4) trial in newly diagnosed RA (ACR 1987 criteria) patients. During 10 years, for each patient 68 joints were assessed three-monthly (41 visits) by trained nurses for swelling (yes/no) and tenderness.

We analyzed the association between local joint swelling at baseline and later swelling of the same joint using a multilevel mixed-effects logistic regression model. Models were adjusted for joint location and for timepoint, with joints clustered within patients. A sensitivity analysis was done for the 25% most affected joints (MCP 1-3, PIP 2-3, wrists and MTP 2-4).

To investigate whether later swelling of a joint is predicted by baseline swelling of that same joint specifically, rather than by baseline swelling in general, a permutation test with 1000 permutations was performed. A p-value <0.05 indicates that joint swelling is better predicted by its baseline swelling than by baseline swelling of randomly selected other joints.

In a separate model, with an interaction term between baseline swelling and previous visit swelling (yes/no), we evaluated if the association between baseline swelling and later local swelling was influenced by whether later swelling was persistent (swelling at both the current and previous visit) or recurrent (swelling at current visit but not at the previous visit).

Results:

The 508 patients had a median (IQR) follow-up duration of 10 (6-10) years. At baseline, 8,137/34,423 (24%) assessed joints were scored as swollen. Baseline swelling was subsequently persistent in 21% of the joints with a median (IQR) duration of 1 (1-2) visit (± 3 months after baseline). In addition, after resolution of initial swelling, swelling recurred at least once in 46% of the joints with baseline swelling.

Baseline swelling was significantly associated with swelling in the same joint during follow-up (OR 2.37, 95% CI 2.30-2.43). A sensitivity analysis of the most affected joints showed similar results (OR 2.10 [95% CI 2.03-2.19]).

The permutation test showed a significant result with p<0.001, indicating that joint swelling is better predicted by baseline swelling of that same joint than by baseline swelling of other joints.

The association between baseline swelling and later local swelling was weaker in case of persistent swelling than in case of recurrent swelling (interaction term baseline swelling * swelling at previous timepoint ‘yes’: OR 0.80 [95% CI 0.75-0.85]).

Conclusion:

In newly diagnosed RA, over median 10 years of treatment to target DAS≤2.4, baseline swelling persisted in 21% of the joints, for median 3 months after baseline. Local recurrence after initial resolution occurred in 46% of the joints. Baseline joint swelling was significantly associated with local joint swelling during follow-up, even when taking into account the higher a priori chance of swelling in the joints that are most often affected, and joint swelling during follow-up was better predicted by baseline swelling of that particular joint than by baseline swelling of other joints. Local persistence and recurrence of joint swelling despite DAS≤2.4 steered treatment adjustments suggest that local joint conditions or even joint memory play a role in mechanisms of joint inflammation.

Acknowledgements:

We would like to thank all patients for their contribution as well as the rheumatologists who participated in the BeSt study group. We would also like to thank all other rheumatologists and trainee rheumatologists who enrolled patients in these studies, and all research nurses for their contributions.

Disclosure of Interests:

Sascha Heckert: None declared, Sytske Anne Bergstra: None declared, Xanthe Matthijssen: None declared, Yvonne Goekoop-Ruiterman: None declared, F. Fodili: None declared, Cornelia Allaart Grant/research support from: The original BeSt study was supported by a government grant from the Dutch insurance companies, with additional funding from Schering-Plough B.V. and Janssen B.V., Thomas Huizinga: None declared

POS0384 A NOVEL MECHANISM LINKING MUCOSAL BACTERIA WITH AUTOANTIBODY RESPONSES IN RA: ACETYLATED BACTERIAL LYSATE AS A MODEL ANTIGEN

Background:

Rheumatoid arthritis (RA) is characterized by autoantibodies against post-translationally modified proteins (AMPA) such as citrullinated, carbamylated and acetylated proteins. Importantly, these antibodies are highly multireactive, as they often recognize more than one of these post-translational modifications. Despite extensive research, the antigens inducing the breach of tolerance remain unknown, although microbial antigens are often suspected. Various bacteria are known to be capable of acetylation, therefore, it is intriguing to know what mechanisms can underlie the breach of tolerance towards acetylated proteins and development of anti-acetylated protein antibodies (AAPA).

Objectives:

To investigate whether acetylated proteins of bacterial origin (1) are recognized by human derived AMPA and AMPA expressing B cells; and (2) can induce AMPA development when used to immunize mice.

Methods:

Acetylated E. coli proteins were acquired with two separate methods (Figure 1A): by culturing E. coli in a condition promoting auto-acetylation (intrinsically acetylated bacterial proteins, IABP), or by directly acetylating lysate-derived proteins via a chemical reaction (extrinsically acetylated BP, EABP). Acetylated ovalbumin (AcOVA) served as positive control for AAPA induction in mice, non-acetylated BP (NABP) and phosphate buffer saline (PBS) served as negative control. Mice were immunized with these proteins and the resulting antibody response was studied by ELISA. Furthermore, EABP/IABP/NABP were investigated for recognition by human-derived AAPA with ELISA and AAPA-expressing B cells with spleen tyrosine kinase (Syk) phosphorylation assay; acetylated human fibrinogen and native fibrinogen served as positive and negative control.

Results:

Repetitive immunization of mice with EABP resulted in an AMPA response recognizing acetylated, carbamylated and citrullinated proteins. AMPA titers in these mice exceeded the titers in the positive control mice immunized with AcOVA and were substantially higher than in the NABP-immunized mice (Figure 1B). Human-derived monoclonal AAPA recognized EABP and IABP (not shown). B cell activation (measured by Syk phosphorylation) assay indicated that AAPA expressing B cells recognized EABP and (to a lesser extent) IABP, but not NABP (Figure 1C).

Conclusion:

Acetylated bacterial proteins are potent antigens that can induce cross-reactive AMPA responses in mice and they are recognized by human AAPA. This suggests that acetylated bacterial proteins could possibly be involved in the breach of tolerance in RA.

Acknowledgements:

We thank Dr. Can Araman and Prof. Chunaram Choudhary for their advice regarding optimization of bacterial auto-acetylation.

Disclosure of Interests:

None declared

POS0714 WHITE MATTER HYPERINTENSITIES LEAD TO REDUCED PSYCHOMOTOR SPEED IN PATIENTS WITH SYSTEMIC LUPUS ERYTHEMATOSUS AND NEUROPSYCHIATRIC SYMPTOMS

Background:

Cognitive impairment is common in patients with systemic lupus erythematosus (SLE) and neuropsychiatric (NP) symptoms, but the exact underlying structural brain correlates are unknown.

Objectives:

We aimed to compare cognitive function between groups of patients with different phenotypes of (NP)SLE and assessed the association between brain volumes, white matter hyperintensity (WMH) volume and cognitive function.

Methods:

Patients who visited the NPSLE clinic of the Leiden University Medical Center between 2007-2015 were included in this study (n=151; 42 ± 13 years, 91% female). In a multidisciplinary consensus meeting, phenotypes were established and neuropsychiatric symptoms were attributed to SLE (NPSLE, inflammatory (n=24) or ischemic (n=12)) or to minor involvement of SLE or other causes (minor/non-NPSLE (n=115)). All patients underwent standardized cognitive assessment of the four cognitive domains: global cognitive functioning (GCF), learning and memory (LM), executive functioning and complex attention (EFCA) and psychomotor speed (PS)). Cognitive dysfunction was defined as a T-score <40, with age, sex, gender and education matched individuals of the Dutch population as reference. In addition, automated volume measurements on brain MRI (white matter, grey matter, white matter hyperintensities (WMH) and total brain volume (TBV)) were performed. Patients with brain infarcts >1.5 cm were excluded. Cognitive function (Z-score) was compared between different NPSLE phenotypes using multiple regression analyses corrected for age, sex and education. Associations between brain volumes, WMH and cognitive function were assessed per phenotype using multiple regression analyses corrected for age, sex and intracranial volume.

Results:

Global cognitive functioning was impaired in 5%, learning & memory in 46%, and executive functioning & complex attention in 39% and psychomotor speed in 46% of all patients. Patients with inflammatory NPSLE showed the most cognitive impairment and reduced cognitive function compared to ischemic NPSLE (all domains) and minor/non-NPSLE (EFCA) (p <0.05).

Lower total brain volume and grey matter volume were associated with lower cognitive functioning in all domains (β: 0.00/0.01 (0.00; 0.01)) and lower white matter volume associated with lower LM, EFCA and PS (β: 0.00/0.01 (0.00; 0.01)) in all patients. Higher WMH volume associated with lower psychomotor speed (β: -0.14 (-0.32; -0.02)), especially in patients with inflammatory NPSLE (β: -0.36 (-0.60; -0.12).

Conclusion:

Reduced brain volume leads to reduced cognitive function in multiple cognitive domains in all patients with SLE and neuropsychiatric symptoms. Increased WMH volume leads to reduced psychomotor speed, especially in patients with inflammatory NPSLE.

Table 1.

Prevalence of cognitive impairment in patients with minor/non-NPSLE, inflammatory NPSLE and ischemic NPSLE

Global cognitive functioningLearning &memoryExecutive functioning & complex attentionPsychomotor speedAll patients(n = 151)8 (5)70 (46)57 (39)69 (46)Inflammatory NPSLE(n = 24) 3 (13) 14 (58) 12 (50) 12 (50)Ischemic NPSLE(n = 12) 0 (0) 6 (50) 2 (17) 2 (17)Minor/non-NPSLE(n = 115) 5 (4) 50 (44) 43 (38) 55 (49)

NPSLE = neuropsychiatric systemic lupus erythematosus.

Data represent n (%) of patients with cognitive impairment in the mentioned cognitive domain. Cognitive impairment was defined as cognitive function at least 1SD lower than the mean of an age, sex and education matched general Dutch population. The percentages were calculated from total number of patients with available scores: Global functioning: 23/24 inflammatory NPSLE and 113/115 minor/non-NPSLE; psychomotor speed 113/115 minor/non-NPSLE, executive function & complex attention: also 113/115. All tests were available for ischemic NPSLE.

Disclosure of Interests:

None declared

POS0843 A NEW RISK MODEL IS ABLE TO IDENTIFY SYSTEMIC SCLEROSIS PATIENTS WITH A LOW RISK OF DISEASE PROGRESSION

Background:

Disease course in Systemic Sclerosis (SSc) ranges from mild, to severe with progressive organ involvement within months. Guidelines for follow-up are mainly based on expert consensus, and advocate annual assessment. So far, no data driven guidelines exist that describe tailormade systematic assessments for individual patients in line with individual disease course.

Objectives:

To develop a prediction model to guide annual assessment of SSc patients tailored in accordance to disease activity.

Methods:

A machine learning approach was used to develop a model that can identify patients without disease progression. SSc patients included in the prospective Leiden SSc cohort and fulfilling the ACR/EULAR 2013 criteria were included. The primary endpoint in the prediction model was disease progression which was defined as progression in ≥1 organ system, and/or start of immunosuppression or death between the two most recent visits. Using elastic-net-regularization, and including 90 independent clinical variables (100% complete), we trained the model on 75% and validated it on 25% of the patients in order to perform internal validation of the final model. We optimized the model on negative predictive value (NPV) to minimize the likelihood of missing progression. By expert assessment of the test characteristics, including swarm plots of the probability scores, cut-offs were identified for low, intermediate and high risk for disease progression.

Results:

Of the 492 SSc patients (range of follow-up: 2-10yrs), disease progression during follow-up was observed in 52% (median time 4.9yrs), including myocardial progression in 29%, lung progression in 23%, skin progression in 16%, and death in 12%. Performance of the model in the test set showed an AUC-ROC of 0.66. Probability score cutoffs were defined: low risk for disease progression (<0.197, NPV:1.0; 29% of patients), intermediate risk (0.197-0.223, NPV:0.82; 27%) and high risk (>0.223, NPV:0.78; 44%). The predictive variables included in the model were: previous use of cyclophosphamide or corticosteroids, start with immunosuppressive drugs, previous gastrointestinal progression, previous cardiovascular event, pulmonary arterial hypertension, modified Rodnan Skin Score, creatinine kinase, and diffusing capacity for carbon monoxide.

Conclusion:

Our machine-learning-assisted model for disease progression enabled us to classify 29% of SSc patients as ‘low risk’. In this group annual assessment programs could be less extensive than indicated by international guidelines.

Baseline characteristicsTotaln=492Non-ProgressorsN=235ProgressorsN=257DemographicsFemale, n (%)389 (79)193 (82)196 (76)Age, mean (SD)55 (14)55 (15)55 (13)Disease duration nonRP, median (IQR)3.2 (0.9-10.3)3.5 (0.8-10.5)3.6 (1.1-9.3)Organ involvementDcSSc, n (%)118 (24)34 (15)84 (33)DLCO% of pred, mean (SD)66 (18)69 (18)64 (17)FVC% of pred, mean (SD)98 (23)96 (24)97 (21)ILD on HRCT, n (%)183 (37)66 (28)117 (46)PAH, n (%)26 (5)10 (4)16 (6)GAVE, n (%)9 (2)4 (2)5 (2)Cardiac involvement, n (%)28 (6)14 (6)14 (5)Myositis, n (%)8 (2)6 (3)2 (1)Renal crisis, n (%)14 (3)6 (3)8 (3)AutoantibodiesAnti-centromere, n (%)194 (39)118 (50)76 (30)Anti-topoisomerase, n (%)116 (24)42 (18)74 (29)

RP=raynaud phenomenon, dcSSc= diffuse cutaneous systemic sclerosis, mRSS=modified rodnan skin score, DU=digital ulcera, DLCO= single-breath diffusing capacity for carbon monoxide, FVC= forced vital capacity, ILD=interstitial lung disease, HRCT= high resolution computed tomography, PAH= pulmonary arterial hypertension, GAVE= gastric antral vascular extasia.

Disclosure of Interests:

None declared

AB0668 THE ONSET OF RHEUMATOID ARTHRITIS AFTER COVID-19 – COINCIDENCE OR CONNECTED?

Background:

COVID-19, caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), has been suggested to induce autoimmune phenomena. Multiple studies have reported the presence of autoantibodies in patients with COVID-19. Also the presence of anti-citrullinated protein antibodies (ACPA) and flaring of rheumatoid arthritis (RA) after COVID-19 has been described.[1, 2] Furthermore, in rheumatology clinics patients may present with polyarthritis compatible with RA shortly after SARS-CoV-2 infection. However, it is unclear how often ACPA occur after COVID-19 and whether preceding COVID-19 impacts on disease presentation of RA and phenotype of the ACPA response.

Objectives:

This study aims to determine the seroprevalence of ACPA after COVID-19 and to investigate the association between preceding COVID-19 infection and disease presentation of new-onset RA, including clinical phenotype and autoantibody response.

Methods:

To estimate the prevalence of ACPA after COVID-19 we measured ACPA IgG in samples from 61 patients visiting the specialized post-COVID outpatient clinic of the LUMC 5 weeks after hospitalization, using routine tests or in-house ELISA. Furthermore, we identified 5 patients presenting with polyarthritis compatible with RA after SARS-CoV-2 infection. To study the impact of COVID-19 on disease presentation, we examined clinical phenotype, autoantibody isotype positivity and ACPA IgG variable domain (V-domain) glycosylation of these patients and compared these features to regular RA patients. Autoantibody isotypes, including rheumatoid factor (RF) IgM/IgA, anti-CCP2 IgG/ IgM/IgA and anti-carbamylated protein antibodies (anti-CarP) IgG were measured using in-house ELISA’s. The percentage of V-domain glycosylation of purified ACPA IgG was measured with UHPLC.

Results:

None of the 61 post-COVID patients tested positive for ACPA 5 weeks after hospitalization, except two patients previously diagnosed with ACPA-positive RA. Thus, we could not observe an increase in ACPA-positivity shortly after COVID-19. Of the 5 patients who developed polyarthritis compatible with RA after SARS-CoV-2 infection, the average age was 63.6 years and 2/5 were female. 4/5 patients had been hospitalized due to severe COVID-19. On average, joint complaints started 6.6 weeks after infection, although two patients reported symptoms before infection. 4/5 patients fulfilled the ACR 2010 criteria for RA. Three patients (patient 1, 4, 5) were phenotypically very similar to regular new-onset RA patients. Patient 3 had a history of seronegative RA and had been in DMARD-free remission for 5 years. She flared 6 weeks after SARS-CoV-2 infection. Patient 2 had a remarkably different presentation. He was admitted with a suspected septic polyarthritis or pneumonia with reactive polyarthritis 6 weeks after COVID-19. ACPA level was low positive. The patient died unexpectedly after two days and autopsy revealed dilating myocarditis of unclear underlying cause. No causative pathogen could be identified.

Previous studies have shown that RA-patients are most often either seronegative or triple-positive for RF, ACPA and anti-CarP antibodies. Autoantibody measurements on sera of the post-COVID polyarthritis patients revealed a similar pattern (Figure 1A) with two patients being completely seronegative, and three patients positive for a range of autoantibodies. In all post-COVID samples, the percentage of ACPA IgG V-domain glycosylation was significantly increased compared to total IgG (Figure 1B), similar as in regular RA.

Conclusion:

In conclusion, we found that the seroprevalence of ACPA is not increased post-COVID and that most patients presenting with polyarthritis after COVID-19 resemble regular RA patients, both regarding clinical phenotype and autoantibody characteristics. Although sample size and follow-up was limited, it appears that RA post-COVID may be coincidence rather than connected.

References:

[1]Vlachoyiannopoulos et al. Ann Rheum Dis, 2020.

[2]Perrot et al. The Lancet Rheumatology, 2020.

Disclosure of Interests:

None declared.

SAT0585 GEO-EPIDEMIOLOGY OF AUTOANTIBODIES IN RA: DIFFERENT PREVALENCES IN FOUR ETHNICALLY DIVERSE RA POPULATIONS

Background:

Rheumatoid arthritis (RA) has been described in virtually every ethnic population. Most RA patients harbor anti-modified protein antibodies (AMPAs), including anti-citrullinated protein (ACPA), anti-carbamylated protein (anti-CarP), anti-malondialdehyde acetaldehyde (anti-MAA), and anti-acetylated protein antibodies (AAPA). However, it is unclear whether differences exist in the AMPA response between different ethnic groups. Such differences could provide new clues to genetic and environmental factors contributing to autoantibody development.

Objectives:

To investigate the prevalence of different AMPA in four ethnically diverse RA populations, and their association with smoking.

Methods:

Enzyme-linked immunosorbent assays were used to measure anti-CarP IgG, anti-MAA IgG (both in-house), and anti-acetylated vimentin IgG (Orgentec) in ACPA-positive sera of Dutch (NL, n=103), Japanese (JP, n=174), Canadian First Nations People (FN, n=100), and black South Africans (SA, n=67) fulfilling the 1987 ACR classification criteria for RA. Ethnicity-matched local healthy controls were used to calculate cohort-specific cut-offs. Logistic regression was used to identify whether ever-smoking was associated with AMPA seropositivity in each cohort, corrected for age, gender, and disease duration. Random-effects meta-analysis was used to pool the resulting odds ratios (OR).

Results:

For all three AMPAs, median levels were higher in FN and especially SA than NL and JP patients (Figure 1). The median autoantibody levels in arbitrary units (in % of patients positive) for NL, JP, FN and SA RA patients were: anti-CarP IgG: 1157 (47%), 994 (43%), 1642 (58%) and 2336 (76%) (p<0.001); anti-MAA IgG: 131 (29%), 179 (22%), 251 (29%) and 257 (53%) (p<0.001); AAPA: 133 (20%), 136 (17%), 153 (38%) and 316 (28%) (p<0.001). Prevalence, meaning positivity, also differed significantly between cohorts for all AMPAs (p<0.001).There were also marked differences in total IgG levels in mean (SD) g/L: 13 (4) for NL, 17 (6) for JP, 18 (6) for FN, and 25 (8) for SA (p<0.001). When the autoantibody levels were normalized to total IgG, the differences in became less pronounced between cohorts (Figure 2). The median arbitrary units per g/L Total IgG for NL, JP, FN and SA RA patients were: anti-CarP IgG: 54, 25, 53, and 79; anti-MAA IgG: 6, 5, 8, and 9; and AAPA: 2, 2, 2, and 3, suggesting that autoantibody level differences may partly correspond to cohort-specific differences in total IgG, although the overall trend of higher levels in SA persisted. There was no association between smoking and anti-CarP or anti-MAA positivity, with pooled OR (95% CI) of 1.31 (0.79-2.18) and 0.85 (0.46-1.56), respectively. However, smoking was positively and consistently associated with AAPA positivity in each cohort: pooled OR (95% CI) of 2.01 (1.06-3.81).

Conclusion:

In these ACPA-positive ethnically diverse RA populations, levels and prevalence of various AMPAs differ, suggesting that ethnic background and environment may influence the development of the autoantibody response in RA. Despite these differences, our results imply smoking as a consistent risk factor for AAPA across different ethnic backgrounds.

Disclosure of Interests:

Emma C. de Moel: None declared, Veerle Derksen: None declared, Leendert A Trouw: None declared, Chikashi Terao: None declared, Mohammed Tikly: None declared, Hani El-Gabalawy: None declared, Holger Bang Grant/research support from: Employee of Orgentec Diagnostika, Thomas Huizinga Grant/research support from: Ablynx, Bristol-Myers Squibb, Roche, Sanofi, Consultant of: Ablynx, Bristol-Myers Squibb, Roche, Sanofi, Rene Toes: None declared, Diane van der Woude: None declared

SAT0104 MAINTENANCE OF SDAI REMISSION AND PATIENT-REPORTED OUTCOMES (PROS) FOLLOWING DOSE DE-ESCALATION OF ABATACEPT IN MTX-NAÏVE, ANTI-CITRULLINATED PROTEIN ANTIBODY (ACPA)+ PATIENTS WITH EARLY RA: RESULTS FROM AVERT-2, A RANDOMISED PHASE IIIB STUDY

Background:

The Phase IIIbAssessingVeryEarlyRATreatment (AVERT)-2 trial (NCT02504268) evaluated SC abatacept (ABA) + MTX vs ABA placebo (PBO) + MTX in ACPA+ patients (pts) with early, active RA.1Results from the 56-wk induction period (IP) showed a significantly greater proportion of pts treated with ABA + MTX (vs MTX alone) reported clinically meaningful improvements in HAQ-DI, global disease activity and pain, which were sustained at 52 wks.2

Objectives:

To report maintenance of SDAI remission and PROs from the AVERT-2 de-escalation (D-E) period.

Methods:

Pts received blinded SC ABA (125 mg once wkly [QW]) + MTX or ABA PBO + MTX induction treatment for 56 wks. In this analysis, pts who completed induction with ABA + MTX and had sustained SDAI remission (≤3.3 at Wks 40 and 52) were re-randomised 1:1:1 to ABA QW + MTX, stepwise D-E (ABA every other wk + MTX for 24 wks then ABA PBO + MTX for 24 wks), or ABA QW + MTX PBO for 48 wks in the D-E period. PROs included physical function (HAQ-DI [0–3; decrease=improvement] and Short-Form 36 [SF-36] v2.0 Physical Functioning Scale [PFS]; 0–100; increase=improvement), and fatigue (Functional Assessment of Chronic Illness Therapy-Fatigue [FACIT-F] score; 0–52; decrease=improvement). Endpoints included: proportion of pts in SDAI remission and pts with HAQ-DI response (decrease from IP Day [D]1 in HAQ-DI ≥0.30); adjusted mean change (adMC) from D-E D1 in HAQ-DI, SF-36 PFS or FACIT-F to D-E Wk 48. adMCs were estimated using a mixed effect model with repeated measures.

Results:

147 ABA + MTX-treated pts were re-randomised in the D-E period. Across re-randomised arms, the range of mean scores was 1.87–2.52 for SDAI and 0.18–0.30 for HAQ-DI at entry into D-E period (D-E D1). 74% of pts receiving ABA QW + MTX maintained SDAI remission at D-E Wk 48 (Fig 1); this proportion was higher than in the ABA withdrawal and ABA QW + MTX PBO arms. Pts continuing ABA QW + MTX maintained HAQ-DI response during D-E (Fig 1), but by D-E Wk 48 the proportion of pts with HAQ-DI response in the ABA withdrawal arm declined by 30%. At D-E Wk 48, a small numerical decrease (adMC –0.04) in HAQ-DI was observed in the ABA QW + MTX arm; increases were seen in the withdrawal (adMC 0.26) and ABA QW + MTX PBO arms (adMC 0.16). By D-E Wk 48, SF-36 PFS increased (adMC 1.68) in the ABA QW + MTX arm but decreased in the withdrawal (adMC –3.34) and ABA QW + MTX PBO (adMC –1.45) arms. FACIT-F score increased during D-E in all arms, but the increase at D-E Wk 48 was lower in the ABA QW + MTX arm (adMC 0.79) vs the withdrawal (adMC 4.12) and ABA QW + MTX PBO (adMC 2.41) arms. Similar trends were seen for other PROs including Work Productivity and Activity Impairment-RA; while activity impairment remained stable in the ABA QW + MTX arm, there was a trend for worsening in the withdrawal arm.

Conclusion:

In the AVERT-2 D-E period, continued combination therapy (abatacept + MTX) resulted in maintenance of benefits on PROs, particularly physical functioning, in seropositive pts with early RA. D-E of abatacept followed by complete withdrawal was associated with the greatest loss of remission as well as worsening of PROs. The PRO results corresponded well to the maintenance of clinical (SDAI) remission.

References:

[1]Emery P, et al. ACR 2018; San Diego, USA: Poster 563.

[2]Emery P, et al. ACR 2019; Atlanta, USA: Poster 1423.

Acknowledgments:

Joanna Wright (medical writing, Caudex; funding: Bristol-Myers Squibb)

Disclosure of Interests:

Paul Emery Grant/research support from: AbbVie, Bristol-Myers Squibb, Merck Sharp & Dohme, Pfizer, Roche (all paid to employer), Consultant of: AbbVie (consultant, clinical trials, advisor), Bristol-Myers Squibb (consultant, clinical trials, advisor), Lilly (clinical trials, advisor), Merck Sharp & Dohme (consultant, clinical trials, advisor), Novartis (consultant, clinical trials, advisor), Pfizer (consultant, clinical trials, advisor), Roche (consultant, clinical trials, advisor), Samsung (clinical trials, advisor), Sandoz (clinical trials, advisor), UCB (consultant, clinical trials, advisor), Yoshiya Tanaka Grant/research support from: Asahi-kasei, Astellas, Mitsubishi-Tanabe, Chugai, Takeda, Sanofi, Bristol-Myers, UCB, Daiichi-Sankyo, Eisai, Pfizer, and Ono, Consultant of: Abbvie, Astellas, Bristol-Myers Squibb, Eli Lilly, Pfizer, Speakers bureau: Daiichi-Sankyo, Astellas, Chugai, Eli Lilly, Pfizer, AbbVie, YL Biologics, Bristol-Myers, Takeda, Mitsubishi-Tanabe, Novartis, Eisai, Janssen, Sanofi, UCB, and Teijin, Vivian Bykerk: None declared, Thomas Huizinga Grant/research support from: Ablynx, Bristol-Myers Squibb, Roche, Sanofi, Consultant of: Ablynx, Bristol-Myers Squibb, Roche, Sanofi, Gustavo Citera Grant/research support from: AbbVie, Amgen, Eli Lilly, Gema, Genzyme, Novartis and Pfizer Inc, Consultant of: AbbVie, Amgen, Eli Lilly, Gema, Genzyme, Novartis and Pfizer Inc, Clifton Bingham Grant/research support from: Bristol-Myers Squibb, Consultant of: Bristol-Myers Squibb, Subhashis Banerjee Shareholder of: AbbVie, Bristol-Myers Squibb, Lily, Pfizer, Employee of: Bristol-Myers Squibb (current); AbbVie, Lily, Pfizer (past), Sean Connolly Shareholder of: Bristol-Myers Squibb, Employee of: Bristol-Myers Squibb, Joe Zhuo Shareholder of: Bristol-Myers Squibb, Employee of: Bristol-Myers Squibb, Robert Wong Shareholder of: Bristol-Myers Squibb, Employee of: Bristol-Myers Squibb, Kuan-Hsiang Gary Huang Shareholder of: Bristol-Myers Squibb, Employee of: Bristol-Myers Squibb, Karissa Lozenski Employee of: Bristol-Myers Squibb, Yedid Elbez Consultant of: Bristol-Myers Squibb, Roy Fleischmann Grant/research support from: AbbVie, Akros, Amgen, AstraZeneca, Bristol-Myers Squibb, Boehringer, IngelhCentrexion, Eli Lilly, EMD Serono, Genentech, Gilead, Janssen, Merck, Nektar, Novartis, Pfizer, Regeneron Pharmaceuticals, Inc., Roche, Samsung, Sandoz, Sanofi Genzyme, Selecta, Taiho, UCB, Consultant of: AbbVie, ACEA, Amgen, Bristol-Myers Squibb, Eli Lilly, Gilead, GlaxoSmithKline, Novartis, Pfizer, Sanofi Genzyme, UCB

SAT0099 BMI AND TREATMENT SURVIVAL IN RA PATIENTS STARTING TREATMENT WITH TNFΑ-INHIBITORS: LONG TERM FOLLOW-UP IN THE REAL LIFE METEOR REGISTRY

Background:

BMI appears to be associated with treatment response on TNFi(nhibitors) in rheumatoid arthritis (RA), but large heterogeneity between studies exists. More extreme BMI categories are rarely studied and it is unclear if differences exist between various TNFi.1

Table 1

Characteristics of RA patients

Female, n (%)935 (79.8)Age, years*51.0 ±13.7Current smokers, n (%)256 (23.2)RF Positivity, n (%)404 (55.6)Anti-CCP Positivity, n (%)430 (58.2)X-ray Erosion, n (%)317 (61.9)ESR, mm/h*31.2±21.9CRP, mg/L*17.2±3.9DAS 28-CRP*3.8±1.6VAS global*46.6±28.6HAQ*0.9±0.7First TNFi Etanercept, n (%)525 (38.7) Adalimumab, n (%)379 (27.9) Infliximab, n (%)118 (8.7) Certolizumab, n (%)188 (13.8) Golimumab, n (%)147 (10.9)

* mean ±S.D

RF, Rheumatoid factor; Anti-CCP, Anti- cyclic citrullinated peptid; ESR, erythrocyte sedimentation rate; CRP, C-reactive protein; DAS28–CRP, Disease Activity Score using 28 joints–CRP; VAS, Visual analog scale; HAQ, Health Assessment Questionnaire

Objectives:

To study whether there is an association between BMI category and drug survival in RA patients starting treatment with various TNFi in a real life longitudinal international registry.

Methods:

Data from 5230 RA patients starting a TNFi were included from the METEOR registry. Timing of follow-up visits was daily practice based. Follow-up was censored at 5000 days (±13.5 years). Patients were divided into 6 BMI categories (WHO definition): underweight BMI <18.5, normal weight BMI 18.5-25, pre-obesity BMI 25-30, obesity class I BMI 30-35, class II BMI 35-40, and class III BMI >40. Missing data were imputed using chained equations. The association between BMI category and time on treatment was investigated using Kaplan-Meier (KM) curves and Cox regression analyses, for time on first TNFi and for the first prescribed course of adalimumab (ADA), etanercept (ETA) and infliximab (IFX) separately. All analyses were adjusted for the potential confounders age, gender, smoking, baseline DAS28, concomitant glucocorticoid use and country. Potential effect modification by reported pain was tested by adding an interaction term between BMI category and baseline pain category (VAS pain 0-25, 25-50, 50-75 and 75-100).

Results:

Most patients had a normal weight (46%) or pre-obesity (32%). 4% of patients were underweight, 10% had obesity class I, 3% obesity class II and 1% obesity class III. N=2936 patients ever started ETA, n=2069 ADA, n=1390 IFX, n=263 certolizumab and n=84 golimumab. The KM curve in fig 1A shows TNFi survival in patient starting their first TNFi per BMI category. Patients with normal weight and pre-obesity had longest drug survival and patients with obesity class II and especially patients with obesity class III had shortest drug survival. The adjusted Cox regression support these findings, with statistically significantly shorter drug survival for patients with obesity class III [HR (95% CI) 1.67 (1.29; 2.18)] and class II [1.28 (1.06; 1.54)], but also for underweight patients [1.3 (1.07; 1.58)], compared to normal weight patients. KM curves for individual TNFi showed shorter drug survival on ADA for patients with obesity class II and III (fig 1B), on ETA for patients with obesity, especially in class III (fig 1C) and on IFX, for patients with obesity class II and III and underweight patients (fig 1D). After adjustment in Cox regression, statistical significant BMI-drug survival associations remained for patients with pre-obesity starting ADA [HR (95% CI) 0.86 (0.75; 0.99)], for patients starting ETA with obesity class II [HR (95% CI) 1.27 (0.98; 1.65) or class III [1.79 (1.25; 2.55)] and for patients on IFX who were underweight [HR (95% CI) 1.82 (1.20; 2.76)] or in obesity class II [1.49 (0.98; 2.26)]. No effect modification was found for reported pain.

Conclusion:

Both underweight (as identified in IFX treated patients) and overweight patients (in ADA, ETA and IFX treated patients) discontinued a first TNFi treatment earlier than normal weight patients. Reported pain was not the main determinant. It remains uncertain what determines TNFi survival in individual patients.

References:

[1] Singh, et al.PloS One2018; 13:e0195123

Disclosure of Interests:

Sytske Anne Bergstra: None declared, David Vega-Morales: None declared, Elizabeth Murphy: None declared, Marieke de Buck: None declared, Karen Solomon-Escoto: None declared, Thomas Huizinga Grant/research support from: Ablynx, Bristol-Myers Squibb, Roche, Sanofi, Consultant of: Ablynx, Bristol-Myers Squibb, Roche, Sanofi, Cornelia Allaart: None declared

THU0327 EFFECT OF IMMUNOSUPPRESSIVE MEDICATION ON GASTRO-INTESTINAL INVOLVEMENT IN SYSTEMIC SCLEROSIS PATIENTS STRATIFIED FOR DISEASE DURATION

Background:

Gastrointestinal tract (GIT) involvement is associated with high morbidity in systemic sclerosis (SSc) but the data on its impact from unselected and well characterized SSc cohorts are scarce. Currently, the effect of immunosuppressive (IS) treatment on GIT involvement is largely unknown.

Objectives:

To evaluate the severity and worsening of GIT involvement in two prospective SSc cohorts. To assess factors associated with severity of GIT involvement, stratified for disease duration. To evaluate effect of IS treatment on worsening of GIT involvement.

Methods:

All SSc patients fulfilling the 2013 SSc classification criteria from two SSc cohorts were evaluated. Incident SSc was defined as disease duration since first symptom non-Raynaud < 24 months at first presentation. GIT involvement was assessed by the UCLA GIT 2.0 score at baseline and after one year to assess worsening of GIT involvement. Worsening was defined as change > minimal clinical important difference for total score and for each of the seven subdomains. GIT involvement was defined as present if the patients reported symptoms resulting in a score of ≥0.01 and was segregated into mild ≥0.01 (<0.5 or for fecal incontinence and distention/bloating <1.01), moderate (≥0.5 or for fecal incontinence and distention/bloating ≥1.01) or severe GI symptoms (> 1.01 or for distension/bloating > 1.61 or for fecal soiling > 2.01). Logistic regression was applied to identify risk factors associated with GIT involvement at baseline. The effect of IS treatment on worsening on each of the subdomains after one year was evaluated with logistic regression, with adjustment for baseline disease duration and severity.

Results:

In total, 834 SSc patients were included; 236 (28%) had incident disease (table 1). Incident cases (IC) showed comparable severity of GIT involvement compared to non-incident cases (NIC) except for significantly less severe reflux and distension/bloating (figure 1). Logistic regression showed female sex (OR 8.5(1.1-36.01)) and smoking (OR 2.9(1.2-7.3)) to be associated with GIT severity at baseline in IC; in NIC anti-centromere antibody (OR 1.7(1.3-2.2)) was additionally associated with GIT severity. The use of IS at baseline did not associate with GI severity at baseline. In total n = 685 (82% never had IS treatment (83% NIC, 81% IC); of these 258 (38%) started with IS after baseline assessment (52% IC, 32% NIC, p =0.02). When comparing change of GIT involvement after one year between those who started IS and those who did not, worsening of GI symptoms occurred more frequently in patients who started IS treatment (figure 2), but notably, patients in this group were also more frequently anti-topoisomerase positive, had ILD, and diffuse disease subset compared to the patients without IS treatment; age and sex were comparable. In the logistic regression with adjustment for disease duration and severity, there were no significant associations between IS treatment and worsening on GIT involvement.

Conclusion:

Regardless disease duration, about 1/3 of all SSc patients had moderate-severe GIT involvement. Disease duration and treatment initiation with IS did not have a significant influence on worsening of GIT involvement.

Table:

Baseline characteristicsNon-inception cohort(n=598)Inception cohort(n=236)Female, n(%)504 (85)180 (76)Age, mean(SD)55 (13)56 (14)Disease duration non Raynaud-Phenomenon, median (IQR)8.8 (4.8-14.4)0.7 (0.3-1.2)Diffuse cutaneous subset, n(%)119 (20)67 (28)Interstitial lung disease, n(%)233 (39)71 (30)Anti-centromere, n(%)296 (50)96 (41)Immunosuppresive treatment at baseline, n(%)102 (17)44 (19))Duration of treatment at baseline in years, mean (SD)4.1 (4.8)1.2 (2.9)Methotrexate, n(%)54 (9)24 (10)Mycophenolate mofetil, n(%)25 (4)13 (6)Hydroxycholoquine, n(%)20 (3)7 (3)Cyclofosfamide, n(%)1 (1)10 (4)Azathioprine, n(%)11 (2)2 (1)Corticosteroids, n(%)58 (10)27 (11)

Acknowledgments:

NA

Disclosure of Interests:

Nina van Leeuwen: None declared, Håvard Fretheim: None declared, Øyvind Molberg: None declared, Thomas Huizinga Grant/research support from: Ablynx, Bristol-Myers Squibb, Roche, Sanofi, Consultant of: Ablynx, Bristol-Myers Squibb, Roche, Sanofi, Jeska de Vries-Bouwstra: None declared, Anna-Maria Hoffmann-Vold Grant/research support from: Boehringer Ingelheim, Consultant of: Boehringer Ingelheim, Actelion, Bayer, GlaxoSmithKline, Speakers bureau: Boehringer Ingelheim, Actelion, Roche

AB1243 TRAINING AND VALIDATION OF A MULTIVARIATE PREDICTOR OF RISK OF RADIOGRAPHIC PROGRESSION FOR PATIENTS WITH RHEUMATOID ARTHRITIS

Background:

The multi-biomarker disease activity (MBDA) score, adjusted for age, sex and adiposity (MBDAadj), has been shown to be better than several conventional disease activity measures for predicting risk for radiographic progression (RP) in patients with rheumatoid arthritis (RA).1Serologic status and other non-disease activity measures are also predictive of RP risk. Combining them with the MBDAadjshould result in a stronger prognostic test for RP than any one measure alone.

Objectives:

Develop a multivariate model for predicting risk for RP that includes the adjusted MBDA score and other known predictors of RP.

Methods:

Four RA cohorts were used, two for training (OPERA and BRASS, n=555) and two for validation (SWEFOT and Leiden, n=397). Each pair of cohorts was heterogeneous in disease duration and treatment history. BMI data were not available for one validation cohort, so a BMI surrogate was modeled using forward selection with the two training cohorts and 3 others (CERTAIN, InFoRM, RACER) (N=1411). An RP risk score was then trained using forward selection in a linear mixed-effects regression, considering disease-related and demographic variables as predictors of change in modified total Sharp score over one year (ΔmTSS), with a random effect on cohort. The RP risk score was validated as a predictor of RP with two cutoffs (ΔmTSS >3 and >5) using logistic mixed-effects regression. Odds ratios (OR) and 95% profile likelihood-based confidence intervals (CI) were calculated from the models and significance was assessed by likelihood ratio tests. Risk curves were generated to show probability of RP as a function of the RP risk score.

Results:

The BMI surrogate included leptin, sex, age and age2and correlated well with BMI (ρ = 0.76). In training, the most significant independent predictors of RP were MBDAadj(p = 0.00020), seropositivity (p = 9.3 x 10-5), BMI surrogate score (p = 0.013) and use of targeted therapy (p = 0.0026). The final model was: RP risk score = 0.024 x MBDAadj+ 0.093 if seropositive – 0.063 x BMI surrogate score – 0.61 if using a targeted therapy. In validation, the OR (95% CI) of the RP risk score for predicting ΔTSS >3 or >5 were 2.2 (1.6, 3.2) (p = 2.6 × 10-6) and 3.1 (2.0, 5.0) (p = 5.7 × 10-8), respectively (Figure 1). The odds of a patient having RP increases by 50% for each 21-unit or 15-unit increase in MBDAadj, for RP defined as ΔTSS >3 or >5, respectively.

Figure 1.

Conclusion:

A multivariate model containing adjusted MBDA score, seropositivity, a BMI surrogate and use of targeted therapy has been trained and validated as a prognostic test for radiographic progression in RA.

References:

[1]Curtis, et al.Rheumatology [Oxford].2018;58:874

Disclosure of Interests:

Thomas Huizinga Grant/research support from: Ablynx, Bristol-Myers Squibb, Roche, Sanofi, Consultant of: Ablynx, Bristol-Myers Squibb, Roche, Sanofi, Michael E. Weinblatt Grant/research support from: BMS, Amgen, Lilly, Crescendo and Sonofi-Regeneron, Consultant of: Horizon Therapeutics, Bristol-Myers Squibb, Amgen, Abbvie, Crescendo, Lilly, Pfizer, Roche, Gilead, Nancy Shadick Grant/research support from: Mallinckrodt, BMS, Lilly, Amgen, Crescendo Biosciences, and Sanofi-Regeneron, Consultant of: BMS, Cecilie Heegaard Brahe: None declared, Mikkel Ǿstergaard Grant/research support from: AbbVie, Bristol-Myers Squibb, Celgene, Merck, and Novartis, Consultant of: AbbVie, Bristol-Myers Squibb, Boehringer Ingelheim, Celgene, Eli Lilly, Hospira, Janssen, Merck, Novartis, Novo Nordisk, Orion, Pfizer, Regeneron, Roche, Sandoz, Sanofi, and UCB, Speakers bureau: AbbVie, Bristol-Myers Squibb, Boehringer Ingelheim, Celgene, Eli Lilly, Hospira, Janssen, Merck, Novartis, Novo Nordisk, Orion, Pfizer, Regeneron, Roche, Sandoz, Sanofi, and UCB, Merete L. Hetland Grant/research support from: BMS, MSD, AbbVie, Roche, Novartis, Biogen and Pfizer, Consultant of: Eli Lilly, Speakers bureau: Orion Pharma, Biogen, Pfizer, CellTrion, Merck and Samsung Bioepis, Saedis Saevarsdottir Employee of: Part-time at deCODE Genetics/Amgen Inc, working on genetic research unrelated to this project, Megan Horton Shareholder of: Myriad Genetics, Inc., Employee of: Myriad Genetics, Inc., Brent Mabey Shareholder of: Myriad Genetics, Inc., Employee of: Myriad Genetics, Inc., Darl Flake Shareholder of: Myriad Genetics, Inc., Employee of: Myriad Genetics, Inc., Rotem Ben-Shachar Shareholder of: Myriad Genetics, Inc., Employee of: Myriad Genetics, Inc., Eric Sasso Shareholder of: Myriad Genetics, Inc., Employee of: Myriad Genetics, Inc., Alexander Gutin Shareholder of: Myriad Genetics, Inc., Employee of: Myriad Genetics, Inc., Elena Hitraya Shareholder of: Myriad Genetics, Inc., Employee of: Myriad Genetics, Inc., Jerry Lanchbury Shareholder of: Myriad Genetics, Inc., Employee of: Myriad Genetics, Inc., Jeffrey Curtis Grant/research support from: AbbVie, Amgen, Bristol-Myers Squibb, Corrona, Janssen, Lilly, Myriad, Pfizer, Regeneron, Roche, UCB, Consultant of: AbbVie, Amgen, Bristol-Myers Squibb, Corrona, Janssen, Lilly, Myriad, Pfizer, Regeneron, Roche, UCB

SAT0297 DIFFERENTIAL PHENOTYPES OF DISEASE-SPECIFIC AUTO-REACTIVE B CELL RESPONSES IN PATIENTS WITH SYSTEMIC SCLEROSIS

Background:

Systemic Sclerosis (SSc) carries the highest mortality burden among the rheumatic diseases. >95% of SSc patients harbor autoantibodies. Anti-topoisomerase antibodies (ATA) and anti-centromere antibodies (ACA) are most prevalent, mutually exclusive in individual patients, associate with distinct disease phenotypes and predict disease. Whether and how these auto-reactive B cell responses contribute to disease, however, is currently unclear.

Objectives:

To delineate phenotypic and functional characteristics of anti-topoisomerase and anti-centromere specific B cell responses in individual patients with SSc.

Methods:

Peripheral blood mononuclear cells (PBMC) obtained from ATA- and ACA-positive SSc patients were cultured without stimulation or in the presence of CD40L-expressing fibroblasts, IL-21 and BAFF. In addition, PBMC were depleted of circulating plasmablasts (CD19+CD20-CD27++) by fluorescence activated cell sorting (FACS), and isolated plasmablasts were cultured separately. ATA- and ACA-IgG and -IgA were measured in culture supernatants by ELISA. B cell subsets were defined by flow cytometry. Healthy donors and patients with rheumatoid arthritis served as controls.

Results:

We observed that ATA- and ACA-positive SSc patients harbour circulating B cells that secrete either ATA-IgG or ACA-IgG upon stimulation, depending on their serotype. In addition, we noted spontaneous secretion of ATA-IgG and, more remarkably, extensive secretion of ATA-IgA in ATA-positive patients. This degree of spontaneous, antigen-specific IgA secretion was specific for the ATA response in ATA-positive patients, while spontaneous ACA-IgA secretion was undetectable in the ACA-positive patient group. FACS experiments showed that spontaneously ATA-IgA secreting B cells were primarily present in the plasmablast compartment.

Conclusion:

Our findings demonstrate that ATA-positive SSc patients harbour an activated ATA-IgG and ATA-IgA B cell response, as indicated by the spontaneous secretion of both ATA isotypes by circulating plasmablasts. Remarkably, the ACA B cell response was far less active and lacked the active IgA component which suggests a difference in the triggers driving these autoreactive B cell responses in patients. Moreover, the remarkable ATA-IgA secretion points towards a potential mucosal origin of the ATA response.

Disclosure of Interests:

Corrie Wortel: None declared, Nina van Leeuwen: None declared, Maaike Boonstra: None declared, Rene Toes: None declared, Thomas Huizinga Grant/research support from: Ablynx, Bristol-Myers Squibb, Roche, Sanofi, Consultant of: Ablynx, Bristol-Myers Squibb, Roche, Sanofi, Jeska de Vries-Bouwstra: None declared, Hans Ulrich Scherer Grant/research support from: Bristol Myers Squibb, Sanofi, Pfizer, Speakers bureau: Pfizer, Lilly, Roche, Abbvie

OP0219 MORTALITY OF RHEUMATOID ARTHRITIS PATIENTS, TREATED TO TARGET AT LOW DISEASE ACTIVITY: 17-YEARS FOLLOW-UP OF THE BEST COHORT

Background:

Rheumatoid arthritis is known to be associated with increased mortality over the years when compared to the general population. In the BeSt study, 508 patients were treated to target (Disease Activity Score ≤2.4) for 10 years between April 2000 and August 2012. At the end of the initial study follow-up, the observed mortality in the BeSt cohort was similar to mortality in the general population. In the current study we evaluated the mortality in the BeSt cohort after 17 years follow-up and compared it to the general Dutch population.

Objectives:

Evaluate long-term mortality in the BeSt study cohort.

Methods:

In the BeSt study 508 patients diagnosed with early RA were randomized to four initial treatment strategies: 1. Sequential monotherapy; 2. Step-up combination therapy; 3. Initial combination therapy with prednisone; or 4. Initial combination therapy with infliximab. During the 10-year follow-up period treatment was steered at low disease activity (DAS ≤2.4) and adjusted every three months when necessary. After 10-years patients were treated and followed-up according to regular care. We explored mortality through the Dutch state registry for mortality (Centrum voor Familiegeschiedenis) and treating rheumatologist. Mortality in the BeSt cohort was compared to the general Dutch population (Statistics Netherlands) matched by gender, age and calendar year using the standardized mortality ratio (SMR). Kaplan-Meier curves and the log-rank test were used to compare survival among the initial treatment strategies.

Results:

The mean duration of follow-up in non-deceased patients was 17 years (range 16-18). In total, 143 patients died (28%) compared to a total of 105 (21%) expected deaths in the reference population. The overall SMR after 17 years was 1.37 (95% CI: 1.16-1.61). Within the study population, no statistically significant difference in survival-curves was observed between the four initial treatment strategies (log-rank p=0.76) (table 1, and figure 1).

Table 1.

BeSt study cohort mortality - stratified for initial treatment strategy

Sequential monotherapyn=126Step-up combination therapyn=121Initial combination therapy with prednisonen=133Initial combination therapy with infliximabn=128N (%) †38 (30)31 (26)41 (31)33 (26)SMR (95% CI)1.41 (1.03–1.94)1.20 (0.84-1.70)1.53 (1.13-2.09)1.31 (0.93-1.85)

SMR: standardized mortality ratio (number observed deaths/number expected deaths); CI: confidence interval.Conclusion:

Figure 1.

Survival curves – stratified for initial treatment strategy

After a mean of 17 years follow-up the mortality was increased in the BeSt study cohort when compared to the general Dutch population. We observed no difference in survival curves among the four treatment strategies.

Disclosure of Interests:

Johanna M. Maassen: None declared, Yvonne Goekoop-Ruiterman: None declared, Hans van Groenendael: None declared, WIllem Lems Grant/research support from: Pfizer, Consultant of: Lilly, Pfizer, Pit Kerstens: None declared, Thomas Huizinga Grant/research support from: Ablynx, Bristol-Myers Squibb, Roche, Sanofi, Consultant of: Ablynx, Bristol-Myers Squibb, Roche, Sanofi, Cornelia Allaart: None declared

FRI0090 MAINTENANCE OF CLINICAL RESPONSE WITH ABATACEPT IN COMBINATION WITH MTX IN INDIVIDUAL PATIENTS WITH EARLY RA WHO ARE MTX-NAÏVE AND ANTI-CITRULLINATED PROTEIN ANTIBODY (ACPA)+: RESULTS FROM THE INDUCTION PERIOD OF AVERT-2, A RANDOMISED PHASE IIIB STUDY

Background:

In the 56-wk induction period (IP) of the Phase IIIbAssessingVeryEarlyRATreatment (AVERT)-2 trial (NCT02504268), more patients (pts) achieved SDAI remission (≤3.3) with abatacept (ABA) + MTX vs ABA placebo (PBO) + MTX at IP Wk 52.1It is unknown whether each individual pt within a treatment (Tx) group achieves and sustains the same efficacy endpoints at all time points during the IP.

Objectives:

To investigate whether ABA effectiveness is sustained by individual pts who achieved SDAI remission (≤3.3), SDAI low disease activity (LDA; >3.3–11), DAS28 (CRP) <2.6, ACR50/70 response or Boolean remission at IP Wk 24 (AVERT-2 primary endpoint) and both Wks 40 and 52 (Wks 40/52).

Methods:

Pts were randomised 3:2 to blinded SC ABA (125 mg/wk) + MTX or ABA PBO + MTX induction Tx for 56 wks. Key inclusion criteria: age ≥18 yrs; RA diagnosis (ACR/EULAR 2010 criteria); RA duration ≤6 mos; SDAI >11; ACPA+; CRP >3 mg/L or ESR ≥28 mm/h; TJC ≥3 and SJC ≥3; DMARD naïve. Response rates were investigated by Tx arm in the cohort 1 analysis population (all randomised pts treated in the IP [intent-to-treat analysis]).

Results:

Of randomised cohort 1, 752 pts were treated during the IP: 451 with ABA + MTX and 301 with ABA PBO + MTX. Baseline characteristics were similar across Tx arms.1Stringent SDAI remission endpoint at IP Wk 24 was achieved by 22% of ABA + MTX-treated pts; of these, 56% sustained SDAI remission at IP Wks 40/52 (Table). A similar proportion of ABA + MTX-treated pts achieved (17%) and sustained (58%) Boolean remission at IP Wks 24 and 40/52. At IP Wk 24, 42% of ABA + MTX-treated pts achieved DAS28 (CRP) <2.6 and 74% sustained DAS28 (CRP) <2.6 to IP Wks 40/52; a high proportion of patients sustained ACR50/70 responses at IP Wks 40/52 (83% and 79%, respectively). A lower proportion of pts sustained SDAI LDA to IP Wks 40/52 vs other endpoints (Table). Most efficacy endpoints were achieved by fewer pts who received ABA PBO + MTX than ABA + MTX (Table); among responders in this Tx group, fewer sustained remission at Wks 40/52, which correlates with a higher proportion of pts sustaining SDAI LDA at Wks 40/52 with ABA PBO + MTX than ABA + MTX.

Conclusion:

The majority of individual pts with RA who achieved clinically stringent endpoints such as SDAI remission, DAS28 (CRP) <2.6 or Boolean remission, as well as clinically meaningful endpoints such as ACR50/70 at IP Wk 24 with weekly SC abatacept, sustained their responses to Wks 40/52. The high proportion of patients achieving early stringent remission and response to SC abatacept by individual pts may be indicative of sustained efficacy over time.

References:

[1]Emery P, et al. ACR 2018; San Diego, USA: Poster 563.

Table .

Proportion of Pts With Response at IP Wk 24 Who Also Achieved Remission at Wks 40/52

EndpointResponders at IP Wk 24, n (%)Responders at IP Wk 24 and Wks 40/52, n/N (%)ABA + MTX(n=451)ABA PBO + MTX(n=301)ABA + MTX*ABA PBO + MTX*SDAI remission (≤3.3)100 (22)40 (13)56/100 (56)17/40 (43)SDAI low disease activity (>3.3–11)167 (37)82 (27)46/167 (28)32/82 (39)DAS28 (CRP) <2.6188 (42)78 (26)139/188 (74)43/78 (55)ACR50 response†260 (58)125 (42)215/260 (83)92/125 (74)ACR70 response†156 (35)66 (22)123/156 (79)42/66 (64)Boolean remission76 (17)29 (10)44/76 (58)8/29 (28)

*% based on number of pts within each Tx group who achieved response at IP Wk 24 (denominator);†Response at IP Wks 24 and 52

Acknowledgments:

Lola Parfitt (medical writing, Caudex; funding: Bristol-Myers Squibb)

Disclosure of Interests:

Paul Emery Grant/research support from: AbbVie, Bristol-Myers Squibb, Merck Sharp & Dohme, Pfizer, Roche (all paid to employer), Consultant of: AbbVie (consultant, clinical trials, advisor), Bristol-Myers Squibb (consultant, clinical trials, advisor), Lilly (clinical trials, advisor), Merck Sharp & Dohme (consultant, clinical trials, advisor), Novartis (consultant, clinical trials, advisor), Pfizer (consultant, clinical trials, advisor), Roche (consultant, clinical trials, advisor), Samsung (clinical trials, advisor), Sandoz (clinical trials, advisor), UCB (consultant, clinical trials, advisor), Yoshiya Tanaka Grant/research support from: Asahi-kasei, Astellas, Mitsubishi-Tanabe, Chugai, Takeda, Sanofi, Bristol-Myers, UCB, Daiichi-Sankyo, Eisai, Pfizer, and Ono, Consultant of: Abbvie, Astellas, Bristol-Myers Squibb, Eli Lilly, Pfizer, Speakers bureau: Daiichi-Sankyo, Astellas, Chugai, Eli Lilly, Pfizer, AbbVie, YL Biologics, Bristol-Myers, Takeda, Mitsubishi-Tanabe, Novartis, Eisai, Janssen, Sanofi, UCB, and Teijin, Vivian Bykerk: None declared, Clifton Bingham Grant/research support from: Bristol-Myers Squibb, Consultant of: Bristol-Myers Squibb, Thomas Huizinga Grant/research support from: Ablynx, Bristol-Myers Squibb, Roche, Sanofi, Consultant of: Ablynx, Bristol-Myers Squibb, Roche, Sanofi, Gustavo Citera Grant/research support from: AbbVie, Amgen, Eli Lilly, Gema, Genzyme, Novartis and Pfizer Inc, Consultant of: AbbVie, Amgen, Eli Lilly, Gema, Genzyme, Novartis and Pfizer Inc, Kuan-Hsiang Gary Huang Shareholder of: Bristol-Myers Squibb, Employee of: Bristol-Myers Squibb, Sean Connolly Shareholder of: Bristol-Myers Squibb, Employee of: Bristol-Myers Squibb, Yedid Elbez Consultant of: Bristol-Myers Squibb, Robert Wong Shareholder of: Bristol-Myers Squibb, Employee of: Bristol-Myers Squibb, Karissa Lozenski Employee of: Bristol-Myers Squibb, Roy Fleischmann Grant/research support from: AbbVie, Akros, Amgen, AstraZeneca, Bristol-Myers Squibb, Boehringer, IngelhCentrexion, Eli Lilly, EMD Serono, Genentech, Gilead, Janssen, Merck, Nektar, Novartis, Pfizer, Regeneron Pharmaceuticals, Inc., Roche, Samsung, Sandoz, Sanofi Genzyme, Selecta, Taiho, UCB, Consultant of: AbbVie, ACEA, Amgen, Bristol-Myers Squibb, Eli Lilly, Gilead, GlaxoSmithKline, Novartis, Pfizer, Sanofi Genzyme, UCB

AB0430 MORTALITY IN PATIENTS WITH SYSTEMIC LUPUS ERYTHEMATOSUS AND NEUROPSYCHIATRIC SYMPTOMS

Background:

Little is known about mortality in patients with systemic lupus erythematosus (SLE) presenting with neuropsychiatric (NP) symptoms.

Objectives:

We aimed to evaluate all-cause and cause-specific mortality in patients with SLE and NP symptoms.

Methods:

All patients with the clinical diagnosis of SLE of 18 years and older that visited the tertiary referral NPSLE clinic of the Leiden University Medical Center between 2007-2018 and signed informed consent were included in this study. Patients were classified as NPSLE if NP symptoms were attributed to SLE and immunosuppressive or anticoagulant therapy was initiated, otherwise patients were classified as non-NPSLE. Municipal registries were checked for current status (alive/deceased). Electronical medical files were studied for clinical characteristics and cause of death. Standardized mortality ratios (SMRs) and 95% confidence intervals were calculated using data from the general Dutch population. In addition, a rate ratio (RR) was calculated using direct standardization to compare mortality in NPSLE with non-NPSLE patients.

Results:

351 patients with the clinical diagnosis of SLE were included, of which 149 patients were classified as NPSLE (42.5%). Compared with the general population, mortality was increased five times in NPSLE (SMR 5.0, 95% CI: 2.6-8.5) and nearly four times in non-NPSLE patients (SMR 3.7, 95% CI: 2.2-6.0), as shown in Table 1. Risk of death due to cardiovascular disease (CVD) was increased in non-NPSLE patients (SMR 6.2, 95% CI: 2.0-14.6) and an increased risk of death to infections was present in both NPSLE and non-NPSLE patients ((SMR 29.9, 95% CI: 3.5 – 105) and SMR 91.3 (95% CI: 18.8 – 266) respectively). However, mortality did not differ between NPSLE and non-NPSLE patients (RR 1.0, 95% CI: 0.5 – 2.0).

Table 1.

All-cause mortality in SLE patients presenting with neuropsychiatric symptoms attributed to SLE (NPSLE) or to other causes (non-NPSLE)

NPSLE(N = 149)Non-NPSLE(N = 202)Deaths (N, %)13 (8.7)17 (8.4)Age at death (median, range)49 (32 – 79)59 (20 – 89)Follow-up time (years)9061047Crude mortality rate (per 1000 PY)14.316.2All-cause mortality*Female5.5 (2.8 – 9.6)3.4 (1.9 – 5.7)Male2.3 (0.1 - 12.8)6.2 (1.3 – 18.2)Combined5.0 (2.6 – 8.5)3.7 (2.2 – 6.0)

*Standardized mortality ratio, ratio of the observed and expected number of deaths

Conclusion:

Mortality was increased in both NPSLE and non-NPSLE patients in comparison with the general population, but there was no difference in mortality between NPSLE and non-NPSLE patients. Risk of death due to infections was increased in both groups.

Disclosure of Interests:

Rory Monahan: None declared, Rolf Fronczek: None declared, Jeroen Eikenboom: None declared, Huub Middelkoop: None declared, L.J.J. Beaart- van de Voorde: None declared, Gisela Terwindt: None declared, Nic van der Wee: None declared, Frits Rosendaal: None declared, Thomas Huizinga Grant/research support from: Ablynx, Bristol-Myers Squibb, Roche, Sanofi, Consultant of: Ablynx, Bristol-Myers Squibb, Roche, Sanofi, Margreet Kloppenburg: None declared, G.M. Steup-Beekman: None declared

FRI0098 ASSOCIATION BETWEEN THE SEROLOGIC STATUS OF ISOTYPE-SPECIFIC AUTOANTIBODIES AND THERAPEUTIC EFFICACY IN RHEUMATOID ARTHRITIS PATIENTS TREATED WITH ABATACEPT: A PROSPECTIVE ULTRASOUND COHORT STUDY IN JAPAN

Background:

The presence of anti-cyclic citrullinated protein antibodies (ACPA) and anti-carbamylated protein (anti-CarP) antibody is specific for rheumatoid arthritis (RA). Recently, it was reported that the serological status of ACPA is associated with the therapeutic response of the T-cell co-stimulation blocker abatacept (1, 2). However, it is currently unclear whether the serological status of each isotype levels of these autoantibodies before treatment introduction or the changes during treatment are associated with the therapeutic response of abatacept.

Objectives:

To evaluate longitudinal changes in the isotypes of ACPA and anti-CarP in RA patients treated with abatacept, and associations between the baseline serological status/ these changes and clinical response/ ultrasonographic response.

Methods:

This study is part of an ongoing non-randomized multicenter prospective cohort study of patients with active RA who received biological or targeted DMARD therapy at 13 participating rheumatology centers from the Kyushu region of Japan since June 2013 (3). As of the present report, we enrolled 43 consecutive Japanese patients with active RA who have introduced treatment with abatacept and had finished the first 12-month observation period. We evaluated disease activity by clinical composite measure and ultrasound score at baseline, 3, 6, 9 and 12 months. In ultrasound of bilateral hands from 22 sites, the findings obtained by gray-scale (GS) and power Doppler (PD) assessments were graded on a semi-quantitative scale from 0 to 3 and the sum of GS or PD scores was used as the total GS or PD score. The serum levels of IgG/IgM/IgA-type of ACPA and anti-CarP were measured by the ELISA method in Leiden University Medical Center. We evaluated the association between serologic status of autoantibodies and clinical /ultrasonographic therapeutic efficacy.

Results:

The median age was 72 years, and the disease duration was 54 months. Methotrexate was concomitant in 22 (51%). Sixteen (37%) patients had a history of previous use of biological DMARDs. Nineteen (44%) and 23 (54%) patients achieved SDAI remission and PD remission (total PD score =0) at 12 months, respectively. The serum levels of all isotypes of ACPA/anti-CarP significantly decreased at 12 months from baseline. The reduction of IgM-ACPA level significantly correlated with the reduction of SDAI (rs=0.33, p=0.031) and total PD score (rs=0.49, p=0.0007). Both clinical and ultrasonographic therapeutic responses were better in patients with the detectable IgM-ACPA at baseline than in patients without that (Figure): the reduction of SDAI (p=0.0078) and that of total PD score (p=0.0079) were significantly larger in the former than in the latter. All isotype of anti-CarP did not associate with therapeutic response.

Conclusion:

Treatment of abatacept induced to the reduction of the autoantibody levels. The IgM-ACPA level at baseline and the change in IgM-ACPA associated with both clinical and ultrasonographic therapeutic response in patients treated with abatacept. IgM-ACPA, compared with usual IgG-ACPA, better reflects the treatment response of abatcept in patients with RA

References:

[1]Ann Rheum Dis. 2016;75:709, 2) RMD Open. 2018;4:e000564, 3)Arthritis Care Res (Hoboken). 2018;70:1719.

Acknowledgments:

We have acknowledged for all the members of Kyushu multicenter rheumatoid arthritis ultrasound prospective observational cohort study group.

Disclosure of Interests:

Shin-ya Kawashiri Grant/research support from: This work was supported by Bristol-Myers Squibb and Ono Pharmaceutical. co., Yushiro Endo: None declared, Ayako Nishino: None declared, Toshimasa Shimizu: None declared, Yukitaka Ueki: None declared, Nobutaka Eiraku: None declared, Akitomo Okada: None declared, Naoki Matsuoka: None declared, Tamami Yoshitama: None declared, Hideki Nakamura: None declared, Mami Tamai: None declared, Tomoki Origuchi: None declared, Rene Toes: None declared, Thomas Huizinga Grant/research support from: Ablynx, Bristol-Myers Squibb, Roche, Sanofi, Consultant of: Ablynx, Bristol-Myers Squibb, Roche, Sanofi, Atsushi Kawakami: None declared