OP0267 A GENOME-WIDE INVESTIGATION OF PERSISTENCE TO TREATMENT WITH METHOTREXATE IN SWEDISH EARLY RHEUMATOID ARTHRITIS PATIENTS

Background

Despite being the anchor drug for treating rheumatoid arthritis (RA), methotrexate (MTX) provides a good response only in some of the treated patients [1]. If MTX treatment outcome has a substantial genetic component, genetic variants could provide useful predictors for identification of patients likely to respond and remain on treatment. So far, studies have focused mainly on primary response, and attempts to explain the inter-patient variability through genetic variants have been inconclusive or underpowered [2,3].

Objectives

We aimed to investigate whether there are genetic variants associated with persistence to treatment with MTX (at one and three years) in early RA, and to estimate any underlying heritability.

Methods

We conducted a genome-wide association study (GWAS) on persistence to treatment with MTX in DMARD-monotherapy. We included participants from the Epidemiological Investigation of RA (EIRA) study and the Swedish Rheumatology Quality Register’s biobank, diagnosed with early RA and treated with MTX as their first ever DMARD. Persistence to MTX was defined as remaining on treatment at one and three years, respectively, with no additional DMARDs added during that period. Estimation of SNP-based heritability was done using restricted maximum likelihood. We performed the analyses for all RA, and two disease subsets: those positive for either ACPA or rheumatoid factor, and those negative for both.

Results

After quality control, 3403 of an initial 3609 early RA patients of European ancestry and above five million SNPs remained for the primary analysis. Among these, 65% were persistent at one year, and 44% were persistent at three years. In secondary analysis we excluded 218 patients due to missing sero-status. Of the remaining, 72% were seropositive. No SNP reached genome-wide significance, neither for persistence at one nor at three years. The SNP-based heritability was estimated to 0.35 (95%CI 0.04-0.65) for persistence at one year and 0.09 (95%CI 0.00-0.37) for persistence at three years. Analyses stratified by sero-status provided results comparable to the main analysis with similar proportions of persistent patients and no SNP reaching genome-wide significance for any of the subgroups. Point estimates of heritability for persistence in the seropositive group at one year (h2 = 0.08, 95%CI 0.00-0.51) was lower than for seropositive persistence at three years (h2 = 0.19, 95%CI 0.00-0.62) while point estimates for the seronegative heritability were zero for both persistence outcomes.

Conclusion

Despite being the largest GWAS on an MTX treatment outcome to date, no genome-wide significant associations were detected. The low heritability observed along with the lack of strong independent associations, may indicate that genetic influence from common genetic variants on persistence to MTX is minor, and of a polygenic nature.

References

[1]Saevarsdottir, S., et al., Predictors of response to methotrexate in early DMARD naive rheumatoid arthritis: results from the initial open-label phase of the SWEFOT trial. Ann Rheum Dis, 2011. 70(3): p. 469-75.

[2]Ling, S., J. Bluett, and A. Barton, Prediction of response to methotrexate in rheumatoid arthritis. Expert Rev Clin Immunol, 2018. 14(5): p. 419-429.

[3]Szostak, B., et al., Using pharmacogenetics to predict methotrexate response in rheumatoid arthritis patients. Expert Opin Drug Metab Toxicol, 2020. 16(7): p. 617-626.

Acknowledgements

The authors would like to thank the participants of the EIRA study and the SRQ biobank as well as deCODE genetics for making this study possible.

Disclosure of Interests

Anton Öberg Sysojev: None declared, Saedis Saevarsdottir Employee of: Part-time employee of deCODE genetics Inc., Lina M. Diaz-Gallo: None declared, Lars Alfredsson: None declared, Lars Klareskog: None declared, - SRQ Biobank Group: None declared, Thomas Frisell: None declared, Leonid Padyukov: None declared, Johan Askling Grant/research support from: Karolinska Institutet has entered into agreements with the following companies, with JA as PI: Abbvie, BMS, Eli Lilly, Galapagos, Janssen, Pfizer, Roche, Samsung Bioepis and Sanofi., Helga Westerlind: None declared

POS0747 ANTI-PHOSPHATIDYLSERINE/PROTHROMBIN ANTIBODIES AND VASCULAR EVENTS ASSOCIATE POSITIVELY WITH HLA-DRB1*13 AND NEGATIVELY WITH HLA-DRB1*03 IN SLE

Background

Anti-phosphatidylserine/prothrombin antibodies (anti-PS/PT) associate with thrombotic events (1). HLA-DRB1 alleles contribute to the occurrence of conventional antiphospholipid antibodies (aPL), including anti-beta2glycoprotein-I (beta2GPI) and anti-cardiolipin (CL) (2).

Objectives

We investigated associations between anti-PS/PT and HLA-DRB1 alleles and thrombosis in patients with SLE. Conventional aPL were included for comparison.

Methods

We included 341 consecutive Swedish SLE patients, with information on general cardiovascular risk factors, including blood lipids, lupus anticoagulant (LAC) and thrombotic events. Anti-PS/PT, anti-beta2GPI and anti-CL of IgA/G/M isotypes were quantified in parallel using particle-based multi-analyte technology. The 99th percentiles among 162 age- and sex-matched populations controls were used as cutoffs. HLA-DRB1 typing was performed using sequence-specific primer PCR.

Results

Anti-PS/PT antibodies associated positively with HLA-DRB1*13 (odds ratio [OR] 2.7, P=0.002), whereas anti-beta2GPI and anti-CL antibodies associated primarily with HLA-DRB1*04 (OR 2.5, P=0.0005; Table 1). These associations remained after adjustment for other significant HLA-DRB1 alleles identified in Table 1 (Figure 1a and b) also for LAC (Figure 1c), and also after adjustment for age and gender (not shown). HLA-DRB1*13, but not DRB1*04, remained as an independent risk factor for thrombosis after adjustment for significant HLA alleles (Figure 1d), and also after adjustment for cardiovascular risk factors in stepwise regression (not shown). Mediation analysis showed that 31.3% of the HLA-DRB1*13-related risk for thrombosis was mediated by anti-PS/PT positivity. HLA-DRB1*03, on the other hand, associated negatively with thrombotic events (Figure 1d) as well as with all aPL (Figure 1a-c). HLA-DRB1*03 had thrombo-protective effect in aPL positive patients (Figure 1d). Additionally, HLA-DRB1*03 positivity was associated with a favourable lipid profile regarding high-density lipoprotein (median 1.4 vs. 1.2 mmol/L, p=0.02) and triglycerides (median 0.9 vs 1.1 mmol/L, p=0.04); whereas no other HLA-DRB1 alleles showed any associations to lipid levels.

Table 1.

Frequency of individual HLA DRB1 and associations with antibody phenotypes. Odds ratios (OR) and confidence intervals (CI) for being antibody positive given a specific HLA allele and corresponding p values were calculated using Chi2 tests, with significant associations underlined.

HLA DRB1HLA-DRB1 n (%) total patientsAnti-PS/PT positive (any isotype) n=48OR (95%CI); PAnti-β2GPI or anti-CL positive (any isotype) n=96OR (95%CI); P*0141 (12.9%)4 (8.3%)0.6 (0.2-1.7); 0.311 (11.4%)0.8 (0.4-1.7); 0.6*03147 (46.5%)13 (27.1%)0.4 (0.2-0.7); 0.00433 (34.4%)0.5 (0.3-0.8); 0.006*0494 (29.7%)18 (37.5%)1.6 (0.8-2.9); 0.241 (42.7%)2.5 (1.5-4.1); 0.0005*0728 (8.9%)6 (12.5%)1.5 (0.6-4); 0.49 (9.4%)1 (0.4-2.4); 0.9*0828 (8.9%)6 (12.5%)1.6 (0.6-4.3); 0.39 (9.4%)1.2 (0.5-2.7); 0.7*099 (2.8%)1 (2.1%)0.7 (0.1-5.5); 0.72 (2.1%)0.6 (0.1-3.0); 0.5*107 (2.2%)0 (0)NA2 (2.1%)0.9 (0.2-4.6); 0.9*1127 (8.5%)6 (12.5%)1.6 (0.6-4.3); 0.38 (8.3%)0.9 (0.4-2.2); 0.8*127 (2.2%)0 (0)NA1 (1%)0.4 (0.05-3.7); 0.4*1379 (25%)21 (43.7%)2.7 (1.4-5.2); 0.00233 (34.3%)2 (1.2-3.4%); 0.01*146 (1.9%)2 (4.2%)3.7 (0.6-23); 0.12 (2.1%)1.4 (0.2-8.9); 0.8*15118 (37.3%)12 (48%)0.5 (0.2-0.9); 0.04527 (28.1%)0.5 (0.3-0.9); 0.01*168 (2.5%)1 (2.1%)0.8 (0.09-6.4); 0.84 (4.2%)2.2 (0.5-9.2); 0.2

Conclusion

HLA-DRB1*13 confers risk for both anti-PS/PT and thrombotic events in SLE. The association between HLA-DRB1*13 and thrombosis is largely, but not entirely, mediated through anti-PS/PT. Due to the negative association of HLA-DRB1*03 with aPL and the positive association with favourable lipid levels, HLA-DRB1*03 seems to identify a subgroup of SLE patients with reduced vascular risk.

References

[1]Elbagir S et al. Lupus 2021;30(8):1289.

[2]Lundström E et al. Ann Rheum Dis 2013;72:1018.

Disclosure of Interests

Sahwa Elbagir: None declared, Lina M. Diaz-Gallo: None declared, Giorgia Grosso: None declared, Agneta Zickert: None declared, Iva Gunnarsson: None declared, Michael Mahler Employee of: Dr Mahler is employee of Werfen., Elisabet Svenungsson Speakers bureau: Dr Svennungson has obtained speaker’s fees from Janssen., Grant/research support from: Dr Svennungson has obtained research grant from Merck., Johan Rönnelid Speakers bureau: Dr Rönnelid has given paid lectures for Thermo Fisher Scientific., Consultant of: Dr Rönnelid has been a member of the Scientific Advisory Board for Thermo Fisher Scientific.

OP0160 HLA-DRB1 ASSOCIATIONS WITH AUTOANTIBODY-DEFINED SUBGROUPS IN IDIOPATHIC INFLAMMATORY MYOPATHIES (IIM)

Background

There is a gap between how IIM patients are classified in practice and current validated classification criteria1. Also, different associations with genetic variations in HLA can inform about different T-cell mechanisms involved in disease pathogenesis.

Objectives

We aimed to systematically study associations between HLA-DRB1 alleles, clinical manifestations, and autoantibody-defined IIM subgroups.

Methods

We included 1348 IIM patients from five European countries. An unsupervised cluster analysis was performed using 14 autoantibodies: anti-Jo1, -PL7, -PL12, -EJ, -OJ, -SRP, -U1RNP, -Ro52, -Mi2, -TIF1γ, -MDA5, -PMScl, -SAE1, and -NXP2 to identify patients’ subgroups. Logistic regressions were used to estimate the associations between HLA-DRB1 alleles, clinical manifestations and the identified subgroups.

Results

Eight subgroups were defined by the autoantibody status (Table 1). Three of the subgroups (1, 2 and 6) have overlapping autoantibodies, while four are almost monospecific (3,4,5 and 7), and one (8) has patients negative for tested autoantibodies. Figure 1 represents the significant associations between HLA-DRB1 alleles and the eight subgroups. Heliotrope rash and Gottron’s sign were significantly more frequent in subgroups 3 (OR:2.2 95%CI:[1.1-4.8], OR:2.6 95%CI:[1.3-5.9], respectively), 4 (OR:12 95%CI:[3.6-75], OR:7.8 95%CI:[2.8-33], respectively) and 7 (OR:22 95%CI:[4.5-385], OR:10 95%CI:[3.1-65], respectively), and Raynaud’s phenomenon was significantly more frequent in subgroup 6 (OR:3.3 95%CI:[1.2-11]).

Table 1.

Autoantibody-defined subgroups using an unsupervised cluster analysis.

Subgroups/ MedoidsVariables1 Ro522 U1RNP3 PMScl4 Mi25 Jo16 Jo1/Ro527 TIF18 None*Alln (%)137 (10)183 (14)107 (8)65 (5)119 (9)140 (10)78 (6)519 (39)1348 (100)Female (%)93 (68)116 (63)79 (74)45 (69)76 (64)96 (69)64 (82)313 (60)882 (65)Age at diagnosis, median (IQR)56 (16)51.5 (23)51 (25)57 (22.5)47.5 (23.25)52 (19.5)53.5 (21.75)58 (22)55 (23)AutoantibodiesAnti-Jo106 (3)01 (2)119 (100)140 (100)00266 (20)Anti-PL77 (5)13 (7)00000020 (1.5)Anti-PL125 (4)3 (2)1 (1)01 (1)00010 (0.7)Anti-EJ2 (2)00000002 (0.1)Anti-OJ07 (4)0000007 (0.5)Anti-TIF110 (7)2 (1)2 (2)00078 (100)092 (7)Anti-Mi21 (1)1 (1)1 (1)65 (100)02 (1)0070 (5)Anti-SAE18 (6)23 (13)00000031 (2)Anti-NXP21 (1)23 (13)1 (1)0000025 (2)Anti-MDA59 (7)10 (6)1 (1)1 (2)01 (1)0022 (2)Anti-SRP8 (6)32 (18)00000040 (3)Anti-Ro52137 (100)16 (9)000140 (100)00293 (22)Anti-PMScl11 (8)1 (1)107 (100)00000119 (9)Anti-U1RNP079 (43)0003 (2)0082 (6)

*IIM patients negative for the tested autoantibodies.

Figure 1.

Forest plot of significant associations of HLA. *DRB1 alleles with autoantibody-defined subgroups. Scandinavia includes patients from Denmark, Norway, and Sweden.

Conclusion

Our study reveals that certain subgroups of IIM patients are characterized by overlap of myositis -specific and -associated autoantibodies, which in turn are associated with different HLA-DRB1 alleles including potential novel associations. These results point to different disease mechanisms in the subgroups, as well as suggest that IIM classification could be improved by integrating broader serological and genetic data.

References

[1]Parker MJS, Oldroyd A, Roberts ME, et al. The performance of the European League Against Rheumatism/American College of Rheumatology idiopathic inflammatory myopathies classification criteria in an expert-defined 10 year incident cohort. Rheumatology (Oxford). 2019;58(3):468-475.

Acknowledgements

We thank all the patients who participated in the study.

Disclosure of Interests

Valerie Leclair: None declared, Angeles Shunashy Galindo-Feria: None declared, Simon Rothwell: None declared, Olga Kryštůfková: None declared, Heřman Mann: None declared, Louise Pyndt Diederichsen: None declared, helena andersson: None declared, Martin Klein: None declared, Sarah Tansley: None declared, Neil McHugh: None declared, Janine Lamb: None declared, Jiří Vencovský Speakers bureau: Abbvie, Biogen, Boehringer, Eli Lilly, Gilead, MSD, Novartis, Pfizer, Roche, Sanofi, UCB, Werfen, Consultant of: Abbvie, Argenx, Boehringer, Eli Lilly, Gilead, Octapharma, Pfizer, UCB, Grant/research support from: Abbvie, Hector Chinoy: None declared, Marie Holmqvist: None declared, Leonid Padyukov: None declared, Ingrid E. Lundberg Shareholder of: Roche and Novartis, Consultant of: Corbus Pharmaceuticals Inc, Astra Zeneca, Bristol Myer´s Squibb, Corbus Pharmaceutical, EMD Serono Research & Development Institute, Argenx, Octapharma, Kezaar, Orphazyme, and Janssen, Grant/research support from: Astra Zeneca, Lina M. Diaz-Gallo: None declared

Association of Toll-like receptor 10 and susceptibility to Crohn's disease independent of NOD2

Impaired innate inflammatory response has a key role in the Crohn's disease (CD) pathogenesis. The aim of this study was to investigate the possible role of the TLR10-TLR1-TLR6 gene cluster in CD susceptibility. A total of 508 CD patients (284, cohort 1 and 224, cohort 2) and 576 controls were included. TLR10-TLR1-TLR6 cluster single-nucleotide polymorphisms genotyping, NOD2 mutations and TLR10 mRNA quantification were performed using TaqMan assays. Nucleotide-binding oligomerization domain containing 2 (NOD2) and Toll-like receptor (TLR) loci interaction was analyzed by logistic regression and multifactor-dimensionality reduction (MDR). Entropy-based analysis was used to interpret combination effects. One TLR10 haplotype (TLR10(GGGG)) was found associated with CD susceptibility in both cohorts, individuals with two copies had approximately twofold more risk of CD susceptibility than individuals having no copies (odds ratio=1.89, P-value=0.0002). No differences in the mRNA levels were observed among the genotypes. The strongest model for predicting CD risk according to the MDR analysis was a two-locus model including NOD2 mutations and TLR10(GGGG) haplotype (P(c)<0.0001). The interaction gain attributed to the combination of both genes was negative (IG=-2.36%), indicating redundancy or independent effects. Our results support association of the TLR10 gene with CD susceptibility. The effect of TLR10 would be independent of NOD2, suggesting different signaling pathways for both genes.

Analysis of the influence of PTPN22 gene polymorphisms in systemic sclerosis

OBJECTIVE

Two functional single nucleotide polymorphisms (SNP) in the PTPN22 gene (rs24746601 and rs33996649) have been associated with autoimmunity. The aim of this study was to investigate the role of the R263Q SNP for the first time and to re-evaluate the role of the R620W SNP in the genetic predisposition to systemic sclerosis (SSc) susceptibility and clinical phenotypes.

METHODS

3422 SSc patients (2020 with limited cutaneous SSc and 1208 with diffuse cutaneous SSc) and 3638 healthy controls of Caucasian ancestry from an initial case--control set of Spain and seven additional independent replication cohorts were included in our study. Both rs33996649 and rs2476601 PTPN22 polymorphisms were genotyped by TaqMan allelic discrimination assay. A meta-analysis was performed to test the overall effect of these PTPN22 polymorphisms in SSc.

RESULTS

The meta-analysis revealed evidence of association of the rs2476601 T allele with SSc susceptibility (p(FDRcorrected)=0.03 pooled, OR 1.15, 95% CI 1.03 to 1.28). In addition, the rs2476601 T allele was significantly associated with anticentromere-positive status (p(FDRcorrected)=0.02 pooled, OR 1.22, 95% CI 1.05 to 1.42). Although the rs33996649 A allele was significantly associated with SSc in the Spanish population (p(FDRcorrected)=0.04, OR 0.58, 95% CI 0.36 to 0.92), this association was not confirmed in the meta-analysis (p=0.36 pooled, OR 0.89, 95% CI 0.72 to 1.1).

CONCLUSION

The study suggests that the PTPN22 R620W polymorphism influences SSc genetic susceptibility but the novel R263Q genetic variant does not. These data strengthen evidence that the R620W mutation is a common risk factor in autoimmune diseases.