Background The use of anti-rheumatic drugs in pregnancy is often complicated by concerns over their potential for adverse effects. Given that rheumatic diseases often affect women of childbearing age and may flare during pregnancy, the safety of anti-rheumatic drugs and immunosuppressant's are of particular importance. Practice has relied on information based mainly on experimental and animal studies. Previous systematic reviews have identified risks with various anti-rheumatic drugs and biologics. This systematic review is an update of consensus papers on anti-rheumatic drugs, biological agents and reproduction published in 2006/8. Objectives This systematic review is an update of consensus papers on anti-rheumatic drugs, biological agents and reproduction published in 2006/8. Methods A systematic search of PubMed and Embase was carried out using relevant keywords for pregnancy, lactation, rheumatoid arthritis (RA), psoriatic arthritis (PsA), inflammatory arthritis (IA), juvenile idiopathic arthritis (JIA) from 2006 onwards and drugs commonly prescribed in patients with rheumatic disease from 1966 onwards. Review articles and non-English language papers were excluded. Results The search identified 43 papers published since the 2006/8 consensus papers describing relevant drug use in pregnant women with RA, PsA, IA and JIA. These consisted of 15 case reports, 7 case series, 16 cohort studies and 5 case-control studies. 4145 pregnancies were reported with an inflammatory arthritis. Over 4530 drug exposures were identified which includes steroids (346) DMARDs (AZA 210, HCQ 189, LFL 80, SSZ 57, MTX 15, MMF 1) and biologic therapies (278). Table 1. illustrates the outcomes in pregnancy with some of the common drugs used in inflammatory arthritis. Table 1 Drug No. of pregnancies exposed to drug Live births related to exposed drug Elective terminations Spontaneous 1st trimester loss Spontaneous 2nd/3rd trimester loss Major malformations Minor malformations No. Studies Prednisolone 176 Min 110 0 1 1 9 1 11 Methotrexate 15 Min 5 2 2 0 0 0 5 Leflunomide 80 Min 78 1 0 0 7 63# 8 Sulphasalazine 57 Min 1 0 0 0 0 0 3 Azathioprine 210 208 14 4 5 8 0 7 HCQ 189 Min 142 0 14 0 10 0 8 Biologics (Inflix, etan, ada)* 118 88 10 29 4 0 4 1 Rituximab 155† 92 28 33 3 3** 1 3 *No increased risk of minor malformations compared to control group * Data is analysed from Verstappen (2010) only. † 29/155 pregnancies were in RA patients. ** One of the major congenital abnormalities (Turners syndrome) was diagnosed pre-administration of RTX. # No increased risk of minor malformations compared to control group. BSRBR: British Society for Rheumatology Biologics Registers. Conclusions Evidence supports the safety of HCQ, AZA, SSZ and steroids in pregnancy. Data from a limited number of pregnancies exposed to anti-TNF alpha drugs predominantly in or immediately prior to the first trimester exposures indicates that the number of spontaneous miscarriages and congenital malformations do not appear to be increased. Further registry data is required for biologic drugs, before the safe use of these drugs can be advocated throughout pregnancy in patients with inflammatory arthritis. Disclosure of Interest None declared DOI 10.1136/annrheumdis-2014-eular.5548

Background Azathiopine (AZA) is commonly used to ensure SLE disease control during pregnancy. There is little information regarding long term outcomes of these children. Objectives We aimed to assess if AZA exposure during pregnancy and lactation was associated with adverse outcomes for children born to mothers with lupus. Methods Women attending lupus clinics, with available pregnancy data, were consented to take part. A standard questionnaire developed for a multi-centre study was used to collect the data. Results Complete data was available for 287 children born alive to 200 women: 66% Caucasian, 15% South Asian, 10% Afro-Caribbean, 1% Chinese, 1% Hispanic, 4% other and 4% not stated. The median (range) age of women at delivery was 32 (19-44) yrs, and duration of disease was 6 (0.5-27) yrs. Two groups were studied: 89 children exposed to AZA during pregnancy &/or breast-feeding (group A) vs 198 children unexposed (group NA). Maternal renal disease (group A 38/67, 57% vs group NA 28/175, 16%), use of maternal prednisolone (group A 78/89, 87% vs group NA 93/198, 47%) and aspirin (group A 70/84, 83% vs group NA 132/200, 66%) were significantly increased in group A compared with group NA (p<0.005). No significant differences in exposure to hydroxychloroquine or heparin between the groups was identified. There was no significant difference in maternal pre-eclampsiaa, hypertension, anti Ro and/or anti-La antibodies, between the two groups. The presence of maternal lupus anticoagulant and/or anticardiolipin antibodies (ACA) was statistically higher in group A 44/73 (60%) vs group NA 79/178 (49%), p=0.03. However the presence of the antibodies individually was not different. Median age of children at enrolment in group A was 3.2, (range 0-17.1) vs group NA 3.1 (0-16.2) years. The median birth weight in group A was 2.8 (0.6-4.4)kg vs 3.1 (0.7-4.7) kg in group NA and gestational age in group A was 36.3 (27-42) wks vs 38 (25-42) wks in NA. These were both statistically significant. Using univariate analysis children exposed to AZA had a significantly increased rate of hospital admissions for infection (group A 25/89, 28% vs group NA 28/198, 14%, p=0.005), but not for overall hospital admissions (group A 29/84, 35% vs group NA 45/192, 23%, p=0.07) or out- patient visits. When the effect of AZA was adjusted by multifactor logistic regression for maternal renal disease and prednisolone use, the relationship became non-significant (OR 1.73 (0.85-3.5), p=0.13, however, due to the lower sample size its power was <80%. Conclusions This study shows that women with SLE who take AZA during pregnancy are more likely to have had renal disease, ACA and/or lupus anticoagulant and to use prednisolone. The median birthweight and duration of pregnancy was lower. There appeared to be an increased risk of hospital admission due to infection in the children of group A mothers. However, after adjusting for markers of maternal lupus disease severity this relationship disappeared. Disclosure of Interest : None declared DOI 10.1136/annrheumdis-2014-eular.1616

OBJECTIVES:The aim of the Systemic LUpus Erythematosus Cost of Care In Europe (LUCIE) study was to evaluate the annual direct medical costs of managing adults with active autoantibody-positive disease on medication for SLE in secondary care. This paper presents the UK analyses only.METHODS:A cost-of-illness study was conducted from the perspective of the National Health Service. Health resource utilization data were retrieved over a two-year period from four centres in England and unit cost data were taken from published sources.RESULTS:At baseline, 86 patients were included, 38 (44.2%) had severe SLE and 48 (55.8%) had non-severe SLE. The mean (SD) SELENA-SLEDAI score was 7.7 (5.7). The mean (SD) annual direct medical cost of was estimated at £3231 (£2333) per patient and was 2.2 times higher in patients with severe SLE compared with patients with non-severe SLE (p < 0.001). Multivariate model analyses showed that renal disease involvement (p = 0.0016) and severe flares (p = 0.0001) were associated with higher annual direct costs.CONCLUSIONS:Improvement of the overall stability of SLE and early intervention to minimize the impact of renal disease may be two approaches to mitigate the long-term direct cost of managing SLE patients in the UK.

Background Studies have shown that 30-50% of SLE patients (pts) develop lupus nephritis (LN). Patients with high affinity anti-dsDNA antibodies (abs) are more likely to develop LN. The Farr radioimmunoassay is the best method of detecting these abs but is rarely available. Instead ELISA tests for anti-dsDNA abs that detect abs of variable affinity and Crithidia test for high affinity abs are used. It is unclear if the Crithidia test needs to be repeated at every visit with C3 and C4 & anti-dsDNA abs by ELISA to assess lupus activity and the risk of developing LN. Objectives To audit the relationship between anti-dsDNA abs by ELISA & Crithidia, & low C3 & C4 at baseline & follow-up in SLE pts with and without biopsy proven LN. Methods Data was recorded prospectively from 1989 including lupus activity, renal biopsy (WHO classification), anti-dsDNA abs by ELISA & Crithidia, C3 and C4 levels. Pts were excluded if not seen in the clinic before 1st renal biopsy, had <2 visits to the clinic or had missing baseline &/or followup data. Pts that developed Class III/IV ±class V were combined. Pure class V LN was analysed separately. Results Of 309 pts, 290 (94%) were eligible: 93%female, 7% male,12.7% Afro-Caribbean, 17% South Asian, 63%Caucasian, 2% Chinese & 3% recorded as mixed/other. The mean±sd age was 49±15.1 & disease duration 17.5±8.3 years. There were 43 eligible pts with renal biopsies. Distribution of each WHO class of LN was Class II 16%, Class III/IV 63%, and Class V 9%. The 27 pts that developed Class III/IV LN (LNIII/IV) were compared with remaining 263 pts. At baseline, 78% of LN III/IV patients had anti ds-DNA abs by ELISA and Crithidia, vs 20% positive (pos) for both without LN III/IV, see table 1; 81% of LN III/IV were pos for each anti-DNA abs test vs 33% without LN III/IV ELISA pos and 22% Crithidia pos. Low C3 & C4 was present in 63% with LN III/IV and 7% without LN III/IV, and 63% with LN III/IV had pos ELISA & low C3/C4 vs 5% without LN III/IV. Pos ELISA, Crithidia and low C3/C4 was found in 78% with LN III/IV vs 5% without LN III/IV. There was no significant relationship between any anti-dsDNA antibody or complement test singly or in combination at baseline and the subsequent development of class V lupus nephritis. Over time more patients developed anti-dsDNA abs measured by ELISAand Crithidia. Developing Crithidia pos later was not helpful in identifying those more at risk of LN III/IV than baseline measurement. The combination of anti-dsDNA by ELISA with low C3/C4 was sensitive for identifying those at risk of LN III/IV. Conclusions This audit suggests that the combination of anti-ds DNA abs by ELISA and Crithidia together with low C3/C4 results at baseline can help to identify patients at most risk of developing class III/IV LN. Results do not support the need to measure anti-dsDNA abs by Crithidia serially, but do support the combination of ELISA, C3 & C4. Disclosure of Interest None Declared

Background: Children and young adults with JIA have increased levels of poor oral hygiene and dental decay [1].Periodontitis and types of arthritis are linked by similar components of blood cytokine profiles.Good dental health can be directly affected in JIA patients due to physical limitations in upper limb movements making brushing and flossing teeth difficult.An important factor in oral care is good dental hygiene and access to dental health practitioners.NHS advice is that all children should be reviewed by a dentist annually and be offered both sealant of their teeth and fluoride varnish at the appropriate time.Our aim was to establish if our patients had any barriers to accessing dental care.Methods: All patients (age 18 and under) diagnosed with JIA in the paediatric rheumatology clinic over a period of 3 months were asked to complete a dental care questionnaire.Parents completed the questionnaire for their children if necessary.Data were analysed using Excel.Results: 30 questionnaires were completed.Demographics were M:F 1:1.3, all children were diagnosed with JIA, average age 10.5 years with range 2-18.27 children were registered with an NHS dentist with the exception of one child with a private dentist.26 children had seen a dentist at least annually and one child in the past 2 years. 2 children, one aged 16, were not registered with a dentist because their parents didn't think it was important.11 children had 25 fillings in total, 9 of these children were not supervised during dental hygiene.13 children admitted to drinking sugary drinks daily and had 16 dental fillings.None of the children admitted to smoking.Conclusions: Whilst our audit showed that most children were registered with a dentist and were reviewed annually, only 1 child had been offered sealant and fluoride varnish.The NHS advises that children with chronic medical conditions can be seen either by their NHS dentist or by the local Community specialist dental service which can be accessed by referral from their rheumatology department or NHS dentist.None of the children were seen by the specialist dental service.We have developed an information leaflet informing parents and children with JIA of the importance of dental health explaining the benefits of both sealant and fluoride for teeth.

Background Recent Canadian data suggested an increased male:female ratio in children born to women with SLE1. However results combined from PROMISSE and RRNL data demonstrated no ignificant gender difference2 but reported hydroxychloroquine(HCQ) was protective against cardiac neonatal lupus(NL) in subsequent pregnancies in mothers with anti-Ro/La antibodies and a prior child with cardiac-NL3 Objectives To address the following questions in UK cohort of children born to mothers with SLE: 1)Does HCQ protect against NL? 2)Is there an increased M:F ratio? Methods A cross sectional, retrospective study recruited women with 4 SLE ACR criteria attending 8 specialist UK clinics with available pregnancy data for children under 17 yrs born after maternal SLE diagnosis. A standard questionnaire developed for multi-centre study was used to collect data on each pregnancy and the child outcome. Data collected included ethnicity, age of mother, disease duration, autoantibody profile, exposure to HCQ and immunosuppressive drugs during pregnancy/lactation, age and gender of child, and medical conditions assessed at a hospital in the child. Results Complete data was available for 288 children born alive to 200 women: 66% Caucasian, 15% South Asian, 10% Afro-Caribbean, 1% Chinese, 1% Hispanic 4% other and 4% not stated. The median (range) age of women at delivery was 32 (19-44) yrs, and duration of disease was 6 (0-27) yrs. There were 128 (44%) male and 160 (56%) female children with M:F gender ratio of 0.80 (95% CI 0.63-1.01). In total 136 (47%) of children were exposed to HCQ and 152 (53%) were unexposed during pregnancy. Congenital heart block (CHB) was observed in 7 (2.5%) children and cutaneous NL was reported in 6 (2%) children. The frequency of CHB or cutaneous NL, was not significantly different in children exposed to HCQ versus those not exposed, OR 1.27 for all types of NL, 95%CI 0.39-4.08 (p=0.69). As expected the only significant risk factor associated with cardiac and cutaneous NL was anti-Ro±La antibodies, OR15.1, 95%CI 1.90-119.8 (p=0.01), which remained significant on multivariate analysis OR16.4 (p=0.01). There were no other significant potential risk or protective factors associated with cardiac or cutaneous NL; none of the other immunosuppressive drugs offered protection. Conclusions Gender ratio was almost equal and consistent with PROMISSE and RRNL data, in contrast to previously published Canadian data.. In contrast to published registry data that HCQ may be protective against recurrence of cardiac NL in subsequent pregnancies, HCQ was not protective against CHB &/or cutaneous NL in this UK cross sectional study. References: Vinet et al. Increased M:F Ratio in Children Born to Women with SLE.A&R.2013. Letter.[Epub ahead of print] Lockshin et al. Sex ratios of children of lupus pregnancies.A&R.2013.65:282 Izmirly et al. Maternal use of HCQ is associated with a reduced risk of recurrent anti-Ro-antibody-associated cardiac manifestations of NL.Circulation.2012:126:76-82 Acknowledgements Lupus UK Disclosure of Interest: None Declared

Background Certain immunosuppressive agents are used in pregnancy in SLE patients to prevent flare and ensure optimal outcome for mother and child. There is little literature regarding long term outcomes of children.US data suggested a link between azathioprine(AZA) and infection1, and between anticardiolipin antibodies(abs) and developmental delay2 have been suggested. Objectives To assess if potential risk factors including maternal disease, autoantibodies, immunosuppression or other medication exposure influence health and behaviour of children. The following will be addressed:1)Does exposure to AZA in pregnancy &/or lactation increase risk of serious infections defined as requiring hospital assessment in the exposed children 2)Do immunosuppressive drugs or anti-phospholipid abs increase risk of developmental problems (delay±special needs±attention deficit disorder(ADD)±special educational needs). Methods A cross sectional, retrospective study recruited women with 4 SLE ACR criteria attending 8 specialist UK clinics with pregnancy data for children under 17yrs born after maternal SLE diagnosis. A standard questionnaire developed for this multi-centre study and used in US was used to collect data. Potential factors influencing infection and developmental problems, assessed were AZA, prednisolone, hydroxychloroquine, aspirin or heparin use in pregnancy, antiphospholipid syndrome, lupus anticoagulant &/or anticardiolipin (IgG&/orIgM) abs. Potential confounders assessed were pregnancy duration, pre-eclampsia, intra-uterine growth restriction, low birth weight, child age at assessment, maternal age at delivery, disease duration, prior renal biopsy and smoking. Results Complete data was available for 288 children born alive to 200 women: 66% Caucasian, 15% South Asian, 10% Afro-Caribbean,1% Chinese, 1% Hispanic 4% other and 4% not stated. The median (range) age of women at delivery was 32 (19-44) yrs, and duration of maternal disease was 6 (0-27) yrs. Infections were reported in 45(16%) of children. The odds ratio (OR) for infection in children with exposure to AZA was 1.63 (95%CI 0.84-3.15, p=0.15). In total 14(4%) of children were reported to have developmental delays, 7(2.5%) special needs, 3(1%) ADD and 6(2%) special educational needs. OR for developmental problems with AZA exposure was 1.22, 95%CI 0.44-3.42, and if mothers had antiphospholipid abs was 0.73, 95%CI 0.22-2.35. The age of child at assessment was associated with a small increased risk of developmental problems OR 1.15, 95%CI 1.05-1.26, p=0.003. No other potential risk factors were associated with infections or developmental problems. Conclusions No risk factors were identified for infection requiring hospital assessment, and only age of child at assessment was associated with increased risk of developmental problems. This study provides data to counsel women with SLE. References Akhtaret al. Maternal Anticardiolipin Affects Childhood Development.A&R.2008.57:9,443 Marderet al. In utero AZA exposure and increased utilization of special educational services in children born to mothers with SLE.A&R.2012:8:288 Acknowledgements LupusUK Disclosure of Interest: None Declared

Background: Patients with SLE are known to have early development of atherosclerotic process. However pathogenic mechanisms of accelerated atherogenesis is not completely detected in patients with SLE, though we think that chronic inflammatory process changes endothelium status and contributes to vascular impairment. Antiphospholipid syndrome (APS) is one of the most important causes of thrombosis in SLE. In addition, recent research indicates association of cardiovascular complications with hyperhomocysteinemia. The aim of this study was to investigate the association of thrombosis with plasma total homocysteine (ptHcy), dyslipidemia, antiphospholipid antibodies (aPL) and other vascular risk factors in SLE patients. Methods: 53 patients with SLE mean age 31,6 ± 6,8 years and 31 healthy persons age and sex-matched were examined. Mean duration of disease was 65 ± 4 months. Clinical and immunological data were obtained from our prospective computerized database. aPL-positivity was defined according to Sapporo criteria. In all patients with SLE activity of disease was evaluated according to disease activity index (SLEDAI). The median SLE disease activity index (SLEDAI) was 16,2 ± 7,5. Results: There was no association between aPL and ptHcy and dyslipidemia. ptHcy was higher in patients with arterial (median 15.06 versus 11.12 micromol/L, P = 0.010) but not venous thrombosis. In the subgroup analysis, this association was only seen in aPL-negative patients. In logistic regression, aPL (OR 6.85, 95% CI 1.93-26.41) and ptHcy (OR 1.12, 95% CI 1.07-1.31) were independently associated with arterial thrombosis. However, when hypertension, smoking and plasma total cholesterol were added to the model, only aPL (OR 7.68, 95% CI 2.02-31.84) and hypertension (OR 7.83, 95% CI 3.25-28.44), but not ptHcy, remained independently related to arterial events. aPL was the only variable independently related to venous thrombosis (OR 7.76, 95% CI 1.70-34.46). ptHcy concentrations and total cholesterol are higher in SLE patients with arterial thrombosis. No interaction between homocysteine, total cholesterol and aPL was found. We believe that increased levels ptHcy and total cholesterol may be a marker of increased vascular risk in aPL-negative SLE patients. Conclusions: In this study we did not find association between aPL with hyperhomocysteinemia and dyslipidemia, although hyperhomocysteinemia and dyslipidemia correlated with disease activity index (SLEDAI). The role of homocysteine and dyslipidemia as a marker of vascular risk may depend on the presence of traditional risk factors, such as man sex, smoking and hypertension, although a modest intrinsic effect cannot be entirely excluded. Disclosure statement: The authors have declared no conflicts of interest.

Individuals with systemic lupus erythematosus (SLE) have an increased susceptibility to certain types of cancer. Of particular concern are haematologic malignancies, specifically non-Hodgkin lymphoma, where a three- to four-fold increased risk is seen in SLE, compared with the general population. There is some evidence that immunosuppressive exposures play a role, although there appear to be other factors driving the risk. Lupus disease activity, with resultant dysregulated lymphocyte proliferation, may itself be a mediator of the association between SLE and lymphoma. Aside from haematologic malignancy risk, lung cancer also is increased in SLE compared with the general population, and smoking likely drives this risk in large part. Last but not least, cervical dysplasia is a concern in women with SLE, particularly with exposure to immunosuppressants; routine screening for this complication should not be neglected.

Pregnancy is an issue that should be discussed with all patients with rheumatic diseases who are in the reproductive age group. Infertility is rarely due to the disease but can be associated with cyclophosphamide therapy. Most rheumatic diseases that are well controlled prior to pregnancy do not deteriorate in pregnancy, providing that the patient continues with appropriate disease-modifying therapy. Some patients with inflammatory arthritis go in to remission during pregnancy. Patients with renal involvement may be at increased risk of disease flare. This needs to be distinguished from pre-eclampsia. Intrauterine growth restriction is more likely in patients with active systemic disease, hypertension, a history of thrombosis and renal involvement. Premature delivery may need to be planned to reduce the risks of stillbirth and can be associated with a variety of neonatal complications. Post-partum flare is common in all the rheumatic diseases.