Background: Psoriatic arthritis (PsA) is a heterogeneous inflammatory disease marked by various clinical features that significantly impact the quality of life for affected patients. While upadacitinib (UPA) has shown promising results in improving the signs and symptoms of PsA in two RCTs, real-world evidence on its effectiveness remains limited. Objectives: This study aims at assessing the real-world effectiveness of UPA in a large multicenter cohort of PsA patients by investigating its impact on key disease activity scores on both axial and peripheral engagement of the disease. Methods: Data of patients meeting CASPAR criteria for peripheral PsA and/or ASAS criteria for axial PsA were collected from 28 Italian rheumatology centers. The study included bio-naïve patients with a history of failure or intolerance to at least one conventional synthetic DMARD, as well as patients who had experienced failure with at least one biologic DMARD (bDMARD). Disease activity scores, such as LEI, DAPSA, MDA, VLDA, and ASDAS-CRP, were assessed at baseline and at 12, 24, 36, and 52 weeks. Adverse events were recorded at each visit. Paired t-tests were employed to compare disease activity indices at baseline and various follow-up times, while McNemar’s test was used to assess differences in the proportions of patients achieving MDA and ASDAS inactive disease at different follow-up intervals. Treatment persistence was estimated using the Kaplan-Meier method. Results: A total of 253 PsA patients, consisting of 181 females (71.5%) and 72 males (28.5%), underwent UPA treatment. The average age was 55.8±10.9 years, with a mean disease duration of 111.2±102.5 months. Peripheral joint involvement was observed in 161 individuals (63.6%), with 49 (33.6%) presenting an oligoarticular and 97 (66.4%) a polyarticular pattern (Table 1). Axial involvement was diagnosed in 92 patients (36.4%). Notably, 89 individuals (94.5%) had previously experienced treatment failure with at least one biologic DMARD (Table 1). Global treatment persistence was 77%, with a mean follow-up duration of 21.82 months. There were no differences observed between the axial and peripheral disease subsets or among the first line and further lines of treatment (Figure 1). Compared to baseline, a significant decrease in DAPSA (mean reduction 11.68; p<0.001) and ASDAS-CRP (mean reduction 1.00; p<0.001) was observed already at 3-month follow-up and throughout the observation period up to 52 weeks (Figure 1). The mean LEI score dropped from 0.64 at baseline to 0.36 at 12 weeks and 0.46 at 52 weeks (p<0.001). Additionally, a noteworthy increase in the percentages of patients achieving MDA (31.1%) and ASDAS-inactive disease (40.7%) was recorded at 52 weeks compared to baseline (Figure 1). During the follow-up period, 51 patients discontinued UPA [lack of efficacy (10 cases), loss of efficacy (25 cases), adverse events (4 cases), and infectious events (3 cases)]. The reasons for the remaining discontinuations were not specified. Conclusion: This study, conducted on a large cohort of PsA patients in a real-life setting, confirms the substantial benefits of UPA on both peripheral and axial disease involvement in the long term. It highlights UPA rapid onset of action, evident as early as the 3-month follow-up. REFERENCES: NIL. Acknowledgements: NIL. Disclosure of Interests: Giuseppe Lopalco: None declared, Eleonora Celletti: None declared, Maria Morrone: None declared, Fabiola Atzeni: None declared, Eleonora Bruschi: None declared, Antonio Carletto: None declared, Alberto Cauli declares consultation fee and meeting expenses from Abbvie, Massimiliano Cazzato: None declared, Maria Sole Chimenti: None declared, Francesco Ciccia: None declared, Francesco Cipollone: None declared, Fabrizio Conti: None declared, Addolorata Corrado: None declared, Francesca Cozzini: None declared, Lorenzo Dagna: None declared, Rosario Foti: None declared, Stefano Gentileschi: None declared, ELISA GREMESE: None declared, Roberto Gorla: None declared, Giuliana Guggino: None declared, Alberto Lo Gullo: None declared, Michele Maria Luchetti: None declared, Carlomaurizio Montecucco: None declared, Roberta Ramonda: None declared, Angelo Semeraro: None declared, Fausto Salaffi: None declared, Carlo Salvarani: None declared, Carlo Selmi: None declared, Emanuela Praino: None declared, Roberto F. Caporali: None declared, Florenzo Iannone: None declared.

Background: Nowadays, numerous therapeutic options are available for patients with rheumatoid arthritis (RA), and the number of patients achieving remission has significantly increased. However, some patients have a disease defined by EULAR as “difficult-to-treat”1 (D2T), which today represents a new challenge for rheumatologists. Objectives: The primary objective was to evaluate the characteristics of the D2T-RA population recorded in the Italian GISEA registry undergoing treatment with b/tsDMARDs. The secondary objective was to assess the effectiveness and changes at 6 and 12 months in disease outcomes, stratifying the analysis by different mechanisms of action. Methods: For this study, data from RA patients treated with b/tsDMARDs recorded in the Italian GISEA registry from January 2017 to December 2023 were analyzed. Disease activity scores and patient-reported outcomes (PROs) were recorded at baseline, and at six- and twelve-month follow-up. D2T-RA patients were defined by meeting these three criteria: 1) failure of ≥ 2 b/tsDMARDs (with different mechanisms of action); 2) signs suggestive of active/progressive disease (at least moderate disease activity according to disease activity scores and/or glucocorticoid treatment ≥ 7.5 mg/day prednisone or equivalent); 3) patient VAS (Visual Analogue Scale) pain and/or PtGA (Patient Global Assessment) and/or PhGA (Physician Global Assessment) > 20. The retention rates were estimated using the Kaplan-Meier method and compared with log-rank test, while repeated measures ANOVA (Analysis of Variance) was used to assess changes in disease activity and PROs during follow-up. Results: The GISEA cohort included 5251 treatment lines with b/tsDMARDs. We were able to assess the D2T category for 3439 cases at the start of a new treatment. Overall, 1060 (30.8%) patients met all three criteria for D2T-RA, while 2379 (69.2%) patients were not categorized as D2T. Table 1 reports the demographic and clinical characteristics of D2T-RA patients. Patients with D2T-RA showed higher disease activity and PRO scores at baseline. Notably, JAK inhibitors (JAKis) were used more frequently in D2T-RA patients (48.8%) compared to non-D2T-RA patients (33.3%, p<0.05).Globally, the 5-year survival rate was significantly lower for D2T-RA patients compared to those with non-D2T-RA (47.5% vs 62.5%, p<0.001). No significant differences in persistence were observed among the classes of b/tsDMARDs used in D2T-RA (log-rank test: 6.76, p=0.15), with a 5-year survival rate of 38.7% for abatacept, 41.2% for TNFi, 48.1% for IL6r inhibitors, 62.3% for anti-CD20, and 49.6% for JAK inhibitors. Figure 1 shows changes from baseline in disease activity scores and VAS pain in D2T-RA according to b/tsDMARD class. DAS28-ESR was reduced in all b/tsDMARD classes, except for TNFi. CDAI was reduced in all b/tsDMARD classes without any differences among the classes. VAS pain was reduced in all classes. For VAS pain, we observed a significant difference between abatacept and JAK inhibitors, with JAK inhibitors showing greater reduction in VAS pain (p<0.05). Also filgotinib, the latest JAK inhibitor approved onto the market, exhibited a significant decrease on pain perception at both time points. Conclusion: Our study provides a snapshot of D2T-RA patients within the GISEA registry. All currently used b/tsDMARDs appear to be effective in this patient cohort. The higher efficacy of JAK inhibitors, particularly in managing pain symptoms in these patients, warrants further investigation. REFERENCES: [1] Nagy G, Roodenrijs NMT, Welsing PM, et al. EULAR definition of difficult-to-treat rheumatoid arthritis. Ann Rheum Dis. 2021 Jan;80(1):31-35. Acknowledgements: We thank Ing. Massimiliano Dellisanti Fabiano Vilardi for his valuable contribution to database creation and management. Disclosure of Interests: None declared.Figure 1

OBJECTIVES:We aimed to investigate the effectiveness of tumour necrosis factor inhibitors (TNFi), anti-interleukin-17 or interleukin-12/23 monoclonal antibodies (anti-IL) on comorbidities in a cohort of patients with spondyloarthritis (SpA), using an average treatment effect (ATE) analysis. METHODS:SpA patients from the multicentre Italian GISEA Registry were divided into groups according to pharmacological exposure: no treatment (G0), TNFi (G1) and non-responders to TNFi switched to anti-IL (G2). In each group, we recorded the prevalence and incidence of infectious, cardiopulmonary, endocrinological, gastrointestinal, oncologic, renal and neurologic comorbidities. Each comorbidity was then fitted for ATE and baseline features were evaluated for importance. RESULTS:The main findings of this study comprising 4458 SpA patients relate to cancer, other gastrointestinal diseases (OGID) and fibromyalgia. ATE showed no increased risk of solid cancer in G1 (0.42 95% CI 0.20-0.85) and G2 (0.26 95% CI 0.08-0.71) vs. G0, with significantly higher incidence in G0 (14.07/1000 patient-years, p=0.0001). Conversely, a significantly higher risk of OGID and fibromyalgia was found in G1 (1.56 95% CI 1.06-2.33; 1.69 95% CI 1.05-2.68, respectively) and G2 (1.91 95% CI 1.05-3.24; 2.13 95% CI 1.14-3.41, respectively) vs. G0. No treatment risk reduction was observed in haematological malignancies, cardiovascular events and endocrinological comorbidities. CONCLUSIONS:Overall, our study confirms the safety of TNFi and anti-IL in SpA patients, albeit with some caveats pertaining to solid cancers, OGID and fibromyalgia. Furthermore, taking into consideration causality with observational data may yield more reliable and relevant clinical information.

Objective We aimed to evaluate ixekizumab (IXE) effectiveness, drug survival and clinical response predictors in moderate-severe psoriatic arthritis (PsA) patients in different clinical scenarios. Methods This was a multicentre real-life observational study based on Gruppo Italiano Studio Early Arthritis (GISEA) registry of IXE treatment in PsA patients (January 2019-June 2023). Data were collected at baseline and every six months. Results 223 PsA outpatients were included. Statistically significant improvement was observed after 6 (T6), 12 (T12) and 24 (T24) months of therapy for tender and swollen joint count (TJC and SJC), Visual Analogue Scale (VAS)-pain and Disease Activity in PSoriatic Arthritis (DAPSA) score. DAPSA remission was reached at T12 in 22% and at T24 in 18.5% of patients. At baseline, higher fibromyalgia and combination therapy with conventional synthetic disease- modifying anti-rheumatic drugs (csDMARDs) in females with respect to males and higher Psoriasis Area Severity Index (PASI) in males than in females were observed. Therapeutic effectiveness showed in males higher DAPSA and VAS-pain reduction, higher percentage of males in DAPSA remission/low disease activity (LDA) at T6, and higher triangle PASI at T6 and T12 than in female patients. At multivariate analysis, male sex was predictive for treatment response at T6 [p=0.02, odds ratio (OR) 2.49 (95% confidence interval 1.11-5.54)], while it lost significance at T12. Conclusion IXE effectiveness was highlighted after 6 months at both joint and skin levels and lasted up to 24 months in different clinical scenarios, making IXE effective in the complexity of managing PsA in a real-life setting.

OBJECTIVES:The revised Fibromyalgia Impact Questionnaire (FIQR) is a widely used fibromyalgia severity assessment tool that was introduced in 2009 prior to the publication of the American College of Rheumatology (ACR) preliminary fibromyalgia criteria in 2010 and its revision in 2016. In 2020, the modified Fibromyalgia Assessment Scale (FASmod) was published. The Polysymptomatic Distress scale (PSD) of the fibromyalgia criteria and FASmod include assessments of pain location severity and can be used for diagnosis as well as in non-fibromyalgia patients. The aim of this study is to provide equations for the conversion of the FIQR scores to PSD and FASmod as an aid to understanding and sharing fibromyalgia severity information.METHODS:3089 patients with fibromyalgia, diagnosed according to the ACR 2010/2011 criteria and belonging to the Italian Fibromyalgia Registry completed FIQR, FASmod and PSD questionnaires. Pearson's correlation coefficient was used to test the correlations between indices. The least square regression approach was used to produce predictive equations for each scale based on the remaining scales.RESULTS:FIQR was correlated with PSD (r=0.714) and FASmod (r=0.801); PSD and FASmod showed the highest correlation (r=0.897), expected since they assess the same constructs. Predictive equations showing a linear model were effective in producing mean cohort values, but individual predictions deviated substantially, precluding prediction in the individual patient.CONCLUSIONS:Conversion equations that allow for interconversion of multiple scales fibromyalgia severity assessment scales are produced. These can be useful in obtaining mean values for cohorts but are not accurate enough for use in individual patients.

Background In the age of targeted-synthetic disease-modifying antirheumatic drugs (tsDMARDs), filgotinib represents the last JAK inhibitor available in Europe for rheumatoid arthritis (RA). Filgotinib is characterized by predominantly inhibition of JAK1 and its efficacy and safety have been highlighted by phase 2/3 studies, but no real-life data in RA are currently available. Objectives The aim of this study was to evaluate the effectiveness and safety profile of filgotinib in real-life setting in RA patients included in Italian GISEA (Group for the Study of Early Arthritis) registry. Methods For this study, data from RA patients treated with filgotinib recorded in Italian GISEA registry were analysed. Disease activity scores and patients reported outcomes (PROs) were compared at baseline and six months follow-up using paired t-tests. The retention rate was estimated by the Kaplan-Meier method, while a cox regression model was used to search for possible factors influencing drug survival. Results One hundred and seventy-nine patients (female 89.4%, age 57.8±12 years, FR/ACPA+ 64.3%, current/former smoker 31.8%) included in GISEA registry started filgotinib for active RA. Most patients were taking filgotinib as second (23.5%) or further (43%) b/tsDMARDs line of treatment. Filgotinib was used in monotherapy in 66.5% of patients, while 52% were not on treatment with glucocorticoids (GCs) at baseline. All demographic and clinical data are reported in Table 1. A follow-up visit was available for 122 patients (mean time of first follow-up visit: 4±2 months). As shown in table 1, we observed a decrease of all disease activity scores and PROs. At first follow-up visit, 67.8% of patients were in remission/low disease activity according to CDAI and 65.4% according to SDAI. Kaplan-Meyer analysis highlighted that drug persistence was similar either in monotherapy or combination therapy (Figure 1a), and irrespective of GCs at baseline (Figure 1b). However, a better persistence was observed in RA patients on first line treatment with filgotinib (Figure 1c). Thirty-five patients stopped filgotinib during follow-up, 10 for lack of efficacy, 4 for loss of efficacy, 4 for adverse events, while for the remaining cases the cause of drug discontinuation was unknown. No major cardiovascular events were reported. Finally, univariate Cox-regression model showed that b/tsDMARD naïve patients had a lower risk of drug discontinuation (naïve vs other lines: HR 0.37, 95%CI 0.20-0.86). Conclusion In Italian real-life setting, filgotinib confirms a good effectiveness and safety profile. REFERENCES: NIL. Acknowledgements: NIL. Disclosure of Interests None Declared.Figure 1Survival analysis in RA patients treated with Filgotinib.Table 1Demographic and disease characteristics at baseline and at first follow-up visit (T1) of RA patients treated with Filgotinib.Variablesbaseline (n. 179)T1 (n. 122)Agemean (SD), years57.8 (12.7)58 (12.6)Gendern. (%), femalen. (%), man160 (89.4)19 (10.6)110 (90.2)12 (9.8)BMIn. (%), underweightn. (%), heathy weightn. (%), overweightn. (%), obese4 (3.8)50 (48.1)21 (20.2)29 (27.9)4 (5.3)36 (47.4)15 (19.7)21 (27.6)Smokersn. (%), currently smokern. (%), former smokern. (%), never smoker13 (13.8)17 (18)64 (68.1)8 (13.5)12 (20.3)39 (66.1)IgG RF/ACPA +n. (%)74 (64.3)54 (71.1)VAS painmean (SD)63.1 (29.1)35 (30.7)***VAS PtGAmean (SD)60 (26.7)35 (29)***VAS PhGAmean (SD)44 (25.4)20.9 (23)***TJC28mean (SD)5.4 (5)2.7 (3.8)**SJC28mean (SD)3.3 (3.5)1.3 (2.2)***DAS28-ESRmean (SD)4.6 (1.3)3.3 (1.4)**CDAImean (SD)19.1 (11.3)9.6 (9.5)**SDAImean (SD)20.5 (12)10.3 (10)**HAQ-DImean (SD)1.3 (0.7)0.9 (0.7)***glucocorticoidn. (%)86 (48)47 (38.5)**Prednisone (equivalent), mg/diemean (SD)5.8 (3.7)3.4 (2.6)*csDMARD (in corso)n. (%)60 (33.5)31 (25.4)b/tsDMARD linen. (%), 1^ linen. (%), 2^ linen. (%), 3^ line or others60 (33.5)42 (23.5)77 (43)/Mean time of first follow-up visit: 4±2 months*p<0.05, **p<0.01, ***p<0.001

OBJECTIVES:To determine the cut-off values of Patient Acceptable Symptom State (PASS) for the revised Fibromyalgia Impact Questionnaire (FIQR), the modified Fibromyalgia Assessment Scale (FASmod), and the Polysymptomatic Distress scale (PSD) and to determine the predictors of PASS in patients with fibromyalgia (FM).METHODS:FM patients belonging to the Italian Fibromyalgia Registry (IFR) completed the FIQR, the FASmod and the PSD. The PASS was assessed using a dichotomous answer. The cut-off values were obtained through the receiver operating characteristic curve (ROC) analyses. A multivariate logistic regression analysis was performed to determine predictors of achieving the PASS.RESULTS:5545 women (93.7%) and 369 males (6.3%) were included in the study. The 27.8% of patients reported an acceptable symptom state. Patients in PASS differed in all patient-reported outcome measures (p <0.001). The FIQR PASS threshold was ≤58 (area under the ROC curve [AUC] = 0.819). The FASmod PASS threshold was ≤23 (AUC = 0.805) and the PSD PASS threshold was ≤16 (AUC = 0.773). In the pairwise AUC comparison, the discriminatory power of the FIQR PASS outperforms both FASmod PASS (p = 0.0124) and PSD PASS (p <0.0001). Multivariate logistic analysis showed that FIQR items related to memory and pain were the only predictors of PASS.CONCLUSIONS:The FIQR, FASmod, and PSD PASS cut-off points for FM patients have never been determined before. This study provides additional information to facilitate interpretation of the severity assessment scales in daily practice and clinical research related to FM patients.

BackgroundJanus kinase inhibitors (JAKi) are efficacious drugs for the control of disease activity in rheumatoid arthritis (RA). The effect on patient-reported outcomes (PROs) has been shown in randomized clinical trials, but is still largely unknown in real-life scenarios.ObjectivesTo assess the time to onset and magnitude of improvement across a number of different PROs in patients with RA treated with JAKi.MethodsPatients were selected from centres involved in the LORHEN Registry. Patients with a diagnosis of RA initiating a JAKi (Baricitinib or Tofacitinib) between May 2019 and February 2020 were clinically assessed at baseline, 8 weeks and 16 weeks with DAS28 and SDAI. PROs were collected at baseline, 2 weeks, 4 weeks, 8 weeks and 16 weeks from JAKi initiation. The PROs assessed at each timepoint were: Patient Global Assessment (PGA), General Health (GH), 0-100 mm visual analogue scale (VAS)-pain, VAS-fatigue, Health Assessment Questionnaire (HAQ), The Clinical Arthritis Activity (PRO-CLARA) questionnaire.ResultsThirty-eight patients (female 76%, mean age 57±15) were enrolled for a total of 152 visits for PROs assessments. Patients had received a median of 2 (IQR 1;3) previous conventional synthetic DMARDs (csDMARD) and a median of 2 (0;3) biological DMARDs. Mean disease duration at time of JAKi initiation was 10±9.5 years. JAKi was prescribed as combination therapy with a csDMARD in 20 (53%), concomitant glucocorticoids (GC) in 28 (73%) patients, at a mean dose of 6±2 mg/day. Twenty-one (55) patients were on regular analgesics. DAS28-ESR reduced from 4.5±0.9 at baseline to 2.6±0.9 at 8 weeks (p<0.001). DAS28-ESR remained stable between 8 and 16 weeks (mean values 2.3±0.1; p=0.17). Similarly, DAS28-CRP reduced from 4.1±0.8 to 2.1±0.6 at 8 weeks (p<0.0001). Between 8 and 16 weeks there was a further significant reduction of DAS28-CRP under the effect of JAKi treatment to a mean value of 1.4±0.5 (p<0.0001). SDAI reduced over the period of observation from 19±8 to 6±5 at 8 weeks (p<0.0001), and to 4±3 at 16 weeks (0.03). By week 8 the intake of concomitant GC reduced to 28% of patients (p=0.0001), and of analgesics to 2% of patients (p<0.0001).By two weeks, all PROs had a significant reduction compared to baseline (Figure 1). For PGA, VAS-pain, VAS-fatigue a further significant reduction was observed by week 4 to then stabilize between week 8 and 16. GH and HAQ had a significant reduction by week 2 and remained stable thereafter. All PROs were significantly lower during all time-points compared to baseline. The PRO-CLARA Road score to assess physical function significantly improved starting from week 2 from a score of 4±2 to 2±2 at the end of follow-up. The PRO-CLARA self-administered tender joint count was stable at week 2 (mean score 4±2) but recorded an improvement from week 4 onwards (p<0.03).Figure 1.Improvement of patient-reported outcomes over the period of observationConclusionTreatment with JAKi ensures a very rapid and persistent improvement of several PROs, including pain, fatigue and physical function, as early as 2 weeks treatment even in patients with long-standing RA treated with several lines of previous DMARDs.Disclosure of InterestsNone declared

In daily clinical practice, it is not rare to observe a relevant discordance between patient’s global assessment (PGA) and physician’s global assessment (PhGA), because of different illness perceptions.To evaluate the presence of PGA/PhGA discrepancy in patients with rheumatoid arthritis (RA) who achieved biological remission and to explore whether this discordance could be influenced by the presence of additional elements affecting patients’ quality of life, such as comorbidities, sleep disturbances and psycho-emotional factors.Our study included adult RA patients, followed in a single rheumatology centre, fulfilling three out of four Boolean remission criteria: tender joint count ≤1, swollen joint count ≤1 and C reactive protein ≤1 mg/dl1. Medical records including demographic data, clinical characteristics and outcomes measures were collected. To evaluate the impact of comorbidities we used the Rheumatic Disease Comorbidity Index (RDCI). Pain assessment, PGA and PhGA were rated on a visual analogue scale (0-100 mm) on the same day of the clinical evaluation. To analyse the discrepancy between PGA and PhGA, the [PGA-PhGA] variable was calculated, considering as discordant a difference ≥ 30 mm. All the subjects completed the following questionnaires: Health Assessment Questionnaire (HAQ), SF36 Health Survey, State-Trait Anxiety Inventory (STAI-Y1/Y2), Self-rating Depression Scale (SDS Zung) and Insomnia Severity Index (ISI). Statistical analysis was performed to compare concordant and discordant groups.The study included 90 patients (64 women, 26 men) with a median age of 60 [50-67] and a median RA duration of 10 [5-16] years; 80% of patients were RF and/or ACPA positive. According to DAS28 (median 1.7 [1.5-2.2]), 90% of patients achieved remission, 9% LDA and 1% MDA. Nevertheless, in 81% of cases PGA>PhGA and in 33% [PGA-PhGA] ≥30. The discordant group showed higher median disease activity scores (DAS28: 2.2 [1.9-2.4] vs 1.5 [1.4-1.9], p<0,001; SDAI: 6.7 [3.9-8.1] vs 2.0 [0.7-3.2], p<0.001). No differences were found in gender, age, comorbidities, RA duration, serology and treatment (cs-/b-DMARDs). Median prednisone dose was higher in the discordant group (0 [0-21] vs 0 [0-7.5] mg/week, p:0,024). Data about patient-reported outcomes and differences between two groups are reported in the table 1.In our study we found discordance between the global disease assessment reported by patients, although considered in biological remission, and their physicians in 33% of the cases. As previously described2, our data seems to confirm that potential causes for this discordance could be pain due to non-inflammatory processes, functional disability and compromised physical health. In the present study also psychological stress, particularly state anxiety, seems to negatively affect PGA/PhGA concordance. These results confirm the importance, in clinical practice, of exploring and managing inflammatory and non-inflammatory parameters separately, also in patients achieving biological remission.[1]Svensson B et al. BMJ Open. 2013. 2 Craig ET et al. Arthritis Care Res. 2019.Table 1.Data are expressed as median (1st-3rd quartile) and compared using Mann-Whitney test.PATIENT-REPORTED OUTCOMEsPATIENT NUMBER=90CONCORDANT GROUP(PGA-PhGA) <30=60 (67%)DISCORDANT GROUP(PGA-PhGA) ≥30=30 (33%)p <0,05PGA, [0-100]20 (10-40)10 (0-20)50 (37-60)-PhGA, [0-100]0 (0-10)0 (0-10)10 (0-10)-PGA-PhGA, [0-100]10 (0-30)10 (0-10)40 (30-52)-VAS pain, [0-100]20 (10-40)10 (0-20)50 (30-60)<0,001HAQ, [0-3]0,13 (0-0,50)0 (0-0,25)0,31 (0,09-0,87)<0,001STAI Y1, [20-80]36 (31-47)35 (30-42)40 (35-51)0,014STAI Y2, [20-80]37 (33-45)37 (33-45)38 (33-47)nsSDS ZUNG, [25-100]45 (40-51)44 (39-49)48 (40-53)nsISI, [0-28]2 (2-11)5 (2-11)8 (3-14)nsSF36 Physical, [0-100]44 (36-52)48 (40-54)36 (28-46)<0,001SF36 Mental, [0-100]50 (38-56)51 (41-57)47 (33-54)nsNone declared

Long-term observational data on the real-life use of JAK inhibitors (JAKis) for rheumatoid arthritis (RA) and their comparison with biological drugs are still very limited. Large population-based registries have been increasingly used to investigate the performance of targeted drugs in a real-life setting.The aim of this study is to evaluate and compare the 3-year retention rate of JAKis, TNF inhibitors (TNFis) and biologic drugs with other mechanisms of action (OMAs) in the large cohort of RA patients included in the Italian national GISEA registry.Data of all RA patients treated with targeted synthetic or biologic drugs were prospectively collected in the Italian multicentric GISEA registry. The analysis was limited to patients who started a first- or second-line targeted drug in the period after the first JAKi was marketed in Italy (1st December 2017). The 3-year retention rate was calculated by the Kaplan-Meier method and compared between different drug classes by a log-rank test. A descriptive analysis of reasons for discontinuation was performed.The study population included 1027 RA patients (79.8% females, mean age [±SD] 56.9 [±13.5] years, mean disease duration 9.8 [±9] years, mean baseline SDAI 17.5 [±11.9], ACPA positive 67.4%, RF positive 62.7%) who received JAKis (baricitinib or tofacitinib, n=297), TNFis (n=365), or OMAs (n=365) as first or second targeted drug. Main baseline characteristics of study population were overall well balanced between treatment groups. Retention rate was numerically but not statistically higher (p=0.18) in patients treated with JAKis compared with TNFis or OMAs (80.6, 78.9 and 76.4% at 1 year and 73, 56.8 and 63.8% at 3 years, respectively) (Figure 1). Drug survival was significantly higher in patients receiving concomitant methotrexate (MTX) compared with monotherapy only in TNFis (66.8 vs 47.1%, p=0.04) but not in JAKis (76.1 vs 70.1%, p=0.54) and OMAs (66.1 vs 61.9%, p=0.41) group. Therapy was discontinued in a total of 211 patients because of ineffectiveness (n=107), adverse events (n=88), or compliance/other reasons (n=16). The most frequent reason for treatment withdrawal was ineffectiveness in both JAKis (n=30 out of 56) and TNFis (n=45 out of 74) groups, whereas OMAs were discontinued more frequently because of adverse events (n=41 out of 81).Our data confirmed in a real-life setting a favorable 3-year retention rate of all available targeted mechanisms of action for RA therapy. As expected, concomitant MTX significantly impacted persistence on therapy of TNFis only. Discontinuations of JAKis for adverse events were infrequent overall, confirming the safety profile observed in randomized clinical trials.Figure 1.Three-year retention rate by treatment groupNone declared

Background:The World Health Organization recommends that infants should be exclusively breastfed until the age of 6 months. The aim of this study was to assess the ratio of breastfeeding patients with Rheumatic Diseases (RD) and to identify possible causes of its discontinuation.Objectives:The aim of this study was to assess the ratio of breastfeeding patients with Rheumatic Diseases (RD) and to identify possible causes of its discontinuation.Methods:This study was embedded in the P-RHEUM.it register, as a nationwide prospective cohort study collecting data of pregnancies in inflammatory RD. Pregnancies, enrolled until the 20th week of pregnancy, are followed from pregnancy until 6 months postpartum. At baseline, sociodemographic parameters, obstetric history, comorbidities are reported. During pregnancy, the course of maternal disease, development of foetus and complications are reported. After delivery, the pregnancy outcome, data on lactation and child development are collected.Results:From May 2018 to May 2020 data of 349 patients were available. Data on lactation were available in 44 pregnancies. Two months after delivery 37 were continuing breastfeeding (n=26) or mixed feeding (n=11), while 7 were using formula feeding. Among patients using formula feeding 2 had a diagnosis of rheumatoid arthritis (RA), 1 of juvenile idiopathic arthritis, 1 of undifferentiated arthritis (UA), 1 of anti-phospholipid syndrome (APS), 1 of vasculitis and 1 of systemic lupus erythematosus. The reasons of formula feeding were the following: 2 for agalactia, 2 for personal preferences, 3 for drug-related concerns (1 for physician’s decision in a patient with APS; 2 for maternal concerns about drugs in patients with RA and UA treated respectively with Adalimumab and Tocilizumab). At 6 months 30 continued breastfeeding (n=23) or mixed feeding (n=7) and 14 formula feeding. The reasons of formula feeding were available in 9 patients: 3 for agalactia, 2 for personal preferences, 2 for physician’s decision in a patient with APS and in a patient positive for anti-phospholipid antibodies; 2 for maternal concerns about drugs.Conclusion:Preliminary data of this prospective study demonstrate a high percentage of breastfeeding/mixed feeding after delivery and after 6 months among women with RD. Drug related concerns are the main reason of discontinuation of breastfeeding, although medication results compatible with lactation. Using our results, strategies supporting patients with RD whishing to breastfeed may be developed.References:[1]Carina Gotestam Sporken et al. The EULAR points to consider for use of antirheumatic drugs before pregnancy, and during pregnancy and lactation. Ann Rheum Dis 2016.[2]Sammaritano et al. 2020 American College of Rheumatology Guideline for the management of reproductive health in rheumatic and muscoloskeletal diseases, Arthritis Rheumatol. 2020.Acknowledgements:Authors would like to thank SIR study center and all patients who accepted to partecipate to our studyDisclosure of Interests:None declared

Background: In SLE pregnancies adverse pregnancy outcomes (APO) are more frequent than in general obstetric population (GOP). In clinical practice, low C3 and C4 levels are associated with active disease and, during pregnancy, complement activation products are shown to be associated with APO. Objectives: To analyse complement variations during SLE pregnancies, focusing on disease flares and APO. Methods: Data on SLE pregnancies prospectively-followed by multidisciplinary team in 2 Italian Centers from 1987 to 2018 were retrospectively analysed. C3 and C4 normal levels were calculated in general obstetric population (GOP) as previously described 1 , and related to maternal and fetal outcome. Non categorical variables were compared using Mann-Whitney test or Wilcoxon test when appropriate. Results: Two hundred forty-six pregnancies in 172 SLE patients were analysed (mean age at conception 31.3 ±4.9 years; mean disease duration 8.3 ±7.1). Anti-Ro antibodies were positive in 64 patients (37%) and anti-phospholipid antibodies (aPL) were positive in 84 (48%), with single positivity in 54%, double in 24% and triple in 21%; 9 patients (5%) had also a diagnosis of obstetric-antiphospholipid syndrome (APS) and 8 (4%) had thrombotic-APS. Seventy-one patients (41%) had history of Lupus Nephritis. Thirty-five flares were recorded in 30 pregnancies (12%). APO occurred in 47 pregnancies (19%) and were: 27 fetal loss (20 early miscarriage <10 th week and 7 intrauterine fetal death), 11 severe preterm birth (<34 th week) and 15 hypertensive disorder (11 pre-eclampsia and 5 pre-eclampsia+HELLP syndrome). In GOP, C3 progressively increased throughout pregnancy and C4 increase from the 1 st trimester to the 2 nd trimester, as well as in SLE pregnancies without flares and without APO, from preconception (Fig 1). In the other SLE groups, C3 and C4 showed a different trend: in pregnancies with flares, they did not increase from preconception to the 1 st trimester; in fetal losses and severe pre-term births, they remained stable throughout pregnancy; in hypertensive disorders they increased only between preconception and the 1 st trimester. C3 and C4 levels were higher in GOP than in all SLE pregnancies groups (including those without flares and without APO) in each trimester. SLE pregnancies without flares showed higher C3 and C4 levels than pregnancies with flares, at preconception and in each trimester. SLE pregnancies without APO had higher C3 and C4 levels than pregnancies with fetal death at 2 nd trimester, higher C3 levels than severe pre-term births in each trimester and higher C4 at 3 rd trimester (Fig.1). At preconception, pregnancies with flares showed a higher frequency of low C3 and of low C4 than in pregnancies without flares (76% vs 42%, p=0.01; 76% vs 26%, p<0.001, respectively). Using the normality range previously calculated in GOP, SLE pregnancies with flares had higher frequency of low C4 in every trimester as compared with pregnancies without flares (1 st : 82% vs 48%, p=0.003; 2 nd : 82% vs 64%, p=0.01; 3 rd : 64% vs 30%, p=0.002). At multivariate analysis, low C4 at preconception was associated with flare (OR [95% CI]: 10.34 [2.52-42.39]; p=0.001). Figure 1. Variations of C3 and C4 median levels (mg/dL) throughout pregnancy in GOP 1 and in SLE pregnancies without and with flare (a) and without and with APO (b). * p <0.05 ^as compared with SLE groups: p<0.05; § as compared with SLE pregnancies without flare: p<0.05; + as compared with SLE pregnancies without APO: p<0.05 Conclusion: In SLE pregnancies, monitoring of C3 and C4 is important: its failure to increase can be useful to recognize potential risk situations which deserve particular monitoring. References: [1]Reggia R. et al. Rheumatology 2012;51:2186-2190. Disclosure of Interests: None declared

Early diagnosis and timely and appropriate treatments positively influence the history of fibromyalgia syndrome (FM), with favourable repercussions at clinical, psychological, social and economic levels. Notwithstanding, there are still significant problems with timeliness of diagnosis, access to pharmacological therapies - particularly to innovative ones - and appropriate and effective taking in charge of patients. All the aforementioned factors have a great impact on FM patients' quality of life. Indeed, even though the World Health Organisation recognised FM as a chronic condition in the International Classification of Diseases 10th edition (ICD-10), many countries still fail to recognise the syndrome, and this negatively influences the capability to appropriately protect and care for patients. This is the case in several European Countries. In Italy, a few Regions have started to put in place precise indications for people suffering from FM, aiming at the implementation of diagnostic-therapeutic pathways. The Diagnostic-Therapeutic Care Pathway (DTCP) provides an important tool to meet the needs of patients suffering from chronic diseases. They present the organisation of an integrated assistance network. This includes a seamless path for disease prevention, diagnosis and treatment, by means of cooperation among physicians and other healthcare professionals.

Background:In rheumatic diseases, specifically in chronic inflammatory arthritis (IA), there are no data about the effectiveness of LDA in preventing pregnancy complications.Objectives:To assess the potential benefit of LDA administration during pregnancy to prevent adverse pregnancy outcomes (APO) in IA.To compare pregnancy outcomes in IA pregnancies treated during pregnancy with LDA with those untreated.Methods:Italian P-RHEUM.it register is designed as a nationwide, web-based longitudinal observational cohort study collecting data on pregnancies in inflammatory rheumatic diseases. At baseline, socio-demographic parameters, obstetric history, comorbidities and clinical/laboratoristic characteristics are registered. During pregnancy, the course of maternal disease, medications, development of fetus and complications are collected for each trimester.Results:From May 2018 to May 2020, 349 patients were enrolled. Forty-five patients with IA had completed their pregnancy with known outcome (20 Rheumatoid Arthritis, 10 Psoriatic Arthritis, 6 Ankylosing spondylitis, 5 Juvenile Idiopathic Arthritis, 4 Undifferentiated Arthritis): 13 (28.89%) treated with LDA during pregnancy and 32 (71.11%) without LDA prophylaxis. LDA was used with a higher frequency in women with previous APO, anti-phospholipid positivity and on biological DMARDs even if it’s not statistically significant (Table 1). All the LDA-treated women had a live-birth rate of 100%, whereas among women not treated with LDA the live birth rate was of 84.4% with 27 live-birth pregnancies, 3 early miscarriages, 1 fetal loss and 1 stillbirth observed. There were no significant differences between the LDA and the not LDA groups regarding pregnancy/peripartum obstetric complications (p=0.14), although less adverse pregnancy/peripartum outcomes were registered in LDA patients’ group (Table 1).Conclusion:The preliminary data of this prospective cohort study show that LDA improve pregnancy outcome in IA even if women treated with LDA had more risk factors for APO. The extension of this cohort will allow us to further investigate these important results.References:[1]Roberge S, et al. AmJObstetGynecol.2017;216:110-120.e6.[2]Andreoli L et al. Annals of the Rheumatic Diseases 2017;76: 476–85.[3]Sammaritano L et al. Arthritis & Rheumatology 2020; 72: 529–56.Table 1.Comparison between inflammatory arthritis pregnancies treated with low-dose aspirin and those untreated.LDA (n 19)Not LDA (n 39)pMaternal age at conception, median (IQR)33 (30.5 - 36.5)33 (28.2 - 35)BMI, median (IQR)25.4 (22.9 - 28.5)22.5 (20.4 - 24.5)Smoking, n (%)3 (15.8%)4/36 (11.1%)0.68Previous APO, n (%)7/11 (63.7%)7/13 (56.1%)0.34aPL+, n (%)3 (15.8%)1/36 (2.8%)0.1CRP+, n (%)5/15 (33.3%)11/27 (40.7%)0.75GC at baseline, n (%)9 (47.4%)15/34 (44.1%)0.82cDMARDs at baseline, n (%)1 (5.3%)3/34 (8.8%)1bDMARDs at baseline, n (%)6 (31.6%)11 (28.2%)0.79Live-birth pregnancy, n (%)18 (94.7%)32 (82.1%)0.25Pregnancy loss, n (%)1 (5.3%)7 (17,9%)0.25Hypertensive disorders, n (%)2/18 (11.1%)6/38 (15,8%)1HELLP syndrome, n (%)0 (0%)0 (0%)1PROM, n (%)0 (0%)3/38 (7.9%)0.54Preterm delivery, 34-37 GW, n (%)2 (10.5%)2 (5.2%)0.59Preterm delivery, <34 GW, n (%)1 (5.3%)1 (2.6%)1IUGR, n (%)0 (0%)3 (7.8%)0.54SGA, n (%)0 (0%)0 (0%)1Pregnancy/Peripartum complications, n (%)3 (15.8%)11 (28.9%)0.35Legend: aPL, Antiphospholipid syndrome; bDMARDs, biological disease-modifying antirheumatic drugs; BMI, Body mass index; cDMARDs, Conventional disease-modifying antirheumatic drugs; CRP, C-reactive protein; GC, Glucocorticoids; GW, gestational week; HELLP, haemolysis, elevated liver enzymes, and low platelets; Hypertensive disorder defined as gestational hypertension, preeclampsia and eclampsia; IQR, Interquartile range; IUGR, intrauterine growth restriction; LDA, Low-dose Aspirin; Pregnancy loss defined as early fetal-loss, miscarriage and still-birth; PROM, Premature rupture of membranes; SGA, Small for gestational age.Acknowledgements:I would like to acknowledge the Epidemiology Unit of the Italian Society for Rheumatology and the Investigators of the Italian Registry.Disclosure of Interests:None declared

Fibromyalgia syndrome is one of the most common causes of chronic widespread pain, but pain accompanies a wide range of ancillary symptoms. To date, its aetiopathogenesis remains elusive, and diagnosis is exquisitely clinical, due to the lack of biomarkers or specific laboratory alterations in fibromyalgia patients. This position paper has the purpose to summarise the current scientific knowledge and expert opinions about the main controversies regarding fibromyalgia syndrome, namely: (i) fibromyalgia definition and why it is still not recognised in many countries as a distinct clinical entity; (ii) fibromyalgia severity and how to evaluate treatment outcome; (iii) how to treat fibromyalgia and which is a correct approach to fibromyalgia patients.

Background: Patients with Spondyloarthritis (SpA) can experience flares during pregnancy and postpartum even though the available data are limited and not conclusive. Objectives: To assess disease activity and treatment modification during pregnancy and postpartum in patients with SpA and to identify risk factors for disease flare. Methods: Data on SpA pregnancies prospectively-followed in a pregnancy clinic from 2010 to 2019 were retrospectively analysed. Disease activity was assessed during each trimester and postpartum using ASDAS-CRP or DAS28-CRP. Flare was defined as an increase of disease activity leading to treatment modification (introduction or increase ≥5mg/day of prednisone, introduction of cDMARD or bDMARD) 1 . Results: Data on 50 pregnancies in 46 patients were collected (mean age at conception 33±4.7 years; median disease duration: 60 months (IQR 24-132); 33 psoriatic arthritis, 6 axialSpA, 2 reactive arthritis, 2 IBD-related SpA; 6 undifferentiated SpA, 1 juvenile idiopathic arthritis). Six pregnancies ended in miscarriage, so they weren’t considered for the analysis of flares during pregnancy (table 1). Fifteen out of 44 (34%) pregnancies had at least one flare during pregnancy (6, 7 and 4 during 1 st , 2 nd and 3 rd trimester respectively; 2 pregnancies had multiple flares). A higher rate of flare was observed in pregnancies of patients with axial involvement (p=0.01), on treatment with bDMARDs at preconceptional visit (p=0.03) and who stopped TNFi at positive pregnancy test (p=0.03). Peripheral involvement was associated with a lower rate of flares (p=0.02). Medications resumed during pregnancy were steroids (in 6 pregnancies), cDMARDs (2 sulfasalazine, 1 cyclosporine) and bDMARDs (4 certolizumab, 4 etanercept). During postpartum period flares were recorded in 46% of patients. Table 1. clinical features, medication and disease activity in pregnancies with flare vs without flare CLINICAL FEATURES FLARE (15) NO FLARE (29) p Axial involvement, n (%) 11/15 (73) 9/29 (31) 0.01 Peripheral arthritis, n (%) 8/15 (53) 26/29 (90) 0.02 Enthesitis, n (%) 5/15 (33) 14/29 (48) ns Dactilitis, n (%) 3/15 (20) 8/29 (28) ns Psoriasis, n (%) 6/15 (40) 17/29 (59) ns IBD, n (%) 2/15 (13) 0 ns Uveitis, n(%) 1/15 (7) 3/29 (10) ns HLAB27 + 7/11 (64) 5/12 (42) ns MEDICATION HISTORY bDMARDs, n (%) 11/15 (73) 7/29 (24) 0.003 bDMARDs at preconception visit, n (%) 8/15 (53) 6/29 (21) 0.04 bDMARDs stopped at positive pregnancy test, n (%) 7/15 (47) 4/29 (14) 0.03 cDMARDs, n (%) 12/15 (80) 25/29 (86) ns DISEASE ACTIVITY ACTIVE DISEASE* preconception visit, n(%) 3/14 (21) 4/23 (17) ns ACTIVE DISEASE 1 st trimester, n(%) 6/15 (40) 1/29 (3) 0.004 ACTIVE DISEASE 2 nd trimester, n(%) 8/15 (47) 2/29 (7) 0.001 ACTIVE DISEASE 3 rd trimester, n(%) 2/15 (13) 1/29 (3) ns *DAS28-CRP>3.2 or ASDAS-CRP≥2.1 Conclusion: In our cohort of prospectively-followed SpA pregnancies, 34% experienced a flare during pregnancy and 46% during postpartum. Flares occurred especially in those patients who discontinued TNFi early in pregnancy and with axial involvement. When resumed during pregnancy, TNFi was able to control the disease. At preconception counselling, the continuation of TNFi during pregnancy should be considered to ensure a better control of disease. References: [1]Fischer-Betz R et al.Arthritis Rheumatol. 2015; 67. Disclosure of Interests: : None declared

OBJECTIVES:We aimed to evaluate the impact of biologic therapy on work productivity outcomes in an Italian real-life cohort of biologic-naïve patients with active rheumatoid arthritis (RA).METHODS:This observational prospective multicentre study enrolled RA patients in working age with an active disease who started their first biologic agent. Every patient completed the RA-specific Work Productivity Survey (WPS-RA) at each clinical evaluation (baseline, 6 and 12 months). The primary outcome of the study was the productivity gain at 12 months from the beginning of the biologic treatment, compared to baseline, assessed in terms of absenteeism and presenteeism reduction, both for employed and unemployed subjects. Linear regression analyses were performed to assess the impact of patient- and disease-related variables on productivity gain.RESULTS:Overall, 100 patients were enrolled and 85 completed the study. All indexes of disease activity and functional ability were significantly improved from baseline already at 6 months. At 12 months, the 55 employed subjects showed a significant reduction in the mean number of days of work missed (absenteeism) and of reduced productivity (presenteeism). A significant reduction in the mean number of days of household work missed was observed for all patients. At multivariate analysis, functional disability had a significant negative impact on all parameters of household work productivity, while the achievement of a low disease activity or remission was inversely correlated with presenteeism.CONCLUSIONS:One year of treatment with a biological drug significantly impacts on the disease activity and work ability of RA patients and allows economic gains due to productivity improvement.

OBJECTIVES:Baricitinib, an oral Janus kinase (JAK) 1-2 inhibitor, is currently used along biologic DMARDs (bDMARDs) after the failure of methotrexate (MTX) in rheumatoid arthritis (RA). We investigated the efficacy and safety of baricitinib in real life. METHODS:We prospectively enrolled 446 RA patients treated with baricitinib from 11 Italian centres. Patients were evaluated at baseline and after 3, 6, and 12 months. They were arrayed based on previous treatments as bDMARD-naïve and bDMARD-insufficient responders (IR) after the failure or intolerance to bDMARDs. A sub-analysis differentiated the effects of methotrexate (MTX) and the use of oral glucocorticoids (OGC). RESULTS:Our cohort included 150 (34%) bDMARD-naïve and 296 (66%) bDMARD-IR patients, with 217 (49%) using baricitinib as monotherapy. Considering DAS-28-CRP as the primary outcome, at 3 and 6 months, 114/314 (36%) and 149/289 (51.6%) patients achieved remission, while those in low disease activity (LDA) were 62/314 (20%) and 46/289 (15.9%), respectively; finally at 12 months 81/126 (64%) were in remission and 21/126 (17%) in LDA. At all-timepoints up to 12 months, bDMARDs-naïve patients demonstrated a better clinical response, independently of MTX. A significant reduction in the OGC dose was observed at 3 and 12 months in all groups. The serum positivity for both rheumatoid factors (RF) and anti-citrullinated protein antibodies (ACPA) conferred a lower risk of stopping baricitinib due to inefficacy. Fifty-eight (13%) patients discontinued baricitinib due to adverse events, including thrombotic events and herpes zoster reactivation. CONCLUSIONS:Real-life data confirm the efficacy and safety profiles of baricitinib in patients with RA and provide evidence that drug survival is higher in bDMARDs-naïve and seropositive patients.

Background:Pregnant patients (pts) with spondyloarthritis (SpA) seem at increased risk for adverse pregnancy outcomes (APO), however limited and conflicting data have been published so far and risk factors for APO in these pts remain poorly understood.Objectives:To assess APO and identify possible risk factors for those in a cohort of SpA pregnant pts.Methods:Data on SpA pts prospectively-followed in a pregnancy clinic from 2010 to 2019 were retrospectively analysed before conception and during each trimester. Pregnancies complicated by APO were compared with those that were uneventful for demographic and clinical variables. Active disease was defined as a DAS-28-CRP>3.2 or an ASDAS-CRP ≥ 2.1 according to peripheral or axial dominant disease respectively.Results:56 pregnancies (mean age 34±5 years; median disease duration 60 months, IQR 24-123) in 47 pts were analysed: 37 psoriatic arthritis, 7 axial SpA, 6 undifferentiated SpA, 3 enteropathic SpA, 2 reactive arthritis and 1 enthesitis-related juvenile idiopathic arthritis. APO were recorded in 23/56 (41%) pregnancies: 5 (9%) early miscarriages, 1 (2%) medical abortion (central nervous system malformation), 3 (5%) preterm births (≥34 gestational week, all for preterm premature rupture of membranes - PROM); 2 (4%) PROM; 7 (13%) small for gestational age newborns (SGA); 3 gestational diabetes and 2 cholestasis of pregnancy. Table 1 displays the comparison between pregnancies with and without APO. A higher number of pts with active disease were detected during the 2ndtrimester in both groups, however differences between those were only significant at the 3rdtrimester (p=0.03). History of inflammatory bowel symptoms (IBS) was also associated with an increased risk for APO (p=0.02). Although not reaching statistical significance, APO occurred more frequently in pts with a previous use of > 1 conventional synthetic (cs) or biological (b) disease-modifying antirheumatic drug (DMARD) (p=0.05), suggesting a more difficult to treat phenotype. Likewise, pts with APO were less often treated with low dose aspirin (LDA) during pregnancy.Conclusion:SGA was the main APO recorded. History of IBS, a more difficult to treat phenotype and the presence of active disease during pregnancy influenced APO in this cohort, reinforcing the need for tight disease control before and during pregnancy. Larger and prospective data are warranted to confirm these results and to assess the potential protective role of LDA.References:[1]Molto 2018; Zbinden 2018.Table 1.Differences between patients with and without APOPregnancieswith APOPregnancies without APOpN, %23, 41.133, 58.9Maternal age at conception,mean±SD (years)33.8±533.8±4.9NSHx previous APO– N, %4,17.43, 9.1NSDisease duration,median (IQR), (months)96 (36-132)48 (24-96)NSAxial dominant disease– N, %6, 26.16, 18.2NSPeripheral dominant disease –N, %11, 47.820, 60.6NSHx enthesitis– N, %9, 39.113, 39.4NSHx dactylitis– N, %5, 21.711, 33.3NSHx psoriasis– N, %10, 43.518, 54.5NSHx uveitis–N, %1, 4.34, 12.1NSHx inflammatory bowel symptoms– N, %4, 17.40, 00.02HLA-B27– N, %8, 34.87, 21.2NS>1 cs or bDMARDs before conception –N, %13, 56.510, 30.30.05LDA during pregnancy– N, %7, 30.417, 51.5NSActive disease before conception- N, %5, 21.72, 6.1NSActive disease 1sttrimester- N, %1, 4.34, 12.1NSActive disease 2ndtrimester- N, %6, 26.15, 15.2NSActive disease 3rdtrimester- N, %3, 130, 00.03Legend: Hx – history of; NS – non significant.Disclosure of Interests:None declared

Background: SLE pregnancies have an increased risk of Adverse Pregnancy Outcomes (APO). In clinical practice, low C3 and C4 levels are associated with active disease and, during pregnancy, complement activation products are shown to be associated with APO. Objectives: To analyse potential association between C3 and C4 variations and disease flares and APO during SLE pregnancies. Methods: Demographic, clinical and laboratory data on SLE pregnancies prospectively-followed by a multidisciplinary team in a pregnancy clinic from 1987 to 2015 were retrospectively analysed at preconception and at each trimester. Hypocomplementemia was defined according to the normality range calculated in healthy pregnancies by Reggia et al 1 . APO were defined as: early miscarriage (<10 th week), intrauterine fetal death (>10 th week), perinatal death (<30 th day of life), pre-eclampsia (PE), severe preterm birth (<34 th week). Results: 134 pregnancies in 98 SLE patients were analysed. APO occurred in 22 (16%) pregnancies: 9 early miscarriages, 4 intrauterine fetal deaths, 3 severe preterm births, 6 PE (hesitated in 1 intrauterine fetal death, 1 perinatal death; 2 preterm birth between 34 th and 37 th weeks and 2 term births). 13 flares (2 renal, 4 articular, 6 cutaneous and 1 neurological) were recorded in 11 (8%) pregnancies. The mean C3 and C4 levels at each trimester are shown in table 1. Table 1. C3 and C4 mean levels (mg/dL) at pre-conceptional visit (T0), 1 st trimester (T1), 2 nd trimester (T2) and 3 rd trimester (T3). C3 T0 C3 T1 C3 T2 C3 T3 p T0-T1 p T1-T2 p T2-T3 Pregnancies with flares 73.2 82.2 83.8 97.5 0.04 0.02 0.06 Pregnancies without flares 85.3 91.8 104.4 114.7 <0.001 <0.001 <0.001 Pregnancies with APO* 84.7 91.5 98.5 106.3 0.03 0.03 0.89 Pregnancies without APO* 84.5 90.9 102.8 112.8 <0.001 <0.001 <0.001 C4 T0 C4 T1 C4 T2 C4 T3 p T0-T1 p T1-T2 p T2-T3 Pregnancies with flares 8.6 11.8 10.9 11.4 0.01 0.43 0.27 Pregnancies without flares 13.2 14.7 16.3 16.6 <0.001 <0.001 0.98 Pregnancies with APO** 13.1 16.1 15.9 14.7 0.01 0.3 0.13 Pregnancies without APO** 12.9 14.2 15.8 16.1 <0.001 <0.001 0.42 Comparison of C3 and C4 mean levels between pregnancies with APO vs without APO: * T0, T1, T2, T3: ns; ** T0, T1, T2, T3: ns Both in pregnancies with flares and with APO, there was no increase of C3 between the 2 nd and the 3 rd trimester and of C4 between the 1 st and the 2 nd trimester. At preconception, mean levels of C4 were lower in pregnancies with flares compared to those without flares (images 1 and 2); during the 2 nd and the 3 rd trimesters the mean levels of both C3 and C4 were lower in pregnancies with flares. In pregnancies with APO, the variation of C4 levels between the 2 nd and the 3 rd trimester was lower than in pregnancies without APO (-3.18 vs 0.27; p=0.01). A higher frequency of low C4 was observed at pre-conceptional visit, 1 st trimester and 3 rd trimester (6/7 vs 25/103 p=0.002; 8/9 vs 56/106 p=0.04; 9/11 vs 33/96 p=0.003) in pregnancies with flare as compared with pregnancies without flares. Figure 1. Image 1: comparison of C3 mean levels between pregnancies with fares vs without flares Figure 2. Image 2: comparison of C4 mean levels between pregnancies with fares vs without flares Conclusion: In our cohort of prospectively-followed SLE pregnancies, low C4 levels at preconception seems to predict flares during pregnancy. Low increase of C4 levels between the 2 nd and the 3 rd trimester could predict an APO. References: [1]Reggia R. et al. Rheumatology 2012;51:2186-90 Disclosure of Interests: None declared