Oral abstracts

OC.01

THE INCIDENCE AND RISK FACTORS FOR NEW ONSET OF INTERSTITIAL LUNG DISEASE IN SYSTEMIC SCLEROSIS: A SENSITIVITY ANALYSIS FROM EUSTAR COHORT STUDY

Liubov Petelytska¹, Arthiha Velauthapillai³, Lorenzo Tofani⁴, Eric Hachulla⁵, Ulf Müller-Ladner⁶, Elise Siegert⁷, Yannick Allanore⁸, Gabriela Riemekasten⁹, Christina Bergmann¹⁰, Radim Becvar¹¹, Kamal Solanki¹², Branimir Anic¹³, Simona Rednic ¹⁴, Bojana Stamenkovic¹⁵, Lisa Stamp ¹⁶, Madelon C Vonk³, Jeska De Vries-Bouwstra¹⁷, Anna -Maria Hoffmann-Vold1^1,18, Marco Matucci-Cerinic^19,20, Oliver Distler¹, Cosimo Bruni^1,19

¹University Hospital Zurich, University of Zurich, Department of Rheumatology, Zurich, SWITZERLAND, ²Bogomolets National Medical University, Dept Internal Medicine #3, Kyiv, UKRAINE, ³Radboud University Medical Center, Department of Rheumatology, Nijmegen, THE NETHERLANDS, ⁴University of Florence, Department of Statistics, Computer Science, Applications, Florence, ITALY, ⁵Huriez Hospital, University of Lille, Department of Internal Medicine and Clinical immunology, Lille, FRANCE, ⁶Justus Liebig University Giessen, Campus Kerckhoff, Department of Rheumatology and Clinical Immunology Bad Nauheim,GERMANY, ⁷Charité University Hospital, Department of Rheumatology, Berlin, GERMANY, ⁸Cochin Hospital, Rheumatology A dpt, Paris 5 University, Paris, FRANCE, ⁹University Clinic Schleswig-Holstein, University of Lübeck,Department of Rheumatology and Clinical Immunology, Lübeck, GERMANY, ¹⁰Friedrich-Alexander-University Erlangen-Nürnberg (FAU) and University Hospital Erlangen, Department of Internal Medicine, Erlangen, GERMANY, ¹¹Institute Of Rheumatology And Department Of Rheumatology 1st Medical School, Prague, CZECH REPUBLIC, ¹²Waikato Hospital, Rheumatology Unit, Hamilton City, NEW ZEALAND, ¹³University of Zagreb, School of Medicine, University Hospital Center Zagreb, Division of Clinical Immunology and Rheuma, Zagreb, CROATIA, ¹⁴Clinica Reumatologie, University of Medicine & Pharmacy Iuliu Hatieganu Cluj, Cluj-Napoca, ROMANIA, ¹⁵Institute For Prevention, Treatment And Rehabilitation Rheumatic And Cardiovascular Disease, Niska Banja, SERBIA, ¹⁶University of Otago, Dept. of Medicine, Christchurch, NEW ZEALAND, ¹⁷Leiden University Medical Center, Department of Rheumatology, Leiden, THE NETHERLANDS, ¹⁸Oslo University Hospital, Rikshospitalet, Department of Rheumatology, Oslo, NORWAY, ¹⁹University of Florence - University Hospital Careggi, Dept Experimental and Clinical Medicine, Division of Rheumatology, Florence, ITALY, ²⁰San Raffaele Hospital, Unit of Immunology, Rheumatology, Allergy and Rare Diseases (UniRAR), Milano, ITALY

Introduction: Our previous post-hoc analysis of the EUSTAR database estimated the overall SSc-ILD incidence rate of 3.7 cases per 100 person-years, with an overall incidence of 20.2% during average 5.1±4.4 years follow-up. Given the heterogeneity between patients on disease duration, number of visits and follow-up duration, we performed a sensitivity analysis restricted to (1) patients who had at least 3 visits with available high-resolution computed tomography (HRCT) within 5 years from baseline and (2) patients with disease duration less 5 years.

Material and Methods: SSc patients classified according to the 2013 ACR/EULAR criteria from the EUSTAR database, with absence of ILD on HRCTs at baseline and having at least 3 visits with available HRCTs within 5 years were included. Patients with pulmonary arterial hypertension diagnosed by right heart catheterization were excluded. Based on follow-up HRCTs, our population was divided into patients with new onset of ILD (incident group) and patients who remained ILD negative (negative group). Incidence of SSc-ILD was calculated as a rate per 100 person-years and presented starting from the first visit to the center.

Results: The baseline characteristics of the ILD-incident and ILD-negative groups are shown in Table 1. In addition to disease duration, the two sub-cohorts differed for prevalence of smoking exposure, peripheral vascular complications and arthritis.

For the first sensitivity analysis focusing on patients with at least 3 visits with available HRCTs within 5 years from baseline, we identified 3358 SSc patients, among whom 631 (18.8%) developed ILD during an average 3.9±1.6 years follow-up. There was a continuous detection of new onset of ILD up to 5 years from baseline (incidence rate 3.6 ranging from 2.7-5.8 per 100 person-years; Figure 1).

In the second sensitivity analysis focusing on disease duration less 5 years, 576/2674 (21.5%), early patients developed ILD with an average of 5.0±4.2 years. Also in this population, ILD onset was detected regularly during the 5 years follow-up, with an incidence rate 4.4 cases per 100 person-years, ranging from 2.7 to 5.8. Interestingly, comparable incidence rates were detected for SSc patients with disease duration >5 years (Figure 2).

Conclusions: Both sensitivity analyses confirmed that ILD can appear at any time after SSc diagnosis, with regular incidence over time during the disease course and not influenced by an earlier disease duration or a tighter follow-up. This supports the need for continuous screening over time for SSc-ILD, to achieve early diagnosis and facilitate immediate treatment.

OC.02

PERIPHERAL BLOOD GENE EXPRESSION PROFILING SHOWS PREDICTIVE SIGNIFICANCE FOR THE COURSE OF INTERSTITIAL LUNG DISEASE IN SYSTEMIC SCLEROSIS

Shervin Assassi¹, Christopher Denton², Ramona Schmid³, Carina Ittrich³, Tobias Litzenburger³, Sudha Visvanathan⁴

¹The University of Texas Health Science Center at Houston, Houston, USA, ²University College London Division of Medicine, London, UNITED KINGDOM, ³Boehringer Ingelheim Pharma GmbH & Co. KG, Biberach, GERMANY, ⁴Boehringer Ingelheim Pharmaceuticals Inc., Ridgefield, USA

Introduction: Peripheral blood cell (PBC) RNA is an easily accessible sample type for biomarker development and clinical use. In Scleroderma Lung Study II, patients with SSc-ILD who had higher baseline lymphoid modules (sets of transcripts linked to specific pathways) had a better response to mycophenolate mofetil (MMF), while those with a higher myeloid cell lineage activation score were less likely to respond (Assassi et al. Ann Rheum Dis 2022). However, all patients in this trial were treated with immunosuppressive agents and it was unclear whether the observed trends were due to the natural history of disease or a response to MMF. In the placebo arm of the SENSCIS trial in patients with SSc-ILD, approximately half of the patients were on a stable dose of MMF. We aimed to determine the predictive significance of baseline PBC transcript scores in patients with and without MMF background treatment in the SENSCIS trial.

Material and Methods: Patients with a baseline PBC RNA sample (stored in PAXgene tubes) enrolled in the placebo arm of the SENSCIS trial were included. These patients were divided into those on stable MMF background treatment (n=120) and no MMF background treatment (n=118, of whom 110 were on no immunosuppressive agents and 8 were on methotrexate). Global RNA sequencing was performed and a modular analysis using 62 curated whole blood modules was pursued as previously described (Chaussabel et al. Nat Rev Immunol 2014). The predictive significance of baseline composite modular scores for decline in forced vital capacity (FVC) at week 52 was investigated.

Results: Consistent with the findings of Scleroderma Lung Study II, among patients on stable MMF background treatment, higher baseline Lymphoid Lineage, and Mitochondrial/Protein Synthesis modules predicted a better FVC course, while higher baseline Myeloid Lineage and Inflammation modules predicted a higher rate of FVC decline (Figure 1A). Among patients without MMF background treatment, only Myeloid Lineage and Inflammation modules predicted a higher rate of decline in FVC; other modules were not predictive (Figure 1B).

Conclusions: In the largest study of PBC gene expression study in patients with SSc-ILD presented to date, modules involved in myeloid lineage predicted a greater rate of FVC decline regardless of MMF treatment status. Consistent with the primary action of MMF on lymphocytes, higher baseline lymphoid modules predicted a better ILD course only in patients on MMF. Blood gene expression profiling might be useful as a prognostic/predictive biomarker in SSc-ILD.

OC.03

IDENTIFICATION OF SERUM PROTEIN BIOMARKERS ASSOCIATED WITH CARDIAC MAGNETIC RESONANCE IMAGING-DEFINED SUBCLINICAL CARDIOVASCULAR ABNORMALITIES IN SYSTEMIC SCLEROSIS PATIENTS CHARACTERISE INFLAMMATION

Wang Chun Cheung¹, Rudresh Shukla^2,3, Raluca Bianca Dumitru⁴, Nicholas Black^2,5, Christopher Miller^3,5,7, Francesco Del Galdo^4,6, Sven Plein⁸, Darren Plant^2,3, Maya H. Buch^2,3,4

¹University of Manchester - School of Biological Sciences, Manchester, UNITED KINGDOM, ²University of Manchester - Centre for Musculoskeletal Health, Manchester, UNITED KINGDOM, ³Manchester University NHS Foundation Trust - NIHR Manchester Biomedical Research Centre, Manchester, UNITED KINGDOM, ⁴University of Leeds - Leeds Institute of Rheumatic & Musculoskeletal Medicine, Leeds, UNITED KINGDOM, ⁵University of Manchester - Division of Cardiovascular Sciences, Manchester, UNITED KINGDOM, ⁶University of Manchester - Wellcome Centre for Cell Matrix Research, Manchester, UNITED KINGDOM, ⁷Leeds Teaching Hospitals NHS Trust - NIHR Leeds Biomedical Research Centre, Leeds, UNITED KINGDOM, ⁸University of Leeds - Multidisciplinary Cardiovascular Research Centre (MCRC) and Leeds Institute of Cardiovascular a, Leeds, UNITED KINGDOM

Introduction: Systemic sclerosis-primary heart involvement (SSc-pHI) accounts for up to one-third of SSc-related deaths and clinically apparent pHI portends poor outcome. Early detection and identification of therapeutic targets is needed to improve the management of SSc-pHI. We have previously shown cardiovascular magnetic resonance (CMR)-detected subclinical abnormalities. In this study, we aimed to identify protein biomarkers associated with subclinical CMR-measured abnormalities in SSc patients, and the predominant inflammatory/cardiometabolic pathways implicated.

Material and Methods: Serum from 78 patients from CONVAS (‘CONnective Tissue Disease and VASculitis Cohort’) and ELCASA (‘ELectrophysiology and CArdiac imaging in SclerodermA’) cohorts with no history of pHI, pulmonary hypertension, diabetes or more than 2 traditional cardiovascular (CV) risk factors was used to measure levels of 355 proteins across inflammation, cardiometabolic and cardiovascular II/III Olink panels. Generalised linear regression (corrected for multiple testing) was used to identify significant proteins associated with CMR measures of myocardial oedema/fibrosis (MO/MF) [native T1, myocardial extracellular volume (ECV) and late gadolinium enhancement (LGE)] and vascular stiffness (VS) [aortic distensibility]. Subsequently, an expanded physical protein-protein interaction (PPI) network was created (String-DB) with k-means clustering applied to identify enriched clusters and functional analysis performed (GeneOntology and KEGG).

Results: Seventy out of 355 proteins were associated with MO/MF (64 proteins; 26 positively, 38 negatively) and VS (6 proteins; 2 positively, 4 negatively). Two overlapping proteins associated with focal (LGE) and diffuse (ECV) fibrosis were identified.

Proteins identified include those involved in coagulation cascade (estimate = -1.97, 95%CI = -3.63 to -0.31, adj.p = 0.039), carbohydrate binding and opsonisation activities (estimate = -1.31, 95%CI = -2.19 to -0.43, adj.p = 0.006); and cancer and neovascular inflammatory conditions (estimate = -75.79 95%CI = -119.81 to -31.76, adj.p = 0.002).

K-means clustering (enrichment p.value < 1.0e-16) identified 4 clusters (Figure 1) with roles in TNF receptor superfamily binding and NF-kappa B signalling pathway (red); Vascular endothelial growth factor-activated receptor activity and Rap1 signalling pathway (yellow); IL-10 receptor activity and JAK-STAT signalling pathway (blue); and FGF activated class receptor binding and ErbB signalling pathway (green). Proteins associated with CMR detected subclinical myocardial tissue oedema/fibrosis mainly mapped to yellow and blue clusters.

Conclusions: In this first proteomic study of subclinical SSc-pHI, we have identified 70 protein biomarkers, with MO/MF associated with inflammation and vascular pathways. The next step is to validate these proteins and test the utility of network proteins in an independent patient cohort that could aid detection and diagnosis of SSc-pHI.

OC.04

UNRAVELING CARDIAC FIBROSIS THROUGH MULTI-OMICS ANALYSIS: DYSF, MXRA5, AND FOXF1 AS PROMISING TARGETS

Ievgeniia Kocherova¹, Elena Pachera¹, Daria Nurzynska², Franca Di Meglio², Oliver Distler¹, Przemyslaw Blyszczuk¹, Gabriela Kania¹

¹Center of Experimental Rheumatology, Department of Rheumatology, University Hospital Zurich, University of Zurich, Schlieren, SWITZERLAND, ²Department of Public Health, University of Naples Federico II, Naples, ITALY

Introduction: Inflammatory dilated cardiomyopathy (iDCM), a frequent finding in systemic sclerosis, is accompanied by progressive myocardial fibrosis and ultimately leads to heart failure (HF).

Fibroblast activation, driven by the activator protein 1 family member Fos-related antigen-2 (FOSL-2), is a critical step in cardiac fibrogenesis. Current treatments demonstrate limited efficacy against myocardial fibrosis. We aimed to identify new candidate targets implicated in cardiac fibrogenesis under immunofibrotic conditions.

Material and Methods: Cardiac fibroblasts were isolated from the left atria of patients (n=5) undergoing heart transplantation due to HF associated with iDCM and from unaffected hearts of brain-dead donors (Ctrl, n=5). Protein quantification was performed using liquid chromatography tandem-mass spectrometry (LC–MS/MS). The data were processed with MaxQuant v1.6.2.3 software. Bulk RNA sequencing was conducted using the Illumina HiSeq platform. Differentially expressed genes were identified using DESeq2. Publicly available single-cell RNA sequencing datasets (GSE109816, GSE121893) on adult hearts from HF patients (n=6) and Ctrl (n=14) were analysed using Seurat package (V.2.3.4). Selected gene knockdown was achieved by siRNA transfection of human foetal cardiac fibroblasts (HCFs), untreated or stimulated with TGF-β for 48-72h. Profibrotic marker expression was determined using RT-qPCR and Western Blot. Cell viability and ATP production were assessed using PrestoBlue HS and CellTiter-Glo assays, respectively.

Results: Our multi-omics analysis identified membrane protein dysferlin (DYSF), matrix remodeling-associated protein MXRA5 and transcription factor FOXF1 as candidate targets significantly upregulated in HF. Further in vitro studies on HCFs (n=4, p<0.05) revealed that TGF-β upregulated DYSF and MXRA5 while downregulating FOXF1. DYSF silencing (n=4) upregulated MXRA5 after 48h of TGF-β stimulation (p<0.05), downregulated ACTA2 (48h and 72h of TGF-β stimulation, p<0.05), and upregulated FOSL-2 protein levels in untreated HCFs and 72h after TGF-β stimulation (n=3, p<0.05). MXRA5 knockdown (n=8) resulted in the upregulation of DYSF (p<0.05), ACTA2 (p<0.05) and COL1A1 (p<0.001) in untreated HCFs and upregulated DYSF (p<0.01) and COL1A1 (p<0.05) after 48h of TGF-β stimulation. FOXF1 silencing in HCFs (n=8) followed by 48h of TGF-β stimulation downregulated MXRA5 (p<0.01) while upregulating DYSF (p<0.001) and COL1A1 (p=0.05). Knockdown of candidate targets (n=4) reduced cell viability in untreated (DYSF: p<0.01, MXRA5: p<0.001, FOXF1: p<0.01) and TGF-β-stimulated (DYSF: p<0.05, MXRA5: p<0.05, FOXF1: p<0.01) HCFs. ATP levels were decreased in TGF-β-stimulated HCFs after DYSF silencing (n=6, p=0.05).

Conclusions: In summary, we identified DYSF, MXRA5 and FOXF1 as potential therapeutic targets implicated in myocardial fibrogenesis, including the regulation of profibrotic transcription factor FOSL-2.

OC.05

THE EVALUATION OF LOWER EXTREMITIES ARTERIAL DISEASE IN SYSTEMIC SCLEROSIS PATIENTS THROUGH RHEUMATOLOGIST-VASCULAR SURGEON MULTIDISCIPLINARY MANAGEMENT: PRELIMINARY DATA FROM AN ITALIAN SINGLE CENTER

Maria Grazia Lazzaroni¹, Liala Moschetti¹, Eleonora Pedretti¹, Paolo Baggi², Stefano Bonardelli², Franco Franceschini¹, Paolo Air¹

¹Scleroderma Unit, Rheumatology and Clinical Immunology, ASST Spedali Civili of Brescia, University of Brescia, Brescia, ITALY, ²Vascular Surgery Unit, ASST Spedali Civili of Brescia, University of Brescia, Brescia, ITALY

Introduction: Microvascular changes represent a key step of pathogenic process in Systemic Sclerosis (SSc). However, macrovascular complications have also been reported, especially configuring lower extremities arterial disease (LEAD). Possible risk factors were identified as digital ulcers (DUs), pulmonary arterial hypertension (PAH), corticosteroids therapy and smoking (active/stopped <3 years).

Our aim was the estimation of LEAD prevalence in: (i) a single-centre prevalent SSc cohort; (ii) at-risk patients.

Material and Methods: Consecutive SSc patients attending our centre for scheduled visits (October 2022-July 2023) were included. Patients without known LEAD with >=1 risk factor, as described above, were referred for lower extremities arterial Doppler ultrasonography (DUS) performed by an expert vascular surgeon.

Results: Among 242 SSc patients evaluated (93% females, 80% lcSSc, 46% ACA+), 19 (8%) were previously diagnosed with LEAD, requiring revascularization interventions in 8 (42%) and amputations in 5 (26%). As compared to the remaining 223, they were older (median[IQR1-3]: 75[70-81] vs 62[52-72] years; p<0.001), had longer disease duration (18[13-26] vs 12[6-19] years; p:0.003), higher frequency of DUs (89% vs 44%; p:0.001), and lower limbs ulcers(68% vs 8%; p<0.001), more comorbidities (Charlson Comorbidity Index: 6[5-7] vs 3[2-5]; p<0.001), and were more frequently treated with prostanoids (68% vs 29%; p:0.001) and steroids (58% vs 30%; p:0.013).

At least 1 LEAD risk factor was identified in 150/223 patients (67%) (DUs (65%); steroid use (40%); smoking (28%); PAH (9%)). Among patients without known LEAD: 53/150 received DUS, with LEAD detected in 17 (32%) (Table 1), classified as 15 stage I (asymptomatic), 1 stage II (claudication) and 1 stage III (chronic permanent ischemia without trophic disorder)(Leriche-Fontaine classification). Anti-platelet/anti-coagulant therapy was already ongoing in 12/17, while in the remaining 5 patients low dose aspirin was added; the patient with stage III was referred to angiographic study. As compared to 36 patients without LEAD, no differences were observed in clinical-demographic features, including traditional cardiovascular risk factors (Table 2).

Conclusions: Based on this preliminary data, the prevalence of diagnosed LEAD in our SSc cohort is 8%. However, screening of at-risk patients (DUs, PAH, steroid use and smoking), showed DUS LEAD signs in 32%, mostly asymptomatic and without higher frequency of traditional risk factors. Therefore, lower extremities arterial DUS is a non-invasive helpful tool to improve SSc diagnostic/therapeutic work-up, in which vascular surgeon should become active part of the dedicated multidisciplinary team.

OC.06

ARTIFICIAL INTELLIGENCE BASED RAYNAUD’S QUANTIFICATION INDEX (ARTIX ©): AN OBJECTIVE MOBILE-BASED TOOL FOR PATIENT-CENTRED ASSESSMENT OF RAYNAUD’S SEVERITY

Marco Di Battista¹, Anna Howard¹, Andrea Rindone¹, Stefano Di Donato¹, Francesca Donadoni², Francesco Del Galdo¹

¹Leeds Institute of Rheumatic and Musculoskeletal Medicine - University of Leeds, Leeds, UNITED KINGDOM, ²Procedure Health Limited, London, UNITED KINGDOM

Introduction: Raynaud’s phenomenon (RP) is a clinical hallmark of systemic sclerosis (SSc) and represents a pivotal sign for the diagnosis of very early forms of the disease (VEDOSS). Patient-centred assessment of RP severity is crucial given its recurrent nature, but objective tools to aid in the quantification of overall disease burden are currently lacking. The global uptake of mobile phone technology is already supporting the diagnosis of RP through patient collected images of the attacks. Here we aimed to develop an artificial intelligence (AI) algorithm for RP detection and quantification, as to inform a patient-centred, image-based quantification of RP burden.

Material and Methods: Consecutive patients attending the Raynaud’s and Scleroderma Clinic were enrolled in this pilot study, along with a group of healthy controls (HC). RP was induced by cold challenge, as previously described. Hands were imaged with a standard photo and a thermal image (FLIR, UK) at baseline, right after cold challenge and then every 2 min up to 10 min. The algorithm was first trained to detect hand contours and precisely segmentate each finger on standard photo. Images were then decomposed in the distinct color channels and processed to identify the signal which best reflected the thermal changes. The resulting signal from single digits was finally combined in a single score (ARTIX) and analysed across groups and over time. All computational tasks related to image processing were executed using Python version 3.7. R software was employed for statistical analysis.

Results: 35 patients with RP (15 SSc and 20 VEDOSS, 88% female, mean age 51 ±17 years) were enrolled along with 15 matched HC. Cold challenge induced the already described classic changes in thermal images over time and the AI algorithm was able to detect such changes in the standard photos. Most interestingly, the combined index ARTIX was lower in RP patients vs HC with an overall 94.2% sensitivity and 53.3% specificity of identifying cold challenge induced RP (AUC= 0.754). Furthermore, ARTIX was significantly worse in SSc patients with previous digital ulcers (p=0.03).

Conclusions: AI based quantification of hands images may offer a quantitative tool for assessment of RP severity. Validation in a real-life scenario in the context of an observational cohort is ongoing to determine the correlation of ARTIX with patient-reported outcome of RP and severity over time.

OC.07

PLASMA DEPHOSPHORYLATED-UNCARBOXYLATED MATRIX GLA-PROTEIN IN SYSTEMIC SCLEROSIS PATIENTS: BIOMARKER POTENTIAL FOR VASCULAR CALCIFICATION AND INFLAMMATION

Judith Potjewijd Potjewijd¹, Rachid Tobal¹, Jan GMC Damoiseaux², Leon J Schurgers³, Pieter van Paassen¹

¹Department of Internal Medicine, division Clinical and Experimental Immunology, Maastricht University Medical Center, Maastricht, THE NETHERLANDS, ²Central Diagnostic Laboratory, Maastricht University Medical Center, Maastricht, THE NETHERLANDS, ³Department of Biochemistry, Cardiovascular Research Institute Maastricht, Maastricht University, Maastricht, THE NETHERLANDS

Introduction: Systemic sclerosis (SSc) patients have an increased risk of cardiovascular disease (CVD), which still persists after adjustment for classical cardiovascular risk factors. Plasma dephosphorylated-uncarboxylated Matrix Gla-protein (dp-ucMGP) is the inactive form of MGP which correlates with increased risk of CVD. It is mainly secreted by smooth muscle cells, which play a role in the pathogenesis of SSc. In this study, dp-ucMGP levels and incidence of first-ever CVD were determined in early-diagnosed SSc patients. We investigated if dp-ucMGP as a biomarker predicts CVD and/or all-cause mortality during long-term follow-up.

Material and Methods: We conducted a cohort study among 87 SSc patients. Patients with pre-existing CVD or dependent on dialysis were excluded. Circulating plasma dp-ucMGP levels were determined at baseline and CVD and cardiovascular risk factors were retrieved from medical records. For validation, we determine dp-ucMGP in a subset of our prospective cohort of treatment-naive SSc patients.

Results: The median dp-ucMGP level was 634 pmol/L (IQR 301), which is greatly increased compared with age-matched healthy controls (dp-ucMGP<393 pmol/L;p<0.001). Validation of elevated dp-ucMGP levels in the prospective cohort of SSc patients (n=29) was achieved (median 589 pmol/L (IQR 370)). Only one patient had prevalent CVD before onset SSc. Nine patients were lost to follow-up, in the remaining 78 SSc patients, 26 (33.3%) patients suffered from first CVD event during the study period, with a median time of 10.5 (IQR 15.2) years from onset SSc disease, corresponding to an incidence rate of 29.9 per 1000 person-years. Cardiovascular risk factors were not significantly different between patients with and without CVD. Odds ratios for sex (OR 2.44; 95%CI 0.89-6.72) and hypertension (OR 2.53; 95%CI 0.86-7.46) tended to predict CVD, but the association did not reach statistical significance. However, Kaplan-Meier analysis showed that elevated dp-ucMGP levels (>634 pmol/L) were associated with an increased risk for CVD and/or death during the first 10 years follow-up (log-rank test: P=0.006).

Conclusions: This study shows increased dp-ucMGP levels in SSc patients compared to age-matched HCs, indicating dp-ucMGP as a biomarker of disease. We confirm the high risk of CVD in SSc patients but traditional cardiovascular risk factors did not predict development of CVD. In contrast, high dp-ucMGP levels revealed an increased risk for CVD and/or death in SSc, suggesting dp-ucMGP as a new inflammatory biomarker of disease.

OC.08

DISTINCT CLINICAL TRAJECTORIES OF GASTROINTESTINAL PROGRESSION AMONG PATIENTS WITH SYSTEMIC SCLEROSIS

Jamie Perin¹, Michael Hughes², Fredrick Wigley³, Christopher Mecoli³, Julie Paik³, Allan Gelber³, Laura Hummers³, Ami Shah³, Scott Zeger⁴, Zsuzsanna Mcmahan³

¹Department of International Health, Johns Hopkins Bloomberg School of Public Health, Baltimore, USA, ²Department of Rheumatology, Northern Care Alliance NHS Foundation Trust, Salford, UNITED KINGDOM, ³Division of Rheumatology, Johns Hopkins University School of Medicine, Baltimore, USA, ⁴Department of Biostatistics, Johns Hopkins Bloomberg School of Public Health, Baltimore, USA

Introduction: Systemic sclerosis (SSc) is very heterogenous in its clinical presentation. Common manifestations cluster together defining subgroups of disease. The goal of this investigation is to characterize and define gastrointestinal (GI) phenotypes.

Material and Methods: We examine a well-established SSc patient cohort, with a modified Medsger GI severity score measured over time, to determine heterogeneity in disease progression. Growth mixture models are used to estimate the phenotype for each patient as well as the trajectory of disease severity over time. We compare the characteristics of estimated phenotypes using non-parametric statistics, and linear and logistic regression to compare patient characteristics between phenotypes while adjusting for disease duration.

Results: We examined GI organ system involvement from 2696 patients with SSc with at least two Medsger GI scores, identifying four unique phenotypes. The most common phenotype (n=2325) (Cluster C, “Stable”) had an average score of 1 that was consistent over time. Two phenotypes were progressive [Cluster A, “Progressive/Fibrotic” (n=142) and Cluster D, “Progressive/Vascular” (n=115)] and had an initial average score of 1. The Progressive/Fibrotic group increased initially and stabilized, and the Progressive/Vascular group worsened slowly over time. A fourth phenotype [Cluster B, “Most Severe GI”; (n=114)] had the most severe GI disease initially (an initial average Medsger GI score of just below 3) and had the highest mortality, but GI severity improved over time even for this cluster after accounting for disease duration and survival. Significant differences between phenotypes were identified, including autoantibody and mortality status.

Conclusions: Clinically distinct GI phenotypes exist among patients with SSc. These phenotypes are not only distinguished by GI and extra-intestinal SSc clinical complications, but they are also temporally distinct. Distinct autoantibody profiles associate strongly with more severe GI disease.

OC.09

FECAL MICROBIOTA TRANSPLANTATION IN PATIENTS WITH SYSTEMIC SCLEROSIS AND LOWER GASTROINTESTINAL TRACT SYMPTOMS: THE RESSCUE PHASE 2 RANDOMIZED CLINICAL TRIAL

Håvard Fretheim¹, Imon Barua¹, Maylen Carstens¹, Henriette Didriksen¹, Vikas Sarna³, Knut Ae Lundin⁴, Oliver Distler ⁵, Dinesh Khanna⁶, Elizabeth Volkmann⁷, Øyvind Midtvedt¹, Tore Midtvedt⁸, Alvilde Dhainaut⁹, Anne-Kristine H Halse¹⁰, Gunnstein Bakland¹¹, Inge Olsen¹², Maiju E Pesonen¹², Diana Domanska¹³, Øyvind Molberg¹, Anna -Maria Hoffmann-Vold¹

¹Oslo University Hospital, Department of Rheumatology, Oslo, NORWAY, ²University of Oslo, Institute of Clinical Medicine, Oslo, NORWAY, ³Oslo University Hospital, Ullevål, Department of Gastroenterology, Oslo, NORWAY, ⁴Oslo University Hospital, Rikshopitalet, Department of Gastroenterology, Oslo, NORWAY, ⁵University Hospital Zurich, Department of Rheumatology, Zurich, SWITZERLAND, ⁶University of Michigan, Division of Rheumatology, Ann Arbor, USA, ⁷University of California, Division of Rheumatology, Department of Medicine, Los Angeles, USA, ⁸Karolinska Institutet, Department of Microbiology, Tumor and Cell Biology, Stockholm, SWEDEN, ⁹St. Olavs hospital, Department of Rheumatology, Trondheim, NORWAY, ¹⁰Haukeland University Hospital, Department of Rheumatology, Bergen, NORWAY, ¹¹University Hospital of Northern Norway, Department of Rheumatology, Tromsø, NORWAY, ¹²Oslo University Hospital Research Support Services, Oslo Centre for Biostatistics and Epidemiology (OCBE), Oslo, NORWAY, ¹³Oslo University Hospital, Department of Pathology, Oslo, NORWAY

Introduction: Lower gastrointestinal tract (GIT) complications are common in patients with systemic sclerosis (SSc), and current treatment alternatives are limited. Patients with SSc have an altered intestinal microbiota composition. This provides a rational for the investigation of fecal microbiota transplantation (FMT) in SSc patients with lower GIT symptoms. In this randomized multicenter double-blind clinical trial (RCT) we assessed the safety and efficacy of a standardized intestinal microbiota infusion in SSc patients with lower GIT symptoms.

Material and Methods: Patients with SSc and moderate to severe bloating and/or diarrhea assessed by the UCLA SCTC GIT score 2.0 were enrolled in a multicentre, double-blind, randomized, placebo-controlled phase 2 trial. Patients were randomized to receive FMT with an intestinal infusion of a standardized fecal microbiota culture (ACHIM) or placebo at weeks 0 and 2. At week 12, all patients received a FMT infusion and were followed in an open label phase until week 20. The primary outcome was change between baseline and week 12 in UCLA GIT score item diarrhea or bloating measured as the average marginal effect (AME), depending on which was the worst symptom at baseline; evaluated separately for each patient. Secondary outcomes were safety and tolerability and total UCLA GIT score. Other outcome measures included the change in UCLA GIT score from week 12 to week 20.

Results: A total of 65 patients were randomized to either ACHIM or placebo. No significant baseline differences were found between the placebo and ACHIM groups (Table). From week 0 to week 12, no significant differences in symptom changes were observed between the groups (AME=0.17 (-0.12, 0.47), p=0.25) (Figure 1). Similarly, from weeks 12 to 20, changes were comparable between the groups (AME= -0.04 (-0.32, 0.23), p=077). No significant difference was observed in the total GIT score between the two groups over the same time period (AME=0.09 (-0.04, 0.23), p=0.17). Side effects were reported by 16 (37%) in the ACHIM group, and 19 (42%) in placebo group. Abdominal pain was the most frequent side effect, occurring in 5 (15%) patients in ACHIM group and 2 (6%) patients in the placebo group. Time to resolved pain was 2 days in both groups. One patient underwent an intramural perforation during gastroscopy, requiring IV antibiotics, and made a full recovery.

Conclusions: We found no significant clinical effect of FMT in SSc patients with lower GIT symptoms.

OC.10

DECONVOLUTION OF METABOLIC BIOMARKER SIGNATURES OF VERY EARLY DIAGNOSIS OF SYSTEMIC SCLEROSIS (VEDOSS) AND ESTABLISHED DISEASE: EXPLORATORY METABOLIC BIOMARKERS OF FAST PROGRESSORS

Sunhwa Kim¹, Vishal Kakkar^2,3, Rebecca Ross^2,3, Yingtao Bi⁴, Thierry Sornasse¹, Francesco Del Galdo ^2,3

¹AbbVie Precision Medicine Immunology, South San Francisco, USA, ²Leeds Institute of Rheumatic and Musculoskeletal Medicine, Faculty of Medicine and Health, University of Leeds, Leeds, UNITED KINGDOM, ³NIHR Leeds Biomedical Research Centre, Leeds Teaching Hospitals, NHS Trust, Chapel Allerton Hospital, Leeds, UNITED KINGDOM, ⁴AbbVie Discovery and Exploratory Statistics, Worcester, USA

Introduction: Systemic sclerosis (SSc) is a complex autoimmune disease characterized by its asynchronous impact on multiple organs, resulting in highly variable fibrotic damage and elevated morbidity and mortality rates. Despite advancements in omics research shedding light on potential genes and proteins contributing to disease heterogeneity, effectively sub-stratifying individuals with rapidly progressing disease requiring urgent medical intervention remains a formidable challenge. In this study, we aim to elucidate previously underexplored metabolic serum biomarkers to identify individuals with rapid disease progression.

Material and Methods: We compared the global metabolic profiles measured by high-performance chemical isotope labeling LC-MS (Dansyl-labeling) of the selected SSc serum samples from the ongoing SSc registry at the University of Leeds, including 33 VEDOSS, 20 progressing diffuse cutaneous SSc (pDcSSc), 27 stable DcSSc (sDcSSc), and 22 limited cutaneous SSc (LcSSc). All metabolites corresponding to each clinical subset of SSc were identified by IsoMS Pro 1.2.19 and NovaMT Metabolite Database v3.0 and classified according to a three-tier identification method. The significant metabolite pathways per each subset were analyzed by MetaboAnalyst 5.0.

Results: We measured a total of n=7538 distinct metabolic features with 362, 672, and 708 metabolites significantly dysregulated in pDcSSc, sDcSSc, and LcSSc as compared to VEDOSS (nominal p-value less than or equal to 0.05), respectively. Along with the disease progression from VEDOSS to established SSc, “Arginine and proline metabolism” a putative potentiator of TGFβ-induced production of matrix protein, was globally and the most significantly dysregulated across the disease subsets (nominal p-value less than or equal to 0.05). “Beta-Alanine metabolism” and “Histidine metabolism” were uniquely and significantly dysregulated in sDcSSc and LcSSc, respectively. Furthermore, we conducted binary classification analyses using baseline biomarkers, leading to the identification of 20 and 15 classified metabolic biomarkers associated with early disease progression in the lung (e.g., 4-Hydroxyphenylglyoxylic Acid and N-Acetyl-5-Hydroxyl-L-tryptophan) and skin (e.g., Dihydronaringenin-O-sulphate and Tryptophyl-Serine), respectively (rate/month for rapid progressor, mRSS >0.2; FVC <-0.5, AUC >0.7).

Conclusions: These findings suggest that metabolic changes in serum samples from SSc patients are closely linked to clinical subsets, and some of these metabolic biomarkers show correlations with clinical manifestations such as skin and lung fibrosis. It is important to note that while our results are promising, further validation studies in an independent cohort are necessary to confirm the reliability and robustness of these metabolic biomarkers. Nonetheless, our findings hold the potential to contribute to the identification of individuals at risk of rapid disease progression in the context of SSc.

OC.11

EXPLORING THE VERY EARLY DIAGNOSIS OF SYSTEMIC SCLEROSIS (VEDOSS) CRITERIA USAGE IN THE REAL-WORLD: A MULTINATIONAL SURVEY OF RHEUMATOLOGISTS

Daniel Twigg, Nicola Massey

Adelphi Real World, Bollington, UNITED KINGDOM

Introduction: The heterogeneous and progressive nature of systemic sclerosis (SSc) make it important to detect early yet difficult to diagnose. Presence of Raynaud’s phenomenon (RP) and puffy fingers, positivity for antinuclear antibodies (ANA) and SSc-specific autoantibodies, and abnormal nailfold capillaries have been identified by SSc experts as markers for the very early diagnosis of SSc (VEDOSS). The multicentre, longitudinal VEDOSS project from 2010 to 2018 supported the predictive value of combinations of these clinical characteristics for SSc development. Whether physicians are determining which criteria suspected SSc patients meet, remains unexplored. We aimed to investigate the usage of VEDOSS-related tests and observations in a real-world clinical setting.

Material and Methods: Data were drawn from the Adelphi Real World SSc Disease Specific Programme™, a cross-sectional survey, with retrospective data collection, of rheumatologists actively managing SSc patients, conducted from April – August 2022 in France, Germany, Japan, Italy, Spain, the United Kingdom and the United States of America. Rheumatologists making treatment decisions for at least five SSc patients within the past year reported data on practice setting, SSc expertise and beliefs, preference for diagnostic tests and vigilance for specific symptoms relating to the VEDOSS criteria. Analyses were descriptive.

Results: Of rheumatologists surveyed (n=354), 49% reported a public hospital as their primary practice setting, 20% a public office and 20% a private office. Over half (51%) reported having a special interest in treating SSc and 83% reported that they believe there is a population of SSc patients that progress more quickly than others over the five years following their first non-RP symptom. RP was the most reported symptom leading rheumatologists to suspect SSc (94%); presence of puffy fingers was reported by 73%. ANA-testing was reported as being routinely used to aid SSc diagnoses by 88% of rheumatologists, SSc-specific autoantibody testing was reported by 89%, and nailfold capillaroscopy by 51%. Only 39% of rheumatologists reported routinely using all four non-RP criteria. When asked what would help facilitate the early identification and diagnosis of SSc, 83% reported increased awareness/education of physicians.

Conclusions: While the predictive value of the VEDOSS criteria in patients with RP has previously been demonstrated, just over a third of rheumatologists surveyed reported routinely using tests or observations to determine the criteria patients meet. Nailfold capillaroscopy was the most underutilised test in the real-world. Increased physician awareness and education would increase the rates of VEDOSS, diagnosing patients earlier and optimising treatment decisions.

OC.12

PULMONARY INVOLVEMENT IN VERY EARLY SYSTEMIC SCLEROSIS (VEDOSS): REPORT FROM A SINGLE CENTER

Marcelo Neto¹, Ana Isabel Maduro¹, André Saraiva¹, JAP da Silva^1,2, Tânia Santiago^1,2, Maria J. Salvador^1,2

¹Department of Rheumatology, Centro Hospitalar e Universitario de Coimbra, Coimbra, PORTUGAL, ²Faculty of Medicine, University of Coimbra, Coimbra, PORTUGAL

Introduction: Previous work has described interstitial lung disease (ILD) in very early systemic sclerosis (VEDOSS), but its progression over time is unknown. This can contribute decisively to the delineation of screening and monitoring strategies. The aim of this work was to evaluate the prevalence of ILD and its course in a cohort of patients who persist in a VEDOSS state.

Material and Methods: Retrospective observational study of patients persistently fulfilling VEDOSS criteria, with sequential measurement of diffusing lung capacity for carbon monoxide (DLCO) and forced vital capacity (FVC), at 0, 1, 5 and 10 years. A DLCO or FVC <80% of predicted values was considered indicative of lung involvement. A decline of FVC or DLCO of >=5% and >=15% in any of the time intervals was considered clinically relevant. Demographic, clinical, and immunological variables at baseline were taken as independent variables, including age, sex, disease duration, arthritis, capillaroscopy pattern, and autoantibody profile. Pearson χ² test, Fisher’s exact test or independent sample t-test were used for independent measures as appropriate. Repeated-measures ANOVA was used to compare variables over time.

Results: Thirty patients (86,7% women, mean baseline age 46,3±13,6 yrs) were included; median time from first symptoms to diagnosis was 4,4 (IQR 7,6) yrs.

At baseline, 4 patients had mildly decreased DLCO values (60-79% of predicted), results being normal in all remaining persons (Table 1). During the first year of follow-up, a >=5% decrease in FVC was observed in 6/30 patients (20%) and a DLCO decrease of >=15% in 2/30 (6,7%). There was a non significant trend for a higher probability of FVC and DLCO decrease in males vs. females (M:50,0% vs. F:15,4% and M:25,0% vs F3,8%, respectively).

Pulmonary tests were available for the 25 and 5 patients who reached 5 and 10 years of follow-up. Mean values of FVC and DLCO consistently declined over time, without reaching statistical significance (Table 1). The incidence of significant change in FVC and DLCO between contiguous evaluations is represented in Figure 1. There is a non-significant trend for a higher probability of decline as follow-up time increased.

Conclusions: Our study provides data on the long-term assessment of FVC and DLCO in VEDOSS patients. There was a relevant decrease in FVC during the first year in 20% of the patients, and 20% had impaired DLCO (<80%) at 5 years, which suggests that a subset of patients who persist in VEDOSS state overtime still endure subclinical pulmonary involvement.

OC.13

IMMUNOSUPPRESSIVE TREATMENT ALONE VERSUS IMMUNOSUPPRESSIVE TREATMENT PLUS ORAL GLUCOCORTICOIDS FOR SKIN FIBROSIS IN EARLY DIFFUSE SSC PATIENTS. A TARGET TRIAL EMULATION STUDY FROM THE EUSTAR DATABASE

Denis Mongin¹, Marco Matucci-Cerinic², Ulrich A Walker³, Oliver Distler⁴, Radim Becvar⁵, Elise Siegert⁶, Lidia P Ananyeva⁷, Vanessa Smith⁸, Juan José Alegre-Sancho⁹, Sule Yavuz¹⁰, Massimiliiano Limonta¹¹, Gabriela Riemekasten¹², Elena Rezus¹³, Madelon Vonk¹⁴, Marie -Elise Truchetet¹⁵, Francesco Del Galdo¹⁶, Delphine S Courvoisier¹, Michele Iudici¹

¹Division of Rheumatology, Department of Medicine, Geneva University Hospitals, Geneva, SWITZERLAND, ²Department of Experimental and Clinical Medicine, University of Florence & Division of Rheumatology AOUC, Florence, ITALY, ³Department of Rheumatology, University Hospital Basel, Basel, SWITZERLAND, ⁴Department of Rheumatology, University Hospital Zurich, Zurich, SWITZERLAND, ⁵Institute Of Rheumatology And Department Of Rheumatology 1st Medical School, Prague, CZECH REPUBLIC, ⁶Department of Rheumatology, Charité University Hospital, Berlin, GERMANY, ⁷V.A. Nasonova Research Institute Of Rheumatology Russian Federation, Moscow, RUSSIA, ⁸University of Ghent, Department of Rheumatology, Ghent, BELGIUM, ⁹Hospital Universitario Dr Peset, Valencia, SPAIN, ¹⁰Istanbul Bilim University, Dept. Of Rheumatology, Altunizade-Istanbul, TURKEY, ¹¹Asst Papa Giovanni XXIII, Bergamo, ITALY, ¹² Klinik Für Rheumatologie Und Klinische Immunologie, Universitätsklinikum Schleswig-Holstein, Lübeck, GERMANY, ¹³Grigore T. Popa University of Medicine and Pharmacy Iasi, Rehabilitation Hospital, Iasi, ROMANIA, ¹⁴Dept. of Rheumatology, Radboud University Medical Center, Nijmegen, THE NETHERLANDS, ¹⁵Department of Rheumatology, National Reference Center for Systemic Autoimmune Rare Diseases, Bordeaux University Hospita, Bordeaux, FRANCE, ¹⁶Leeds Institute of Rheumatic and Musculoskeletal Medicine, University of Leeds, Leeds, UNITED KINGDOM

Introduction: Glucocorticoids are commonly used to control active disease in patients with early diffuse systemic sclerosis (SSc), but there is limited evidence supporting their effectiveness on skin fibrosis. The only high-quality randomized controlled trial designed to evaluate the benefit of combining oral glucocorticoids with immunosuppressive therapy on skin fibrosis was prematurely stopped due to slow patient enrollment. The risk-benefit ratio of glucocorticoids use in early diffuse SSc patients remains uncertain, especially since they may also increase the risk of scleroderma renal crisis.

Material and Methods: We performed a comparative effectiveness study emulating a randomized controlled trial comparing the changes from baseline in skin fibrosis assessed by the modified Rodnan skin score (mRSS)(primary outcome) at 12±3 months in early (<5 years disease duration from non-Raynaud onset) diffuse SSc patients receiving either oral glucocorticoids (<15 mg/day of prednisone equivalent) combined with immunosuppressive treatment or immunosuppressive treatment alone. We used data from the European Scleroderma Trials and Research Group (EUSTAR) collected between January 2013 and December 2022. Groups were determined according to treatments received, without any intervention from the investigators. Secondary endpoints were the difference in the percentage of patients having progressive skin fibrosis; progressive lung fibrosis; regressive skin and lung fibrosis. Multivariable linear regression was used to adjust for baseline imbalance between groups. Sensitivity analyses were performed, including multiple imputation addressing the missing data.

Results: Among 230 patients included (mean age 49 years; 33% male; 61% anti-Scl70 positive, mean ± standard deviation mRSS 18.7 ± 8.6), 126 received immunosuppressive treatment monotherapy and 104 immunosuppressive treatment plus oral glucocorticoids. Mean ± standard deviation change in mRSS from baseline to 12-month was -2.57 ± 8.00 in immunosuppressive plus oral prednisone group versus -2.88 ± 6.65 in immunosuppressive monotherapy group. The adjusted effect of prednisone on the mRSS change was 0.1 (95% CI [-1.8; +2.0], p = 0.41). Similar results were observed in the subgroups of patients with shorter disease duration (< 3 years from non-Raynaud’ onset) or having a mRSS ranging from 7 to 22. We did not find any between-group difference for all the prespecified secondary outcomes. No case of scleroderma renal crisis was observed.

Conclusions: Within the limitation of the analysis of observational data, this study did not find any significant effect of low to moderate dose oral glucocorticoids on skin fibrosis at 1-year when prescribed in combination with immunosuppressive treatment to early diffuse patients. RCTs are required before conclusions can be drawn.

OC.14

TYPE I INTERFERON ACTIVATION OF MONOCYTES IN SYSTEMIC SCLEROSIS IS CGAS-STING DEPENDENT AND CAN BE DIRECTLY INDUCED BY DERMAL FIBROBALSTS: A PROMISING THERAPEUTIC TARGET FOR SYSTEMIC SCLEROSIS

Stefano Di Donato¹, Rebecca Ross¹, Christopher Wasson¹, Jochen Schmitz², Jun Li², Sudha Visvanathan², Francesco Del Galdo^1,3

¹Leeds Institute of Rheumatic and Musculoskeletal Medicine - Faculty of Medicine and Health, Leeds University, Leeds, UNITED KINGDOM, ²Boehringer-Ingelheim Pharmaceuticals Inc, Mount Kisco, New York, USA, ³NIHR Leeds Biomedical Research Centre, Leeds Teaching Hospitals NHS Trust, Leeds, UNITED KINGDOM

Introduction: Type I Interferon (IFN) activation has been shown to correlate with disease activity and progression in Systemic Sclerosis (SSc). In vitro studies have shown that SSc dermal fibroblasts induce pro-inflammatory activation of monocytes. Here we aimed to study the source of Type I IFN activation in SSc.

Material and Methods: Dermal fibroblasts from SSc and healthy controls (HC) skin biopsies were co-cultured, in a ratio 1 to 5, for 48 hours with THP1 monocytes stably transfected with Lucia gene. Lucia is a secreted luciferase reporter gene under the control of an ISG54 minimal promoter in conjunction with five IFN-stimulated response elements (ISRE), that is activated by several pathways, including cGAS-STING-IRF3. After co-culture, luminescence-fibre assay was conducted on supernatants to quantify Lucia luciferase activity as an indirect measure of ISRE transcription. In ex vivo validation studies, human peripheral blood mononuclear cells (PBMCs) from SSc patients (n=12) were isolated and cultured in RPMI 1640 with 10% FBS and treated with DMSO, a cGAS, or a STING inhibitor for 16 hours. mRNA expression of IFN-related genes was assessed via real-time PCR. Protein phosphorylation was assessed by Western blot. Paired Wilcoxon test was used for statistical analysis.

Results: SSc fibroblasts induced more than 2-fold transcription of ISRE in THP1 compared to healthy fibroblasts, as assessed by Luciferase activity (mean [SD] luminescence, 4234 [1927] vs 1852 [425], p=0.02). The activation of ISRE was associated with significant phosphorylation of IRF3. In turn, P-IRF3 levels were suppressed after cGAS or STING inhibition. Accordingly, targeting cGAS/STING pathway abrogated SSc fibroblasts-induced activation of ISRE elements by 91% (360 [96] vs 4234 [1927], p-value<.0001) and 89% (434 [95] vs 4234 [1927], p-value<.0001) for cGAS and STING inhibitors, respectively. Ex vivo, SSc PBMCs showed an average 70% increase in basal Type I IFN activation compared to HC PBMC, as assessed by rt-PCR of 5 Interferon inducible genes (namely, CXCL10, IFIT1, MX1, OAS, ISG15). Treatment of SSc PBMCs with cGAS and STING inhibitor suppressed the increased ISG expression by overall 21% and 40% (p-value=0.002), respectively, the latter bringing ISG to levels not significantly different from HC PBMCs.

Conclusions: SSc fibroblasts induce Type I IFN activation in THP1 cells in a cGAS/STING dependent way. Ex-vivo, inhibition of cGAS/STING pathway in SSc PBMCs suppressed the upregulation of Interferon inducible genes. cGAS/STING pathway activation is a promising target to modulate Type I IFN activation in SSc.

OC.15

TRANSCRIPTOMES AND OPEN CHROMATIN LANDSCAPES IMPLICATE TRANSCRIPTION FACTORS REGULATING INJURY AND REPAIR PHENOTYPES IN SYSTEMIC SCLEROSIS ENDOTHELIAL CELLS

Mengqi Huang¹, Tracy Tabib¹, Dinseh Khanna², Shervin Assassi³, Robyn Domsic¹, Robert Lafyatis¹

¹Division of Rheumatology and Clinical Immunology, Department of Medicine, School of Medicine, University of Pittsburgh, Pittsburgh, USA, ²Division of Rheumatology, Department of Internal Medicine, University of Michigan, Ann Arbor, USA, ³Division of Rheumatology, University of Texas Health Science Center Houston, McGovern Medical School, Houston, USA

Introduction: Typically, vasculopathy is the earliest clinical manifestation in patients with systemic sclerosis (SSc). It is associated with endothelial cell (EC) injury and angiogenesis. Thus, elucidating transcriptomes and epigenomes of ECs is critical to better understand the mechanisms underlying SSc-associated vasculopathy.

Material and Methods: Transcriptomic and epigenetic alterations of EC populations in SSc skin were assessed by single-cell RNA sequencing (scRNA-seq; SSc n=27, Control n=10) and single-nucleus transposase-accessible chromatin sequencing (snATAC-seq; SSc n=8, Control n=6). Immunofluorescent staining in skin and proteomics assay of patients’ sera confirmed the altered SSc EC phenotypes. Human dermal microvascular endothelial cells were transduced with lentivrius carrying genes of ETS1, ELK4, ELK3 and ERF to perform gain-of function assay in vitro, followed by bulk RNA-seq and matrigel tube formation assay to assess the regulated genes and angiogenic effects by the overexpressed factor.

Results: Transcriptomic signatures of classical vascular linages (arterial, capillary and venous ECs) and lymphatic ECs were shared by both control and SSc ECs. Compared to control ECs, arterial ECs in SSc skin were decreased in number and showed increased expression of genes associated with apoptosis. Two small EC subpopulations, tip (PGF, CXCR4) and proliferating (MKI67, TOP2A) ECs, were found markedly upregulated in SSc. They showed enhanced proangiogenic and proliferative activities implied by pathway analysis of their enriched genes. Systematic analysis of molecular features with clinical metadata associated these aberrant SSc-ECs with disease pathogenesis and SSc clinical traits, such as skin fibrosis and the presence of digital ulcer. Ligand-receptor analysis demonstrated altered intercellular networks of SSc EC subpopulations with neighboring cells, including perivascular and immune cells. Furthermore, integration of open chromatin profiles with transcriptomic analysis suggested an increased accessibility of DNA binding sites of ETS family transcription factors in both SSc arterial and tip ECs. Gain-of-function assay of those top enriched ETS factors suggested that ELK4, ERF, and ETS1 may orchestrate arterial apoptosis and dysregulated angiogenesis in SSc.

Conclusions: These data provide new insights into both transcriptional and chromatin alterations in driving endovascular remodeling in SSc and suggest previously unrecognized EC targets, ELK4, ERF and ETS1, for ameliorating vascular disease in SSc.

OC.16

PAIRED TRANSCRIPTOMIC ANALYSIS OF TRANSITIONAL B CELLS IN SYSTEMIC SCLEROSIS HIGHLIGHTS DEFECTIVE PERIPHERAL TOLERANCE

Claire Beesley¹, Nina Goldman¹, David Abraham¹, Christopher Denton¹, Rizgar Mageed², Voon Ong¹

¹Royal Free Hospital, Division of Medicine, University College London, London, UNITED KINGDOM, ²The William Harvey Research Institute, Queen Mary University of London, London, UNITED KINGDOM

Introduction: Transitional B cells comprise a distinct population of B cells that have recently migrated to the periphery from the bone marrow. In systemic sclerosis (SSc), there are differences in the number and function of transitional B cells compared with healthy controls (HCs). In addition, we have previously isolated transitional B cells from SSc patients which are seropositive for anti-topisomerase I autoantibodies (ATA) and demonstrated that these cells evade tolerance. To investigate mechanisms that underpin this defect we performed a paired bulk and single-cell RNA-sequencing study of transitional B cell subsets in treatment-naive ATA-positive diffuse cutaneous SSc and HCs.

Material and Methods: 5000 CD19+CD24hiCD38hi transitional B cells were sorted from four HCs and four treatment-naive ATA-positive dcSSc patients. Single-cell RNA-sequencing was performed on the sorted B cells and the data was analysed using Seurat in RStudio. Paired bulk RNA-sequencing of the sorted transitional T1 (CD19+CD24hiCD38hiCD27-IgMhiIgDmed) and T2 (CD19+CD24hiCD38hiCD27-IgMhiIgDhi) subsets was undertaken in a subset of this cohort.

Results: Single cell transcriptomics identified at least four transitional B cell clusters in SSc patients and HCs (Figure 1). This included a cluster of marginal zone precursor cells expressing PLD4 and MZB1 which is expanded in SSc (cluster 2 in Figure 1A). We also identified and interesting cluster (cluster 3) which was diminished in the SSc group and expresses genes such as TNFRSF13B (BAFF), AICDA (AID) and S100A10 (Figure 1B). In the HCs this cluster also expresses genes such as CD27 and IGHA1 which are absent in SSc indicating that B cell development is impaired in the disease. Subsequent pathway analyses provided evidence for defective tolerance in SSc patients which may be related to dysregulated NF-kB signalling and abhorrent extrafollicular B cell development as evidenced through pseudotime trajectory analysis. Differential gene expression analyses also identified five candidate gene signatures which are significantly enriched in the T2 subset. This latter finding is in congruence with the single cell data and provides further evidence to support the existence of a peripheral tolerance/selection checkpoint at the transitional B cell stage.

Conclusions: Through this data we have identified four transitional B cell subsets and have characterised their transcriptomic profile. Additionally, we provide further evidence for a potential breach in peripheral tolerance of SSc patients. Further studies are underway to identify key pathways that underpin B cell pathogenesis in SSc and explore the relevance of such findings for better targeted therapies.

OC.17

AM1476 – A HIGHLY SELECTIVE 5-HYDROXYTRYPTAMINE 2B (5-HT2B) RECEPTOR ANTAGONIST AMELIORATING BOTH PULMONARY AND DERMAL FIBROSIS IN THE CHRONIC GRAFT VERSUS HOST DISEASE MODEL OF SYSTEMIC SCLEROSIS

Christina Wenglén, Helena Arozenius, Lars Pettersson

AnaMar AB, Lund, SWEDEN

Introduction: The primary event in systemic sclerosis (SSc) pathogenesis is thought to be injury to endothelial cells, followed by aberrant vascular and immune responses leading to excessive deposition and accumulation of extracellular matrix (ECM).

5-hydroxytryptamine (5-HT), released from platelets at sites of vascular damage, is a known stimulator of tissue fibrosis and a significant role of peripheral 5-HT2B receptors in fibrosis has been suggested. Activation of the 5-HT2B receptor on fibroblasts results in increased myofibroblast differentiation and ultimately fibrosis. 5-HT regulates macrophage plasticity and activation of the 5-HT2B receptor promotes pro-fibrotic M2 macrophages. Pharmacologic inhibition of 5-HT2B receptor signalling thus represents a promising treatment strategy for fibrotic disorders including SSc. In this study the effects of a selective 5-HT2B receptor antagonist on fibrosis development were evaluated in the chronic graft versus host disease (cGvHD) model of SSc. RNA sequencing analysis was performed on skin biopsies to get insight in target cells and molecular pathways modulated by 5-HT2B receptor inhibition.

Material and Methods: The murine cGvHD model was used to evaluate anti-fibrotic effects of the 5-HT2B receptor antagonist, AM1476. Oral, once and twice daily treatment with AM1476 was applied day 21 to 49. Pulmonary fibrosis was evaluated using hydroxyproline content, Sirius Red staining, and Ashcroft score. Dermal thickness, myofibroblast counts, collagen production and number of M2 macrophages were used to evaluate dermal fibrosis. The number of phosphorylated Smad3 (pSmad3) positive cells in dermal tissue was used to evaluate inhibition of TGF-β signalling. Total RNA was extracted from skin biopsies. RNA sequencing was performed, followed by statistical bioinformatic analyses.

Results: The 5-HT2B receptor antagonist AM1476, significantly reduced all measured dermal and pulmonary fibrosis readouts in the cGvHD model using an oral therapeutic treatment approach. The number of pSmad3 positive cells were significantly reduced in skin samples isolated from treated animals. RNA sequencing analysis of skin biopsies showed that treatment with AM1476 ameliorated several processes related to fibroblast and macrophage activation. Further analysis defined a set of genes that offers potential to define a response signature for 5-HT2B receptor inhibition in clinical trials.

Conclusions: Inhibition of 5-HT2B receptor activity resulted in anti-fibrotic effects in pulmonary and dermal fibrotic tissues. Amelioration of several processes related to fibroblast and macrophage activation was observed. The highly selective 5-HT2B receptor antagonist AM1476, represents a promising drug candidate for treatment of fibrotic conditions and is currently in phase II development for systemic sclerosis.

OC.18

GRB2 SERVES AS A VIABLE TARGET AGAINST SKIN FIBROSIS IN SYSTEMIC SCLEROSIS BY REGULATING ENDOTHELIAL CELL APOPTOSIS

Yan Huang¹, Han Zhao ^1,2, Xiangguang Shi³, Jing Liu¹, Jui -Ming Lin³, Qianqian Ma¹, Shuai Jiang¹, Weilin Pu¹, Yanyun Ma^4,5, Jianlan Liu³, Wenyu Wu^{3, 6}, Jiucun Wang^1,3,7, Qingmei Liu ^3,5

¹School of Life Sciences, and Human Phenome Institute, Fudan University, Shanghai, CHINA, ²Nanjing Intellectual Property Protection Center, Nanjing, CHINA, ³Division of Dermatology, Huashan Hospital, Fudan University, Shanghai Institute of Dermatology, Shanghai, CHINA, ⁴Institute for Six-sector Economy, Fudan University, Shanghai, CHINA, ⁵Ministry of Education Key Laboratory of Contemporary Anthropology, School of Life Sciences, Fudan University, Shanghai, CHINA, ⁶Department of Dermatology, Jing' an District Central Hospital, Shanghai, CHINA

Introduction: Systemic Sclerosis (SSc) is an autoimmune disease characterized by vascular and immune system dysfunction, along with tissue fibrosis. Our previous study found GRB2 was downregulated after the treatment of salvianolic acid B (SAB), a bioactive component extracted from Salvia miltiorrhiza that could effectively attenuate skin fibrosis of SSc and experimental pulmonary fibrosis. In this study, we aim to investigate the role of GRB2 in the fibrosis skin of SSc.

Material and Methods: To validate the expression of GRB2 in the skin tissue of SSc, we analyzed the microarray data of SSc skin biopsies in Caucasians from the Gene Expression Omnibus (GEO) database. The expression of GRB2 was further detected in Chinese SSc and healthy controls. Bleomycin (BLM)-induced skin fibrosis mice were used to explore how Grb2 downregulation affected fibrosis. H2O2-challenged EA.hy926 cells were used to study the effects of GRB2 on the permeability of endothelial cells. The apoptosis of EA.hy926 endothelial cells was induced by H2O2 and apoptosis ratio was measured by flow cytometric. Transcriptome and phosphoproteomic analyses were performed on EA.hy926 cells transfected with NC and GRB2 siRNA to explore the GRB2 regulated pathway.

Results: The gene expression of GRB2 was significantly enhanced in SSc patient skin, 1.51-fold in Caucasians and 1.40-fold in Chinese. IHC staining showed more alpha-SMA, S100A4, and GRB2 positive cells in SSc patients than in normal controls. Double immunofluorescence staining showed the endothelial cells of SSc patient’s skin highly expressed GRB2. The in vivo study revealed that Grb2 knockdown alleviated skin fibrosis and apoptosis of endothelial cells in BLM mouse skin. The in vitro study showed that decreased GRB2 expression level could up-regulate tight junction related genes expression and inhibit apoptosis to protect endothelial cells from H2O2-induced hyperpermeability. Moreover, enrichment analysis performed on GRB2 siRNA transfected endothelial cells showed that the focal adhesion was enriched both in transcriptome and phosphoproteomic sequencing.

Conclusions: Our results demonstrated GRB2 highly expressed in endothelial cells of SSc skin, and inhibiting GRB2 could effectively attenuate BLM-induced skin fibrosis and endothelial cell apoptosis. The focal adhesion pathway may be the mechanism of GRB2 in regulating endothelial cells. GRB2 is expected to be a new therapeutic target for SSc.

OC.19

THE LARGEST JAPANESE AND FRENCH GENETIC STUDIES FOR SYSTEMIC SCLEROSIS IDENTIFIED A NOVEL HIGH-RISK CAUSAL LOCUS AND THE HLA ASSOCIATIONS

Yuki Ishikawa¹, Working Group², Anne Cauvet-Burgevin³, Luc Mouthon⁴, Marie -Elise Truchetet⁵, Eric Hachulla⁶, Yannick Allanore^3,7, Chikashi Terao^1,8,9

¹RIKEN, Center for Integrative Medical Sciences, Laboratory for Statistical and Translational Genetics, Yokohama, JAPAN, ²Japan Ministryof Health, Labor, and Welfare, Tokyo, JAPAN, ³INSERM U1016, INSTITUT COCHIN, Paris, FRANCE, ⁴Internal Medecine, Cochin Hospital, Université Paris Cité, Paris, FRANCE, ⁵ImmunoConcEpt, CNRS, UMR 5164, University of Bordeaux, Talence, FRANCE, ⁶Internal Medicine, Lille University School of Medicine, Lile, FRANCE, ⁷Rheumatology a Department, Cochin Hospital, Paris Descartes University, Paris, FRANCE, ⁸Clinical Research Center, Shizuoka General Hospital, Shizuoka, JAPAN, ⁹School of Pharmaceutical Science, University of Shizuoka, Shizuoka, JAPAN

Introduction: The genetic architecture of SSc in East Asian populations has been less well characterized compared to those of Caucasian populations. This study aimed to identify novel causal SNPs specific to Japanese and those shared with European populations and clarify their mechanistic effects on SSc pathology.

Material and Methods: A total of 1,499 cases and 112,609 controls and 1,593 cases and 9,528 controls were enrolled from Japan and France, respectively. Japanese genotype data was imputed referring to the high-depth WGS data of 3,256 Japanese subjects combined with the phase 3v5 1,000 genome project data. We conducted GWAS for Japanese SSc and then a trans-ancestry meta-analysis with the largest European meta-GWAS dataset. For the MHC region, SNPs, HLA-alleles, and amino acids were imputed by SNP2HLA (tapa) for both Japanese and French SSc datasets, and an association test and omnibus test were conducted.

Results: We identified three (FCRLA-FCGR, TNFAIP3, PLD4) and four (EOMES, ESR1, SLC12A5, TPI1P2) novel loci by the Japanese GWAS and the trans-ethnic meta-analysis, respectively. One of the novel risk SNPs of Japanese GWAS, rs6697139, was a rare variant located within FCGR gene clusters demonstrating a strong effect size (MAF 0.04, OR 2.05, P=4.9×10-11). The complete LD variant, rs10917688, was a part of the cis-regulatory element and a binding motif of IRF8 in B cells. IRF8 was another genome-wide significant locus and, notably, the association of rs10917688 was significant only in the presence of the risk allele of IRF8. Furthermore, significant heritability enrichment of active histone marks was found in B cells both in European and Japanese populations, implying a shared genetic architecture across populations. PRS developed by using effects sizes of the trans-ancestry meta-analysis moderately fit in the disease prediction in Japanese (AUC 0.604) with further improvement by prioritizing the top 5% IMPACT-annotated SNPs for IRF8-biding in the B-cell line (AUC 0.610). The HLA analysis identified the 76th amino acid position of the HLA-DPbeta1 molecule, a part of the antigen binding pocket previously reported for its association with GVHD, showed the strongest association in Japanese, while the 37th amino acid position of the HLA-DRbeta1 was the strongest association in French SSc. Specific associations were observed in the clinical subtypes of SSc including lcSSc, dcSSc, ACA+SSc, ATA+SSc., and SSc-ILD and there were also overt differences of association patterns between Japanese and French SSc

Conclusions: Our largest Asian GWAS and HLA association study provide novel insight into disease mechanisms of SSc and population differences of genetic architectures.

OC.20

A HIGH-THROUGHPUT PHARMACOLOGICAL SCREENING AS NOVEL IN VITRO APPROACH TO IDENTIFY MYOFIBROBLASTS TARGETS IN SYSTEMIC SCLEROSIS

Elena Pachera¹, Isabelle Claerr², Christina Merz-Stoeckle², Enrico Schmidt², Oliver Distler¹

¹Center of Experimental Rheumatology, Department of Rheumatology, University Hospital Zurich, University of Zurich, Zurich, SWITZERLAND, ² Novartis Institutes for Biomedical Research, Basel, SWITZERLAND

Introduction: Myofibroblasts are extracellular matrix (ECM) producing cells, considered major effector cells in systemic sclerosis (SSc) and other fibrotic diseases. Here, we used an in vitro high-throughput screening platform employing a library of pharmaceutical compounds to target selectively pathogenic myofibroblasts.

Material and Methods: Systemic sclerosis dermal fibroblasts (n=4) were seeded and starved for 24h. TGFb stimulation was performed to exacerbate the myofibroblasts phenotype. A high-throughput pharmacological screening was implemented using the Novartis Mode of Action library, consisting of 2464 compounds with annotated targets. 72h after the compound treatment, a-smooth muscle action (aSMA) immunofluorescent staining was performed as a measure of myofibroblasts differentiation. The images were acquired with the Yokogawa CV7000 confocal scanner. Values of the area and the intensity of aSMA staining for each compound were recorded and compared to values for TGFb only treated cells.

Massively parallel single-cell RNA sequencing (MARS-seq) data obtained from CD90+ skin cells of early diffuse SSc (dSSc, < 5 years of disease duration, n=24) patients and healthy controls (HC, n=60) were extracted from a publicly available dataset (GSE195452). Data were analyzed with the Seurat 4.3.0 package in R.

Results: The screening identified 82 compounds that were significantly reducing the area of the aSMA staining corresponding to 176 targets. Similarly, 151 compounds, corresponding to 299 targets, were significantly reducing the intensity of aSMA staining. Analysis using the KEGG and WikiPathway databases showed that the targets were significantly enriched in “PI3K-Akt signaling pathway” and “Calcium signaling pathway”, while the analysis using MSigDB Hallmark indicated a strong association with “Apoptosis”, “Complement” and “IL-2/STAT5 Signaling” among others.

To validate which candidates might be used to target myofibroblasts in the early active phases of the disease, we analyzed the transcriptome profiles of early dSSc fibroblasts, obtained from MARS-seq experiments. In this dataset, we determined that the major contributors of the ECM gene transcription were the subsets Fibro_LGR5, and Fibro_MYOC2.

Next, we matched the results from the compound library screening with the genes overexpressed by each ECM overproducing fibroblast subtype. With this approach, we were able to identify several undescribed candidates, targeting fibroblasts in SSc, as displayed in the figure. Interestingly, some of the candidate such as EGFR, endothelin receptors and DPP4 are well characterized antifibrotic targets confirming the validity of our approach.

Conclusions: We have developed an innovative strategy to identify uncharacterized candidates for the selective targeting of ECM producing myofibroblasts. The most promising novel targets will undergo additional functional validation.

OC.21

PRIMARY CILIA OF SYSTEMIC SCLEROSIS (SSC) DERMAL FIBROBLASTS ARE DISRUPTED BY DOWNREGULATION OF CAVEOLIN-1 AND ABERRANT ACTIVITY OF AURORA A KINASE, INDEPENDENT OF TRANSFORMING GROWTH FACTOR BETA

Rebecca Wells¹, Rebecca Ross^1,2, Alex Timmis³, Ioanna Georgiou¹, Colin Johnson⁴, Natalia Riobo Del Galdo³, Francesco Del Galdo^1,2

¹Leeds Institute of Rheumatic and Musculoskeletal Medicine, University of Leeds, Leeds, UNITED KINGDOM, ²NIHR Leeds Musculoskeletal Biomedical Research Centre, Chapel Allerton Hospital, Leeds, UNITED KINGDOM, ³School of Molecular and Cellular Biology and Astbury Centre for Structural Molecular Biology, University of Leeds, Leeds, UNITED KINGDOM, ⁴Leeds Institute of Medical Research, University of Leeds, Leeds, UNITED KINGDOM

Introduction: Systemic Sclerosis (SSc) is characterised by abnormal activation of tissue fibroblasts. The cellular and molecular mechanisms driving this activation remain unclear, but aberrant TGFβ signalling activation and downregulation of cell membrane component Caveolin-1 are implicated. Both TGFβ and Caveolin-1 have roles at the primary cilium (PC), an antenna-like organelle which forms a signalling hub for pathways involved in cellular homeostasis. Several of these pathways are dysregulated in SSc. This study aimed to examine PC structure in SSc dermal fibroblasts to determine its potential contribution to SSc pathogenesis via signalling dysregulation, and investigate TGFβ activation and Caveolin-1 downregulation in any abnormal ciliary phenotype.

Material and Methods: Dermal fibroblasts from healthy control (HC) and SSc patients (limited, diffuse, and VEDOSS) were isolated. PC were visualised by fluorescent confocal microscopy. Average PC length was quantified from at least 100 PC across 3 fields per condition. Fibroblasts were treated with exogenous TGFβ, TGFβR1 inhibitor SD208, ROCK2 inhibitor KD025, and Aurora A kinase (AURKA) inhibitor MLN8054. Canonical TGFβ signalling was inhibited with siRNA targeting SMAD3. HC fibroblasts were treated with shRNA against Caveolin-1 (shCAV) to mimic SSc-associated downregulation, or Scramble control shRNA (shSCR).

Results: SSc patient fibroblast PC were, on average, 45% shorter than HC (2.5±0.6µm vs 4.6±1.3µm (P<0.0001)), regardless of SSc subtype. HC shCAV fibroblasts displayed significantly shorter PC (3.0±0.8µm) together with already described increased fibrotic markers and sensitivity to TGFβ. TGFβ treatment significantly reduced PC length in both HC and SSc. The effect of TGFβ was fully abolished by TGFβR1 inhibition, partially abolished by ROCK2 inhibition, and unaffected by SMAD3 knockdown, but none of these treatments were able to ‘rescue’ SSc or shCAV cilia length to that of HC. Inhibition of AURKA, known to be involved in ciliary disassembly, did not prevent TGFβ-induced shortening, but did restore PC length in SSc and shCAV fibroblasts to lengths resembling HC and shSCR controls, leaving HC and shSCR cilia unaffected.

Conclusions: PC length is stably reduced in SSc dermal fibroblasts, early in disease, with Caveolin-1 downregulation and aberrant AURKA activity contributing to the shorter cilium. While TGFβ activation can shorten PC via a non-canonical pathway involving ROCK2, this is not responsible for the short cilia phenotype in SSc. These findings support the notion that SSc profibrotic activation may involve mechanisms beyond TGFβ, and highlight the potential significance of Caveolin-1 and AURKA in modulating PC length as contributors to profibrotic activation, and potential therapeutic targets for tissue fibrosis.

OC.23

Reticularized skin phenotype in Systemic Sclerosis- potential mediators

Sara Chenguit Fakhouri¹, Honglin Zhu², Laura Konstantinidis¹, Moritz Ronicke¹, Aleix Rius Rigau¹, Andrea -Hermina Györfi³, Alexandru -Emil Matei³, Yi -Nan Li³, Markus Eckstein¹, Carol Geppert¹, Clara Dees¹, Alexander Kreuter⁴, Michael Sticherling¹, Carola Berking¹, Georg Schett¹, Jörg Distler³, Christina Bergmann¹

¹Universitätsklinikum Erlangen-Nürnberg, Erlangen, GERMANY, ²Central South University, Changsha, CHINA, ³Universitätsklinikum Düsseldorf, Düsseldorf, GERMANY, ⁴Helios Privatklinik Oberhausen, Oberhausen, GERMANY

Introduction: Alteration of skin morphology is a central hallmark of Systemic Sclerosis (SSc). In healthy adult human skin, the dermis is comprised of a papillary layer (papillae, nerves, mechanoreceptors, vessels, loose collagen bundles) and a reticular dermis (coarse collagen bundles). In SSc, skin architecture is disrupted and shifted towards a “reticularized” phenotype with a loss of dermal papillae and excessive collagen deposition. Little is known on the mechanisms necessary to maintain physiological adult human skin morphology. Evidence from murine experiments suggests that short-distance morphogen gradients are necessary for three-dimensional pattern formation in embryonic skin morphogenesis in rodents.

Material and Methods: Dermal skin morphology was quantitatively assessed using HE and trichrome staining of skin sections of 40 SSc patients and 18 healthy controls. Enrichment of papillary and reticular marker genes and WNT target genes in healthy and SSc skin transcriptome were assessed by gene set enrichment analyses using prepublished datasets (Zhu, et al, manuscript submitted). Spatial WNT/β-catenin activation was analyzed by RNAscope®-in situ hybridization detection of AXIN2. Immunofluorescence (IF) staining was performed to visualize β-catenin distribution.

Results: Dermal papillae were markedly decreased in number, area and height in SSc compared to controls. Consistently, gene expression analyses revealed a shift of papillary/reticular marker gene expression towards a reticular profile in SSc skin compared to controls. Furthermore, we analyzed the ratio between papillary and reticular marker genes, which was shifted in certain fibroblast subpopulations that have recently been described (Zhu et al, manuscript submitted; Tabib et al, J Invest Dermatol. 2018). Next, we observed an enrichment of WNT3A-regulated target genes in papillary gene sets of healthy skin. Consistently, we observed a distinct spatial β-catenin distribution pattern in healthy skin that is disturbed in SSc: In healthy skin, β-catenin-positive fibroblasts were predominantly located in the papillary dermis and reduced in the reticular part. In SSc skin, we observed a 2-fold increase of β-catenin-positive fibroblasts throughout the dermis. A gradient with an enrichment of β-catenin-expressing fibroblasts in the papillary layer was no longer detectable in SSc. The loss of spatial WNT activation in SSc was confirmed on a transcriptional level.

Conclusions: Here, we demonstrated a loss of a physiological β-catenin gradient in SSc, in correlation with a loss of papillary dermal structures. A thorough analysis of spatial distribution of WNT agonists and antagonists and advanced in vitro models of papillary skin structure will be needed to gain further understanding on the mechanisms underlying skin structure maintenance in healthy adult human skin.

OC.24

DEVELOPMENT OF PULMONARY HYPERTENSION-LIKE OCCLUSIVE VASCULOPATHY IN INTERLEUKIN-13 TRANSGENIC MICE

Jacqueline Wax¹, Lifang Wen¹, Xiaoyang Yue¹, Christoph Hölscher², Gabriela Riemekasten³, Xinhua Yu¹, Frank Petersen¹

¹Reasearch Center Borstel, Department of Pulmonary Immune Diseases, Borstel, GERMANY, ²Research Center Borstel, Department of Infection Immunology, Borstel, GERMANY, ³University of Lübeck, University Clinic of Schleswig-Holstein, Department of Rheumatology and Clinical Immunology, Lübeck, GERMANY

Introduction: Systemic sclerosis (SSc) is a chronic autoimmune disease characterized by inflammation, fibrosis and vasculopathy. In previous studies, we demonstrated that autoantibodies against angiotensin II receptor type 1 (AT1R) play a functional role in the pathogenesis of the disease by mediating the development of skin inflammation and fibrosis as well as inflammation in the lung. However, two further critical features of SSc, lung fibrosis and vasculopathy, are not induced in this mouse model, suggesting a further key player in the manifestations of the disease. In this study, we further investigate the TH2 cytokine IL-13 as a potential second key player in the disease manifestations.

Material and Methods: IL-13 transgenic (IL-13tg) mice overexpressing IL-13 under control of the human CD2 promoter as well as their littermate controls were immunized with membrane extracts overexpressing human AT1R. Nine weeks later, tissues were analyzed for pathological changes.

Results: The immunized IL-13tg mice exhibited inflammation and occlusive vasculopathy in the lung while littermate controls did not show any pathological changes. Severity of vasculopathy was positively correlated with lung inflammation, IL-13 expression in the lung, number of neutrophils and IL-6 concentration in BAL fluids. Surprisingly, the same pathological changes were observed in both IL-13tg and IL-13tg/IL-4R-/- mice in a tuberculosis mouse model. Neutralizing antibodies against IL-13 inhibited the development of occlusive pulmonary vasculopathy in immunized IL-13tg mice, confirming a causal role of IL-13 in the development of occlusive pulmonary vasculopathy in this mouse model.

Conclusions: IL-13 is a central player for the development of occlusive pulmonary vasculopathy and thus represents a potential therapeutic target in the treatment of SSc.

OC.25

UNRAVELING THE UNEXPECTED IMPACTS OF IL-6 ON PULMONARY VASCULOPATHY AND CYTOKINE DYSREGULATION IN EXPERIMENTAL SYSTEMIC SCLEROSIS

Afsaneh Mehrpouyan¹, Jacqueline Wax¹, Christoph Hölscher², Gabriela Riemekasten^1,3, Frank Petersen¹, Xinhua Yu¹

¹Research Center Borstel, Department of Pulmonary Immune Diseases, Borstel, GERMANY, ²Research Center Borstel, Department of Infection Immunology, Borstel, GERMANY, ³University of Lübeck, University Clinic of Schleswig-Holstein, Department of Rheumatology and Clinical Immunology, Lübeck, GERMANY

Introduction: Recently, our laboratory has established a novel mouse model for systemic sclerosis (SSc) by immunizing mice with a membrane extract of CHO cells overexpressing human angiotensin II receptor type I (AT1R). Notably, after the immunization, IL-13 transgenic (IL-13tg) mice that overexpress IL-13 specifically in activated T cells develop pulmonary occlusive vasculopathy compared to AT1R-immunized littermate controls. Furthermore, levels of interleukin-6 (IL-6) in bronchoalveolar lavage (BAL) fluid are significantly elevated and positively correlated with the severity of the occlusive vasculopathy. In this study, we aimed to investigate the role of IL-6 in this context and its potential as a therapeutic target.

Material and Methods: IL-13tg mice were immunized with membrane extract of hAT1R on Day 0 and Day 21 and treated with anti-IL-6 neutralizing antibody or its isotype control once per week. Six weeks after the first immunization, mice were sacrificed, and peripheral blood and tissues were collected for further evaluation.

Results: The preliminary data showed that anti-IL-6 treatment effectively depleted IL-6 in peripheral blood but not in the lungs. Unexpectedly, IL-6 depletion led to the development of mild pulmonary vasculopathy in non-immunized IL-13tg mice. Furthermore, in AT1R-immunized female IL-13tg mice, anti-IL-6 neutralizing antibody treatment exacerbated pulmonary occlusive vasculopathy compared to isotype control IgG treatment.

Conclusions: Contrary to our expectations, our findings suggest that IL-6, often considered pro-inflammatory, may play a protective role in developing pulmonary occlusive vasculopathy in IL-13tg mice. Further investigations need to be carried out to validate the protective role of IL-6 and to explore the underlying molecule mechanism.

OC.26

SINGLE CELL RNA-SEQ OF SCLERODERMA-ASSOCIATED INTERSTITIAL LUNG DISEASE LUNG EXPLANTS REVEALS AN ACTIVE, CYTOTOXIC NATURAL KILLER CELL POPULATION

Cristina Padilla¹, Eleanor Valenzi¹, Tracy Tabib¹, Banafsheh Nazari¹, John Sembrat¹, Mauricio Rojas², Patrizia Fuschiotti¹, Robert Lafyatis¹

¹University of Pittsburgh, Pittsburgh, USA, ²Ohio State University, Columbus, USA

Introduction: Scleroderma-associated interstitial lung disease (SSc-ILD) is the leading cause of death among SSc patients. Loss of alveolar epithelial cells (AEC) and bronchial cell hyperplasia, i.e., bronchiolization, are principal pathological features of the disease. While inflammation is known to be another primary disease feature, little is known about the potential role of lung lymphoid populations and particularly, natural killer (NK) cells on bronchiolization. What is known about NK cells in SSc comes from peripheral blood and bronchoalveolar lavage fluid and has not provided a consistent vision of their role in SSc. Using single cell RNA-sequencing (scRNA-seq), we have identified an activated, cytotoxic NK cell population in SSc lung tissue.

Material and Methods: Lung tissues from 13 SSc-ILD and 6 healthy control (HC) lungs collected were processed and analyzed by 3’ 10X single cell RNA chemistry. Samples were normalized and batch-corrected before unsupervised UMAP clustering and subclustered further into specific lymphoid subpopulations. Differentially expressed genes were derived to identify lymphoid subclusters. Connectome, a program that analyzes ligand-receptor interactions within and between cell types in lung tissue, was utilized. Immunohistochemistry stains for CD56 and granzyme B (GZMB) in lung tissues were completed.

Results: 16 lymphocyte clusters in both SSc-ILD and HC, including three populations of NK cells, several populations of T cells, a proliferating population of lymphoid cells, and a cluster of interferon-activated T cells were identified. We observed striking shifts in lymphoid cell populations in SSc-ILD lungs. SSc-ILD lungs showed a shift in CD56dim CD16+ NK cells toward an activated phenotype with strong upregulation of GZMB, CD226, interferon-gamma, and its associated genes. Amphiregulin (AREG), a growth factor ligand, that binds epidermal growth factor receptor (EGFR), was shown for the first time to be highly upregulated in the SSc-ILD NK cells, demonstrating high-yield interaction with EGFR on basal and goblet cells via Connectome. Higher proportions of CD8+ T cells and regulatory T cells were also observed in SSc-ILD lungs.

Conclusions: SSc-ILD lungs are infiltrated by active, cytotoxic CD56dim CD16+ NK cells, showing increased expression of IFN-gamma, GZMB, and AREG. These activated NK cells might kill AEC, contributing to the observed loss of these cells. The highly upregulated expression of AREG suggests that these cells might also contribute to bronchial epithelial cell hyperplasia as well as fibroblast activation via the AREG-EGFR pathway. Ongoing studies will clarify the pathological role of these cells in the bronchiolization and formation of honeycomb cysts in SSc-ILD.

OC.27

INTERFERON SCORE AS A NOVEL TRIAL ENRICHMENT TOOL TO PREDICT CLINICAL TRIAL OUTCOMES IN PATIENTS WITH LIMITED CUTANEOUS SYSTEMIC SCLEROSIS

Stefano Di Donato^1,2, Rebecca Ross¹, Ranjita Karanth¹, Vishal Kakkar^1,2, Enrico De Lorenzs^1,3, Giuseppina Abignano^1,4, Kristina Clark⁵, Christopher Denton^5,6, Francesco Del Galdo^1,2

¹Leeds Institute of Rheumatic and Musculoskeletal Medicine, University of Leeds, Leeds, United Kingdom, ²NIHR Leeds Biomedical Research Centre, Leeds Teaching Hospitals NHS Trust, Leeds, United Kingdom, ³Division of Rheumatology, Catholic University of the Sacred Heart, Fondazione Policlinico Univ Agostino Gemelli IRCCS, Rome, Italy, ⁴Istituto Reumatologico Lucano, Azienda Ospedaliera Regionale San Carlo., Potenza, Italy, ⁵UCL Division of Medicine, University College London, London, United Kingdom, ⁶UCL Centre for Rheumatology and Connective Tissue Diseases, Royal Free Hospital, London, United Kingdom

Introduction: Limited cutaneous SSc (lcSSc) is the understudied clinical subset of SSc, despite representing more than 60% of patients. Recently, the analysis of a large single centre observational cohort, has identified nature and rate of clinically meaningful events in lcSSc. This has informed the design of a combined morbi-mortality endpoint for clinical trials in lcSSc (MINIMISE, EudraCT: 2019-004139-21). Type I IFN activation is associated with high disease activity and poor outcome in diffuse cutaneous SSc. Here we aimed to study the value of serum Type I IFN in stratifying for outcome in lcSSc according to the MINIMISE combined morbi-mortality endpoint.

Materials and methods: A retrospective, longitudinal cohort of lcSSc patients was identified within a national, multicentre observational cohort. The MINIMISE, combined morbi-mortality endpoint was used as clinical outcome from time of serum analysis. IFN score was calculated as previously described (Ross et al.) derived from the serum concentration of CCL2, CCL8, CCL19, CXCL9, CXCL10, and CXCL11 (Myriad RBM). Patients were characterized according to EUSTAR MEDS dataset. Student's t, Wilcoxon's, Pearson Chi-squared, and Fisher's tests were employed for analysis, as appropriate. Kaplan-Meier curves, Log-rank test and Cox Proportional Hazard regression models were employed for time to endpoint analysis.

Material and Methods: 149 lcSSc patients (median [IQR] follow up 65 [39] months) were identified for analysis (146 female, mean [SD] age 59 [12] years). Median (IQR) disease duration was 13 (12) years. 67 (45%) patients had high IFN score (above Healthy Control [HC] threshold), whereas 82 (55%) were within HC range "IFN low". 37/67 (55%) patients with high IFN score vs 10/82 (12%) (p<0.001) met the combined endpoint during follow-up. Overall, 79% of patients meeting the endpoint were IFN High. Accordingly, IFN Low patients presented a higher event-free survival (Log rank Chi-Sq=12, df=1, p<0.001). Age, presence of ACA antibodies alone, ILD, DU, or PAH were the other baseline variables predicting outcome over time, independently of IFN score. Cox Proportional Hazard models defined High IFN score being associated with 5.5 Hazard Ratio (HR) compared to IFN low (95% C.I. 2.68-11.3, p<0.001). IFN score maintained a HR of 2.2 (95% C.I. 1.38-3.49, p<.001) also when used as continuous variable, independently of other clinical variables.

Conclusion: Serum assessment of Type I IFN activity could be a useful test to stratify for severe outcome in lcSSc and as such, contribute to enrichment and/or stratification strategies in clinical trial design.

OC.28

RESULTS FROM OUTCOME SELECTION FOR AN FDA- AND EMA-ACCEPTABLE COMBINED RESPONSE INDEX FOR LIMITED CUTANEOUS SYSTEMIC SCLEROSIS: THE CRISTAL PROJECT

Alain Lescoat¹, Yen Chen², Susan L. Murphy², Michael Lang², Nadia Vann², Paul Didio², Sarah Hiya², Tiffany Phanhdone², Sue Farrington³, Yannick Allanore⁴, David Cella ⁵, Lorinda Chung ⁶, Philip Clements⁷, Christopher Denton⁸, Francesco Del Galdo⁹, Oliver Distler¹⁰, Monique Hinchcliff¹¹, Michael Hughes¹², Laura K Hummers¹³, John D Pauling¹⁴, Janet Pope¹⁵, Virginia Steen¹⁶, John Varga², Peter A. Merkel¹⁷, Maya H. Buch¹², Dinesh Khanna ²

¹Internal Medicine and Clinical immunology, CHU Rennes, Rennes, FRANCE, ²University of Michigan Scleroderma Program, Ann Arbor, MI, USA, ³Federation of European Scleroderma Associations, Copenhagen, DENMARK, ⁴Department of Rheumatology, Cochin institute, Paris, FRANCE, ⁵Department of Medical Social Sciences, Feinberg School of Medicine, Northwestern University, Chicago, USA, ⁶Division of Immunology and Rheumatology, Stanford University School of Medicine, Palo Alto, CA, USA, ⁷Division of Rheumatology, David Geffen School of Medicine, University of California, Los Angeles, CA, USA, ⁸Centre for Rheumatology, UCL Division of Medicine, Royal Free Campus, London, UNITED KINGDOM, ⁹Leeds Institute of Rheumatic and Musculoskeletal Medicine, University of Leeds, Leeds, UNITED KINGDOM, ¹⁰Department of Rheumatology, University Hospital Zurich, University of Zurich, Zürich, SWITZERLAND, ¹¹Department of Internal Medicine, Section of Rheumatology, Allergy and Immunology, Yale School of Medicine, New Haven, USA, ¹²Division of Musculoskeletal and Dermatological Sciences, The University of Manchester, Manchester, UNITED KINGDOM, ¹³Johns Hopkins University, Division of Rheumatology, Baltimore, MD, USA, ¹⁴North Bristol NHS Trust, Bristol, UK; Bristol Medical School, University of Bristol, Bristol, UNITED KINGDOM, ¹⁵Division of Rheumatology, St Joseph's Health Care, London, Ontario, CANADA, ¹⁶Division of Rheumatology, Department of Medicine, Georgetown University School of Medicine, Washington, DC, USA, ¹⁷University of Pennsylvania, Philadelphia, PA, USA

Introduction: The CRISTAL project, supported by the World Scleroderma Foundation, Scleroderma &. Raynaud’s UK, the Scleroderma Clinical Trial Consortium, and the Scleroderma Research Foundation, aims to create a combined response index for limited cutaneous systemic sclerosis (lcSSc) for use in clinical trials; the CRISTAL index. The most bothersome domains were identified through patient-centered focus groups and existing outcomes were identified through a scoping literature review (presented at the 2022 World Scleroderma Congress).

The main objective was to select candidate outcomes for inclusion in a planned longitudinal prospective observational study to assess the instruments’ psychometric properties. Aims included a) enrich the list of candidate clinician-reported outcomes (ClinROs) and performance outcomes (PerfOs), and use this list to select the most appropriate candidate outcomes for CRISTAL; and b) select patient-reported outcomes (PROs) for these domains

Material and Methods: Experts in lcSSc were asked to suggest additional items/outcome measures in a two-round international online Delphi exercise. ClinROs and PerfOs were selected during a two-day meeting utilizing a nominal group technique (NGT) comprising 3 patient research partners and 8 international experts. Selection of PROs is ongoing through an online survey and cognitive debriefing.

Results: 100 experts were invited to the Delphi exercise and 71 provided answers for at least one round. Participants from round 1 were invited to round 2, rating each item on scales (range: 1-9) for feasibility, face validity, content validity, and overall appropriateness for the CRISTAL Index. 59/71 participants provided answers to at least one survey. Items endorsed by more than 50% of the experts were included for discussion at the 2 day-NGT discussion. During the NGT meeting, participants discussed and ranked items from “most appropriate for the CRISTAL index” to “least appropriate”. At the end of each ranking, voting members were asked if they agreed with the ranking; 80% agreement was required to move to the next domain. Across the 10 lcSSc-specific domains discussed, 19 items (17 ClinROs and 2 PerfOs) were identified and retained as draft items for the planned observational cohort study. Ongoing work with an online survey and cognitive debriefing will provide data on which PROs are deemed the most appropriate from patients’ perspectives to assess lcSSc, with results expected by the end of 2023.

Conclusions: The CRISTAL project is steadily progressing. Next steps include final selection of PROs and a prospective cohort study to assess the psychometric properties of selected items to lead to the development of the CRISTAL index.

OC.29

DECREASSC: AN OBSERVATIONAL STUDY TO ASSESS THE VALIDITY OF HOME MONITORING TO DETECT PROGRESSION OF INTERSTITIAL LUNG DISEASE IN SYSTEMIC SCLEROSIS

Arthiha Velauthapillai¹, C Moor², J De Vries-Bouwstra³, M S Wijsenbeek-Lourens², C H M Van Den Ende¹, M C Vonk¹

¹Radboud University Medical Center - Department of Rheumatology, Nijmegen, THE NETHERLANDS, ²Department of Pulmonology, Erasmus Medical Center, Rotterdam, THE NETHERLANDS, ³Department of Rheumatology, Leiden University Medical Center, Leiden, THE NETHERLANDS

Introduction: Timely detection of progression of interstitial lung disease (ILD) in systemic sclerosis(SSc) is warranted to start or intensify treatment to halt disease progression. Home spirometry, enabling frequent monitoring, can potentially identify progressive disease earlier. We aim to assess the validity of home monitoring to detect progressive SSc-ILD.

Material and Methods: Forty-three adult SSc patients, from our ongoing DecreaSSc study, fulfilling ACR-EULAR criteria, with a disease-duration < 5 years and maximum 8 weeks immunotherapy are included. Patients performed weekly home spirometry (forced vital capacity(FVC)) using a Bluetooth-enabled handheld spirometer for one year. Results were collected through a mobile application. During semi-annual study visits, patients performed pulmonary function testing(PFT) in the hospital. This interim analysis focuses on patients with 6 month’s follow-up data. Pearson correlation coefficients between home and hospital measurements at baseline and 6 months was calculated. Next, we performed regression analyses with FVC as dependent variable for each individual patient to estimate the absolute decline in FVC% assessed by home spirometry at 6 months . Based on this estimated decline, the sensitivity and specificity of home spirometry to detect progression of SSc-ILD (defined as 5% decline in FVC (yes/no)) was assessed using the change in hospital PFT as gold standard

Results: Twenty-five patients completed 6 months follow-up, with a mean(SD) age of 59(11) years, 54% being female, 35% having dcSSc and a mean(SD) disease duration of 1.5(1.2) years. The mean(SD) FVC% predicted at baseline based on the hospital PFT was 91(13)%. During the study, 9 patients were initiated on immunomodulation. A strong correlation between home and hospital measurements was observed at baseline (r=0.90, p<0.001) and at 6 months (r= 0.85, p=<0.001). The mean(SD) change in FVC % predicted estimated with linear regression analysis for the home measurements was -1.7(5.8)% and the change in FVC% predicted in the hospital measurements was -0.32(5.8)%. Five patients showed disagreement between hospital and home measurements for a decline of 5% in FVC (Table 1). The sensitivity of home spirometry to detect progression of SSc-ILD was 75%(95% CI 58-92%) and the specificity was 81% (95% CI 66-96%).

Conclusions: These preliminary data indicate that home spirometry might be a valid tool to monitor lung function and to detect progressive SSc-ILD.

OC.30

BEST CLINICAL PRACTICE IN THE TREATMENT OF JUVENILE SYSTEMIC SCLEROSIS: EXPERT PANEL GUIDANCE - THE RESULT OF THE INTERNATIONAL HAMBURG CONSENSUS MEETING DECEMBER 2022

Ivan Foeldvari¹, Kathryn Torok², Jordi Anton³, Michael Blakley⁴, Tamas Constantin⁵, Maurizio Cutolo⁶, Christopher Denton⁷, Kim Fliglstone⁸, Bernd Hinrichs⁹, Suzanne Li¹⁰, Susan Maillard ¹¹, Edoardo Marrani¹², Pia Moinzadeh¹³, Catherine Orteu¹⁴, Clare Pain¹⁵, John Pauling¹⁶, Clarissa Pilkington¹¹, Franziska Rosser², Vanessa Smith¹⁷, Dan Furst¹⁸

¹Hamburg Centre for Pediatric and Adolescent Rheumatology, Hamburg, GERMANY, ²University of Pittsburgh, Pittsburgh, USA, ³Hospital Sant Joan de Deu, Barcelona, SPAIN, ⁴Indiana University School of Medicine, Indianapolis, USA, ⁵Semmelweis University, Budapest, HUNGARY, ⁶University of Genoa and IRCCS San Martino Polyclinic Hospital, Genoa, ITALY, ⁷Royal Free London NHS Foundation Trust, London, UNITED KINGDOM, ⁸Scleroderma and Raynauds United Kindgom, London, UNITED KINGDOM, ⁹Children's pulmonology, Asklepios Klinik Nord Heidberg, Hamburg, GERMANY, ¹⁰Hackensack University Medical Center, Hackensack, USA, ¹¹Great Ormond Street Hospital, London, UNITED KINGDOM, ¹²Meyer Children's Hospital, Florence, ITALY, ¹³Universtät Köln, Cologne, GERMANY, ¹⁴Royal Free London, London, UNITED KINGDOM, ¹⁵Alder Hey Childrens Foundation NHS Trust, Liverpool, UNITED KINGDOM, ¹⁶North Bristol NHS Trust, Bristol, UNITED KINGDOM, ¹⁷Ghent University, Ghent, BELGIUM, ¹⁸University of California, Los Angeles, USA

Introduction: Juvenile systemic sclerosis (jSSc) is an orphan disease with a prevalence of 3 in 1,000,000 children. Currently there is only one consensus treatment guideline concerning skin, pulmonary and vascular involvement for jSSc, the jSSc SHARE (Single Hub and Access point for paediatric Rheumatology in Europe) initiative, which was based on data procured up to 2014. Therefore, an update of this guideline, with more recent literature and expert experience, and extension of the guidance to more aspects of the disease is needed.

Material and Methods: In the consensus meeting at the Hamburg Symposium for Juvenile Systemic Sclerosis (jSSc) the treatment options were reviewed on Day 1. The consensus meeting was held on Day 2, where > 80% agreement among voting members present had to be reached for consensus, as defined a priori, using the nominal group technique. The consensus meeting was moderated by Dan Furst, an experienced adult SSc rheumatologist.

Results: Treatment opinions were provided for most facets of jSSc including general management, some which differs from adult systemic sclerosis, such as the use of corticosteroids, and specific organ involvement, such as skin, musculoskeletal, pulmonary, gastroenterology etc. (Table 1.)

Conclusions: We are suggesting the treat to target strategy to treat early to prevent cumulative disease damage in jSSc. Conclusions are derived from both expert opinion and available literature, which is mostly based on adult systemic sclerosis (aSSc), given shared pathophysiology, extrapolation of results from aSSc studies was judged reasonable.

OC.31

CIGARETTE EXPOSURE IN SYSTEMIC SCLEROSIS: IMPACT ON AUTOANTIBODY EXPRESSION AND DISEASE MANIFESTATIONS: AN ANALYSIS OF THE EUSTAR COHORT

Jacopo Ciaffi¹, Sophie Liem², Saad Ahmed², Eva Hoekstra², Piotr Wiland³, Tatsuya Atsumi⁴, Gabriella Szucs⁵, Alexandra Balbir-Gurman⁶, Laszlo Czirjak⁷, Elisabetta Zanatta⁸, Ina Koetter⁹, Jorg Henes¹⁰, Marco Matucci-Cerinic¹¹, Paolo Airo'¹², Francesco Ursini¹, Thomas Huizinga², Jeska De Vries-Bouwstra²

¹Medicine and Rheumatology Unit, IRCCS Istituto Ortopedico Rizzoli, Bologna, ITALY, ²Department of Rheumatology, Leiden University Medical Center, Leiden, THE NETHERLANDS, ³Wroclaw University of Medicine, Department of Rheumatology and Internal Diseases, Wroclaw, POLAND, ⁴Hokkaido University Hospital, Department of Rheumatology, Endocrinology and Nephrology, Sapporo, JAPAN, ⁵University of Debrecen, Faculty of Medicine, Department of Rheumatology, Debrecen, HUNGARY, ⁶Rheumatology Institute, Rambam Health Care Campus, Haifa, ISRAEL, ⁷University of Pecs, Department of Rheumatology and Immunology, Pecs, HUNGARY, ⁸Padova University Hospital, Rheumatology Unit, Padova, ITALY, ⁹Asklepios Clinic Altona, Medical Department 4, Rheumatology, Immunology, Nephrology, Hamburg, GERMANY, ¹⁰University Hospital and Faculty of Medicine, University of Tübingen, Internal Medicine II - Oncology, haematology, clini, Tubingen, GERMANY, ¹¹University of Florence, Dipartimento di Medicina Sperimentale e Clinica, Florence, ITALY, ¹²Asst Spedali Civili of Brescia, Rheumatology and Clinical Immunology Unit, Brescia, ITALY

Introduction: The literature describes a lower frequency of anti-topoisomerase antibodies (ATA) in never-smokers than in ever-smokers in systemic sclerosis (SSc). Therefore, we used the EUSTAR database to evaluate the effect of smoking on ATA expression and how sex, smoking and ATA/anticentromere (ACA) interact and impact on disease progression.

Material and Methods: All SSc patients with complete information about smoking status available were included and categorised as “never-smokers” or “ever-smokers”. Chi-square test and logistic regression were used to evaluate differences in ATA and ACA expression between ever-smokers and never-smokers. A possible dose-response effect was evaluated using logistic regression with pack/years or smoking duration as predictors. We evaluated the association between disease progression, smoking and antibody expression using Kaplan-Meier curves and multivariate Cox regression models corrected for age and sex and with stratification for ACA/ATA status.

Results: In total, 12314 patients were included. In women, 34% of never-smokers were ATA+ compared to 21% of ever-smokers (p <0.001). Ever-smoking females had lower risk of ATA+ [OR 0.523 (95% CI 0.473 – 0.578), p <0.001] and higher exposure further decreased the risk of being ATA+ [OR number of pack/years 0.988 (95% CI 0.978 – 0.997), p= 0.009; OR years smoking duration 0.963 (95% CI 0.952 – 0.975), p <0.001]. Ever-smoking females had increased risk to be ACA+ [OR 1.320 (95% CI 1.212 – 1.437), p<0.001] and a higher number of pack/years further increased risk of ACA expression [OR 1.012(95% CI 1.002 – 1.021), p= 0.014]. In men, the proportion of ATA+ was comparable between non-smokers and ever-smokers, but higher exposure decreased the risk of ATA expression (OR 0.989 (95% CI 0.979 – 0.999), p= 0.032]. We observed different associations with disease progression depending on antibody status (Figure). In ACA-positive patients, smoking was associated with increased risk of mortality (HR 1.3, 95% CI 1.0 – 1.6, p= 0.033), of cardiac involvement (HR 1.2, 95% CI 1.1 – 1.4, p= 0.001) and of “any organ progression” (HR 1.2, 95% CI 1.1 – 1.3, p= 0.002). In ATA-positive patients, smoking was a risk factor for mortality (HR 1.4, 95% CI 1.1 – 1.8, p= 0.006) and for development of digital ulcers (HR 1.2, 95% CI 1.0 – 1.3, p= 0.029).

Conclusions: We observe a clear association between smoking and ATA in SSc, particularly in females. We describe a negative dose-response relationship between cumulative cigarette exposure and ATA expression. Intriguingly, smoking exposure had different impact on disease manifestations depending on ATA/ ACA status.

OC.32

INCIDENT INTERSTITIAL LUNG DISEASE IN SYSTEMIC SCLEROSIS IN THE EUSTAR DATABASE: INCREASED RISK OF MORTALITY DESPITE LESS FUNCTIONAL PROGRESSION

Liubov Petelytska^1,2, Arthiha Velauthapillai³, Lorenzo Tofani⁴, Patricia E. Carreira⁵, Giovanna Cuomo⁶, Eric Hachulla⁷, Ivan Castellvì⁸, Radim Becvar⁹, Alexandra Balbir-Gurman¹⁰, Paolo Air¹¹, Ira Litinsky¹², Lesley Ann Aaketkoo¹³, Madelon C Vonk³, Jeska De.Vries-Bouwstra¹⁴, Anna -Maria Hoffmann-Vold^1,15, Marco Matucci-Cerinic^16,17, Oliver Distler¹, Cosimo Bruni^1,16

¹Department of Rheumatology, University Hospital Zurich, University of Zurich, Zurich, SWITZERLAND, ²Bogomolets National Medical University, Dept Internal Medicine #3, Kyiv, UKRAINE, ³Department of Rheumatology, Radboud University Medical Center, Nijmegen, THE NETHERLANDS, ⁴Department of Statistics, Computer Science, Applications, University of Florence, Florence, ITALY, ⁵Servicio de Reumatología, Hospital 12 de Octubre, Avda. De Córdoba, S/N, 28041, Madrid, SPAIN, ⁶Department of Precision Medicine, Luigi Vanvitelli University of Campania, Naples, ITALY, ⁷Department of Internal Medicine and Clinical immunology, Huriez Hospital, University of Lille, Lille, FRANCE, ⁸Department of Rheumatology. Hospital Universitari de la Santa Creu i Sant Pau, Universitat Autnoma de Barcelona, Barcelona, SPAIN, ⁹Institute Of Rheumatology And Department Of Rheumatology 1st Medical School, Charles University Na Slupi 4, Prague, CZECH REPUBLIC, ¹⁰B. Shine Rheumatology Institute, Rambam Health Care Campus, Rappaport Faculty of Medicine, Technion, Haifa, ISRAEL, ¹¹Spedali Civili di Brescia, Scleroderma UNIT, UOC Reumatologia ed Immunologia Clinica, Piazzale Spedali Civili 1, 25123, Brescia, ITALY, ¹²Centre Tel-Aviv Sourasky. Rheumatology Institute, 6 Weizmann St Tel Aviv, 6423906, Tel-Aviv, ISRAEL, ¹³Tulane University and Louisiana State University Schools of Medicine / University Medical Center Scleroderma and Sarcoid, New Orleans, USA, ¹⁴Leiden University Medical Center, Department of Rheumatology, Leiden, THE NETHERLANDS, ¹⁵Department of Rheumatology, Oslo University Hospital, Rikshospitalet, Oslo, NORWAY, ¹⁶Dept Experimental and Clinical Medicine, Division of Rheumatology, University of Florence - University Hospital Careggi, Florence, ITALY, ¹⁷San Raffaele Hospital, Unit of Immunology, Rheumatology, Allergy and Rare Diseases (UniRAR), Milan, ITALY

Introduction: Interstitial lung disease (ILD) determines a negative prognosis in systemic sclerosis (SSc). Although most SSc patients present ILD at baseline, new onset of ILD can also occur during the follow-up. We aimed to compare the prognosis of incident and prevalent SSc-ILD.

Material and Methods: SSc patients from the EUSTAR database were eligible if classified according to the 2013 ACR/EULAR criteria, with available information for ILD on high-resolution computed tomography at baseline and follow-up, with at least 6 months of observation. Patients with pulmonary arterial hypertension on right heart catheterization were excluded.

Based on HRCTs, we separated patients with ILD at baseline (ILD-prevalent), patients with new onset of ILD during the follow-up (ILD-incident) and patients who remained ILD negative (ILD-negative).

Functional progression of ILD at 12±3 months intervals was tested with three different definitions: (A) absolute FVC decline>=5% predicted, (B) absolute FVC decline>=5% predicted or absolute DLCO decline>=10% predicted and (C) relative FVC decline>=10% predicted or 5-9% with relative DLCO decline>=15% predicted.

Logistic regression with backward selection was used to identify independent predictors, comparing ILD-prevalent vs ILD-incident cases. The three ILD groups were tested as a predictor of mortality using multivariable Cox regression with backward selection. Known predictors of SSc-ILD progression and mortality were chosen as covariates from the literature and based on expert opinion.

Results: We included 9003 SSc patients: 4265 (47.3%) ILD-prevalent, 735 (8.2%) ILD-incident and 4003 (44.5%) ILD-negative.

ILD progression was variably observed among 7337 follow-ups of 2803 SSc-ILD patients (25% vs 35% vs 19%, according to the respective above-mentioned definitions). Despite a milder baseline functional impairment (Table 1), the ILD incident group showed a decreased risk of ILD progression by definition B (OR 0.85, 95%CI 0.73-0.97) and C (OR 0.83, 95%CI 0.69-0.98), but not by definition A (Figure.1).

During a median 4.1 (IQR2.0-7.7) years follow-up, 708/9003 deaths were recorded, more frequent in the ILD prevalent (506/4265, 11.9%) compared to ILD incident (57/735, 7.8%) or ILD negative (145/4003, 3.6%) groups (p<0.01 by Log-rank test).

In the Cox regression model (Figure.2), the ILD-incident group showed a lower risk of mortality in comparison to the ILD-prevalent (HR 0.67, 95%CI 0.51-0.89), but higher than the ILD-negative group (HR 1.94, 95%CI 1.41-2.67).

Conclusions: ILD-incident patients present a higher risk of mortality compared to ILD-negative cases, despite representing a milder phenotype compared to ILD-prevalent patients. Therefore, screening for SSc-ILD should be continued after a negative baseline, aiming at early detection and treatment.

OC.33

EVALUATING THE ROBUSTNESS OF 2D AND 3D DEEP LEARNING-BASED MODELS FOR DIAGNOSING IDIOPATHIC PULMONARY FIBROSIS WITHIN AN ILD POPULATION UNDER DIFFERENT CT IMAGING PROTOCOLS

Grace Hyun Kim¹, Wenxi Yu¹, Jin Woo Song², Michael McNitt-Gray¹, Jonathna Goldin¹

¹David Geffen School of Medicine, UCLA, Los Angeles, USA, ²Department of Pulmonary and Critical Care Medicine, Asan Medical Center, University of Ulsan College of Medicine, Seoul, SOUTH KOREA

Introduction: Deep learning-based systems have not yet been broadly implemented in clinical practice, in part due to unknown robustness across multiple imaging protocols. We aim to access the robustness of several 2D and 3D models using CT from non-idiopathic pulmonary fibrosis (non-IPF) ILD subjects, acquired using various CT protocols.

Material and Methods: Three deep learning-based models, including one 2D and two 3D models, have been previously developed to classify ILD patients into IPF or non-IPF based on chest CT scans. Three models were trained on CT image data from 389 IPF and 700 non-IPF ILD patients, retrospectively obtained from five multi-center studies (model construction stage, these CT series are referred to as reference conditions). In one non-IPF ILD study, due to its specific study protocol, many patients had multiple CT image data sets that were acquired under both prone and supine positions and/or reconstructed under different imaging parameters from their clinical visit. To assess the robustness of the model under different imaging protocols, we used available CT data from this study (343 subjects) that had paired sets of CT images on the same patient between model construction stage and model evaluation stage (evaluation conditions). Among these 343 non-IPF ILD subjects, there are four subtypes of ILD subjects including 80 (23.3%) systemic sclerosis (SSc)-ILD. Generalized linear mixed effects model (GLMM) was utilized to identify the significant CT technical and clinical parameters that were associated with getting conflicting diagnostic results between reference and evaluation conditions. Selected parameters include effective tube current-time product (known as “effective mAs”), reconstruction kernels, slice thickness, patient orientation, manufacture model name, and clinical diagnosis.

Results: For all three deep learning models, the overall specificity of the non-IPF ILD diagnosis model decreased when that model was applied to the evaluation conditions (two sample proportional test P<0.05 for all three models). GLMM further suggests that for two out of three models, mean effective mAs across the scan is the key factor that leads to the decrease in model predictive performance (P<0.001); the difference of mean effective mAs between the reference and evaluation conditions, patient position, and slice thickness (3mm) are flagged as significant factors for one out of three models (P<0.05).

Conclusions: Preliminary findings demonstrated the lack of robustness of deep learning model when applying to CT series collected under different imaging protocols, which indicated that imaging standardization should be taken when developing and deploying deep learning models into clinical practice.

OC.34

RECOMMENDATIONS FOR THE EXECUTION AND REPORTING OF SKIN ULTRASOUND IN SYSTEMIC SCLEROSIS: AN INTERNATIONAL COLLABORATION UNDER THE WSF SKIN ULTRASOUND GROUP

Tania Santiago¹, Eduardo Ferreira¹, Barbara Ruaro ², Gemma Lepri^3,4, Lorraine Green⁵, Marie Wildt⁶, Shinji Watanabe⁷, Alain Lescoat⁸, Roger Hesselstrand⁶, Francesco del Galdo⁵, John D Pauling¹⁰, Lucy Jean Reeve⁹, Maria Antonieta D'Agostino¹¹, Marco Matucci-Cerinic^3,4, Annamaria Iagnocco¹², José A. Pereira da Silva¹

¹Rheumatology, Centro Hospitalar e Universitario de Coimbra EPE, Coimbra, PORTUGAL, ²Pulmonology, University Hospital of Cattinara, Triestre, ITALY, ³Experimental and Clinical Medicine, University of Florence, Florence, ITALY, ⁴Department of Internal Medicine, University of Florence, Florence, ITALY, ⁵Leeds Institute of Molecular Medicine Section of Musculoskeletal Disease, Leeds, UNITED KINGDOM, ⁶Department of Rheumatology, Lund University, Lund, SWEDEN, ⁷Department of Allergy and Rheumatology, Nippon Medical School Graduate School of Medicine, Tokyo, JAPAN, ⁸Univ Rennes, CHU Rennes, Inserm, EHESP, Irset (Institut de Recherche en Santé, Environnement et Travail) - UMR_S 1085, Rennes, FRANCE, ⁹Scleroderma Programme, NIHR Leeds Musculoskeletal Biomedical Research Centre, Leeds, UNITED KINGDOM, ¹⁰Royal National Hospital for Rheumatic Diseases, Royal United Hospital Bath NHS Trust, Bath, UNITED KINGDOM, ¹¹Rheumatology, Fondazione Policlinico Universitario Agostino Gemelli IRCCS, Roma, ITALY, ¹²Academic Rheumatology Centre, Department of Clinical and Biological Science, University of Turin, Turin, ITALY

Introduction: Ultrasound is a promising tool to foster much- needed improvement of skin assessment in systemic sclerosis (SSc). However, there is a remarkable methodological heterogeneity and lack of information in a variety of technical aspects in skin ultrasound studies. Despite these advances, the role of skin ultrasound in clinical practice and research is not yet established.

Our aim was to develop evidence and expert opinion-based recommendations to promote the standardisation and harmonisation of technical execution and reporting of skin ultrasound studies in SSc.

Material and Methods: A multidisciplinary task force of 16 members from five European countries and Japan was convened under the auspices of World Scleroderma Foundation.

First, a systematic literature review (SLR) was performed. Then, each member proposed and formulated items to the overarching principles, recommendations and research agenda. Two rounds of mails exchange for consensus as well as an on-line meeting were performed to debate and refine the proposals.

Two Delphi rounds of voting resulted in the final recommendations. Levels of evidence and strengths of recommendations were assigned, and task force members voted anonymously on the level of agreement with each of the items.

Results: Five overarching principles and seven recommendations were developed, based on an SLR and expert opinion, through consensus procedures.

The overarching principles highlight the promising role of skin ultrasound in SSc assessment, the need for standardisation of technical aspects, sufficient training and adequate equipment.

The recommendations provide standards for the execution and reporting of skin ultrasound in SSc. The research agenda includes the need for more research into unmet needs according to Outcome Measures in Rheumatology Algorithm requirements.

Conclusions: These are the first recommendations providing guidance on the execution and reporting of skin ultrasound in SSc patients, aiming at improving the interpretability, reliability and generalisability of skin ultrasound, thus consolidating its role in research and practice.

This will represent a major advance in topics such as identification of subclinical or early disease, and assessment of response to immunosuppressive or antifibrotic therapies.

Well-designed observational studies and randomised clinical trials are now required aiming at providing resolution to research agenda.

OC.35

DIFFERENTIATION BETWEEN ‘SCLERODERMA’ WITH ‘NON-SCLERODERMA’ PATTERNS IN NAILFOLD CAPILLARY MICROSCOPY: IS ARTIFICIAL INTELLIGENCE HERE APPLICABLE?

Viktor Korendovych¹, Natalina Sugiarto², Jan Rademacher³, Marielle Heide⁴, Manuela Oettler⁵, Andrea von Figura⁶, Peter Korsten⁷

¹University Medical Center Göttingen, Department of Nephrology and Rheumatology, Göttingen, GERMANY, ²University Medical Center Göttingen, Department of Nephrology and Rheumatology, Göttingen, GERMANY, ³University Medical Center Göttingen, Department of Nephrology and Rheumatology, Göttingen, GERMANY, ⁴University Medical Center Göttingen, Department of Nephrology and Rheumatology, Göttingen, GERMANY, ⁵Rheumatological Medical Practice, Göttingen, GERMANY, ⁶University Medical Center Göttingen, Department of Nephrology and Rheumatology, Göttingen, GERMANY, ⁷Clinic of Rheumatology and Clinical Immunology, Sendenhorst, GERMANY

Introduction: Nailfold video capillaroscopy (NVC) is a useful diagnostic tool in rheumatology. Pathological changes are included in the 2013 ACR/EULAR classification criteria for systemic sclerosis (SSc). For quick differentiation between ‘scleroderma’ and ‘non-scleroderma’ patterns, a fast-track algorithm was suggested in 2019 based on the capillary density and the presence of megacapillaries. Digitalization and artificial intelligence play an important role nowadays in many spheres of daily life and are consequently being implemented into medicine. In our study, we aimed to construct a deep learning model based on convolutional neural networks and determine if such a model may help physicians differentiate ‘scleroderma’ from ‘non-scleroderma’ patterns.

Material and Methods: NVC-images were captured using an Optilia Digital Video Capillaroscope and classified as ‘scleroderma’ and ‘non-scleroderma’ patterns. A few pictures were randomly excluded to build two groups with an equal image number. Subsequently, we assigned the pictures to a training and a test set. A deep learning model with 5x cross-validation was trained (Figure 1A). Three physicians with practical experience in NVC and three doctors without practical experience were asked to differentiate the images from the test set after a ten-minute tutorial, with and without the support of the developed deep-learning model serving as a ‘second opinion’. We developed the software application and estimated accuracy and Light's kappa for these purposes (Figure 2B).

Results: A total of 1568 capillaroscopy images from 80 patients were analyzed. After the randomized exclusion, 1176 images were included in the study: 976 and 200 pictures in the training and test sets, respectively. The model had an accuracy of 89.44%, a sensitivity of 95.08% (95% CI 0.9277 to 0.9682), a specificity of 83.81% (95% CI 0.8024 to 0.8697), and an area under the curve (AUC) of 0.96 in the validation set and an accuracy of 95%, a sensitivity of 98% (95% CI 0.9296 to 0.9976), a specificity of 92% (95% CI 0.8484 to 0.9648), and an AUC of 0.99 in the test set. Using a deep learning model as a second opinion by the physicians showed better interrater reliability without a significant difference in accuracy (Table 1).

Conclusions: We successfully developed a deep learning model for the differentiation between ‘scleroderma’- vs. ‘non-scleroderma’ NVC patterns. Using such a model in clinical practice may help physicians achieve better interrater reliability, especially in inexperienced providers.

OC.36

COMPARING DEEP NEURAL NETWORK ANALYSIS TO MODIFIED RODNAN SKIN SCORE IN A TRIAL OF BELUMOSUDIL FOR SYSTEMIC SCLEROSIS PATIENTS

Baran Gunes¹, Lucy Duran Camacho¹, Shawn Cowper ¹, Gauri Panse¹, Elizabeth Bundschuh¹, Alyssa Williams ¹, Nicolas Page¹, Mary Carns², Kathleen Aren², Niki Pradhan³, Elana Bernstein³, Sarah Fantus⁴, Chase Correia², Francis Wilson¹, Seamus Mawe⁵, J. Matthew Mahoney⁵, Monique Hinchcliff¹, Rui Wang⁶

¹Yale University - Department of Rheumatology, New Haven, USA, ²Northwestern University- Divisions of Rheumatology and Pulmonary and Critical Care Medicine, Evanston, USA, ³Columbia University- Department of Rheumatology, New York, USA, ⁴Kansas City Physician Partners- Rheumatology, Kansas, USA, ⁵The Jackson Laboratory- Computational Sciences, Bar Habor, USA, ⁶Sanofi- Oncology Clinical Translation Medicine, Paris, FRANCE

Introduction: We previously published a proof-of-principle study demonstrating the potential utility of computer vision (Deep Neural Network/DNN) applied to stained skin biopsy sections from patients with systemic sclerosis (SSc) as a novel skin outcome. We compared the ‘DNN Fibrosis Score’ with histopathologic and modified Rodnan Skin Score (mRSS) changes for participants enrolled in a clinical trial.

Material and Methods: Ten adults with early (<= 6 y) diffuse cutaneous SSc (15 <= mRSS <= 35) in an open-label belumosudil (ROCK2 inhibitor, 200 mg PO BID) trial had mRSS and two, 4mm, dorsal arm skin biopsies performed at weeks 0, 24, and 52. Biopsies were stained with CD34, CD3, CD8, alpha smooth muscle actin (ASMA), H&E, and trichrome. Two blinded dermatopathologists assessed biopsies for 16 histopathological parameters. CD3+, CD8+ were counted and visual analogue scales (VAS) were used to score CD34, ASMA, and relative SSc severity on H&E and trichrome (Figure. 1). DNN algorithm, AlexNet, was applied to trichrome-stained images to generate a ‘DNN Fibrosis Score.’ DNN Fibrosis Scores were compared to mRSS using Spearman correlation. Parameters were compared to mRSS and DNN Fibrosis Scores using logistic regression models. p<= 0.05 was considered significant.

Results: Five patients had paired biopsies. The median (interquartile range) mRSS change between 0 - 52W was -2.5 (-11 to 7.5) while the median DNN Fibrosis Score change (W0 - last follow-up) was -6 (-10.5 to 6.5) (Figure. 3). Of the histopathological scores, subcutaneous (SC) fat loss (p = 0.012), eccrine entrapment (p = 0.008), % CD8+ among CD3+ cells (p = 0.006) changed most during treatment. The correlation between mRSS and DNN-Fibrosis Score for the 5 paired biopsies was 0.18 [at higher mRSS, i.e. 25-51, the correlation was weaker] (Figure. 3b). Per 1-unit mRSS increase, the parameter odds ratios (OR); p-values were: telangiectasia =2.01; 0.001, perivascular CD3+ =1.03; 0.015, and % of CD8+ among CD3+ =1.08; 0.031. Per 1-unit DNN Fibrosis Score increase, OR; p-values for parameters were: hyalinized collagen =1.1; 0.00033, SC fat loss =1.47; 0.00033, intima wall =1.21; 0.005, and eccrine entrapment =1.14; 0.046 (Figure. 3c).

Conclusions: The DNN Fibrosis Score exhibited sensitivity to histopathologic changes. DNN Fibrosis Scores tended to be lower for participants with higher mRSS (>25). We attribute this divergence to batch effects from staining protocols and may be overcome with analyses of larger cohorts. However, the DNN Fibrosis Score significantly correlated with a set of histopathological variables that were distinct from those correlated with mRSS.

OC.37

VALIDATION OF A LUNG ULTRASOUND INTERPRETATION CRITERIA FOR INTERSTITIAL LUNG DISEASE SCREENING IN PATIENTS WITH SYSTEMIC SCLEROSIS, MIXED CONNECTIVE TISSUE DISEASE, AND INFLAMMATORY MYOPATHY

Robert Fairchild, Diane Mar, Mariani Deluna, Lorinda Chung

Stanford University, Stanford, UNITED STATES MINOR OUTLYING ISLANDS

Introduction: Introduction: Interstitial lung disease (ILD) is highly prevalent in systemic sclerosis (SSc), mixed connective tissue disease (MCTD), and inflammatory myopathies (IM), and early detection is important to reduce damage and improve outcomes. Lung ultrasonography (LUS) is a potential screening tool for ILD with low-cost, potential for point-of-care implementation, and lack of ionizing radiation. We previously developed the first set of LUS interpretation criteria for SSc-ILD screening with excellent sensitivity and specificity. In this study, we validate these criteria in SSc, MCTD, and IM patients.

Material and Methods: Methods: Patients with SSc, MCTD, and IM scheduled to undergo CT chest for any reason were enrolled over a 3-year period to undergo LUS. We followed our previously reported LUS acquisition technique (14 lung positions) and obtained LUS within 10 days of CT scan. LUS interpretation was performed by 3 blinded readers. The sensitivity and specificity for ILD detection was assessed for the overall cohort, patients with “suspected ILD” (no prior ILD diagnosis), and subsets of SSc/MCTD and IM. Agreement among readers was measured using percent concordance and Fleiss Kappa statistic. Spearmen’s correlation was assessed between number of positive lung zones and both percent predicted diffusion capacity (DLCO) and forced vital capacity (%FVC).

Results: Results: We enrolled 100 patients with SSc/MCTD (71%) or IM (29%); 46% were referred for suspected ILD and 46% had a prior diagnosis of ILD. ILD on CT was identified in 66% of participants, with 16% of new diagnoses occurring in the suspected ILD cohort. After excluding 5 patients with non-specific infectious and/or inflammatory findings on CT, LUS was positive for ILD using our interpretation criteria in 71 of the 95 cases with a sensitivity of 97% and specificity of 83% versus CT. LUS identified all new diagnoses of ILD and remained highly sensitive with good specificity across SSc/MCTD and IM cohorts. The sum of positive lung zones across readers inversely correlated with DLCO (-0.433, p<0.0001) and FVC (-0.212, p = 0.030). ILD detection by individual lung zone was consistent across the 3 readers in 93.7% of cases with kappa = 95.8. The average scanning time per patient was 7.3 minutes.

Conclusions: We validated our previously proposed LUS interpretation criteria in a large cohort of patients with SSc, MCTD, and IM, showing high sensitivity and specificity for ILD detection in both cross-sectional and suspected ILD cohorts. These findings support the use of LUS for ILD screening in these populations.

OC.38

DETECTION OF SCLERODERMA FACIAL FEATURES BY ARTIFICIAL INTELLIGENCE: A PILOT STUDY

Yossra Suliman¹, Julius Lund-Jacosen², Daniel Eric Furst ³

¹Rheumatology and rehabilitation Dept, Assiut, Assiut, EGYPT, ²Viceron, Copenhagen, DENMARK, ³Rheumatology division, University of California Los Angeles, California, USA

Introduction: Systemic sclerosis (SSc )is a rare autoimmune multisystem fibrosing disease with high rates of morbidity and mortality . Dermatologists and/or rheumatologists typically diagnose scleroderma. However, since internists and family physicians are unfamiliar with the disease and its symptoms, a delay in their referral of SSc pts to a rheumatologist may result in a delayed disease detection. We hypothesize that facial features of SSc patients are distinctive and can be detected by a trained artificial intelligence (AI) system after processing a mobile phone picture of SSc patient’s face through Convolutional Neural Networks (CNN). This system could be used by family practioner and internists, aiding them to increase suspicion of SSc and referral of patients in a timely manner. We aimed to evaluate the ability of an AI facial recognition system to identify SSc related facial features.

Material and Methods: Photographs of 60 SSc pts from the internet were evaluated versus a group of age and sex matched normal faces . Artificial intelligence (AI) algorithms evaluated all the pixels in the facial map and identified their utility in the facial recognition prediction models, utilizing the Viceron company AI system. (This is a well-established and experienced algorithm for AI facial feature recognition based in Denmark and has been used for facial age recognition in > 1 million general public faces). AI weighed multiple layers of mathematical models, either isolated or pooled to create a predictive model and eliminate unnecessary data. Protocols of smoothing and uniformity were established, through preprocessing with the CNN. Multiple AI models were used as a training set in the first 40 pts and as a partial validation step in the other 20 pts.Multiple models were developped among the 60 SSc images and matched controls that were able to distinguish SSc distinct facial features with variable specificity and sensitivity.

Results: The best model, pre-trained on the development set and fine-tuned on the training set, achieved 80-90% accuracy on the respective datasets.

AI programs, when used on small numbers of patients, develop solutions tailored to the individual patients in the examined data set. A more realistic estimate of accuracy can be derived when estimated for a larger data set, allowing for more generalized facial recognition.

Conclusions: Artificial intelligence algorithms utilized in SSc face identification, gave encouraging pilot results(80-90% accuracy), further testing to develop a larger protocol to establish the effects of race/ethnicity, sex, age, disease duration, disease activity is warranted

OC.39

DRIVERS OF DISEASE BURDEN IN PATIENTS WITH SYSTEMIC SCLEROSIS REPORTED BY THE SCLEROID QUESTIONNAIRE – A CROSS-SECTIONAL ANALYSIS ACROSS CLINICAL AND DEMOGRAPHIC SUBGROUPS

Sinziana Muraru¹, Alexandru Garaiman¹, Kim Fligelstone², Ann Tyrrell Kennedy², Annelise Roennow², Yannick Allanore³, Patricia E Carreira⁴, László Czirják⁵, Christopher Denton⁶, Roger Hesselstrand⁷, Gunnel Sandqvist⁷, Otylia Kowal-Bielecka⁸, Cosimo Bruni^1,9, Marco Matucci-Cerinic⁹, Carmen -Marina Mihai¹, Ana Maria Gheorghiu¹⁰, Ulf Mueller-Ladner¹¹, Tore K Kvien ^12,13, Turid Heidberg¹⁴, Oliver Distler¹, Mike Becker¹, Rucsandra Dobrota¹

¹Department of Rheumatology, University Hospital Zurich, University of Zurich, Zurich, SWITZERLAND, ²Federation of European Scleroderma Associations (FESCA), Saint Maur, BELGIUM, ³Department of Rheumatology, University Paris Descartes and Cochin Hospital, Paris, FRANCE, ⁴Department of Rheumatology, Hospital Universitario 12 de Octubre, Madrid, SPAIN, ⁵Department of Immunology and Rheumatology, Medical School, University of Pécs, Pécs, HUNGARY, ⁶Centre for Rheumatology, University College London, Royal Free Campus, London, UNITED KINGDOM, ⁷Department of Rheumatology, Lund University, Lund, SWEDEN, ⁸Department of Rheumatology and Internal Medicine, Medical University of Bialystok, Bialystok, POLAND, ⁹Division of Rheumatology Careggi University Hospital, University of Florence, Florence, ITALY, ¹⁰Department of Internal Medicine and Rheumatology, Cantacuzino Hospital, Carol Davila University of Medicine and Pharmacy, Bucharest, ROMANIA, ¹¹Department of Rheumatology and Clinical Immunology, Justus-Liebig University Giessen, Campus Kerckhoff, Bad Nauheim, Bad Nauheim, GERMANY, ¹²Center for treatment of Rheumatic and Musculoskeletal Diseases (REMEDY), Diakonhjemmet Hospital, Oslo, NORWAY, ¹³Faculty of Medicine, University of Oslo, Oslo, NORWAY, ¹⁴Regional Research Support, Oslo University Hospital, Oslo, NORWAY

Introduction: Patient-reported outcome measures (PROMs) offer valuable insights into the disease burden of patients living with systemic sclerosis (SSc). A joint team of patients with SSc and medical experts developed ScleroID as a specific disease impact questionnaire for research and clinical use.

In this study, we investigated clinical and demographic associations and drivers of disease burden in patients with SSc, as reported by ScleroID.

Material and Methods: This is a post-hoc, cross-sectional analysis of the initial cohort on which ScleroID (Figure 1A) was developed and validated (previously published), including patients with SSc from 9 European expert centers.

For correlations between ScleroID and continuous variables, we computed the Spearman coefficient. To analyze differences between groups we used the Wilcoxon rank sum or the Kruskal-Wallis test, followed by pairwise comparisons with adjusted significance level (Bonferroni).

Results: We included 471 patients, with a median age of 57 years, 84.7% female, median disease duration of 9 years; 62% of the patients had limited cutaneous SSc.

We observed a strong correlation between the ScleroID score and the global patient reported disease activity (VAS scale) (r= 0.78, p<0.001), but not with the physicians’ global assessement (r= 0.29, p<0.001). Moreover, the ScleroID score showed a good correlation with patient-reported current disease status as assessed by a 5-points Likert scale, with increasing scores from “very good” to “very bad” (Figure 1B).

ScleroID weakly correlated with the first European Scleroderma Study Group index (r= 0.23, p<0.001).

Women had higher global ScleroID scores than men (p=0.001), suggesting a higher disease burden. Among the individual health domains, higher scores for Raynaud`s phenomenon (p=0.001), pain (p<0.001), fatigue (p<0.001) and both upper- (p=0.003) and lower gastrointestinal symptoms (<0.001) were reported in women.

Higher global ScleroID scores were observed in patients with a disease duration >5 years (p=0.011), digital ulcers (p=0.02), dyspnea stages NYHA 3-4 (p<0.001), interstitial lung disease (ILD) (p=0.03), and esophageal symptoms (p<0.001). Interestingly, not only the corresponding ScleroID health dimension, but also multiple others, were affected (Table 1).

There were no significant differences in the ScleroID score between patients with limited versus diffuse SSc.

Conclusions: A disease duration >5years, female gender, digital ulcers, severe dyspnea, ILD and esophageal reflux were the most important drivers of a high disease burden as reflected by higher ScleroID scores. The weak correlation with physician’s reported global disease activity and with disease activity scores highlights the importance of including PROMs in clinical care and research to adequately reflect the patients’ view.

OC.40

EARLY CHARACTERIZATION OF RENAL IMPAIRMENT IN PATIENTS WITH SYSTEMIC SCLEROSIS BY MULTI-PARAMETRIC MAGNETIC RESONANCE QUANTITATIVE IMAGING

Xinyu Tong¹, Huilin He², Zihan Ning³, Rui shen³, Zuoxiang He¹, Dong Xu², Xihai Zhao³

¹School of Clinical Medicine, Tsinghua University, Beijing, CHINA, ²Department of Rheumatology and Clinical Immunology, Peking Union Medical College Hospital, Beijing, CHINA, ³Center for Biomedical Imaging Research, School of Medicine, Tsinghua University, Beijing, CHINA

Introduction: Systemic sclerosis (SSc) is a systemic autoimmune diseases, which leads to fibrosis of multiple organs including kidneys. However, the invasive procedures may increase the risk of complications, and the microstructural impairment may occur prior to clinical symptoms. This study aims to characterize early impairment of renal microstructure in SSc patients using multiparametric MR quantitative imaging.

Material and Methods: Patients who were diagnosed with SSc and age- and sex-matched healthy volunteers were recruited and underwent renal MR imaging. Demographic and clinical characteristics of recruited patients were collected. Renal MR imaging was performed on a 3T MR scanner (Ingenia CX, Philips Healthcare, The Netherlands) with a 16-channel dStream Torso coil and a 12-channel embedded posterior coil. The imaging protocol includes diffusion weighted imaging (DWI), blood oxygen level dependent (BOLD) and saturated multi-delay renal arterial spin labeling. Above renal quantitative measurements were compared among diffuse cutaneous SSc (dcSSc) and limited cutaneous SSc (lcSSc) patient groups and healthy controls using One-way ANOVA and logistic regression analysis.

Results: Forty-six SSc patients and 22 healthy volunteers were recruited in this study. Compared to healthy volunteers, SSc patients with normal eGFR (eGFR>90 mL/min/1.73m2) (n=40) had significantly lower T2* value (p=0.026) in the left renal cortex, longer T1 value (right: p=0.015; left: p=0.023), lower RBF (right: p<0.001; left: p<0.001), and shorter tBAT (right: p<0.001; left: p=0.005) in both right and left renal cortex after adjusting for demographics (Table 1, Figure 1). The dcSSc patients (n=23) had significantly lower RBF in both right (226.7±65.2 mL/100g/min vs. 278.2±73.5 mL/100g/min, p=0.022) and left (194.5±71.5 mL/100g/min vs. 252.7±84.4 mL/100g/min, p =0.020) renal cortex compared to lcSSc patients (n=23) after adjusting for demographics, but the significance of the difference was attenuated after further adjusting for modified Rodnan skin score and digital ulcers. No significant differences were found in other quantitative measurements (p>0.05).

Conclusions: SSc patients with normal eGFR had significantly lower T2* value, longer T1 value, lower RBF, and shorter tBAT of renal cortex than healthy volunteers. According to the previous studies, the increase of T1 value may be associated with vasculitis or adventitial and peri-adventitial fibrosis, while the decrease of renal RBF and T2* may be related to intimal thickening due to endothelial cell injury and episodic vasospasm named renal Raynaud’s phenomenon. Only tBAT instead of aBAT appeared a significant decrease in SSc patients, which may imply the impairment of renal microcirculation. Multi-parametric MR quantitative imaging might be a more sensitive modality in characterizing early impairment of renal microstructure.

OC.41

EFFICIENCY OF ANNUAL PULMONARY ARTERIAL HYPERTENSION SCREENING BY ECHOCARDIOGRAPHY IN PATIENTS WITH SYSTEMIC SCLEROSIS

Yuichiro Shirai¹, Hironari Hanaoka², Masataka Kuwana^1,2

¹Nippon Medical School, Tokyo, JAPAN, ²Keio University, Tokyo, JAPAN

Introduction: Current guidelines recommend annual screening of pulmonary arterial hypertension (PAH) by transthoracic echocardiogram (TTE) in patients with systemic sclerosis (SSc), but evidence to support this statement is scarce. This study was aimed to examine prevalence of newly diagnosed PAH at serial screening, and initial predictors for future development of PAH using a single-centre SSc cohort.

Material and Methods: This study enrolled 477 consecutive SSc patients who visited our center between 2007 and 2022 without definitive diagnosis of pulmonary hypertension (PH), and underwent serial PAH screening at an interval of <2 years. Estimated systolic pulmonary arterial pressure (esPAP) was calculated from tricuspid regurgitation velocity and presumptive right atrial pressure. Patients who had esPAP >50 mmHg, esPAP 37-50 mmHg with echocardiographic variables suggestive of PH, or unexplained dyspnea were subjected to right heart catheterization. The mean pulmonary arterial pressure >=25 mmHg at rest was regarded as having PH, and clinical classification of PH was made by comprehensive evaluations. Univariate and multivariable analyses were conducted to identify initial predictors for future development of PAH. Receiver-operating characteristic curve (ROC) analysis was performed to determine an optimal cut-off.

Results: A total of 1,600 TTE screening was conducted with a median of 3 times (IQR 1-5). PAH was newly diagnosed in 12/477 patients (2.5%) at first screening, and 4/347 (1.2%) at second screening, but PAH diagnosis was made in 2/776 (0.3%) at third to tenth screening combined together. In addition, incident PAH was diagnosed in 3 patients due to worsening of dyspnea prior to the next screening timing. In the univariate analysis, higher esPAP and lower percent predicted diffusing capacity for carbon monoxide (DLCO) were identified as initial predictors for future PAH development. Multivariable analysis found that DLCO was the primary contributor (hazard ratio 0.941, 95% confidence interval 0.909-0.973), while esPAP was the subordinate contributor (hazard ratio 1.069, 95% confidence interval 0.997-1.148). ROC analysis revealed optimal cut-off values of esPAP and DLCO as 27 mmHg and 53%, respectively. When patients were stratified into 4 groups based on above or below optimal cut-offs of esPAP and DLCO at initial screening, cumulative rates of no PAH development were significantly lower in a group of esPAP >=27 mmHg and DLCO =<53% than in other groups.

Conclusions: Prevalence of newly diagnosed PAH at serial screening was decreased with time. Intervals of PAH screening need to be adjusted based on initial risks for PAH development.

OC.42

CLINICAL SIGNIFICANCE OF ANTI-RO/SSA ANTIBODIES IN PATIENTS WITH SYSTEMIC SCLEROSIS: A STUDY FROM THE EUSTAR DATABASE

Blaz Burja¹, Marouane Boubaya², Patricia E. Carreira³, Christina Bergmann⁴, Lidia P. Ananieva⁵, Gabriela Riemerkasten⁶, Okada Masado⁷, Jeska De Vries-Bouwstra⁸, Edoardo Rosato⁹, Marie -Elise Truchetet¹⁰, Nicoletta Del Papa¹¹, Antonella Marcocccia¹², Fabiola Atzeni¹³, Tim Schmeiser¹⁵, Madelon Vonk¹⁵, Francesco Del Galdo¹⁶, Oliver Distler¹, Muriel Elhai¹

¹Center of Experimental Rheumatology, Department of Rheumatology, University Hospital Zurich, University of Zurich, Zurich, SWITZERLAND, ²Department of Clinical Research, CHU Avicenne, APHP, Bobigny, FRANCE, ³Rheumatology Department, Hospital Universitario 12 De Octubre, Madrid, SPAIN, ⁴Department Internal Medicine, University Hospital Erlangen, Erlangen, GERMANY, ⁵V.A. Nasonova Research Institute Of Rheumatology Russian Federation, Moscow, RUSSIA, ⁶Klinik Für Rheumatologie Und Klinische Immunologie, Universitätsklinikum Schleswig-Holstein, Lübeck, GERMANY, ⁷St. Luke's International Hospital Immuno-Rheumatology Center, Tokyo, JAPAN, ⁸Center Department Of Rheumatology, Leiden University Medical Center, Leiden, THE NETHERLANDS, ⁹Sapienza University Of Rome-Department Of Translational And Precision Medicine, Rome, ITALY, ¹⁰Chu De Bordeaux, Rheumatology Department, Bordeaux, FRANCE, ¹¹Scleroderma Clinic, UOC Reumatologia Clinica, ASST G. Pini-CTO, Milan, ITALY, ¹²Centro Di Riferimento Interdisciplinare Per La Sclerosi Sistemica, Rome, ITALY, ¹³Rheumatology Unit, University Of Messina, Messina, ITALY, ¹⁴Krankenhaus St. Josef, Wuppertal-Elberfeld, GERMANY, ¹⁵Radbound University Medical Centre, Nijmengen, THE NETHERLANDS, ¹⁶Leeds Institute of Rheumatic and Musculoskeletal Medicine, Faculty of Medicine and Health, University of Leeds, Leeds, UNITED KINGDOM

Introduction: Despite the progress in understanding the patient heterogeneity in systemic sclerosis (SSc) based on SSc-specific antibodies, better risk stratification is needed. SSc non-specific antibodies might represent surrogate markers to improve the stratification of SSc patients. Here, we aim to evaluate the prevalence of anti-Ro/SSA antibodies in the largest available cohort of established SSc patients and study their association with disease phenotype and clinical outcomes, focusing on lung involvement.

Material and Methods: Patients from the EUSTAR database fulfilling the ACR 2013 classification criteria for SSc with available data on anti-Ro/SSA antibodies were included. Clinical characteristics of patients with or without anti-Ro/SSA antibodies were compared at baseline using t-test and chi-squared according to the distribution of the variable. The progression of lung fibrosis was defined (i) in patients with lung fibrosis (FVC% decline from baseline of > 10% or an FVC% decline of 5-9% in association with a DLCO% decline of > 15%) or (ii) by a decline of FVC >5% in patients with lung fibrosis or (iii) by the development of lung fibrosis de novo (on HRCT scan). Prognostic factors for lung fibrosis progression and death during the follow-up were tested by multivariate Cox proportional hazards regression. Covariates were selected according to literature evidence. Multiple imputation was used to impute missing data in these models.

Results: Among the 4.421 patients fulfilling the inclusion criteria, 661 (15.2%) had positive anti-Ro/SSA antibodies. Anti-Ro/SSA antibodies were more frequently observed among Asians and Africans and less prevalent in Caucasians (Table 1). Anti-Ro/SSA antibodies were positively associated (p<0.001) with anti-SSB, anti-U1RNP antibodies, and rheumatoid factor. Patients with anti-Ro/SSA antibodies more frequently presented with muscular involvement (18% vs 12.5%, p<0.001), PAPs>45 mmHg on echocardiography (9.2% vs. 6.5%, p=0.058) and lung fibrosis on HRCT (56.2% vs 47.8%, p=0.001). Specifically, the percentage predicted of DLCO in patients with lung fibrosis was significantly lower in patients with anti-SSA antibodies (59.0±18.6% vs. 61.9±20.2, p=0.041). Over a median follow-up of 2.4 years [95CI: 2.2-2.9], anti-SSA antibodies did not predict lung fibrosis progression ((i) HR: 1.03 [0.8-1.33], (ii) HR: 1.05 [0.83-1.31] and (iii) HR: 0.98 [0.72-1.32] or death (HR: 1.27 [0.8-2]).

Conclusions: In the large EUSTAR cohort, anti-SSA antibodies are detected in 15% of SSc-patients and are associated with more severe lung involvement. These data support the inclusion of anti-SSA antibodies in clinical practice for better SSc-patient risk stratification.

OC.43

TREATMENT REGIMENS AND MORTALITY IN SYSTEMIC SCLEROSIS-ASSOCIATED PULMONARY ARTERIAL HYPERTENSION IN LIGHT OF THE 2022 ESC/ERS GUIDELINES

Hilde Jenssen Bjørkekjær^1,2, Cosimo Bruni^3,4, Cathrine Brunborg^2,5, Patricia Carreira⁶, Paolo Air⁷, Carmen Pilar Simeón-Aznar⁸, Marie -Elise Truchetet⁹, Alessandro Giollo¹⁰, Alexandra Balbir-Gurman¹¹, Mickael Martin¹², Christopher Denton¹³, Armando Gabrielli¹⁴, Håvard Fretheim¹⁵, Imon Barua^2,15, Helle Bitter¹, Øyvind Midtvedt¹⁵, Torhild Garen¹⁵, Kaspar Broch^16,17, Arne Andreassen¹⁶, Yoshiya Tanaka¹⁸, Gabriela Riemekasten¹⁹, Ulf Müller-Ladner²⁰, Oliver Distler⁴, Anna -Maria Hoffmann-Vold^4,15 Marco Matucci-Cerinic^3,21, Ivan Castellví²², Elise Siegert²³, Eric Hachulla²⁴

¹Hospital of Southern Norway, Kristiansand, Department of Rheumatology, Kristiansand, NORWAY, ²University of Oslo, Institute of Clinical Medicine, Oslo, NORWAY, ³Azienda Ospedaliero Universitaria Careggi, University of Florence, Department of Experimental and Clinical Medicine, Division of Rheumatology, Florence, ITALY, ⁴University Hospital Zurich, University of Zurich, Department of Rheumatology, Zurich, SWITZERLAND, ⁵Oslo Centre for Biostatistics and Epidemiology, Research Support Services, Oslo University Hospital, Rikshospitalet, Oslo, NORWAY, ⁶12 de Octubre University Hospital, Department of Rheumatology, Madrid, SPAIN, ⁷9 Spedali Civili di Brescia, Scleroderma UNIT, UOC Reumatologia ed Immunologia Clinica, Brescia, ITALY, ⁸University Vall d'Hebron Hospital, Systemic Autoimmune Diseases Unit. Department of Internal Medicine, Barcelona, SPAIN, ⁹Bordeaux University Hospital, Rheumatology Department, Bordeaux, FRANCE, ¹⁰University of Verona, Rheumatology Section, Department of Medicine, Verona, ITALY, ¹¹Rambam Health Care Campus, Rappaport Faculty of Medicine, Technion-Institute of Technology, B Shine Rheumatology Institute, Haifa, ISRAEL, ¹²Poitiers University Hospital, Department of Internal Medicine, Poitiers, FRANCE, ¹³University College London Division of Medicine and Royal Free Hospital, Centre for Rheumatology and Connective Tissue Diseases, London, UNITED KINGDOM, ¹⁴Fondazione di Medicina Molecolare e Terapia Cellulare, Università Politecnica delle Marche, Ancona, ITALY, ¹⁵Oslo University Hospital, Rikshospitalet, Department of Rheumatology, Oslo, NORWAY, ¹⁶Oslo University Hospital, Rikshospitalet, Department of Cardiology, Oslo, NORWAY, ¹⁷KG Jebsen center, Institute for Experimental Medical Research, University of Oslo, Oslo, NORWAY, ¹⁸University of Occupational and Environmental Health, Japan, First Department of Internal Medicine, Kitakyushu, JAPAN, ¹⁹University of Lübeck, Department of Rheumatology and Clinical Immunology, Lübeck, GERMANY, ²⁰Justus-Liebig University Giessen, Campus Kerckhoff, Dept of Rheumatology and Clinical Immunology, Bad Nauheim, GERMANY, ²¹Hospital San Raffaele Hospital, UNIRAR, Milan, ITALY, ²²Hospital de la Santa Creu i Sant Pau, Department of Rheumatology, Barcelona, SPAIN, ²³Charit University Hospital, Rheumatology, Berlin, GERMANY, ²⁴Department of Internal Medicine and Clinical Immunology, Referral Centre for Centre for rare systemic autoimmune diseases North and North-West of France (CeRAINO), CHU Lille, Univ. Lille, Inserm, U1286 - INFINITE - Institute for Translational Research in Inflammation, F-59000 Lille, FRANCE

Introduction: The 2022 ESC/ERS Guidelines recommend upfront dual combination therapy for low- and intermediate-risk patients and triple therapy for high-risk patients with systemic sclerosis-associated pulmonary arterial hypertension (SSc-PAH). There is no treatment recommendation for patients with mean pulmonary arterial pressure (mPAP) of 21–24 mmHg and pulmonary vascular resistance (PVR) of 2–3 Wood units (WU). We aimed to assess current treatment regimens, stratified by hemodynamic thresholds and risk stratification, and their association with mortality.

Material and Methods: We included patients in the EUSTAR database who fulfilled the criteria for SSc-PAH according to the new hemodynamic definition (Project Number: CP122). Patients who were already on PAH-specific treatment or who had clinically significant interstitial lung disease (ILD) were excluded. Initial treatment strategies were defined as (1) upfront monotherapy with endothelin receptor antagonists, phosphodiesterase-5 inhibitors, or prostanoids; (2) upfront combination therapy; or (3) no therapy. We assessed treatment regimens stratified by mPAP and PVR thresholds (lower threshold: mPAP 21-24 mmHg or PVR 2-3 WU vs. higher threshold: mPAP >=25 mmHg and PVR >=3 WU) and by risk as assessed by COMPERA, which allocates patients to one of four risk groups. We assessed the impact of the initial treatment regimens on mortality using Cox regression adjusted for age, sex, DLCO, mPAP and PVR thresholds, pre-existing treatment for digital ulcers and Raynaud’s, and risk group. Transplant-free survival was evaluated using Kaplan-Meier analysis and the log-rank test.

Results: Of 890 patients who had right heart catheterization, 366 were eligible (Table). Upfront combination therapy was used more frequently in patients with higher vs. lower mPAP and PVR thresholds and in patients in the high-risk group vs. lower risk groups (Figure 1A-B). In multivariable Cox regression analysis, upfront monotherapy was significantly associated with lower mortality compared with no treatment. (Figure 2). This result was independent of higher mPAP and PVR thresholds and intermediate-high and high-risk groups, both of which were significantly associated with mortality. Upfront combination therapy seemed to be associated with lower mortality compared with no treatment, but this result was not statistically significant. Survival by hemodynamic thresholds and COMPERA risk group are presented in Figure 3.

Conclusions: These results show that few patients have received the currently recommended upfront treatment, regardless of hemodynamic thresholds and risk stratification. Our findings indicate that upfront treatment is associated with improved survival, independent of higher mPAP and PVR thresholds and intermediate-high and high-risk groups. Based on these results, we suggest considering more aggressive SSc-PAH treatment.

OC.44

PREDICTORS OF SYSTEMIC SCLEROSIS-RELATED PRIMARY HEART INVOLVEMENT AT CARDIAC MAGNETIC RESONANCE: A MONOCENTRIC RETROSPECTIVE STUDY

Veronica Batani¹, Corrado Campochiaro¹, Anna Palmisano², Gaia Mancuso¹, Davide Vignale², Lorenzo Dagna¹, Antonio Esposito², Marco Matucci Cerinic¹, Giacomo De Luca¹

¹UnIRAR, Vita-Salute San Raffaele University & San Raffaele Hospital, Milan, ITALY, ²Radiology Unit, ita-Salute San Raffaele University & San Raffaele Hospital, Milan, ITALY

Introduction: Primary heart involvement(pHI) is frequent in systemic sclerosis(SSc) and it is associated with a bad prognosis, therefore early recognition is pivotal. Cardiac magnetic resonance(CMR) represents the non-invasive tool of choice for diagnosing pHI-SSc. The aim of the study was to identify the predictors of pHI-SSc at CMR in a large monocentric cohort.

Material and Methods: Among SSc patients followed-up at our centre, those who underwent CMR for suspected pHI (clinical signs and/or symptoms or instrumental signs suggestive for pHI) were identified. Clinical, laboratory and instrumental features were compared between pHI-SSc patients and non-pHI-SSc patients and between CMR-positive and CMR-negative patients. Univariable and multivariable logistic regression were used to identify predictors of pHI-SSc at CMR.

Results: 259 SSc patients (females 89.6%;median age 49[39-61]years; median disease duration 99[60-153]months; diffuse disease 17%; anti-topoisomerase-I 28.2%) were included. CMR was performed in 56 patients (21%), and it was diagnostic for pHI in 37 of them(66%). Comparative features of pHI-SSc and non-pHI-SSc patients are summarized in Table-1.

At univariate analysis, age at SSc onset, diffuse cutaneous subset, increase of inflammatory markers, cardiac enzymes and creatinphosphokinase, and interstitial lung disease(ILD) were significantly associated with SSc-pHI, whereas an inverse correlation emerged with anti-centromere antibodies, VEDOSS and female gender. At multivariate analysis, increased serum levels of troponin T (TnT[OR 14.108, CI 2.772-71.808,p=0.014]) and C-reactive-protein (CRP[OR 11.237, CI 2.265-55.763,p=0.0031]), and presence of ILD(OR 6.275, CI 1.240-31.763,p=0.0264]), emerged as predictors of pHI-SSc.

In the subgroup of 56 patients who underwent CMR, only the presence of an arrhythmic event at 24h-ECG-Holter emerged at univariate analysis as a predictor of pHI-SSc (OR 3.75, IC 1.041-13.5,p=0.04).

Considering the 37 patients with positive CMR, signs of active myocardial inflammation (myocardial edema, increased T2 mapping and/or early gadolinium enhancement[EGE]) were disclosed in 20 cases(54%), while 17 patients(46%) had signs suggestive for chronic involvement/fibrotic changes (increased T1 mapping and/or extracellular volume[ECV], late gadolinium enhancement[LGE] with non-ischemic pattern) (Figure 1).

In the subgroup of 37 patients with positive CMR, an arrhythmic presentation, disclosed in 17(46%) patients, emerged at univariate analysis as a predictor of chronic/fibrotic changes at CMR (OR 3.750, CI 1.108-12.694,p=0.003). No other associations were found between clinical features and CMR patterns of active or chronic myocardial involvement.

Conclusions: Increased serum levels of TnT and CRP, and the presence of ILD and arrhythmias at 24h-ECG Holter should be considered as red flags for pHI in SSc patients. The presence of arrhythmic burden is associated with a chronic myocardial damage at CMR.

OC.45

RITUXIMAB RETENTION RATE IN SYSTEMIC SCLEROSIS: A REAL-LIFE ITALIAN MULTICENTER STUDY

Giacomo De Luca¹, Enrico De Lorenzis², Corrado Campochiaro¹, Fabio Cacciapaglia³, Nicoletta Del Papa⁴, Elisabetta Zanatta⁵, Paolo Airo'⁶, Maria Grazia Lazzaroni⁶, Dilia Giuggioli⁷, Maria De Santis⁸, Gabriella Alonzi², Stefano Stano³, Marco Binda⁵, Beatrice Moccaldi⁵, Antonio Tonutti⁸, Veronica Batani¹, Silvia Cavalli⁴, Florenzo Iannone³, Maria Antonietta D'agostino², Lorenzo Dagna¹, Marco Matucci Cerinic¹, Silvia Laura Bosello²

¹Scleroderma Unit, Vita-Salute San Raffaele University & San Raffaele Hospital, Milan, ITALY, ²Division of Rheumatology, Fondazione Policlinico Universitario A. Gemelli IRCCS, Rome, ITALY, ³Rheumatology Unit - DiMePRe-J University of Bari, Bari, ITALY, ⁴Scleroderma Clinic, ASST Gaetano Pini-CTO, Milan, ITALY, ⁵Rheumatology Unit, Padova University Hospital, Padua, ITALY, ⁶UOC Reumatology and Clinical Immunology, ASST Spedali Civili, Brescia, ITALY, ⁷Scleroderma Unit, University of Modena and Reggio Emilia, Modena, ITALY, ⁸Rheumatology and Clinical Immunology, IRCCS Humanitas Research Hospital, Rozzano, ITALY

Introduction: Rituximab(RTX) has been widely used for the management of systemic sclerosis(SSc), thanks to accumulating evidence about its efficacy for skin and lung fibrosis in some disease subsets. Long-term data on RTX safety and efficacy are limited. We present the real-life RTX retention rate as an indicator of safety and efficacy in a multicentric cohort of Italian SSc patients.

Material and Methods: SSc patients treated with RTX and with a follow-up for at least 36 months were included and clinically characterized. Patients were longitudinally monitored for up to 5 years. A competing risk analysis with sub-Hazard Ratio(sHR) definition was performed to explore the clinical variables linked to specific cause of RTX discontinuation.

Results: 152 SSc patients (mean age 47.3±12.3 years; females 79.6%; diffuse cutaneous SSc[dcSSc] 77.6%) were evaluated for a median(IQR) time of 3.3(1.7-5.0) years (Table 1). The main reasons for RTX prescription were lung fibrosis(38.8%), worsening of skin fibrosis(36.8%), and synovitis(13.8%); 138 patients(90.8%) were treated with more than 1 RTX course(median 4[3-8] courses). Five-year RTX retention rate was 59.9%(44.6-64.7%). A clinical response was the most common cause of RTX discontinuation[5.7(3.7-8.4) per 100 patient-year] and was associated with a shorter disease duration[sHR 0.8(0.7-0.9)], anti-Scl70 negativity[sHR 0.4(0.2-0.9)], previous digital ulcers[sHR 2.6(1.1-6.2] and no history of synovitis[sHR 0.3 (0.1-0.8)]. At 6 months and at final follow-up, skin score significantly improved either in the entire cohort(mRSS: 11.9±8.0 and 8.3±7.4 vs 16.5±9.6) and in dcSSc patients with progressive skin involvement (mRSS: 15.6±9.6 and 9.2±7.9 vs 21.0±10.8)(p<0.001 for all). At 6 months and at final follow-up, mean predicted(%) FVC and DLCO remained stable both in the entire cohort(FVC: 87.3±21.5 and 87.0±23.5 vs 87.4±19.8; DLCO: 65.2±20.9 and 61.1±19.6 vs 62.1±19.7) and in ILD patients(FVC: 84.7±19.2 and 82.5±21.6 vs 81.6±17.8; DLCO: 64.5±19.0 and 57.9±18.5 vs 60.6±18.7)(p=ns for all).

Treatment failure was the second cause of RTX discontinuation [3.7(2.2-6.0) per 100 patient-year], and was associated with anti-centromere antibody positivity[sHR 2.8(1.1-7.4)] and anti-Scl70 negativity[sHR 0.2(0.1-0.6)]. Occurrence of adverse events (7 infusion reactions, 4 severe infections, 2 cardiovascular events and one cytopenia) was the less common cause of RTX discontinuation[3.1(1.7-5.2) per 100 patient-year], associated with limited cutaneous subset [sHR 3.4(1.2-9.7)] and previous mycophenolate mofetil treatment[sHR 4.5(1.2-16.3)].

Conclusions: Our data showed that RTX helps in controlling skin and lung involvement in SSc patients. RTX showed a good retention rate and safety is satisfactory in a real life setting for the treatment of SSc. Indeed, in the long-term follow-up, only few adverse events were shown to limit treatment retention.

OC.46

CHANGES IN TREATMENT PATTERNS AND THEIR INFLUENCE ON THE OUTCOME OF SYSTEMIC SCLEROSIS-INTERSTITIAL LUNG DISEASE (SSC-ILD) PATIENTS: AN EUSTAR ANALYSIS

Corrado Campochiaro¹, Marie -Elise Truchetet², Madelon Vonk³, Giovanna Cuomo⁴, Lidia P Ananieva⁵, Eric Hachulla⁶, Vanessa Smith⁷, Ana Maria Gheorghiu⁸, Radim Becvar⁹, Patricia Carreira¹⁰, Nicolas Hunzelmann¹¹, Daniel Furst¹², Vera Ortiz-Santamaria¹³, Francesco Del Galdo¹⁴, Marco Matucci-Cerinic¹⁵, Anna -Maria Hoffmann-Vold¹⁶

¹IRCCS San Raffaele Hospital, Milano, ITALY, ² Bordeaux Hospital, Bordeaux, FRANCE, ³Radboud University, Nijmegen, THE NETHERLANDS, ⁴Naples University, Naples, ITALY, ⁵Nasonova Reserach Institute, Moscow, RUSSIA, ⁶Lille University, Lille, FRANCE, ⁷University of Ghent, Ghent, FRANCE, ⁸Cantacuzino Clinical Hospital, Bucarest, ROMANIA, ⁹Charles Univerisity, Prague, CZECH REPUBLIC, ¹⁰Universidad Complutense de Madrid, Madrid, SPAIN, ¹¹Cologne University, Cologne, GERMANY, ¹²University of Los Angeles, Los Angeles, USA, ¹³Universitat Internacional de Catalunya, Barcelona, SPAIN, ¹⁴University of Leeds, Leeds, UNITED KINGDOM, ¹⁵Florence University, Florence, ITALY, ¹⁶Oslo University, Oslo, NORWAY

Introduction: The immunosuppressive treatments(ISTs) for interstitial lung disease(ILD) in systemic sclerosis(SSc) has expanded over the past decades. The implementation in clinical practice and the impact of IST on ILD progression is to date unclear. This is of importance for the clinical management and inclusion of patients into clinical trials. Here, we assess treatment regimens in SSc-ILD over time and their impact on ILD progression and progression-free survival

Material and Methods: SSc patients registered in the EUSTAR database with presence of ILD on imaging, available FVC%, DLCO% predicted and treatments at baseline and at 12±3 months were included. We segregated treatment regimens into 4 periods based on patient’s first assessment: 1)<=2006; 2)2007-2011; 3)2012-2016; 4)<=2017 and assessed clinical characteristics and type of IST. Next, we evaluated the impact of ISTs across a 3-year follow-up period on the 2 outcomes, progressive ILD events and progression free survival. Absolute decline of FVC>=5% or DLCO>=10% over 12±3 months was defined as a progressive event. Progression-free survival analysis was defined as survival without progression within 36 months. Next, we assessed the impact of ISTs in anti-topoisomerase I (ATA) patients to enrich for progressors.

Results: 1409 SSc–ILD patients were included across the 4 periods: 236(17%); 558(40%); 338(42%) and 277(20%). The clinical characteristics were comparable (Table 1). The number of patients on ISTs changed significantly from 13.6 %(period 1) to 57.4%(period 4) as well as the type of IST, the frequency of switchs and stops in ISTs (Figure 1). When assessing the impact of ISTs over the 3-year follow-up(excluding period 1 due to low number of patients on IST) we identified significantly less progressive events: 115/540(21.3%), 146/1061(13.8%) and 160/1320(12.1%) in period 2, 3 and 4, respectively(p <0.001). We identified improved progression-free survival from period 2 to 3 and 4: 46.3%, 64.9% and 55.0%;p=0.007(Figure 2A). In the subgroup analysis of ATA+ SSc-ILD, we included 786(56%) patients across the 4 periods(136, 311, 185 and 154). Excluding period 1, we identified significantly less progressive events on IST: 73/318(23.0%), 84/563(14.9%) and 83/654(12.7%) in period 2, 3 and 4(p<0.001). Progression-free survival improved from 46.9%, to 67.1% and 57.6% in period 2, 3 and 4;p=0.045(Figure 2B).

Conclusions: The majority of SSc-ILD patients are currently started on IST. Over time there has been a significant change in treatments regimens. This has reduced the number of progressive ILD events, but progression-free survival is still unsatisfying (45% at 3 years). This represents a clear need for the development of novel treatment options and treatment regimens.

OC.47

CHASING CARS: CLINICAL FOLLOW UP DATA ON DISEASE ACTIVITY, SEROLOGICAL MARKERS AND ORGAN INVOLVEMENT OF PATIENTS WITH SEVERE SYSTEMIC SCLEROSIS AFTER TREATMENT WITH CD19-TARGETING CAR T CELLS

Janina Auth^1,6, Fabian Müller^2,6, Jule Taubmann^1,6, Ioanna Minopoulou^1,6, Simon Völkl^2,6, Michael Aigner^2,6, Armin Atzinger³, Nadine Bayerl⁴, Andreas Wirsching^1,6, Clara Dees^1,6, Andrea -Hermina Györfi⁵, Jörg Distler⁵, Andreas Mackensen^2,6, Georg Schett^1,6, Christina Bergmann^1,6

¹University Hospital Erlangen - Department of Rheumatology and Immunology, Erlangen, GERMANY, ²University Hospital Erlangen - Department of Haematology and Oncology, Erlangen, GERMANY, ³University Hospital Erlangen - Nuclear Medicine, Erlangen, GERMANY, ⁴University Hospital Erlangen - Radiology, Erlangen, GERMANY, ⁵University Hospital Düsseldorf - Department of Rheumatology, Düsseldorf, GERMANY, ⁶University Hospital Erlangen - Deutsches Zentrum Immuntherapie (DZI), Erlangen, GERMANY

Introduction: CD19-targeting CAR-T-cells showed remarkable improvements in autoimmune diseases including refractory lupus erythematodes and inflammatory myopathy and first data of a patient with diffuse Systemic Sclerosis (SSc) have recently been reported. Here we present 4 months to 1 year follow up data of the first four patients with severe diffuse SSc who received CAR-T-cell treatment.

Material and Methods: T-cells were acquired by peripheral blood apheresis and transfected with a lentiviral vector encoding a CAR against CD19 (Miltenyi Biotec) using the CliniMacs Prodigy system (Mitenyi Biotec). Immunosuppression was stopped before CAR-T-cell infusion and the therapy was performed upon lymphodepletion with fludarabine (25 mg/m2 on days -5, -4, -3) and cyclophosphamide (1g/m2 on day -3, 50% dose reduction in patient 3 due to renal insufficiency) as single infusion. Outcomes were assessed before baseline and regularly during the follow up period.

Results: Four patients who had failed state of the art SSc treatments were treated with CD19-targeting CAR-T-cells: patient 1 (male, 60 years, mRSS 24, lung- and myocardial involvement), patient 2 (male, 36 years, mRSS 31, lung- and myocardial involvement), patient 3 (female, 37 years, mRSS 32, lung- and kidney involvement) and patient 4 (male, 47 years, mRSS 17, lung involvement). CAR-T-cell expansion peaked between day 8 and 14. B cells were completely absent in peripheral blood within 7 days and returned after 80 to 170 days (still absent in patient 4 after 50 days). ANA titres were reduced by 10- and 30-fold in patients 2 and 3, patient 1 converted to negative ANA status and patient 4 maintained a stable titre so far throughout the follow up period. Anti-RNAP-III antibodies in patient 1 turned negative, whereas anti-Scl70 antibodies in patients 2, 3 and 4 were reduced by 50% to 75% as determined by ELISA analysis. mRSS in all four patients showed a reduction by 30% to 50% within the first 3 months after treatment and the EUSTAR activity index improved by up to 5,75 points during follow up so far. Lung function parameters remained stable in all patients and average 68GA-FAPI-04-uptake in lung and heart as an indicator of fibroblast activity was reduced by 55% and 42% respectively within 1 to 8 months after therapy.

Conclusions: These data on the first four SSc patients receiving CD19-targeting CAR-T-cell therapy show that the procedure can lead to stabilization of SSc disease activity without additional immunosuppression during 4 months to 1 year follow up.

OC.48

OUTCOMES IN PROGRESSIVE SYSTEMIC SCLEROSIS TREATED WITH AUTOLOGOUS HEMATOPOIETIC STEM CELL TRANSPLANTATION COMPARED WITH COMBINATION THERAPY

Doron Rimar¹, Israel Henig², Tsila Zuckerman², Gleb Slobodin¹, Lisa Kaly¹, Abid Awisat¹, Aniela Shouval¹, Michael Rozenbaum¹, Itzhak Rosner¹, Yair Molad³, Firas Sabbas⁴, Mohammad Naffaa⁵, Ariela Dortort LAzar³, Emilia Hardak¹, Shiri Keret¹

¹Bnai_ZIon Medical Center, Haifa, ISRAEL, ²Rambam Medical Center, Haifa, ISRAEL, ³Rabin Medical center, Petah Tikva, ISRAEL, ⁴Baruch Padeh medical Center, Poriya, ISRAEL, ⁵Galil Maaravi Medical center, Naharia, ISRAEL

Introduction: Autologous hematopoietic stem cell transplantation (AHSCT) has been shown to improve long-term survival for early diffuse progressive systemic sclerosis (SSc), compared to cyclophosphamide. Cyclophosphamide, however, does not provide a long-term benefit in SSc. The combination of mycophenolate mofetil (MMF) and rituximab is a potent alternative regimen.

The aim of our study was to retrospectively compare the outcomes of SSc patients who underwent AHSCT to patients who met the eligibility criteria for AHSCT but received upfront combination therapy with MMF and rituximab because of personal preferences.

Material and Methods: SSc patients in our cohort, eligible for transplantation but treated with combination therapy of Rituximab and MMF, were compared to SSc patients who underwent AHSCT. Repeated longitudinal assessments at baseline and every 6 months for 24 months were compared between groups using a linear mixed model. Clinical improvement (CI) at 12 months was defined as a decrease in mRSS by more than 25% or an increase in FVC by more than 10%. Event-free survival (EFS) was defined as until the occurrence of death, or the development of persistent major organ failure (heart, lung, kidney).

Results: Twenty-one SSc patients in the combination therapy group were compared to sixteen in the AHSCT group. Age, sex and disease duration were similar between the two groups (table 1). Clinical improvement at 12 months was seen in 18 (86%) patients in the combination group, compared to 13 (81%) in the AHSCT group (p=0.7). The hazard ratio for EFS at 24 months favored the combination group (HR=0.09, P =0.04). During follow-up, both groups exhibited a significant and comparable reduction in mRSS and an increase in FVC values at each time interval up to 24 months (Figure 1). Treatment-related mortality after AHSCT was 3/16 (18.7%). Overall, there were 43 adverse events (AE) in the AHSCT group (2.7±0.5 per patient), vs. 25 in the combination group (1.2±0.4 per patient) (p<0.001)

Conclusions: MMF and rituximab combination therapy compared to AHSCT in SSc patients eligible for AHSCT, resulted in similar skin and lung clinical improvement with a better safety profile at 24 months. We suggest that CYC can no longer be used as the gold standard in clinical trials. Future prospective RCT are needed to evaluate AHSCT compared to rituximab, MMF and nintedanib combination therapy.