A BAYESIAN APPROACH TO DETERMINE THE ROLE OF ORAL ANTICOAGULANTS FOR THE OCCURRENCE OF DIGITAL ULCERS IN SYSTEMIC SCLEROSIS – A EUSTAR OBSERVATIONAL STUDY

A. Garaiman, C. Gebhard, C. Mihai, R. Dobrota, C. Bruni, M. Matucci-Cerinic, J. Henes, J. De Vries-Bouwstra, V. Smith, A. Doria, Y. Allanore, L. Dagna, B. Anic, C. Montecucco, O. Kowal-Bielecka, M. Martin, Y. Tanaka, A. M. Hoffmann-Vold, O. Distler, M. O. BeckerUniversity Hospital of Zürich, University of Zurich, Department of Rheumatology, Zürich, Switzerland University Hospital Zurich, University of Zurich, Department of Nuclear Medicine, Zurich, Switzerland University of Zurich, Center for Molecular Cardiology, Zurich, Switzerland University Hospital Bern, University of Bern, Department of Cardiology, Zurich, Switzerland University of Florence, Scleroderma Unit, AOUC, Department of Experimental and Clinical Medicine, Division of Rheumatology, Florence, Italy IRCCS San Raffaele Scientific Institute, Unit of Immunology, Rheumatology, Allergy and Rare Diseases (UnIRAR), Milan, Switzerland University Hospital Tuebingen, Centre for Interdisciplinary Clinical Immunology, Rheumatology and Autoinflammatory Diseases and Department of Internal Medicine II (Hematology, Oncology, Immunology and Rheumatology), Tuebingen, Germany Leiden University Medical Center, Department of Rheumatology, Leiden, Netherlands Ghent University, Ghent University Hospital, Department of Rheumatology, Department of Internal Medicine, Ghent, Belgium University of Padova, Rheumatology Unit, Department of Medicine, Padova, Italy Descartes University, APHP, Cochin Hospital, Department of Rheumatology, Paris, France IRCCS San Raffaele Scientific, Institute; Vita-Salute San Raffaele University, Unit of Immunology, Rheumatology, Allergy and Rare Diseases (UnIRAR), Milan, Italy University Hospital Centre Zagreb and University of Zagreb, School of Medicine, Division of Clinical Immunology and Rheumatology, Department of Internal Medicine, Zagreb, Croatia Fondazione IRCCS Policlinico San Matteo, University of Pavia, Department of Rheumatology, Pavia, Italy Medical University of Bialystok, Department of Rheumatology and Internal Medicine, Bialystok, Poland Poitiers University Hospital, Department of Internal Medicine, Poitiers, France School of Medicine, University of Occupational and Environmental Health Japan, The First Department of Internal Medicine, Kitakyushu, Japan Oslo University Hospital, Department of Rheumatology, Oslo, Norway  Background In systemic sclerosis (SSc), the role of oral anticoagulants (OACs) for treating or preventing the occurrence of digital ulcers (DUs) is not well understood. Given the low prevalence of OACs used in these patients, standard frequentist statistics are not suitable to capture the effect of OACs on DUs conclusively. However, a Bayesian statistical approach, which incorporates prior knowledge, can deal with a low prevalence of outcome or exposure and might provide conclusive evidence. Objectives To assess the effectiveness of a therapy with OACs on DUs in patients with SSc. Methods This study used prospectively collected data from the European Scleroderma Trials and Research group (EUSTAR) registry. Patients fulfilling the 2013 ACR/EULAR SSc classification criteria with complete longitudinal data on the presence of DUs and OACs were included in the analysis. The Bayesian prior probability distribution was created by capturing SSc experts’ opinion on the probability of absence of DUs with and without OACs and their degree of uncertainty using a questionnaire. In addition, participants ranked factors that influence the presence or absence of DUs as confounding factors using a numeric rating scale. The outcome was the absence of DUs at the last follow-up visit in SSc patients from the EUSTAR cohort, after one year of exposure to OAC. A nearest neighbor matching algorithm using Bayesian Additive Regression Trees with uniform priors generated a cohort with two groups of patients (treated and non-treated with OACs), which were balanced for age, sex, anti-Scl-70 antibody positivity, joint contractures, use of platelet inhibitors, use of prostacyclin analogues, use of any endothelin receptors antagonists or any phosphodiesterase type 5 inhibitor. The matching algorithm generated two balanced groups of treated and untreated patients (imbalance < 10%). A Bayesian logistic regression was implemented to estimate the effect of OACs on DUs. This model incorporated the priors elicited from experts’ beliefs and was adjusted for the above mentioned factors that influence the absence of DUs, as well as modified Rodnan skin score, disease duration, season, left ventricular ejection fraction, presence of pulmonary hypertension, C-reactive protein elevation, arterial hypertension and any previous DUs. Results Of the 6,424 patients enrolled in the EUSTAR registry, 663 (10.3%) had current digital ulcers, 2556 (39.8%) had previous DUs and 143 (2.2%) were exposed to OACs at their last follow-up visit. Mean age was 59 (SD: 13.6) years, 933 (14.5%) were males, mean disease duration was 13.1 (SD: 8.7) years, 4,454 (69.3%) had a limited cutaneous SSc and 1510 (23.5%) had C-reactive protein elevation. The median probability for the absence of DUs after exposure of one year to OACs was 70-75%, the pessimistic and optimistic probability were 70-75% and 80-85% respectively. After incorporation of the prior experts’ opinion, the unadjusted analysis revealed that the use of OACs was associated with absence of DUs at the last follow-up visit (OR 2.05, 95%CrI 1.63 to 2.46). In the adjusted analysis, the effect is OACs was slightly stronger (OR 2.15, 95%CrI, 1.71 to 2.52). The results of the analysis where the priors from experts were withheld (and uniform priors were set) are illustrated in figure 1. Conclusion This study indicates with incorporation of Bayesian priors that represent experts’ beliefs, a positive effect of OACs for the DUs in SSc. Image/graph:Figure 1. Effect of OACs on the absence DUs related to SSc at the last follow-up visit. Bayesian (priors): Analysis where the prior knowledge from the experts was incorporated into the model. Bayesian (uniform): Analysis where the prior knowledge from the experts was not incorporated into the model and a uniform/uninformative/neutral prior was used. OACs: Oral anticoagulants. REFERENCES: NIL. Acknowledgements: NIL. Disclosure of Interests Alexandru Garaiman: None declared, Catherine Gebhard: None declared, Carina Mihai Consultant of: Consulting fees and/or honoraria from Boehringer Ingelheim, Janssen, MED Talks Switzerland, Mepha, and PlayToKnow AG and consulting fees and/or honoraria from Boehringer Ingelheim, Janssen, MED Talks Switzerland, Mepha, and PlayToKnow AG, outside the submitted work, Rucsandra Dobrota Consultant of: consulting or speaker fees from Actelion and Boehringer-Ingelheim, congress support from Amgen, outside the submitted work, Grant/research support from: research grants from Articulum Fellowship, sponsored by Pfizer (2013-2014), Actelion, outside the submitted work, Cosimo Bruni Speakers bureau: Eli-Lilly, Consultant of: Boehringer Ingelheim, Consultant of: Boehringer Ingelheim, Grant/research support from: Gruppo Italiano Lotta alla Sclerodermia (GILS), European Scleroderma Trials and Research Group (EUSTAR), Foundation for research in Rheumatology (FOREUM), Scleroderma Clinical Trials Consortium (SCTC). Educational grants from AbbVie., Marco Matucci-Cerinic: None declared, Jörg Henes Consultant of: Abbvie, Boehringer Ingelheim, GSK, BMS, Janssen, Novartis, Pfizer, UCB, Jeska de Vries-Bouwstra Consultant of: ABBvie, Janssen, Boehringer-Ingerlheim, Grant/research support from: Janssen-Cilag, Galapagos, Roche, Vanessa Smith: None declared, Andrea Doria: None declared, Yannick Allanore Consultant of: AbbVie, AstraZeneca, Bayer, Boehringer-Ingelheim, Mylan, Janssen, Medsenic, Prometheus, Sanofi, Roche, Grant/research support from: Alpine Immunosciences, Medsenic, OSE Immunotherapeutics, Lorenzo Dagna: None declared, Branimir Anic: None declared, Carlomaurizio Montecucco: None declared, Otylia Kowal-Bielecka Consultant of: Abbvie, Boehringer Ingelheim, CSL Behring, Janssen, Medac, MSD, Novartis, Pfizer, Sandoz, Mickael Martin: None declared, Yoshiya Tanaka Speakers bureau: Behringer-Ingelheim, Eli Lilly, Abbvie, Gilead, AstraZeneca, Bristol-Myers, Chugai, Daiichi-Sankyo, Eisai, Pfizer, Mitsubishi-Tanabe, GlaxoSmithKline, Grant/research support from: Asahi-Kasei, Abbvie, Chugai, Eisai, Takeda, Daiichi-Sankyo, Behringer-Ingelheim., Anna-Maria Hoffmann-Vold Speakers bureau: Boehringer Ingelheim, Jannsen, Medscape, Merck Sharp & Dohme and Roche, Consultant of: ARXX, Boehringer Ingelheim, Genentech, Jannsen, Medscape, Merck Sharp & Dohme and Roche, Grant/research support from: Boehringer Ingelheim, Jannsen, Oliver Distler Speakers bureau: Bayer, Boehringer Ingelheim, Janssen, Medscape, Consultant of: Abbvie, Acceleron, Alcimed, Amgen, AnaMar, Arxx, AstraZeneca, Baecon, Blade, Bayer, Boehringer Ingelheim, Corbus, CSL Behring, 4P Science, Galapagos, Glenmark, Horizon, Inventiva, Kymera, Lupin, Miltenyi Biotec, Mitsubishi Tanabe, MSD, Novartis, Prometheus, Roivant, Sanofi and Topadur, Grant/research support from: Kymera, Mitsubishi Tanabe, Boehringer Ingelheim, Mike O. Becker Speakers bureau: Mepha, Bayer, MSD, GSK, Amgen, Novartis and Vifor, Consultant of: Mepha, Bayer, MSD, GSK, Amgen, Novartis and Vifor. Keywords: Systemic sclerosis, Artificial Intelligence, Rare/orphan diseases DOI: 10.1136/annrheumdis-2023-eular.3915Citation: , volume 82, supplement 1, year 2023, page 1624Session: Scleroderma, myositis and related syndromes (Publication only)