Background:Idiopathic inflammatory myopathies are a heterogenous group of systemic autoimmune diseases. Several phenotypes have been linked to specific autoantibodies. Clinically amyopathic dermatomyositis with rapidly progressive interstitial lung disease, the most severe form of ILD, is associated with the anti-MDA5 antibodies. However not all the patients with dermatomyositis and anti-MDA5 positive antibodies develop this severe condition.Objectives:We aim to define different phenotypes from a large cohort of patients diagnosed with dermatomyositis who were positive to anti-MDA5 antibodies.Methods:We retrospective analyzed the clinical and immunological data of 90 anti-MDA5 patients [50 female, 55.6%, mean (SD) age at diagnosis 47 (15.4) yrs.] with dermatomyositis recruited from a multicenter register in Spain (MEDRA5) including 30 hospitals. All the patients fulfill de International Myositis Classification Criteria (EULAR/ACR) for dermatomyositis (score >90%). Anti-MDA5 were detected by means of commercial immunoblot (EUROIMMUN©). The chi-square test was used to assess the relationships between qualitative variables. The Kruskal-Wallis test was used to compared medians between groups.Results:Sixty-six patients (73.3%) were diagnosed with clinically amyopathic dermatomyositis. Three different phenotypes linked with the anti-MDA5 antibody were identified. Group 1: patients with rapidly-ILD phenotype (28 patients, 31.1%), group 2: antisynthetase-like phenotype (23 patients, 25.5%), and group 3: non-ILD phenotype (39 patients, 43.3%). Clinical and immunological comparison between the groups disclosed that age at disease onset was higher (median, IQR) in patients from group 1 [53 (43-60)] vs. group 2 [46 (40-56)] or group 3 [42(41-51)] (p=0.01); disease onset was more frequent in spring in patients from group 1 (46.5%) than in the rest of the groups (21.7% and 28.9%) (p<0.01). Cancer was detected in 7 patients, only associated with myositis in 3 cases (3 years interval between cancer and dermatomyositis) without significant differences between phenotypes. Vasculitis (one case ANCA positive) was detected in 9 cases (6 limited to skin, 1 renal and 1 intestinal), 6 of them in the group 3 (statistical significance, in comparison with group 1 and 2, p<0.01). Mortality rate was higher in group 1 (51.9%, 16 out of 17 due to refractory respiratory failure) vs group 2 (12.5%) or 3 (0%) (p<0.001). Anti Ro52 positivity was more frequent in group 1 (65.4%) vs. group 2 (25%) or 3 (35.5%) (p<0.017), although it did not reach statistical significance in terms of mortality (p=0.173) or patients admitted in the intensive care unit (p=0.173). Mechanic hands were more frequent in group 2 (40.6%) than in groups 1 (25%) and 3 (34.4%) (p=0.05). Fever was significantly most frequent in group 1(52.6%) than in group 2 (21.1%) and 3 (26.3%) (p=0.001). Other clinical or immunological features such as arthritis, myositis, or the number of characteristic skin lesions among others were not more frequent in one group or another.Conclusion:Three different phenotypes of patients positive to anti-MDA5 were identified. The presence or not of ILD, or the different type (rapidly progressive or not) of ILD were the main feature that allow to differentiate these phenotypes, which are relevant in clinical practice.References:[1]Allenbach Y, Uzunhan Y, Toquet S, et al; French Myositis Network. Different phenotypes in dermatomyositis associated with anti-MDA5 antibody: Study of 121 cases. Neurology. 2020;95: e70-e78.Acknowledgements:List of contributors of MEDRA5 group: Aguilar-García J (Internal Medicine, Hospital Costa del Sol, Marbella), Carrión-Barberá I (Rheumatology, Hospital del Mar, Barcelona), Cobo-Ibañez T (Rheumatology, Hospital Infanta Sofía, San Sebastián de los Reyes), de Escalante-Yangüela B (Internal Medicine, Hospital Clínico Lozano Blesa, Zaragoza), Fonseca-Aizpuru EM (Internal Medicine, Hospital de Cabueñes, Gijón), González-Cubillo L (Intensive Medicine, Hospital Universitario de Cruces, Barakaldo), González-Gay MA (Rheumatology, Hospital Marqués de Valdecilla, Santander), Prieto-González S (Internal Medicine, Hospital Clinic, Barcelona), Ruiz-Román A (Rheumatology, Hospital Universitario Virgen del Rocío, Sevilla), Calero-Paniagua I (Internal Medicine, Hospital Virgen de la Luz, Cuenca), Callejas-Rubio JL (Internal Medicine, Hospital Clínico San Cecilio, Granada), Gil-Vila A (Internal Medicine, Hospital Vall d’Hebron, Barcelona), de Miguel-Campo B (Internal Medicine, Hospital Doce de Octubre, Madrid), García-Sevilla R (Pneumology, Hospital General Universitario de Alicante, Alicante), Iriarte-Fuster A (Internal Medicine, Hospital de Bellvitge, Hospitalet de Llobregat), Jovani-Casano V (Rheumatology, Hospital General Universitario de Alicante, Alicante), Lozano-Rivas N (Rheumatology, Hospital Virgen de la Arritxaca, Murcia), Martín-Gascón M (Internal Medicine, Hospital Morales Meseguer, Murcia), Martinez-González O (Rheumatology, Hospital Universitario de Salamanca, Salamanca), Monteagudo-Jiménez M (Internal Medicine, Hospital Parc Taulí, Sabadell), Mora-Ortega GM (Pneumology, Hospital Universitario Infanta Sofía, San Sebastián de los Reyes), Moral-Moral Pedro (Internal Medicine, Hospital Universitari i Politecnic La Fe, Valencia), Pérez-De Pedro I (Interna Medicine, Hospital Regional Universitario de Málaga, Málaga), Picazo-Talavera MR (Rheumatology, Hospital del Sureste, Madrid), Rubio-Rivas M (Internal Medicine, Hospital de Bellvitge, Hospitalet de Llobregat)Disclosure of Interests:None declared