Objective: To help identify homogeneous subgroups among patients with anti-PM-scleroderma-antibodies (PM-Scl-Abs) positive auto-immune diseases regardless of diagnostic classifications.
Material and methods: This multicentric (four hospitals) retrospective study collected all consecutive patients (from 2011 to 2021) with positive testing for anti-PM-Scl-Abs in a context of CTD. Subgroups of patients with similar clinico-biological phenotypes were defined using unsupervised multiple correspondence analysis and hierarchical clustering analysis of the features recorded in the first year of follow-up.
Results: One hundred and forty-two patients with anti-PM-Scl-Abs were evaluated and 129 patients were included in the clustering analysis and divided into three clusters. Cluster 1 (n = 47) included patients with frequent skin thickening, digestive involvement and interstitial lung disease (ILD) with non-specific interstitial pneumonia (NSIP). They were more likely to develop progressive fibrosing ILD. Cluster 2 (n = 36) included patients who all featured NSIP with frequent organizing pneumonia-associated pattern and mechanic's hands. This subgroup had increased risk of relapse and ILD was characterized by a good functional outcome. Cluster 3 (n = 46) was characterized by predominant or isolated musculoskeletal involvement and frequently matched UCTD criteria. Although very frequent among anti-PM-Scl-Abs positive patients, muscle involvement was less discriminating compared with skin thickening and ILD pattern to classify patients into subgroups.
Conclusion: Anti-PM-Scl-Abs associated auto-immune diseases are segregated into three subgroups with distinct clinical phenotype and outcomes. Skin thickening and NSIP are determinant predictors in segregation of theses populations.
Keywords: SSc; anti-PM-Scl antibodies; cluster study; idiopathic inflammatory myopathies.
© The Author(s) 2022. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For permissions, please email: journals.permissions@oup.com.