Amandine Laporte1, Kubéraka Mariampillai2, Yves Allenbach2,3, Nicoletta Pasi1, Victoria Donciu4, Dan Toledano1, Benjamin Granger5, Olivier Benveniste2,3, Philippe A Grenier6, Samia Boussouar7. 1. Cardiovascular and Thoracic Imaging Unit, Pitié-Salpêtrière Hospital, AP-HP, Sorbonne University and Laboratoire d'Imagerie Biomédicale, INSERM, CNRS, Institute of Cardiometabolism and Nutrition, Sorbonne University, Paris, France. 2. MRSU 974, INSERM, Research Center in Myology, Sorbonne University, Paris, France. 3. Department of Internal Medecine and Clinical Immunology, Referral Center for Rare Neuromuscular Diseases, Pitié-Salpêtrière Hospital, AP-HP, Sorbonne University, Paris, France. 4. Department of Radiology, Pitié-Salpêtrière Hospital, AP-HP, Sorbonne University, Paris, France. 5. Department of Public Health (INSERM UMR 1136) and Pharmaco-Epidemiology Center, Pitié-Salpêtrière Hospital, AP-HP, Sorbonne University, Paris, France. 6. Referral Center for Neuro-Muscular Diseases, DHUi2B, Paris and Department of Clinical Research and Innovation, Foch Hospital, Suresnes, France. 7. Cardiovascular and Thoracic Imaging Unit, Pitié-Salpêtrière Hospital, AP-HP, Sorbonne University and Laboratoire d'Imagerie Biomédicale, INSERM, CNRS, Institute of Cardiometabolism and Nutrition, Sorbonne University, Paris, France. samiaboussouar@gmail.com.
Abstract
OBJECTIVES: Interstitial lung disease (ILD), one of the most common extramuscular manifestations of idiopathic inflammatory myopathies (IIMs), carries a poor prognosis. Myositis-specific autoantibody (MSA)-positivity is a key finding for IIM diagnosis. We aimed to identify IIM-associated lung patterns, evaluate potential CT-ILD finding-MSA relationships, and assess intra- and interobserver reproducibility in a large IIM population. METHODS: All consecutive IIM patients (2003-2019) were included. Two chest radiologists retrospectively assessed all chest CT scans. Multiple correspondence and hierarchical cluster analyses of CT findings identified and characterized ILD-patient subgroups. Classification and regression-tree analyses highlighted CT-scan variables predicting three patterns. Three independent radiologists read CT scans twice to assign patients according to CT-ILD-pattern clusters. RESULTS: Among 257 IIM patients, 94 (36.6%) had ILDs; 87 (93%) of them were MSA-positive. ILD-IIM distribution was 54 (57%) ASyS, 21 (22%) DM, 15 (16%) IMNM, and 4 (4%) IBM. Cluster analysis identified three ILD-patient subgroups. Consolidation characterized cluster 1, with significantly (p < 0.05) more frequent anti-MDA5-autoantibody-positivity. Significantly more cluster-2 patients had a reticular pattern, without cysts and with few consolidations. All cluster-3 patients had cysts and anti-PL12 autoantibodies. Clusters 2 and 3 included significantly more ASyS patients. Intraobserver concordances to classify patients into those three clusters were good-to-excellent (Cohen κ 0.64-0.81), with good interobserver reliability (Fleiss's κ 0.56). CONCLUSION: Despite the observed IIM heterogeneity, CT-scan criteria enabled ILD assignment to the three clusters, which were associated with MSAs. Radiologist identification of those clusters could facilitate diagnostic screening and therapeutics. Interstitial lung disease in patients with idiopathic inflammatory myopathy could be classified into three clusters according to CT-scan criteria, and these clusters were significantly associated with myositis-specific autoantibodies. KEY POINTS: • Cluster analysis discerned three homogeneous groups of interstitial lung disease (ILD) for which cysts, consolidations, and reticular pattern were discriminatory, and associated with myositis-specific autoantibodies. • Like muscle- and extramuscular-specific phenotypes, myositis-specific autoantibodies are also associated with specific ILD patterns in patients with idiopathic inflammatory myopathies.
OBJECTIVES: Interstitial lung disease (ILD), one of the most common extramuscular manifestations of idiopathic inflammatory myopathies (IIMs), carries a poor prognosis. Myositis-specific autoantibody (MSA)-positivity is a key finding for IIM diagnosis. We aimed to identify IIM-associated lung patterns, evaluate potential CT-ILD finding-MSA relationships, and assess intra- and interobserver reproducibility in a large IIM population. METHODS: All consecutive IIM patients (2003-2019) were included. Two chest radiologists retrospectively assessed all chest CT scans. Multiple correspondence and hierarchical cluster analyses of CT findings identified and characterized ILD-patient subgroups. Classification and regression-tree analyses highlighted CT-scan variables predicting three patterns. Three independent radiologists read CT scans twice to assign patients according to CT-ILD-pattern clusters. RESULTS: Among 257 IIM patients, 94 (36.6%) had ILDs; 87 (93%) of them were MSA-positive. ILD-IIM distribution was 54 (57%) ASyS, 21 (22%) DM, 15 (16%) IMNM, and 4 (4%) IBM. Cluster analysis identified three ILD-patient subgroups. Consolidation characterized cluster 1, with significantly (p < 0.05) more frequent anti-MDA5-autoantibody-positivity. Significantly more cluster-2 patients had a reticular pattern, without cysts and with few consolidations. All cluster-3 patients had cysts and anti-PL12 autoantibodies. Clusters 2 and 3 included significantly more ASyS patients. Intraobserver concordances to classify patients into those three clusters were good-to-excellent (Cohen κ 0.64-0.81), with good interobserver reliability (Fleiss's κ 0.56). CONCLUSION: Despite the observed IIM heterogeneity, CT-scan criteria enabled ILD assignment to the three clusters, which were associated with MSAs. Radiologist identification of those clusters could facilitate diagnostic screening and therapeutics. Interstitial lung disease in patients with idiopathic inflammatory myopathy could be classified into three clusters according to CT-scan criteria, and these clusters were significantly associated with myositis-specific autoantibodies. KEY POINTS: • Cluster analysis discerned three homogeneous groups of interstitial lung disease (ILD) for which cysts, consolidations, and reticular pattern were discriminatory, and associated with myositis-specific autoantibodies. • Like muscle- and extramuscular-specific phenotypes, myositis-specific autoantibodies are also associated with specific ILD patterns in patients with idiopathic inflammatory myopathies.
Authors: Jessica E Hoogendijk; Anthony A Amato; Bryan R Lecky; Ernest H Choy; Ingrid E Lundberg; Michael R Rose; Jiri Vencovsky; Marianne de Visser; Richard A Hughes Journal: Neuromuscul Disord Date: 2004-05 Impact factor: 4.296
Authors: Sabina Oreska; Hana Storkanova; Jaroslav Kudlicka; Vladimir Tuka; Ondrej Mikes; Zdislava Krupickova; Martin Satny; Eva Chytilova; Jan Kvasnicka; Maja Spiritovic; Barbora Hermankova; Petr Cesak; Marian Rybar; Karel Pavelka; Ladislav Senolt; Herman Mann; Jiri Vencovsky; Michal Vrablik; Michal Tomcik Journal: Front Med (Lausanne) Date: 2022-05-03