To the Editor It was with interest that we read the article by Mariampillai et al¹ describing a new classification approach for idiopathic inflammatory myopathies (IIM) based on clinical manifestations and myositis-specific autoantibodies (MSAs).¹ The authors describe their classification system as complementary to the recently published European League Against Rheumatism/American College of Rheumatology (EULAR/ACR) classification criteria, as they focus on defining IIM subsets (clusters) rather than IIM diagnoses. The EULAR/ACR criteria attempted to subgroup as well, but an immune-mediated necrotizing myopathy (IMNM) subgroup could not be isolated, which is worrisome given the clinical consensus on its status as a subgroup and inclusion in previous classification approaches.^2,3 The 4 clusters proposed by Mariampillai et al were significantly associated with some MSAs (eg, anti–3-hydroxy-3-methyl-glutaryl-coenzyme A reductase and cluster 2, corresponding to IMNM). This is reflected in the major role of MSAs in allocating patients to a specific cluster in their decision tree. In the EULAR/ACR criteria, only the presence of anti–Jo-1 was included as a classification criterion, generating debate on the role of MSAs in the classification of IIMs.^2,4