Classification Criteria for Dermatomyositis
The 2017 EULAR/ACR classification criteria represent the current gold standard for classifying dermatomyositis, using a weighted scoring system where a total score ≥5.5 (without biopsy) or ≥6.7 (with biopsy) corresponds to ≥55% probability of idiopathic inflammatory myopathy (IIM), with subsequent sub-classification into dermatomyositis using a classification tree. 1
Overview of the Scoring System
The EULAR/ACR criteria employ a data-driven, probability-based model that assigns weighted scores to clinical, laboratory, and histopathologic features 1:
- Each variable receives a specific weighted score that contributes to a total score corresponding to probability of having IIM 1
- Probability cutoffs define classification certainty:
- Sub-classification into dermatomyositis occurs after establishing IIM diagnosis using a classification tree algorithm 1
Key Variables in the Classification System
The criteria incorporate easily accessible and operationally defined elements 1:
Clinical Features
- Age at onset (dermatomyositis typically presents across all age groups, unlike IBM which requires age ≥40 years) 2
- Pattern of muscle weakness (proximal and symmetrical in most cases, though can be distal or asymmetrical) 2
- Characteristic skin manifestations including heliotrope rash, Gottron papules, Gottron sign, V-sign, and shawl sign 1
Laboratory Features
- Muscle enzyme elevations (creatine kinase, aldolase, AST, ALT, LDH) 1
- Myositis-specific autoantibodies (when available, though not required for classification) 1
Histopathologic Features
- Muscle biopsy findings including perifascicular atrophy, inflammation patterns, and muscle fiber changes 1
- Note: Biopsy improves classification accuracy but is not mandatory 1
Advantages Over Historical Criteria
The EULAR/ACR criteria address critical limitations of older classification systems 1:
- The Bohan and Peter criteria (most widely used historically) frequently misclassified inclusion body myositis as polymyositis and lacked explicit operational definitions 1, 2
- The new criteria perform better than all existing criteria including Bohan and Peter, Tanimoto, Targoff, Dalakas and Hohlfeld, and ENMC criteria 1
- They correctly classify dermatomyositis and juvenile dermatomyositis while also distinguishing other IIM subgroups 1
Validation and Performance
The criteria have been extensively validated 1:
- Derived from 976 IIM patients (74% adults, 26% children) and 624 non-IIM patients with mimicking conditions 1
- External validation in the Euromyositis register (592 cases) showed 100% sensitivity with no misclassification when sufficient data were available 1
- Validation in the UK/Ireland Juvenile Dermatomyositis register (332 cases) demonstrated 92% could be classified with 100% sensitivity using the 55% cutoff 1
- Cross-validated area under the curve = 0.933 without biopsy and 0.942 with biopsy 1
Practical Application
A web-based calculator is available at www.imm.ki.se/biostatistics/calculators/iim to facilitate use of these criteria in clinical practice and research 1
The criteria provide flexibility for different study types 1:
- Clinical trials requiring high specificity should use higher probability cutoffs (≥90%)
- Epidemiological studies requiring high sensitivity can use lower probability cutoffs (≥55%)
Important Caveats
- These are classification criteria, not diagnostic criteria - they are designed for research studies and clinical trials to ensure homogeneous patient populations 1
- The criteria apply to both adult and juvenile dermatomyositis, making them universally applicable across age groups 1
- Missing data can prevent classification - in external validation cohorts, 8-17% of cases could not be classified due to incomplete data 1
- Perifascicular atrophy on muscle biopsy is not specific for dermatomyositis - it can occur in overlap myositis with dermatomyositis features 3