Abstract
Objective.
To develop widely approved and validated classification criteria for anti-synthetase syndrome (ASyS) supported by the European Alliance of Associations for Rheumatology (EULAR) and the American College of Rheumatology (ACR).
Methods.
The CLASS project is a global, multicenter effort to develop data- and consensus-driven classification criteria for ASyS. Using a hybrid approach, the study integrated a systematic literature review, expert consensus, real-world clinical data from 103 centers, and multi-criteria decision analysis (MCDA). Key clinical and serological variables were identified and tested through both linear regression models and expert-rated clinical vignettes. Cohorts were stratified for derivation, internal validation, and independent validation, including a gold-standard centrally antibody-tested cohort by immunoprecipitation. Final classification criteria were developed with expert consensus to ensure face validity and feasibility, validated on large real world dataset with psychometric properties, and integrated with input from patients for clinical relevance.
Results.
A total of 1953 ASyS cases and 2093 controls were analyzed to develop and validate new classification criteria. Using data-driven and MCDA-based approaches, key variables included interstitial lung disease, myositis, arthritis, mechanic’s hands (or hiker’s feet), inflammatory skin rashes, unexplained fever, Raynaud’s phenomenon and anti-aminoacyl-tRNA synthetase autoantibodies (anti-ARS). Three classification levels - “definite,” “probable,” and “possible”- were proposed, with thresholds based on weighted scoring. In the independent validation cohort, definite ASyS showed 86.8% sensitivity and 99.2% specificity; in the gold standard cohort, sensitivity and specificity reached 94.3% and 99.7%, respectively. Probable ASyS reached 92.7%/87.3% (independent) and 97.5%/87.5% (gold standard); possible ASyS achieved 96.5%/75.4% and 98.7%/80.1%, respectively.
Conclusion.
We propose novel, validated classification criteria for ASyS, integrating clinical and serological domains. Developed through international consensus and data-driven methods, these criteria enhance accuracy in ASyS classification, enabling advancement in research and clinical trials.
To develop widely approved and validated classification criteria for anti-synthetase syndrome (ASyS) supported by the European Alliance of Associations for Rheumatology (EULAR) and the American College of Rheumatology (ACR).
Methods.
The CLASS project is a global, multicenter effort to develop data- and consensus-driven classification criteria for ASyS. Using a hybrid approach, the study integrated a systematic literature review, expert consensus, real-world clinical data from 103 centers, and multi-criteria decision analysis (MCDA). Key clinical and serological variables were identified and tested through both linear regression models and expert-rated clinical vignettes. Cohorts were stratified for derivation, internal validation, and independent validation, including a gold-standard centrally antibody-tested cohort by immunoprecipitation. Final classification criteria were developed with expert consensus to ensure face validity and feasibility, validated on large real world dataset with psychometric properties, and integrated with input from patients for clinical relevance.
Results.
A total of 1953 ASyS cases and 2093 controls were analyzed to develop and validate new classification criteria. Using data-driven and MCDA-based approaches, key variables included interstitial lung disease, myositis, arthritis, mechanic’s hands (or hiker’s feet), inflammatory skin rashes, unexplained fever, Raynaud’s phenomenon and anti-aminoacyl-tRNA synthetase autoantibodies (anti-ARS). Three classification levels - “definite,” “probable,” and “possible”- were proposed, with thresholds based on weighted scoring. In the independent validation cohort, definite ASyS showed 86.8% sensitivity and 99.2% specificity; in the gold standard cohort, sensitivity and specificity reached 94.3% and 99.7%, respectively. Probable ASyS reached 92.7%/87.3% (independent) and 97.5%/87.5% (gold standard); possible ASyS achieved 96.5%/75.4% and 98.7%/80.1%, respectively.
Conclusion.
We propose novel, validated classification criteria for ASyS, integrating clinical and serological domains. Developed through international consensus and data-driven methods, these criteria enhance accuracy in ASyS classification, enabling advancement in research and clinical trials.
| Original language | English |
|---|---|
| Journal | Annals Of Rheumatic Diseases |
| Publication status | Accepted/In press - 25 Jan 2025 |