Background/Purpose: Idiopathic inflammatory myopathies (IIM) are a heterogeneous group of diseases characterized by muscle inflammation and internal organ involvement. The Bohan & Peter (B&P) criteria (1) have been widely used for both, diagnostic and classification criteria for IIM, up until recently, when newer classification criteria were developed by the European League Against Rheumatism/American College of Rheumatology (EULAR/ACR) (2). These criteria are based on data from children and adults with different ethnicities from Europe, America and Asia, but the original study did not include patients from Latin America. We sought to determine the performance of the 2017 EULAR/ACR classification criteria for IIM in a cohort of Chilean patients in Latin America.

Methods: We performed a retrospective study of all patients from a tertiary referral medical center in Chile from 2014 to 2019 who had an associated ICD-10 code of M33 (dermatomyositis (DM)/polymyositis(PM)), M60.9 (myositis), and G72.4 (inflammatory myopathy). Demographic data, clinical and serological features were collected from medical records according to a pre-defined protocol in REDCap.(3) Patients were classified according to B&P criteria and 2017 EULAR/ACR classification criteria for IIM. Sensitivity, specificity, and Cohen’s Kappa coefficient of agreement (κ) were calculated to assess the performance of the 2017 EULAR/ACR classification criteria when compared to B&P criteria.

Results: 150 patients were included in the study. Our cohort was 77% female, mean age was 43.2±22.6 years, 61% presented with heliotrope rash, 53% with Gottron papules, 61% with proximal muscle weakness, 64% patients had elevated CK (mean CK 3386±9391 U/L), 60% had positive ANA, 5% had a positive Jo-1 antibody. 42 patients had myositis-associated antibodies (MAA) and 37 patients had at least one myositis-specific antibody (ASM). Muscle biopsy was available in 58 patients, electromyography was available in 76 patients, and was suggestive of myositis in 38 patients. 139 patients met B&P criteria; 34 (24.5%) were classifiable as PM (12 patients definite, 9 probable, 13 possible) and 105 (75.5%) as DM (30 definite, 48 probable, 27 possible). 134 patients (89.3%) met EULAR/ACR criteria; 105 (78.4%) patients were classifiable as definite IIM, 27 (20.2%) patients as probable IIM, and 2 (1.5%) as possible. Subclassification of IIM was as follows: 22 (16.4%) patients were classified as juvenile DM, 51 (38%) as adult DM, 32 (23.9%) as amyopathic DM, 28 (21%) as PM, and 1 (0.75%) as inclusion body myositis. Sensitivity of EULAR/ACR criteria to detect IIM was 0.92, and specificity was 0.53 when compared to B&P criteria. Kappa coefficient of agreement was weak between the two classification criteria (0.39, SD 0.15-0.64). Sensitivity and specificity of ACR/EULAR criteria to diagnose DM was 0.92 and 0.82, respectively. Sensitivity of ACR/EULAR criteria for PM was 0.85 and specificity was 0.9 (Table 1).

Conclusion: The current most accepted classification criteria for IIM have good sensitivity but low specificity in a Latin American cohort of patients. The new EULAR/ACR criteria performed well in this cohort to classify DM and PM correctly.

Table 1. Sensitivity and specificity of 2017 ACR/EULAR criteria for IIM

« Back to ACR Convergence 2020

ACR Meeting Abstracts - https://acrabstracts.org/abstract/sensitivity-and-specificity-of-the-2017-eular-acr-criteria-for-idiopathic-inflammatory-myopathies-in-a-cohort-of-patients-from-latin-america/