Background/Purpose:

New 2016 EULAR/ACR classification criteria have been established for idiopathic inflammatory myopathies (IIM). The highest weighted score is associated with the presence of anti-Jo-1 antibody and dermatomyositis (DM) classic rashes (Gottron papules/sign and heliotrope rash). The aim of our study was to evaluate if the addition of other myositis specific antibodies (MSAs) and/or other DM rashes would improve the accuracy of the criteria.

Methods:

A prospective, longitudinal IIM dataset from a single myositis center was evaluated using the 2016 EULAR/ACR criteria as well as the Bohan & Peter classification criteria. All variables required for both these criteria were ascertained at the patientÕs first visit and subsequently updated prospectively on follow up visits. Expert physician clinical diagnoses (RA, CVO) were used as the gold standard to evaluate the criteria and proposed modifications. The latter included a) all other MSAs with anti-Jo1, b) all other DM rashes with Gottron papules (i.e. shawl sign, V-neck, holster sign, malar rash not sparing nasolabial fold, mechanics hands, periorbital edema), c) both a & b. This ongoing project thus far only includes data collection for DM. Percent agreement between the criteria and itsÕ proposed modifications with expert physician diagnosis were evaluated along with sensitivity and specificity.

Results:

213 DM patients (70% female, 89% Caucasian) have been analyzed to date. 2016 EULAR/ACR and Bohan & Peter showed % agreement of 84.3% and 82.9%, respectively as compared to the expert physician diagnosis, with sensitivity and specificity of 95.3% and 34.2% for EULAR/ACR and 83.8% and 82.9% for Bohan & Peter. The low specificity of EULAR/ACR criteria was partially explained by high false positives, as the current criteria fail to differentiate between DM and amyopathic DM, whereas the expert physician assessment did. The specificity of EULAR/ACR criteria improves to 44.8% without loss of sensitivity, if amyopathic DM is included under DM in the gold standard (Table 1). With the inclusion of other MSAs, there is a small gain in sensitivity to 98.3%, however, there is a marked decrease in specificity and a mild reduction of % agreement (Table 1). Similarly, with modification of inclusion of other DM rashes, the sensitivity increases to 97.7% at the cost of specificity and % agreement. Applying both other MSA and other rash modifications, the sensitivity increases to almost 99%, with a severe loss of specificity and % agreement.    

Conclusion:

Contrary to popular belief, addition of other MSAs or other DM rashes or both in the 2016 EULAR/ACR criteria did not improve the criteria. There is a small gain in sensitivity but a marked decrease in % agreement and specificity. Further analysis including PM, IBM and other muscle disease control subjects is required to provide conclusive answers.

Table 1