Objectives To evaluate the diagnostic and discriminative ability of the new 2012 European League Against Rheumatism (EULAR)/American College of Rheumatology (ACR) polymyalgia rheumatica (PMR) classification criteria compared to previous five diagnostic/classification criteria for PMR in a multi-centre prospective study.
Methods Patients older than 50 years of age (n=163), presenting with new onset (symptom duration ≤24 weeks) bilateral shoulder pain with elevated acute phase reactants were enrolled from 15 rheumatology clinics in Turkey. Patients were prospectively followed and the diagnosis of PMR was established when the diagnosis was maintained without an alternative diagnosis at 6 months of follow-up. Those who were diagnosed as other than PMR at the 6th month were designated as control group. All patients were classified by each of the six different criteria for PMR and 2010 ACR/EULAR Rheumatoid arthritis (RA) classification criteria.
Results Of the 163 patients with new-onset (mean symptom duration 10.9±7.5 weeks) bilateral shoulder pain 92 (56.4%) patients were diagnosed as PMR and 71 (43.6%) were diagnosed as nonPMR (RA: n=26). The discriminative ability as estimated by the area under the receiver operating characteristic (ROC) curve, was better for the Chuang criteria (0.83) than 2012 EULAR/ACR clinical criteria for PMR (0.69), Jones (0.68), Bird (0.68) and Nobunaga (0.77) criteria (Table 1). The 2012 EULAR/ACR clinical criteria for PMR had a sensitivity of 90.2% and a specificity of 49.3%. Jones and Chuang criteria had the highest specificity (91.5 and 87.3%, respectively). The specificity of the new 2012 EULAR/ACR clinical criteria for PMR slightly increased to 53.8% in RA patients. Although the new 2010 ACR/EULAR RA classification criteria classified only 9 out of 92 PMR patients as RA, the new 2012 EULAR/ACR clinical criteria for PMR classified 12 out of 26 RA patients as PMR.
Conclusions The new 2012 EULAR/ACR clinical classification criteria for PMR can classify PMR patients with high sensitivity, however, its ability to discriminate PMR from other inflammatory conditions with shoulder pain, especially RA is poor. Other criteria sets, Jones or Nabunaga criteria, despite involvement of similar clinical parameters, perform better in discriminating PMR from RA and other inflammatory/noninflammatory articular diseases. Our results, therefore, suggest that in seronegative patients with bilateral shoulder pain and acute-phase response, differential diagnosis of PMR and RA may require imaging (as proposed also in the other version of the new criteria) or biomarker studies.
Disclosure of Interest None declared