Performance of the 2022 ACR/EULAR Classification Criteria in Comparison With the European Medicines Agency Algorithm in Antineutrophil Cytoplasmic Antibody-Associated Vasculitis.

Abstract

This study aimed to compare the 2022 American College of Rheumatology (ACR)/European Alliance of Associations for Rheumatology (EULAR) classification criteria with the European Medicines Agency (EMA) algorithm for antineutrophil cytoplasmic antibody (ANCA)-associated vasculitis (AAV). All consecutive, newly diagnosed patients with AAV according to the 2012 Chapel Hill Consensus Conference who visited Keio University Hospital between March 2012 and May 2022 were retrospectively reviewed. Patients were reclassified according to the EMA algorithm and the 2022 ACR/EULAR criteria, and their clinical characteristics were statistically analyzed. A total of 114 patients with AAV were included in the analyses. Using the EMA algorithm as a reference, reclassification of the patients revealed sensitivity and specificity of the 2022 ACR/EULAR criteria of 100% and 96% for eosinophilic granulomatosis with polyangiitis, 40% and 97% for granulomatosis with polyangiitis (GPA), and 90% and 49% for microscopic polyangiitis (MPA), respectively. Approximately half of patients classified as EMA-GPA or EMA-unclassifiable were reclassified as 2022-MPA; these patients were older, were more disposed to be positive for myeloperoxidase (MPO)-ANCA, and had interstitial lung disease (ILD) more frequently than patients with 2022-GPA or non-2022-MPA. Further, some patients positive for MPO-ANCA with biopsy-proven granulomatous inflammation were also reclassified from EMA-GPA to 2022-MPA. Over the mean observation period of 4.0 years, 16 patients died. Overall survival for each classification group differed significantly from the 2022 ACR/EULAR criteria (P = 0.02), but not with the EMA algorithm (P = 0.21). Among the patients classified as EMA-GPA or EMA-unclassifiable, older patients with MPO-ANCA and ILD tended to be reclassified as 2022-MPA. The 2022 ACR/EULAR criteria were more useful in prognostic prediction than the EMA algorithm.