POS0708 VALIDATION OF THE REVISED 2022 AMERICAN COLLEGE OF RHEUMATOLOGY/ EULAR CLASSIFICATION CRITERIA FOR TAKAYASU ARTERITIS

The 2022 American College of Rheumatology (ACR)/European Alliance of Associations for Rheumatology (EULAR) classification criteria for Takayasu arteritis (TAK) were recently published [1].To validate and evaluate the 2022 ACR/EULAR TAK classification criteria in the light of the 1990 ACR TAK classification criteria [2].Clinical data of TAK patients from four referral centers (two from Italy and two from India) were reviewed to assess the fulfillment of 2022 ACR/EULAR and 1990 ACR TAK criteria. Control subjects included large-vessel giant cell arteritis (LV-GCA), large vessel vasculitis (LVV) other than TAK or GCA, or non-inflammatory arterial disorders. The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), likelihood ratio of a positive test (LR+), likelihood ratio of a negative test (LR-), area under the receiver operating characteristics curve (AUC) at the cut-offs of ≥3 points for 1990 ACR criteria and ≥5 points for 2022 ACR/EULAR criteria were calculated. AUC was also calculated using the actual scores for the 2022 ACR/EULAR criteria. Secondary analyses were conducted on the basis of sex (male/female), using only LV-GCA as controls, using subjects ≤60 years, and stratified on age (<40, 40-60, >60 years).504 TAK [404 females, mean (SD) age at diagnosis 31.7 (12.6) years] and 222 controls [144 LV-GCA, 151 females, mean (SD) age at diagnosis 61.9 (15.4) years] were identified. The 2022 ACR/EULAR criteria had better sensitivity and NPV but had poorer specificity, PPV, LR+, LR-, and AUC at predetermined cut-offs than the 1990 ACR criteria (Table 1). Similar performance of 2022 ACR/EULAR criteria was observed with only LV-GCA as controls (sensitivity 95.83%, specificity 60.42%, AUC 0.781) or in subjects ≤60 years old (sensitivity 95.81%, specificity 61.90%, AUC 0.789). The 2022 ACR/EULAR criteria had a greater specificity (76.06% vs 57.62%) and AUC (0.845 vs 0.771) with similar sensitivity (93% vs 96.53%) in males than in females. Stratified for age [<40 years, n=399, 374 TAK, 25 controls; 40-60 years, n=186, 127 TAK, 59 controls; >60 years, n=141, 3 TAK, 138 controls], sensitivity remained similar (96.26%, 94.49%, 100%, respectively), whereas, specificity was higher for older age groups (52%, 66.10%, 64.49%, respectively). Cut-offs of ≥6 (sensitivity 91.87%, specificity 82.88%) and ≥7 (sensitivity 86.71%, specificity 86.49%) greatly improved balance between sensitivity and specificity (Figure 1).In this first validation study, the 2022 ACR/EULAR TAK criteria had poorer specificity in real-life than in the development cohort. Higher cut-offs (6 or 7) might improve the performance of these criteria. Higher PPV but lower NPV in the Indian than in the Italian cohort might reflect the different performance of the criteria in different ethnic groups.[1]Grayson PC, et al. 2022 American College of Rheumatology/EULAR Classification Criteria for Takayasu Arteritis. Ann Rheum Dis. 2022;81(12):1654-1660.[2]Arend WP, et al. The American College of Rheumatology 1990 criteria for the classification of Takayasu arteritis. Arthritis Rheum 1990;33(8):1129-34.Table 1.Performance of the criteriaOverall (n=726, 504 TAK, 222 controls)Italian cohort (n=401, 201 TAK, 200 controls)Indian Cohort (n=325, 303 TAK, 22 controls)ACR 1990 criteriaACR EULAR 2022 criteriaACR 1990 criteriaACR EULAR 2022 criteriaACR 1990 criteriaACR EULAR 2022 criteriaSensitivity82.94%95.83%75.12%94.53%88.12%96.70%Specificity90.54%63.51%93.50%63.50%63.64%63.64%PPV95.22%85.64%92.07%72.24%97.09%97.34%NPV70.03%87.04%78.90%92.03%28.00%58.33%LR+8.772.6311.562.592.422.66LR-0.190.070.270.090.190.05AUC (95% CI)0.867 (0.842 – 0.893)0.797 (0.764 – 0.830)0.843 (0.809 – 0.878)0.790 (0.753 – 0.827)0.759 (0.654 – 0.863)0.802 (0.698 – 0.905)Correctly classified (%)85.26%85.95%84.29%79.05%86.46%94.46%Figure 1.AUC using actual 2022 ACR/EULAR scores.None Declared.