AB1254 DIAGNOSTIC PERFORMANCE OF IgG4 CLASSIFICATION CRITERIA IN A REAL-LIFE COHORT FROM A SINGLE TERTIARY HOSPITAL CENTER

Background: The diagnosis of Immunoglobulin IgG4-related disease (IgG4-RD) lacks a definitive gold standard; however, the prevailing diagnostic criteria to date include those proposed by Umehara1 and Okazaki2. Umehara delineates diagnostic categories as definitive, probable, or possible, while Okazaki establishes the presence or absence of the disease. In 2019, ACR and EULAR3 introduced novel classification criteria necessitating a sequential application of exclusion and inclusion criteria, followed by a point count. IgG4-RD is classified if the score surpasses 20 points. Objectives: To assess the sensitivity and specificity of Okazaki, Umehara, and ACR/EULAR 2019 classification criteria for IgG4-RD in a patient cohort, and to compare them with the definitive diagnosis based on medical criteria. Methods: A retrospective study conducted at a national tertiary university center. Eligible patients were those with elevated serum IgG4 in any analytical determination, reviewing a total of 719 medical records between January 2000 and December 2023. Of these, inclusion criteria were patients with a high clinical suspicion of IgG4-RD due to compatible organ involvement or elevated IgG4. Demographic, clinical, serological, histological data, as well as variables related to classification criteria, were collected. The medical criterion, based on the criteria outlined in Figure 1, served as the reference for a definitive diagnosis. Sensitivity, specificity, positive and negative predictive values, as well as the ROC curve, were calculated. Results: A total of 39 patients with a high suspicion of IgG4-RD were included. Clinical characteristics are detailed in Figure 2. Of the 39 suspected patients, 26 (66.6%) met the definitive diagnosis according to medical criteria. Of the 26 IgG4-RD patients, 24 (92.3%) met Umehara's criteria; 16 (66.66%) possible, 2 (8.33%) probable, 6 (25%) definitive. Seventeen (65.4%) met Okazaki's criteria, and 18 (69.2%) met the ACR/EULAR 2019 criteria. The sensitivity of Umehara's criteria was 69.2% if the disease was categorized as probable or definite (NPV 61.9%, PPV 100% and accuracy 79.5%). Sensitivity for Okazaki's criteria was 65.4% (NPV 59.1%, PPV 100%, and accuracy 76.9%), and 69.2% for the new ACR/EULAR 2019 criteria (NPV 61.9%, PPV 100%, and accuracy 79.5%). For all three criteria, the specificity was 100%. The area under the curve was 0.962, 0.827, and 0.901 for Umehara, Okazaki, and ACR/EULAR 2019 criteria, respectively (p <0.000). Approximately 8% of patients were not classified by any of the three criteria but were accurately classified by the medical criteria. Conclusion: Sensitivity and the area under the curve were superior for Umehara and ACR/EULAR 2019 criteria. Eight percent of patients were not classified by any criteria, underscoring the significance of considering medical criteria. REFERENCES: [1] Umehara, Okazaki, Masaki, Kawano, Yamamoto, Saeki et. al. A novel clinical entity, IgG4-related disease (IgG4RD): general concept and details. Mod Rheumatol. 2012 Feb;22(1):1-14. [2] Okazaki, Umehara. Are Classification Criteria for IgG4-RD Now Possible? The Concept of IgG4-Related Disease and Proposal of Comprehensive Diagnostic Criteria in Japan. Int J Rheumatol. 2012; 2012: 357071. [3] Wallace, Naden, Chari, Choi. The 2019 American College of Rheumatology/European League Against Rheumatism Classification Criteria for IgG4-Related Disease. Arthritis Rheumatol. 2020 Jan;72(1):7-19. Acknowledgements: NIL. Disclosure of Interests: None declared.