Comparative Analysis of Classification Criteria in IgG4-Related Disease and Evaluating Diagnostic Accuracy from a Retrospective Cohort in Clinical Practice

Le résumé fourni par la source

Introduction: We conducted a comprehensive comparative analysis of the Okazaki, Umehara, and American College of Rheumatology/European League Against Rheumatism (ACR/EULAR) classification criteria for diagnosing immunoglobulin G4-related disease (IgG4-RD). Materials and Methods: A retrospective study was conducted in a single tertiary hospital, using expert clinical judgment as the gold standard. We compared the diagnostic accuracy of the Okazaki, Umehara, and ACR/EULAR criteria in a cohort of 41 patients with suspected IgG4-RD. We assessed sensitivity, specificity, and positive and negative predictive values for each criterion, and conducted a separate analysis based on four IgG4-RD subtypes. Results: A total of 30 patients were confirmed to have IgG4-RD and 11 were identified as mimickers. The Umehara criteria demonstrated the highest sensitivity (83.33%), followed by the ACR/EULAR 2019 (66.67%) and Okazaki (60.0%) criteria. All three criteria exhibited 100% specificity, with overall diagnostic accuracy ranging from 70% to 88%. The areas under the curve (AUC) were 0.917 (Umehara), 0.800 (Okazaki), and 0.833 (ACR/EULAR 2019), indicating significant diagnostic effectiveness (p < 0.000). Subtype analysis revealed that the Umehara and ACR/EULAR 2019 criteria were more effective in diagnosing pancreato-hepato-biliary involvement (subtype 1), while the Okazaki and ACR/EULAR 2019 criteria were more effective in diagnosing retroperitoneal fibrosis and/or aortitis (subtype 2). Conclusions: Our study provides valuable insights into the diagnostic performance of the Okazaki, Umehara, and ACR/EULAR criteria for a cohort of patients with suspected IgG4-RD. The Umehara criterion demonstrated the highest sensitivity, suggesting its potential utility for screening purposes, while all three criteria showed consistent specificity.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.