Assessing IgG4-related ophthalmic disease and its mimics: a comparison of ACR/EULAR, organ-specific and revised comprehensive diagnostic criteria.

Abstract

Aims: Diagnosis of IgG4-related ophthalmic disease (IgG4-ROD) rests on the correlation of clinical features, serological testing and histopathology, using internationally accepted diagnostic criteria for objective interpretation; however, several mimickers of IgG4-RD overlap in clinical presentation and histopathology. We assess histopathological features in a series of presumptive IgG4-ROD cases, with emphasis on histopathological mimics and comparison of three IgG4-ROD diagnostic/classification criteria (organ-specific (OS), revised comprehensive diagnostic (RCD) and American College of Rheumatology/European Alliance of Associations for Rheumatology (ACR/EULAR) criteria).

Methods: The histopathology database was screened for cases with clinical/histopathological suspicion of IgG4-ROD. Slides were reviewed, OS, RCD and ACR/EULAR criteria were applied, and the final clinicopathological diagnosis was recorded.

Results: 37 patients (24 females, 13 males; 19-73 years) were diagnosed as either IgG4-ROD (n=18) or non-IgG4-related disease (n=19). Non-IgG4-related disease group showed elevated serum IgG4 (55.5%), fibrosis (100%), dense lymphoplasmacytic inflammation (92.8%), with an increase in tissue IgG4+plasma cells (57.1%) and elevated IgG4:IgG+plasma cell ratio (14.3%). ACR/EULAR missed 50% (9/18, sensitivity-52.8%) of true IgG4-ROD cases, while OS and RCD criteria missed 11.1% (2/18, sensitivity-88.9%) of IgG-ROD cases. ACR/EULAR criteria mislabelled 7.14% (1/14, specificity-90.9%) while OS and RCD criteria wrongly categorised 71.4% (10/14, specificity-47.4%) and 50% (7/14, specificity-63.2%) specific non-IgG4-ROD cases as IgG4-ROD. Storiform fibrosis, obliterative phlebitis, increased IgG4:IgG+plasma cell ratio and elevated serum IgG were statistically significant in distinguishing IgG4-ROD from its mimics.

Conclusion: ACR/EULAR criteria showed high specificity but were cumbersome and sensitivity was low, while RCD and OS criteria showed low specificity. Stringent clinicopathological correlation to exclude mimics is critical in avoiding diagnostic errors in IgG4-ROD.