Recurrence of Idiopathic Orbital Inflammation: An 11-year Retrospective Study.

PURPOSE: The high recurrence rate of idiopathic orbital inflammation (IOI) has been reported. This study aims to determine existing predictive factors for the recurrence of IOI.

METHODS: This was an 11-year retrospective study with at least a 12-month follow-up. Fifty patients with biopsy-proven IOI admitted between 2006 and 2017 at our tertiary hospital were observed. We compared the clinical characteristics, histopathological profile, and biomarker expressions (mast cell, immunoglobulin G4, tumor necrosis factor-alpha, and transforming growth factor-beta) of 16 patients with recurrence (Group I) and 34 patients with no recurrence (Group II). Statistical comparison and multivariate analysis were performed to establish the predictive factors.

RESULTS: We discovered five recurrence predictive factors: presentation of proptosis (odds ratio [OR] 4.96, 95% confidence interval [CI] 1.36-18.03), visual impairment (OR 15, 95% CI 1.58-142.72), extraocular muscle (EOM) restriction (OR 3.86, 95% CI 1.07-13.94), nonanterior involvement (OR 7.94, 95% CI 1.88-33.5), and corticosteroid (CS) alone treatment (OR 7.20, 95% CI 1.87-27.8). On multivariate analysis, nonanterior involvement and CS alone treatment were validated as predictive factors (area under the curve = 0.807 [95% CI 0.69-0.92]). Histopathological profile and biomarker expressions were not associated with recurrence. However, there was a 22-fold higher recurrence risk for granulomatous-type patients given CS alone treatment.

CONCLUSION: Unlike the five clinical characteristics mentioned, both histopathology and biomarker variables were not associated with recurrence. CS alone treatment for patients with nonanterior involvement or granulomatous type is proven to increase the risk of recurrence. Therefore, we suggest not giving CS without any combination treatment with other modalities for this group of patients.