Swept-source optical coherence tomography angiography reveals vascular changes in intermediate uveitis

Abstract

Purpose: To evaluate retinal and choroidal vascular changes in patients with intermediate uveitis with/without concomitant retinal vasculitis using wide field swept-source optical coherence tomography angiography (OCTA). Methods: In this study consecutive patients with intermediate uveitis ± vasculitis and healthy age-matched controls were evaluated using central 3 × 3 mm OCTA scans and montage scans. Differences among the groups as well as the association between central changes assessed by 3 × 3 scans and wide field OCTA were evaluated. Results: 93 eyes of 58 patients with intermediate uveitis and 33 healthy age-matched controls were included. The presence of a cystoid macular oedema (CME) was associated with capillary non-and reduced perfusion of the superficial capillary plexus (SCP) (p = 0.03 and 0.02, respectively), and deep capillary plexus (DCP) (p = 0.02 and p ≤ 0.0001, respectively) of the 3 × 3 mm scans. The raw length, circularity and size of foveal avascular area (FAZ) significantly differed among the three groups. (p = 0.01, p = 0.045 and p = 0.004, respectively). Multivariable regression analysis revealed that the presence of epiretinal membranes (ERM) and CME rather than the disease entity contributed to the changes of FAZ size (R2 = 0.15, p = 0.0003) The mean vessel density (VD) of the SCP significantly differed among the three groups (intermediate uveitis with concomitant vasculitis: 16.8 ± 3.8 mm−1 vs intermediate uveitis only: 15.6 ± 4.4 mm−1 versus control: 18 ± 3.5 mm−1, p = 0.046). Multivariable regression analysis showed that the presence of CME rather than the disease entity impacted vessel density of SCP (R2 = 0.1, p = 0.016). There was no association between peripheral non- or reduced perfusion of the wide field OCTA slabs of the SCP and DCP and the VD of the 3 × 3 slabs. Conclusion: Although patients with intermediate uveitis and vasculitis present with reduced central vessel density compared to healthy age-matched controls, these changes are presumably contributed to the presence of CME.