In this study, we identified SRDs in 15 % of all patients with uveitis who underwent time-domain OCT imaging due to suspected macular pathology. Intermediate and panuveitis were the most common anatomic sites of inflammation, which is consistent with prior research evaluating uveitic macular edema [1
]. Moderate or severe visual impairment was identified in 71 % of patients in this series. We also found that increased CST was correlated with poorer logMAR visual acuity in patients with SRD. Whether decreased vision was a direct result of the SRD is difficult to ascertain due to high prevalence of concurrent intraretinal macular edema, which may also cause reduced visual acuity. Other factors including photoreceptor integrity, presence or absence of foveal atrophy, and chronicity of macular edema likely contributed to visual impairment, but these factors were not specifically addressed in the context of this study.
Diffuse macular edema and focal cystoid macular edema were the OCT features most commonly associated with SRD. The focal subtype of macular edema was most commonly identified while the diffuse subtype was also seen in a large proportion of patients. Patients with intermediate uveitis and cystoid macular edema comprised the majority of patients with OCT evidence of SRD and is consistent with prior series of patients with macular edema in tertiary referral settings [3
A variety of pathogenic mechanisms appear to be involved in the development of SRD in the uveitis population. Macular edema, the retinal architectural feature and clinically identifiable feature typically prompting OCT evaluation, is caused by inflammatory disruption of the normal permeability of the blood–retinal barrier.
Uveitis therapy with corticosteroids or other immunosuppressive medications leads to reduction in inflammation, restoration of vascular permeability, and subsequent resolution of the SRD and macular edema. However, because of the variety and diversity of etiologies associated with uveitis, other mechanisms are likely involved as well.
In VKH, breakdown of the blood–retinal barrier from iris and ciliary body inflammation is likely compounded by choroidal inflammation, overlying RPE injury and poor RPE pump function leading to the massive exudative detachments found in this disease process. In patients with multifocal choroiditis and punctate inner choroidopathy, subretinal fluid may accumulate following the formation of a choroidal neovascular complex, a consequence of both inflammatory and angiogenic pathways. Increased levels of vascular endothelial growth factor (VEGF) have been identified in the aqueous humor of uveitis patients [13
], and recently, anti-VEGF therapy with and without concomitant immunosuppressive medication has been successfully used for the treatment of uveitis-associated choroidal neovascular membrane (CNVM) [14
Within this series of patients, the specific therapy required to treat the process and associated SRD varied according to the disease process. Specifically, corticosteroid treatment was indicated for treatment of SRD secondary to inflammation as in intermediate uveitis and VKH; however, in patients with SRD due to CNVM, anti-VEGF and anti-inflammatory medications were required. Central serous retinopathy, although not specifically addressed in the context of this study, is another important etiology to consider, as corticosteroid use is nearly ubiquitous in the uveitis patient population and requires corticosteroid withdrawal to expedite its resolution.
The limitations of this study include the retrospective, cross-sectional nature of the study. Any study from tertiary referrals centers is prone to ascertainment bias, potentially skewing the proportion of uveitis patients with SRD. Conversely, some patients with advanced uveitis and severe vitritis and retinitis were excluded by the nature of the severity of their disease, which may cause significant media opacity precluding OCT imaging. Only a subset of patients with uveitis was characterized by OCT, and this subset mostly included patients with suspected macular edema. Lastly, our study exclusively used time-domain OCT, which utilizes 400 axial A-scans per second to achieve an axial resolution of 10 μm, whereas newer spectral domain OCT technology utilizes 20,000 A-scans per second to achieve a resolution approaching 5 μm. Moreover, because the retinal segmentation algorithms of spectral domain (SD)-OCT may result in greater macular thicknesses when compared with the algorithms of the time domain-OCT protocol used in this investigation, the CST measured in this study may be underestimated in patients with SRD.
Despite these limitations, we were able to identify specific uveitis syndromes in which SRD was a key clinical feature and described OCT characteristics valuable for planning future longitudinal studies employing SD-OCT strategies in the measurement of structural outcomes. Future prospective studies are needed and could provide more objective data by comparing the presence of SRDs in uveitis patients without macular edema to those with macular edema and might better define the visual prognostic significance of different subtypes of uveitic macular edema and their response to therapy.