Optical Coherence Tomography with Angiography in Behçet Uveitis

Purpose: Behçet uveitis is a cause of retinal vasculitis, with fluorescein angiography being the gold-standard diagnostic exam. Optical coherence tomography with angiography (OCT-A) is a non-invasive exam that analyzes macular retinal circulation with high resolution, and therefore, could be a useful diagnostic exam in these patients. Material and methods: Analysis of 14 eyes of 8 patients with Behçet uveitis from our institution that were examined by OCT A (AngioPlex®, Zeiss). Macular morphology, area, perimeter and circularity of the foveal avascular zone, as well as central internal and global macular vascular density were evaluated. We compared our results with healthy controls. Results: Mean visual acuity in the patient’s subgroup was 0,84±0,25 and 3 eyes had previous diagnosis of retinal vasculitis with fluorescein angiography. Morphologically we found attenuation of the superficial and deep capillary plexus in 8 eyes, and in 1 of these eyes atrophy of the internal retinal layers was also found in a localized territory. In the foveal avascular area, no statistically significant changes were observed in its area and perimeter (p=0.804 and p=0.380 respectively), although we did find significant changes in its circularity (p=0.027). Vascular density was significantly altered in the internal (p=0.023) and global (p=0.028) macular parameters. Conclusion: We found changes in the macular microcirculation in patients with Behçet uveitis, even in the absence of angiographic changes. Future studies with more patients are necessary to understand the clinical significance of these changes.


Introduction
Behçet uveitis is a chronic autoimmune vasculitis with multi-systemic involvement and unknown etiology.1 Ocular involvement may be present in up to 70% of cases. 1,2Prominent manifestations include retinal vasculitis, occlusive or non occlusive periphlebitis, the presence of retinal infiltrates and optic disc edema.1 Fluorescein angiography is the gold-standard method for evaluation and monitoring of retinal vasculitis in Behçet uveitis patients, revealing contrast diffusion throughout affected vessels, while also facilitating the detection of other retinal alterations such as ischemia, neovascularization and cystoid macular edema, among others.Although of unquestionable clinical value, fluorescein angiography is a mildly invasive exam requiring a dye, which might constitute a limitation in some patients.Light dispersion phenomena also limit the exam's resolution. 3tical coherence tomography angiography (OCT-A) is a recent noninvasive imaging method that allows imaging of retinal and choroidal microvascularization without requiring contrast.The method is based on detecting fluctuations of red blood cells' signal in successive B-scan images obtained in the same location. 4,5][8][9][10][11][12][13][14][15][16][17][18] However, there is a limited number of recently published studies (Somki-ORIGINAL ARTICLE jrungroj et al. 2017 e Khairallah et al. 2017) 19,20 that have analyzed Behçet uveitis patients with the OCT-A, finding alterations in macular perfusion in the superficial, and deep retinal plexus.
The goal of our study is to study retinal microvasculature at the macular level using OCT-A on a population of Behçet uveitis patients from our institution's ophthalmology clinic.

Methods
Our study complies with the principles of the Declaration of Helsinki, and informed consent was obtained from all patients.This is a retrospective study, where we included patients previously diagnosed with Behçet uveitis (according to the international criteria of Behçet uveitis's international consensus) 21 under ophthalmology follow-up in our institution.Patients with high refractive error (>5D spherical equivalent), posterior capsule opacification, and history of retinal disease or Diabetes Mellitus were excluded.Clinical records (n=11) were consulted for best recorded corrected visual acuity (BCVA), previous ocular involvement pattern, the timeline of the disease's evolution, the disease's activity and fluorescein angiography findings in the last 6 months.All the patients that we included did not have active uveitis in the last 6 months, to avoid limitations associated with imaging associated with anterior segment and vitreous inflammation.Among the 11 eligible patients called for OCT-A, only 8 attended.It was only possible to perform the exam in 14 eyes, due to counting fingers vision in 2 eyes, therefore excluding the findings of these eyes.
Every patient underwent a routine ophthalmic exam.We performed OCT-A with the CIRRUS™ HD-OCT Model 5000 AngioPlex®, Zeiss machine.The machine is a spectral domain OCT working at a speed of 68,000 A-scans per second.Microvasculature images en face are obtained based on the OCT microangiography-complex algorithm, that takes into account variations in phase and intensity of sequential B-scans of the same location.We obtained images of the macular cube in the spectral domain OCT mode and images of a 3x3mm macular cube in the angiography mode.We analyzed the morphology of retinal layers regarding their structural integrity in the macular cube obtained by spectral domain.With the angiography mode images, we analyzed the scans from the superficial and deep capillary plexus that were measured automatically by the software.We searched for areas of disruption in the perifoveal capillary network, capillary alterations (thinning, dilation, shunts, telangiectasia), non-perfusion/hypoperfusion areas and localized or dif-fuse areas of disrupted capillary network architecture in the superficial or deep plexus's.We performed quantitative analysis of the foveal avascular zone's area, perimeter and circularity, as well as the central, internal and global macular vascular density, using the machine's automated analysis software.
A control group with a similar demographic profile also underwent the same evaluation.In the selection process, individuals with high refractive error (>5.0D spherical equivalent) or any ocular disease were excluded.
Quantitative data were analyzed with IBM SPSS Sta-tistics® (Version 2.2) software.Each studied parameter was tested for normality and subjected to a paired sample T-Test if normal, or Mann-Whitney's test otherwise.Specifically, avascular foveal zone area and circularity were compared by Mann-Whitney's test and the remaining analyses by paired sample T-Test.We calculated the correlation between visual acuity parameters, as well as disease duration, and avascular foveal zone area, perimeter and circularity, central, internal and global macular vascular density.The calculated correlation coefficient was Spearman's rho.

Results
The studied population and the control group were demographically equivalent with an average age of Behçet uveitis patients of 47.75±11.59,and 43.00±12.90 in the control group.The sex distribution in the Behçet uveitis patients was 8 females and 6 males, with the control group having 9 females and 5 males.
Behçet uveitis patients had a best-corrected visual acuity of 0.84±0.25 in the Snellen visual acuity scale.Among the 8 patients, five were found to have had previous ocular involvement with posterior uveitis, three of which with retinal vasculitis/branch retinal vein occlusion identified by angiography (Table 1).The patients with posterior uveitis without retinal vasculitis had retinitis (2 cases) and papillitis (3 cases).
Morphologic analysis of retinal layers found retinal atrophy in just 1 eye, affecting the internal retinal layers, corresponding to one of the eyes with retinal vasculitis (patient 7, Figure 1).
Analysis of the 3x3mm macular cube revealed vascular rarefaction in the superficial and deep plexus in 8 eyes, 4 of which corresponding to patients showing previous ocular involvement with retinal vasculitis (Figures 2 and  3); the remaining 4 did not have an angiographic diagnosis of vasculitis.

ORIGINAL ARTICLE » Silva D et al. OCT-A and Behçet
Average avascular foveal zone area measured at the superficial capillary plexus was 0.25±0.79mm 2 , average perimeter was 2.33±0.43mmand average circularity was 0.61±0.12.Comparing these parameters with the control group's parameters, we found no statistically significant difference for the avascular foveal zone area (p=0.804) and perimeter (p=0.380).A statistically significant difference was found for circularity (p=0.027)(Table 2).
Average central macular vascular density at the superficial capillary plexus was 11.60±4.17mm-1,average internal density was 19.97±3.50mm-1,and average global macular density was 19.13±3.23mm-1.We found statistically significant differences in the internal vascular density (p=0.023) and global vascular density (p=0.028),although not for the central vascular density (p=0.531)(Table 3).
We did not find statistically significant correlations between visual acuity or disease duration and the quantitative parameters of the avascular foveal zone, namely area, perimeter, circularity, central, internal and global macular density (Table 4).

Discussion
OCT-A allows us to obtain high spatial resolution imagery of retinal circulation in a non-invasive way, and as such is being used as a method for studying retinal conditions, to identify vascular alterations not visible in fluorescein angiography, especially in the deep capillary plexus.
Our study found vascular rarefaction in the superficial and deep capillary plexus in 8 eyes, 4 of which belong to patients without a previous diagnosis of retinal vasculitis by angiography.We also found statistically significant lower circularity index for these patients in comparison with the control group, as well as statistically significant alterations in internal and global vascular density in the superficial capillary plexus.Our findings suggest that Behçet uveitis patients might have microvascular alterations at the macular level in the absence of previously diagnosed retinal vasculitis.The clinical significance of these alterations is not evident, however.Our results agree with two previously published studies.Khairallah et al (2017) analyzed 44 eyes of 25 Behçet patients with ocular involvement in the posterior segment, comparing imaging findings obtained by fluorescein angiography, swept-source spectral domain OCT and OCT-A. 19reas with perifoveal microvascular alterations such as disruption of the perifoveal capillary network, retinal capillary perfusion/non-perfusion areas and perifoveal capillary anomalies were more frequently observed by OCT-A than by fluorescein angiography (40.9 vs. 25%; p=0.039, 86.4 vs. 34.1%;p <0.001, and 84.1 vs. 36.4%;p=0.039, respectively). 19Additionally, avascular foveal area was not significantly larger in Behçet patients than in the control group, either in the superficial or in the deep plexus.(0.4 vs. 0.34mm 2 ; p=0.23 and 0.72 vs. 0.53mm 2 ; p=0.053, respectively). 19Capillary density in the deep capillary plexus was significantly lower in Behçet patients than in the control group (13.7 vs. 17.2mm 2 , p=0.004).

ORIGINAL ARTICLE » Silva D et al. OCT-A and Behçet
The work of Somkijrungroj et al focuses on 37 eyes of 21 Behçet uveitis patients.Avascular foveal area of the superficial capillary plexus was found to be significantly smaller than in the deep capillary plexus (0.47 vs. 1.94mm2; p<0.001), with a capillary perfusion ratio between the superficial and deep capillary plexus of 0.57±0.27,correlated with visual acuity, thereby suggesting that the deep capillary plexus may be more intensely affected than the superficial capillary plexus. 20However, there was no comparison with a control group, limiting the findings of the study. 20e results of our study, along with the two referred studies, suggest microcirculation alterations at the macular level for these patients, which can be shown by OCT-A.The clinical significance of these alterations is not clear given the absence of prospective studies evaluating the stages where these alterations occur, and their relationship with the severity of the disease.Additionally, given that other vascular retinal diseases have been associated with microvascular changes in the OCT-A, these changes could likely be observed in other types of vasculitis.
Previous studies on patients with ocular manifestations of Behçet uveitis concluded that the posterior pole was spared until the late stages of the disease, when general occlusive vasculitis sets in. 22However, fluorescein angiography does not allow the detection of vascularization in the deep capillary plexus and thus may underestimate observed macular ischemia.Our study shows the presence of macular microcirculation alterations even in patients with no documented retinal vasculitis by fluorescein angiography.The work of Somkijrungroj et al. also found selective decreases in vascular density for the deep capillary plexus, even in eyes with normal avascular foveal zone as determined by angiography. 20hese findings suggest a degree of macular ischemia may be present in several stages of the disease.Being an important cause of decreased visual acuity, early detection by use of this new technology may be clinically relevant.
We did not observe statistically significant correlations between visual acuity or duration of the disease and quantitative parameters of the avascular foveal zone, namely area, perimeter, circularity, central, internal and global macular density.However, we did not undertake a longitudinal prospective analysis of these parameters throughout the timeline of the disease, nor have we studied its relationship with the type of immunosuppressive therapy, which constitutes an indirect measure of the severity of the disease.
The limitations of our study are the sample size/heterogeneity of previous ocular manifestations of BU, the inability of obtaining OCT-A scans in patients with low vision and the inability of the AngioPlex® analysis software to measure vascular density parameters, such as area and circularity of the avascular foveal zone in the deep capillary plexus.The comparison of these parameters with the parameters of the superficial capillary plexus would be highly relevant, given that previous studies on the use of OCT-A in Behçet uveitis show vascular alterations may be more pronounced in the deep capillary plexus. 19,20Additionally, studies on the use of OCT-A in other pathologies, namely diabetic retinopathy [6][7][8][9][10][11][12] , retinal venous occlusion [13][14][15]18 and sickle-cell disease associated retinopathy 16,17 suggest hypoperfusion alterations are more profound in the deep capillary plexus than in the superficial capillary plexus and may also be more strongly correlated with decreased visual acuity. Th2][13][14][15][16][17] We also acknowledge the limitation associated with retinal morphologic analysis being qualitative, thus operator-dependent.The comparison with the control group was always made against an individual of similar age to mitigate the introduction of an age-related vascular alterations bias.Despite these limitations, documenting non-observable (through traditional imaging methods) macular microcirculation alterations by OCT-A, in a non-invasive fashion, is an innovation in the study of this disease.Until now, identifiable macular changes were thought to be a reflex of generalized and severe retinal involvement 22 , and our study shows that these can be present even in patients with good visual acuity and without a previous diagnosis of vasculitis.Observing these changes, highlights the possibility of different pathways of retinal damage in this disease, which has pathophysiology mechanisms that are still unclear.
In conclusion, our study found that OCT-A allows us to identify retinal microcirculation alterations at the macular level in Behçet uveitis patients even in the absence of previous angiographic diagnosis of vasculitis.The clinical significance of these alterations remains unclear, and further large sample studies of a prospective nature are needed, where patients are clinically assessed with this exam and an analysis of the correlation between the degree of vascular alterations and the patients' disease severity can then be conducted.

ORIGINAL ARTICLE »Figure 1 .
Figure 1.OCT-A image of a patient with occlusive vasculitis 2 years ago

Figure 2 .Figure2.Figure 3 .
Figure 2. OCT-A image of a patient with occlusive vasculitis 2 years ago

Table 2 -
Comparison of the foveal avascular area parameters in patients with Behçet uveitis and controls.

Table 3 -
Comparison of the macular vascular density in patients with Behçet uveitis and controls.

Table 1 -
Characterization of the Behçet uveitis population (RE-Right Eye, LE-Left Eye, BCVA-Best corrected visual acuity with the Snellen scale).

Table 4 -
Correlation analysis between visual acuity or disease duration time; and the quantitative parameters of area, perimeter and circularity of the foveal avascular zone aswell as the parameters of macular vascular density.