In this current study, we found a decreased VD at the posterior pole in eyes with active non-infectious PU without ME, compared to healthy controls. The VD decreased significantly in the SCP and slightly but insignificantly in the DCP. All the changes were quantitative, obtained with the algorithms built-up in the software of the device. The VD analysis was based on the ETDRS grid; thus, the decrease was evident in the whole image; foveal, parafoveal, and perifoveal regions for both plexuses. These changes represent an inflammation-induced disturbance in the foveal microvasculature. A literature search reveals similar findings with some discrepancies.
Waizel M and co-workers in their patients with PU of various etiologies, reported a significant enlargement of FAZ – regardless of ME - in the DCP compared to healthy controls; whereas, this enlargement was insignificant in SCP. They reported a significant effect of disease activity on FAZ enlargement only in the DCP; however, in contrast to the current study, their group consisted mostly of inactive eyes (3 vs. 23) and eyes with ME were included as well [15].
As a specific uveitis entity, mainly involving retinal vessels, Behcet’s disease was the most common in which OCTA features were evaluated in a number of reports [7,16-21]. In their report, Khairallah and co-workers found that the microvascular changes were more prominent in the DCP than the SCP. The authors used a manual measurement for quantitative analysis and eyes with ME were not excluded [7]. In another study, which also did not exclude ME, the VD was significantly lower in the DCP of Behcet patients compared to healthy controls; in the SCP, VD was significantly decreased in all ETDRS sectors [16]. Optical coherence tomography angiography depicted significant changes in Behcet’s patients compared to healthy controls in the report by Pei and co-workers: the VD in both plexuses were decreased and the FAZ was enlarged; the capillary disruptions were found to be more frequent in the SCP [17]. Accorinti and co-workers found significant VD decrease in both plexuses, being more severe in the DCP which was prominently correlated with duration of the disease. They also reported enlarged FAZ areas [18]. A report correlating microvascular damage and outer retinal disruption reported VD decrease was prominent in DCP and limited – albeit significantly – in SCP of eyes with Behcet’s disease; however, the enlargement in FAZ did not reach significance [19].
In one study analyzing the OCTA features of intermediate uveitis, in patients which retinal vasculitis accompanied, the VD decreased significantly in the SCP and insignificantly in the DCP, compared to controls, in accordance with our findings. The decrease was not resolved after removal of the projection artifacts in DCP [22]. Projection Artifact Removal algorithm could have helped better evaluation of the DCP in our group, as DCP was reported to be more prone to projection artifacts, resulting in an artificial increase in flow [23,24]. Moreover, Fenner and co-workers reported a higher repeatability of VD measurements for the SCP, than the DCP. The authors suggested that, this was likely the result of higher resolution of the SCP compared with the DCP [25].
The decreased VD is the sign of the perfusion deficit and the enlargement in FAZ shows macular ischemia; thus, OCTA which depicts the microvasculature of the posterior fundus successfully is rapidly becoming an important diagnostic tool. Optical coherence tomography angiography has been reported to be superior to FA to demonstrate the inflammation-induced microvascular changes in uveitis [7]. The changes in the SCP were concerned to be more strictly related to the active stage of the disease, and the change in DCP was correlated with the duration of the disease. [18]. Our results reporting a decreased VD in both SCP and DCP – though the latter is insignificant - are in accordance with the findings of the aforementioned studies. Our patients presented with a short duration of the disease as it was the first attack they experienced. It could be interpreted that, the VD change in the SCP was significant due to disease activity, whereas it did not reach significance in the DCP due to a relatively short disease duration.
Perhaps, exclusion of eyes with ME helped us to obtain accurate measurements of the DCP, as the cystoid spaces were depicted to be devoid of flow, localized in the DCP, causing peripheral displacement of retinal capillaries [7,16]. Foveal avascular zone in the DCP could not even be determined in all of the eyes with ME and as ME resolved with treatment, the VD in DCP was reported to increase significantly [23]. Being apart from the SCP, the DCP is not directly connected to the retinal arterioles; thus, it was concerned that this structural diversity makes it prone to ischemic attack in uveitis as well as retinal vascular disorders [7,18,26]. On the other hand, it was reported that, in hypoxic conditions oxygen was supplied to the DCP from the choroid in a rat model. They asserted that this could possibly be the underlying mechanism why the DCP was less affected [27]. Considering our results, this hypothesis is more likely to be admissible. Perhaps, further studies would help to elucidate these concerns.
We found that, FD-300 which was recently reported to demonstrate the VD in a 300 µ width zone around the FAZ, was significantly decreased indicating reduced perfusion. This parameter combines the SCP and the DCP and has been introduced to detect early signs of diabetic retinopathy [12,28]. We believe, this is a promising parameter which OCTA offers. Depending on a thorough literature search ours is the first report to depict FD-300 in uveitis patients.
We made a correlation analysis of severity score based on FA findings as previously reported [14]. Covering FA findings regarding the whole retina, we believe this system is a favorable indicator of severity of inflammation. Our results revealed a positive correlation with FAZ area and negative correlation with FD-300. We consider that these findings which could be interpreted as the severity of inflammation were associated with enlarged FAZ, reduced VD in a 300 µ width zone around the FAZ. More severe the inflammation was the more decreased perfusion in the retinal microvasculature.
As we excluded any concomitant diseases like retinal vascular diseases or glaucoma, we believe these microvascular changes could be attributed to a particular consequence of inflammation. All patients included were treatment-naïve with recent-onset active inflammation; thus, this eliminated a potential effect of therapy on the results. We conducted an automated, quantitative analysis which could eliminate possible interobserver variability of semi-automated or manual measurements. Automated algorithms, taking the mean brightness of the central FAZ area as threshold, were reported to decrease a confounding effect of tissue reflectance on quantification of VD [19].
Our study has several limitations. The sample size was relatively small and the study group comprises a variety of patients with different etiologies. It is considerable that, the investigation in a set of patients with a particular diagnosis could give more consistent results. However, in PU, all entities cause an inflammation in the posterior eye and it is not always possible to delineate these entities strictly. The lack of search for a potential correlation between the microvascular changes and visual outcome is another limitation.
In conclusion, OCTA depicted a decreased VD in the capillary plexuses and enlarged FAZ in eyes with non-infectious PU. With the non-invasive nature and repeatability, OCTA seems to be a promising method for the diagnosis, management, and follow-up of PU. Perhaps, further studies focusing on various uveitis entities would provide more precise data.