Anti-VEGF Maintains The Function of Retinal Pigment Epithelium In Patients With Retinal Vein Occlusion

Background: To evaluate the drainage function of retinal pigment epithelium (RPE) in patients with macular edema secondary to retinal vein occlusion (RVO-ME) after anti-vascular endothelial growth factor (anti-VEGF) treatment by optical coherence tomography angiography (OCTA). Patients and Methods: Twenty-eight eyes with RVO-ME were imaged. Best-corrected visual acuity (BCVA), the area of the foveal avascular zone (FAZ), the nonperfusion areas (NPAs), the central macular thickness (CMT), the vascular density of the supercial/deep capillary plexus (SCP/DCP), the height of subretinal uid (SRF), and the number of hyperreective foci (HRF) were compared before and after treatments. Results: Compared with pre-treatment, the CMT, NPAs, number of HRF and the height of SRF were decreased signicantly after anti-VEGF treatment. There was no difference for BCVA, vascular density and FAZ area. Conclusion: The absorption of SRF and the decrease of HRF indicated the restoration of RPE drainage function and anti-inammation by anti-VEGF treatment.


Background
Retinal vein occlusion (RVO) is a relatively common retinal vascular disease which caused by the obstruction in central or branch retinal venous system, which is characterized by tortuous and dilatation of the retinal veins [1]. Clinically, RVO is divided into central retinal vein occlusion (CRVO) and branch retinal vein occlusion (BRVO) by the occlusion site. Hemiretinal vein occlusion (HRVO), also considered by some ophthalmologists to be a subgroup of CRVO, is an occlusion that causes about half of the retina to be affected [2]. According to statistics, the prevalence of BRVO is higher than CRVO, but the degree of visual impairment in patients with CRVO is more severer than that of BRVO [2]. Macular edema (ME) is the accumulation of the intraretinal or subretinal uid in the macular region, and is one of the most common complications of RVO (RVO-ME) [3]. Although the pathogenesis of RVO-ME is not yet fully understood, the most accepted hypothesis is that venous compression leads to blood ow disorders, endothelial damage and thrombosis, and in ammation especially vascular endothelial growth factor (VEGF), is crucial in this process [4]. Based on the role of VEGF in the pathogenesis of RVO, anti-VEGF has become an effective treatment and rst-line therapy for the treatment of RVO-ME [2].
The destruction of the blood-retinal barrier (BRB) combined with the subsequent increased vascular permeability usually leads to ME, which results in the visual impairment [3,5]. BRB is composed with inner BRB (iBRB, mainly consists endothelial cells and pericytes) and outer BRB (oBRB). The oBRB consists retinal pigment epithelium (RPE), and their tight junctions among RPE form the molecular components of oBRB [6]. Besides barrier function, RPE also drainage the subretinal uid to the choroids to maintain relatively "dry" environment of subretinal space [7] through active ionic, aquaporins and other Optical coherence tomography angiography (OCTA) is widely used technique during the clinical practice due to the rapid and non-invasive imaging in patients [9]. With the help of OCTA, the vascular density in both SCP and DCP, NPA and hyperre ective foci (HRF) could be well de ned in fundus diseases, including diabetic macular edema [10], RVO [11] and neovascular age-related macular degeneration [12].
In this study, we aimed to evaluate the drainage function of RPE before and after anti-VEGF treatment in patients with RVO-ME with the aid of OCTA. Besides the improvement in visual acuity and ME reduction, the SRF reduction after anti-VEGF indicated the improvement of drainage function of RPE by anti-VEGF treatment.

Patients
Twenty-eight consecutive treatment-naïve eyes from 28 RVO-ME patients who were diagnosed with comprehensive ophthalmologic examinations, between March 16, 2020, and July 1, 2021. The exclusion criteria were eyes with other diseases, such as diabetic retinopathy, hypertensive retinopathy, age-related macular degeneration, uveitis and other ocular diseases. In this retrospective study, the participants with two consecutive monthly intravitreal injections of anti-VEGF agents were enrolled.
All participants underwent full ophthalmologic examinations, including best-corrected visual acuity (BCVA), intraocular pressure (IOP) measurement, and anterior segment evaluation using slit-lamp biomicroscope, fundus photography, and OCTA imaging at the baseline and 1 week after treatment.
All intravitreal injections were performed by one experienced ophthalmologist. The intravitreal injection of anti-VEGF reagents (ranibizumab, Novartis Pharma Stein AG, Switzerland, 0.5 mg/0.05 mL or conbercept, Chengdu Kang Hong Biotech Co, Ltd, Sichuan, China, 0.5 mg/0.05 mL) was conducted at the infratemporal quadrant through the eyeball's pars plana under aseptic conditions. About 1-week variation was allowed for every injection interval.
The examination with OCTA was refered to a published method [12]. Retinal microvasculature was imaged using the RTVue XR Avanti OCT system (Optovue, Inc., Fremont, CA, USA), and measurements were acquired using the manufacturer's AngioVue software.
The NPA were outlined manually in the enface image of the SCP and analyzed automatically by the OCTA auto-segmentation software. The number of HRF was manually counted within a 6 mm diameter centered on the foveal area using a fovea spanning horizontal B-scan. Measured the longest vertical distance between the highest point of the subretinal uid and the interdigitation zone (IZ) / retinal pigment epithelial (RPE) layer as the height of the SRF. Poor quality images with a signal strength index< 4/10 were excluded from further analysis. The quanti cation of HRF was carried out by two experienced physicians independently.

Statistical Analyses.
The data were analyzed by using GraphPad Prism 6.0 software. All values are presented as a number or mean ± standard deviation. The visual acuity was presented as the logarithm of the minimum angle of resolution (logMAR). A paired t-test was used to compare BCVA, the vascular density of SCP and DCP, the FAZ area, NPA, the number of HRF and the height of SRF between pre-treatment and after 2 consecutive monthly anti-VEGF injections. A P value < 0.05 was considered as statistical signi cance.

Results
Demographics and clinical characteristics of the patients Twenty-eight eyes from 28 patients (12 males and 16 females) were enrolled in this study, as shown in Table 1. The mean age of patients was 64.18 ± 10.97 years old, ranging from 37 to 88 years old. The RVO includes 15 CRVO and 13 BRVO. All patients underwent two consecutive monthly injections of anti-VEGF. The mean interval between baseline and nal follow-up was 62.4 ± 24.8 (range 28 -147) days.
The microstructure of outer retina was partially restored after anti-VEGF injections.
Besides the absorption of SRF or the resolution of ME, the microstructures of the outer retina, including ellipsoid zone (EZ), interdigitation zone (IZ)/RPE layer, were partially restored, which became more discernible after anti-VEGF treatment (Fig. 3).

No alterations for Microvascular Density and FAZ after anti-VEGF treatment
Regarding retinal microvascular density in the SCP and DCP analyzed with OCTA, no signi cant difference was found for the microvascular density in the foveal, parafoveal, and whole macular sectors before and after treatment as shown in Table 2.

Discussion
RVO is the second common retinal vascular disease that impairs vision after diabetic retinopathy [7].
Although there are many treatment measures, such as laser photocoagulation, pharmacological agents, and surgical options [1], intravitreal injection of anti-VEGF improved visual acuity and reduce the macular edema, which is becoming the rst-line therapy for the treatment of RVO-ME [4]. The research showed that VEGF not only activates vascular endothelial growth factor receptor 2 (VEGFR2), leading to the reduction of tight junctions and the destruction of BRB, but also stimulates endothelial cells to production of a vasodilator, nitric oxide (NO), and increase intercellular cell adhesion molecule-1 (ICAM-1) expression [13,14]. Besides VEGF is also an important stimulator for the activation of in ammatory cells, such as microglia and mononuclear phagocytes. Through binding VEGF, anti-VEGF therapy becomes the rst-line therapy in the treatment of RVO-ME [15].
In this study, the improvement of visual acuity and decrease of macular edema were observed after anti-VEGF treatments, demonstrating that anti-VEGF therapy is very effective for patients with RVO-ME. Besides, BCVA and CMT, other parameters related to RVO were also analyzed, including vessel density, FAZ, HRF, and SRF, etc. The vessel density of SCP and DCP can be well identi ed and calculated with OCTA auto-segmentation software. Winegarner A. et.al [16] reported that the relationship between the decreased vessel density and visual impairment in RVO. The FAZ area was suggested as a biomarker to for local retinal ischemia within macular region [17,18], and previous study showed that progressive enlargement of FAZ after anti-VEGF therapy in cystoid ME secondary to diabetic retinopathy or RVO [19], indicating the macular ischemia. However, in our study, no signi cant changes for vessel density and FAZ area before and after anti-VEGF treatments, which is consistent with the previous study [20].
The mechanism of NPA formation remains unclear. Some researchers speculated the causative may compression of capillaries by swollen retinal tissues, arterial insu ciency based on back pressure by the obstructed veins, or capillary occlusion due to leukocyte accumulation at the endothelium. NPA usually considered similar to ischemia, and the size of NPA can distinguish the ischemic and nonischemic types in RVO [4]. Here, the decrease of NPA was observed after anti-VEGF treatment. The present study indicated that early and promptly treatment will bene t the patients more due to alleviate the ischemic condition of macula. It is suggested that anti-VEGF improved blood ow and oxygen supply, reduced the expression of VEGF and ICAM-A, and suppressed the accumulation of leukocyte, leading to reduction of NPA. The detailed mechanisms for NPA decrease after anti-VEGF treatment remains further exploration.
Chen et.al [21] reported that HRF associated with acquired vitelliform lesions were of RPE origin from a donor eye. HRF might as a biomarker associated with the activity and severity of choroideremia. It related to macrophages/microglial activation or progressive PRE degeneration [22]. HRF position was positively associated with nal visual outcome, and varied in different diseases depending on various pathologic mechanisms [23]. With aid of OCTA, we found HRF distributed in all the retina layers in patients with RVO-ME, and the number of HRF were decreased signi cantly after anti-VEGF treatment. At the same time, the restoration of microstructure of the outer retina was observed, including the myoid zone (MZ), ellipsoid zone (EZ), the outer segments of the photoreceptors (OS) and interdigitation zone (IZ) / RPE after anti-VEGF therapy. It also showed that SRF absorption might contribute to the repair of the microstructure and function of retina. Kang, J. W. et.al [24] reported that the more rupture of the external limiting membrane (ELM) structure, the less intact, and the greater the amount of HRF, no matter CRVO or BRVO.
The CMT, de ned as the mean thickness of the central 1 mm circle centered on the fovea, is the most used parameter to evaluate ME [25]. The signi cant reduction of CMT in our study implied that ME or SRF were absorbed rapidly after anti-VEGF. RPE functions as to maintain the permeability of transepithelial gradient generated by ion pumps and ion channels, and to avoid paracellular ow. The uid movement between cells follows the principle of osmotic gradient, owing from the retina toward to the choroid [6].
In RVO-ME, the function of RPE is impaired, resulting in restricted drainage function, causing SRF. The decrease of SRF after anti-VEGF treatment indicated the improvement of RPE drainage function by anti-VEGF, although the detailed mechanisms need further study.
Pervious study showed that the level of in ammatory cytokines, including VEGF, was elevated in either aqueous or vitreous samples, which was associated with the severity of ME [4,26]. VEGF has also been reported to increase vascular permeability by increasing the phosphorylation of tight junction proteins. As mentioned before, the mechanism for RVO-ME is complex [27], with other factors involved, such as interleukin 6 (IL-6), monocyte chemoattractant protein-1 (MCP-1), and pigment epithelial-derived factor (PEDF) [3,7]. Thus, it would be more accurate if personized treatment for RVO-ME is initiated. More indepth research and more precise combination treatments are needed in the future.
In summary, anti-VEGF is effective for the treatment of RVO-ME, in terms of visual acuity improve and reduction of macular edema. The absorption of SRF indicated the restoration of RPE drainage function by anti-VEGF treatment; while the decrease of HRF and NPA might indicate anti-in ammatory effect by anti-VEGF treatment to deactivate the in ammatory cells (Fig. 4).

Consent for publication
Written informed consents were obtained from all participants for the publication of clinical information and images.

Availability of data and materials
All data generated or analyzed during this study are included in this published article and its supplementary information les.

Competing interests
The authors declare that they have no competing interests.  Figure 1 Page 12/13

Funding
The changes of hyperre ective foci (HRF) in 66-years-old female patient with CRVO-ME before and after anti-VEGF treatments. The representative OCTA images at the 6×6 mm 2 sections centered on the fovea showed the gradual decrease of the number of HRF (purple triangle arrow) after anti-VEGF treatment. HRF were distributed throughout the whole retina before treatment. HRF in the outer retina (beneath the yellow line and above the blue line) was decreased signi cantly after treatment. The macular edema was also decreased after 2 consecutive anti-VEGF treatments.

Figure 2
The changes of subretinal uid (SRF) in 54-years-old female patient with BRVO-ME before and after anti-VEGF treatments. The representative OCTA images at the 6×6 mm 2 sections centered on the fovea showed the resolution of SRF (yellow arrow head) and macular edema (ME, green arrow) before and after intravitreal injection of anti-VEGF treatments. The SRF was completely absorbed after 2 consecutive anti-VEGF treatments. Diagram of the possible mechanism of anti-VEGF treatment for RVO-ME. On the one hand, anti-VEGF can reduce the number of HRF by inactivating microglia/macrophages cells. On the other hand, by restoring the drainage function of RPE and repairing the oBRB barrier to reduce the accumulation of SRF, and nally achieving the treatment of RVO-ME.