Role of Soluble Vascular Endothelial Growth Factor Receptor Signaling and Other Factors or Cytokines in Central Retinal Vein Occlusion With Macular Edema

Abstract

Purpose.: The association between aqueous humor levels of various growth/inflammatory factors or cytokines and the severity of macular edema was investigated in patients with central retinal vein occlusion (CRVO).

Methods.: Aqueous humor levels of 11 factors or cytokines were measured in CRVO patients with macular edema (n = 38) and in cataract patients (n = 15) as controls. Aqueous humor samples were obtained from the patients during surgery and the levels of VEGF, its receptors, and other factors or cytokines were measured by the suspension array method. The severity of macular edema was determined by measuring the central macular thickness, neurosensory retinal thickness, and subfoveal serous retinal thickness with optical coherence tomography.

Results.: Aqueous humor levels of growth factors, soluble VEGF receptor 1 (sVEGFR-1), sVEGFR-2, and inflammatory factors or cytokines were significantly higher in the CRVO group than in the control group. The aqueous level of sVEGFR-1 was significantly correlated with the neurosensory retinal thickness, as well as with the levels of growth factors (VEGF, placental growth factor, and platelet-derived growth factor-AA) and inflammatory factors/cytokines (monocyte chemotactic protein-1, interleukin-6, and interleukin-8). Aqueous humor levels of the three growth factors also were significantly correlated with each other, as were levels of the two sVEGFRs.

Conclusions.: These findings suggest that growth/inflammatory factors and cytokines have an important role in macular edema associated with CRVO. Better understanding of the mechanisms involved may lead to development of new treatments, such as anti–VEGFR-1 therapy.

Introduction

Central retinal vein occlusion (CRVO) involves vascular obstruction associated with atherosclerosis that leads to mechanical, ischemic, and inflammatory changes. Retinal ischemia and hemorrhage cause inflammation after retinal vein occlusion, while increased adhesion of leukocytes to the vein walls leads to stagnation of blood flow.¹ The processes of retinal ischemia, vascular remodeling, and atherosclerosis are all reported to be associated with inflammation.^2,3 Each of these processes is mediated by inflammatory cytokines, chemokines, and growth factors that regulate the behavior of lymphocytes, macrophages, and endothelial progenitor cells from the bone marrow. Macular edema is frequent in CRVO patients and is the most common cause of visual impairment associated with CRVO. Macular edema is mediated by angiogenic and inflammatory cytokines, including VEGF.^4–7 Thus, the onset and progression of macular edema in patients with CRVO seems to depend on various intraocular mediators of angiogenesis and inflammation. 

Since the development of anti-VEGF agents, treatment of CRVO has changed dramatically. Intravitreal injection of bevacizumab (a monoclonal antibody for VEGF) or ranibizumab (a Fab fragment that neutralizes all isoforms of VEGF-A) improves macular edema in CRVO patients.^8,9 However, macular edema sometimes persists or recurs after intravitreal treatment with these agents,^10,11 which suggests that other mechanisms contribute to macular edema in addition to VEGF. We recently reported that various cytokines may have a role in macular edema associated with branch retinal vein occlusion.¹² Aflibercept (VEGF Trap-Eye) is a fusion protein combining key domains from human VEGFR-1 and -2 with the constant region (Fc) of human immunoglobulin G that binds multiple VEGF-A isoforms, and it has been reported to improve macular edema in CRVO patients.^13,14 These reports suggest that VEGF family members may interact to cause macular edema in CRVO, but the actual process remains unclear. In the present study, we simultaneously measured the aqueous humor levels of various growth factors, soluble VEGF receptor (sVEGFR)-1, and sVEGFR-2, and several inflammatory factors/cytokines in CRVO patients with macular edema and control patients with cataract using suspension array technology for the first time, although we already have studied several cytokines using enzyme-linked immunosorbent assay. Then, we investigated the association between each factor/cytokine and the severity of macular edema. 

Methods

Subjects

This study was conducted in the Department of Ophthalmology at the Hachioji Medical Center of Tokyo Medical University. Approval was obtained from the Ethics Committee of Hachioji Medical Center. Study procedures conformed to the tenets of the Declaration of Helsinki and all patients signed an informed consent form before inclusion. 

We studied 38 patients with CRVO (38 eyes) who were scheduled to undergo intravitreal injection of bevacizumab (13 eyes, 1.25 mg Avascin; Genentech and Hoffmann La Roche, Basel, Switzerland), ranibizumab (16 eyes, 0.5 mg Lucentis; Genentech, Inc., South San Francisco, CA, USA), or aflibercept (9 eyes, 2 mg Eylea; Regeneron Pharmaceuticals, Inc., Tarrytown, NY, USA, and Bayer HealthCare Pharmaceuticals, Berlin, Germany) between January 2013 and July 2014. Aqueous humor was collected just before the injection of anti-VEGF agents. The criteria for performing intravitreal anti-VEGF therapy were macular edema involving the fovea (retinal thickness > 300 μm) and best-corrected visual acuity (BCVA) < 20/30. The BCVA was converted to the logarithm of the minimum angle of resolution (logMAR). 

Exclusion criteria were a history of glaucoma, uveitis, retinal disease other than CRVO, diabetes mellitus, rubeosis iridis, ocular infections, laser photocoagulation, and intraocular surgery (including cataract surgery) on the study eye within 6 months before the planned start of anti-VEGF therapy. Four eyes were pseudophakic, but cataract surgery had been performed more than 6 months earlier. 

Ocular Examinations

All patients underwent a complete ophthalmic examination that included decimal BCVA and fluorescein angiography (FA; Digital Retinal Camera CF-1; Canon, New York, NY, USA). A blinded grader independently assessed ischemic retinal vascular occlusion by examining the fluorescein angiograms and measuring the ischemic region of the retina using the public domain Scion Image program (Scion Corporation, National Institutes of Health, Bethesda, MD, USA), as reported previously.^6,7 If the nonperfused area divided by the disc area gave a value of 10 or more, this was defined as retinal ischemia.^15–17 Two subjects were excluded from this analysis because assessing the nonperfused area was difficult due to severe retinal hemorrhage. The 36 remaining CRVO patients included 11 patients with retinal ischemia (4 women and 7 men aged 69.3 ± 11.6 years) and 25 patients without ischemia (14 women and 11 men aged 73.7 ± 10.2 years). 

Optical coherence tomography (OCT) was performed in each subject within 1 week before intravitreal injection of bevacizumab, ranibizumab, or aflibercept, using a spectral-domain OCT apparatus (Spectralis; Heidelberg Engineering, Heidelberg, Germany). The severity of macular edema was classified on the basis of the central macular thickness (CMT), thickness of the neurosensory retina (TNeuro), and serous retinal thickness (SRT).¹² These parameters were measured as follows: The CMT was calculated as the distance from the inner limiting membrane to the basal membrane of the retinal pigment epithelium (including all compartments between), the TNeuro was the thickness of the subfoveal neurosensory retina, and SRT was the subfoveal SRT. Measurements were performed with calipers incorporated into the software of the OCT machine by two retinal specialists who were blinded to the BCVA status and cytokine levels of the subjects. 

Sample Collection

All of the CRVO patients received intravitreal injection of anti-VEGF agents (1.25 mg bevacizumab, 0.5 mg ranibizumab, or 2 mg aflibercept) and aqueous humor samples were obtained at the same time. A mean volume of 0.1 mL aqueous humor was collected by anterior chamber limbal paracentesis with a 30-gauge needle attached to an insulin syringe. Then injection of bevacizumab, ranibizumab, or aflibercept was performed through the pars plana at 3.5 mm from the limbus. Antibiotic ointment was given after surgery for 7 days. Immediately after collection, aqueous humor samples were transferred to sterile plastic tubes and stored at −80°C until analysis. Control aqueous samples were collected from 15 patients undergoing routine cataract surgery by limbal paracentesis, and the samples were stored frozen at −80°C. 

Measurement of Cytokines and Growth Factors

Samples were analyzed using suspension array technology (xMAP; Luminex Corp., Austin, TX, USA).¹² Capture bead kits (Beadlyte; Upstate Biotechnology, Lake Placid, NY, USA) were used for the detection of sVEGFR-1, sVEGFR-2, VEGF, placental growth factor (PlGF), soluble intercellular adhesion molecule (sICAM)-1, monocyte chemotactic protein 1 (MCP-1), platelet-derived growth factor (PDGF)-AA, IL-6, IL-8, IL-12 (p70), and IL-13. Samples of undiluted aqueous humor (25 μL) were incubated overnight (16–18 hours) for PlGF and sICAM1 or for 2 hours to measure the other factors. Kits were used according to the manufacturer's instructions. Standard curves for each cytokine were generated (in duplicate) by using the reference set of cytokine concentrations supplied in each kit. All incubation steps were performed at room temperature and in the dark. Samples were read on the suspension array system. To avoid between-run imprecision, we measured cytokines in the samples from all patients in a single run. Control samples were included in all runs. The levels of these factors in the aqueous humor samples were within the detection ranges of the assays, with the minimum detectable concentration being 1.59 pg/mL for sVEGFR-1, 44.81 pg/mL for sVEGFR-2, 0.64 pg/mL for VEGF, 0.37 pg/mL for PlGF, 0.03 ng/mL for sICAM-1, 1.2 pg/mL for MCP-1, 0.64 pg/mL for PDGF-AA, 0.29 pg/mL for IL-6, 0.14 pg/mL for IL-8, 0.14 pg/mL for IL-12(p70), and 0.12 pg/mL for IL-13. 

Statistical Analysis

Analyses were performed with SAS System 9.3 software (SAS Institute, Inc., Cary, NC, USA). Student's t-test was used to compare normally distributed unpaired continuous variables between the two groups, while the Mann-Whitney U test was used for other variables with a skewed distribution. The χ² test or Fisher's exact test was used to compare discrete variables. Differences between the median aqueous levels were assessed with the Wilcoxon single-rank test. To examine relationships among the variables, Spearman's rank-order correlation coefficients or Pearson's correlation coefficients were calculated. Two-tailed P values of less than 0.05 were considered to indicate statistical significance. 

Results

The CRVO group (19 men and 19 women) was aged 72.2 ± 10.7 years (mean ± SD), while the control group (seven men and eight women) was aged 67.9 ± 3.2 years. There was no significant difference of age (P = 0.126) or sex (P = 0.827) between the control and CRVO groups. The mean duration of macular edema was 46 ± 43 days (range, 11–180 days). The mean CMT, TNeuro, and SRT of the eyes with CRVO was 827 ± 248 μm, 615 ± 140 μm, and 213 ± 172 μm, respectively. Mean BCVA in the eyes with CRVO was 0.82 ± 0.58 logMAR units. Of the 38 CRVO patients, 23 (61%) had hypertension and 11 (29%) had hyperlipidemia. Two of the 15 control patients (13%) had hypertension, while 3 (20%) had hyperlipidemia. There was a significant difference in the prevalence of hypertension between the control and CRVO groups (P = 0.002), but there was no significant difference of hyperlipidemia (P = 0.630) between the two groups. 

Measurement of the aqueous humor cytokine levels showed significantly higher concentrations of sVEGFR-1, sVEGFR-2, VEGF, PlGF, sICAM-1, MCP-1, PDGF-AA, IL-6, IL-8, IL-12(p70), and IL-13 in the CRVO group than in the control group in descending order (Table 1). 

Table 1

 

Aqueous Humor Factors in the Control and CRVO Groups

	Control	CRVO	Fold Change in CRVO vs. Control	P Value
VEGF, pg/mL	0.64 (0.64–53.1)	98.8 (49.8–195)	×154	<0.001
sICAM-1, ng/mL	0.14 (0.04–0.20)	6.75 (1.10–13.3)	×48.2	<0.001
IL-8, pg/mL	1.84 (1.44–3.45)	34.3 (16.2–76.6)	×18.6	<0.001
IL-12, pg/mL	0.14 (0.14–0.14)	2.28 (0.14–3.12)	×16.3	<0.001
PlGF, pg/mL	0.37 (0.37–3.99)	5.36 (1.40–12.5)	×14.5	<0.001
IL-6, pg/mL	1.04 (0.54–3.75)	14.3 (6.50–28.1)	×13.8	<0.001
MCP-1, pg/mL	577 (474–865)	2351 (1504–3208)	×4.07	<0.001
sVEGFR-2, pg/mL	161 (51.0–229)	619 (479–762)	×3.84	<0.001
sVEGFR-1, pg/mL	648 (357–943)	1778 (1330–2697)	×2.74	<0.001
IL-13, pg/mL	0.63 (0.12–2.02)	1.60 (0.50–4.55)	×2.54	0.033
PDGF-AA, pg/mL	10.1 (5.68–15.6)	23.5 (17.2–35.4)	×2.33	<0.001

 Values are the median (interquartile range).

Aqueous humor levels of sVEGFR-1, VEGF, PlGF, MCP-1, IL-6, and IL-8 were significantly higher in the CRVO patients with retinal ischemia than in those without ischemia (Table 2). In contrast, the aqueous humor levels of sVEGFR-2, PDGF-AA, sICAM-1, IL-12, and IL-13 did not show a significant difference between CRVO patients with or without retinal ischemia (Table 2). 

Table 2

 

Aqueous Humor Factors in Patients with Ischemic CRVO or Nonischemic CRVO

Variable	Nonischemic	Ischemic	Fold Change in Ischemic vs. Nonischemic	P Value
PlGF, pg/mL	3.52 (0.91–7.07)	22.1 (8.22–29.2)	×6.28	0.002
IL-13, pg/mL	0.97 (0.43–4.55)	3.84 (0.66–4.78)	×3.98	0.299
VEGF, pg/mL	76.6 (46.2–124)	302 (194–894)	×3.94	<0.001
IL-8, pg/mL	21.1 (12.8–44.7)	70.1 (43.2–143)	×3.32	<0.001
IL-6, pg/mL	10.5 (5.67–27.2)	20.3 (13.0–32.8)	×1.93	0.047
sVEGFR-1, pg/mL	1667 (1261–2282)	2997 (1650–3998)	×1.80	0.012
MCP-1, pg/mL	1874 (1362–2990)	2943 (2351–3992)	×1.57	0.010
sICAM-1, ng/mL	5.20 (1.02–13.1)	7.80 (2.16–21.8)	×1.50	0.593
IL-12, pg/mL	1.91 (0.14–3.12)	2.72 (0.92–3.38)	×1.42	0.286
PDGF-AA, pg/mL	20.2 (15.4–35.4)	28.5 (21.1–35.2)	×1.41	0.197
sVEGFR-2, pg/mL	620 (496–763)	597 (440–731)	×0.96	0.887

 Values are the median (interquartile range).

The relationship between aqueous humor levels of the above factors or cytokines and the three OCT parameters (CMT, TNeuro, and SRT) also was assessed. As a result there were significant correlations between CMT and the aqueous levels of VEGF, PlGF, MCP-1, IL-6, and IL-8 in the CRVO group (Table 3). Significant correlations also were found between TNeuro and the levels of sVEGFR-1, VEGF, PlGF, MCP-1, IL-6, and IL-8 in this group (Table 3). Furthermore, there were significant correlations between SRT and the aqueous levels of MCP-1 and IL-8 (Table 3). 

Table 3

 

Correlations Between Aqueous Humor Factors and the OCT Parameters

Variable	sVEGFR-1	sVEGFR-2	VEGF	PlGF	PDGF-AA	sICAM-1	MCP-1	IL-6	IL-8	IL-12(p70)	IL-13
	r, P Value	r, P Value	r, P Value	r, P Value	r, P Value	r, P Value	r, P Value	r, P Value	r, P Value	r, P Value	r, P Value
CMT	0.25, 0.125	0.02, 0.933	0.34, 0.037	0.42, 0.010	0.03, 0.847	0.21, 0.209	0.50, 0.002	0.37, 0.023	0.52, <0.001	0.05, 0.761	−0.08, 0.617
TNeuro	0.36, 0.027	0.27, 0.107	0.41, 0.012	0.42, 0.011	0.12, 0.461	0.16, 0.326	0.43, 0.009	0.37, 0.024	0.40, 0.014	0.29, 0.067	0.21, 0.195
SRT	0.10, 0.549	−0.18, 0.255	0.12, 0.461	0.22, 0.174	−0.05, 0.747	0.19, 0.241	0.45, 0.005	0.22, 0.181	0.48, 0.002	−0.17, 0.303	−0.29, 0.071

r  , correlation coefficient.

In the CRVO group, significant correlations were noted between the level of sVEGFR-1 in the aqueous humor and the levels of sVEGFR-2, VEGF, PlGF, PDGF-AA, MCP-1, IL-6, and IL-8 (Table 4). There also were significant correlations between the aqueous VEGF level and the levels of PlGF, PDGF-AA, MCP-1, IL-6, and IL-8 (Table 4). Furthermore, there were significant correlations between the PlGF level and the levels of PDGF-AA, sICAM-1, MCP-1, IL-6, and IL-8 (Table 4). Moreover, significant correlations were found between the PDGF-AA level and the levels of MCP-1, IL-6, and IL-8 (Table 4), as well as a significant correlation between sICAM-1 and MCP-1, or IL-8 (Table 4). There also was a significant correlation between MCP-1 and IL-6 or IL-8 (Table 4), as well as between IL-6 and IL-8 (Table 4). Finally, there was a significant correlation between IL-12 (p70) and IL-13 (Table 4). 

Table 4

 

Correlation Matrix For Aqueous Humor Factors

Variable	sVEGFR-1	sVEGFR-2	VEGF	PlGF	PDGF-AA	sICAM-1	MCP-1	IL-6	IL-8	IL-12	IL-13
	r, P Value	r, P Value	r, P Value	r, P Value	r, P Value	r, P Value	r, P Value	r, P Value	r, P Value	r, P Value	r, P Value
sVEGFR-1		0.49, 0.003	0.58, <0.001	0.34, 0.039	0.43, 0.009	−0.06, 0.697	0.32, 0.049	0.42, 0.010	0.48, 0.002	−0.11, 0.461	−0.01, 0.968
sVEGFR-2			0.17, 0.309	0.25, 0.127	0.30, 0.066	0.29, 0.077	0.21, 0.199	0.19, 0.259	0.10, 0.533	−0.01, 0.943	−0.16, 0.317
VEGF				0.60, <0.001	0.46, 0.005	−0.07, 0.649	0.54, 0.001	0.47, 0.004	0.59, <0.001	−0.15, 0.374	−0.07, 0.681
PlGF					0.32, 0.049	0.39, 0.019	0.68, <0.001	0.53, 0.001	0.69, <0.001	−0.06, 0.731	−0.06, 0.727
PDGF-AA						0.28, 0.090	0.46, 0.005	0.42, 0.010	0.34, 0.037	−0.03, 0.834	−0.21, 0.199
sICAM-1							0.44, 0.008	0.07, 0.687	0.36, 0.029	−0.06, 0.693	−0.20, 0.228
MCP-1								0.68, <0.001	0.77, <0.001	0.01, 0.967	−0.13, 0.431
IL-6									0.57, <0.001	0.06, 0.718	0.07, 0.682
IL-8										−0.12, 0.464	−0.18, 0.286
IL-12											0.74, <0.001

We also assessed the relationship between each factor or cytokine and two possible confounders, which were a history of hypertension and the duration of CRVO. When each factor/cytokine was compared between subjects with or without hypertension, there were no significant differences (data not shown). In addition, none of the factors/cytokines was significantly correlated with the duration of CRVO (data not shown). When each factor/cytokine was compared between the 4 subjects with and the 34 subjects without pseudophakia, there were no significant differences (data not shown), indicating that pseudophakia did not affect the levels of these factors/cytokines. 

Discussion

In the present study, we investigated the association between retinal ischemia or the severity of macular edema and aqueous humor levels of VEGFRs, their ligand, other growth factors, and inflammatory factors or cytokines in patients with CRVO. First, we found that the aqueous humor levels of sVEGFR-1 and sVEGFR-2 were significantly higher in eyes with CRVO than in control eyes with cataract, and that there was a significant correlation between sVEGFR-1 and sVEGFR-2, suggesting a concordant increase in the expression of both receptors in CRVO patients with macular edema. Soluble VEGFR-1 and -2 are produced by alternative mRNA splicing, with the same gene encoding the transmembrane forms of these receptors^18–20 or the soluble forms that are released from the cell surface.²¹ In human umbilical vein endothelial cells (HUVEC) stimulated by phorbol 12-myristate 13-acetate or bFGF, northern blotting demonstrated a correlation between expression of sVEGFR-1 and transmembrane VEGFR-1.²² Therefore, expression of transmembrane VEGFR-1 and VEGFR-2 may be upregulated along with that of sVEGFR-1 and sVEGFR-2 in the regulation of vascular permeability. Previous studies have demonstrated that VEGF signaling is tightly regulated by VEGFR-1 and VEGFR-2 in vivo.²³ The VEGF binds to VEGFR-1 and VEGFR-2, which have tyrosine kinase activity and are expressed by vascular endothelial cells.^24,25 Vascular endothelial growth factor receptor 1 is also expressed by monocytes/macrophages^26,27 at the mRNA and protein levels,^27,28 and VEGFR-1 signaling has a role in recruitment of these cells to sites of angiogenesis and inflammation by VEGF.^27,29–31 In contrast, VEGFR-2 is exclusively expressed by endothelial cells and VEGF signaling via membrane-bound VEGFR-2 influences normal endothelial cell function, vascular permeability, and angiogenesis.^23,32–34 Therefore, these VEGFRs may well be important for the development of macular edema associated with CRVO. 

There was a significant correlation between TNeuro and the aqueous level of sVEGFR-1 in the CRVO group. Macular edema is promoted by an increase of VEGF, which binds to VEGFRs expressed on vascular endothelial cells, monocytes, and macrophages. It has been reported that sVEGFR-1 promotes inflammation.^27,35,36 This is supported by our finding that the aqueous humor level of sVEGFR-1 was significantly correlated with the aqueous levels of various inflammatory factors (MCP-1, PDGF, IL-6, and IL-8). There is evidence that sVEGFR-1 influences vascular permeability in CRVO. On the other hand, there was no significant correlation between the aqueous humor level of sVEGFR-2 and the severity of macular edema in this study, as well as no significant correlation between the vitreous fluid level of sVEGFR-2 and macular edema in a previous study.⁷ Moreover, this study demonstrated that sVEGFR-2 was not correlated with inflammatory factors in aqueous humor, although the sVEGFR-1 level was correlated with various inflammatory factors. These results suggested that macular edema in CRVO patients may be more strongly influenced by inflammation mediated via VEGFR-1 than by changes of vascular permeability mediated via VEGFR-2. Inflammation may be dominant in the pathogenesis of macular edema associated with CRVO because there is widespread retinal damage and persistent injury to retinal cells, but further investigations will be required to confirm the roles of sVEGFR-1 and sVEGFR-2 in this condition. 

This study demonstrated that the aqueous humor level of PlGF was significantly higher in the CRVO group than in the control group. The PlGF is a member of the VEGF family,^37,38 and it modulates inflammation as a specific ligand of VEGFR-1, stimulating tissue factor production and chemotaxis by monocytes/macrophages.²⁷ After binding to VEGFR-1, PlGF also increases the production of proinflammatory factors by cultured monocytes via a calcineurin-dependent pathway,³⁹ suggesting a direct influence on the inflammatory response. This is supported by our finding that the aqueous humor level of PlGF was correlated with the levels of sVEGFR-1, sICAM-1, MCP-1, IL-6, and IL-8. It was reported that PlGF induces the secretion of VEGF by mononuclear cells,⁴⁰ and this report is supported by our finding of a significant correlation between the aqueous humor levels of PlGF and VEGF. In addition, the PlGF level in aqueous humor was significantly correlated with the CMT and TNeuro, and there also was a correlation of VEGF with CMT and TNeuro. These results suggest that PlGF may act together with VEGF when macular edema develops in patients with CRVO. 

We also found that PDGF-AA was significantly increased in aqueous humor from the CRVO group compared with the controls. The PDGF is a growth factor involved in regulating the migration of mesenchymal cells. Expression of PDGF-AA by endothelial cells was reported to increase when blood flow was decreased by arteriosclerosis.⁴¹ Because the aqueous level of PDGF-AA was correlated with those of MCP-1, sICAM-1, IL-6, and IL-8 in the present study, it seems that PDGF-AA also contributes to inflammation. In fact, the aqueous level of PDGF-AA was not significantly correlated with the severity of macular edema in our CRVO patients, although it was reported that PDGF-AA promotes the formation of tight junction.⁴² In addition, the aqueous level of PDGF-AA was not significantly higher in the CRVO patients with retinal ischemia than in those without ischemia. Therefore, these findings suggest that PDGF-AA influences inflammation rather than vascular permeability or retinal ischemia in CRVO patients with macular edema. 

We previously reported that the levels of three inflammatory factors (sICAM-1, MCP-1, and IL-6) were significantly higher in vitreous fluid samples from CRVO patients compared with controls and were significantly correlated with each other, suggesting that these inflammatory factors had an important role in macular edema associated with CRVO.⁷ In this study, we likewise found that aqueous levels of inflammatory factors/cytokines (IL-6, IL-8, sICAM-1, and MCP-1) were significantly elevated in the CRVO group compared to the controls and that most of these inflammatory factors were significantly correlated with each other. In particular, the aqueous levels of MCP-1 and IL-8 were significantly correlated with all three parameters of macular edema (CMT, TNeuro, and SRT), suggesting that MCP-1 and IL-8 may be the most important inflammatory factors involved in the development of macular edema associated with CRVO. 

As detailed above, various factors and cytokines (VEGFRs, their ligands, growth factors, and inflammatory factors/cytokines) are involved in the pathogenesis of macular edema associated with CRVO. This study demonstrated that the aqueous level of sVEGFR-1 was correlated with the levels of ligands and growth factors (VEGF, PlGF, and PDGF-AA) and with the levels of inflammatory factors/cytokines (MCP-1, IL-6, and IL-8), while sVEGFR-2 was not correlated with any of these factors. Therefore, upregulation of growth factors and inflammatory factors/cytokines in CRVO patients may depend on sVEGFR-1 more than sVEGFR-2. This suggests that multiple factors could be inhibited by targeting VEGFR-1, so it may be worth considering anti-VEGFR-1 therapy for treating macular edema in CRVO patients. Indeed, it has already been reported that VEGFR-1 inhibitors are effective for diabetic macular edema and age-related macular degeneration.^43–45 Of course, a prospective trial would be required to determine the effectiveness of such therapy in CRVO. 

In conclusion, the aqueous humor levels of various factors/cytokines were significantly higher in patients with CRVO than in control subjects. The aqueous level of sVEGFR-1 was correlated with TNeuro and with the aqueous levels of its ligand and other growth factors (VEGF, PlGF, and PDGF-AA), as well as with inflammatory factors/cytokines (MCP-1, IL-6, and IL-8). Aqueous levels of the growth factors (VEGF, PlGF, and PDGF-AA) also were significantly correlated with each other, as were the aqueous levels of sVEGFR-1 and sVEGFR-2. These findings suggest the importance of further investigation into the relations among growth factors, inflammatory factors, and cytokines, and may contribute to better understanding of macular edema in CRVO patients and to development of new treatments targeting VEGFR-1. 

Acknowledgments

The authors thank Katsunori Shimada (Department of Biostatistics, STATZ Corporation, Tokyo, Japan) for assistance with the statistical analysis. 

Supported in part by a Grant-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology of Japan. The authors alone are responsible for the content and writing of the paper. 

Disclosure: H. Noma, None; T. Mimura, None; K. Yasuda, None; M. Shimura, None