Simultaneous Analysis of Multiple Cytokines in the Vitreous of Patients with Sarcoid Uveitis

Abstract

Purpose.: Levels of some cytokines are significantly higher in the vitreous fluid of patients with acute uveitis than in normal vitreous fluid. The authors sought to determine which proinflammatory cytokines were upregulated in the vitreous fluid of patients with ocular sarcoidosis.

Methods.: Samples of vitreous fluid were collected from patients with sarcoid uveitis and from nonsarcoid control patients with idiopathic epiretinal membrane. The levels of 27 proinflammatory cytokines were measured with a multiplex beads array system. Postvitrectomy macular thickness was also measured by using spectral domain optical coherence tomography (SD-OCT). To assess the relationship between cytokine levels and disease stage, the authors divided patients into three groups based on macular thickness 1 month after operation.

Results.: The vitreous levels of 17 cytokines were significantly higher in patients with ocular sarcoidosis than in nonsarcoid controls. Serum levels of interferon γ–induced protein 10 (IP-10) were also higher in ocular sarcoidosis patients than in nonsarcoid controls. Conversely, serum levels of interleukin (IL) 15 in ocular sarcoidosis patients were lower than in the control group. Analysis of cytokine levels and macular thickness revealed that IL-1ra, IL-4, IL-8, IFN-γ, IP-10, monocyte chemotactic protein (MCP)-1, macrophage inflammatory protein (MIP)-1β, and regulated on activation, normal T-cell expressed and secreted (RANTES) were significantly upregulated in patients with thin cystoid macular edema group.

Conclusions.: Patients with ocular sarcoidosis had elevated levels of proinflammatory cytokines in vitreous fluids. Different cytokines might contribute to different stages of macular edema.

Introduction

Sarcoidosis is a chronic multisystem granulomatous disorder of unknown etiology. ^1–4 It is thought to result from an exaggerated cellular immune response to a variety of self-antigens or nonself antigens. ⁵ The lung is the organ most frequently affected by this disease, and when affected, it is characterized by bilateral hilar lymphadenopathy and/or pulmonary infiltration as revealed by chest radiography. Ocular and skin lesions are also common aspects of sarcoidosis, but other organs (including the heart, the central nervous system, and the spleen) may also be affected. ⁶ Although the exact cause of sarcoidosis is currently unknown, previous reports have discussed environmental (e.g., spatial ⁷ or climatic ⁸ ) factors, occupational factors, or infectious agents (specifically, propionibacteria ^9–11 ) as possible causes. 

Reportedly, 30% to 60% of patients with sarcoidosis suffer from ocular involvement. ^12–15 Bilateral anterior and/or posterior uveitis is common, but the conjunctiva, lacrimal gland, and orbit of the eye can also be affected. The clinical presentation of sarcoidosis-related uveitis is characteristically marked by iris nodules, mutton-fat keratic precipitates, and tent-shaped peripheral anterior synechia in the anterior segment of the eye. Snowball-like vitreous opacity, which results from phlebitis and vitreitis, is a common posterior segment finding. Chronic uveitis can result in formation of an epiretinal membrane (ERM) and cystoid macular edema (CME); consequently, severe visual impairment can occur. ¹⁶ Uveitis is commonly treated with topical (or occasionally systemic) administration of corticosteroids, but in cases accompanied by the formation of an ERM or long-standing vitreous opacity, surgical treatment is necessary. 

An internationally acknowledged set of criteria for the diagnosis of ocular sarcoidosis has been established. ¹⁷ Previously, it has been reported that the types of cytokines in ocular fluids are dependent on the disease. ^18,19 The authors thus sought to assess vitreous cytokine levels and correlate them with disease status in ocular sarcoidosis. The multiplex bead analysis system is a new technique that combines the principle of the sandwich immunoassay with fluorescent bead-based technology. Levels of many types of cytokines in small sample volumes (such as vitreous samples) can be measured simultaneously with this system. Sato et al. ²⁰ have used this method to show that the levels of vascular endothelial growth factor A (VEGF–A) are much higher than those of any type of cytokine in infants with retinopathy of prematurity (ROP). There are also several reports about the levels of various cytokines in vitreous fluid from uveitis patients. ^21–25 Yoshimura et al. ¹⁸ reported that interleukin (IL) 6, IL-8, and monocyte chemotactic protein-1 (MCP-1) were elevated in the vitreous body of patients with diabetic retinopathy and retinal vein occlusion.  

Analyses of many cytokines in peripheral blood or bronchoalveolar lavage fluid (BALF) have elucidated the pathogenesis of sarcoidosis. For example, high amounts of interferon γ–induced protein 10 (IP-10) are secreted from sarcoid-dependent alveolar macrophages and T cells. ²⁶ However, to the authors' knowledge, investigations of cytokine levels in both vitreous fluid and serum from patients with sarcoid uveitis have not been conducted with multiplex bead analysis. 

Here, the vitreous and serum levels of 27 types of cytokines in patients with sarcoid uveitis were measured and compared with those from patients with idiopathic ERM. Moreover, the authors investigated the relationship between vitreous cytokine levels and CME severity. 

Methods

Patients

This study was performed in accordance with the tenets of the Declaration of Helsinki, and the procedures were approved by the Institutional Review Board of the Kyoto Prefectural University of Medicine Hospital. Twenty-one patients gave informed consent for participation in this study. 

For the multiplex bead analysis of vitreous fluid, a total of 24 samples were enrolled. Of the 20 patients, 15 (1 man and 14 women) had diagnoses of ocular sarcoidosis for one or both eyes (19 eyes total), based on the international criteria. Vitreous fluid was collected from five eyes of five idiopathic ERM patients (3 men and 2 women), who were enrolled as nonsarcoid controls. 

General examinations of all patients were performed by using the algorithm designed for the diagnosis of ocular sarcoidosis.¹⁷  

The group of patients with ocular sarcoidosis comprised 1 man and 14 women; the idiopathic ERM (nonsarcoid control) group comprised 2 men and 3 women. The mean age of the ocular sarcoidosis patients was 66.8 ± 2.2 years, and that of the idiopathic ERM patients was 69.5 ± 3.1 years. Statistical differences were not found in mean age.  

Vitreous Sample Collection

Vitreous specimens were obtained from each of 20 patients at the start of a conventional 25-gauge pars plana vitrectomy by using either a CV-24,000 (NIDEK Co., Ltd., Aichi, Japan) or an Accurus (Alcon Laboratories, Inc., Fort Worth, TX) vitrectomy system. A three-way stopcock was attached to the connector on the suction-tube line of the cutter probe, and a 5-ml syringe was connected to the free end of the three-way stopcock. Dry vitrectomy without perfusion of balanced salt solution (Alcon Laboratories, Inc.) was conducted with a cut rate of 500 cpm so as not to damage cells infiltrating into the vitreous. At least 500 μL dry vitreous sample was collected from each patient. Each sample was divided into 5 microtubes for each analysis (200 μL for PCR, 200 μL for multiplex bead analysis, and 100 μL for culture). No intraoperative complications occurred in these patients. 

Multiplex Bead Analysis System (Multiplex-ELISA)

The vitreous levels of 27 types of cytokines were determined by using a commercially available multiplex bead analysis system (BioPlex Pro Suspension Array System; BioRad Laboratories, Tokyo, Japan). The 27 cytokines measured were IL-1b; IL-1 receptor antagonist (IL-1ra); IL-2; IL-4; IL-5; IL-6; IL-7; IL-8; IL-9; IL-10; IL-12; IL-13; IL-15; IL-17; eotaxin; basic fibroblast growth factor (bFGF); granulocyte colony-stimulating factor (G-CSF); granulocyte macrophage colony-stimulating factor (GM-CSF); IFN-γ; IFN-γ–IP-10; MCP-1; macrophage inflammatory protein (MIP)-1α and MIP-1β platelet-derived growth factor BB (PDGF-BB); regulated on activation, normal T-cell expressed and secreted (RANTES); tumor necrosis factor α (TNF-α); and VEGF.  

Multiplex PCR

Human herpes virus (HHV) genomic DNA was assayed in vitreous fluids by using two independent PCR assays (a qualitative multiplex PCR assay and a quantitative real-time PCR assay) as described previously. ^27,28 DNA was extracted from samples by using an E21 virus minikit (QIAGEN, Inc., Valencia, CA) installed on a robotic workstation for automated purification of nucleic acids (BioRobot E21; QIAGEN). Multiplex PCR was designed to qualitatively identify the genomic DNA of the following 8 types of HHV: herpes simplex virus type 1 (HSV-1 or HHV-1) and type 2 (HSV-2 or HHV-2), varicella zoster virus (VZV or HHV-3), Epstein-Barr virus (EBV or HHV-4), cytomegalovirus (CMV or HHV-5), HHV6, HHV7, and HHV8. Other ocular pathogens were also tested, such as Propionibacterium acnes , Toxoplasma, Toxocara, Bartonella henselae , Chlamydia trachomatis , Treponema pallidum , Mycobacterium tuberculosis , Candida (18s rRNA), Aspergillus (18s rRNA), and bacterial 16s rRNA. PCR was performed with a LightCycler (Roche Diagnostics [Schweiz] AG, Rotkreuz, Switzerland). Primers and probes used to detect HHV1 to HHV8 and the PCR conditions have all been described previously. ²⁷ Specific primers for each virus were used with Accuprime Taq (Invitrogen, Carlsbad, CA). The templates were subjected to 40 cycles of PCR amplification, and probes were then mixed with the PCR products. Subsequently, real-time PCR was performed only for the HHVs, with the genomic DNA detected by multiplex PCR. The real-time PCR was performed with AmpliTaq Gold (Applied Biosystems, Foster City, CA) and the Real-Time PCR 7300 system (Applied Biosystems). All of the templates obtained were subjected to 45 cycles of PCR amplification. The value of the viral copy number in the sample was considered to be significant when more than 50 copies per tube (5 × 10³ copies/mL) were observed.  

Macular Thickness Measurement in Ocular Sarcoidosis Patients

The patients who received surgical intervention and cytokine measurements were also evaluated for macular thickness. Macular 3D scans with the spectral-domain 3D-OCT 1000 (TOPCON, Itabashi, Japan) were performed at 1 month after surgery. Patients were divided into two groups on the basis of macular thickness, which was calculated as the average thickness of macular area (fovea 1000 μm across) ± SD. The thin CME range (group 1; 10 patients and 12 eyes) was 119.60 μm to 359.5 μm, and the thick CME range (group 2; 9 patients and 10 eyes) was 359.5 μm to 599.40 μm. Some patients overlapped because of different status of disease in each eye.  

Statistical Analysis

Statistical comparisons of the cytokine concentrations of the two patient groups (sarcoidosis and ERM) were carried out by nonparametric analysis with the Mann-Whitney U test. The statistical evaluation of the data was performed by using Bonferroni correction. A P value <0.025 was considered to be statistically significant. Statistical comparisons of the macular thickness of three groups (CME grade) were carried out by nonparametric analysis with the Mann-Whitney U test. 

Results

Vitreous and Serum Levels of 27 Types of Cytokines

The vitreous and serum levels of cytokines are shown in Tables 1 and 2. The levels of PDGF-BB, IL-1ra, IL-4, IL-5, IL-6, IL-8, IL-9, IL-10, IL-12, G-CSF, IFN-γ, IP-10, MCP-1α, MIP-1β, RANTES, TNF-α, and VEGF were significantly higher in the vitreous from patients with ocular sarcoidosis than in that of ERM (nonsarcoid control) patients. Interestingly, although the levels of these cytokines were much higher in vitreous fluid of patients with ocular sarcoidosis, the differences in the serum cytokine levels between these two groups (with only two exceptions) were not significant. Serum IP-10 was significantly higher in patients with ocular sarcoidosis than in those with ERM, but the serum levels of IL-15 were significantly lower than in patients with ERM. 

Table 1.  

 

Vitreous Fluid Levels of 27 Types of Cytokines Determined by Using a Multiplex Bead Analysis System

	Sarcoidosis Vitreous, Mean ± SE (N = 19)	ERM Vitreous, Mean ± SE (N = 5)	P Value
PDGF-BB	74.54 ± 12.89	1.29 ± 0.81	0.0043*
IL-1β	0.60 ± 0.08	1.07 ± 0.78	0.1766
IL-1ra	225.1 ± 69.88	7.93 ± 1.76	0.0008†
IL-2	0.885 ± 0.36	0	0.2526
IL-4	0.67 ± 0.09	0.02 ± 0.02	0.0021*
IL-5	0.41 ± 0.07	0.02 ± 0.02	0.0061*
IL-6	736 ± 276.5	13.8 ± 4.2	0.0008†
IL-7	26.93 ± 3.59	9.68 ± 3.27	0.0646
IL-8	161.3 ± 65.95	16.78 ± 7.52	0.0045*
IL-9	28.87 ± 9.04	1.96 ± 1.01	0.0014*
IL-10	13.31 ± 2.79	0.44 ± 0.14	0.0008†
IL-12	33.10 ± 12.02	1.40 ± 1.23	0.0055*
IL-13	45.35 ± 8.74	5.67 ± 4.51	0.0251
IL-15	10.66 ± 1.46	2.68 ± 1.46	0.0257
IL-17	1.49 ± 0.49	0	0.3515
Eotaxin	3.77 ± 0.79	0.49 ± 0.36	0.1409
bFGF	5.98 ± 2.28	2.28 ± 2.28	0.8413
G-CSF	94.88 ± 13.57	7.20 ± 5.70	0.0055*
GM-CSF	130.9 ± 7.63	95.72 ± 10.1	0.088
IFN-γ	57.89 ± 8.69	0.98 ± 0.98	0.0008†
IP-10	58,249 ± 4564	163.5 ± 48.72	0.0008†
MCP-1	1556 ± 232	321.1 ± 48.9	0.0018*
MIP-1α	2.413 ± 0.56	0	0.052
MIP-1β	66.98 ± 9.89	6.06 ± 2.04	0.0008†
RANTES	71.21 ± 15.75	0.52 ± 0.52	0.0017*
TNF-α	17.69 ± 2,12	3.24 ± 1.62	0.0055*
VEGF	155.0 ± 99.59	18.88 ± 17.67	0.0190‡

*  More significant difference (P < 0.01).

†  Most significant difference (P < 0.001).

‡  Significant difference (P < 0.025).

Table 2.  

 

Serum Levels of 27 Types of Cytokines Determined by Using a Multiplex Bead Analysis System

	Sarcoidosis Vitreous, Mean ± SE (N = 19)	ERM Serum, Mean ± SE (N = 5)	P Value
PDGF-BB	8703 ± 511.7	8526 ± 1857	0.887
IL-1β	3.72 ± 0.58	5.91 ± 1.91	0.2268
IL-1ra	206.9 ± 22.42	223.3 ± 42.87	0.887
IL-2	5.12 ± 3.04	54.21 ± 49.40	0.721
IL-4	6.96 ± 1.25	5.76 ± 2.04	0.4342
IL-5	2.76 ± 0.23	25.99 ± 23.60	0.8868
IL-6	28.51 ± 10.50	116.1 ± 67.50	0.1355
IL-7	9.79 ± 0.89	8.12 ± 2.45	0.3935
IL-8	415.7 ± 193.3	791.3 ± 340.3	0.2554
IL-9	115.7 ± 51.05	59.81 ± 22.51	0.5695
IL-10	5.325 ± 0.91	5.43 ± 1.64	0.8869
IL-12	40.26 ± 6.12	83.49 ± 46.87	0.6698
IL-13	5.98 ± 1.59	7.79 ± 3.54	0.4772
IL-15	0	1.522 ± 1.27	0.0061*
IL-17	55.23 ± 8.65	48.43 ± 13.96	0.4339
Eotaxin	117.80 ± 18.12	117.0 ± 28.97	0.9433
bFGF	23.11 ± 5.93	43.42 ± 21.66	0.5426
G-CSF	28.46 ± 4.19	28.52 ± 8.69	1
GM-CSF	8.56 ± 2.67	77.93 ± 62.45	0.1803
IFN-γ	63.06 ± 7.97	124.9 ± 62.16	0.6187
IP-10	3485 ± 555.2	1053 ± 108.3	0.0105†
MCP-1	47.02 ± 10.83	64.83 ± 29.95	1
MIP-1α	68.39 ± 31.38	269.9 ± 161.6	0.1176
MIP-1β	752.7 ± 171.5	1777 ± 789.1	0.2007
RANTES	4318 ± 180.5	5077 ± 520.2	0.1768
TNF-α	52.21 ± 8.79	172.2 ± 76.81	0.0941
VEGF	103.9 ± 13.71	109.4 ± 30.01	0.9433

*  More significant difference (P < 0.01).

†  Significant difference (P < 0.025).

The cytokine ratios of vitreous/serum in ocular sarcoidosis and ERM are shown in Table 3. The vitreous/serum ratios of PDGF-BB, IL-1ra, IL-4, IL-5, IL-6, IL-12, IL-13, G-CSF, IFN-γ, IP-10, MCP-1, MIP-1β, RANTES, TNF-α, and VEGF were significantly higher in ocular sarcoidosis samples than in those of ERM (nonsarcoid control) patients. 

Table 3.  

 

Vitreous/Serum Ratio of 27 Types of Cytokines

	Sarcoidosis Vitreous/Serum, Mean ± SE (N = 19)	ERM Vitreous/Serum, Mean ± SE (N = 5)	P Value
PDGF-BB	0.006 ± 0.001	0.0001 ± 0.000009	0.0054*
IL-1β	0.61 ± 0.33	0.22 ± 0.20	0.0753 ns
IL-1ra	2.59 ± 1.26	0.03 ± 0.01	0.0022*
IL-2	4.45 ± 2.83	0	0.1470 ns
IL-4	0.39 ± 0.14	0.002 ± 0.002	0.0017*
IL-5	0.21 ± 0.03	0.01 ± 0.01	0.0096*
IL-6	90.69 ± 25.11	0.93 ± 0.72	0.0018*
IL-7	2.97 ± 0.59	1.91 ± 0.79	0.3555
IL-8	4.31 ± 1.29	0.09 ± 0.05	0.0229 ns
IL-9	1.54 ± 0.81	0.04 ± 0.03	0.0262 ns
IL-10	15.44 ± 8.79	4.49 ± 4.46	0.0283 ns
IL-12	1.12 ± 0.44	0.008 ± 0.005	0.0015*
IL-13	21.53 ± 8.94	0.53 ± 0.24	0.0129†
IL-15	Impossible to measure	0	Not measured
IL-17	0.36 ± 0.27	0	0.6944 ns
Eotaxin	0.36 ± 0.27	0	0.6944 ns
bFGF	0	0	ns
G-CSF	12.83 ± 5.06	0.27 ± 0.13	0.0028*
GM-CSF	21.86 ± 5.08	9.88 ± 7.57	0.1631 ns
IFN-γ	1.88 ± 0.61	0.02 ± 0.02	0.0008‡
IP-10	24.64 ± 3.86	0.13 ± 0.05	0.0008‡
MCP-1	58.14 ± 15.07	12.62 ± 5.51	0.0157†
MIP-1α	0.17 ± 0.05	0	0.0787
MIP-1β	0.20 ± 0.04	0.013 ± 0.01	0.0028*
RANTES	0.02 ± 0.004	0.0001 ± 0.0001	0.0017*
TNF-α	0.38 ± 0.09	0.03 ± 0.03	0.0017*
VEGF	0.81 ± 0.30	0.01 ± 0.007	0.0025*

 ns, not specified.

*  More significant difference (P < 0.01).

†  Significant difference (P < 0.025).

‡  Most significant difference (P < 0.001).

The Relationship between Postoperative Macular Thickness and Cytokine Concentrations

Data from the analysis of vitreous cytokines were compared with macular thickness 1 month after surgery (Fig.). Macular thickness in group 1 (10 patients and 12 eyes) was <359.5 μm, whereas measurements in group 2 (9 patients and 10 eyes) were ≥359.5 μm. The levels of most cytokines tend to increase in patients within the thin CME group. Notably, IL-1ra (P = 0.02), IL-2 (P = 0.03), IL-4 (P = 0.03), IL-8 (P = 0.03), IFN-γ (P = 0.03), IP-10 (P = 0.01), MCP-1 (P = 0.03), MIP-1β (P = 0.005), and RANTES (P = 0.006) were significantly upregulated in group 1. In contrast, the levels of PDGF-BB, IL-12, IL-13, bFGF, G-CSF, and VEGF seemed to be elevated in group 2 (no statistical difference).  

Discussion

To the best of the authors' knowledge, this is the first study to investigate both vitreous and serum levels of cytokines, using multiplex ELISA analysis, in patients with ocular sarcoidosis. These patients had high levels of several types of cytokines, especially T helper 1 (Th-1)–related cytokines, in their vitreous body. In their investigation of the relationship between cytokine and clinical stage, the authors noticed that advanced pathologic stage in the posterior segment (e.g., complications such as severe CME) might be influenced by G-CSF, VEGF, and PDGF-BB, which are highly associated with vascular leakage, as previously reported. ^29,30 In fact, only the cytokines that are associated with vascular permeability tended to undergo upregulation in the advanced CME stage. Previous reports have suggested that VEGF is strongly associated with blood–retina barrier (BRB) breakdown. ¹⁸ In contrast, almost all of the proinflammatory cytokines were downregulated in vitreous fluid from patients with advanced-stage ocular sarcoidosis (Fig.). It appears that numerous cytokines may be increased in order to increase the lymphocyte infiltration in vitreous fluid or macrophage infiltration into granulomas at the retinal layer. There is the potential for an increase in BRB breakdown from the high vitreous/serum ratio of cytokines, such as PDGF-BB, IL-6, G-CSF, and VEGF; if so, all inflammatory and angiogenic cytokine levels should increase in vitreous fluid after onset of the clinical condition. Moreover, the cytokine level in vitreous fluid for herpes virus or parasite infection, which might be associated with high BRB breakdown, was much higher than that for both nonocular sarcoidosis and ocular sarcoidosis. However, the number of groups studied is too limited to allow a comprehensive discussion for the present experiment. Further investigations should be performed to analyze the BRB breakdown, with comparison between blood and vitreous albumin concentration.  

Interestingly, IP-10 was the only cytokine that was found to be elevated in both the vitreous body and serum of sarcoid uveitis patients. IP-10, also known as C-X-C motif chemokine 10, is a chemokine that is induced by IFN-γ and secreted from monocytes and macrophages stimulated with IFN-γ. Because IP-10 promotes the migration of T cells to sites of inflammation, it is an attractive candidate for further investigation of the mechanisms promoting the development of sarcoid granulomas. IP-10 is strongly expressed in sarcoid granuloma tissue, ²⁶ and IP-10 secretion from alveolar macrophages is highly correlated with the CD4⁺ T lymphocyte population. This finding may indicate that IP-10 secreted from macrophages in the granuloma tissue can regulate migration of CD4⁺ T lymphocytes to the site of sarcoidosis during inflammatory processes in the posterior segment of the eye. The amount of CD4⁺ T lymphocytes is related to IP-10 levels in BALF ²⁶ ; the authors observed a similar phenomenon in vitreous fluid samples in an ongoing experiment (manuscript submitted). The vitreous fluid from patients with ocular sarcoidosis had high amounts of CD4⁺ T lymphocytes, but not CD8⁺ T lymphocytes (data not shown; manuscript submitted). Reportedly, serum levels of IP-10 in patients with sarcoidosis were significantly higher than those of healthy volunteers; however, peripheral blood mononuclear cells did not increase the IP-10 production. ³¹ The present data indicated that IP-10 might be secreted from local macrophages that had infiltrated granuloma tissue in retina, vitreous fluid, or any other place in the body. In fact, the authors found that retinas from patients with sarcoid uveitis involved granulomas that contained CD68⁺ macrophages (data not shown). Therefore, serum levels of IP-10 probably reflect the clinical state of sarcoid uveitis. High levels of Th-1–type cytokines in vitreous fluid suggest that IP-10 cooperates with Th-1–type cytokines, which act as local factors that promote T cell activation and proliferation. 

Several cytokines have angiogenic/vascular permeability properties; for example, IL-6 increases vascular permeability. ³² Moreover, VEGF is a well-known angiogenic factor that is highly associated with vascular leakage in retinas of patients suffering with diabetic retinopathy, retinal vascular occlusion, or ROP. Vitreous levels of VEGF are significantly higher in eyes of patients with proliferative ROP than in those of control individuals, and VEGF is the only cytokine for which the vitreous level is correlated with vascular activity in ROP eyes. ³³ Here, the authors intended to determine which cytokine(s) were correlated with critical clinical issues, such as the development of macular edema. The authors found that vitreous cytokines that are proangiogenic and/or associated with vascular leakage, such as VEGF, PDGF-BB, and G-CSF, tended to be upregulated in patients with severe macular edema (group 2), but not in patients with thin macular edema (group 1). In contrast, other proinflammatory and proangiogenic cytokines, such as MCP-1, ³⁴ RANTES, IL-1ra, ³⁵ and TNF-α, ³⁶ which are all associated with vascular permeability, were downregulated in patients with severe macular edema. Moreover, IP-10, which has antiangiogenic properties, was significantly higher in the vitreous fluid of group 1 than in that of group 2. IP-10 prevents corneal hemangiogenesis ³⁷ and inhibits growth and metastasis of lung carcinoma ³⁸ ; based on these findings, it is thought that high VEGF concentrations in vitreous fluids may be associated with macular edema and may enhance chronic inflammation. Furthermore, the balance of both proangiogenic and antiangiogenic cytokines was maintained until severe macular edema developed. Most patients in group 2 were treated with long-term steroids, or abandoned disease status before surgery. The visual acuity did not improve after surgery, most likely because of irreversible change of the sensory retina caused by long-standing macular edema (data not shown). Surprisingly, even in the patients who were treated with steroids before surgery, the vitreous fluid had higher cytokine levels than the vitreous fluid from the control group. The results suggest that proinflammatory cytokines that are upregulated in the vitreous fluid of ocular sarcoidosis patients might be influenced by BRB breakdown. It proved problematic to test for this in the present study because the limited amount of vitreous sample precluded comparison between serum and vitreous protein; however, VEGF, PDGF-BB, and G-CSF levels should be increased in primary status, not in the late stage of disease such as in group 2. Therefore, the authors believe that surgical intervention should be performed at an early stage of macular edema (before development of severe macular edema) for the purpose of removing proinflammatory cytokines from vitreous fluid.  

Figure. 

 

The relationship between postoperative macular thickness and cytokine concentration in vitreous fluid at 1 month after operation. The y-axis indicates the concentration of cytokines in vitreous fluid (pg/mL). The x-axis indicates the level of macular thickness. Patients were divided into two groups by macular thickness (average thickness of macular area [fovea 1000 μm across] ± DEV): thin CME range (1 = group 1) from 119.60 μm to 359.5 μm and thick CME range (2 = group 2) from 359.5 μm to 599.40 μm. Some patients overlapped because of different status of disease in each eye. *P < 0.05.

Previous reports have shown that surgical intervention is beneficial for improvement of CME or ERM caused by sarcoid uveitis. ^39,40 Here, the authors performed vitrectomy for acute-phase uveitis without CME or ERM. Diamond and Kaplan ⁴¹ have suggested that removal of inflammatory mediators (such as T cells) that accumulate in vitreous fluid in patients with uveitis may have a beneficial effect on macular edema. Likewise, the present study showed that removal of inflammatory cells and cytokines may lead to suppression of ocular inflammation. This report demonstrated that several proinflammatory cytokines persist at highly elevated levels in the vitreous fluid of patients with ocular sarcoidosis. Moreover, the authors confirmed that CD4⁺ T cells also infiltrated into the vitreous fluid of patients with sarcoid uveitis (manuscript submitted). 

The authors found that the retinal granulomas also had P. acnes DNA (data not shown). As with pulmonary sarcoidosis, the infectious theory should be considered for the etiology of ocular sarcoidosis. However, because the sample number was limited, further investigations and proper control samples are required to elucidate the role of P. acnes in ocular sarcoidosis. Therefore, eliminating the continuous exposure to proinflammatory cytokines, proinflammatory cells, and infectious agents (which were the key factors for inducing macular edema) by surgical intervention should be a beneficial treatment for preventing progression of uveitis. On the other hand, some anti-inflammatory cytokines, such as IL-4 and RANTES, were also upregulated in vitreous fluid in the present cases. These cytokines seemed to increase in moderate inflammatory status either to balance the Th1 and Th2 status, or to recruit the T regulatory cells for suppressing inflammation. Notably, RANTES is likely to be a key chemokine for suppressing inflammation in an experimental autoimmune uveitis model for recruiting anti-inflammatory CD8⁺ T cells. ⁴² However, in this study, the number of vitreous CD8⁺ T lymphocytes did not differ significantly in patients who had high RANTES level in vitreous fluid (data not shown). Because the number of patients was limited in this study, further studies are required to investigate and analyze the relationship between RANTES level and CD8⁺ T lymphocytes in vitreous fluid. 

The limitations of this study include the lack of an accurate control group and the variability in the duration of postoperative follow-up. In comparison, the level of IL-1β in the vitreous in infectious uveitis (153.24 ± 117.33 pg/mL) was significantly higher than in ocular sarcoidosis (0.60 ± 0.08 pg/mL; P < 0.01). However, no specific elevation of cytokine was found in ocular sarcoidosis. Further investigations should be performed to analyze the specific elevation of cytokine in ocular sarcoidosis. Despite this, these findings strongly suggest that highly elevated levels of proinflammatory cytokines are present in the vitreous fluid of patients with sarcoid uveitis, and that surgical intervention may be beneficial by removing excess proinflammatory cytokines from the vitreous fluid. Further investigations are warranted. 

Acknowledgments

The authors thank Kentaro Kojima, Hideki Komori, and Toru Yasuhara for the helpful sample collections. Moreover, the authors thank Wendy Chao for editing and critical reading of this manuscript. 

Footnotes

 Supported by KAKEN 22791679 and KAKEN 10433244.

Footnotes

 Disclosure: K. Nagata, None; K. Maruyama, None; K. Uno, None; K. Shinomiya, None; K. Yoneda, None; J. Hamuro, None; S. Sugita, None; T. Yoshimura, None; K.-H. Sonoda, None; M. Mochizuki, None; S. Kinoshita, None