Glaucoma is a condition in the eye characterized by damage to the optic nerve. It is predominantly caused by an elevation of pressure in the eye as a result of a defect in the drainage of the aqueous humor due to disease of the outflow pathway known as the trabecular meshwork found in the angle between the cornea and iris. The glaucoma may be either closed or open angle glaucoma (OAG), depending on the status of the angle. POAG is more predominant in the United States of America the highest prevalence being in the African and Caribbean–derived populations. This population group develops glaucoma earlier than the Caucasian population and responds less well to treatment either by medical or surgical means. Uncontrolled glaucoma inevitably leads to blindness. 

Polypeptide growth factors are pro-inflammatory substances found in normal aqueous, with raised amounts in samples of aqueous humor obtained at the time of glaucoma surgery. ^1–3 TGF-β2 has been the predominant cytokine studied in glaucomatous aqueous. Accumulating evidence suggests that this cytokine may mediate the fibrotic process seen in patients undergoing glaucoma filtration surgery and glaucoma implant surgery. The use of an anti–TGF-β2 antibody, to improve the outcome of glaucoma filtration surgery was studied, resulting in the finding that it did not improve surgical outcomes. ⁴ This suggested the possible presence of other pro-inflammatory cytokines in the aqueous, which would not have been affected by the anti–TGF-β2 antibody. Increased levels of TNF-α, ⁵ interleukins IL-6, ^6,7 IL-8, ^8,9 and VEGF ⁸ have also been found in glaucomatous aqueous. A recent study has reported that IL-9, IL-10, IL-12, IFN-γ, and IFN-α, and chemokine (C-X-C motif) ligand 9 (CXCL9) are significantly elevated in glaucoma aqueous, compared with cataract aqueous. ¹⁰  

Additional cytokines, not as yet described, may also be present in glaucomatous aqueous, and could in addition affect the outcome of glaucoma surgery. This study reports the attempt to find additional pro-inflammatory cytokines in glaucomatous aqueous, by quantifying 23 cytokines using an ELISA. The relationship of IOP to the levels of cytokines was also studied. 

The institutional review board (IRB) of the State University of New York Downstate Medical School approved the study, which adhered to the tenets of the Declaration of Helsinki. 

Aqueous humor was removed from 23 eyes with OAG undergoing glaucoma filtration surgery, 24 eyes with OAG during the hypertensive phase of Molteno implants, and 13 eyes without glaucoma undergoing cataract removal. 

The Molteno implant is nonvalved, in order to prevent postoperative hypotony, the tube is partially occluded with a supramid suture, and the occlusion is made complete by cinching off the tube by placing a vicryl suture around it, thus, not allowing any aqueous to reach the plate surface. The intratubal suture is made available for removal by placing its free end at the edge of the conjunctiva. This occlusion is allowed to remain for a minimum of 4 weeks, and often longer if the IOP remains controlled. This time period allows a capsule to form over the plate. When the eye pressure rises the stent is pulled from the tube and now, for the first time, aqueous reaches the plate surface. The pressure in all cases then becomes normalized, and will become elevated some two to three weeks after the stent is removed, and the aqueous has had a chance to induce an inflammatory response. This is the hypertensive phase referred to in the manuscript. The average time from surgery to the hypertensive phase in the study was 6 to 10 weeks, far removed from any effect that the surgery may have had on the aqueous. It is this aqueous present when the pressure in the bleb is very high that was examined for cytokines. 

The racial profiles, age, comorbidities, and lens status of the patients is shown in Table 1. IOP measurements were recorded. In the patients undergoing filtration surgery and cataract extraction, this consisted of an average of the pressure measurements taken for 2 weeks prior to the surgery. The pressure measurement for the Molteno implant patients was recorded immediately prior to aqueous removal. The patients undergoing filtration surgery as well as the patients with encysted blebs were all on two glaucoma medications prior to aqueous removal. The combination was different in both groups. Consisting of a prostaglandin analog travaprost (Travatan; Alcon, Fort Worth, TX), latanoprost (Xalatan; Pfizer, New York, NY), bimatoprost (Lumigan; Allergan, Irvine, CA), and either a combination of timolol and carbonic anhydrase inhibitor (CAI), dorzolamide (Merck Sharpe & Dohme, Rahway, NJ) or timolol and an alpha agonist brimonidine (Combigan; Allergan) in the filtration patients, and a combination of Cosopt or Combigan and a CAI, dorzolamide (Trusopt; Merck Sharp & Dohme) or brinzolamide (Azopt; Alcon), but no prostaglandins in the bleb patients. The cataract patients had no medications prior to aqueous removal. The type of surgery employed in the filtration and cataract patients was not relevant to the study, as the aqueous tested was that prior to the surgical procedures. 

The aqueous was collected under sterile conditions in all patients. The aqueous was removed with a 30-gauge needle via a clear corneal limbal entry in cataract and glaucoma patients, and by a transbleb surface approach in the Molteno implant patients. Informed consent was obtained from patients prior to aqueous removal. The aqueous was stored immediately or within 1 hour in a freezer at −70°C until analysis occurred. 

TGF-β2 aqueous concentration was measured with RayBio Human TGF-β2 ELISA kit (RayBiotech, Inc., Norcross, GA). This test detects the amount of the active (mature) TGF-β2, however, the majority of the protein is in biologically inactive (latent) form as it is secreted. Latent TGF-β2 was activated with 1 N HCL (10 minutes at room temperature) with following neutralization with 1.2 N NaOH/ 0.5 M HEPES. 

Additionally, we analyzed pro-inflammatory cytokines using the multiplex ELISA KIT Human Cytokine–Screen IR (16-Plex), and chemokines with Human Chemokine (9-plex; Quansys Bioscience, Inc., Logan, UT). We evaluated interleukins IL-1a, IL-2, IL-4, IL-5, IL-6, IL-8, IL-10, IL-12p70, IL-13, IL-15, IL-17, IFN-γ, TNFα, TNF-β2, and chemokines C-C motif chemokine 11 (CCL11; eotaxin eosinophil chemotactic protein), CXCL1, CCL1 (I-309), CXCL10, CCL2, monocyte chemotactic protein-1 [MCP-1]), CCL8 (MCP-2), and CCL5 (RANTES regulated upon activation, normal T-cell). The plates were read with LI-COR ODYSSEY (Biosciences, Lincoln, NE), and analyzed with Q-View software (Quansys Bioscience, Inc.). 

Average sample volumes were 91 ± 68 μL, 148 ± 118 μL, and 207 ± 146 μL for glaucoma eye, bleb, and cataract eye aqueous specimens, respectively. 

This was done using the IBM SPSS-21 statistical package (IBM SPSS Statistics for Windows, Version 21.0; IBM Corp., Armonk, NY). To compare averages, we used one-way ANOVA with post hoc multiple comparisons without assuming the equal variances Dunnett's T3, and a significance level of 0.05. 

Patients were all adults. The glaucoma and Molteno bleb patients all had OAG. The cataract patients were all normotensive patients. The average IOP in the three groups was 13 mm Hg in the cataract group, 26 mm Hg in the glaucoma group, and 38 mm Hg in the bleb group. The duration of the glaucoma was not a confounding factor. 

Aqueous cytokines: 23 cytokines were measured from aqueous concentrations derived from cataract, glaucoma filtration, and hypertensive bleb patients. The presence of the cytokines was determined in each group, and any significant difference noted. Trace to significant amounts of each cytokine was found only in the hypertensive bleb group. The glaucoma group had trace to significant levels of 18 of the cytokines tested, whereas the cataract group had trace to minimal levels of 15 cytokines. The cytokine levels were lowest in the cataract group, with the levels in the glaucoma group all being lower than the levels in the hypertensive bleb group. 

A statistically significant difference (P ≤ 0.05) versus controls (cataract group), was noted in the amount of TGF-β2, IL-6, and CXCL1 from the bleb aqueous (Fig.; Table 2). Within this group only IL-10 was found exclusively in the bleb aqueous. The levels of these cytokines were also higher in the glaucomatous aqueous, but not significantly so. A significant difference when compared with controls was noted in levels of CCL2 in both the glaucoma and bleb aqueous (Fig.). The level of cytokines was noted to increase with an increase in IOP (Table 2). 

In a previous study we reported higher levels of TGF-β2 and Prostaglandin E2 (PGE2), from both glaucomatous aqueous and hypertensive phase bleb aqueous obtained from Molteno implants. ¹¹ The results from that study postulated that a sustained rise in IOP may result in the production of cytokines especially in encysted blebs. The present study was performed to confirm the previous finding, as well as to ascertain the possible presence of other cytokines in glaucomatous aqueous. The present study confirms the presence of additional cytokines in glaucomatous aqueous, a finding that has been previously reported. ^5–10 However, this is the first report of these additional cytokines in the aqueous taken from encysted blebs over glaucoma implants. This paper also confirms that as the IOP rises, the highest levels of cytokines are in the aqueous taken from eyes with the highest pressure group. This finding, once again, suggests that the IOP itself may be responsible for the synthesis of the cytokines. Whether these bioactive substances are produced within the eye, in particular within the trabecular meshwork remains to be determined. A recent review of the role of TGF-β in glaucoma has indicated that this cytokine is produced locally by ocular tissues and is responsible for causing changes in the cells of the trabecular meshwork resulting in a decrease in aqueous outflow, thereby causing an elevation of IOP. It has been shown that mechanical stress also causes an elevation of TGF-β. ¹² The wall of the encysted Molteno blebs contain fibroblasts and macrophages that produce cytokines. ¹³ This is thought to be a response to mechanical stress, namely elevated pressure, in the bleb. Studies using a rat air pouch, a surrogate model for a synovial cavity, have found that the mesenchymal cells comprising the wall of the pouch have the capacity, when appropriately stimulated, to secrete relatively large amounts of cytokines. In this rat model repetitive injections of air, forms the pouch whose walls are composed of mesenchymal cells. ^14,15 The relative amounts of collagen and noncollagen matrix increase with age as a result of continued pressure from the air, resulting in a corresponding increase in cytokine content .We believe that a similar process occurs in an encysted glaucoma implant bleb. The encysted bleb of glaucoma implants are therefore analogous to the rat air pouch, in that cells in the bleb wall when exposed to the stress of pressure in the bleb, will secrete cytokines, resulting in enhanced fibrosis. The increase in cytokine levels seen with the increase in IOP is highly suggestive of IOP being a potential stimulus for cytokine production. The high IOP within the bleb cavity produces a shear force, which in turn could stimulate secretion of cytokines from cells within the bleb wall and possibly from the eye itself. 

This study also demonstrates, significantly elevated levels of CCL2 in both glaucomatous and hypertensive bleb aqueous, with the levels being highest in the hypertensive bleb eyes, once again suggesting that high IOP is associated with production of this cytokine. All of the cytokines, which had higher levels in the bleb aqueous, and CCL2, which had higher levels in the glaucomatous aqueous as well, are pro-inflammatory cytokines. 

This study indicates that there are five pro-inflammatory cytokines that increase with an increase in IOP. The presence of these pro-inflammatory cytokines in hypertensive glaucoma implant blebs may result in fibrosis of the bleb lining, thus, decreasing the efficacy of the bleb, which remains the major cause of glaucoma filtration failure. A recent study reports the development of a mouse model for glaucoma, enabling the study of wound healing in vivo. ¹⁶ There is an early “acute inflammatory” phase characterized by a significantly increased transcript expression of CXLCL1, CCL2, CXCL5, 3, and 4. A later fibrotic phase was characterized by an increased expression of TGF-β2. This experimental in vivo study correlates with the findings of our study. The presence of these cytokines during the early formation of a bleb, as is seen in valved glaucoma implants, where glaucomatous aqueous reaches the plate surface immediately, results in a more severe and prolonged hypertensive phase and finally a less successful bleb. ¹⁷ Removing aqueous from the bleb during the hypertensive phase, has the dual effect of both removing cytokines as well as the stimulus for their production, thus, resulting in less fibrosis and a more successful bleb, as has been previously reported. ¹⁸ Molteno et al. demonstrated that if aqueous was prevented from reaching the episcleral tissue over the implant plate, then a minimal inflammatory response occurred, lasting less than three weeks with virtually no fibrous tissue deposition around the implants. ¹⁸ Molteno also noted that a “long period of normalization of IOP is associated with an improvement of facility of outflow; while elevation of IOP to 25 or 30 mm Hg, over a long period during the stable stage is associated with increased bleb fibrosis and decreased facility of outflow.” ^19,20 The finding that there appears to be a relationship between IOP and cytokine production within the eye would support these heretofore clinical observations. This also suggests that there is a reason for maintaining a low IOP during the formative period of glaucoma filtration or glaucoma implant blebs. The presence of multiple pro-inflammatory cytokines in glaucomatous aqueous, could also explain the reason for the failure of anti–TGF-β2 antibody to improve results of glaucoma surgery procedures. 

A secondary clinical observation was noted during the course of this study. Following failure of an initial glaucoma implant our standard procedure is to insert a second implant in a different quadrant of the affected eye. The pressure lowering effect of the second implant was generally noted to be minimally effective. High levels of cytokines were detected from the aqueous of the first implant. In two patients, it was decided to remove the first implant, followed by insertion of the second implant into a supra-Tenon's pocket. The pressure lowering effect of the second implant was dramatic. The postulate was that the first implant was acting as a factory for the production of cytokines, and as the implants used were nonvalved, the first implant was feeding high levels of cytokines into the second implant, resulting in the rapid fibrosis of the bleb wall and ultimate failure of the second implant. This observation confirms the importance of the pathophysiology of bleb formation, wherein decreasing the effects of pro-inflammatory cytokines, as well as the added effect of the tissue reaction from Tenon's capsule, can improve bleb outcome. When failure of an initial Molteno implant is associated with a return of the IOP to the initial preoperative level, or, as has occurred, become even higher, our current approach is to remove the tube of the first implant from the anterior chamber before placing the second implant. 

In conclusion, this is the first study to show the relationship between increased IOP and the presence of multiple cytokines in glaucomatous aqueous. The finding of significant quantities of CCL2, a markedly pro-inflammatory cytokine, in glaucomatous aqueous has not been previously reported. The clinical implications of the study support the observations that maintaining a low IOP during the formative stage of an implant bleb may result in a more efficient final bleb. Furthermore, preventing the initial “glaucomatous” aqueous from reaching the plate of a glaucoma implant may result in a less severe hypertensive phase. Topical antiglaucoma medications especially the prostaglandin analogs, may influence the aqueous immune content. This could possibly be a confounder to some extent in the interpretation of our findings. A study of aqueous, prior to commencement of glaucoma medications, could determine the possible role of these medications, but ethical considerations would make such a study difficult. A study investigating the relationship between preoperative cytokine levels, and the success of the final bleb is ongoing. 

The authors thank Sara Ferri for obtaining institutional review board permission for the study. 

Disclosure: J. Freedman, None; P. Iserovich, None