Improving patient outcomes following glaucoma surgery: state of the art and future perspectives

Abstract

Of all the treatments currently used to lower intraocular pressure in glaucoma patients, filtration surgery is known to be the most effective. However, in a significant percentage of cases, the constructed channel closes due to excessive scar formation, resulting in surgical failure. The process of postoperative wound healing is characterized by the coagulative and inflammatory phase, followed by the proliferative and repair phase, and finally the remodeling phase. Perioperative antimitotic agents, such as mitomycin C and 5-fluorouracil, are known to modulate the process of wound healing and to improve surgical outcome, but they carry a risk of vision-threatening complications. New alternative strategies to prevent filtration failure, such as inhibition of transforming growth factor-β, vascular endothelial growth factor, and placental growth factor, have shown promising results in the improvement of surgical success. However, it remains necessary to broaden the therapeutic approach by focusing on combined therapies and on extended drug delivery.

Keywords:

• glaucoma filtration surgery
• filtration failure
• wound healing
• wound modulating agents

Introduction

The most effective treatment to lower intraocular pressure (IOP) in glaucoma patients is glaucoma filtration surgery. Trabeculectomy, the reference procedure, is a surgical technique by which a tunnel is created between the anterior chamber and the subconjunctival space, with removal of a part of the trabecular meshwork. A controlled leaking area for aqueous humor is formed, called a filtration bleb. Although this procedure may be crucial in the management of glaucoma,^1,2 it carries a risk of complications. Excessive postoperative wound healing of the conjunctiva and Tenon’s capsule, with subsequent scarring, is known to frequently lead to surgical failure,³ associated with poor postoperative IOP control and consequent progression of glaucomatous disc cupping and visual field loss.

Pharmacological enhancement of trabeculectomy using different antiscarring agents was found to significantly improve surgical success rates. Indeed, the use of 5-fluorouracil and mitomycin C, known as the gold standards in clinical practice, has improved the success rate significantly by enhancing bleb survival. However, long-term success rates for blebs supplemented with these antimitotics were not as promising as desired. Moreover, the nonspecific mechanism of action of these agents may result in severe vision-threatening side effects, such as corneal toxicity, thin-walled avascular blebs, blebitis, endophthalmitis, and hypotony. Therefore, there is still a need for alternative strategies to prevent filtration failure. In this review, an overview of the current wound healing modulation agents is provided, and new approaches to wound modulation that may improve glaucoma filtration surgery are discussed.

Process of wound healing

Wound healing comprises a cascade of different processes that are closely linked, and includes a coagulative and inflammatory phase, followed by a proliferation and repair phase, and finally a remodeling phase.⁴ Figure 1 gives a schematic overview of the wound healing process and the various molecules (growth factors, cytokines, and proteinases) involved in the different phases. After injury, leakage of plasma proteins (fibrinogen, fibronectin, and plasminogen) and blood cells (red and white blood cells and platelets) from disrupted blood vessels takes place. Clotting factor activation leads to conversion of fibrinogen to fibrin, to reduce blood loss. Clots as well as hormones (histamine, serotonin, prostaglandins and leukotrienes), cytokines (eg, interleukin-1 and interferon-α2b) and growth factors (eg, vascular endothelial growth factor [VEGF], placental growth factor, platelet-derived growth factor, fibroblast growth factor, and transforming growth factor [TGF]-β), released during vascular injury, lead to cellular migration and attraction of neutrophils, macrophages, and lymphocytes to the damaged tissue during the inflammatory phase. This results in a proliferative and repair phase during which endothelial cells and fibroblasts migrate into the site of injury. In this phase, angiogenesis and formation of granulation tissue occurs. Finally, remodeling of tissue takes place and involves the formation of scar tissue. Blood vessels regress over time, and fibroblasts induce crosslinking of collagen type I and elastin, which leads to collagen supercoil formation and dense scar tissue.^4,5 After trabeculectomy, wound healing under the scleral flap may play a role in some cases, but Tenon fibroblasts are regarded as the key players in the initiation of wound healing and fibrotic scar formation.⁶ Although it is known that bleb failure in humans normally happens at around 2–3 months,^7,8 proliferation of subconjunctival fibroblasts already occurs on postoperative days 3–5.⁹ The further development of conjunctival fibrosis, regulated by the release of a number of cytokines and growth factors, appears as a result of inflammation, angiogenesis, progressive fibroblast proliferation and migration, and collagen deposition at the site of filtration.⁴

Figure 1 Schematic overview of the process of wound healing and the most important molecules (growth factors, cytokines, and proteinases) involved.

Abbreviations: FGF, fibroblast growth factor; IL, interleukin; MMP, matrix metalloproteinase; PDGF, platelet-derived growth factor; PIGF, placental growth factor; TGF-β, transforming growth factor beta; VEGF, vascular endothelial growth factor; INFα2b, interferon alpha 2b.

Modulators of wound healing in clinical practice

Gentle handling of tissue and control of intraoperative bleeding lead to reduced fibroblast activity, but such preventive maneuvers are usually insufficient to prevent scarring in the long term.⁴ Therefore, various anti-inflammatory and antiproliferative agents are used in clinical practice to inhibit wound healing and to improve bleb survival.

Anti-inflammatory agents

Inhibition of inflammation and wound healing by corticosteroids is largely mediated by suppression of leukocyte concentration and function, as well as by their effects on vascular permeability. This leads to less local tissue disruption, reduced mitogen and growth factor release, and less clot and fibrin production, resulting in diminished fibroblast activity and wound healing.⁴ Corticosteroids are therefore frequently used as anti-inflammatory agents after glaucoma filtration surgery. Theoretically, they carry a risk of inducing a steroid response, resulting in elevated IOP. It is known that approximately 18%–36% of the general population is steroid-responsive, but this prevalence can increase to 46%–92% in patients with primary-open angle glaucoma.¹⁰ It is clear that steroid-induced ocular hypertension is associated with specific morphological changes in the trabecular meshwork, such as an increased deposition of extracellular matrix and altered trabecular meshwork cell function, which in turn will cause disturbed aqueous humor outflow and an increase in IOP.¹¹ Although these morphological changes in the trabecular meshwork still exist after glaucoma filtration surgery, it is known that steroid response is less common after surgery, presumably because the aqueous humor bypasses the trabecular meshwork via the created channel. Indeed, it has been shown that aqueous flow after surgery is reduced through the trabecular pathway, and that after a successful trabeculectomy, the diameter of Schlemm’s canal decreases, most likely due to underperfusion of the trabecular meshwork.^12,13 This may explain why only 17%–36% of operated glaucoma patients develop a steroid-induced rise in IOP after trabeculectomy,¹⁴ whereas the risk in nonoperated glaucoma patients is much higher (46%–92%).¹⁰ Moreover, a ten-year follow-up of a prospective randomized trial of postoperative corticosteroids after trabeculectomy showed that IOP was stabilized in 82.8% (38 of 46 eyes) of steroidtreated eyes after successful glaucoma surgery.¹⁵ Finally, it has been suggested that the risk of a steroid response might increase after filtration failure, since the aqueous humor must again flow through the altered trabecular meshwork.¹⁶ Also, nonsteroidal anti-inflammatory drugs suppress the inflammatory response in the process of wound healing, but are known to be less potent as compared with corticosteroids.¹⁷

Antimitotics

In the early 1980s, antimitotic agents, such as 5-fluorouracil and mitomycin C, were found to be effective in inhibiting activation of fibroblasts. Although these agents have impacted the success rate of glaucoma filtration surgery, they carry a risk of vision-threatening complications. With lower levels of exposure, bleb-related complications are less frequent, but filtration failure is more common, representing a trade-off between efficacy and safety.

5-Fluorouracil

The antiproliferative effect of 5-fluorouracil (Hoffman-La Roche Ltd., Basel, Switzerland), a chemotherapeutic agent, is mediated by antagonizing pyrimidine metabolism. It interferes with the synthesis of thymidine nucleotides, resulting in inhibition of DNA synthesis and ultimately in cell death.¹⁸ Khaw et al showed that 5- fluorouracil is an effective inhibitor of fibroblast growth. A five-minute exposure resulted in arrest of growth and had a long-lasting effect on human Tenon fibroblasts.¹⁹ Animal studies demonstrated that adjunctive 5-fluorouracil following trabeculectomy resulted in prolonged bleb survival with decreased scarring and fibroblast proliferation compared with no 5-fluorouracil.²⁰ Use of 5-fluorouracil at high levels, however, frequently leads to corneal toxicity and apoptosis, since 5-fluorouracil is toxic to all actively replicating tissues, such as corneal epithelium.²¹

Mitomycin C

Mitomycin C (Kyowa Hakko Kirin Co., Ltd., Princeton, NJ, USA), is another chemotherapeutic agent with antiproliferative properties. It undergoes metabolic activation via reduction into an alkylating agent that cross-links DNA. Mitomycin C not only inhibits DNA replication, but also mitosis and synthesis of proteins, since it can interfere with any phase in the cell cycle.⁴ Inhibition of the proliferative phase of the wound healing pathway by mitomycin C increases the success rate of glaucoma surgery, mainly through inhibition of proliferation of fibroblasts and endothelial cells.²² Importantly, different aspects of mitomycin C administration, including dose and exposure time, can determine the outcome of success and affect the wound healing process. It is indeed known that brief intraoperative application of mitomycin C (2–2.5 minutes) is as effective for improvement of surgical outcome as a long exposure (5 minutes),^23,24 whereas a low dose of mitomycin C (0.002%) shows reduced effectiveness as compared with a higher dose (0.02%).²⁵ In addition to dose and duration of application, the size of the surface area can also affect the surgical outcome. Indeed, a large area treated with mitomycin C is more effective in achieving and maintaining target IOP, and the incidence of bleb scarring is lower compared with eyes with small areas of application.^26–28 Moreover, Cordeiro et al reported that diffuse noncystic blebs were observed more often in rabbits treated with large sponges (8×10 mm) than in those treated with small sponges (4×2 mm).²⁹ Khaw et al showed that mitomycin C is more potent than 5-fluorouracil.^30,31 They found that fibroblasts cultured from rabbit samples treated with mitomycin C still induced inhibition of growth at one month, while fibroblasts from 5-fluorouracil-treated rabbits demonstrated full recovery from growth arrest after 7 days. This can be explained by their different mode of action, since it is known that mitomycin C induces permanent apoptosis through activation of caspase-3 and caspase-9, Fas, Bad, and phosphorylated p53 in human Tenon fibroblasts, while this is not the case after administration of 5-fluorouracil.³² Mitomycin C-related cytological toxicity and apoptosis has been associated with the development of thin-walled avascular blebs, which carry a risk of hypotony and endophthalmitis.³³ Comparative clinical trials demonstrated that 5-fluorouracil was associated with fewer side effects than mitomycin C, but seemed to be less efficient in improving the surgical outcome.^34–36

Altogether, although these anti-inflammatory and antimitotic agents were found to be effective in inhibiting the process of wound healing in clinical practice, they are not sufficiently effective to satisfactorily inhibit scarring after filtering surgery, and some carry a risk of severe side effects. Therefore, there is still a need for new approaches in modulation of wound healing to prevent filtration failure.

State of the art in modulation of wound healing

The aqueous humor flowing through the constructed channel into the bleb after glaucoma surgery contains a large number of growth factors. It is known that the presence of growth factors in the aqueous humor of patients with glaucoma can increase the proliferation of Tenon fibroblasts by 60% compared with the aqueous humor of normal persons.³⁷ As such, pre-existing elevated aqueous levels of these growth factors may put patients with glaucoma at a higher risk of scarring after filtration surgery. Therefore, the upregulated growth factors in the aqueous humor of patients with glaucoma can be considered as potential targets in the development of new wound modulation agents. Indeed, a number of growth factor inhibitors have been investigated to improve the surgical outcome after glaucoma filtration surgery. Table 1 represents an overview of the new wound modulation agents.

Table 1 New wound modulation agents

Inhibitor	Mechanism of action	Phase of wound healing	Study	Reference
CAT-152	Antibody against TGF-β2	Inflammation, fibrosis	In vitro (HTF), rabbit, phase III trial	^Citation40,Citation41
Decorin	Inhibition of TGF-β2	Inflammation, fibrosis	Rabbit	^Citation42
Suramin	Inhibition of TGF-β, PDGF, and FGF	Inflammation, fibrosis	Rabbit, case series	^Citation43,Citation44
Tranilast	Inhibition of TGF-β, IL-1, and PGE2	Inflammation, fibrosis	In vitro (RTF), rabbit, pilot study	^Citation45^–^Citation47
Lovastatin	Inhibition of Rho pathway of TGF-β	Inflammation, fibrosis	In vitro (HTF)	^Citation48
ALK inhibitor	Inhibition of ALK signaling of TGF-β	Inflammation, fibrosis	In vitro (RTF), rabbit	^Citation49,Citation50
microRNA	Gene suppression of TGF-β	Fibrosis	In vitro (HTF)	^Citation51
Pegaptanib	Aptamer against VEGF165	Angiogenesis	In vitro (HTF and RTF), rabbit	^Citation61
Ranibizumab	Fab fragment against all VEGF isoforms	Angiogenesis, fibrosis	Pilot study	^Citation99
Bevacizumab	Antibody against all VEGF isoforms	Angiogenesis, fibrosis	In vitro (HTF and RTF), rabbit, case reports and series, pilot study, RCT	^Citation53,Citation59,Citation60, ^Citation62^–^Citation75,Citation97, ^Citation98,Citation100
5D11D4	Antibody against PIGF	Inflammation, angiogenesis, fibrosis	In vitro (MTF), mouse	^Citation76
Y-27932	Inhibition of ROCK	Inflammation, fibrosis	In vitro (HTF), rabbit	^Citation93,Citation94
PDT	Phototoxicity of Tenon fibroblasts	Angiogenesis, fibrosis	In vitro (HTF), rabbit, pilot study, RCT	^Citation101^–^Citation104
Saratin	Inhibition of platelet adhesion	Coagulation, fibrosis	Rabbit	^Citation105,Citation106
Sonepcizumab	Antibody against SIP	Fibrosis	Rabbit	^Citation107
Doxycycline	Inhibition of MMP-1, MMP-2, MMP-8, MMP-9, MMP-13	Inflammation, fibrosis	Rabbit	^Citation108
Pirfenidone	Inhibition of TGF-β, CTGF, PDGF, TNF-α	Inflammation, fibrosis	In vitro (HTF), rabbit	^Citation109,Citation110
Tacrolimus	Immunosuppressive agent	Inflammation, fibrosis	Rabbit	^Citation111
Octreotide	Octapeptide of somatostatin	Inflammation, fibrosis	Rabbit	^Citation111

Abbreviations: HTF, human Tenon fibroblasts; RTF, rabbit Tenon fibroblasts; RCT, randomized controlled trial; MTF, murine Tenon fibroblasts; PDT, photodynamic therapy; PIGF, placental growth factor; MMP, matrix metalloproteinase; VEGF, vascular endothelial growth factor; IL, interleukin; PGE2, prostaglandin E2; FGF, fibroblast growth factor; TNF-α, tumor necrosis factor alpha; TGF-β, transforming growth factor beta; PDGF, platelet-derived growth factor; CTGF, connective tissue growth factor; ALK, anaplastic lymphoma kinase; CAT-152, lerdelimumab; ROCK, Rho-associated protein kinase.

Inhibition of TGF-β

TGF-β is a key cytokine in the process of wound healing and has been found to be present at significantly higher levels in the aqueous humor of patients with glaucoma than in that of normal individuals.³⁸ Moreover, TGF-β2 is known to stimulate proliferation of human Tenon fibroblasts in vitro and to enhance fibroblast-mediated collagen contraction.³⁹ Therefore, the strategy of inhibiting this growth factor has been extensively investigated by a group led by Professor PT Khaw (Moorfields Eye Hospital and UCL Institute of Ophthalmology, London, UK). Blocking TGF-β2 with a recombinant human monoclonal antibody (CAT-152, Cambridge Antibody Technology, Cambridge, UK) seemed promising with regard to proliferation of Tenon fibroblasts and in an animal model. Khaw’s group investigated CAT-152 as an adjunct to glaucoma filtration surgery in a rabbit model and showed that repeated subconjunctival injections (1 mg/mL on postoperative days 0, 1, 2, 3, and 7) significantly improved the surgical outcome and bleb survival compared with controls. Histologically, CAT-152-treated eyes showed less collagen deposition and clear evidence of bleb formation, without side effects.⁴⁰ A clinical study was initiated but was terminated prematurely due to lack of efficacy at the dose used (1 mg/mL).⁴¹ Decorin, a naturally occurring TGF-β inhibitor, also significantly altered conjunctival scarring and the surgical outcome of glaucoma filtration surgery in rabbits, but no clinical data are available as yet.⁴² Inhibition of TGF-β2 by other agents, such as suramin (Bayer AG, Leverkusen, Germany),^43,44 tranilast (Kissei Pharmaceutical Co., Ltd., Matsumoto City, Japan),^45–47 lovastatin (Merck & Co., Inc., Whitehouse Station, NJ, USA),⁴⁸ and ALK5 inhibitors,^49,50 has also been attempted in in vitro and animal studies, with promising results. A novel approach for targeting the TGF-β signaling pathway, currently under investigation, is the use of microRNA;⁵¹ however, confirmation in animal studies is required.

Inhibition of VEGF

VEGF is known to play an important role in scar formation.⁵² Increased concentrations of VEGF have been demonstrated in the aqueous humor from eyes of patients with glaucoma when compared with nonglaucomatous controls.^53–55 In addition to being a potent inducer of angiogenesis, VEGF promotes the migration of inflammatory cells (such as neutrophils and macrophages) and fibroblasts. This growth factor indirectly stimulates fibrosis via angiogenesis, and also has a direct effect on the activity of fibroblasts.^56–58 Therefore, VEGF inhibitors were investigated by several groups as a potential antiscarring adjunct to filtration surgery.

Our group showed for the first time in a rabbit model of glaucoma filtration surgery that a single subconjunctival (2.5 mg) and intracameral (5 mg) injection of bevacizumab (Genentech, Inc., San Francisco, CA, USA) was effective in improving the outcome of glaucoma surgery by increasing bleb area and survival. This was associated with a reduction in angiogenesis during the initial phase of healing and with diminished fibrosis in the later stages.⁵³ When compared with blebs treated with 5-fluorouracil (5 mg), those treated with the VEGF inhibitor (1.25 mg) survived longer,⁵⁹ and combining bevacizumab (2.5 mg) and 5-fluorouracil (5 mg) resulted in 100% bleb survival and a greater antifibrotic effect compared with monotherapy using each inhibitor.⁶⁰ Importantly, the various VEGF inhibitors available differ in their isoform selectivity, with implications for both efficacy and safety. Based on their VEGF isoform targets, two types of VEGF inhibitors have been identified, ie, selective and nonselective. The effect of pegaptanib (Pfizer Inc., New York, NY, USA), a selective VEGF₁₆₅ inhibitor, on surgical outcome in a rabbit model of glaucoma surgery was less pronounced than that of nonselective VEGF inhibition, due to a retained action of the VEGF₁₂₁ and VEGF₁₈₉ isoforms. It is indeed known that Tenon fibroblasts proliferation is mainly induced by these isoforms, while VEGF₁₆₅ has less pronounced effects on Tenon fibroblasts.⁶¹

Meanwhile, a number of clinical case reports and case series have also shown a beneficial effect of subconjunctival (1, 1.25, and 2.5 mg) and topical (5 mg/mL) bevacizumab, a nonselective VEGF inhibitor, in reducing IOP and improving the surgical outcome.^62–67 These studies also demonstrated that the antibody against VEGF is safe for ocular use. An early pilot study to evaluate a single subconjunctival bevacizumab injection (1.25 mg) demonstrated that this VEGF inhibitor is a potential adjunctive treatment for reducing the incidence of bleb failure after trabeculectomy. Six-month outcomes showed that trabeculectomy was successful in terms of IOP control in eleven of 12 eyes (92%), with an average IOP reduction of 52%.⁶⁸ This beneficial effect was confirmed in more recent studies indicating no significant differences between subconjunctival bevacizumab (1.25 and 2.5 mg) and mitomycin C (0.03% for 3 minutes) in terms of morphological bleb features and IOP-lowering.^69,70

On the other hand, four further pilot studies showed that administration of bevacizumab (0.2 and 1.25 mg) during trabeculectomy did not affect IOP differently from placebo⁷¹ or 5-fluorouracil (5 and 7.5 mg).^72–74 Notably, all these studies included small numbers of patients or had a short follow-up duration, meaning that they lacked the necessary power to draw firm conclusions. Therefore, our group at the University Hospitals of Leuven performed a prospective, randomized, double-blind, placebo-controlled trial in which a 12-month follow-up was undertaken in 138 patients, 69 of whom were treated with bevacizumab (1.25 mg by intracameral injection). The absolute success rate was higher in the bevacizumab group than in the placebo group (71% versus 51%, respectively), with the need for IOP-lowering interventions (needlings) being lower after bevacizumab treatment (12% versus 33%).⁷⁵ Figure 2 shows blebs from the bevacizumab and placebo groups at 6 months of follow-up. Bevacizumab-treated eyes showed less vascularity than those treated with placebo. These recent insights open up promising perspectives for the use of anti-VEGF therapy to prevent filtration failure after glaucoma surgery.

Figure 2 Representative blebs from (A) bevacizumab group and (B) placebo group showing less vascularity in the bevacizumab-treated eye at 6 months postoperatively.

Note: Reproduced from intracameral bevacizumab as an adjunct to trabeculectomy: a 1-year prospective, randomised study. Vandewalle E, Abegao Pinto L, Van Bergen T, et al, Vol 98, pages 73–78, Copyright © 2014 with permission from BMJ Publishing Group Ltd.⁷⁵

Inhibition of placental growth factor

Importantly, although bevacizumab could improve the surgical outcome by inhibiting angiogenesis and deposition of collagen, and VEGF is known to have proinflammatory properties, neither selective nor nonselective VEGF inhibition could reduce inflammation, another important process in postoperative healing.^53,61 Our group showed that the upregulated levels of aqueous placental growth factor after administration of bevacizumab might explain why anti-VEGF therapy was not sufficient to reduce the inflammatory response after glaucoma surgery. Indeed, aqueous placental growth factor levels were significantly upregulated in patients with glaucoma compared with control subjects. Moreover, intracameral injection of anti-placental growth factor antibody (5.2 μg, ThromboGenics NV, Leuven, Belgium) effectively improved the surgical outcome in a mouse model of trabeculectomy, by increasing the bleb area and survival after surgery compared with an irrelevant antibody control and with the VEGF-R2 antibody. The effect on surgical outcome was associated with a decreased angiogenesis, fibrosis, and importantly, inflammation. Therefore, inhibition of placental growth factor might theoretically be even more effective than inhibition of VEGF, due to the additional effect on inflammation.⁷⁶

Other growth factors, cytokines, and proteinases

Several other molecules, including growth factors (eg, platelet-derived growth factor) cytokines (eg, interleukin-6, interleukin-8, and tumor necrosis factor-α), and matrix metalloproteinases (eg, MMP-2 and MMP-3) have been reported to be upregulated after glaucoma filtration surgery.^77–80 Different studies have shown that these factors can stimulate proliferation of Tenon fibroblasts in vitro.^81,82 Moreover, inhibition of some of these factors reduced growth of Tenon fibroblasts⁸³ and improved surgical outcome in a rabbit model of glaucoma.^84,85 However, none of these inhibitors have reached clinical trials. This can probably be explained by the fact that upregulation of TGF-(β2 and VEGF in patients with glaucoma correlates strongly with surgical failure, whereas other factors (eg, MMP-1, interleukin-6, and interleukin-8) are associated with surgical success.⁸⁶

Inhibition of ROCK

Rho kinases (ROCK 1 and 2) are major downstream effectors of Rho-GTPase proteins and mediate a number of important cellular functions associated with cytoskeletal rearrangement, such as cell morphology, motility, adhesion, contraction, and cytokinesis.⁸⁷ It is known that ROCK plays an important role in a number of wound healing processes, ie, inflammation, angiogenesis, and fibrosis. Administration of ROCK inhibitors on lipopolysaccharide-stimulated inflammatory cells decreased the activation of NF-κβ and subsequently blocked the generation of proinflammatory cytokines, such as interleukin-1β/6 and tumor necrosis factor-α.^88–90 Increased expression of RhoA in endothelial cells also significantly enhanced migration and angiogenic capacity in vitro, which suggests an important role for the Rho/ROCK pathway in these processes.⁹¹ Furthermore, van Niew-Amerongen et al showed demonstrated that inhibition of the Rho/ROCK pathway attenuated VEGF-mediated migration and angiogenesis in vitro.⁹² Importantly, in vitro studies also showed that contractility of Tenon fibroblasts is essential in the transdifferentation into myofibroblasts. Inhibition of the Rho/ROCK pathway by specific kinase inhibitors (Y-27632 [Novartis AG, Basel, Switzerland], HA-1077 [Santen Pharmaceutical Co., Ltd., Osaka, Japan], H-1152 [Calbiochem^®; Merck KGaA, Darmstadt, Germany], and ML-7 [Merck KGaA]) is known to counteract this contraction by blocking TGF-β or lysophospholipid acid-induced activation of Tenon fibroblasts.^93,94 Secondly, Honjo et al showed that topical treatment with the ROCK inhibitor Y-27632 had profound effects on surgical outcome in a rabbit model of glaucoma surgery. Histological examination revealed that blebs treated with Y-27632 lacked significant collagen deposition compared with vehicle treatment.⁹³ As such, ROCK inhibitors could strengthen our armamentarium to improve the effectiveness of filtration surgery, because these inhibitors have the potency to interfere at different levels of the wound healing process.

Future perspectives

Although bevacizumab has been proven to be effective in reducing wound healing after filtering glaucoma surgery in a prospective clinical trial, mitomycin C is still the most common antifibrotic agent used during trabeculectomy in routine clinical practice. This study may represent a breakthrough in the quest for novel antifibrotic strategies in filtering surgery; however, some additional questions remain to be answered. The most optimal route of administration of anti-VEGF drugs is still unknown. Nomoto et al showed that, despite lower maximal concentrations of bevacizumab after subconjunctival injection in the rabbit eye, the intraocular half-life was higher after subconjunctival injection as compared with intravitreal injection.⁹⁵ Ambati et al showed that immunoglobulin G antibodies have a relatively high scleral permeability.⁹⁶ Since bevacizumab is an immunoglobulin G1 antibody, it should penetrate intraocular tissues via the sclera, which might then work as a sustained-release mechanism. Recent studies in a rabbit model of glaucoma surgery showed that postoperative subconjunctival injection of bevacizumab was more effective than intravitreal administration.⁹⁷ It also remains unknown whether bevacizumab could replace antimitotics in clinical practice, or if it should be considered as an adjunctive to antimitotics. Although some small comparative^69,70,98 and complementary studies^99,100 have been performed, large randomized clinical trials are still lacking.

A variety of alternative antiscarring strategies (Table 1), such as photodynamic therapy,^101–104 saratin (Merck KGaA),^105,106 sonepcizumab (Lpath, Inc., San Diego, CA, USA),¹⁰⁷ doxycycline (Pfizer Inc.),¹⁰⁸ pirfenidone (InterMune, Brisbane, CA, USA),^109,110 tacrolimus (Astellas Pharma Inc., Tokyo, Japan),¹¹¹ and octreotide (Novartis),¹¹¹ have also shown promising results by improving the outcome of surgery in a standard rabbit model. Although many agents have been evaluated in preclinical models for glaucoma filtration surgery, few have been found to be beneficial in large prospective comparative clinical trials and none have replaced mitomycin C or 5-fluorouracil in clinical practice so far. Therefore, it is necessary to broaden the therapeutic approach, given that treatment with a single agent may lead to drug resistance. Indeed, because of the complexity of the wound healing process,^4,5 it is possible that targeting one molecule may be insufficient to impede the wound healing process due to the upregulation of other growth factors. It has been suggested that mitomycin C can upregulate proinflammatory, proangiogenic, and/or profibrotic factors. Therefore, combining mitomycin C with drugs that block these upregulated profibrotic growth factors may offer promising complementary efficacy and allow a lower dose of mitomycin C to be used, with fewer side effects.

Finally, wound healing is a relatively slow process involving interaction between the antiscarring agent and the target tissue (Tenon’s capsule) for several weeks to months. Therefore, future research should focus on the development of targeted drug delivery systems that improve and/or prolong contact between the target tissues and the antiscarring molecules. A few studies have already been published in which a slow-release formulation of 5-fluorouracil is described. Collagen plugs,¹¹² polyesters,¹¹³ and amniotic membranes¹¹⁴ containing the antimitotic agent were shown to induce a prolonged effect on bleb survival in the standard rabbit model. The effect of a sustained-release dexamethasone implant (Ozurdex^®, Allergan, Irvine, CA, USA) on postoperative wound healing was also investigated in this animal model, and was found to be effective in prolonging bleb survival. However, mitomycin C-treated blebs showed significantly longer survival compared with Ozurdex-treated blebs.¹¹⁵

Khalili et al developed a new solid tissue tablet that provides prolonged release of bevacizumab in vitro, and an in vivo study showed substantially prolonged bleb survival compared with mitomycin C in the rabbit model.¹¹⁶ Thus, this antibody tablet has the potential to prolong local antibody release and exert a significant biological effect. These data are very promising, but further studies are required. If successful, such improved drug delivery systems might significantly improve the efficacy of known antiscarring molecules and possibly decrease their side effects.

Conclusion

Failing filtering surgery due to an excessive healing reaction is still a considerable problem in ophthalmology, and contributes to progressive vision loss in patients with glaucoma. Although a number of studies have demonstrated that many antifibrotic agents can reduce postoperative wound healing in the preclinical setting, bevacizumab is the only agent that has been proven to be effective in a prospective randomized clinical trial. Further research should focus on targeting multiple steps of the wound healing process by use of combination therapy and also on new methods of extended drug delivery to improve the efficacy of the known antiscarring agents. Finally, we should be cautious when extrapolating preclinical findings to the management of human disease. Animal models are not precise replicas of human pathology, and the pharmacokinetics and half-lives of the therapeutic agents might differ between animals and human eyes. Therefore, further clinical trials in humans with long-term follow-up and inclusion of comparative controls are required.

Disclosure

The authors report no conflicts of interest in this work.